,P
u.
0 .5‘ I unit‘s

3?...
.

‘
. . .

~v.......u
.5.

u. 3.! L.

. u
a:
a. I:.!.1!931.I.l.l.r
:2.
3

"u, v

[(I1Y 1. I

 

 

 

.‘ . {III}... 1| arl’llE...“

      

 

t A -‘35‘ I l' 3].! '2 l"! ‘1
I ' , . . . i.
u: . . Ola. fissfitazsznu X INA-unfit?” v4.5 .(htnjr. LT ,..r§vt..vl#..t9..v.._. F :2. w d: .. 4...... u. _ . y: .|.,. 2% 1.. . .. , .1... .. AK... HI .AA... ......Tl..: .P‘rr I’h... Ain.“.~w....m..,-__l..r9.s35'. gnu. v'uo . ‘ ‘ s.
.. . . . , ‘

 

SETAT UNIVERSITY LIBRARIES

'llllllll W l liill lli ill

3 1293 008820

 

 

 

 

 

 

 

 

 

 

 

 

                     

 

 

 

 

 

This is to certify that the

dissertation entitled

ONCOGENE COOPERATION AND CELLULAR DIFFERENTIATION IN
THE IN VIVO AND IN VITRO TRANSFORMATION OF MURINE PRE-B CELLS

presented by

Shu-Chih Chen

has been accepted towards fulfillment
of the requirements for

 

 

PhoDo degree in MicrObiOlogy

'7 t ,

Major Professor 6/
(

 

 

Date March 25, 1992

 

MS U i: an Affirmatiw Action/Equal Opportunity Institution 0- 12771

_____ _ ,V_ _ . ._ . — A- I, _ A — _ r i

 

' _ LIBRARY
Michigan State

'1 University

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

 

J

DATE DUE DATE DUE DATE DUE l

__i J

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative ActiorVEqual Opportunity Institution
cmmunii-ni

 

 

ONCOGENE COOPERATION AND CELLULAR DIFFERENTIATION IN
THE IN VIVO AND IN VITRO TRANSFORMATION OF MURINE PRE-B CELLS

BY

Shu-Chih Chen

A DISSERTATION

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

DOCTOR OF PHILOSOPHY

Department of Microbiology and Public Health

1 992

(77-a7g'0

ABSTRACT

ONCOGENE COOPERATION AND CELLULAR DIFFERENTIATION IN
THE IN VIVO AND IN VITRO TRANSFORMATION OF MURINE PRE-B CELLS

BY
ShU-Chih Chen

An in vivo and an in vitro model system were used to study oncogene
cooperation which presumably occurs in multi-step tumorigenesis. Both model
systems involved transforming murine B cells using the Whitlock and Witte culture
system in combination with tumor challenge. In the in vivo model system, c-myc
gene activation was found in tumors derived from a v-Ha-ras cell line. Two tumors
possessed a MoMuLV provirus integration immediately upstream and in a reverse
transcriptional orientation to c—myc. Elevated expression of c-myc was found in
these two tumors and another two tumors with no gross gene alteration. This
finding parallels the synergy of v—Ha—ras and v—myc observed in the in vitro
transformation of murine B lymphoid cells and validates synergy as a model for in
vivo tumor progression. The insertional activation of the c-myc gene by MoMuLV
in B cell lymphomas is novel. The flanking region, bone-1, of one of these non-c-
myc tumor-specific viral integration sites was characterized. Sequence homology

to this locus was found in other mammals, and chicken.

In the in vitro model system, lL-7, a pre-B cell growth factor, was found to
be incapable of cooperating with the v-Ha-ras oncogene in inducing a fully
transformed phenotype in murine B cells. A discrepancy between the oncogene
cooperation in a co-infection procedure and a sequential addition procedure was
found. The clonal nature of the cell lines generated from the coinfection procedure
suggests the selection of additional oncogenic events. The oncogenic potential
of lL-7 expression, in itself, and those of other genes are probably best assessed
in well-characterized individual cell lines.

A “lineage switching" from v-Ha-ras transformed pre-B cells to "macrophage-
like" cells was also found in this study. These pre-B cells have gained the capacity
to effectively present antigen in an MHC-restricted fashion. These cells have also
rearranged their kappa light chain immunoglobulin locus, suggesting that
macrophage differentiation and immunoglobulin rearrangement are not mutually
exclusive processes. The existence of both lymphoid and myeloid characteristics
in a cell suggests greater plasticity in hematopoietic lineage commitment than

conventionally thought to be the case.

TO MY MOM AND DAD

iv

ACKNOWLEDGMENTS

I would like to express my deepest appreciation to my major professor, Dr.
Richard Schwartz for his direction, encouragement, and patience during my years
as a graduate student.

I also would like to thank the members of my committee for their guidance
and invaluable time, Drs. Jerry Dodgson, Susan Conrad, Michele Fluck, and Lyman
Crittenden.

I thank Diane Redenius for her technical assistance and friendship, and
James Bretz and Timothy Weichert for their companionship. They and Rich have
taught me lots about American culture and made my stay in Michigan State
University fun and colorful. Thanks are due the members of Dr. Dodgson’s
Laboratory for their generosity in letting me share equipments and materials.

Last, but not least, I want to thank my parents Mr. S.C. Chen and Mrs. H.F.
Tsai, my sister H.H. Chen, my brothers Y.S. Chen and Y.Y. Chen for their endless
love and support. Thanks are to all my friends, especially H.-C. Li, in East

Lansing, too. With them, life has been delightful for the past five years.

TABLE OF CONTENTS

Page

List of tables ....................................................................................................... viii
List of figures ......................................................................................................... ix
Introduction .......................................................................................................... 1

References ................................................................................................ 4
Chapter 1 Literature review

1. Tumorigenesis is a multi-step process ......................................... 5

2.1 The search for genes that are involved in tumorigenesis .............. 8

2.1.1 Identification of cellular oncogenes by DNA transfer experiment..8

2.1.2 Characterization of viral flanking regions ...................................... 11

2.1.3 Molecular characterization of chromosomal aberrations .............. 15

2.1.4 Identification of new oncogenes by homologous recombination..18
2 2 Experimental models for transformation and tumor progression..19

2.2.1 Use of animal model versus human models ................................. 19
2.2.2 Use of in vivo versus in vitro systems .......................................... 20
2.2.3 Model systems used to test the oncogenic potential of

oncogenes ..................................................................................... 20
2.2.4 Model systems used to study tumor progression ........................ 22
3. Multiple genes or factors that are involved in B cell

lymphomas/leukemias of human, mouse, and avian ................... 24
3.1 Human B cell neoplasia ................................................................ 24
3.1.1 Burkitt lymphoma .......................................................................... 24
3.1.2 Other human B-cell leukemias and lymphomas ............................ 30

B-cell chronic lymphocytic leukemia ............................................. 30

Follicular lymphomas ..................................................................... 33

Pre-B cell acute lymphocytic leukemia .......................................... 34
3.2 Murine plasmacytoma .................................................................... 35
3.3 Avian lymphoid leukosis ................................................................ 39
3.4 In vitro transformation of murine B cells ....................................... 42
Literature cited ......................................................................................... 48

Chapter 2 Tumorigenesis of a v-Ha-ras-expressing pre-B cell line selects
for c-myc activation .......................................................................... 61

Abstract .................................................................................................... 61
Materials and methods ............................................................................ 62
Results ..................................................................................................... 62
Discussion ................................................................................................ 66
References ............................................................................................... 67

vi

Chapter 3 Cloning and characterization of a viral flanking region
common to B lymphoid tumors derived from

a v-Ha-ras infected pre-B cell line .................................................... 69
Abstract .................................................................................................... 70
Introduction .............................................................................................. 71
Materials and methods ............................................................................ 73
Results ..................................................................................................... 76
Discussion ................................................................................................ 85
References ............................................................................................... 87

Chapter 4 IL-7 expression in a v-Ha-ras transformed pre-B cell line is not
sufficient for tumorigenicity: differing assessments in

clonal versus heterogeneous populations ........................................ 89
Abstract ..................................................................................................... 90
Introduction .............................................................................................. 91
Results ..................................................................................................... 93
Discussion ................................................................................................ 107
Materials and methods ........................................................................... 109
Acknowledgement .................................................................................. 1 12
References .............................................................................................. 1 13
Appendix A Lineage switch macrophages can present antigen ......................... 115
Abstract ................................................................................................... 1 15
Introduction ............................................................................................. 1 16
Results .................................................................................................... 1 17
Discussion .............................................................................................. 135
Materials and methods ........................................................................... 136
Acknowledgement .................................................................................. 141
References .............................................................................................. 142
Appendix B ....................................................................................................... 145
Introduction ............................................................................................ 145
Materials and methods ........................................................................... 145
Results .................................................................................................... 147
Discussion .............................................................................................. 159
References .............................................................................................. 162
Summary and discussion .................................................................................. 163
References .............................................................................................. 172

vii

LIST OF TABLES

Page

Table
Chapter 1
1.1 Oncogenes identified by DNA transfer experiment ...................................... 10
1.2 Cellular genes activated by proviral insertion ............................................... 12
1.3 Viral flanking regions without viral oncogenes analogues ............................ 13
1.4 Genes identified from chromosomal aberrations .......................................... 17
Chapter 2
1. Restriction fragments in the 5’ c-myc region of tumors 1 and 3 ................... 65
Chapter 4
1. Growth without cellular feeder layers ........................................................... 101
2. Soft agar growth and tumorigenesis ............................................................ 101
Appendix A
1. LPS-induced cytokine released by tumor 4 macrophage subclones .......... 124
Appendix B
1. Tumor challenge in Balb/c mice .................................................................... 150
2. Probes that were used in searching for the putative secondary events

that may be involved in the tumor progression of R2 tumors ............... 151
3. Relative stability of c-myc mRNA in R2 and tumors .................................... 155

viii

LIST OF FIGURES

Page
Figure
Chapter 2
1. Viral ras integrations ....................................................................................... 63
2. MoMuLV integrations ..................................................................................... 64
3. Rearrangement of c-myc in tumor 1 and 3 .................................................... 64
4. MoMuLV integration near c-myc .................................................................... 65
5. env and c-myc cohybridize ............................................................................ 66
6. Elevated c-myc mRNA levels .......................................................................... 66
Chapter 3
1. Tumor specific viral integration in the bone-1 locus ...................................... 77
2. Cloning stragegy and restriction map of the bone-1 locus ............................ 79
3. Bone-1 is evolutionary conserved .................................................................. 81
4. DNA sequences of the 03P2 fragment .......................................................... 82
5. Putative open reading frames in the sequence of the 03P2 fragment .......... 83
6. RFLP mapping of bone-1 ................................................................................ 84
Chapter 4
1. Proviral integration .......................................................................................... 94
2. Retroviral transcription .................................................................................... 98
3. Immune loci define a pre-B phenotype ......................................................... 100
4. Proviral integration ........................................................................................ 105
5. Retroviral transcription ................................................................................. 106
Appendix A
1. Viral integrations ........................................................................................... 118
2. Nonspecific phagocytosis of latex beads ..................................................... 120
3. RNA analyses of c-myc, c-myb, and c-fms ................................................... 122
4. Kappa light chain rearrangement ................................................................. 125
5. Expression of mu chain RNA ....................................................................... 127
6. RT-PCR analysis of CD45 ............................................................................ 131
7. Antigen presentation .................................................................................... 132
8. l-ACl expression ............................................................................................. 133
Appendix B
1. Growth of R2 and seven tumors in the presence and absence of IL-7 ....... 149
2. Nuclear run-on transcription ........................................................................ 153
3. Kappa chain gene rearrangement ................................................................ 157
4. Mu chain gene rearrangement ..................................................................... 158

Introduction

The focus of my thesis has been the identification of putative oncogenes
that can cooperate with v-Ha-ras in tumor progression or in the in vitro
transformation of murine B lymphoid cells. Two model systems involving the use
of a long term bone marrow culture (Whitlock and Witte, 1982) were used in my
studies.

The first model system is derived from the study of Schwartz and coworkers
(1986a). Using a murine long term bone marrow culture system, Schwartz et al.
(1986a) were able to show that v-Ha-ras and v-myc have a synergistic effect in
transforming murine pre-B cells. In their study, pre-B cell lines carrying both
oncogenes were capable of causing lymphoma in syngeneic mice at high
frequency and with a short latency, whereas pre—B cell lines carrying v-Ha-ras
alone only gave rise to tumors occasionally and with a prolonged latency. The
latter result indicates that v-Ha-ras is not sufficient to cause pre-B cell lymphomas
and other secondary events are required to facilitate tumor formation. It is this
finding that prompted us to search for these secondary events by the following
approach. First, Whitlock-Wine high density bone marrow was infected with a
Moloney murine leukemia virus (MoMuLV) based vector carrying a v-Ha-ras
oncogene with MoMuLV as a helper virus. Usually pre-B cell cells with a single
specific v-Ha-ras integration were established. This in vitro transformation step
presumably facilitates multi-step tumorigenesis and provides a molecular marker

for further studies. Secondly, these pre-B cell lines were subjected to tumor

“y. e I

2

challenges, and they gave rise to tumors occasionally. This step provided the
opportunity to accumulate other mutations which may cooperate with v-Ha-ras in
vivo in tumorigenesis. The involvement of secondary events was examined by
southern and northern hybridization analyses with various probes of known
oncogenes and growth-related genes. Nearly thirty DNA probes (as listed in the
Appendix B) of oncogenes, tumor suppressor genes, growth factor genes, or the
flanking region of frequent viral integration sites were used to screen for gene
rearrangements.

The second model system involved the introduction of v-Ha-ras with or
without other oncogenes or genes with oncogenic potential into Whitldck and Witte
high density bone marrow culture. In this model system, the transforming effect
of each gene was studied directly. This model system involves less labor than the
first model, although it suffers the shortcoming of all in vitro transforming systems,
the lack of interactions with the full range of in vivo cell types. It is worth
mentioning that the transformation potential of any oncogene identified from the
first model can be easily tested in this system. A pre-B cell growth factor,
interleukin-7 (IL-7), was chosen to study.

In the first chapter of this thesis, I will summarize briefly the methodologies
that have been used to identify putative oncogenes that may be involved in
tumorigenesis, and to test the oncogenic potential of these putative oncogenes.

This is followed by a description of the known involvement of oncogenes and other
molecular events in human B cell neoplasia as well as B cell tumors of other

animals.

3

The results of studies on oncogene cooperativity between v-Ha-ras and
other oncogenes in the in vivo and in vitro transformation of murine B cells will be
presented in Chapters 2, 3, and 4 in this thesis. Chapter 2 details a c-myc
activation found in the in vivo tumor progression of some of the B cell tumors
obtained from the first model. This work has been published and will be presented
as a manuscript. Data that were not shown in the manuscript because of space
limitations will be included in the Appendix B.

Chapter 3 describes the isolation and characterization of a viral integration
flanking region common to tumor cell lines. That locus may encode a putative
oncogene. This work involves a collaboration with Marge Strobel and Nancy
Jenkins (NCI) who did the chromosomal mapping, and will be presented as a
manuscript and submitted for publication at a later date.

Chapter 4 reports the effects of lL-7 in cooperation with v-Ha-ras in
transforming mouse pre-B cells with the second model system. It will be also
presented as a manuscript and will soon be submitted to Molecular and Cellular
Biology.

In examining the transformed phenotypes of pre-B tumor cell lines, at least
two independent tumor cell lines showed an interesting lineage switch
phenomenon, in which macrophage specific characteristics were found. Since
lineage infidelity has also been described in a minority of hematological
malignancies (McCulloch, 1983), we decided to further characterize these two lines
with the hope that some correlations can be drawn to the clinical cases, and to
provide some insights on lineage determination in hematopoiesis. This work

includes molecular, cytological, and functional analysis of the lineage switched

HA

w I

4

lines. I had made the discovery of the morphological changes of these tumor cell
lines and subsequently subcloned one of these tumor cell lines, the T4 cell lines.
I also proved that the T4 subclones were indeed descendants of their parental pre-
8 cell line, the R2 cell line, and did some of the cytochemical and
immunocytological staining analyses on the T4 subclones. The demonstration of
macrophage specific gene expression of T4 cells and kappa chain rearrangement
of T4 subclones were also performed by me. Other portions of this work were
contributed by the other authors listed on this manuscript. This work is in press
in Developmental Immunology. I will present it as the submitted manuscript in
Appendix A.

At the end, a section of summary and discussion has been included to
cover points that can be integrated from all of the above work but were not

specified in each individual chapter.

Reference:

McCulloch EA. (1983). Stem cells in normal and leukemic hemopoeisis. Blood 62,
1-13.

Schwartz R.C., L.W. Stanton, S.C. Riley, K.B. Marcu, and ON. Witte. (1986a).
Synergism of v-myc and v-Ha-ras in the in vitro neoplastic progression of murine
lymphoid cells. Mol. Cell. Biol. 6, 3221-3231.

Whitlock C.A., O.N. Witte. (1982). Long-term culture of B lymphocytes and their
precursors from murine bone marrow. Proc. Natl. Acad. Sci. USA 79, 3608-3612.

Chapter 1
Literature review
In this chapter, evidence in support of the multi-step nature of tumorigenesis
will be described first. It is followed by a brief overview of methodologies used in
the identification of oncogenes and the evaluation of their oncogenic potential.
Comparison of the use of in vivo versus in vitro model systems, and animal versus
human systems is included in this section. Then, I will discuss in detail the
literature on oncogenes known to be involved in the formation of B lymphoid
tumors. The discussion will encompass oncogene activation observed in tumors

of human and other animals, as well as in in vitro transformation of B cells.

1. Tumorigenesis Is a multiple-step process

Cancer cells differ from their normal counterparts in that they no longer
respond to normal growth controlling mechanisms. Since the proliferation and
differentiation of somatic cells in higher organisms are regulated by multiple
controls, cancer cells probably are the end result of multiple changes which may
take years to develop. Several lines of evidence confirmed that the natural history
of spontaneously occurring human and animal cancers is usually a multi-step
process.

First, statistical analyses have shown a rapid increase in the incidence of
cancer with age among most of the important human cancers (Fisher and
Holloman, 1951; Armitage and Doll, 1954; Dix, 1989). Results of these analyses
suggest that multiple events may accumulate with time, and thus are responsible

for the increased cancer rates in old age. Peto et al., (1975) further demonstrated

9)

kn

”u.

..
”he

6

that the intrinsic effects of ageing such as falling immunological surveillance or age-
related hormonal changes were not required to explain the vastly increased
incidence of cancer in old age, and that age was a reflection of the duration of
exposure to carcinogenic agents. They used mice at 10, 25, 40, and 55 weeks of
age in a skin carcinogenesis experiment with benzpyrene. The rate of malignant
epithelial tumors in each group increased steeply with time and the increase was
independent of age at the start of exposure.

Second, studies of chemical carcinogenesis have defined stages into which
carcinogenesis may be divided: initiation, promotion and progression (Diamond at
al., 1980; Pitot, 1990). Each stage is induced independently by a different class of
chemical agents.

Third, histological evidence can somewhat arbitrarily be used to divide the
neoplastic development into three phases: initiating, intermediate, and advanwd
(Foulds, 1975). The initiating phase is either clinically "silent" or manifested only
by apparently trivial and dubiously neoplastic lesions such as hyperplasia. The
intermediate stage is characterized by the emergence of "precancerous" or
"premalignant" lesions which may progress into the advanced stage, or persist
indolently for a long time with minimal growth and no qualitative change, or regress
completely. The last stage is characterized by the presence of malignant
carcinomas or sarcomas. The phenomenon of tumor regression suggests the
requirement of multiple factors for preneoplastic cells to develop a fully malignant
phenotype.

Fourth, cytogenetic studies of tumors reveal a multiplicity of chromosomal

abnormalities in all human cancers (Mitelman, 1991; Solomon et al., 1991).

r- JP
:60! U

test;

("I

7

Examination of colorectal carcinoma (Fearon and Vogelstein, 1990) and
neuroblastoma (Knudson and Meadows, 1980) indicate that additional cytogenetic
lesions are associated with the advanced stage of these malignancies.

Finally, the most direct evidence comes from the discovery of chicken RNA
tumor viruses, AMV-E26, MH2, and AEV-ES4 strains, which carry dual oncogenes
(ets-myb, myc-mil, and erb-A-erb—B, respectively). The full oncogenic potential of
these viruses requires both oncogenes (Cole et al., 1983 and Jansen et al., 1983
for MHZ; Leprince et al., 1983, Nunn et al., 1983 and Kan et al., 1983 for E26;
Frykberg et al., 1983 for ES-4). Viruses possessing only one of these oncogenes
shows more limited tissue specificity and longer latency for tumor development.
Similar oncogene cooperation in tumor development has also been demonstrated
in DNA transfection experiments. Land and coworkers (1983a; 1983b) showed
that myc and ras could fully transform primary rat embryo fibroblasts when
transfected together but not singly. Schwartz et al. (1986a) further extended the
synergistic effect of the myc and ras oncogene to the transformation of murine B
cells. Other examples of oncogene cooperation involved in tumorigenesis have
recently been summarized by Hunter (1991).

In conclusion, through decades of studies, researchers have been able to
correlate the multifactorial characteristics of neoplasia from the histological and
cellular levels to the molecular level. Identification of genetic lesions and their
effects in tumorigenesis not only benefits our understanding of the growth and
differentiation of normal and cancerous cells, but also provides insights on

prevention or therapy of cancer.

8

2.1 The search for genes that are involved in tumorigenesis

Studies of mechanisms of chemical carcinogenesis (Pitot, 1990; Balmain
and Brown, 1988) and heritability of cancers (Ponder, 1990) have strongly
suggested that the lesions of tumors reside in the genetic material. The search for
putative oncogenes in a genome size of 3 x 109 base pair containing approximately
10‘ to 105 genes was greatly simplified by the discovery of viral oncogenes in
animal RNA tumor viruses. The first RNA tumor virus was isolated by Rous (1910),
but it was not until the development of molecular biology that the identification of
the v-src oncogene became possible (Stehelin et al., 1976a). Twenty six viral
oncogenes have been identified from animal RNA tumor viruses (Varmus, 1989),
and shown to exhibit sequence similarity with cellular genes (later termed
protooncogenes) by hybridization and sequencing analyses. Since the cellular
protooncogenes are also found to be conserved among different species, it
stresses the important roles that these genes may play in regulating growth and
differentiation. Thus, alteration or destruction of these genes may contribute to
tumor progression.

The list of protooncogenes has been expanded by four major approaches
(i) gene transfer, (ii) analysis of known chromosomal translocations or
amplifications, (iii) mapping of viral integration sites in virally induced tumors, (Iv)
sequence homology to the known oncogenes. Some of these genes were

repeatedly identified by different methods.

2.1.1 Identificatlon of cellular oncogenes by DNA transfer experiments

The most common method of gene transfer has utilized DNA transfection

9
and focus formation (Graham and van der Eb, 1973; Shih et al., 1979). DNA is

isolated from tumor cells and introduced into recipient cells, usually an
immortalized mouse fibroblast line, NIH3T3. NIH3T3 cells have two major
advantages other than their easy cultivation: (i) they are flat, contact-inhibited cells
that form a monolayer culture, and (ii) their DNA can be distinguished in the
hybridization experiments from the DNAs of other species, thus allowing
identification of donor DNA. The first property allows one to examine the
transfected cultures for the appearance of morphologically altered (transformed)
foci due to the expression of an introduced oncogene. Examples of cellular
oncogenes identified by this procedure are N-ras from neuroblastoma (Shimizu et
al., 1983), trk from colon carcinoma (Martin-Zanaca et al., 1986), and others as
listed in Table 1.1.

Oncogenes of the ras family predominate among those found fin about 20%
of human tumors tested)(Der et al., 1982; Parada et al., 1982; Santos et al., 1982),
whereas cellular analogues of other viral oncogenes are less frequently detected
by this method. This could be due, at least in part, to the ability of NIH3T3 cells
to respond morphologically to a given oncogene product, to the need for the gene
to be genetically dominant, and to the need for the gene to be sufficient for the
expression and the maintenance of the transformed state. In other words, NIH3T3
cells may be the wrong lineage or species to respond to the effects of certain
oncogenes. In addition, NIH3T3 cells may have a preneoplastic phenotype as they
already have the ability to grow continuously in culture. Thus, they may be more
susceptible to genes whose effects are manifest during the later stages of tumor

progression.

10

Table 1.1 Oncogenes identified by DNA transfer experiment

 

 

Gene Sourcec Function Reference

N-ras Neuroblastoma G-protein-like Shimizu et al., 1983

neu Rat neuroglloblastoma EGF receptor like Shih et al., 1981
Bargmann et al., 1986

mas Epidermoid carcinoma Angiotensin receptor Young et al., 1986

hst Kaposi’s sarcoma FGF family member Delli Bovi et al., 1987
Taira et al., 1987

trk Colon carcinoma Receptor-like Martin-Zanaca at al., 1986

met Osteosarcoma cell line Receptor-like Cooper et al., 1984a

ret T cell lymphoma Receptor-like Takahashi et al.. 1985

dbla Diffuse B cell lymphoma Cytoskeletal matrix Eva and Aaronson. 1985

associated phosphoprotein Graziani et al., 1989

mcf.2a Mammary carcinoma cell line as that of dbl, a Fasano at al., 1984

Ice Hepatocellular carcinoma b Ochiya et al., 1986

mel Melanoma cell line b Padua at al., 1984

raf Stomach cancer Serine/threonine kinase Shimizu et al.. 1985

res Mammary carcinoma cell line Receptor-like Birchmeier at al.. 1985

 

a: these two genes were found to be two different activated versions of the same protooncogene (Noguchl et al., 1988)
b: not yet defined
c: human tumors if not specified

A

2.1.2 I

11

In another aspect, oncogenes that are isolated from RNA tumor viruses
carrying a single oncogene probably contain multiple mutations within each
oncogene and, therefore, are capable of inducing tumors as a single gene. Thus,
the failure of detecting cellular counterparts of other viral oncogenes by this
method may simply be attributed to the fact that the cellular counterparts did not

contain multiple mutations to induce focus formation of NIH3T3 cells.

2.1.2 Characterization of viral flanking regions

Studies of RNA tumor viruses such as Avian leukosis virus (ALV), Moloney
murine leukemia virus (MoMuLV), and Feline leukemia virus (FLV) that do not carry
oncogenes reveal another mode of transformation. Insertion of the viral genome
into a cellular genome may activate adjacent cellular genes through enhancer or
promoter sequences of the viral LTR (Hayward et al., 1981; Fung et al., 1981; for
a review see Nusse, 1986a). Therefore, cloning of flanking regions of frequent viral
integration sites may allow the identification of putative oncogenes. A variety of
genes and “loci" have been isolated by this method. Protooncogenes which
contain homology to known viral oncogenes and other known cellular genes are
shown in Table 1.2, and "loci" that do not show homology to known viral
oncogenes and other known cellular genes are shown in Table 1.3. These “loci"
may correspond to novel oncogenes or may be linked to protooncogenes at a
distance.

Among the "loci" listed in Table 1.3, int-1, int-2, and pim-1 have been shown
to contain detectable transcriptional activity and to encode proteins with distinct

functions (Nusse et al., 1984; Dickson et al., 1984; Cuypers et al., 1984; Selten et

Tabl

J

«n-
p
v

I?

Table 1.2 Cellular genes activated by proviral Insertion

12

 

 

 

Gene Source Virus Reference

c-myc Chicken bursal lymphoma ALV Hayward et al., 19813
Mouse T cell lymphoma Souls-MuLV Adams et al., 1982‘
Cat T cell lymphoma FeLV Nell et al., 1984a

N-myc Mouse T cell lymphoma Moloney MuLV van Lohuizen et al., 1989a

c-erbB Chicken erythroleukemia ALV Fung et al., 1983

c~myb Mouse lymphosarcoma defective MoMuLV Shen-Ong et al., 1984

c-Kl—ras Mouse myeloid cell line Friend MuLV George et al., 1986

c-Ha-ras Mouse T cell leukemia Moloney MuLV Ihle et al., 1989

lL2 Ape T cell lymphoma cell line GaLV Chen et al., 1985

lL-3 Mouse myelomonocytlc leukemia IAP Ymer et al., 1985

CSF-1 Mouse monocyte tumor endogenous Baumbach et al., 1988

ecotropic provirus

GM-CSF Mouse promyelocytic cell line lAP,R-MuLV,F-SFFV Stocking et al., 1988

c-mos Mouse plasmacytoma line IAP Canaanl et al., 1983

p53 Mouse erythroleukemlc cell line Friend MuLV Hicks and Mowat, 1988

abbreviation:

MuLV: murine leukemia virus
IAP: intracisternal A particle

FeLV: Feline leukemia Virus
ALV: Avian Leukosis Virus

R-MuLV: Rauscher murine leukemia virus
F-SFFV: Friend spleen focus forming virus
GaLV: Gibbon Leukemia Virus

‘2 only the earliest reference is shown

T:

.K e ‘ s A.
a C-
A\.V

13

Table 1.3: VIraI flanking regions without viral oncogene analogues.

 

 

 

Name Source Virus mRNA Reference
Evi-I AKXD myeloid tumor Cas—Br-M MuLV Mucenskl et al.,1988a
MCF yes Morlshita et al., 1988
MM NIH/Swiss or SIMS Moloney MuLV nd Poirier et al., 1988
pre-B cell lymphoma
Fis-f AKR lymphoma Friend MuLV nd Silver and Kozak, 1986
' or BXD-2 leukemia
Dsi-f Fisher rat thymoma Moloney MuLV nd Vijaya et al., 1987
Pim-2 Balb/c or Moloney MuLV nd Breuer et al., 1989a
CS7BL10 lymphoma
Pim-I Balb/c or AKR T lymphoma MCF yes Cuypers et al., 1984
Int-1 03H Mammary carcinoma MMTV yes Nusse and Vannus, 1982
Int-2 C3H Mammary carcinoma MMTV yes Dickson et al., 1984
Int-41 mouse mammary and MMTV yes Garcia et al., 1986
kidney adenocarcinoma
MIvi-t' rat thymoma Moloney MuLV nd Tsichlls et al., 1983a
(PW-1)
MIvi-2 rat thymoma Moloney MuLV nd Tsichlls et al., 1984
MIvl-3 rat thymoma Moloney MuLV nd Tsichlls et al., 1985a
MIvi-4 rat thymoma Moloney MuLV nd Lazo et al., 1990
Gin-1 mouse thymoma Moloney MuLV nd Villemur et al., 1987
Spi-t mouse erythroleukemia SFFV yes Moreau-Gachelln at al.,1988
Bic Avain B cell lymphoma ALV yes Clunnan and Hayward, 1989
Abbreviation:

MuLV and ALV: same as those in Table 1.2
MMTV: Mouse mammary tumor virus
MCF: mink cell focus forming virus
Cas-Br-M MuLV: Casitas Brain Mousetropic MuLV
SFFV: spleen focus forming virus
‘: also known as mis-f

14

al., 1986; Nusse, 1988), and to exhibit transforming ability in transgenic mice
(Adams and Cory, 1991) or the in vitro transformation of an epithelial cell line (Int-1
only; Brown et al., 1986). On the other hand, many of them do not have
detectable transcriptional activity. Since the enhancer element of the viral LTR may
function over a long distance, it is possible that certain putative oncogenes may
be located distal to the breakpoints of viral integration sites. Techniques such as
chromosomal walking, jumping and YAC cloning may facilitate the identification of
such genes. Additionally, exon trapping (Duyk et al., 1990; Buckler et al., 1991),
a method used to identify any exon sequence from cloned genomic DNA may also
be used to facilitate the identification of any transcription unit. Once a transcription
activity is associated with a “locus, it is, however, necessary to test for its
transformation potential. Either the in vitro focus formation assay (see DNA
transfer experiment) or the transgenic animal model (see below) can be used for
that purpose.

The spectrum of oncogenes that can be identified by this method relies on
the randomness of viral integration. Although retrovirus integration has been
shown to have preferred target sites (Shih et al., 1988), these preferred sites
comprise only 20% of all the viral integration sites. Thus, this preference does not
seriously affect the variety of oncogenes that may be identified by this method. In
addition, the nature of this preference is unknown. Perhaps it resides in the active
transcription of preferred regions. The fact that both cellular counterparts of viral
oncogenes and novel oncogenes are isolated by this method makes it a feasible
way of searching for a diverse class of genes that are involved in the regulation of

growth and differentiation. It has clearly allowed identification of a wider variety of

3‘1

or

15

oncogenes than transfection and focus formation experiments.

Interestingly, no human oncogene identified thus far is activated or
inactivated by viral integration. However, the discovery of multiple copies of
endogenous virus-related sequences in human cells suggests that gene activation

through this mode may still be possible (Bonner et al., 1982; Callahan et al., 1985).

2.1.3 Molecular Characterization of chromosomal aberrations

Chromosomal aberrations have long been associated with human cancers.
Karyotype analyses of metaphase chromosomes has found some abnormalities
specific for distinct tumor types (for a recent review, Mitelman, 1991). The
discovery of the Philadelphia (Ph1) chromosome in bone marrow cells of patients
with chronic myelogenous leukemia (CML) actually laid the foundation for the
association of specific chromosomal anomalies with a particular neoplasm (Nowell
and Hungerford, 1960). It was not until recently, however, that the molecular basis
of these cytogenetic lesions has been elucidated as a result of the improvement
of both cytogenetic and molecular cloning methods. Methotrexate treatment
(aminopterin) (Yunis, 1976; Hagemeijer et al., 1979) is to block DNA synthesis
pathways is blocked so that cells can be collected at the S phase of a cell cycle.
The block is then released by adding thymidine to allow DNA synthesis using the
salvage pathway. A large number of cells enter mitosis synchronously following
these treatments. These synchronized cells are then treated briefly with colchicine
followed by standard banding procedures (Chaudhuri et al., 1971; Caspersson et
al., 1970; Hsu, 1974). Chromosomes obtained by this procedure are less

condensed, allowing 1200 bands to be identified. The increased banding

IVA
Ni V

t:-

Va;

16

resolution not only increases the accuracy of identification of chromosomes, but
also leads to the recognition of many karyotypic changes and the association of
these changes with particular neoplasias. Cloning of genes located in
translocation junctions or other chromosomal aberrations has been done by
conventional cloning methods, ie. use of probes for genetic markers near the
locus of interest, and chromosomal walking using a genomic library until the
authentic gene is identified by in situ hybridizations or southern blot analyses.

The introduction of microdissection and microcloning should advance this
process. Microdissection was first used by Scalenghe et al.(1981) to clone DNA
from Drosophila polytene chromosomes. It was later used in the human genome
mapping project (Bates et al., 1986; Kaiser et al., 1987), and to clone genes
involved in human genetic diseases (LOdecke et al., 1989; Kondo et al., 1984). In
this technique, a very small segment of the chromosome of interest can be
dissected with needles from banded chromosomes by using an electronically
controlled micromanipulator. Use of this method in combination with microcloning
in which the technique of polymerase chain reaction (PCR) was incorporated
(Edstrom et al., 1986) can reduce the number of genomic clones required to
screen for the locus of interest compared to that in conventional cloning
procedures. Adapting these approaches may greatly facilitate the cloning of genes
that reside at the site of chromosomal lesions.

A recent review by Solomon et al. (1991) summarizes genes identified from
chromosomal aberration sites (Table 1.4). Many of them are either analogues of

viral oncogenes, or genes involved in the cell cycle or differentiation. More

Tab

(5 )
(b
-3
1 I

I!" "
a... c
A
Axe“:
v. _
l
L‘
. 1
“1
Ta' :
L .U
A“.
IU- ‘:
N’tnd
U 4

F.
,tf
U l

33:4

17

Table 1.4 Genes Identified from chromosomal aberrations

 

Gene

Source

Protein type

Reference

 

chromosomal amplification

 

N-myc
L-myc

Neuroblastoma
Small-cell
lung cancer

Nuclear protein
Nuclear protein

Schwab et al., 1983
Little et al., 1983

 

chromosomal translocation or inversion

(adapted from Table 1 In Solomon et al. (1991))

 

 

 

Gene Disease Rearrangement Protein type
myc Burkitt lymphoma t(8;14)(q24;q32) HLH domain
t(2;8)(pi 1 ;q24)
t(8;22)(<124;q11)
T-ALL t(8;14)(q24;q11)

BcI—I B-CLL t(11;14)(q13;q32) GI cyclln-Ilke

Bel-2 Follicular lymphoma t(14;18)(q32;q21) Inner mitochondria membrane

BcI-3 BCLL t(14;19)(q32;q13) CDC 10 motif

lL-3 Pre-B ALL t(5;14)(q31;q32) Growth factor

Lylt T-ALL t(7;19)(q35;p13) HLH domain

TcI-s T-AII t(1 ;14) (p32) (q1 1) HLH domain

Rbntl T-ALL t(11;14)(p15;q11) UM domain

Rbnt2 T-ALL t(11;14)(p13;q11) UM domain

Tan1 T-ALL t(7;9)(q35;q34) Notch homolog

Hox11 T-ALL t(10;14)(q24;q11) Homeo domain

Pth Parathyroid adenoma inv(11)(p15;q12) Deregulate myc

Big-1 B-CLL t(8;12)(q24;q22) Deregulate myc

Bcr-Abl CML, B-ALL t(9;22)(q34;q11) Bcr. Gap for p21'“
Abl, Tyrosine kinase

Pml-Rara APL t(15;17)(q22;q11-12) Pml, Zinc finger
Rare, Zinc finger

Dek-Can AML-M2,-M4 t(6;9)(p23;q34) Dek, nuclear protein
Can, cytoplasmic protein

EZA-Pbx Pre-B ALL t(1;19)(q23;p13) 52A, HLH domain
Pbx, homeodomain

Rel-Nrg NHL lns(2;2)(p13;p11-I4) Rel, NF-kB family
Nrg, no homology

Abbreviation:

CLL: chronic lymphocytic leukemia
ALL: acute lymphocytic leukemia
CML: chronic myelogenous leukemia
AML: acute myelogenous leukemia
NHL: non Hodgkin’s lymphoma

{1‘0 "

at":

i839

MJ‘:
U‘v

2.1.

(L)
r I)

I..."
U

18

significantly, tumor suppressor genes (such as RB and WT-I; Marshall, 1991,
recently reviewed by Weinberg, 1991) and genes associated with apoptosis (such
as bcI-2) were also discovered. Alteration of the genomic structure of these genes
may deregulate their expression and participate in the transformation of cells, and

consequently lead to tumor formation.

2.1.4 Identification of new oncogenes by homologous sequences

Another approach not commonly used, but potentially useful, is to search
for sequences homologous to known oncogenes assuming the conservation of
important regulatory genes. Given the fact that all the genes in the src family
(Hunter and Cooper, 1985) and ras family (Barbacid, 1987) share significant
homologies among certain genetic domains (such as the effector domain of ras
genes and the carboxy-terminal region of src-related genes), hybridization probes
for those regions may identify new members of these gene families. The Ick
oncogene (Marth et al., 1985) was first isolated from a murine T cell lymphoma by
using an oligonucleotide probe for the conserved major tyrosine phosphorylation
site.

Other proliferation and differentiation related genes may also be identified
from cells on the basis of reversion from the transformed phenotype. Such cells
have been generated either by cell-fusion with normal cells (Harris, 1988) or DNA
transfection. An example of a gene identified by the latter method is the Krev-I
gene, which is responsible for the revertant phenotype of Kirsten sarcoma virus-
transformed NIH3T3 cells (Kitayama et al., 1989; Noda et al., 1989).

The oncogenic potential of the putative oncogenes identified from the above

19

procedures should be tested to verify their active roles in oncogenesis. The

following sections describe general approaches used for this purpose.

2.2 Experimental models for transformation and tumor progression
2.2.1 Use of animal versus human models

Animal model systems are obviously required for experimental in vivo
carcinogenesis studies, since it is ethically unacceptable to use human subjects.
The compatibility of results from animal models with those obtained in humans has
been supported by several observations. First, the conservation of oncogenes
among species suggests their common properties and regulation (Varmus, 1989).
Additionally, human oncogenes are able to transform mouse fibroblasts in DNA
transfection experiments (see above) and induce tumors in transgenic mice
(Palmiter and Brinster, 1986). Second, growth factors and growth regulatory
mechanisms do not appear to be species specific since transplanted tumors from
a great range of species can grow in immunodeficient, athymic (nude) mice while
normal tissue can not (Stiles and Kawahara, 1978). These results suggest the
universality of the genetic events causing neoplasia, and the validity of animal
models in studying human carcinogenesis. The fact that laboratory animals can
be environmentally and genetically controlled makes them an even better system
to work with.

On the other hand, the use of human tissues and cells does offer some
unique advantages. First, some rare forms of human cancer reflect inherited,
predisposing conditions. Their genetic basis and perhaps common pathways to

carcinogenesis may be understood through the study of nontumorous cells from

ANN

K...

20

affected individuals (Ponder, 1990). Moreover, the fact that human cells are
genetically more stable in vitro than most rodent cells (DiPaolo, 1983) makes them
especially suitable for studying the multiple steps of tumorigenesis. Lastly. the
findings from studies of human cells complement and validate the results derived

from studies of laboratory animals.

2.2.2 Use of in vivo versus in vitro systems

The role of the host, not only in controlling, but also in facilitating the
development and spread of cancer has already been established (for a review, see
Alexander, 1987). These host factors interact with the genetic lesions that reside
in the tumor cells themselves. The in vivo system therefore has an essential role
in cancer studies because the integral multi-systemic interactions of the organism
remain intact.

The disadvantage of using in vivo models for studies of tumor progression
is its inconvenience for dissecting events within the progression. Most often an
endpoint must be chosen for analysis and intervening events must be implied. An
in vitro model using cell or tissue culture, however, provides opportunities to
dissect the processes controlling growth, differentiation, neoplastic transformation
and tumor progression of cells, and to describe their mechanisms in biochemical
or molecular terms.

2.2.3 Model systems used to test the oncogenic potential of oncogenes

One commonly used in vitro culture system, NIH3T3 cells, although allowing
the detection of certain oncogenes, presents limited sensitivity since this method

tends to identify oncogenes in the ras gene family as discussed previously.

LEG.”

AT; 4

3""
' U

A~F
‘ v

't‘ J

21

Various tissue culture systems including those of epithelial and hematopoietic
origins (Gabrielson and Harris, 1985; Rheinwald and Green, 1975; Deeh, 1985;
Lechner

“T21%* ’eck, 1985; Dexter et al., 1977; Whitlock and Witte, 1982) were established
and have allowed the study of gene interactions within differentiated cell types.
Among them, the long-term B cell culture system of Whitlock and Witte (1982) is
exclusively used in our studies. This culture system can provide not only a wide
range of targets for transformation within the B cell developmental series, but also
allows the culture of cells that are not transformed or of an intermediate
transformed phenotype. It thus provides an excellent tool to test the transforming
potential of oncogenes in the B cell lineage, and to elucidate the mechanisms of
transformation on the molecular level.

In vitro human B cell lines such as the lymphoblastoid cell lines (LCLs) are
commonly used for the same purpose. These cell lines are derived from the
infection of human peripheral blood with Epstein-Barr virus (EBV) (for review see
Nilsson and Klein, 1982). Examples of the uses of these lines in testing oncogenic
potential and oncogene cooperativity of putative B cell oncogenes are described
in human B cell neoplasia (see below).

Transgenic mice can also be used to assess the transforming potential of
various oncogenes (for review see Palmiter and Brinster, 1986; Adams and Cory,
1991). Using tissue specific regulatory elements, one can examine the oncogenic
potential of the gene of choice in a particular tissue. For example, the enhancer
sequences of immunoglobulin genes (Eu) have been used to test oncogenes such

as c-myc and bcI-2 in the oncogenesis of B-cell neoplasia (Adams et al., 1985;

22

McDonnell et al., 1989). In addition to its advantages as an In vivo system (as
described previously), it also provides the only means for studies of tissue specific

oncogenesis where an in vitro culture system is lacking.

2.2.4 Model systems used to study tumor progression

Two of the most established genetic systems, the mouse skin carcinoma
(for a review see Balmain and Brown, 1988) and human colon carcinoma (Fearon
and Vogelstein, 1990), share the advantage of ease in access to and recognition
of tumor samples at different stages of progression. In contrast, tumors of
lymphoid and other organs lack a good animal model to study molecular changes
during the tumor progression period, mainly due to the difficulty of identifying
preneoplastic cells and obtaining cells at varying stages of neoplastic development
in a single host (i.e., the collection of neoplastic samples of these tissue types
often requires termination of the donor animal). Even if the experimental animals
were blindly sacrificed at different time points to collect tumors at different stages
of neoplastic development, one still encounters the problem of insufficient numbers
of cells to perform molecular analysis. In vitro cultures would be required to
expand these isolates and such in vitro tissue culture systems are lacking for many
of these tissue types. Although the results generated from the above in vivo
systems may be applicable to tumors of other origins, evidence of the association
of certain oncogenes with tumors of particular tissue types from studies of
transgenic mice and inherited cancers (Adams and Cory, 1991; Marshall, 1991;
Lanes et al., 1981 & 1982; Cooper and Neiman, 1980; Padhy et al., 1982) have

suggested the need for tissue specific model systems. Although the use of

23

transgenic mice has proven to be a useful in vivo model to test the ability of an
oncogene to initiate tumor formation, little knowledge is gained about the
secondary events that have contributed to the progression of tumors. Tests of
oncogene cooperativity either involve the construction of transgenic mice with
multiple oncogenes or performing crosses with transgenic mice carrying different
oncogenes (Adams and Cory, 1991). These procedures may involve intensive
labor. Moreover, the cooperative effect is unavoidably under the influence of
unnatural environments because cells of all tissue types may express the activated
oncogenes. The use of tissue-specific enhancer sequences may avoid this
problem; however, these sequences are not available for all tissue types. One
approach used to study oncogene cooperation in the tumor progression of
transgenic mice is to infect transgenic mice with retroviruses to facilitate tumor
progression (van Lohuizen et al., 1991).

The Whitlock and Witte culture system (1982) allows certain advantages in
the study of oncogene cooperativity. By comparing the transformed phenotypes
of B cells generated after the introduction of two or more oncogenes of interest
into this culture to those of B cells generated from cultures with a single oncogene,
one can determine whether one oncogene can complement another in
transforming B cells. This procedure is relatively easy as compared to that of
transgenic mice. It also provides a faster way of testing whether there are
temporal effects on the acquisition of genetic alterations in tumor progression.

In summary, using the methodologies described above, extensive studies
have been performed to uncover the basis of tumorigenesis. Among them,

hematopoietic neoplasias are the best characterized tumors at the molecular

WE

24

genetic level due to their frequent occurrence in animal models and their high
mitotic index, allowing for cytogenetic analyses (see above). Many alterations of
protooncogenes have been specifically associated with different subtypes of
leukemias/lymphomas of different species, and are summarized elsewhere
(Schwartz and Witte, 1988; Solomon et al., 1991; Sawyers et al.,1991). Here, I will
describe exclusively the molecular characterization of B cell lymphomas/leukemias

of human, mouse, and avian origin, since this has been the focus of my thesis.

3. Multiple genes or factors that are Involved in B cell lymphomas/leukemias
of human, mouse, and avian

3.1 Human B cell neoplasia
3.1.1 Burkitt’s lymphoma

Burkitt’s lymphoma (BL) is the best characterized human B cell tumor. It
was first described by Burkitt (1958) as a distinct clinicopathological entity
occurring with high frequency in the jaws of children from Central Africa. Non-
endemic (or sporadic) BL is also found in areas outside of Africa but the rate of
incidence is low (O’Conor et al., 1985; Philip et al., 1982). BL is a malignant
lymphoma comprising a monomorphic outgrowth of B lymphocytes. ‘lhe stage of
maturation of BL cells varies among individual tumors, from a near pre-B
phenotype (Preud’homme et al., 1975) to a more mature phenotype with
expression of predominantly lgM (Klein et al., 1967), and occasionally IgD, IgG, or
even IgA (Preud’homme et al., 1985). However, the cells never fully differentiate
to a plasmacytic phenotype.

The etiology of BL involves at least two steps. Infection with Epstein-Barr

Virus (EBV) appears to be the primary event in the development of African Burkitt’s

(TI

.E=l
VII

25

lymphoma from both genetic (Geser et al., 1983) and epidemiological evidence
(de-The’ et al., 1978; Geser et al., 1982). Geser and coworkers found that 96% of
African Burkitt’s lymphomas carry multiple cepies of the EBV genome in all their
cells. Both groups showed that a high multiplicity EBV infection early in the life of
African children contributes to the tumor development. Raab—Traub and Flynn
(1986) and Neri et al. (1991) have studied the termini of EBV episome genomes
in both endemic BL primary tumors and cell lines. The linear EBV DNA has
variable numbers of direct tandem 500 bp repeats at each terminus. Restriction
endonuclease analyses have indicated that the termini are uniformly clonal in all
cases. This result again strongly suggests that EBV infection has preceded, and
thus, most likely contributed to clonal expansion in these malignancies. Although
the lack of viral markers in 4% of the endemic BL and approximately 85% of the
non-endemic BL cases (Lenoir et al., 19843) may argue the against a necessary
causal relationship between EBV and BL, the fact that EBV can transform human
peripheral B cells into established cell lines (T akada and Osato, 1979; Robinson
and Smith, 1981) and that EBV can induce B cell lymphomas in marmosets (zur
Hausen, 1980) implies an important role for EBV in BL.

The immortalizing mechanism of EBV has been suggested to be mediated
through autocrine stimulation by the finding of the release of a Bee" growth factor
(BCGF) from Burkitt lymphoma cells and EBV infected cell lines (Gordon et al.,
1984). The precise consequences of this autocrine stimulation are unknown.
However, two EBV-encoded proteins are found to be involved in the process of
oncogenesis. The EBNA2 gene product, a nuclear protein, is required for

immortalizing B cells since EBV deleted for this gene fails to do so (Menezes et al.,

26
1975; Delius and Bornkamm, 1978; Bomkamm et al., 1980; Heller et al., 1981).

EBNA2 can also transiently stimulate cellular DNA synthesis upon transfection into
B cells (Volsky et al., 1984). Another EBV encoded protein, LMP1 (a latent
membrane protein), can transform established rodent cells (Wang et al., 1985;
Baichwald and Sugden, 1988). The linkage between the release of the BCGF and
the biochemical effects of these two proteins are becoming clear through the
following studies. EBNA2 is a multiple-function protein, and may act similarly to T
antigen of SV40 virus. EBNA2 was found to induce CD21, CD23, and LMP1
expression upon introducing this gene into cell lines which are infected with EBV
strains, deletion mutants that do not encode EBNA2 protein (Wang et al., 1990;
Abott et al., 1990). CD21 is a surface protein involved in human B cell
differentiation (T edder et al., 1984). CD23 is a membrane-bound B cell activation
marker, and acts as an autocrine BCGF for normal and transformed B cells when
shed (Swendeman and Thorley-Lawson, 1987). LMP1 was found to cooperatively
Induce 0023 expression, and to induce the expression of several cellular adhesion
molecules (Wang et al., 1990). The increased expression of the cellular adhesion
molecules may explain the phenotypic changes of LMP1 transfected cells including
growth in large tight cell clumps (Wang et al., 1988). Taken together, these
findings suggest that EBNA2 may immortalize B lymphoid cells by a autocrine
mechanism through induction of CD23. This autocrine stimulation is augmented
through the cooperation with LMP1, which may act via cellular second
messengers. Whether EBNA2 acts directly or through other cellular intermediates
in the upregulation of CD23 transcription remains to be determined.

The second step in the development of Burkitt’s lymphoma may involve

' A;
fls'y\
U“ '.

MFA

I‘y’Y‘. 7'

F x
e":

Crag

27

deregulation of the c-myc gene through chromosomal translocations of
chromosome 8 on which it residues. The almost universal presence of
chromosomal translocations between chromosome 8 and one of three other
chromosomes (chromosome 14, 2 and 22) is another characteristic of Burkitt
lymphoma. Only one exception to this has been reported (Zech et al., 1976).
Translocation t(8;14)(q24;q32) was found in approximately 80% of cases of Burkitt
lymphoma (Croce and Nowell, 1986), while the remainder of these tumors carry
t(2;8)(p11;q24) or t(8;22)(q24;q11) translocations. It is striking that each of these
translocations involves the cytogenetic location of one of the immunoglobulin (lg)
loci. The heavy-chain gene is located at chromosome 14q32 (Croce et al., 1979);
the kappa and the lambda genes are at 2p11 (Malcolm et al., 1982) and 22q11
(Erikson et al., 1981), respectively. This may suggest a molecular relationship
between the rearrangement of Ig genes and oncogenesis, perhaps through a
aberrant recombination between two chromosomes. The identification of the
crossover point of the translocation on chromosome 8 was facilitated by the
discovery in murine plasmacytomas of a similar translocations (see below).

The involvement of the c-myc gene in these chromosomal translocations
was suggested by the finding of c-myc gene activation in an abortive
immunoglobulin gene recombination in a mouse plasmacytoma (Shen-Ong et al.,
1982). Additionally, the c-myc gene was mapped to chromosome 8q24 using
human/rodent somatic cell hybrids by Dalla Favera et al. (1982). Finally, southern
blot analyses showed that the c—myc gene was rearranged in approximately 50%
of Burkitt’s lymphomas examined (Bernard et al., 1983; Dalia Favera et al., 1983;

Taub et al., 1982). Molecular cloning ultimately enabled the detailed analysis of the

28

translocation break points and demonstrated that the translocations join sequences

from the lg genes to regions surrounding c-myc (reviewed by Haluska et al., 1987).

The result of all three translocations is the constitutive expression of the c-
myc gene. In contrast to the high levels of c-myc expression associated with the
viral transformation of avian B-cell lymphomas (see below) and human tumor cell
lines that contain amplified c-myc genes (Collins and Groudine, 1982; Delia Favera
et al., 1982), high levels of c-myc expression are not found consistently in all
Burkitt’s lymphoma cell lines. A variety of mechanisms have been proposed to
account for the myc/lg juxtaposition in its contribution to tumorigenesis (review by
Klein and Klein, 1985). These include abnormally high transcription rates,
abnormal transcription size, changed promoter usage, mutations, and lack of
transcriptional pausing at exon 1 of c-myc gene. None of the above mechanisms
seems to be applicable universally to all or even most of the tumors. The current
idea is that c-myc is deregulated as a consequence of cis-acting sequences
associated with the constitutively active lg region. The result of such deregulation
would be to keep the cells in continuous division through a loss of c—myc’s normal
regulation.

The oncogenic potential of c-myc genes in Burkitt’s lymphoma was tested
by Lombardi et al. (1987). They introduced an expression vector containing
constitutively expressed c-myc into EBV-infected human B lymphoblastoid cell lines
(LCLs)(Nilsson and Klein, 1982). The resulting myc-transfeCted LCLs displayed a
reduced serum requirement for growth, an increased soft-agar cloning efficiency,

and an increased tumorigenicity in nude mice as compared to those of vector-only-

29

transfected LCLs. Their results indicate a contribution of the c-myc gene in the
tumorigenesis of Burkitt’s lymphoma, and a cooperation between two putative
immortalizing functions. However, the inability of myc-transfected LCLs to form
colonies as efficiently as BL cell lines, and their low tumorigenicity have suggested
that additional events are required for a fully tumorigenic phenotype.

In addition to the common events of EBV infection and c-myc translocation,
p53 mutations were found associated with 9/27 biopsies of BLs and 17/27 BL cell
lines (Gaidano et al., 1991). The p53 gene encodes a 53-kDa nuclear
phosphoprotein that may be involved in the negative regulation of cell growth
(Lamb and Crawford, 1986; for review see Iman and Harris, 1991). Additionally,
p53 may differ from prototypical tumor suppressor genes in that at least some p53
mutant alleles can behave as dominant oncogenes by transforming target cells in
vitro and causing tumorigenesis in transgenic mice even in the presence of the
normal allele (Halevy et al., 1990; Lavigueur et al., 1989). Gaidano and coworkers’
results suggest both loss of function and dominant negative mutations are present
in BLs.N-ras activation was found in a sporadic Burkitt’s lymphomas (Murray et al.,
1983; Lenoir et al., 1984b). The effect of the ras oncogene in the tumorigenesis
of Burkitt’s lymphoma has been studied by Seremetis et al. (1989). Introduction
of activated N-ras or H-ras oncogenes into EBV immortalized LCLs has led to
malignant transformation of these cells. However, their results show that these ras
genes are also capable of inducing terminal differentiation of LCLs into plasma
cells, and therefore, may have implications in the pathogenesis of terminally
differentiated B-lymphoid malignancy such as multiple myeloma rather than in

Burkitt’s lymphoma.

30

3.1.2 Other human B-cell Ieukemlas and lymphomas

Most of the human B-cell leukemia or lymphoma associated oncogenes that
will be described here were identified by molecular cloning of frequent or
nonrandom translocation sites, and are recently reviewed by Solomon et al. (1991).
It should be noted that unlike the tight association of the c-myc gene with both
Burkitt’s lymphoma and murine plasmacytoma, some of the genes described in
this section are not invariably found to be activated in all tumors of a particular
type. Their impact on the progression of particular neoplasia requires further
investigation.
B-cell chronic lymphocytic leukemia

Four translocations t(1 1 ,14)(q13,q32), t(14,19) (q32,q13), t(8,12)(q24,q22),
and t(18;22) have been found in B cell chronic lymphocytic leukemia (B-CLL). The
t(11,14)(q13,q32) translocation also occurs in some diffuse small-cell lymphocytic
leukemias and diffuse large-cell lymphomas (Yunis, 1983), and multiple myelomas
(van der Berghe et al., 1984). The chromosome 11 breakpoints in two CLL
patients have been shown to occur only seven nucleotides away from each other,
whereas the breakpoint in a diffuse B-cell lymphoma is approximately 0.9 kb
distant from those characterized in CLL (T sujimoto et al., 1984a, 1985a). The
break point cluster region of this translocation has been denoted as bcI-1. No
transcription unit has been detected in this region other than the closely linked
PRAD1 gene (Lammie et al., 1991), a gene which was found to be a putative
oncogene in parathyroid adenoma (Arnold et al., 1989; Friedman et al., 1990;
Rosenberg et al., 1991). PRAD1 encodes a G1 cyclin-like protein (Motokura et al.,

1991). Cyclins can form a complex with and activate p34°""2 protein kinase,

31

thereby regulating progress through the cell cycle (for review see Nurse, 1990).
PRAD1 mRNA is expressed in many tissues and is highly conserved in bovine and
murine tissues (Rosenberg et al., 1991). PRAD1 has been implicated in non-
parathyroid neoplasia, squamous cell and mammary carcinomas, and is invariably
amplified and overexpressed in these tumors (Lammie et al., 1991). No direct
demonstration of the altered expression of PRAD1 has been reported in B-CLL as
this thesis is being written. Presumably, the disruption of the cell cycle after
alteration of this gene may contribute to the course of B-CLL as in the other
tumors.

The t(14,19)(q23,q13) translocation involves a deregulation of the col-3 gene
(McKeithan et al., 1987; Ohno et al., 1990). As a result of this translocation,
chromosome 19 sequences including the bcI-3 gene are juxtaposed to the 5’ end
of the 0:1 switch region of the IgH gene on chromosome 14 in a head to head
manner. The bcI-3 transcription unit is not disrupted by the translocation, and a
more than 3.5 fold increase of the mRNA level was found in total RNA from the
peripheral blood of two CLL patients with the t(14;19) translocation as compared
to RNA from a patient with the prolymphocytic variant of CLL, which does not
contain this translocation. The bcI-3 gene encodes seven tandem copies of the
cdc10 motif, a proline rich N-terminal, and a proline-serine rich C-tenninal (Ohno
et al., 1990). The cdc10 motif was previously identified in yeast genes that regulate
events at the start of the cell cycle (for review see Simanis et al., 1987) and in
invertebrate transmembrane proteins involved in cell differentiation pathways
(Austin and Kimble, 1987; Seydoux and Greenwald, 1989; Stemberg and Horvitz,

1989). It is not clear which of these two classes of proteins that the bcI-3 gene

32

resembles, but it is clear that it is not a transmembrane protein. The proline rich
region has been shown to have transcription activating properties (Mermod et al.,
1989). Thus, it is plausible that the bcI-3 gene could be a transcriptional activating
factor. The involvement of the bcI-3 gene in B-CLL was later analyzed in a large
series of patients by Raghoebier et al. (1991). Unexpectedly, none of the forty four
B-CLL studied had a rearrangement within 15 kb of the bcI-3 locus. However,
mutations in the bcl-3 gene undetectable in their assays may exist. Whether the
bcI-3 gene contributes to the oncogenesis of B-CLL awaits further investigation.

The t(8,12)(q24,q22) translocation links c-myc not with an lg enhancer but
rather with a locus termed BTG1 on chromosome 12 that presumably deregulates
myc (Rimokh et al., 1991). The breakpoint is located in the 3’ end of the myc
locus, and increased c-myc expression has been found. Sequences cloned from
the breakpoint recognize a 1.8 kb transcript in the CLL cells and in tissues of
lymphoid origin. In addition, this chromosome 12 coding sequence is conserved
in evolution and a transcript of similar size is present in murine tissues. However,
little is known about the function of this putative gene. Whether or not this gene
activation represents a general feature for the B-CLL is also unknown.

In addition to the possible involvement of the PRAD1, col-3, and c-myc
genes in this malignancy, bcI-2 was found rearranged in 3/34 B-CLL through a
variant translocation t(18;22). In this translocation, the bcI-2 gene is juxtaposed
to the lgtt on: genes in a head to head configuration (Adachi et al., 1990). Bel-2
was also found rearranged in 3/44 B-CLL by Raghoebier and coworkers (1991).
The role of bcI-2 in tumorigenesis will be discussed in the following section.

The relatively low frequency of any individual oncogene activation in B-CLL

U0:

Fol

WEi

33

may relate to the accuracy of clinical diagnosis, and the broad range of cell types
involved in this malignancy. For example, CLL may sometimes be difficult to
distinguish from non-Hodgkin’s lymphoma (Bennett et al., 1989; Deegan, 1989).
Systemic analyses on a large scale of samples may be required for a better
understanding of the molecular basis of this malignancy.
Follicular lymphoma

The t(14,18)(q32,q21) translocation which occurs in 85% of follicular lymphomas
was first described by Fukuhara et al. (1979). The interchromosomal junction has
been cloned from several follicular lymphomas (T sujimoto et al, 1985b; Cleary and
Sklar, 1985). The recombination region of chromosome 18, bcI-2, is rearranged
into the heavy chain enhancer region on chromosome 14 resulting in deregulation
of bcI-2 expression (Tsujimoto et al., 1984b, 19850; Cleary and Sklar, 1985;
Bakhshi et al., 1985). The normal bcI-2 gene is quiescent is resting B cells,
expressed in proliferating B cells, and downregulated in differentiated cells
(Graninger et al., 1987; Reed et al., 1987). Inappropriately high levels of bcI-2-
immunoglobulin chimeric RNA are present in t(14;18) follicular lymphoma
considering their mature B-cell stage (Seto et al., 1988). This indicates that the
translocated bcI-2 allele has escaped normal control mechanisms. Nucleotide
sequence analysis and biochemical studies from one group suggest that bcI-2 is
a GTP-binding protein located on the cytoplasmic surface of cell membranes
(Haldar et al., 1989), but a second group has localized it to the inner mitochondrial
membrane (Hockenbery et al., 1990). Bel-2 has been shown to prolong cell
survival by blocking programmed cell death (Nunez et al., 1990). A similar effect

may thus contribute to the formation of follicular lymphomas. Activated bcI-2 genes

34

proved capable of transforming or enhancing the survival of a cultured human B-
cell line (T sujimoto, 1989) or NIH3T3 cells (Reed et al., 1988). It is also capable
of inducing follicular hyperplasia in transgenic mice (McDonnell et al., 1989) which
progresses to a malignant diffuse large-cell lymphoma after a long latency
(McDonnell and Korsmeyer, 1991). The long latency, progression from polyclonal
to monoclonal disease, and histological conversion, are all suggestive of second
alterations. C-myc activation is suggested to be a candidate oncogene both for
its occurrence of rearrangement in the diffuse lymphoma, and its well known
involvement in B cell neoplasia. In fact, the cooperativity of the c-myc oncogene
and 001-2 In tumorigenesis has been shown by Vaux et al. (1988). The authors
found the bcI-2 gene can promote the proliferation of bone marrow cells of Eu-myc
transgenic mice, some of which become tumorigenic. Another oncogene, c-Ha-
ras, was found to complement bcI-2 in malignant transformation of rat embryo
fibroblasts (Reed et al., 1990).
Pre-B cell acute lymphocytic leukemia

Two translocations t(1,19)(q23,p13) and t(5,14)(q31,q32) have been found
in pre-B cell acute lymphocytic leukemias (pre-B ALL). The t(1,19)(q23,p13)
translocation was described as a common feature for pre-B ALLs (Carroll et al.
1984; Michael et al., 1984; Williams et al., 1984). A fusion protein, E2A-PBX,
results from this translocation which links the E2A gene on chromosome 19 to the
homeobox containing PBX gene on chromosome 1 (Kamps et al., 1990; Nourse
et al., 1990). The E2A gene encodes for two similar immunoglobulin enhancer
binding factors, each with a 5’ effector domain and a 3’ DNA binding domain

(Murre et al., 1989). The translocation switches the DNA binding domain of the

35
E2A transcription factor with that of PBX, thus placing those genes usually

regulated by PBX under the trans-activational control of E2A. Because PBX is not
normally transcribed in pre-B cells, the translocation results in ectopic expression
of the PBX DNA binding domain and therefore implicates the fusion protein in the
tumorigenesis of pre—B cells.

The other translocation, t(5,14)(q31,q32) was identified by Grimaldi and
Meeker (1989). At the breakpoint of this translocation, interleukin-3 (IL-3) on
chromosome 5 is positioned next to the heavy chain enhancer region on
chromosome 14 (Meeker et al., 1990). This has led to the hypothesis that the
overproduction of IL-3 may result in an autocrine loop that favors leukemogenesis

(Meeker et al., 1990).

3.2 Murine plasmacytoma

Murine plasmacytoma is induced by the application of a mineral oil such as
pristane into the intraperitoneal cavity of a mouse (with a restriction to Balb/c or
NZB mice) (Potter and Boyce, 1962; Anderson and Potter, 1969). Balb/c mice
develop plasmacytoma with a latency period of 6-12 months after the injection of
mineral oil. The long latency has suggested a multi-step nature for the formation
of this tumor. Similar to BL, two common characteristics have been associated
with almost all plasmacytomas.

First, the involvement of a paracrine stimulation is critical for the formation
of plasmacytomas. Prior to formation of a plasmacytoma, induction of chronic
granuloma tissues that consist primarily of macrophages and neutrophils is found.

Granuloma formation has been shown to play an important role both in the

36

development (Potter and MacCardle, 1964) and in the maintenance (Cancro and
Potter, 1976) of the primary plasmacytoma. A growth factor (25-kD) that is
expressed by a macrophage cell line was found to stimulate the growth of
plasmacytoma cells in vitro (Nordan and Potter, 1986). This growth factor is likely
to be interleukin-6 (IL-6) as judged by its size and biological activity. lL-6, a
pleiotropic cytokine, induces the terminal differentiation of B cells into antibody-
secreting cells, and also acts on a variety of other cell types (Kishimoto and
Hirano, 1988). Pristane treatment presumably results in the secretion of this
growth factor from the chronic granuloma tissue. The involvement of IL-6 in the
in vivo growth of plasmacytomas is further implied by the following finding.
Overexpression of lL—6 in transgenic mice, although not sufficient to cause the
development of mouse plasmacytomas, induces a massive plasmacytosis in
thymus, lymph node, spleen, and other tissues (Suematsu et al., 1989).
Transfection with the IL-6 gene into lL-6-dependent plasmacytoma cells has been
shown to increase their tumorigenicity (Vink et al., 1990). The involvement of
growth factor stimulation is similar to that seen in Burkitt’s lymphoma by EBV with
the exception of its being paracrine instead of autocrine and its being induced

chemically rather than through viral infection.

Chromosomal translocation has also been found in almost all
plasmacytomas (Yosida et al., 1970; and references listed below).
Two types of translocations were first described by Shepard and coworkers
(1974a, 1974b, 1978) in transplanted plasmacytoma lines. One consists of a

reciprocal translocation between chromosomes 12 and 15, and was later termed

O?"
vay

0556

THE

195:

37

the "typical" translocation by Ohno et al. (1979) because of its higher occurrence.
The other is a reciprocal translocation between chromosomes 6 and 15, and was
named the "variant" translocation by the same group because of its relatively lower
occurrence. Both translocations are associated with lg gene loci; the "typical“ one
usually involves the alpha switch region of the heavy chain gene (Kirsch et al.,
1981; Adams et al., 1982; Harris et al., 1982), whereas the "variant“ one involves
the kappa light chain gene (van Ness et al., 1983; Webb et al., 1983; Cory et al.,
1985). Analyses of the breakpoint on chromosome 15 in these translocations have
revealed the involvement of the c-myc gene (Shen-Ong et al., 1982; Adams et al.,
1983; Marcu et al., 1983). The consequence of the translocation has been
proposed to be similar to that of Burkitt’s lymphoma, i.e., the deregulation of c-myc
gene expression (Adams et al., 1983; Bernard et al., 1983).

Other secondary events have been observed in some but not all of
plasmacytomas. In a few plasmacytomas, activation of the c-mos gene by
insertion of an intracisternal A particle element was found in addition to c-myc
activation (Rechavi et al., 1982; Canaani et al., 1983; Gattoni-Celli et al., 1983).
The v-mos protein exhibit activities of both serine/threonine autophosphorylation
(Maxwell and Arlinghaus, 1985) and ATP-dependent DNA binding (Seth et al.,
1987) in vitro. Two different biological activities have been associated with the mos
protein: inducing monocyte differentiation into macrophages (Kurata et al., 1989)
and transforming both NIH3T3 cells and normal kidney cells into malignant cells
(Kurata et al., 1987). The biochemical basis of these two different potentialities is
not clear. Its cytoplasmic location and kinase activity suggest a role in a signaling

pathway. Alteration of a signaling pathway may contribute to the progression of

malignancy.

Perlmutter and associates (1984) found a t(6,10) recombination besides
t(12,15) in the NS-1 murlne plasmacytoma line and another plasmacytoma.
Transcription from this locus was detected at a high level, and homologous
sequences could be detected in mouse, rabbit, and human. Therefore, it is
possible that t(6,10) may encode a gene that plays a role in the course of B cell
tumor progression.

Lastly, chromosome 11 trisomy is another frequent secondary alteration in
murlne plasmacytomas (Ohno et al., 1984). Several oncogenes (Rel, ErbA, ErbB,
p53), and cytokine genes (G-CSF, GM-CSF, lL—3, lL-4, IL-5) have been mapped to
chromosome 11 (Buchberg et al., 1989; Wilson et al., 1990). Among these genes,
lL-4 is a factor involved mainly in the activation of resting B cells, and IL-5 is a
factor for the growth and maturation of activated B cells (Kishimoto, 1985). An
increased dosage of these two genes may contribute to the formation of
plasmacytoma, and perhaps act cooperatively with IL—6.

The involvement of multiple oncogenes in the development of murine
plasmacytoma is not only suggested by the above findings, but also by the rapid
induction of murlne plasmacytoma with viral oncogenes. Infection of mice with
Abelson murine leukemia virus after application of pristane greatly reduces the
latency of tumor induction (Ohno et al., 1984). These tumors not only express v-
abl, but show the same translocations activating c-myc that are observed in tumors
induced by pristane alone. Infection with a recombinant retrovirus expressing an
avian v-myc after pristane application also accelerated plasmacytoma formation

(Potter et al., 1986, 1987). In this case, expression of v-myc seems to replace the

39

requirement for c-myc translocation. As with Burkitt’s lymphoma, no single gene

is sufficient for plasmacytoma formation.

3.3 Avian lymphoid leukosis

Lymphoid leukosis, a B-cell lymphoma induced by avian leukosis viruses
(ALVs), progresses through a series of clinically distinct stages (Cooper et al.,
1968). Three stages have been identified in ALV-induced lymphoid leukosis
(Cooper et al., 1968; Baba and Humphries, 1985). The earliest detectable lesion
is the transformed follicle, a hyperplastic bursal follicle in which the normal follicular
architecture is obscured by an abnormal proliferation of lymphoblasts. The
lymphoblasts are confined to their follicle of origin during this early stage. Up to
100 transformed follicles may be present in a single bursa, but most of these
regress during bursal involution. Then, one or sometimes more than one of the
transformed follicles gives rise to a bursal nodule, which is readily identifiable at
necropsy. Ultimately, tumor cells from the bursal nodule disseminate to other
tissues, resulting in widespread metastasis in the liver, kidney, and spleen. The
instance of regression after the first stage suggests the requirement of additional
events for further progression.

Molecular analyses of ALV-induced lymphomas show that the majority of
them contain proviral integrations within or near the c-myc gene resulting in
deregulated expression of c-myc (Hayward et al., 1981; Neel et al., 1981; Payne
et al., 1982). The insertional activation of c-myc appears to be an early event in
lymphomagenesis according to studies of Neiman et al. (1985), but additional

proto-oncogene activations are required for progression to the late stages of the

4O

disease. Neiman and coworkers have found that HB-1, a v-myc-containing virus,
induces transformed follicles when virus-infected cells are used to reconstitute a
chemically ablated bursa (Neiman et al., 1985; Thompson et al., 1987). Like ALV-
induced transformed follicles, only a small portion of these progress to become
lymphomas. These results again indicate that additional genetic events beyond c-
myc induction may be required for late stages of tumor progression. Clurman and
Hayward (1988) have used a double-infection protocol to facilitate the occurrence
of multiple insertional activation in order to identify those genetic events that may
function in cooperation with c-myc to induce late stages of progression in ALV-
induced lymphomas. In their study, c-myc rearrangement was found in 70% of
lymphomas (both primary and metastatic) identified 3 to 4 months after hatching.
Additionally, a frequent viral integration locus, c-bic, was found in 14% of the
primary lymphomas, and 50% of the metastatic lymphomas. These results confirm
that the c-myc gene is frequently a target for insertional activation in ALV-induced
lymphomas. They also indicate that c-bic may act synergistically with c-myc during
lymphomagenesis. The increased frequency of viral integration at c-bic in
metastatic tumors suggests that c-bic is involved in late stages of tumor
progression.

In summary, studies on oncogene alteration in B-cell lymphomas/leukemias
of three species have revealed the mutli-factorial nature of tumor formation. These
gene alterations caused by either insertional mutagenesis or chromosomal
translocations show different degrees of association with B-cell
lymphomas/leukemias. The most commonly seen gene altered, c—myc, shows a

more broad range 'of association with tumors of different cell types, whereas other

41

genes altered, such as bcI-2 and IL-6, show a more restricted linkage to lymphoid
tumors. The bcI-2 gene alteration has been implicated in lymphoid tumors of
multiple developmental stages such as follicular lymphoma, diffuse small and large
cell lymphoma, and chronic lymphocytic leukemia. The IL—6 gene alteration,
however, couples specifically to plasmacytoma. The significance of these
differences is not yet known. Maybe only mature cells have lL-6 receptors. A
systemic analysis with various probes on a greater number of tumor samples will
be required to obtain a clear sense of the tissue and developmental specificity of
these oncogene activations.

These affected genes found in B cell neoplasias encode a variety of proteins
including growth factors (e.g., IL-6), membrane associated proteins (e.g., N-ras,
bcI-2 ), signal transduction proteins (e.g., c—mos), and many cell cycle or
differentiation-related regulatory proteins (e.g., c-myc, bcI-I, bcI-3 etc). Since
these proteins are presumably involved in the regulation of cell growth and
differentiation, deregulation of these genes should be important in the oncogenesis
of B-cell lymphomas/leukemias. Although their roles in these malignant tumors
seem obvious, a direct test of their oncogenic potential should be provided to
ultimately distinguish their causative effects in tumorigenesis from their simply being
the result of malignancy. Furthermore, analyses of end point specimens, such as
malignant tumors, do not provide direct information on the effects of differentiation
upon oncogene activation or the nature of genes that may be activated in
lymphomas/leukemias in addition to those in an obvious translocation. To address
these questions, a few model systems, including the use of in vitro culture

systems, have been used. Results of these studies are summarized in the

3.4 I

rm:
Us

vim

TECE

“Mp

V-at

42

following section.

3.4 In vitro transformation of murlne B cells

A number of oncogenes or growth factors have been found to induce
transformation of B lymphoid cells in vitro. These may include (i) oncogenes of the
ras oncogene family which code for protein with GDP-binding activity (for review
see Barbacid, 1987), (ii) oncogenes of the src oncogene family which code for
tyrosine-specific protein kinases (for review see Hunter and Cooper,1985), (iii) the
v-fms oncogene whose cellular counterpart encodes a colony stimulating factor-1
receptor, and (iv) the interleukin-7 (IL-7) gene which codes for a pre-B cell growth
factor gene (Namen et al., 1988).

V-abI is the first viral oncogene that is capable of transforming murlne pre-B
cells in vitro (Sklar et al., 1974; Rosenberg et al., 1975; Rosenberg and Baltimore,
1976). Infection of cell cultures derived from murlne fetal liver, bone marrow or
spleen, but not thymus with Abelson murine leukemia virus (A-MuLV) carrying the
v-abl oncogene give rise to permanently growing, neoplastic cell lines. These in
vitro isolated cell lines are similar morphologically to A-MuLV-induced tumor cells
display a characteristic of B cells, and have been found to be preferentially pre-B
cells (Siden et al., 1979). Another oncogene of the src gene family, v-fes, also
exhibits oncogenic potential on murine pre-B cells in vitro (Pierce and Aaronson,
1983). In this study, bone marrow cells were infected with a v-fes containing
retrovirus (Snyder-Theilen feline sarcoma virus, ST-FeLV) and selected for their
ability to grow on soft agar, a characteristic of transformed cells. In both cases,

A-MuLV or ST-FeLV infection alone appears to be sufficient to induce aggressive

If

I‘ll".

re.
‘U

S"
“U
U

tie

43

tumors in syngeneic mice with a 1-3 weeks latency. However, results from
Whitlock and Witte (1981) using a fresh bone marrow cell suspension as target for
the A-MuLV infection and a feeder layer to support the growth of the initial infected
population, show that the A-MuLV infected cells are initially poorly oncogenic and
that they become progressively tumorigenic and growth independent only after
progression on normal adherent bone marrow feeder layers. This latter result
suggests that a minor subpopulatlon capable of unrestricted growth is present at
the initiation of the culture and gradually becomes the predominant population
through a slight growth advantage. Alternatively, the A-MuLV-infected cells may
undergo secondary changes which further alter their growth properties. Using
clonal cell lines isolated from the A-MuLV infected cells, Whitlock et al. (1983b)
have shown that the progression of these clonal cell lines to a more malignant
growth phenotype occurs with no changes in viral related properties. The
expression of cellular genes appears to alter the growth properties of lymphoid
cells after their initial transformation by A-MuLV. The discrepancy between
Rosenberg’s and Whitlock’s results may residue in the properties of cells used for
the initial infection, and the stringency of the growth parameters used to determine
transformation. Since the cells used by Rosenberg’s group have been cultured in
vitro for a period of time, cells with a slight growth advantage may be preselected
for viral infection. The lack of a feeder layer again selects for cells with a more
transformed phenotype. A similar explanation may be applied to the results from
Pierce and Aaronson (1983), in which the soft agar assay may select for events
additional to the ST-FeLV infection. In conclusion, these results suggest that v-abl

or v-fes may, by itself, be sufficient to initiate transformation of B cells but may

44

require additional events, such as the activation of cellular oncogenes, for
expression of the fully transformed state.

The involvement of the v-Ha-ras and v-bas oncogenes in the in vitro
transformation of lymphoid cells was demonstrated by Pierce and Aaronson (1982;
1984) using a similar approach as that described for the v-fes oncogene (see
above). The Harvey murine sarcoma virus carrying a v-Ha-ras oncogene, and the
Balb-murine sarcoma virus carrying a v-bas oncogene were used instead. The
resulting transformed cell lines were as tumorigenic as those infected by A-MuLV.
Cells transformed by v-Ha-ras or v-bas displayed characteristics of immature
lymphoid cells: high terminal deoxynucleotidyl transferase (T dt) activity, the
presence of F0 receptor, and mercaptoethanol dependence for growth. The lack
of u chain expression, and of Thy-1 antigen on these transformed cells suggested
that these cells might be at an earlier stage of differentiation than pre-B or pre-T
lymphoid cells. Holmes et al. (1986) further confirmed the role of oncogenes in the
ras gene family in the in vitro transformation of B cells. They demonstrated that
transformed cell lines obtained from cells infected with viruses containing v-Ha-ras,
v-bas, and v-K-ras display characteristics of pro-B and pre-B cells. Their finding

suggest that a wide range of oncogene can induce B cell transformation in vitro.

The difference in the target cells for the ras oncogene family in the results
of Holmes et al. and, Pierce and Aaronson may be a reflection of a continuous
differentiation, which may occur under the influence of the microenvironment of a
particular experimental setup, of a lymphoid progenitor (or a hematopoietic

progenitor in a broader aspect). This hypothesis is supported by the isolation of

Will
a i1

01

DIE
A3;
by

*0
O.

45

macrophage-like cell lines from pre-B cell lines containing the v-Ha-ras oncogene
obtained in the same study (Holmes et al., 1986). This lineage-switch
phenomenon is also found in pre-B cells transformed by the v-fms oncogene
(Borzillo et al., 1990; see below), and the v-raf oncogene (Klinken et al., 1988), as
well as in our study (Appendix A).

Another disparity is that viruses carrying v-Ha-ras, v—bas, and v-Ki-ras can also
induce sarcomas and erythroleukemias in susceptible animals (Peters et al., 1974;
Harvey, 1964; Scher et al., 1975; Hankins and Scolnick, 1981) and transform
myeloid cells in vitro (Pierce and Aaronson, 1985), whereas A-MuLV seems to have
a restricted target cell for transformation. This tissue-preference of tumorigenesis
of A-MuLV appears to be directly associated with the v-abl oncogene from the
studies of transgenic mice in which tumors of a B lymphoid origin are found
predominately in transgenic mice carrying the v-abl oncogene (for review, see
Adams and Cory, 1991). These results suggest that the tyrosine kinase encoded
by the v-abl oncogene may belong to a pre-B cell specific signal transduction
pathway, whereas the G-protein activity encoded by the ras gene family may
represent a common signal pathway shared by several cell types.

Borzillo and Sherr (1989) have shown that the v-fms gene is capable of
inducing transformation of pre-B cells. As with v-abl transformants, only the late-
passage cultures give rise to factor-independent variants that proliferate in the
absence of feeder layers, and become tumorigenic in syngeneic mice. The v-fms
oncogene encodes an analog of CSF-1R that retains a functional ligand-binding
domain but acts constitutively as a tyrosine kinase because of mutations. The

mechanism of this mononuclear-macrophage—lineage related receptor functioning

46

in pre-B cells is not well known. The fact that some of the v-fms transformed pre-B
cell line may undergo a lineage switch to macrophages when transferred from
RPMI1640 to Iscove modified Dulbecco medium suggests that v-fms or its cellular
analogue may be a common intermediate in the signal transduction pathways of
both the lymphoid and myeloid differentiation (Borzillo et al., 1990).

In addition to the above oncogene, a pre-B cell growth factor, IL-7 also
shows transforming potential on pre-B cells (Young et al., 1991; Overell et al.,
1991). However, the low cloning frequency of IL-7 transformants and the long
latency for tumorigenesis suggest that additional events are required in generating
a fully transformed phenotype.

In summary, a variety of oncogenes, growth factors or receptors have been
implicated in the in vitro transformation of B cells as in the in vivo tumorigenesis
of B cell tumors from different species. Consistent with the multi-step
tumorigenesis dogma, none of them are tumorigenic by themselves, and additional
events must exist. These results validate the use of the in vitro model to examine
the events that underlie tumor progression.

In my thesis, oncogene cooperation was used as a model to define the
multi-factorial molecular events in the tumor progression of murine pre-B cells. The
advantage of using oncogene cooperativity as a model to dissect the events
contributing to tumor progression is the enhanced likelihood of identifying genes
that by themselves are not capable of demonstrable transformation in growth factor
independence assays, soft agar assays, and tumor challenges. However, genes
with a subtle or negligible phenotype in the above assays may be able to transform

8 cells in cooperation with other oncogenes in a synergistic manner, for example

47

with v-myc or v-Ha—ras, and thus are potentially involved in the course of tumor
progression. In setting up a system to examine cooperative transformation of B
lymphoid cells, v-myc was the first candidate because c-myc activation was found
in many of the B cell malignancies described above. Unfortunately, the v-myc
gene was not able by itself to transforme primary bone marrow cells in the
Whitlock and Witte culture system and may be lethal (data not shown, Stevenson
and Volsky, 1986). We therefore chose v-Ha-ras, partly because v-Ha-ras has
been found to transform murlne B cells into an intermediate transformed
phenotype (Schwartz et al., 1986a; Holmes et al., 1986), and partly because ras
gene aberration was found in some 8 cell neoplasias (see previous section). The

two model systems used in my studies are described in the “Introduction".

48
Literature cited:

Adachi M., A. Tefferi, P.R. Greipp, T.J. Kipps, and Y. Tsujimoto. (1990). Preferential linkage of bcI-2
to immunoglobulin light chain gene in chronic lymphocytic leukemia. J. Exp. Med. 171, 559-564.

Adams J.M., and 8. Cory. (1991). Transgenic models of tumor development. Science 254,
1161-1167.

Adams J.M., E. Webb. J. Mitchell, 0. Bernard, and S. Cory. (1982). Transcriptionally active DNA
region that rearranges frequently in murlne lymphoid tumors. Proc. Natl. Acad. Sci. USA 79.
6966-6970.

Adams J.M., S. Gerondakls, E. Webb, LM. Corcoran, and S. Cory. (1983). Cellular myc gene is
altered by chromosome translocation to an immunoglobulin locus in murlne plasmacytomas and
Is rearranged similarly in human Burkitt’s lymphomas. Proc. Natl. Acad. Sci. USA 80. 1982-1986.

Adams J.M., A.W. Harris, C.A. Pinkert, LM. Corcoran, W.S. Alexander, S. Cory, R.D. Palmiter, and
BL Brlnster. (1985). The c-myc oncogene driven by lmmunoglobulln enhancers Induce lymphoid
malignancy in transgenic mice. Nature 318, 533-538.

Arnold A., H.G. Kim, R.D. Gaz, P.L. Eddy, Y. Fukushima, MG. Byer, TB. Shows, and HM.
Kronenberg. (1989). Molecular cloning and chromosomal mapping of DNA rearranged with the
parathyroid hormone gene in a parathyroid adenoma. J. Clin. Invest. 83, 2034-2040.

Austin J.. and J. Kimble. (1987). ng-t is required in the germ line for regulation of the decision
between mitosis and meiosis in C. elegans. Cell 51, 589-599.

Barbacld M. (1987). Fias genes. Ann. Rev. Biochem. 56, 779-827.

Bargmann C.I., M.-C. Hung, and RA Weinberg. (1986). The neu oncogene encoded an epldennal
growth factor receptor-related protein. Nature 319, 226-230.

Bates G.P., B.J. Wainwright, R. Williamson, and S.D.M. Brown. (1986). Microdissection and
microcloning from the short arm of human chromosome 2. Mol. Cell. Biol. 6, 3826-3830.

Baumbach W.R., EM. Colston, and MD. Cole. (1988). Integration of the Balb/c ecotropic provlms
Into the colony-stimulating factor-1 growth factor locus in a myc retrovirus induced murlne monocyte
tumor. J. Virol. 62, 3151-3155.

Bennett J.M., D. Catovsky, M.T. Daniel, G. Flandrin. D.A.G. Galton, H.R. Gralnick and C. Sulatn.
(1989). Proposal for the classification of chronic (mature) B and T lymphoid leukemias and
lymphomas. J. Clin. Pathol. 42, 567.

Bernard 0., S. Cory, S. Gerondakls, E. Webb, and J. M. Adams. (1983). Sequence of the murlne and
human cellular myc oncogene and two modes of myc transcription resulting from chromosome
translocation in B lymphoid tumours. EMBO J. 2, 2375-2383.

Birchmeier C., D. Broek, T. Toda, S. Powers, T. Kataoka, and M. Wigler. (1985). Conservation and
divergence of RAS protein function during evolution. Cold Spring Harb. Symp. Quant. Biol. 50,
721 -725.

Bonner T.l., C. O’Connell, and M. Cohen. (1982). Cloned endogenous retrovlral sequences from
human DNA. Proc. Natl. Acad. USA 79. 4709-4713.

49

Borzlllo G.V., and OJ. Sherr. (1989). Early pre-B-cell transformation Induced by the v-fms oncogene
In long-term mouse bone marrow cultures. Mol. Cell. Biol. 9, 3973-3981.

Borzillo G.V., R.A. Ashman, and OJ. Sherr. (1990). Macrophage lineage switching of murlne early
pre-B lymphoid cells expressing transduced fms genes. Mol. Cell. Biol. 10, 2703-2714.

Breuer M.C., H.T. Cuypers, and A. Bems. (1989a). Evidence for the Involvement of pIm-2 a new
common proviral insertion site in progression of lymphomas. EMBO J. 8, 743-747.

Buchberg A.M., E. Brownell, S. Nagata, N.A. Jenkins, and MG. Copeland. (1989). A comprehensive
genetic map of murlne chromosome 11 reveals extensive lineage conservation between mouse and
human. Genetics 122, 153-161 .

Burkitt DP. (1958). A sarcoma involving the jaws in African children. Br. J. Surg. 46, 218-223.

Canaani E., O. Dreazen, A. Klar, Rechave, D. Ram, J.B. Cohen, and 0. leol. (1983). Activation of
the c-mos oncogene in a mouse plasmacytoma by insertion of an endogenous intracisternal
A-particle genome. Proc. Natl. Acad. Sci. USA 80, 7118-7122.

Cancro M. , and M. Potter. (1976). The requirement of an adherent substratum for the growth of
developing plasmacytoma cells in vivo. J. Exp. Med. 144, 1554-1566.

Carroll A.J., W.M. Crist, R.T. Parmley, M. Roper, M.D. Cooper, and WM. Finley. (1984). Pre-B cell
leukemia associated with chromosome translocation 1:19. Blood 63, 721-724.

Chaudhuri J.P., W. Vogel, I. Volculescu, and U. Wolf. (1971). A simplified method of demonstrating
Glemsa band pattern in human chromosomes. Human Genetlk 14, 83.

Chen S.J., N.J. Holbrook, KF. Mitchell, C.A. Vallone, J.S. Greengard, G.R. Crabtree, and Y. L].
(1985). A viral long terminal repeat In the Interleukin-2 gene of a cell line that constitutively produces
interleukin-2. Proc. Natl. Acad. Sci. USA 82, 7284-7288.

Clurman BE. and W.S. Hayward. (1989). Multiple proto-oncogene activation in avian leukosis
virus-induced lymphomas: evidence for stage-specific events. Mol. Cell. Biol. 9, 2657-2664.

Cole J.M., C. Righi. C. Dissous, G. Gegonne, and D. Stehelin. (1983). Molecular cloning of the avian
acute transforming retrovirus MH2 reveals a novel cell-derived sequence(v-mll) In addition to the
myc oncogene. EMBO J. 2, 2189-2194.

Collins 8., and M. Groudine. (1982). Amplificatlon of endogenous myc-related DNA sequences in
a human myeloid leukemia cell line. Nature 298, 679-681.

Cooper M.D., LE. Payne, P.B. Dent, B.P. Burmester, and RA. Good. (1968). Pathogenesis of avian
leukosis. J. Natl. Cancer Inst. 41, 374-389.

Cooper 08., M. Park, D.G. Blair, M.A. Tainsky, K. Huebner, CM. Croce, and G.F. Vonde Woude.
(1984a). Molecular cloning of a new transforrnlng gene from a chemically transformed human cell
line. Nature 311, 29-33.

Croce CM, and RC. Nowell. (1986). Molecular genetics of human B cell neoplasia Adv. Immunol.
38, 245.

Croce C.M., M. Shander, J. Martinis, L Clcurel, G.G. D’Ancona T.W. Dolby. and H. Koprowskl.
(1979). chromosomal location of the genes for human immunoglobulin heavy chains. Proc. Natl.
Acad. Sci. USA 76, 3416-3419.

50

Cuypers H.T., G. Selten, W. Quint, M. Zijlstra, ER. Maandag, W. Boelens, P. van Wezenbeek, C.
Melief, and A Bems. (1984). Murine leukemia virus-induced T cell lymphomagenesis: integration of
proviruses in a distinct chromosomal region. Cell 37, 141-150.

Delll Bovi P., A.M. Curatola, F.G. Kern, A. Greco, M. Ittmann, and C. Basilica. (1987). An oncogene
ls isolated by transfection of Kaposi’s sarcoma DNA encoded a growth factor that is a member of
the FGF family. Cell 50, 729-737.

Der C.J., T.G. Krontlris, and GM. Copper. (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.

de The’ G., A Geser, N.E. Day. P.M. Tukel, E.H. Williams, D. Beri, P.G. Smith, AG. Dean, G,W.
Bornkamm, P. Feorlno, and W. Henle. (1978 ). Epidemiological evidence for a causal relationship
between Epstein-Barr Virus and Burkitt's lymphoma: result of Uganda prospective study. Nature 274,
756-761.

Diamond L. ,T.G. O’Brien, and WM. Baird. (1980). Tumor promoters and the mechanism of tumor
promotion. Adv. Can. Res. 32, 1-74.

Dickson C., R. Smith, S. Brookers. and G. Peters. (1984). Tumorigenesis by mouse mammary tumor
virus: proviral activation of a cellular gene in the common integration region int-2. Cell 37, 529-536.

DiPaolo J.A. (1983). Relative difficulties In transforming human and animal cells In vino. J. Natl.
Cancer Inst. 70, 3-7.

Duyk G.M., S. Kim, R.D. Myers, and DR. Cox. (1990). Exon trapping: a genetic screen to Identify
candidate transcribed sequences in cloned mammalian genomic DNA Proc. Natl. Acad. Sci. USA
87, 8995-8999.

Edstrom J.-E., R. Kaiser, and D. Rohme. (1986). Microcloning of mammalian metaphase
chromosomes. Methods Enzymol. 151, 503-516.

Erikson J., J. Martinis, and CM. Croce. (1981). Assignment of the genes for human lambda
lmmunoglobulln chains to chromosome 22. Nature 294, 173-175.

Eva. A.,, and SA. Aaronson. (1985). Isolation of a new human oncogene from a diffuse B-cell
lymphoma. Nature 316, 273-275.

Fasano D. , D. Blmbaum, L Edlund, J. Fogh, and M. Wigler. (1984). New human transforrnlng genes
detected by a tumorigenecity assay. Mol. Cell. Biol. 4, 1695-1705.

Fearon ER, and B. Vogelstein. (1990). A genetic model for colorectal tumorigenesis. Cell 61,
759-769.

Fisher J.C., and J.H. Holloman. (1951 ). A hypothesis for the origin of cancer loci. Cancer 4, 916-918.
Foulds L (1975). Neoplastic development. Vol. II. (New York: Academic press).

Fung Y.-KT., W.L Louis, LB. Crittenden. and H.J. Kung. (1983). Activation of the cellular oncogene
c-erbB by LTR Insertion: Molecular basis of erythroblastosls by Avian Leukosis Virus. Cell 33,
357-368.

Gabrielson E.W., and CC. harrls. (1985). Use of cultured human tissues and cells In carcinogenesis
research. J. Cancer Res. Clin. Oncol. 110, 1-10.

51

Gaidano G.. P. Ballerini, J.Z. Gong, G. lnghlraml, A Neri, E.W. Newcomb, I.T. Magrath, D.M.
Knowles, and R. Dalia-Favera. (1991). p53 mutations In human lymphoid malignancies: Association
with Burkitt lymphoma and chronic lymphocytic leukemia. Proc. Natl. Acad. Sci. USA 88, 5413-5417.

Garcia M., R. Wellinger, A. Vessaz, and H. Diggelmann. (1986). A new site of Integration for mouse
mammary tumor virus proviral DNA common to Balb/cf(CSH) mammary and kidney
adenocarcinomas. EMBO J. 5, 127-134.

Geroge D.L, B. Glick, S. Trusko, and N. Freeman. (1986). Enhanced c-Kl-ras expression associated
with Friend Virus Integration is a bone-marrow derived mouse line. Proc. Natl. Acad. Sci. USA 83,
1651-1655.

Geser A., G. Lenoir, M. Anvret, G.W. Bornkamm, G. Klein, E.H. willlams, D.H. Wright, and G. de The'
(1983). Epstein-Barr Virus markers in a series of Burkitt's lymphoma from West Nile District of
Uganda. European J. Cancer. Clin. Oncol. 19, 1394-1404.

Gordon J., S.C. Ley, M.D. Melamid, LS. English, and NC. Hughes Jones. (1984). lmmortallzed B
lymphocytes produces B cell growth factor. Nature 310, 145-147.

Graham FL, and A.J. van der Eb. (1973). Transformation of rat cells by DNA of human adenovirus
5. Virology 54, 536-539.

Graninger W., M. Seto, B. Boutain, P. Goldman, and SJ. Korsmeyer. (1987). Expression of bcI-2 and
bcI-2-Ig fusion transcripts in normal neoplastic cells. J. Clin. invest. 80, 1512-1515.

Grazianl D, D. Ron, A. Eva, and K. Srivastavas. (1989). The human dbl-protooncogene product is
a cytoplasmic phosphoprotein which Is associated with the cytoskeletal matrix Oncogene 4,
823-829.

Grimaldi J.C., and TC. Meeker. (1989). The t(5;14) chromosomal translocation In a case of acute
lymphocytic leukemia joins the interleukin-3 gene to the immunoglobulin heavy chain gene. Blood
73, 2081 -2085.

Haldar S., C. Beatty, Y. Tsujimoto. and CM. Croce. (1989). The bcI-2 gene encodes a novel G
protein. Nature 342, 195-198.

Halevy 0., D. Michalovitz, and M. Oren. (1990). Different tumor-derived p53 mutants exhle distinct
biological activities. Science 250, 113-116.

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

Hayward W., 8.6. Neel, and S. Astrin. (1981). Activation of a cellular one gene by promoter Insertion
in ALV-Induced lymphoid leukosis. Nature 290, 475-480.

Hicks G.G.. and M. Mowat. (1988). Integration of Friend Murine Leukemia Virus Into both alleles of
the p53 oncogene In an erythroleukemia cell line. J. Virol. 62, 4762-4755.

Hockenbery D., G. Nunez, C. Milliman, R.D. Schreiber, and J. Korsmeyer. (1990). Bcl-2 Is an Inner
mitochondrial membrane protein that blocks programmed cell death. Nature 348, 334-336.

Holmes KL, J.H. Pierce, W.F. Davison, and HG Morse, III. (1986). Murine hematopoietic cells with
pre-B or pre-B/myelold characteristics are generated by In vitro transformation with retrovirus
containing fes, ras, abI, and src oncogenes. J. Exp. Med. 160, 443-459.

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

swim“

New. lame

Id be m

52
Hunter T., and J.A. Cooper. (1985). Protein tyrosine kinases. Ann. Rev. Biochem.
Iman, OS. and CC. Harris. (1991). Crit. Rev. Oncogen. 2, 161.

lhle J.N., 8.8. White, B. Sisson, D. Parker, D.G. Blair, A. Schultz, C. kozak, RD. Cunsford. D. Askew,
Y. Weinstein, and R.J. lsofort. (1989). Activation of c-Ha-ras protooncogenes by retrovirus Insertion
and chromosomal rearrangement In a Moloney leukemia virus-Induced T cell leukemia. J. Virol. 63,
2959-2966.

Kamps M.P., C. Murre, X.-H. Sun, and D. Baltimore. (1990). A new homeobox gene contributes the
DNA binding domain of the t(1;19) translocation protein in pre-B ALL Cell 60, 547-555.

Kan W.C., C.S. Flordellis, C.F. Garon, P. Duesberg, and TS. Papas. (1983). Avian carcinoma virus
MH2 contains a transformation-specific sequence, mht, and share the myc sequence with M029,
CMII, and OKIO viruses. Proc. Natl. Acad. Sci. USA 80, 6656-6670.

Kirsch I.R., J.V. Ravetch, S.-P. Kwan, E.E. Max, R.L Ney, and P. Leder. (1981). Mdtlple
lmmunoglobulln switch region homologies outside the heavy chain constant region locus. Nature
293. 585-587.

Klshimoto T. (1985). Factors affecting B-ceII growth and differentiation. Ann. Rev. Immunol. 3,
133-157.

Klshimoto T., and T. Hirano. (1988). Molecular regulation of B Iyrnphocyte Response. Ann. Rev.
Immunol. 6, 485-512.

Kitayama H., Y. Sugimoto, T. Matsuzakl, Y. Ikawa. and M. Noda. (1989). A ras related gene with
transformation suppressor activity. Cell 56, 77-84.

Klein 6.. and E. Klein. (1985). Evolution of tumors and the impact of molecular biology. Nature 315,
190-195.

Klein E., G. Klein, J.S. Nadkarni, J.J. Nadkaml, H. ngzell, and P. Clifford. (1967). Surface lgM
specificity on cells derived from a Burkitt’s lymphoma. Lancet 2, 1068-1070.

Klinken S.P., W.S. Alexander, J.M. Adams. (1988). Hemopoletlc lineage switch: v-raf oncogene
converts Eu-myc transgenic B cells into macrophages. Cell 53, 857-867.

Knudsen AG., Jr., and AT. Meadows . (1980). Regression of neuroblastoma IV-S: a genetic
hypothesis. N. Engl. J. Med. 302, 1254-1256.

Kurata S., N. Kurata, and Y. Ikawa. (1987). Production of recombinant rat vlmses as a method of
oncogene Isolation in coculture medium. Cancer Res. 47, 5908-5912.

Kurata N., H. Aklyama, T. Tanlyama, and T. Marunouchi. (1989). Dose-dependent regulation of
macrophage differentiation by mos mRNA In a human monocytlc cell line. EMBO J. 8, 457-463.

Lamb P., and L Crawford. (1986). Characterization of the human p53 gene. Mol. Cell. Biol. 6,
1379-1385.

Lammie G.A., V. Fantl, R. Smith, E. Schuuring, S. Brookes, R. Mlchalides, C. Dickson, A. Arnold, and
G. Peters. (1991 ). D1 13287, a putative oncogene on chromosome 11q13, is amplified and expressed
in squamous cell and mammary carcinomas and linked to BCL-I. Oncogene 6, 439-444.

53

Lenoir G.M., T. Philip, and R. Sohier. (1984a). Burkitt-type lymphomas-EBV association and
cytogenetic markers In cases from various geographic locations. In Pathogenesis of Leukemlas and
lymphomas: environmental influences. I.J. Magrath, G.T. O’Conor, and B. Ramot. eds. (New York:
Raven Press). pp. 283-295.

Little C.D., D.N. Carney, and AF. Gazdar. (1983). Amplificatlon and expression of the c-myc
oncogene In human lung cancer cell lines. Nature 306, 194-196.

Lombardi L, E.W. Newcomb, and R. Dalia-Favera. (1987). Pathogenesis of Burkitt lymphoma:
expression of an activated c-myc oncogene causes the tumorigenic conversion of EBV-Infected
human B lymphoblasts. Cell 49, 161-170.

Lfidecke H.-J., G. Senger, U. Claussen, and B. Horsthemke. (1989). Cloning defined regions of the
human genome by microdissection of banded chromosomes and enzymatic amplification. Nature
338. 348-350.

Malcolm 8., P. Barton, C. Murphy, M.A. Ferguson-Smith, D.L Bentley, and TH. Rabbltts . (1982).
Localization of human immunoglobulin K chain variable region to the short arm of chromosome 2
by in situ hybridization. Proc. Natl. Acad. Sci. USA 79, 4957-4961.

Marth J.D., R. Reet, E.G. Krebs, and RM. Pertmutter. (1985). A lymphocyte-specific protein tyrosine
kinase gene is rearranged and overexpressed in murine T cell lymphoma LSTRA. Cell 43, 393-404.

Martin-Zanaca D., S.H. Hughes, and M. Barbacid. (1986). A human oncogene formed by the fusion
of truncated tropomyosin and protein tyrosine kinase sequences. Nature 319, 743-748.

Maxwell SA, and RB. Arlinghaus. (1985). Serine kinase activity associated with Moloney murlne
sarcoma virus-124-encoded p37 mos. Virology 143, 321-333.

McDonnell T.J., and SJ. Stanley. (1991). Progression from lymphoid hyperplasia to high-grade
malignant lymphoma in mice transgenic for the t(14;18). Nature 349, 254-256.

McDonnell T.J., N. Deane, F.M. Platt, G. Nunez, U. Jaeger, J.P. Mckeam, and SJ. Korsmeyer.
(1989). bcI-2-Immunoglobulin transgenic mice demonstrate extended 8 cell survival and follicular
Iymphoproliferation. Cell 57, 79-88.

McKeithan T.W., J.D. Rowley, T.B. Shows, and MO. Diaz. (1987). Cloning of the chromosome
translocation breakpoint junction of the t(14;19) In chronic lymphocytic leukemia. Proc. Natl. Acad.
Sci. USA 84. 9257-9260.

Meeker T.C., D. Hardy, C. Willman, T. Hogan, and J. Abrams. (1990). Activation of the Interleukin-3
gene by chromosome translocation in acute lymphocytic leukemia with eosinophllia Blood 76,
285-289.

Menezes J., W. Leibold, and G. Klein. (1975). Biological differences between Epstein-Barr Virus
strains with regard to lymphocyte transforming ability, superlnfection and antigen induction. Exp.
Cell. Res. 92. 478-484.

Mermod N., E.A. O’Nell. T.J. Kelly, and R. Tijan. (1989). The proline-rich transcriptional activator of
CTF/NF-I is distinct from the replication and DNA binding domain. Cell 58, 741 -753.

Mltelman F. (1991). Catalogue of chromosome aberrations In cancer. In Progress and Topics In
Cytogenetlcs. Vol. 5, AA. Sandberg. ed. (New York: WIIey-Liss Press).

Moreau-Gacheiin F., A Tavitian, and P. Tambourln. (1988). Spi-I Is a putative oncogene in virally
induced murlne erythroleukemias. Nature 331, 277-280.

54

Morishita K, 0.8. Parker, M. Mucenski, NA Jenkins, N.G. Copeland; and J.N.
lhle. (1988). Retroviral activation of a novel gene encoding a zinc finger protein in lL-3 dependent
leukemia cell lines. Cell 59, 831-840.

Motokura T., T. Bloom, H.G. Kim, H. Juppner, J.V. Ruderrnan, HM. Kronenberg. and A Arnold.
(1991). A novel cyclin encoded by a bcI-1 linked candidate oncogene. Nature 350, 512-515.

Mucenski M.L., B.A. Taylor, J.M. lhle. J.N. Hartley, H.C. Morse, lli, N.A. Jenkins, and MG. Copeland.
(1988a). identification of a common ecotropic viral integration site evi-i in the DNA of AKXD murlne
myeloid tumors. Moi. Cell. Biol. 8, 301-308.

Murray M.J.. J.M. Cunningham, LF. Parada, F. Daultry, P. Lebowitz, and RA Weinberg. (1983). The
HL-60 transforming cell sequence: a ras oncogene coexisting with altered myc gene in
hematopoietic tumor. Cell 33, 749-757.

Murre C., P.S. McCaw, H. Vaessin, M. Caudy, LY. Jan, Y.N. Jan, C.V. Cabrera, J.N. Buskln, S.D.
Hauschka, AB. Lassar, H. Weintraub, and D. Baltimore. (1989). Interactions between heterologous
helix-ioop-helbi proteins generates complexes that bind specifically to a common DNA sequence.
Cell 58. 537-544.

Namen A.E., A.E. Schmierer, C.J. March, R.W. Overell. LS. Park, D.L Urdal, and D.Y. Mochizukl.
(1988). B cell precursor growth-promoting activity: purification and characterization of a growth
factor active on lymphocyte precursors. J. Exp. Med. 107, 988-1002.

Neil J.C., R. McFarlane, N.M. Wilkie, D.e. Onions, G. Lees, and O. Jarett. (19884). Transduction and
rearrangement of the myc gene by feline leukemia virus in naturally occurring T -cell leukemias.
Nature 308, 814-820.

Neiman P.E., C. Wolf, P.J. Enrletto, and GM. Cooper. (1985). A retrovlral myc gene induces
preneoplastic transformation of lymphocytes in a bursal transplantation assay. Proc. Natl. Acad. Sci.
USA 82, 222-226.

Nilsson K, and G. Klein. (1982). Phenotyplc and cytogenetic characteristics of human B-lymphold
cell lines and their relevance for the etiology of Burkitt’s lymphoma. Adv. Cancer. Res. 37, 31943).

Noguchi T., F. Galland, M. Batoz, M.-G. Mattel, and D. Bimbaum. (1988). Activation of a mcf.2
oncogene by deletion of amino-terminal coding sequences. Oncogene 3, 709-715.

Nordan RP, and M. Potter. (1986). A macrophage-derived factor required by plasmacytomas for
survival and proliferation In vitro. Science 233, 566-569.

Nowell PC, and DA Hungeriord. (1960). A minute chromosome in human chronic grandocytic
leukemia . Science 132, 1497.

Nunez G., L London, D. Hockenbery, M. Alexander, J.P. McKeam, and SJ.
Korsmeyer. (1990). Deregulated bcI-2 gene expression selectively prolongs survival of growth
factor-deprived hematopoietic cell lines. J. lmmunoi. 144, 3002-3010.

Nunn M.F., P.H. Seeberg, C. Moscovici, and P. Duesberg. (1983). Tripartite structure of the avian
erythroblastosls virus E26 transforming gene. Nature 306, 391-393.

Nusse R., and HE. Varmus. (1982). Many tumors induced by the mouse mammary tumor virus
contain a provirus integrated in the same region of the host genome. Cell 31, 99-109.

Nusse R., A. van Ooyen, D. Cox, Y.K.T. Fung, and H. Varmus. (1984). Mode of proviral integration
of a putative mammary oncogene(int-1) on chromosome 15. Nature 307, 131-136.

55

Ochiya T., A. Fujiyama, S. Fukushige, i. Hatakada, K Matsubara. (1986). Molecular cloning of an
oncogene from a human hepatocellular carcinoma. Proc. Natl. Acad. Sci. USA 83, 4993-4997.

O'Conor G.T., H. Rappaport, and EB. Smith. (1985). Childhood lymphoma resembling Burkitt tumor
in the United States. Cancer 18, 411-417.

Ohno S., M. Babonits, F. Wiener, J. Spira, G.Klein, and M. Potter. (1979). Nonrandom chromosome
changes including the lg gene-carrying chromosomes 12 and 6 in pristane-induced mouse
plasmacytomas. Cell 18, 1001-1007.

Ohno S., S. Migita, F. Wiener, M. Babonits, G. Klein, J.F. Mushinski, and M. Potter. (1984).
Chromosomal translocations activating myc sequences and transduction of v-abl are critical events
in the rapid induction of plasmacytomas by pristane and Abelson Virus. J. Exp. Med. 159,
1762-1777.

Ohno H., G. Takimoto, and T.W. McKeithan. (1990). The candidate proto-oncogene bcI-3 is related
to genes implicated in cell lineage determination and cell cycle control. Cell 60, 991-997.

Padua 6.8., N. Barrass, and GA. Currie. (1984). A novel transforming gene in a human malignant
melanoma cell line. Nature 311, 671-673.

Palmiter RD, and R.L. Brinster. (1986). Germ-line transformation of mice. Ann. Rev. Genet. 20,
465-499.

Payne G.S., J.M. Adams, and HE. Varmus. (1982). Multiple arrangements of viral DNA and an
activated host oncogene in bursal lymphomas. Nature 295, 209-213.

Perlmutter R.M., J.L Kiitz, D. Pravtcheva, F. Ruddle. and L Hood. (1984). A novel 6:10 chromosomal
translocation in the murlne plasmacytoma NS-1. Nature 307, 473-476.

Peters R.L, LS. Rabstein, R. van Vleck, G.J. Keiloff, and R.J. Huebner. (1974). Naturally occurring
sarcoma viruses of the Baib/cCr mouse. J. Natl. Cancer inst. 53, 1725.

Peto R., F.J. Roe, PM Lee, L. Levy, and J. Clark. (1975). Cancer and ageing in mice and men. Br.
J. Cancer 32, 411-426.

Pierce J.H., and S.A. Aaronson. (1983). In vitro transformation of murlne pre-B lymphoid cells by
Snyder-Theiien feline sarcoma virus. J. Virol. 46, 993-1002.

Pierce J.H., and S.A. Aaronson. (1985). Myeloid cell transformation by ras-containing murlne
sarcoma viruses. Mol. Cell. Biol. 5, 667-674.

Pltot H.C. (1990). Carcinogenesis by chemicals: a multifaceted process. in The Cellular and
Molecular Biology of Human Carcinogenesis. RK Boutwell, and LL
Reigei. eds. (San Diego: Academic Press). pp. 81-109.

Poirier Y., C. Kozak, and P. Joliculeur. (1988). Identification of a common helper provirus integration
site in Abelson murlne leukemia virus induced lymphoma DNA. J. Virol. 62, 3985-3992.

Ponder BAJ. (1990). inherited predisposition to cancer. Trends Genet. 6, 213-218.

Potter M., and C. Boyce. (1962). induction of plasma cell neoplasms in strains Balb/c mice with
mineral oil and mineral adiuvants. Nature 193, 1086-1087.

Potter M.. and RC. MacCardie. (1964). Histology of developing plasma cell neoplasia induced by
mineral oil in Baib/c mice. J. Natl. Cancer. institute 33, 497-515.

56

Potter M., J. Wax, E. Mushinski, S. Brust, M. Babonits, F. Wiener, J.F. Mushinski, D. Mezeblsh, R.
Skuria, U. Rapp, and H.C. Morse ill. (1986). Rapid induction of plasmacytomas in mice by pristane.
and a murlne recombinant retrovirus containing an avian v-myc and a defective raf oncogene. Curr.
Topics. Micro. lmmunol. 132, 40-43.

Preud’homme J.L., G. Fiandrin, M.T. Daniel, and JG Brouet. (1975). Burkitt’s tumor cells In acute
leukemia. Blood 46, 990-992.

Preud’homme J.L, K. Deilagi, P. Gugliemi, LB. Vogler, F. Danon, GM. Lenoir, F. Valensi, and JG
Brouet. (1985). Immunologic markers of Burkitt’s lymphoma cells. in Burkitt’s lymphoma: a human
cancer model. G.M. Lenoir, G. O’Conor, and C.L.M. Olweny. eds. (Lyon: lARC Scientific publication
No. 60). pp. 47-64.

Raghoebier S., J.H.J.M. van Krieken, J.C. KIuin-Neiemans, A. Gillis, G.J.B. van Ommen, AM.
Ginsberg, M. Raffeid, and PM. Kluin. (1991). Oncogene rearrangements in chronic B-cell leukemia.
Blood 77, 1560-1564.

Rechavi G., D. Givoi, and E. Canaani. (1982). Activation of a cellular oncogene by DNA
rearrangement: possible involvement of an lS-like element. Nature 300, 607-611.

Reed J.C., M. Cuddy, T. Slabiak, C.M. Croce, and PC. Nowell. (1988). Oncogenlc potential of bcI-2
demonstrated by gene transfer. Nature 336, 259-261.

Reed J.C., S. Hardar, C. M. Croce, and MP. Cuddy. (1990). Complementatlon by bcI-2 and c-Ha-ras
oncogenes in malignant transformation of rat embryo fibroblasts. Mol. Cell. Biol. 10, 4370-4374.

leokh R., Rouauit J.P. Wahbi K Gadoux M. and others. (1991). A chromosome 12 coding region
is juxtaposed to the MYC protooncogene locus in a t(8; 12)(q24;q22) translocation in a case of
B-cell chronic lymphocytic leukemia. Genes Chromosom. Cancer 3, 24-36.

Scalenghe F., E. Turco. J.-E. Edstrom, V. Pirrotta, and M. Meili. (1981). Microdissection and cloning
of DNA from specific region of Drosophila meianogaster polytene chromosomes. Chromosoma 82,
205-216.

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

Schwartz RC, and ON. Witte. (1988). The role of multiple oncogenes in hematopoietic neoplasia.
Mutation Research 195, 245-253 .

Schwartz R.C.. L.W. Stanton. S.C. Riley. K.B. Marcu, and ON. Witte. (1986a). Synergism of v-myc
and v-Ha-ras in the in vitro neoplastic progression of murlne lymphoid cells. Mol. Cell. Biol. 6,
3221-3231.

Seremetis S., G. lnghlraml, D. Ferraro, E.W. Newcomb, D.M. Knowles, G.-P. Dotto, and R.
Dalia-Favera. (1989). Transforrnatlon and plasmacytoid differentiation of EBV-Infected human B
lymphoblasts by ras oncogenes. Science 243. 660-663.

Seth A., E. Priel, and Vande Woude GP. (1987). Nucleotide triphosphate—dependent DNA-binding
properties of mos protein. Proc. Natl. Acad. Sci. USA 84, 3560-3564.

Seto M., U. Jaeger, R.D. Hockett, W. Graninger, S. Bennett, P. Goldman, and SJ. Korsmeyer.
(1988). Alternative promoters and axons. somatic mutation and deregulation of the bcI-2-lg fusion
gene in lymphoma. EMBO J.

57

Shen-0ng G.L, E.J. Keath, S.P. Piccoli, and MD. Cole. (1982). Novel myc oncogene RNA from
abortive immunoglobulin-gene recombination in mouse plasmacytomas. Cell 31, 443-452.

Shen-Ong G.LC., M. Potter, J.F. Mushinski, S. Lavu, and E.P. Raddy. (1984). Activation of the myc
locus hybridization Insertionai mutagenesis in piasmacytoid iymphosarcomas. Science 226,
1077-1080.

Shepard J.S., D.H. Wurster-Hili. 0.S. Pettenglil, and GD. Sorenson. (1974a). Glemsa banded
chromosomes of mouse myeloma in relationship to oncogeniclty. Cytogenet. Cell Genet. 13,
279-304.

Shepard J.S., 0.6. Pettenglil, D.H. Wurster-Hiil, and GD. Sorenson. (1974b). Karyotype marker
formation and oncogenicity in mouse plasmacytomas. J. Natl. Cancer inst. 56, 1003-1011.

Shepard J.S., 0.6. Pettenglil, D.H. Wurster-Hill, and D. Sorenson . (1978). A specific chromosome
breakpoint associated with mouse plasmacytomas. J. Natl. Cancer inst. 61, 225-256.

Shih C., L Padhy, M. Murray, and RA. Weinberg. (1981). Transforming genes of carcinomas and
neuroblastomas introduced into mouse fibroblasts. Nature 290, 260-264.

Shih C.-C., J.P. Stoye. and J.M. Coffin . (1988). Highly preferred targets for retrovirus integration.
Cell 53, 531-537.

Shimizu K, M. Goldfarb, Y. Suard, M. Perucho, Y. Li, T. Kamata, J. Feramisco, E. Stavnezer, J.
Fogh, and M.H. Wigler. (1983). Three human transforming genes are related to the viral ras
oncogenes. Proc. Natl. Acad. Sci. USA 80. 2112-2116.

Shimizu K, Y. Nakatsu, M. Seklguchl, K Hokamura, K Tanaka, M. Terada, and T. Sugimura. (1985).
Molecular cloning of an activated human oncogene. homologous to wet. from primary stomach
cancer. Proc. Natl. Acad. Sci. USA 82, 5641-5645.

Siden E.J., D. Baltimore, D. Clark. and NE. Rosenberg. (1979). lmmunoglobulln synthesis by
lymphoid cells transformed in vitro by Abelson murlne leukemia virus. Cell 16. 389-396.

Silver J., and C. Kozak. (1986). Common proviral integration region on mouse chromosome 7 in
lymphomas and myelogenous leukemias induced by Friend murlne leukemia virus. J. Virol. 57,
526-533.

Skiar M.D., B.J. White, and WP. Rowe. (1974). Initiation of oncogenic transformation of mouse
lymphocytes in vitro by Abelson Leukemia Vinis. Proc. Natl. Acad. Sci. USA 71, 4077-4081.

Stehelin 0, RV. Guntaka, H.E. Varmus. and J.M. Bishop. (1976a). Purification of DNA
complimentary to nucleotide sequences required for neoplastic transformation of fibroblasts by
Avian sarcoma viruses. J. Mol. Biol. 101, 349-365.

Stiles CD, and AA. Kawahara. (1978). The growth behavior of virus-transforrned cells in nude mice.
in The Nude Mouse in Experimental and Clinical Research. J. Fogh, and B. Gioranella. eds. (New
York: Academic Press). pp. 385-409.

Stocking C., C. Loliger, M. Kawai, S. Suciu, N. Gough, and W. 0stertag. (1988). identification for
genes involved in growth autonomy of hematopoietic cells by analysis of factor-independent
mutants. Cell 53. 869-879.

Suematsu S., T. Matsuda, K Aozasa, S. Aklra, N. Nakano, S. 0hno, J.-l. Mlyazakl, K-l. Yarnamura,
T. Hirano, and T. Klshimoto. (1989). lgG 1 plasmacytosis in interleukin 6 transgenic mice. Proc. Natl.
Acad. Sci. USA 86, 7547-7551.

58

Swendeman S., and DA. Thoriey-Lawson. (1987). The activation antigen BLAST-2, when shed, is
an autocrine BCGF for normal and transformed B cells. EMBO J. 6, 1637-1642.

Taira M., T. Yoshida, K. Miyagawak. and H. Sakamoto. M. Terada. and T. Sugimura. (1987). CDNA
sequence of human transforming gene hst and identification of the coding sequence required for
the transforming activity. Proc. Natl. Acad. Sci. USA 84. 2980-2984.

Takada D., and T. Osato. (1979). Analysis of the transformation of human lymphocytes by
Epstein-Barr Virus l. sequential occurrence from the virus-deterrnined nuclear antigen synthesis, to
blastogenesis, to DNA synthesis. lntervirology 11, 30-39.

Takahashi M., J. Ritz, and GM. Cooper. (1985). Activation of a novel human transforming gene, ret,
by DNA rearrangement. Cell 42, 581-588.

Tedder T.F., LT. Clement, and MD. COOper. (1984). Expression of C3d receptor during human B
cell differentiation: lmmunofluorescence analysis with the HB5 monoclonal antibody. J. Immunol.
133. 668-673.

Tsichlls P N., G. Strauss, and PH Liu. (1983a). A common region for proviral DNA integration in
MoMuLV-induced rat thymic lymphomas. Nature 302, 445-449.

Tsichlls P.N., P.G. Strauss. and CA. Kozak. (1984). Cellular DNA region involved in induction of
thymic lymphoma(MIvi-z) maps to chromosome 15. Moi. Cell. Biol. 4, 997-1000.

Tsichlls P.N., P.G. Strauss, and MA. Lohse. (19853). Concerted DNA rearrangements in Moloney
murlne leukemia virus-induced thymomas: a potential synergistic relationship in oncogenesis. J.
Virol. 56, 258-267.

Tsujimoto Y. (1989). Overexpression of the human bcI-2 gene product results in growth
enhancement of Epstein-Barr virus-immortalized B cells. Proc. Natl. Acad. Sci. USA 86, 1958-1962.

Tsujimoto Y., J. Yunis, L 0norato-Showe, J. Erikson, P.C. Nowell, and CM. Croce. (1984a).
Molecular cloning of the chromosomal breakpoint of B-ceii lymphomas and leukemias with the
t(11;14) chromosomal translocation. Science 224, 1403-1406.

Tsujimoto Y., LR. Finger, J. Yunis, P.C. Nowell, and CM. Croce. (1984b). Cloning of the
chromosome breakpoint of neoplastic B cells with the t(14;18) chromosome translocation. Science
226, 1097-1099.

Tsujimoto Y., E. Jaffe, J. Cossman, J. Gorham, P.C. Nowell. and CM. Croce. (1985a). Clustering
of breakpoints on chromosome 11 in human B-cell neoplasms with the t(11;14) chromosome
translocation. Nature 315, 340-343.

Tsujimoto Y., J. Gorham, J. Cossman, E. Jaffe, and CM. Croce. (1985b). The t(14;18) chromosomal
translocations involved in B-ceii neoplasms result from mistakes in VDJ joining. Science, 229
1390—1392.

Tsujimoto Y., J. Cossman, E. Jaffe, and CM. Croce. (1985c). involvement of the bcI-2 gene in
human follicular lymphoma. Science 228, 1440-1443.

van der Berghe H., K Vermaelen, A. Louwagie, A. Criei, C. Mecuccl, and J.-P.
Vaerman. (1984). High incidence of chromosome abnorrnallties In lgGa myeloma. Cancer Genet.
Cytogenet. 11, 381-387.

van Lohuizen M., M. Breuer, and A. Bems. (1989a). N-myc is frequently activated by proviral
integration in MuLV induced T cell Lymphomas. EMBO J. 8. 133-136.

59

van Lohuizen M., S. Verbeek, S. Scheijen, E. Wientjens, H. van der Guiden, and
A Bems. (1991). Cell 65, 737-752. identification of cooperating oncogenes in Eu-myc transgenic
mice by provirus tagging.

van Ness B., M. Shapiro, P.E. Kelley, R.P. Perry, M. Weigert, P. D’Eustachio, and F. Ruddle.
Varmus H. (1989). A historical overview of oncogenes. ln Oncogenes and the Molecular Origins of
Cancer. R.A. Weinberg ed. (New York: Cold Spring Harbor Laboratory Press). pp. 3-44.

Vaux D.L, 8. Cory and J.M. Adams. (1988). Bel-2 gene promotes haemopoietlc cell survival and
cooperates with c-myc to immortalize pre-B cells. Nature 335, 440-442.

Vijaya S., D.L. Steffen, C. Kozak, and H.L Robinson. (1987). A region with frequent proviral
insertions in Moloney murine leukemia virus induced rat thymomas. J. Virol. 61, 1164-1170.

Villemur R., Y. Monczak, E. Rassart, C. Kozak, and P. Jolicoeur. (1987). identification of a new
common proviral integration site in cross passage a murlne leukemia virus-induced mouse thymoma
DNA. Mol. Cell. Biol. 7, 512-522.

Vlnk A, P. Coulie, G. Warnler, J.-C. Renauld, M. Stevens, D. Donckers, and J. Van Snick . (1990).
Mouse plasmacytoma growth in vivo: enhancement by interleukin 6 (lL-6) and inhibition by
antibodies directed against iL-6 or its receptor. J. Exp. Med. 172, 997-1000.

Voisky D.J., T. Gross, F. Sinangil, C. Kuszynski, R. Bartzatt, T. Dambaugh, and E. Kieff. (1984).
Expression of Epstein-Barr Virus(EBV) DNA and cloned DNA fragments in human lymphocytes
following Sendai virus envelop-mediated gene transfer. Proc. Natl. Acad. Sci. USA 81, 5926-5930.

Wang D., D. Liebowitz, and E. Kieffe. (1985). The EBV membrane protein expressed in immortalized
lymphocytes transforms established rodent cells. Cell 43, 831-840.

Wang D., D. Liebowitz, F. Wang, 0. Gregory, A. Rickinson, R. Larson, T. Springer, and E. Kieff.
(1988). Epstein-Barr virus latent infection membrane protein alters the human B-lymphocyte
phenotype: deletion of the amino terminus abolishes activity. J. Virol. 62, 4173-4184.

Wang F., OD. Gregory, 0. Sample, M. Rowe, D. Liebowitz, R. Murray, A. Rickinson, and E. Kieff.
(1990). Epstein-Barr virus latent membrane protein (LMP1) and nuclear protein 2 and 3c are
effectors of phenotypic changes in B lymphocytes: EBNA-2 and LMP1 cooperatively induce CD23.
J. Virol. 64, 2309-2328.

Whitlock CA. and 0N. Witte. (1981). Abelson virus-infected cells can exhibit restricted in vitro
growth and low oncogenic potential. J. Virol. 40. 577-584.

Whitlock C.A., S.F. Ziegler, and 0N. Witte. (1983b). Progression of the transformed phenotype in
clonal lines of Abelson virus-infected lymphocytes. Mol. Cell. Biol. 3, 596-604.

Ymer S., W.Q.J. Tucker, C.J. Sanderson, A.J. Hapel, H.D. Campbell, and LG. Young. (1985).
Constitutive synthesis of interleukin-3 by leukemia cell line WEHi-3B is due to retrovlral insertion near
the gene. Nature 317, 255-259.

Yosida T.H., H.T. lmai, and U. Moriwaki. (1970). Chromosomal alteration and development of tumors
XXI cytogenetic studies of primary plasma cell neoplasms induced in Balb/c mice. J. Natl. Cancer
Inst. 45, 411-418.

Young D., G. Waitches, Birchmeier, 0. Fasano, and M. Wigler. (1986). isolation and characterization
of a new cellular oncogene encoding a protein with multiple potential transmembrane domains. Cell
45, 711-719.

Youm;
DGCE’S

Yunis
Yunis
Zach E
in bio:

Canoe

zur H;

60

Young J.C., C. Mikttail, B. Gishizky, and 0N. Witte. (1991). Hyperexpresslon of interleukin-7 is not
necessary or sufficient for transformation of a pre-B lymphoid cell line. Mol. Cell. Biol. 11, 854-864.

Yunis J.J. (1976). High resolution of human chromosomes. Science 191, 1268-1270.

Yunis J.J. (1983). The chromosomal basis of human neoplasia. Science 221, 227-236.

Zech L, U. Haglund, K Nilsson, and G. Klein. (1976). Characteristics chromosomal abnormalities
in biopsies and lymphoid cell lines from patients with Burkitt and non-Burkitt lymphomas. int. J.
Cancer 17, 47-56.

zur Hausen H. (1980). Oncogenlc herpesviruses. in DNA Tumor Virus. J. Tooze ed. (New Work:
Cold Spring Harbor Laboratory). pp. 755-757.

Chapter 2

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
August15,1991

Tumorigenesis of a V—Ha-gas-Bxpressing
Pre-B Cell Line Selects for c-ugg Activation

Shu-Chih Chen, Diane Redenius, and Richard C. Schwartz

Department of Microbiology, Michigan State University, East Lansing, MI
48824-1101

Received June 24, 1991

 

Seven tumors independently derived from a v-Ha-ggg-expressing pre-B cell
line were examined to determine the oncogene activations cooperating with
v-Ha-gag in 13 vivo tumor progression. The pre-B cell line was generated by
infection with Moloney murine leukemia virus (MoMuLV) and a MoMuLV-derived
recombinant expressing v-Ha-ggg. Two of seven tumors possessed a MoMuLV
integration immediately upstream and in reverse transcriptional orientation to
c-m. This correlated with a 3-fold increased level of c-myg mRNA. Two
other tumors displayed elevated c-myg mRNA levels, although the mechanism of
enhanced expression was unclear. Thus the tumor progression of a v-Ha-gag-
expressing murine pre-B cell line selects for the activation of c-myg. elem

Academic Press . Inc .

 

It is generally accepted that tumorigenesis proceeds through multiple
events involving the altered expression or structure of specific oncogenes
that function in the regulation of cellular growth. The requirement of two
cooperating' oncogenes for the 13_ giggg neoplastic transformation of many
primary cells may be a reflection of this process. Several examples of
oncogene cooperativity in lymphoid transformation have been described in
experimental systems. Infection by Abelson murine leukemia virus accelerates
the induction of plasmacytomas by pristane (1). These tumors not only express
v-g_l, but possess the same c-myg translocations observed in tumors induced by
pristane alone. Cooperativity between v-ggg and v-myg has been observed in
the induction of both 3 lymphomas (2) and plasmacytomas (3). A synergy of
v-Ha-gag and v-myg has been observed in the i3 giggg transformation of pre-B
lymphoid cells (4). Either v-Ha-ggg or v-ggg can cooperate with activated my;
in the transformation of Ep-myg transgenic pre-B lymphoid cells (5,6). While
in 315:9 manipulations have revealed pairs of oncogenes capable of eliciting
full neoplastic transformation, there have been few tests of what oncogenes
might be selected 13 yigg to achieve transformation given the prior expression
of a single oncogene that is incapable by itself of eliciting transformation.

The retroviral activation of c-fms has been observed in a monocytic tumor

61

62

Vol. 178, No. 3. 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

induced by a c-myg-expressing retrovirus (7). 3;; mutations were observed in
some lymphomas generated in Ep-myg transgenic mice (8).

In earlier studies, we found that v-Ha-m—expressing pre-B lymphoid
cells displayed an intermediate transformed phenotype and were infrequently
tumorigenic (4). The irregular occurrence of tumors derived from these cells
suggested continued neoplastic progression in gigg, or selection in zigg for
pro-existing subpopulatlons containing additional neoplastic mutations. Since
v-myg was found to cooperate with v-Ha-ggg in the in giggg transformation of
pre-B lymphoid cells, we decided to investigate the status of c-myg in tumors

derived from a v-Ha-gag-expressing pre-B cell line.

IAIBRIALS AND METHODS

W. R2, a v-Ha-m-transformed murine pre-B cell line, and
tumors derived from R2 are described in Schwartz et al. (4). Tumor cell lines
were produced from explanted tumors by dispersal onto feeder cultures of
adherent bone marrow cells (9). Cell lines were cultured over feeder cells in
RPMI 1640 with St fetal calf serum and 5 x 10-5 H 2-mercaptoethanol.

Nucleic acid analyses. Cytoplasmic RNA was isolated by a sodium dodecyl
sulfate-urea procedure as described by Schwartz et al. (10). Poly A+ RNA was
selected by oligo-dT cellulose chromatography. RNA was denatured,
electrophoresed in a formaldehyde-1t agarose gel and transferred to Nytran
(Schleicher and Schuell). High molecular weight DNA was isolated from nuclei
collected in the preceding RNA isolation procedure as described in Schwartz et
al. (4). DNA was digested with restriction enzymes as noted, electrophoresed
through agarose and transferred to Nytran.

Hybridization probes were prepared by nick translation through the
incorporation of [a-32P1dATP (3000 Ci/mmol; ICN). The v-Ha-ggg probe was a
0.46 kb EcoRI fragment corresponding to v-Ha-gag encoding sequences (11). The
en! probe was a 0.8 kb BamHI fragment from the gag region of Friend murine
leukemia virus (12). The c-myg probes were the 4.7 kb HindIII fragment of
murine c-myg encompassing exons 1, 2 and 3 (Figures 3 and 6) and the 0.8 kb
SmaI-Sac]: fragment encompassing S' upstream flanking sequences and part of
exon 1 (Figures 4 and 5). The rat glyceraldehyde-3-phosphate dehydrogenase
(rGAPDH) probe was a cloned 1.3 kb cDNA (13). All hybridizations were
performed under aqueous conditions in 5 x SSC at 65°C and washed to a
stringency of 0.1 x SSC at 65°C.

RESULTS

Seven tumors independently derived from a v-Ha-ggg-expressing pro-8 cell
line were studied in regard to their retroviral integration sites and status
of c-mxg structure and expression. These tumors were derived from the clonal
R2 cell line that was generated by infection of fresh murine bone marrow with
a recombinant v-Ha-ggg-expressing retrovirus and MoMuLV (4). This cell line
is dependent upon a bone marrow-derived cellular feeder layer for growth and
gives rise to tumors infrequently.

r a e v d fro . In order to verify that the tumors were
derived from R2, sites of integration of the vbfla-ras-expressing retrovirus
were compared between that cell line and the tumors. Southern hybridization

of BcoRI-digested DNA with a v—Ha-zag probe showed that the tumors contained

63

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

 

FIGURI 1. Viral :3; integrations. Southern blot analysis of DNAs from liver
(L), R2, and seven tumors (1-7). DNA was digested with lcoRI and lOpg of each
sample was electrophoresed through O.BI agarose. The blot was probed for
v-Ha-ggg. Size markers are the positions of an ethidium bromide-stained Hind
III digest of bacteriophage l and are denoted in kilobases.

the same 5.3 kb proviral integration fragment as R2 (Figure 1). In addition
to the 5.3 kb fragment, there is a 23 kb fragment representing the endogenous
c-Ha-ggg gene in all the DNAs, and an additional 4.2 kb fragment in tumor 7.
This 4.2 kb fragment represents an additional v-Ha-zgg proviral insertion site
(data not shown).

a v v s . If the tumors resulted from
genetic lesions subsequent to the introduction of v-Ra;;gg, then they would be
expected to be clonal outgrowths from the parental R2 cell line. R2 was
generated by infection with a recombinant v-Ha-gag—expressing retrovirus and
MoMuLV. since proviral integrations can occur subsequent to the original
integration event, it is reasonable to expect clonal outgrowths to contain
unique sites of proviral integration in addition to any common sites derived
from the parent. Examination of integrations of the v-Ha-ggg retrovirus
revealed only one new site in tumor 7 (Figure 1). We then examined the sites
of MoMuLV integration in R2 and the tumors by Southern hybridization of BglII-
digested DNA, using a probe for the ecotropic MuLV egg gene. All of the
tumors possessed gag-hybridizing BglII fragments unique to the tumors and not
present in R2 (Figure 2). Conversely, some fragments present in R2 were
absent in the tumor cell lines. Fragments unique to the tumors represent
either new proviral integrations or rearrangements of pre-existing
integrations. The tumors (with the exception of tumor 4) clearly possess
guy-hybridizing BglII fragments more similar to each other than to R2,
although individual tumors also possess unique BglII fragments. These data
suggest that the tumors (with the exception of tumor 4) are clonal outgrowths
derived from a single subclone of R2. Tumor 4 appears more closely related to

the original R2 isolate.

64

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

      

fix
its”
~ “istflyr‘i’lfis
goals 2:. MoMuLV integrations. Southern blot analysis of blue from BALD/c
mouse liver (1.), R2, and seven tumors (1-7). DNA was digested with lglII and
lOug of each sample was electrophoresed through 0.8! agarose. The blot was
probed for murlne ecotropic m sequences. Size markers are the positions of

an ethidium bromide-stained Hindu! digest of bacteriophage .\ and are denoted
in kilobases.

Rearranganent of cm in tumors 1 and 3. Southern blot analysis
of Blue from liver (L), R2 and seven tumors (1-7). DNA was digested with
Icon: and long of each sample was eletrophoresed through 0.3\ agarose. Size
markers to the left of photographs are the positions of an ethidium bromide-
stained flindIII digest of bacteriophage l and are denoted in kilobases.

 

Raving demonstrated that v-m can cooperate with v-aa-m in the in mm
neoplastic progression of pre-B lymphoid cells (4), we investigated the status
of the c-m locus in the tumors. Southern hybridization of lcoRI-digested
blue with a c-m probe revealed that tumors 1 and 3 possessed a m-
hybridizing fragment of mid: in addition to the normal 23kb fragment (Fig. 3).

Since the tumors displayed several proviral integrations, we
investigated whether the rearrangement of c-m in tumors 1 and 3 was caused
by integration of MoMuLV. Assuming a proviral integration site near cm,
and with complete restriction maps of both MoMuLV and cm, we could' predict
the sizes of restriction fragments that would be revealed by a c-m probe in
a Southern analysis (Fig. 4). Assuming that an intact MoMuLV provirus had
integrated adjacent to exon 1 of c-m in a reverse transcriptional
orientation (the 5' long terminal repeat (LTR) juxtaposed to exon 1), a series
of expected sizes were calculated (Table 1). We performed double digestions
with xbaI/xhoI, SacI/xhoI and XpnI/XhoI to examine the presence of a MoMuLV
LTR pattern of restriction sites in the vicinity of c-m (Fig. 4). The data
generated by digestions with restriction enzymes specific for LTR sites were
in excellent agreement with the predicted sizes (Fig. 4 and Table 1). These
data suggested integration of a MoMuLV genome within lOObp of the 5'-terminue
of c-m. In order to further verify the presence of KoHuLV near c-m in
tumors 1 and 3, BglII-digested DNA was analyzed by Southern blot with a probe
for the MuLV m gene and a probe for c-m that included exon 1 and 5'

65

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

XbaI—Xhol SacI-Xhol Kpnl-Xho

 

 

as I a a5 1‘ a 'sa' f ’f?
23. - I
3'3: . .‘- t as -.
4L4. ‘ ‘_ ..
.. K -——3.7
-
2. ' 9'
:13: .- ‘No .
s w- +——t9
‘.4 #1.?
-~ o——4190
‘ s——1Lss
(176
2..— L P
sum
'"“'5 w
a.
i 1 i I ITII'I 1 1 fit. It I
= H» wit a a JIL L
I 9 A. p g n g g.
rxgggg 4. MoMuLV integration near c-myc. Southern blot analysis of DNAs from
R2 and tumors 1 and 3. DNA was digested and lOug of each sample was
electrophoresed through 1.6t agarose. Size markers to the left are the

positions of an ethidium bromide-stained BindIII digest of bacteriophage A and
are denoted in kilobases. The positions of rearranged c-myg fragments are
marked to the right of photographs and denoted in kilobases. DNA was doubly
digested with either XbaI and XhoI, or SacI and XhoI, or anI and XhoI. The
blot was probed with a SmaI-SacI fragment encompassing' upstream flanking
sequences and part of the first exon of murine c-m. A schematic of the
murlne c-myg locus with MoMuLV integrated upstream in a reverse
transcriptional orientation is included. The following restriction sites are
marked: KpnI (K), SacI (S), XbaI (Xb), SmaI (Sm), BclI (8c), AatII (A), NotI
(N), XhoI (Xh) and 89111 (89). The probe is delineated by a line labeled '9'
above the map. Note that probe ”P“ is interrupted by MoMuLV integration.

flanking sequences extending to a SmaI site about SOObp upstream of c-m
(Fig. 5). Tumors l and 3 displayed two identical rearranged c-myg-hybridizing

fragments, one of which (see arrowheads) co-hybridized with the envelope

Table 1. Restriction fragments in the 5'
c-myg region of tumors 1 and 3

 

 

Fragment Rxpected Size Observed Size
lxon l/LTR:
XhoI-XDaI 809bp 750bp
SacI-SacI 919 840
XhoI-anI 992 900
Upstream Region/LTR:
XbaI-XbaI 1797 1700
SacI-SacI 1954 1870
xpnI-xpnt 4000‘ 3700

 

‘ Estimated value since the region between the upstream KpnI site and c-myc
is not completely sequenced.

66

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

guises

c-mr wuwzauuasu

a: ”1, 2 a 4 5 _07 7
.« '
C-MVC ..' ‘-._.“

 

'DIII ..III—sQ’ <::> GAFDN an. pl. 1.. ‘l’ II. 'II' J

FIGURE 5. En! and c-mxg co-hybridize. southern blot analyses of DNAs from
liver (L), R2 and tumors 1, 2, 3 and 5. DNA was digested with BglII and lOug
of each sample was electrophoresed through 0.8t agarose. Size markers are the
positions of an ethidium bromide-stained HindIII digest of bacteriophage A and
are denoted in kilobases. Arrowheads mark co-hybridization. (A) The blot was
probed for murlne ecotropic virus guy sequences. (3) The blot was probed for
c-myg with a SmaI-SacI fragment encompassing upstream flanking sequences and
part of exon 1.

FIGURE 6. Elevated c-myg mRNA levels. Northern blot analyses of polyA+ RNA
from R2 and seven tumor cell lines (1—7). Each sample of RNA was the polyA+
fraction selected from lSOug of total cytoplasmic RNA. The blot was probed
for both c-myg and rGAPDH. Relative levels of c-myg mRNA were determined by
densitometry and normalized to rGAPDH.

region probe. An LTR probe detected fragments that co-migrate with both
rearranged c-mJLc-hybridizing fragments indicating integration of both viral
LTRs (data not shown). Clearly, tumors 1 and 3 had suffered MoMuLV
integration in close proximity to c—myg.

MoMuLV ' “on causes i ‘ 'ann of c-mvc. Having
observed proviral insertion near c-mg in two tumors, we decided to test
whether expression of c-myg mRNA was altered in these or any of the other
tumors. Cytoplasmic poly A+ RNAs of R2 and the seven tumor cell lines were
prepared from actively growing cells and examined by Northern hybridization
(Fig. 6). The same blot was hybridized successively with c-M and rGAPDH
probes. Hybridization to the rGAPDH probe provided a control for gel loading.
Densitometry revealed that tumors 1 and 3 had 3—fold elevated levels of c-m
mRNA. Tumors 6 and 7, which displayed no gross abnormalities in c-m, had 4-
fold elevated levels of c-m mRNA. The c-M mRNAs were all 2.4kb in length,

suggesting normal promoter usage and an unaltered RNA structure.

DISCUSSION
The data presented here demonstrate selection for c-M activation in
the lymphomagenesis of a v-Ha-m—transformed murine pre-B lymphoid cell line.
Four of seven tumors examined had 3 to 4—fold elevated levels of c—m mRNA.

In the case of tumors 1 and 3, MoMuLV had integrated immediately 5' to c-m

VoLW

in a
natu

{330

othe
(22)

I'll

1m
in 1
I!!!
V-Ha
0th

ever

Vit!

67

Vol. 178. No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

in a reverse transcriptional orientation, suggesting enhancer activation. The
nature of the events leading to elevated levels of c-myg mRNA in the other two
tumors is unclear.

The activation. of c-myg, by retroviral integration is 'the prominent
feature of lymphomas induced by avian leukosis virus (14,15). MoMuLV

activation of c-myg is also a frequent event in T lymphomas induced by that
virus (16,17). This report of MoMuLV activation of c-myg in a murine B

lymphoma is novel. The structure of the MoMuLV integration in tumors 1 and 3
is reminiscent of retroviral integrations observed in murine T lymphomas (16-
19). The provirus is located upstream of c-myg in a reverse transcriptional
orientation. The modest 3-fold increase in the steady state level of c-myg
mRNA is similar to that observed by others in T lymphomas (18,20,21).

We have examined v-Ha-m-induced B lymphomas for rearrangements at
other frequent MoMuLV integration sites found in T lymphomas. Neither ulyi;1
(22), 513i;z (23), fllvi-B (24), Mlvi-4 (25), nor pim;1 (26) were found to be
rearranged (data not shown). We have not yet examined a large enough panel of
tumors to assess the importance of these frequent integration sites in B cell
lymphomagenesis, but examination of MoMuLV integration sites may be fruitful
in this system. The seven tumors studied possess MoMuLV integration-related
restriction fragments unique to the tumors and absent from 32, the parental
v-Ha-m-expressing cell line. This indicates that the tumors are clonal
outgrowths presumably having advanced in tumor progression due to mutagenic
events. The patterns of viral integration in six of the tumors are more
similar to each other than to R2. This suggests that a single subclone of R2
was predisposed toward tumorigenesis. The presence of MoMuLV integration near
c-myg in two of these six tumors indicates that retroviral activation of c-myg
is at least the third event in the progression of these tumors. This suggests
that the activation of oncogenes other than gag and my; may be important for
tumorigenesis. Our finding of common MoMuLV integration sites in addition to
two instances of integration near c-myg presents the possibility that sites of
MoMuLV integration may identify additional genetic elements that can cooperate
with v-Ha-gag in the tumor progression of B lymphoid cells.

Acknowledgment: This work was supported by grant CA45360 from the National
Cancer Institute.

REFERENCES

1. Ohno, s., Higata, 5., Weiner, P., Babonits, N., Klein, 6., flushinski, J.
1., and Potter, H. (1984) J. Exp. Med. 159:1762-1777.

2. Rapp, U. R., Cleveland, J. L., Fredrickson, T. N., Holmes, K. L., Morse
III, R. C., Jansen, H. w., Patschinsky T., and Sister, x. (1985) J.
Virol. 55:23-33.

3. Troppmair, J., Potter, M., wax, J. 8., and Rapp, U. R. (1989) Proc.
Natl. Acad. Sci. USA 86:9941-9945.

68

Vol. 178, No. 3, 1991 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

4. Schwartz, R. C., Stanton, L. 10., Riley, 8. C., Marcu, K. B., and
Witte, O. N. (1986) Mol. Cell. Biol. 633221-3231.

5. Alexander, W. 8., Adams, J. N., and Cory, 8. (1989) hol. Cell. Biol.
9:67-73.

6. Langdon, W. Y., Harris, A. N., and Cory, S. (1989) Oncogene Res.

' :253-258.

7. Baumbach, N. R., Colston,B. N., and Cole, H. D. (1989) J. Virol.
62:3151-3155.

8. Alexander, W. 8., Bernard, 0., Cory, 8., and Adams, J. N. (1989)
Oncogene 4:575-581.

9. Whitlock, C. A., Ziegler, S. F., Treiman, L. J., Stafford, J. 1., and
Witte, O. N. (1983) Cell 32:903-911.

10. Schwartz, R. C., Sonenshein, C. 3., Bothwell, A. and Gefter, N. L.
(1981) J. Immunol. 126:2104-2108.

11. Ellis, R. N., De Feo, D., Maryak, J. 14., Young, H. A., Shih, T. Y.,
Chang, 8. H., Lowy, D. R., and Scolnick, E. H. (1980) J. Virol.
36:408-420.

12. Silver, J., and Kozak, C. (1986) J. Virol. 57:526-533.

13. Port, P., Marty, L., Piechaczyk, N., 31 Salrouty, 8., Dani, C.,
Jeanteur, J., and Blanchard, J. H. (1985) Nucl. Acids Res. 13:1431-
1442.

14. Hayward, W. 8., Neel, B. C., and Astrin, S. N. (1981) Nature 290:475-
480.

15. Payne, G. 8., Bishop, J. N., and Varmus, H. E. (1982) Nature 295:209-
214.

16. Selten, C., Cuypers, H. T., Zylstra, N., Hilief, C. and Berns, A.
(1984) EHBO J. 3:3215-3222.

l7. Steffen, D. L. (1984) Proc. Natl. Acad. Sci. USA 81:2097-2101.

18. Corcoran, L. N., Adams, J. N., Dunn, A. P., and Cory, S. (1984) Cell
37:113-122.

19. Li, Y., Holland, C. A., Hartley, J. H. and Hopkins, N. (1984) Proc.
Natl. Acad. Sci. USA 81:6806-6811.

20. Reicin, A., Yang, J. 9., Marcu, K. B., Fleissner, R., Xoehne, C. P. and
O'Donnell. P. V. (1986) Mol. Cell. Biol. 6:4088-4092.

21. Steffen, D. L., and Nacar, E. Q. (1988) Virology 164:55-63.

22. Tsichlis, P. N., Strauss, P. C., and flu, L. F. (1983) Nature 302:445-
449.

23. Tsichlis, P. N., Strauss, P. C., and Lohse, H. A. (1985) .J. Virol.
56:258-267.

24. Tsichlis, P. N., Lohse, N. A., Szpirer, C., Szpirer, J., and Levan, G.
(1985) J. Virol. 56:932-942.

25. Lazo, P. A., Lee, J. 8., and Tsichlis, P. N. (1990) Proc. Natl. Acad.
Sci. USA 87:170-173.

26. Selten, C., Cuypers, H. T., and Berna, A. (1985) EMBO J. 4:1793-1798.

Chapter 3

Cloning and characterization of a viral flanking region
common to B lymphoid tumors derived from

a v-Ha-ras infected pre-B cell line

Shu-Chih Chen,1 Marge Strobel? Neal Copeland,2 Nancy Jenkins,2

and Richard C. Schwartz‘

lDepartment of Microbiology, Michigan State University,

East Lansing, MI 48824-1 101

2Mammalian Genetics Laboratory, BRI-Basic Research Program,
NCl-Frederick Cancer Research Facility,

Frederick, Maryland 21 701

69

70
Abstract:

A viral flanking region, designated as bone-1, common to B lymphoid tumors
derived from a v-Ha-ras oncogene-transformed cell line, R2 (Chen et al., 1991) was
cloned and characterized. This locus may be associated with the growth
advantage of tumors derived from the R2 cell line. We were unable to detect any
transcriptional activity over a 20 kb region in the vicinity of the bonc-1 locus.
However, sequence homology within bonc—1 was found to the genome of other
mammals, and chicken. The high evolutionary conservation within the bone-1
region and the presence of several open reading frames suggest that one or more
sequences of function importance exist therein. Bone-1 was mapped to mouse
chromosome 19, closely linked to the Ipr locus, and distinct from another frequent
viral integration site, gin-1, on the same chromosome. The Ipr locus is linked to
a recessive lymphoproliferation trait. Whether bonc-l is identical to Ipr awaits
further investigation. The fact that it is close to the lpr locus provides a use for this

locus in the search for genes responsible for the Ipr lymphoproliferation trait.

71

Introduction:

Oncogenlc transformation by nonacute transforming retroviruses such as avian
leukosis virus (ALV), mouse mammary tumor virus (MMTV), and murlne leukemia
virus (MLV) depends on the activation of cellular protooncogenes by viral
integration. Studies of frequent viral integration sites in animal tumors have been
fruitful both in identifying new gene loci that could be involved in tumorigenesis,
and in reconfirming the transformation potential of known protooncogenes (for
review see Nusse, 1986a). Examples of the first case are pim-l (Cuypers et al.,
1984), int-2, (Nusse and Varmus, 1982) and int-2 (Dickson et al., 1984). Pim-1
was identified as a common region of insertion for mink cell focus-forming (MCF)
virus or other MuLV in T cell lymphomas of several mouse strains (Cuypers et al.,
1984; Selten et al., 1985). Pim-1 encodes a serine kinase (Selten et al., 1986), and
may thus participate in cell signaling. Studies of transgenic mice carrying the pim-
1 gene have verified its role in T lymphoid neoplasia (van Lohuizen et al., 1989b;
Breuer et al., 1989a; Breuer et al., 1991). Increased expression of int-1 and int-2
as a result of MMTV insertion is thought to contribute to the formation of mouse
mammary tumors through an autocrine loop (Aaronson, 1991). Both int-1 and int-
2 are expressed temporally in a very restricted pattern during early development
of the mouse, and the protein products show the characteristics of genes
encoding secreted factors (for review see Nusse, 1988). Studies of transgenic
mice bearing activated int-1 (T sukamoto et al., 1988) or int-2 (Muller et al., 1990)
confirm their oncogenic properties. Such animals develop mammary hyperplasia,
and those with an int-1 transgene gradually develop mammary carcinomas with

age. A wide range of cellular oncogenes activated by viral integration has been

72

found in a variety of tumors and these integration events have been implicated in
the malignant phenotypes of the tumors. These include lL-3 (Morishita et al.,
1988), GM-CSF (Stocking et al., 1988), CSF-1 (Baumbach et al., 1988), c-Ha-ras
(lhle et al., 1989), c-erb B (Fung et al., 1983), c-myb (Shen-Ong et al., 1984), N-
myc ( van Lohuizen et al., 1989a), p53 (Hicks and Mowat, 1988) and, most
frequently, c-myc (Hayward et al., 1981; Neel et al., 1981; Payne et al., 1982;
Westaway et al., 1984; Swift et al., 1985; Corcoran et al., 1984; Selten et al, 1984;
Li et al., 1984; Steffen, 1984; O’Donnell et al., 1985; Neil et al., 1984).

Here we describe a locus that we have designated as bone-1. Bonc-1 is one
of several common integration sites found in 6 of 7 Balb/c B lymphoid tumor cell
lines (Chen et al., 1991) (Chapter 2). These 7 tumor cell lines were derived from
independent tumors which were collected after tumor challenge of Balb/c mice
with a cell line (R2) established after infecting Balb/c bone marrow cells with a
retroviral vector carrying the v-Ha-ras oncogene and helper Moloney murine
leukemia virus (MoMuLV) (Schwartz et al., 1986a). The long latency and infrequent
occurrence of tumor development within these animals led us to hypothesize the
involvement of secondary events in the tumorigenesis of these 7 cell lines (Chen
et al., 1991). The association of gene activation through viral integration to these
secondary events was suggested by the observation of common viral integration
sites in 6 of these 7 tumor cell lines (Chen et al., 1991). This result indicates the
outgrth of a transformed subclone from the original R2 cell population during
the period of tumor progression. Studies of the viral flanking regions may reveal
the molecular basis of the growth advantage possessed by this subclone, and,

furthermore, the putative roles of the flanking regions in oncogenesis of B lymphoid

73

tumors.

Materials and Methods:

Mice. Balb/c.lpr mice were a gift from The Jackson Laboratories (Ban Harbor,
Maine).

Cell culture. R2 and tumor cell lines were maintained on cellular feeder cultures
as described in Chen et al. (1991). Only short term cultures of two weeks or less
were used for molecular analyses to avoid accumulation of genetic alterations in
vitro.

Nucleic acid isolation. High-molecular-weight DNA was isolated from nuclei
collected as described by Schwartz et al. (1986a). Cytoplasmic RNA was isolated
by a sodium dodecyl sulfate (SDS)-urea procedure as described by Schwartz et
al. (1981). Poly A" RNA was selected by oligo-dT cellulose chromatography (Aviv
and Leder. 1975). Testicle RNA was prepared by a LiCl/urea method with a motor
driven polytron homogenizer (Auffray and Rougeon, 1980).

Molecular cloning. DNA from tumor 5 (T 5) was partially digested with Sau3A and
partially end-filled with dGTP and dATP, and then cloned into a Xhol half-site
LambdaGEM-11 vector (Promega). The ligated DNA was packaged using a
Packagene kit(Promega). The packaged DNA was titered and plated onto LB
plates using LE392 (Promega) as host. The library was screened by a env or LTR
enhancer-specific probe as described (Chen et al., 1991). Positive clones were
purified and used to prepare phage DNA (Sambrook et al., 1989). Viral flanking
regions were identified by hybridization and restriction mapping analyses. DNA

fragments which contain viral flanking regions and with appropriate restriction site

74
ends were then subcloned into pBIuescript IlKS+ (Stratagene). A Pstl DNA

fragment of bone-1 was used to "walk" through a NIH3T3 phage genomic library
(a gift from Dr. DeWitt at Michigan State University). This library was generated in
Lambda FixTM ll (Stratagene). Clones 11 and 32 derived from this library were
isolated by similar approaches to those described above. Three SacI DNA
fragments of clone 11 and clone 32, respectively, were also subcloned into the
pBIuescript KSII + vector for fine structure restriction mapping. Sequences unique
to clone 11 and clone 32 were identified by hybridization of DNA blots used in the
restriction mapping with a labelled liver DNA probe. Since 1/3 of total genomic
DNA consists of repetitive sequences, restriction fragments that hybridize strongly
to labelled liver DNA contain repetitive sequences unlikely to encode gene
products. Therefore, sequences of these restriction fragments were excluded in
northern analyses for detecting mRNA expression of bonc-1. Sequences unique
to bone-1 included the retire insert fragment of clone 11 except the two Xbal
fragments in the very ends of the insert fragment, and included the 2.2 kb SacI
fragment in the right side of the clone 32 insert fragment as shown in Figure 2.

Blots and Hybridization analyses. Southern and Northern blot hybridization
procedures were performed as described by Chen et al. (1991) except that a
oligomer random priming kit (USB) was occasionally used as the probe labelling
method. Plaque lifts were prepared as described in Sambrook et al. (1989). All
washing conditions were performed to a stringency of 0.1 x SSPE at 65° C except
in the case ofthe zoo blot where conditions are indicated in the figure legend. The
rat gcheraldehyde-S-phosphate dehydrogenase (rGAPDH) and env probes were

as described (Chen et al., 1991). The 03P2 probe was a 1.3 kb Pstl fragment

75

from the bone-1 locus (this paper).

Restriction mapping. Restriction enzyme sites in the multiple cloning regions of
the various vectors were used as reference points for mapping. Double digestion
with these enzymes and sites in the cloned DNA were performed to construct the
restriction map of bonc-l.

DNA sequenclng. The 03P2 fragment was subcloned into Mp19 for single-strand-
sequencing. Subclones which produced either (+) or (-) strand of the recombinant
M13 phage DNAa were obtained. DNA sequences were determined by the
dideoxy sequencing procedure (Sanger et al., 1977) using a sequencing kit from
USB. The predicted sequence was analyzed with the sequence analysis software
package of the University of Wisconsin Genetics Computer Group (Devereux et al.,

1986) or Genepro4.2 (Riverside Scientific, Seattle).

76
Results:

Cloning of a tumor specific viral flanking region: bonc-1

A genomic library, lambda T5, was constructed from DNA of tumor cell line 5
(T 5) to isolate tumor specific viral flanking regions. T5 was chosen because it
apparently contains the highest number of viral integration sites as judged by the
number of hybridized fragments on a southern blot hybridization analysis with a
virus-specific probe for the envelope gene of ecotropic viruses (Chen et al., 1991).
Four positive clones (A-D) were originally identified from 2 x 105 recombinant
phages. Viral flanking regions of these clones were subcloned by first identifying
the integration junction fragment and then isolating the non-viral subfragment from
the junction fragment. Subsequent analyses of viral flanking sequences of
individual clones revealed that clones A and B were derived from a viral integration
site that was already present in the R2 parental cell line (data not shown); clone
D consisted of the viral flanking region of an endogenous virus (data not shown);
and clone C contained the flanking region of a tumor-specific virus integration site
(Figure 1). Southern blot hybridization of EcoRl-digested DNA with a probe
derived from the flanking region of clone C (CSP2) demonstrated that six of the
seven tumor cell lines (lanes 3, 4, 5, 7, 8, and 9) contained a 20 kb fragment in
addition to the 11 kb germ-line fragment, whereas the R2 cell line and the tumor
4 (T 4) contained only the same germ-line fragment (lanes 2 and 6) as that in liver
(lane 1). The size difference (9 kb) of these two fragments was about the size of
a MoMuLV provirus (8.8 kb). Since EcoRI does not out within the MoMuLV
sequence, the 20 kb fragment probably resulted from an integration of an intact

virus. This locus was designated as bone-1, a putative B cell oncogene.

77

123456789

23> . ..

9.4. w ‘ " a ‘1‘? H ‘. ~4- ez.-.a ‘N g.

44'

Figure 1. Tumor specific viral integration in the bonc-1 locus. Southern blot
analysis of DNAs from liver (lane 1), R2 (lane 2) and tumor cells (lanes 3—9). DNA
was digested with EcoRI and 10 pg of each sample was electrophoresed through
0.8 % agarose. The blot was probed for bone-1. Size markers are the positions

of an ethidium bromide-stained Hindlll digest of bacteriophage l, and are denoted
in kilobases.

Expr
from 401
expressi
kb C3P£

Sinc
activate
we dec
walking.
rescree!
since Ni
donesl
clone 3:
revealec
20.5 kb
side on
and Clo
restrict“
Schlon
in FigUr.

Nor
20.5 kb

099(31in

78
Expression of bonc-1 was examined by northern blots of poly A” RNA isolated

from 400 micrograms of cytoplasmic RNA to ensure the detection of low levels of
expression. No mRNA was detectable in R2 or the tumor cell lines using the 1.3
kb CSP2 fragment as a probe (data not shown).

Since the viral LTR may function over a relatively long distance, the putative
activated gene may be located distal to the integration junction region. Therefore,
we decided to clone the distal regions of the flanking sites by chromosomal
walking. We failed to obtain positive clones from the Lambda T5 library, when we
rescreened this library with the CSP2 probe. A NIH3T3 library was used instead
since NIH Swiss mice are closely related genetically to Balb/c mice. Two different
clones were isolated. Clone 11 consisted of an 18.2 kb genomic fragment and
clone 32 consisted of a 14.5 kb genomic fragment. Restriction mapping analyses
revealed that the genomic fragments of these two overlapping clones covered a
20.5 kb region at the bcnc-1 locus, with about 10.5 kb flanking sequence at each
side of the viral integration (Figure 2). Three SacI fragments derived from clone 11
and clone 32 were subsequently cloned into plasmid vectors to perform finer
restriction mapping. The results obtained from restriction mapping these
subclones and the original clones correlated well, and yielded the map illustrated
in Figure 2.

Northern analysis was repeated with probes to the unique regions within the
20.5 kb region (Figure 2E). RNAs from different tissues were included for the

detection of transcripts. Still, no mRNA was detected.

 

 

 

 

 

A :91: ‘ I if E MoMuLV
env 1
a a
e i“ t 'i 11‘ iii if“ c
1 1 I
x a '
“ p e
C L l i ‘3
D 'I 'I can
‘4
1kb
E
cam-i 1‘11 iii? ITITIII iiiI...rm
1““ E:- ‘uh fl bent-l
s— — “algae

 

Figure 2. Cloning strategy and restriction endonuclease mapping of the bone-1
locus. (A) Map of MoMuLV. (8) Map of the clone C isolated from the lambda T5
library. (C) Map of the CS fragment This env-containing Bglll DNA fragment was
subcloned for the isolation of the flanking cellular region. (D) The Pstl fragment
(03P2) containing the flanking cellular region was subcloned for use as a
hybridization probe. (E) Map of genomic DNA from the bone-1 locus. The big
arrow represents the site of viral integration in Tumor 5. Restriction fragments
containing sequences unique to this region are indicated by filled boxes. Restriction
endonucleases: B, BamHl; E, EcoRl; G, Bglll; H, Hindlll; K, Kpnl; N, Notl; P, Pstl;
Sau, Sau3A; S, SacI; X, Xbal. '

80

Bone-1 sequence was conserved among some species

A "200" blot containing DNAs from chicken, rat, human, hamster, bovine, and
goat was used to examine the sequence conservation of the bonc-1 locus.
Surprisingly, we could detect discrete bands by hybridization with C3P2 in all of the
species after washing to a stringency of 0.2 x SSPE at 65° C (Figure 3) despite the
fact that no mRNA was detected in this region.
Sequence analyses of bone-1

The sequence conservation suggested that bone-1 might contain important
gene sequences. We therefore sequenced the C3P2 fragment. Figure 4 shows
the DNA sequence of the entire 1327 bp 03P2 fragment, with some uncertainty in
nucleotide numbers 444 and 446. This uncertainty is due to the strong pause of
the sequence reaction at this region. Several methods including the use of ITP
and temperature modification in the sequencing reactions have been used to
resolve this sequence ambiguity without success. Sequences of the CSP2
fragment were compared to those in the GeneBank and EMBL databases. No
homology was found (data not shown). When the sequence was examined for the
presence of possible open reading frames (ORFs), several possible ones were
identified (Figure 5). The longest ORF, from nucleotide 565 to 957, encoded 131
amino acids. When the protein sequences of these putative ORFs were examined
for the presence of known functional domains (motifs) such as leucine-zipper, no

known protein motifs were found (data not shown).

81

23> i, j:
9.4» ’ . ;.
66>: f 3 4,. I”, are U
. 5“? 5' W:
,3) ,
4.4»: ‘
‘1 '. a
2-3'3 .

Figure 3. bonc-1 is evolutionary conserved. Southern blot analysis of DNAs from
human (lane 1), hamster (lane 2), bovine (lane 3), goat (lane 4), chicken (lane 5),
and rat (lane 6). DNA was digested with EcoRI and 10 pg of each sample was
electrophoresed through 0.8 % agarose. The blot was probed for bone-1. Size
markers are the positions of an ethidium bromide-stained Hindlll digest of
bacteriophage l, and are denoted in kilobases.

 

82
GCCCTCGCCGGGACGACGGGCGCCTCCGGAGCCGGGGCCGCGGCCGAGGCGCCGA
CGCCGCGCGAGCTCGCCCAGACGGGGGCCCGCCCAGCCTGTCAGCGCGGCGGGAG
GCCGGCGGTGGCCGCTGCTGCCCCCTTTCTGCTCTGCCTCAGCTTCCTGCAGCCC
ACGCCACGCCCACCGCGCGCCGAGCCCGTCTCCGCCCTCCAGGGGCCGCACGCCG
CCTCCGCTCGGCCCCGCGACGCCGGCTCAGCTGCCCTGCTCGGCGGGCTCCGGCG
CGGTGCAGCTTCGGGAAAGCGGCCCCGAGCGGCGGACGGGGCGTGGGCAAGCCGG

CGGGGGTTGGCGGGGGGGCGGAGGCGGCACGCCCCACTGCGCCTGCGCGCCCGCG

GCCGCTCTGCGGGCTGGGCCGGGACGGGGAGGCGGCCGCGGGCTCCGGGGAAGCG

GAGNCNCGCGTGGAGATTCCCGGGGCAGCCCCCGCGAGAGCGCGCGAGGAGGAGG

AGGAGCGGGGCGGGTGCGGGTTGGCGCAGCGTGCTTGCGGCCTCGGCTCGCGGAG

GAGACGGCTGGGAATGAGTCAGCCCGGCCGGGAAGGCCCCGCTGCGTCCGAGCTG

ATAGGATTGGCGGCGCTGCCTCCGCGGAGACTTTCCCCTTCCTGCTTTCGTTTCC

ACAGGCGGCTCCCTGCCCTAAGCGCTCGGCCCAGGGGACGTGGCACCGTGGACCG

GGCGCTGAGACCCAAGTACCTGACTTCAGTTAACCACCACTTCTGCGAAGGGACC

GCTTGGAAGAAGAAACGACGTATAGGGTCCCAAAGAGCCACGTTCATTCATTCAA

GAATTTGTTCCCTGCTAGGCAAGGGCTCCCAGGAGAGGAGACGAAGGCAGCACCC

ATTCTGCATGTGCTATTCTCATAGAAGCCAGAAAAGAGGCTTTAAATCTGCACTG

CTCTTTGTGTGTGATTGCCAGTAACGGGTACACACACACACACAAAGGTGAGTGG

AGATGTTATTTTGAGATGACCAGAAAAGACCTCTTAGAGTGGCATTTGAGCAGAA

ATGTGAAGGATGAGAACGATCCAGTCCTGGAACTATCTGGTCAAGAAAGAAGCTA

ACAATGAAAAGGATTCTGAAAAGGGAATGTGTTTTAAGGACCAAGAAAGAACCCT

CAATGGAAGACAGTAAAAGGAAAGGGGCTAAATAAGAAGACAAAGAAATAACTGG

GGTCATTCTGTATAGGGCTTCTTATTTTTTCTCTTAGCCCCAGGTTTGTCAACCG

AAATGAAGCATTCACAACACCCACGTCCCCTTATATACGTGACGGATGCTGTTAT

TCAGATC

55

110

165

220

285

330

395

440

495

550

605

660

715

770

825

880

935

990

1045

1100

1155

1210

1265

1320

1327

Figure 4. DNA sequences of the C3P2 fragment. X indicates

undetermined nucleotides.

EKB

soan.ao. u.

on 05.3 3 Co

a 8.9: 9.53.. coco 2:.

«notes:

«98.. E .0»

.5 pea-050.3 9.2. 38......

332.3 u

use... 2.63.. coco of 432033.. admit—mu:

32:3,. Eon

8330303..

 

 

 

053s mo— mo

403633: 53.33 «2630.. Bus 3...... 3. to 329.3 < 8.6.; 3:80.. coco of
2.2! mean... 9.53.. coco 309.3 09.5 of e... u use. a .< .333 539.253 E3959. .89 p23 of
.2863 533:5 9.30.59. xon 3.3 of gone 9.3 .3223 of .23.: 3:63.. code so c233 30:02.0...
2: 695.... 9.33.. 3338 5a of acomocooc n- oca .~- .7 .n .N J .sstooad c3303”.
9:30.. coco £5.59; ~anu of co seconds... as. E «95.... 9.33.. coco 0383:... .m
1 _ _ _ __ _ _:-:.o._ '..-...o...._ _ _ ml
EBB...— __ —. =— ”U _ _ NI
_ . _ _ _ 1 _ 1.2:...— _ _ MI
3— _ _ F3438...— _ _d_ _ _ ‘_ m
.3. _ :................_,___ _ _ _ _ N
_ _ _ __ _ _ . .

 

— — —P-Ie=e.eeefle—

 

 

03.33
339.3

.ucao.

:33 one...»

o..u.uc.

~ .eodecoo e5

otsm_m

Frede
fragrr
abou1
abou1
There
(ngr
anec
fieque
that g
maDD
to Chr

D
iympr
RFLP
restric
{Nobe

ShOWr

chm

Fl9Ure

84

Bone-1 was mapped to chromosome 19 by RFLP analyses
In collaboration with Marge Strobel, Neal Copeland and Nancy Jenkins (NCl,
Frederick), the bonc-1 locus was mapped to mOuse chromosome 19 by restriction
fragment length polymorphism (RFLP) analyses. Bone-1 was mapped to a region

about ,-_16.0 t 2.7 cM from Porno-2 and 2.3 i 1.6 cM from cypzc. Pomc-2 is

H

._ about, 6.0 1- 1.7 cM from Ly-1, and CypZC is about 2.3 1.3 cM from Tdt.
Therefore, bonc-1 was 4.7 i 3.0 cM from Tdt, and 22.0 i 4.5 cM from Ly-1
(Figure 6). Two other loci were also mapped near this region: Ipr, a genetic loCus
linked to a recessive lymphoproliferatiOn'trait (Lyon and Scarle, .1989), and a
frequent virus integration site, gin-1 (Villemur et al., 1987). RFLP analyses showed
that gin-1 is not identical to bone-1 (Figure 6). Watanabe et al. (1990) have-
mapped Ipr to 6.1 cM from Tdt and 19.3 cM from Ly—44, a locus physically mapped
to chromosome 11q12-13 which is very close to Ly-t in chromosome 11513.
Due to the close linkage of bone-1 to Ipr and their common association with .
lymphoid disorders, we examined the possibility of both loci being identical by an
RFLP analysis. DNA of Balb/c.lpr mice and Balb/c were digested with 7 different
restriction enzymes, and subjected to a southern hybridization with the C3P2

probe. No DNA rearrangements were found in DNA of Balb/c.lpr mice (data not

shown).

ll-le-u
5 cM

[gnu-2
lens-l
meta-I
Ii!

 

Chromosome 19

Figure 6. RFLP mapping ofbonc-1.

)i

J!

(U

85

Discussion:

A tumor-specific viral integration site (bone-1) which may be associated with
the growth advantage of tumors derived from the R2 cell line was isolated.
Sequence homology to other species was found within this locus, significance of
which remains unknown. Similar results have been observed at other loci: evi-1,
which is a frequent viral integration site found in AKXD murine myeloid tumors
(Mucenski et al., 1988a); and dsi-t, which is also a frequent viral integration site
identified from Fisher rat thymomas (Vijaya et al., 1987). Evl-1 was later found to
encode a zinc finger protein (Morishita et al., 1988), whereas no transcripts have
been associated with dsi-1 thus far.

Although no mRNA was found over a region extending 10 kb to each side of
the site of viral integration, we cannot rule out the possibility of the activation of a
gene beyond this 20 kb region. Long distance activation of the myc
protooncogenes has been reported by Lazo et al. (1990). They found that
provirus integrations 30 kb 3’ of c-myc (mlvi-4) and 270 kb 3’ of c-myc (mlvi-1)
enhanced c-myc expression in rat T lymphomas. Further chromosomal walking
may resolve this question. it is also possible that the tumor-specific viral
integration at this locus is coincidental with another true activation event. For
example, there are still three or four tumor-specific viral integration sites remaining
to be analyzed. Nevertheless, the conservation of the bone-1 locus across species
and the presence of open reading frames leaves open the possibility that it may
represent a gene expressed at low levels, restricted to a tissue type, or restricted
temporally.

Bone-1 was mapped to mouse chromosome 19 near a previously described

86

frequent viral integration site, gin-1 (Villemur et al., 1987), and the Ipr locus which
is linked to a recessive lymphoproliferation trait (Lyon and Scarle, 1989). Results
from RFLP analyses suggest that bone-1 and gin-1 are not identical. Our results
from RFLP analyses on the DNA of Balb/c and Balb/c.lpr mice did not suggest or
exclude the possibility of bone-1 and Ipr being identical. However, because Ipr is
linked to a recessive trait and we only detected the gene rearrangement of bone-1
on one allele, they are likely to be non-identical. The fact that it is close to the Ipr
locus provides a use for this locus in the search for genes responsible for the Ipr

lymphoproliferation trait.

RE‘

86'.
mi:
IUTT

Chl

Hit
the

Ihl

87

References:

Aaronson S. (1991). Growth factors and cancer. Science 254, 1146-1153.

Auffray C. , and F. Rougeon. (1980). Purification of mouse immunoglobulin heavy—chain messenger
RNAs from total myeloma tumor RNA. Eur. J. Biochem. 107, 304-314.

Aviv l-l., and P. Leder. (1975). Purification of biologically active globulln message RNA by
chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci USA 69, 1408.

Baumbach W.R., E.M. Colston, and MD. Cole. (1988). Integration of the Balb/c ecotropic provirus
into the colony-stimulating factor-1 growth factor locus in a myc retrovirus induced murlne monocyte
tumor. J. Virol. 62, 3151-3155.

Chen S.-C., D. Redenius, and RC. Schwartz. (1991 ). Tumorigenesis of a v-Ha-ras-expressing pre-B
cell line selects for c-myc activation. Biochem. Biophy. Res. Comm. 178, 1343-1350.

Cuypers H.T., G. Selten, W. Quint, M. Zijlstra, E.R. Maandag, W. Boelens, P. van Wezenbeek, C.
Melief, and A Bems. (1984). Murine leukemia virus-induced T cell lymphomagenesis: integration of
proviruses In a distinct chromosomal region. Cell 37, 141-150.

Devereux J., P. Haeberl, and O. Smithies. (1986). A comprehensive set of sequence analysis
programs for the VAX. Nucl. Acids Res. 12, 387-395.

Dickson C., R. Smith, S. Brookers, and G. Peters. (1984). Tumorigenesis by mouse mammary tumor
virus: proviral activation of a cellular gene in the common integration region int-2. Cell 37, 529-536.

Fung Y.-K.T., W.L Louis, LB. Crittenden. and H.J. Kung. (1983). Activation of the cellular oncogene
c-erbB by LTR insertion: Molecular basis of erythroblastosls by Avian Leukosis Virus. Cell 33,
357-368.

Hayward W., E.G. Neel, and S. Astrin. (1981). Activation of a cellular onc gene by promoter insertion
in ALV-induced lymphoid leukosis. Nature 290, 475-480.

Hicks G.G.. and M. Mowat. (1988). integration of Friend Murine Leukemia Virus into both alleles of
the p53 oncogene in an erythroleukemia cell line. J. Virol. 62, 4762-4755.

lhle J.N., 8.8. White, B. Sisson, D. Parker, D.G. Blair, A Schultz, C. Kozak. R.D. Cunsford. D. Askew.
Y. Weinstein, and R.J. lsofort. (1989). Activation of c-Ha-ras protooncogenes by retrovirus insertion
and chromosomal rearrangement in a Moloney leukemia virus-induced T cell leukemia. J. Virol. 63,
2959-2966.

Lazo P.A. , J.S. Lee, and RN. Tschilis. (1990). Long-distance activation of the myc protooncogene
by provirus insertion in MIvi-1 or MIvi-4 in rat T-cell lymphoma. Proc. Natl. Acad. Sci. USA 87,
170-173.

Lyon M.F., A.G. Scarle. (1989). Genetic variants and strains of the laboratory mouse. 2nd ed. pp.
209.

Morishita K., 03. Parker, M. Mucenski, N.A. Jenkins, N.G. Copeland; and J.N. lhle. (1988). Retroviral
activation of a novel gene encoding a zinc finger protein in lL-3 dependent leukemia cell lines. Cell
59. 831 840.

 

Mucenski l
(1988a) ldi
myeloid tut

Muller W J
product ac

Nusse R. -
contain a t

Payne GS

Sambrook
Spring Har

Sanger F .,
Proc. Natl

Schwartz l
lambda ch

Schwartz i
am V-Ha-l
3221-3231

Selten G..
and A Be
homology

locus hyt
1077-1 030

Siocking (
genes inv
Williams (

TSUkamot‘
"‘9 inll
309!)er

Win 8.
'nsemons

com"ion
DNA Moi.F

Chm-”080'

88

Mucenski M.L, B.A. Taylor, J.M. lhle, J.N. Hartley, H.C. Morse, Ill, N.A. Jenkins, and MG. Copeland.
(1988a). Identification of a common ecotropic viral integration site evi-1 in the DNA of AKXD murlne
myeloid tumors. Mol. Cell. Biol. 8, 301—308.

Muller W.J., F.S. Lee, C. Dickson, G. Peters, P. Pattengae, and P. Leder. (1990). The int-2 gene
product acts as an epithelial growth factor in transgenic mice. EMBO J. 9. 907-913.

Nusse R., and HE. Varmus. (1982). Many tumors induced by the mouse mammary tumor virus
contain a provirus integrated in the same region of the host genome. Cell 31, 99-109.

Payne G.S., J.M. Adams, and HE. Varmus. (1982). Multiple arrangements of viral DNA and an
activated host oncogene in bursal lymphomas. Nature 295, 209-213.

Sambrook J., E.P. Fritsch, and T. Manlatls. (1989). Molecular cloning: A laboratory manual, Cold
Spring Harbor Laboratory . (New York: Cold Spring Harbor).

Sanger F., S. Nicklen, and AR. Coulson. (1977). DNA sequencing with chain-terminating inhibitors.
Proc. Natl. Acad. Sci. USA 74, 5463-5467.

Schwartz R.C., G.E. Sonenshein, A. Bothweli, and ML. Gefter. (1981). Multiple expression of lg
lambda chain encoding RNA species in murine plasmacytoma cells. J. immunol. 126, 2104-2108.

Schwartz R.C., LW. Stanton, S.C. Riley, K.B. Marcu, and ON. Witte. (1986a). Synergism of v-myc
and v-Ha-ras in the in vitro neoplastic progression of murlne lymphoid cells. Mol. Cell. Biol. 6,
3221-3231.

Selten 6., HT. Cuypers, W. Boelens, E. Robanus-Maandag, J. Verbeek, J. Domen, C. van Beveren
and A Bems. (1986). The primary structure of the putative oncogene pim-1 shows extensive
homology with protein kinase. Cell 46, 603-611.

ShenCng G.LC., M. Potter, J.F. Mushinski, S. Lavu, and E.P. Reddy. (1984). Activation of the myc
locus hybridization insertional mutagenesis in piasmacytoid iymphosarcomas. Science 226,
1077-1080.

Stocking C., C. L6llger, M. Kawai, S. Suciu, N. Gough, and W. Ostertag. (1988). Identification for
genes involved in growth autonomy of hematopoietic cells by analysis of factor-independent
mutants. Cell 53. 869-879.

Tsukamoto A.S., R. Grosschedl, R.C. Guzman, T. Parslow. and HE. Varmus. (1988). Expression of
the inf-1 gene in transgenic mice is associated with mammary gland hyperplasia and
adenocarcinomas in male and female mice. Cell 55, 619-625.

Vijaya S., D.L. Steffen, C. Kozak, and H.L Robinson. (1987). A region with frequent proviral
insertions in Moloney murlne leukemia virus induced rat thymomas. J. Virol. 61, 1164-1170.

Villemur R., Y. Monczak, E. Rassart, C. Kozak, and P. Jolicoeur. (1987). identification of a new
common proviral integration site in cross passage a murlne leukemia virus-induced mouse thymoma
DNA. Mol. Cell. Biol. 7, 512-522.

Watanabe T., A. Shimizu, and Y. Sakai. (1990). A molecular genetic lineage map of mouse
chromosome 18 and 19. Fourth international workshop on mouse genome mapping. Abstrct 101.

Chapter 4

IL-7 expression in a v-Ha-ras transformed pre-B cell line
is not sufficient for tumorigenicity:

differing assessments in clonal versus heterogeneous populations

Shu-Chih Chen,1 Diane Redenius,1 Judy C. Young,2

and Richard C. Schwartz‘

‘Department of Microbiology, Michigan State University,

East Lansing, MI 48824-1101

2Department of Microbiology and Molecular Genetics,
University of California-Los Angeles,

Los Angeles, CA 90024-1570

89

Abstract:

A recombinant retrovirus expressing IL-7 was superinfected into a murine
pre-B cell line previously infected with a retrovirus expressing v-Ha-ras.
Populations of cells polyclonal for integration of the iL-7 virus and independent of
iL-7 for growth were generated. lL-7 expression caused only a minimal
enhancement of tumorigenicity. Surprisingly, the introduction of v—myc expression
also had only a minimal effect. in contrast, co-infection of murine bone marrow
with retroviruses expressing iL-7 and v-Ha-ras generated clonal B lymphoid
outgrowths that were tumorigenic. We propose that although iL-7 and v-Ha-ras
may contribute to tumor progression, their introduction into a heterogeneous
population of bone marrow cells allows the selection of additional oncogenic
events occurring independently in that population. The oncogenic potential of lL-7
expression, in itself, and deregulated expression of other genes is probably best

assessed in well-characterized individual cell lines.

91

Introduction:

The neoplastic transformation of hematopoietic cells is generally accepted
to be a multi-step process (Schwartz and Witte, 1988). This process can involve
the overexpression of hematopoietic growth factors, as well as events involving
other classes of oncogenes (Sawyers et al., 1991). For example, a translocation
activating lL-3 expression has been observed in a form of human acute pre-B
leukemia (Meeker et al., 1990). Overexpression of lL-3 by itself has been shown
to cause a myeloproliferative disorder rather than leukemia (Chang et al., 1989;
Wong et al., 1989) suggesting that its role in leukemogenesis requires cooperation
with other events. iL-7, a cytokine for pre-B cells (Namen et al., 1988) and T cells
(Morrissey et al., 1989), has recently been examined for its role in the
transformation of pre-B cells. Young et al. (1991) found that hyperexpression of
lL-7 in a pre-B cell line was neither necessary nor sufficient for neoplastic
transformation. Although the cell line’s dependence upon exogenous iL-7 was
alleviated, its growth in soft-agar medium was not enhanced and only one of six
IL-7-expressing derivatives of the cell line was tumorigenic. In contrast, Overell et
al. (1991) found that a pre-B cell line made growth factor-independent by
hyperexpression of lL-7 was tumorigenic. However, their data indicated that events
in addition to lL-7 expression were required for factor independence. These
seemingly contradictory findings leave unresolved the potential for autocrine lL-7
expression to contribute to the tumor progression of pre-B cells. On the other
hand, both studies suggest that hyperexpression of lL-7 may have a role in the
cooperative transformation of pre-B cells with other oncogenes.

in previous studies, we found that v-Ha-ras-expressing pre-B lymphoid cells

 

92

displayed an intermediate transformed phenotype and were infrequeme
tumorigenic (Schwartz et al., 1986 a, b). The tumors that did arise were not
dependent upon IL-7 for growth, unlike the pre-B cells from which they arose
(unpublished observations, Chen and Schwartz). This led us to investigate
whether high level autocrine expression of lL-7 would be sufficient for the
tumorigenicity of a v-Ha-ras-expressing pre-B cell line. Since we had previously
demonstrated the cooperativity of v-myc and v-Ha-ras in pre-B cell transformation
(Schwartz et al., 1986 a, b) and a role for myc activation in the tumor progression
of a v-Ha-ras-expressing pre-B cell line (Chen et al., 1991), we directly compared
the oncogenic potential of lL-7 and v-myc in a v—Ha—ras-expressing pro-B cell line.
While expression of lL-7 may partially alleviate dependence on exogenous growth
factors for growth, it is clearly not sufficient for tumorigenesis. Curiously, we have
found a disparity for both lL-7 and v-myc between the consequences of
introducing their expression into a v-Ha-ras-expressing cell line and the
simultaneous introduction of these oncogenes by co-infection of primary bone
marrow cells. Pre-B cell lines derived by co-infection of bone marrow were fully
tumorigenic (see results reported here and Schwartz et al., 1986a), while lines
derived by sequential addition to a cell line were not. Neither lL-7 nor v-myc in
combination with v-Ha-ras are truly sufficient in and of themselves for cooperative

transformation of pre-B cells.

93
Results:

Superlnfection of a v-Ha-ras-expressing pre-B cell line

in order to test the oncogenic potential of IL-7 expression in cooperation
with v-Ha-ras expression, we generated a v-Ha-ras-expressing cell line that could
easily be superinfected by a retrovirus carrying the gene for IL-7. Fresh murine
bone marrow was infected with a helper-free stock of SV(X)-Ha-ras, a v-Ha-ras-
expressing retrovirus (Schwartz et al., 1986a), and the bone marrow was placed
under long-term B cell culture conditions (Whitlock and Witte, 1982). A clonal
pre-B cell outgrth designated w2R4 resulted from this infection. This cell line
expresses v-Ha—ras from a single proviral integration site and was found to be
dependent on a cellular feeder layer or IL-7 for growth, unable to grow in soft agar
medium, and not to be tumorigenic (data not shown). ¢2R4 was subjected to
three different retroviral infections:

1. An lL-7-expressing retrovirus, IL-7(SH) (Young et al., 1991) and Moloney
murine leukemia virus (MoMuLV).

2. A v-myc-expressing retrovirus, MMCV-neo (Wagner et al., 1985) and
MoMuLV.

3. MoMuLV.

Three weeks post infection, DNA and RNA were isolated from the infected
cultures and analyzed to assess infection and integration of the above mentioned
viruses and to verify expression of v-Ha-ras, iL-7 and v-myc in the infected cells.
Southern blots of DNAs isolated from two cultures each of MoMuLV, “IL-7" and
“v-myc" infected cells were hybridized with probes for iL-7 (Figure 1A and B),

v-myc (Figure 1C and D) and v-Ha-ras (Figure 1E). Hindlil digestion (Figure 1A)

Figure 1. Proviral integration. DNA (10 pg) was digested with the indicated
restriction enzymes, electrophoresed through 0.8% agarose, and transferred to
nylon membranes. Lane 1, NiH3T3/lL-7(SH); lane 2, $2R4/MoMuLV-1; lane 3,
w2R4/MoMuLV-2; lane 4, w2R4/Myc-1; lane 5, w2R4/Myc-2; lane 6, t2R4/lL7—1;
lane 7, w2R4/lL7—2; lane 8, RM2. Size markers are the positions of an ethidium
bromide-stained Hindlli digest of bacteriophage lambda DNA. (A) HindIII-digested
DNA was hybridized with an IL—7 probe. The position of the retroviral iL-7 gene
is marked at 0.47 kb. (8) BamHI-digested DNA was hybridized with an IL-7 probe.
(C) BamHi-digested DNA was hybridized with a v—myc probe. The position of the
retroviral v—myc gene is marked at 2.5 kb. (D) Hindlii-digested DNA was
hybridized with a v-myc probe. (E) EcoRl-digested DNA was hybridized with a
v-Ha-ras probe. The position of the v—Ha—ras integration-related restriction
fragment is marked at 7 kb.

95

Figure 1. Proviral integrations.

 

        

A: IL'7/Hll‘ld III C: MYC/Baml-l I
1 2 3 4 5 s 7 a
23»: I ‘5 " “ i
9.4}

6» -u'L-.'.g,-.i

D: MYC/Hlnd III
2 3 4 5

(I ‘1' .

678

   

96
revealed the diagnostic viral 0.47 kb lL-7-specific restriction fragment in both

iL-7-infected cultures (lanes 6 and 7), as well as in the NIH3T3 cell line that
produced the iL-7 virus (lane 1). The intensity of hybridization to this fragment is
as great in the infected cultures as it is in the NiH3T3 virus-producing cell line,
suggesting a high proportion of infected cells. The other restriction fragments
represent the endogenous lL-7 gene. BamHl digestion, which cleaves once within
the IL-7 virus genome, was used to assess the number of viral integrations and the
clonaiity of the iL—7-infected populations. This analysis (Figure 1B) revealed the
lL-7-infected cultures to contain multiple viral integration-related restriction
fragments of less than haploid abundance (lanes 6 and 7) indicating polyclonal
populations of IL-7-infected w2R4 pre-B cells. All cells possessed two iL-7-specific
restriction fragments of “'4 kb and “18 kb. Rehybridization of the BamHl-digested
DNAs with a v-myc probe (Figure 10) revealed a 2.5 kb restriction fragment
diagnostic of v-myc infection in those infected populations (lanes 4 and 5), as well
as in RM2 (lane 8), a v-Ha-ras/v-myc transforrnant previously characterized
(Schwartz et al., 1986a). The restriction fragment of ~6 kb represents endogenous
c-myc. The v-myc infected populations were evaluated for viral integration by
rehybridization of the HindiIl-digested DNAs with a probe for v-myc (Figure 1D).
Hindlli cleaves the v-myc viral genome only once. Similarly to the case of iL-7
infection, the v-myc-infected cultures (lanes 4 and 5) were shown to be polyclonal
with respect to v-myc viral integration with a smear of v-myc-specific restriction
fragments as opposed to the single viral fragment in the clonal RM2 cell line (lane
8). The "5 kb restriction fragment represents endogenous c-myc. in order to

verify that the outgrowths in the infected cultures were all derived from ¢2R4, the

97

v-Ha-ras viral integration sites were compared among the cultures. An EcoRl
digestion cleaves once within the viral genome to yield integration site specific
fragments. All of the infected cultures contained the same size v-Ha-ras-related
restriction fragment of ~7 kb confirming their derivation from ¢2R4 (Figure 1E,
lanes 2-7). The ”20 kb restriction fragment represents the endogenous c-Ha-ras
gene.

Cytoplasmic RNAs of the infected ¢2R4 cells were examined by Northern
hybridization analysis. The expected genome-length retroviral transcript of "5.4 kb
was observed in all the populations with a probe for v-Ha-ras (Figure 2A).
Rehybridization with an IL-7 probe revealed the expected 4.0 kb genome-length
retroviral transcript in both iL-7-infected populations (Figure 2B, lanes 5 and 6) and
a v—myc probe showed the expected genomic "8.0 kb and subgenomic "5.5 kb
RNAs in both v-myc-infected populations (Figure 2C, lanes 3 and 4).
Rehybridization to a probe for rat giyceraldehyde phosphate dehydrogenase
(GAPDH) verified that similar quantities of RNA were analyzed among the various
infected populations (Figure 2D). Therefore the levels of mRNA expression for
v-Ha-ras, v-myc and lL-7 were roughly equivalent between populations infected

with viruses expressing those genes.

The superinfected o2R4 populations consist of pro-B cells

Given that the activity of lL-7 within the B lineage is restricted to pre-B cells,
it was important to confirm that the superinfected cells had maintained the pre-B
phenotype of ¢2R4. Southern hybridization analysis of EcoRI-digested DNAs

revealed that all of the populations were identically rearranged in their mu heavy

FeruFLICleraHdli.

1 2 3 4 5 6

A: RAS “fiuflnb 42885'

 

e: lL-7 i
l

(x IWYC

 

E. El ‘__55
"A 9 ‘ ; 428$

‘w‘w-Hi urn-V4 mwmmvv wt}

'Huflufl

HALAszJ EL: 31""V‘h‘

    

«185

o: GAPDH .__15

 

 

Figure 2. Retroviral transcription. Cytoplasmic RNA (20 pg) was denatured,
electrophoresed in a 1% agarose-formaldehyde gel, and transferred to a nylon
membrane. Lane 1, w2R4/MoMuLV-1; lane 2, ¢2R4/MoMuLV-2; lane 3,
1|:2R4/Myc-1; lane 4, 1|:2R4/Myc-2; lane 5, w2R4/lL7-1; lane 6, w2R4/IL-2. The
positions of ethidium bromide-stained 28S and 183 rRNAs are marked. (A)
Hybridization with a v-Ha-ras probe. The position of the SV(X)-Ha-ras
genome-length RNA is marked at 5.4 kb. (B) Hybridization with an lL—7 probe.
The position of the IL-7(SH) genome-length RNA is marked at 4.0 kb. (C)
Hybridization with a v—myc probe. The positions of the MMCV-neo genome-length
and subgenomic v-myc RNAs are marked at 8.0 and 5.5 kb, respectively. (D)
Hybridization with a GAPDH probe. The position of GAPDH mRNA is marked at
1.5 kb.

Sti
en
Wt

wit!

Con

99

chain locus (Figure 3A), again verifying that all of the cells were derived from 1p 2R4.
Analysis of BamHI-digested DNAs revealed all of the populations to have agermline
configuration in the kappa light chain locus (Figure 38) confirming a pre-B
phenotype. This confirmed that the cells should indeed retain responsiveness to

lL-7.

lL-7 and v-myc infected populations show enhanced growth factor
independence

Transformation in a complex culture system, such as that employed here
where there are both nonadherent lymphoid cells and adherent stromal cells,
requires one to distinguish between effects mediated by events occurring in the
lymphoid cells versus those occurring in the stromal layer. To that end, the
infected populations were examined for growth in liquid culture without an adherent
stromal layer (T able 1). While the MoMuLV-infected controls showed no growth,
or rather indolent growth, at plating densities up to 5x104 cells/ml, the IL-7- and the
v-myc-infected populations displayed a 3 to 8 fold increase in growth over the
MoMuLV infected control when seeded at 5 x 10‘ cells/ml. At 10’5 cells/ml, the
differences in growth were not nearly as apparent suggesting the autocrine
stimulation of growth factors other than lL—7. lL-7-infected cells do not show
enhanced growth in soft agar medium, while v—myc-infected cells do (T able 2).
When plated in soft agar medium over a cellular feeder layer, populations infected
with iL-7 had a plating efficiency of about 0.1% (Table 2). This is actually less than
the plating efficiency of the MoMuLV-infected control populations, about 0.2%. In

contrast, v-myc infection elevated the plating efficiency of it: 2R4 about 5-foid to 1%.

A:.iH

1234567
23»-

94>
6.6» .

4.4>'

   

 

-’. . . -' 1 n
' . . . . -. 1'-
.. $3- u-.. 1-3.. . «M

Figure 3. Immune loci define a pre-B phenotype. DNA (10 ng) was digested with
the indicated restriction enzymes, electrophoresed through 0.8% agarose, and
transferred to nylon membranes. Lane 1, fibroblasts; lane 2, i|r2R4/MoMuLV-1;
lane 3, w2R4/MOMuLV-2; lane 4, 1|r2R4/Myc-1; lane 5, w2R4/Myc-2; lane 6,
w2R4/lL7—1; lane 7, ¢2R4/|L7-2. Size markers are the positions of an ethidium
bromide-stained Hindlll digest of bacteriophage lambda DNA. (A) EcoRl-digested
DNA was hybridized with a probe for heavy chain J regions. (B) BamHl-digested
DNA was hybridized with a probe for the kappa light chain constant region.

Table 1.

Cell population

RM2

t 2R4/MoMuLV-1
it 2R4/MoMuLV-2
t 2R4/Myc-1

1r 264/Myc-2

t 2R4/IL7-1

t 2R4/IL7-2
RIL7.1

RIL7.2

101

Growth without cellular feeder layers.

 

 

initial Density13
5 x 103/ml 1 x 104/ml 5 x 104/mi 1 x105/ml
Density on Day 7
2.2 t 0.1(106) 4.0 a: 0.2(106) 6.6 i 0.500“) 7.4 i 1.000")
< 103 < 103 5.3 : 0400‘) 2.0 t 0000“)
< 103 1.4 : 0.0(103) 6.0 : 0.5(105) 2.1 : 0.300“)
1.0 : 0.0(105) 4.5 i 0500‘) 2.0 : 0.000“) 3.6 i 0.3(106)
2.6 : 2.6(103) 2.1 : 0.7(103) 1.7 .1: 0.2(105) 3.6 x 0000“)
< 103 7.7 :t 35003) 2.3 1' 0.300“) 4.2 1 0.200")
1.0 t 0.8(103) 3.6 .4.- 0.0(103) 4.2 i 0.0(105) 5.7 : 0.5(106)
NDb 2.6 : 0.0(103) ND 2.0 : 0.2(105)
ND 2.0 : 0.2(105) ND 1.6 : 0.1(105)

 

3Cells were plated onto 6 cm culture dishes in 4 ml media. On day 4, each dish received 2 ml fresh
medium. Values are the average (with the range) of duplicate cultures.

 

 

 

bNot done.

Table 2. Soft agar growth and tumorigenicity.

Cell population % Cloning efficiency Animals with tumors/ Average days until
in soft agar animals tested dead or moribund

RM2 >5 13/14° 25

t2R4/MoMuLV-1 0.26 0/6b -

t2R4/MoMuLV-2 0.20 0/6b --

t2R4/Myc-1 1.06 1 /8b 42

1: 2R4/Myc-2 0.64 1 /6'° 73

t2R4/lL7-1 0.09 3/6" 36

t2R4/IL7-2 0.10 0/6b -

RIL7.1 ND“ 3/4" 53

RIL7.2 ND 7/8d 40

“Not done.

bThese values have no statistically significant differences among them by Fisher's exact test (p s

0.05).

°This value is statistically different from those in b by the same test in b.
dThese two values are combined, and the resulting value is statistically different from those in b by
the same test in b.

102
A positive control of RM2 cells had a plating efficiency greater than 5%. Curiously,

cells produced by the introduction of v-myc expression subsequent to v-Ha-ras
expression have poorer plating efficiencies in soft agar than cells produced

byco-infection of primary bone marrow cells (is. RM2; Schwartz et al., 19866)

The lL-7 and v-myc-infected populations are not tumorigenic

The abilities of the superinfected 1|:2R4 populations to form tumors In vivo
was tested by intraperitoneal injection of these cells into syngeneic BALB/c mice.
While positive control RM2 cells formed tumors quite consistently, both the v-myc-
and iL—7-infected populations exhibited no statistically significant difference in
tumorigenicity from the MoMuLV-infected 1|:2R4 cells with Fisher’s exact tests
(Table 2). These results suggested that neither lL-7 nor v-myc were not capable

of cooperating with v-Ha-ras in the induction of B cell tumors.

Co-infectlon of primary bone marrow cells with lL-7 and v-Ha-ras retrovirus
yields clonal outgrowths that are tumorigenic

The surprisingly low tumorigenicity of v-myc-infected populations of a
v—Ha—ras-expressing pre-B cell contrasts sharply with our earlier observation of
highly tumorigenic cell lines generated by the co-infection of primary bone marrow
cells with v-Ha-ras and v-myc retroviruses 0.6. RM2; Schwartz et al., 19866). We
therefore decided to examine the effects of co-infecting freshly explanted murine
bone marrow cells with v-Ha-ras and IL-7 retroviruses. Bone marrow cells from
BALB/c mice were infected singly with the lL-7 retrovirus, the v—Ha—ras retrovirus

or doubly infected with both retroviruses. These cells were then plated in liquid

103
culture by the procedure of Whitlock and Witte (1982). In 20 co-infections, only

two cell lines containing both lL-7 and v—Ha—ras retroviral integrations were
obtained (RIL7.1 and RIL7.2) after about 2 months in culture. Other outgrowths
contained only the v-Ha-ras retrovirus. Single infections with the lL-7 retrovirus
never resulted in outgrowths containing that virus, while single infections with the
v-Ha-ras retrovirus consistently produced v-Ha-ras-expressing cell lines similar to
those previously reported (data not shown; Schwartz et al., 19866). Southern blot
analysis of DNAs isolated from RIL7.1 and RIL7.2 revealed these outgrowths to be
clonal, unlike the isolates derived by superlnfection of1p2R4. BamHl digestion
(Figure 4A) revealed one iL-7 viral integration-related restriction fragment for RIL7.1
(lane 1) and two integration-related fragments for RIL7.2 (lane 2). DNA from
uninfected (lane 3) and infected cells possessed two endogenous lL-7-specific
restriction fragments of “4 kb and ~18 kb. EcoRl digestion of the DNAs (Figure
48) showed one v-Ha-ras-specific viral integration-related restriction fragment for
RIL7.1 (lane 1) and two integration-related fragments for RIL7.2 (lane 2) in addition
to the endogenous c-Ha-ras fragment of "20 kb. Subcloning of RIL7.2 in soft agar
medium demonstrated that the multiple viral restriction fragments represent multiple
integrations in a single clone rather than several clones (data not shown).
Northern analyses detected v-Ha-ras mRNA in both RIL7.1 and RIL7.2 (Figure 5A,
lanes 1 and 2), but detected iL-7 mRNA only in RIL7.2 (Figure 58, lane 2).
Southern blot and Northern blot analyses showed RIL7.1 to be a B cell rearranged
in both its mu and kappa loci, and expressing the kappa locus (data not shown).
RIL7.2 was shown to be a pre-B cell with germline configuration of its kappa DNA

(data not shown). Perhaps the progression of RIL7.1 to a B cell phenotype

104

dependent upon factors other than IL-7 for growth eliminated any selective
pressure to maintain IL-7 expression in that cell line. RIL7.1 showed only indolent
growth when plated without a feeder layer at 105 cells/ml (T able 1). Not
surprisingly, RIL7.2, which expresses lL-7 mRNA, was growth factor independent
(T able 1).

Analysis of the tumorigenicity of RIL7.1 and RIL7.2 revealed these cell lines
to be more obviously transformed than populations of ¢2R4 superinfected with the
lL-7 retrovirus. Similar to cell lines produced by the co-infection of bone marrow
with v-myc and v—Ha—ras retroviruses (i.e. RM2; Schwartz et al., 19866), the two
cell lines produced by co-infection with IL-7 and v-Ha-ras retroviruses were highly
tumorigenic (T able 2). RIL7.1 produced tumors in three of four animals challenged

and RIL7.2 produced tumors in seven of eight animals.

105

A:lL-7 BzRAS
_1 2 3 1 2 3

.‘4

 

Figure 4. Proviral integration. DNA (10 pg) was digested with the indicated
restriction enzymes, electrophoresed through 0.8% agarose, and transferred to
nylon membranes. Lane 1, RIL7.1; lane 2, RIL7.2; lane 3, uninfected cells. Size
markers are the positions of an ethidium bromide-stained Hindill digest of
bacteriophage lambda DNA. (A) BamHl-digested DNA was hybridized with an IL-7
probe. (B) EcoRl-digested DNA was hybridized with a v-Ha-ras probe.

106

 

Figure 5. Retroviral transcription. Cytoplasmic RNA (20 pg) was denatured,
electrophoresed in a 1% agarose-formaldehyde gel, and transferred to a nylon
membrane. Lane 1, RIL7.2; lane 2, RIL7.1; lane 3, NIH3T3/iL-7(SH); lane 4,
NIH3T3/SV(X)-Ha-ras. (A) Hybridization with a v-Ha-ras probe. (B) Hybridization
with an lL-7 probe. (C) Hybridization with a GAPDH probe.

107

Discussion:

The experiments presented here find autocrine expression of iL-7 insufficient
to cause tumorigenicity of a v-Ha-ras-expressing pre-B cell line. This contrasts with
previous studies by Overell et al. (1991) that found a pre-B cell line made growth
factor-independent by hyperexpression of IL-7 to be tumorigenic. Our results are
more consistent with those of Young et al. (1991) that found a pre-B cell line
infected with an lL-7-expressing retrovirus to be converted to tumorigenicity only
rarely.

While it may capable of enhancing growth factor independence of a v-Ha-ras
transformant, lL-7 expression had no direct correspondence to tumorigenicity. The
rare tumors recovered had a long latency suggesting that ultimately other events
were required for tumorigenicity. The lack of lL-7 dependence in tumors derived
from v-Ha-ras-transformed pre-B cells is likely to be a consequence of tumor
progression rather than its cause.

Kremer et al. (1991) have recently found c-src and c-ras to be downstream
effectors of signal transduction by nerve growth factor and fibroblast growth factor.
if the effects of both lL-7 and v-Ha-ras expression were upon the same signal
transduction pathway, cooperativity in transformation might not be expected.
Perhaps was is a downstream effector of lL-7, as well as nerve growth factor and
fibroblast growth factor.

Overell et al. (1991) found the induction of IL-7 independence in a pre-B cell
line to be tumorigenic. However, very few of the cells infected with their
lL-7-expressing retrovirus actually attained lL-7 independence. Thus, it seems that

in their system, too, lL-7 expression is not sufficient for tumorigenicity. Other

108
events are required. Overell et al. (1991) do find that IL-7 independence (as

opposed to lL-7 expression) does correlate with tumorigenicity and this result is
more difficult to reconcile with our findings. This difference probably lies in the
intrinsic properties of the individual cell lines used in the several studies. The
v-Ha-ras-transformant used in our studies may require growth factors in addition
to iL-7 for optimal growth. A similar v-Ha-ras-expressing pre-B cell line, as well as
the cell line used by Young et al. (1991), requires a non-lL-7 factor(s) released by
bone marrow stromal cells for Optimal growth (Mulrhead et al., 1990). The cell line
used by Overell et al. (1991) may have more simple growth factor requirements,
having been selected for its utility in lL-7 assays.

Comparing lL-7 directly to v-myc, we found that neither was able to confer
tumorigenicity to ¢2R4, although v-myc expression enhanced the ability to grow
in soft agar medium. This result was surprising in light of our earlier studies
(Schwartz et al., 19866; 1986b), where the in vitro co-infection of bone marrow with
retroviruses expressing v-myc and v-Ha-ras yielded fully tumorigenic cell lines.
When we performed co-infections of bone marrow cells with lL-7 and v-Ha-ras-
expressing retroviruses, tumorigenic cell lines were similarly recovered. The
co-infection of primary bone marrow cells seems to yield more transformed cell
lines than does sequential addition of the same oncogenes to a cell line. The fact
that the outgrowths from co-infection of a heterogeneous population were clonal,
while those produced in sequential infection of a cell line were polyclonal may point
towards an explanation. The clonal outgrowths of co-infection may represent the
dominance of a rare and more highly transformed cell over many other cells

carrying one or both oncogenes. The fact that only two of 20 cultures contained

109
co-infected clones and that their outgrth required 2 months may reflect the

necessity for other oncogenic events. While lL-7 and v-Ha-ras expression may
contribute to the transformed phenotype, the cooperation of other rare oncogenic
events is critical to the tumorigenic phenotype. in the superinfection of1|12R4 cells,
the more homogeneous population offers less opportunity for detecting rare
transforming events. in addition, superinfected populations were recovered more
rapidly (3 weeks). The sequential introduction of oncogenes into clonal cell lines
seems to provide 6 more accurate assessment of their contribution to cooperative

transformation than co-infection of a heterogeneous population.

Materials and methods:

Viruses. lL-7(SH) (Young et al., 1991) is derived from the Moloney murine
sarcoma virus vector, pMV6tKneo (Kirschmeier et al., 1988). A cDNA for murine
lL-7 that lacks translation-inhibiting flanking regions is expressed from the viral long
terminal repeat (LTR), as well as the Tn5 neo gene (G418-resistant) from the
herpesvirus thymidine kinase promoter.

MMCV-neo (Wagner et al., 1985) is derived from Moloney murine leukemia
virus (MoMuLV) and Harvey murine sarcoma virus. it expresses the v-myc gene
of retrovirus OK10 from the viral LTR as a subgenomic mRNA and the Tn5 neo
gene from the herpesvirus thymidine kinase promoter.

SV(X)-Ha-ras (Schwartz et al., 19866) is derived from the MoMuLV vector,
leP-NEOSV(X)1 (Cepko et al., 1984). it expresses v-Ha-ras from the viral LTR
and the Tn5 neo gene as a subgenomic mRNA from the viral LTR.

Virus stocks consisted of culture medium from NIH3T3 cell lines that had

110
been cotransfected with a proviral clone of MoMuLV and lL-7(SH), MMCV-neo, or

SV(X)-Ha-ras. For the helper-free SV(X)-Ha-ras stock, culture medium was
collected from the 1112 cell line (Mann et al., 1983) that had been transfected with
SV(X)-Ha-ras. Virus titers of 2x105 to 8x105 G418-resistant colonies per ml were
obtained for lL-7(SH); 2.4x106 for MMCV-neo; 1.0x105 to 1.5x105 for SV(X)-Ha-ras.
Helper-free SV(X)-Ha-ras had a titer of 5x104 G418-resistant colonies per ml.
Cell culture and viral Infections. Bone marrow from 3— to 5-week old BALB/c
mice was cultured by the procedure of Whitlock and Witte (1982) with the addition
of a viral infection step. Bone marrow was suspended at 2.0x10° cells per ml in
RPMI 1640 medium supplemented with 5% fetal calf serum and 5x10"5 M
2-mercaptoethanol. To this was added an equal volume of virus stock and
Poiybrene (Sigma Chemical Co., St Louis, MO) to a final concentration of 8 pg /mi.
After 3 hours of incubation at 37° C, the cells were pelleted and resuspended at
106 cells per ml in RPMI 1640 supplemented with 5% fetal calf serum and 5x105
M 2 mercaptoethanol. 5 ml of this suspension was plated per 6-cm culture dish.

1|:2R4 cells were cultured in RPMI 1640 supplemented with 5% fetal calf
serum and 5x10‘5 M 2-mercaptoethanoi over a feeder culture of adherent bone
marrow cells (Whitlock et al., 1983). For superinfection, these cells were
suspended in the same medium at 105 cells per ml and mixed with an equal
volume of virus stock and Poiybrene as described above.

All cultures were fed twice weekly with RPMI 1640 supplemented with 5%
fetal calf serum and 5x10'5M 2-mercaptoethanol. Once a week approximately 80%
of the spent medium was replaced with fresh medium. The cultures were

expanded by transfer of nonadherent cells to feeder cultures (Whitlock et al.,

111

1983). Growth in soft agar medium was performed over a feeder culture as
described by Whitlock et al. (1983).

Nucleic acid isolation and analysis. Cytoplasmic RNA was isolated by a sodium
dodecyl sulfate-urea procedure as described by Schwartz et al. (1981). High
molecular weight DNA was isolated from nuclei collected in the preceding
procedure by a method described in Schwartz et al. (19866).

Restriction enzyme-digested DNAs were electrophoresed through 0.8%
agarose. RNAs were electrophoresed through 1% agarose-formaldehyde gels
(Rave et al., 1979). Transfer and hybridizations were washed to a stringency of 0.1
x SSPE in 0.1% SDS.

Hybridization probes were prepared by random priming using a kit from
United States Biochemical Corp. (Cleveland, OH) with the incorporation of
5’-[a-32P]dATP (3,000 ci/mmol; lCN, Costa Mesa, CA). The lL-7 probe was a 0.5
kb Pstl fragment of the murine IL-7 cDNA (Young et al., 1991). The v-Ha-ras probe
was a 0.46 kb EcoRI fragment corresponding to v-Ha-ras-encoding sequences of
Harvey murine sarcoma virus (Ellis et al., 1980). The v-myc probe was a 0.8 kb
CIal-BamHI fragment corresponding to 3’-terminal v-myc-encoding sequences of
retrovirus OK10 (Hayflick et al., 1985). The mu heavy chain probe was a genomic
1.9 kb BamH1-EcoRl fragment, which corresponds to the JH2, JH3, and JH4
regions that are 5’ to the mu heavy chain constant region gene (Early et al., 1980).
The kappa light chain probe was a genomic 0.48 kb Hpal-Bglll fragment extending
from a point about 50 bp within the 5’- terminus of the kappa light chain constant
region gene to the 3’-terminal poly(A) addition site (Seidman and Leder. 1978).

The GAPDH probe was a 1.3 kb cDNA (Fort et al., 1985).

112

Tumor challenges. Cells were washed twice in RPMI 1640 and were then
resuspended in the same at 8x106 cells per ml. BALB/c mice, 3 to 5 weeks old,
were injected intraperitonealiy with 0.25 ml of the cellular suspension. Animals
were observed for a maximum of 73 days post injection. Animals were sacrificed
and autopsied when they became moribund or at 10 weeks. Tumors were verified
as being derived from the challenging cell lines by Southern blot analysis of

retroviral integration sites (data not shown).

Acknowledgments:
This work was supported by Grant CA45360 from the National Cancer
Institute. We thank Dr. Owen Witte for helpful discussions and critical review of this

manuscript.

1 13
References:
Cepko, C.L, Roberts, B.E. & Mulligan, RC. (1984). Cell, 37, 1053-1062.
Chang, J.M., Metcalf, D., Lang, R.A., Gonda, R.J. 81 Johnson, GR. (1989). Blood, 73, 1487-1497.

Chen, S.-C., Redenius, D. & Schwartz, RC. (1991). Biochem. Biophys. Res. Commun., 178,
1343-1350.

Early, P., Huang, H., Davis, M., Calame, K. 8: Hood, L (1980). Cell, 19, 981-992.

Ellis, R.W., DeFeo, D., Maryak, J.M., Young, HA, Shih, T.Y., Chang, E.H., Lowy,
DR. 81 Scolnick, EM. (1980). J. Virol., 36, 408-420.

Fort, P., Marty, L., Piechaczyk, M., El Salrouty, S., Dani, C., Jeanteur, J. & Blanchard, J.M. (1985).
Nucl. Acids Res, 13, 1431-1442.

Hayfiick, J. Seeburg, P.H., Ohlsson, R., Pfeifer-Ohlsson, 8., Watson, 0., Papas, T. 81 Duesberg, PH.
(1985). Proc. Natl. Acad. Sci. USA. 82, 2718-2722.

Kirschmeier, P.T., Housey, G.M., Johnson, MD, Perkins, AS. 81 Weinstein, LB. (1988). DNA, 7,
219-225.

Kremer, N. E., D’Arcangeio G., Thomas S.M., DeMarco M., Brugge J.S., and Halegoua S. (1991).
J. Cell Biol. 115, 809-819.

Mann, R., Mulligan, R.C. & Baltimore, D. (1983). Cell, 33, 153-159.

Meeker, T.C., Hardy, D., Willman, C., Hogan, T. 8: Abrams, J. (1990). Blood, 76, 285-289.
Morrissey, P.J., Goodwin, R.G., Nordan, R.P., Anderson, 0., Grabstein, KH., Cosman, 0, Sims. J.,
Lupton, 8., Acres, B., Reed, S.G., Mochizukl. D., Eisenman, J., Conion, P.J. 81 Namen, AE. (1989).
J. Exp. Med., 169, 707-716.

Muirhead, M., Davis, R., Schwartz, R.C., Waldschmidt, T.J., Ackerrnann, L, Palumbo, G. 8 Smith,
R.G. (1990). intl. J. Cell Cloning, 8, 392—408.

Namen, AE., Lupton, 8., Hjerrild, K., Wignall, J., Mochizukl. D.Y., Schmierer, A., Mosby, B., March,
C.J., Urdal, D., Gillis, 8., Cosman, D. 8: Goodwin, RG. (1988). Nature, 333, 571-573.

Overell, R.W., Clark, L, Lynch, 0., Jerzy, R., Schmierer, A., Weisser, KE., Namen, A.E. 81 Goodwin,
R.G. (1991). Moi. Cell. Biol, 11, 1590-1597.

Rave, N., Crkvenjakov, R. & Boedtker, H. (1979). Nucl. Acids Res, 6, 3559-3567.
Sawyers, C.L, Denny, CT. 81 Witte, ON. (1991). Cell, 64, 337-350.

Schwartz, R.C., Sonenshein, G.E., Bothweli, A. & Gefter, ML (1981). J. lmmunol., 126, 2104-2108.

Schwartz, R.C., Stanton, LW., Riley, S.C., Marcu, KB. 81 Witte, O.N. (19866). Mol. Cell. Biol., 6,
3221-3231.

Schwartz, R.C., Stanton, LW., Marcu, KB. 8: Witte, O.N. (1986b). Curr. Topics Micro. Immunol., 132,
75-80.

Schwartz, R.C. & Witte, ON. (1988). Mutation Res., 195, 245-253.

1 14
Seidman, JG. 8: Leder. P. (1978). Nature, 276, 790-795.
Wagner, E.F., Vanek, M. & Vennstrom, B. (1985). EMBO J., 4, 663-666.
Whitlock, C.A. & Witte, ON. (1982). Proc. Natl. Acad. Sci. USA, 79, 3608-3612.
Whitlock, C.A., Ziegler, S.F., Trieman, L.J., Stafford, J.|. & Witte, ON. (1983). Cell, 32, 903-911.

Wong, P.M.C., Chung, S., Dunbar, C.E., Bodine, D.M., Ruscetti, SR. 8: Nelnhuis, AW. (1989). Mol.
Cell. Biol., 9, 797-808.

Young, J.C., Gishizky, M.L. & Witte, ON. (1991). Mol. Cell. Biol., 11, 854-863.

APPENDICES

APPENDIX A
Lineage Switch Macrophages Can Present Antigen‘
ABSTRACT:

Recent reports of "lineage switching" from a lymphoid to macrophage
phenotype have left unresolved the question of whether such cells are functional
macrophages or nonfunctional products of differentiation gone awry. This study
demonstrates that several "macrophage-like" cell lines derived from
v—Ha—ras-transformed pre-B cells have gained the capacity to effectively present
antigen in an MHC-restricted fashion. Using an assay involving the co-cultivation
of putative antigen-presenting cells with chicken ovalbumin (cOVA) and a
cOVA-specific T cell hybridoma, "lineage switch” cell lines were found to present
antigen as effectively as macrophage-containing peritoneal exudates. Neither the
original pre-B cell precursors nor B cell lymphomas derived from them present
antigen. Thus we have demonstrated that these “lineage switch" macrophages are
capable of antigen presentation, a mature differentiated function.

While gaining macrophage characteristics, these cells have also rearranged
their kappa light chain immunoglobulin locus, suggesting that macrophage
differentiation and immunoglobulin rearrangement are not mutually exclusive
processes. The existence of both lymphoid and myeloid characteristics in a cell
fully capable of antigen presentation suggests greater plasticity in hematopoietic

lineage commitment than conventionally thought to be the case.

 

1James Bretz, Shu-Chih Chen, Diane Redenius, Hsun-Lang Chang, Walter J. Esseiman, and
Richard Schwartz. Department of Microbiology, Michigan State University, East Lansing, Mi 48824-
1 101

115

1 16
INTRODUCTION:

The concept that hematopoietic differentiation involves an early and
irreversible lineage commitment is brought into question by numerous observations
of leukemias and lymphomas that express myeloid or lymphoid markers outside
of their respective lineages. The coexpression of differentiation markers has been
interpreted as being either an aberrant phenomenon caused by leukemogenesis
(McCulloch, 1983) or a reflection of the normal but transient existence of bipotentiai
progenitors in hematopoiesis (Greaves et al., 1986). in particular, the existence of
a number of transformed cell lines with both lymphoid and macrophage
characteristics has suggested a close relationship between these lineages. Murine
macrophage cell lines have been derived from lymphoid tumors and from in vitro
transformants induced either by murine leukemia viruses or chemical carcinogens
(Boyd and Schrader, 1982; Holmes et al., 1986; Hanecak et al., 1989). Three
groups have studied systems where a transition from a lymphoid to a macrophage
phenotype could be induced. Klinken et al. (1988) demonstrated that B lymphoid
cells from transgenic mice that express c—myc using the immunoglobulin mu
enhancer could be induced to take on macrophage-like characteristics when
infected with a retrovirus expressing v—raf. Davidson et al. (1988) showed that a
v—Ha—ras transformed lymphoid cell line could be stimulated by Iipopolysaccharide
(LPS) to differentiate along either the lymphoid pathway into pre—B-like cells or
along the myeloid pathway into macrophage-like cells. Recently, Borzillo et al.
(1990) reported the CSF—1 dependent macrophage lineage transition of a pre—B
cell line expressing the human CSi‘—1 receptor.

The macrophage-like cell lines that have been derived from B lymphoid cells

117

have been classified as macrophage on the basis of their morphology, expression
of MAO—1, MAC—2, a-naphthyi acetate esterase and iysozyme, and their ability to
phagocytose latex beads. More functional assays for antigen presentation and
tumoricidal activity that would establish whether these cells could act in vivo
similarly to authentic macrophages have not been presented. in this paper, we
demonstrate the ability of several macrophage cell lines derived from

v—Ha—ras-transformed pre—B cells to present antigen to a T helper cell hybridoma.

RESULTS:

A tumor consisting of adherent cells with a macrophage morphology was
identified during our studies on the tumor progression of a pre-B lymphoid cell line
expressing v-Ha-ras (Chen et al., 1991). This tumor, designated tumor 4, was
derived from a clonal cell line, designated R2, that was generated by infection of
fresh murlne bone marrow with a mixture of a v-Ha-ras-expressing retrovirus and
Moloney murine leukemia virus (MoMuLV) (Schwartz et al., 1986a). The R2 cell
line was classified as being a pre-B cell on the basis of several criteria. It
possessed 6 blast cell morphology with a large nucleus and scant cytoplasm. it
expressed the B lineage—specific marker, B220 (Coffman and Weissman, 1981).
While not expressing detectable immunoglobulin mu chain, R2 showed a
rearrangement in the DNA of that locus. The immunoglobulin kappa chain locus

was in a germiine configuration.

Tumor 4 is Derived from the R2 Cell IJne.

in order to ascertain whether we had identified a probable instance of

118

   

Figure 1. Viral integrations. Southern blot analysis of DNAs from liver (L), R2 and
tumor 4 (T4). (A) DNA was digested with EcoRI and 10 pg of each sample was
electrophoresed through 0.8% agarose. The blot was probed for v—Ha—ras. (B)
DNA was digested with Bglll. The blot was probed for murine ecotropic env
sequences. Size markers are the positions of an ethidium bromide-stained Hindill
digest of bacteriophage A and are denoted in kilobases.

119

lineage switching, it was necessary to demonstrate that tumor 4 was derived from
R2. To that end, the sites of integration of the v-Ha-ras-expressing retrovirus and
MoMuLV were compared between the tumor and the cell line. Southern
hybridization analysis of EcoRI-digested DNA with a v-Ha-ras probe showed that
tumor 4 contained the same 5.3 kb proviral integration fragment as R2 (Figure 1A).
This proviral integration fragment is defined by a 3’ EcoRI site internal to the viral
genome and a 5’ EcoRI site peculiar to the site of integration. In addition to the
5.3 kb fragment, there is a 23 kb fragment representing the endogenous c-Ha-ras
in all the DNAs. Southern hybridization analysis of Bglll-digested DNA, using a
probe for the ecotropic MuLV env gene, revealed similar MoMuLV integration
fragments in R2 and tumor 4 (Figure 1B). The MoMuLV genome possesses a
Bglll site within env, such that the above hybridization would detect a fragment
extending from that Bglll site to a Bglll site in the host cell genome flanking the 3’
terminus of the provirus. These data demonstrate that the putative macrophage

tumor was derived from the pre-B cell line.

Tumor 4 Cells Possess Macrophage Characteristics.

Tumor 4 was initially suspected to be a macrophage because of the large
size of its cells and its adherent growth in cell culture. Microscopic examination
of Wright-Glemsa stained cells confirmed their large size and revealed the cells of
tumor 4 (Figure 28) to have a much more extensive and granular cytoplasm than
R2 (Figure 2A). An immunoperoxidase detection procedure found tumor 4 cells
to have retained some expression of B220, and to have gained expression of high

levels of MAC-1 (data not shown). MAC-1 is generally considered to be a marker

 

Figure 2. Nonspecific phagocytosis of latex beads. The cells were Wright-Glemsa
stained and photographed at 200x magnification. (A) R2; (B) tumor 4.

121

for cells of the myeloid lineage (Springer et al., 1979). Histochemical
proceduresrevealed a high level of a-naphthyl acetate esterase activity in tumor 4
cells, which is not found in R2 cells (data not shown). This is an enzyme activity
generally associated with cells of the monocyte-macrophage lineage (Rogers et al.,
1980). Tumor 4 cells (Figure 2B) were positive for the nonspecific phagocytosis
of latex beads, while R2 cells (Figure 2A) were not. Nonspecific phagocytosis is
another marker of the monocyte-macrophage lineage (Raschke et al., 1978).
These data strongly suggest a macrophage phenotype for tumor 4 cells.

At late stages of myeloid differentiation, the levels of c-myc and c-myb
mRNA decrease, while the level of c-fms mRNA increases (Gonda and Metcalf,
1984; Sheng-Ong et al., 1987). The levels of mRNA from these protooncogenes
detected in tumor 4 cells was consistent with tumor 4 having advanced to a late
stage of myeloid differentiation. Cytoplasmic poly A+ RNAs of the parental R2 cell
line, tumor 4 and six other tumors derived from R2 that had lymphoid
characteristics were examined by Northern hybridization analysis (Figure 3). One
blot was hybridized successively with c-myc and fl 2-microglobulin probes. Another
blot was hybridized successively with c-myb, c-fms and rat
gcheraldehyde-B-phosphate dehydrogenase (GAPDH) probes. Hybridization to
the fi2-microglobulin and GAPDH probes provided a control for gel loading.
Tumor 4 cells clearly show reduced levels of c-myc and c-myb expression in
comparison to R2 cells and lymphoid tumors. in contrast c-fms expression is
elevated in tumor 4 cells. A cell line with macrophage characteristics has also
been isolated from tumor 5 cells, which show elevated c-fms expression (Figure

3).

122

c-Mvc~
I! 1’1 17 11 1" 1'5 1'! ‘7
”fl” 9.5 p

m

nannn'nnnn

Deals-rel.”

 

Figure 3. RNA analyses of c—myc, c—myb and c—fms. Northern blot analyses
were performed on polyA+ RNA from R2 and seven tumors (T 1—17). Each
sample of RNA was the polyA+ fraction selected from 150 pg of total cytoplasmic
RNA. One blot (upper panel) was probed successively for both c-myc and
BZ-microgiobulin, while the other blot (lower panel) was probed successively for
c—myb, c—fms and rGAPDH.

123

Another aspect of macrophage function is the ability to release cytokines in
response to LPS stimulation. We examined the presence of iL-1, lL—6 and TNF
in the media of cells cultured in the presence or absence of 10 pg/ml of LPS for
24 hours. Cellular proliferation assays for lL—1 and lL-6, and a cytotoxicity assay
for TNF revealed varying levels of cytokine release for six subclones of tumor 4
(see below), while the parental R2 pre—B cell line did not elaborate any of these
cytokines except low levels of lL-1 (Table 1). The LPS-inducible release of
cytokines was again consistent with a macrophage phenotype for tumor 4 cells.
Tumor 4 Cells Also Show Differentiated Lymphoid Characteristics. Davidson et. al.
(1988) found that a v-Ha-ras-tranformed lymphoid cell line could be stimulated to
differentiate along either the myeloid or lymphoid pathways. Since tumor 4 cells
showed a variety of DNA rearrangements in the kappa light chain locus (data not
shown), it was of interest to determine whether the cells that had gone on to
rearrange the kappa locus were the same cells that had progressed toward a
macr0phage phenotype or whether the tumor 4 cells were a mixed population of
B cells and macrophages. To that end, tumor 4 cells were plated in soft agar
medium and six subclones were recovered. Southern hybridization analysis of
EcoRI-digested DNA isolated from the subclones showed that they all contained
the same 5.3 kb v-Ha-ras proviral integration fragment as R2 and tumor 4 cells
(data not shown; see Figure 1A). The six subclones of tumor 4 possessed the
same myeloid characteristics described above for the uncloned tumor, but varied
in their pattern of kappa light chain gene rearrangement. Southern hybridization
analysis of BamHl-digested DNAs with a kappa probe revealed that subclones 3

and 5 possessed one germiine and one rearranged kappa allele, while subclones

124

Table 1. LPS-induced cytokine release by tumor 4 macrophage subclones.

 

 

lL-1 (U/ML) IL-6 (U/ML) TNF (U/ML)
-LPS +LPS -LPS +LPS -LPS +LPS
r12 0 2 0 0 0 0
T4.1 0 2 0 1 0 0
74.2 0 6 0 100 0 40
T43 0 6 0 1500 0 65
74.4 0 19 0 39 0 34
T45 1 4 0 0 0 0
T4.6 o 4 0 0 o 0

 

Table 1. The capacity of cell lines to release the cytokines lL—1, IL—6 and tumor
necrosis factor (T NF) was determined by assaying culture supernatants. For this
purpose, cell lines were incubated for 24 hours at 2.5x10‘5 cells/mi with 10 pg/ml LPS
in RPMI 1640 supplemented with 10% fetal calf serum and 5x10’5 M
2-mercaptoethanol. Culture supernatants were collected, passed through a 0.2
micron filter and stored 6t—70° C until assayed. lL—1 activity was assayed by its ability
to induce proliferation of D10.G4.1 cells in the presence of Concanavalin A as
described by Ayala et al. (19906). lL-6 activity was determined by its ability to induce
the proliferation of the 7TD1 B-cell hybridoma as previously described by Huitner at
al. (1989). TNF activity was assessed by its cytotoxicity to WEHl—164 clone 13 cells
as previously described by Ayala et al. (1990b). The relative units of cytokine activity
were determined by comparison of the activity of dilution series of experimental
supernatants to the activities of dilution series of purified human lL—1 (Genzyme),
recombinant human lL—6 (Amgen Corp.) or murlne TNF—alpha (Amgen Corp.)
standards.

125

 

xiii»: ’5‘" .A .51

14.1
.

v
(‘0

E
3%
5%

2’;

 

 

Flgure 4. Kappa light chain rearrangements. Southern blot analysis of DNAs from
liver (L), R2 and six subclones of tumor 4 (1—6). DNA was digested with BamHI
and 10 pg of each sample was electrophoresed through 0.8% agarose. The blot
was probed for kappa light chain constant region sequences. Size markers are
the positions of an ethidium bromide-stained Hindill digest of bacteriophage A and
are denoted in kilobases.

126

1,2,4 and 6 possessed rearrangements in both alleles (Figure 4). All the subclones
possessed a rearranged BamHI fragment of approximately 7 kb. Tumor 4 was
apparently derived from an outgrth of R2 that had undergone this
rearrangement. Some of the subclones then proceeded to rearrange their other
kappa allele. Clearly, tumor 4 contained cells which individually had differentiated
along both the lymphoid and myeloid pathways.

Having observed kappa light chain rearrangements in macrophage
subclones of tumor 4, we next examined the status of immunoglobulin expression
by Northern blot analysis. Kappa light chain transcript could not be detected (data
not shown). A mu heavy chain probe revealed a diverse range of RNAs in the
macrophage subclones (Figure 5) that correspond in size to 1.9, 2.1, 2.3 and 2.9
kb transcripts reported to be initiated in the mu switch region of myeloid cell lines
(Kemp et al., 1980). The R2 pre-B cell line possesses predominantly larger RNA
species that include those that correspond in size to mature mu mRNAs of 2.4 and
2.7 kb. These species are diminished upon lineage switch. Comparison to a
hybridization of the same blot with a probe for GAPDH (Figure 5) shows the R2
RNA to be underloaded and thus the diminution of mu transcription in the
macrophages is even more dramatic than apparent from casual inspection of the
data. Apparently, the macrophage subclones of tumor 4 lose the capacity to
transcribe functional mu mRNA, even though the rearrangement of the kappa locus
suggests progress in lymphoid differentiation.

Since CD45 isoforms have been reported to be lineage specific (Saga et al.,
1987; Streuli et al., 1987; Ralph et al., 1987), the expression of this surface marker

was examined among the subclones of tumor 4. An immunoperoxidase detection

127

MU
14 Melon“

 

221234.16

 

Figure 5. Expression of mu heavy chain RNA. Northern blot analysis was
performed on poly A+ RNA from R2 and six subclones of tumor 4 (1-6). Each
sample of RNA was the poly A+ fraction selected from 100 pg of total cytoplasmic
RNA. The blot was probed for mu heavy chain. The positions of ethidium
bromide-stained rRNAs are noted on the right. The positions of mu RNA species
are marked on the left and denoted in kilobases. The lower panel shows the same
blot probed for GAPDH as a control for loading.

128
procedure detected 8220, the B lymphoid isoform of CD45, in tumor 4 cells. The

expression of CD45 was further examined among the tumor 4 subclones in orderto
determine the relative expression of the B220 isoform in comparison to the isoform
that predominates in myeloid cells. Recently, Chang et. al. (1989) described the
use of a reverse transcription-polymerase chain reaction (RT-PCR) technique to
determine the pattern of alternate exon use in CD45 expression of hematopoietic
cells. They found that B lymphoid cell lines uniquely expressed a form of CD45
mRNA possessing 3 optional exons, while two myeloid cell lines (a macrophage
and a mast cell) predominantly expressed a form lacking these exons. We utilized
RT—PCR to examine CD45 expression among R2 and the subclones of tumor 4
(Figure 6). All of the cell lines expressed multiple species of CD45 mRNA.
Subclones 2,3,4 and 6 expressed a CD45 mRNA containing 3 optional exons,
typical of B lymphoid cells, while subclones 1 and 5 predominantly expressed
mRNA lacking these exons, typical of myeloid cells. R2 expressed the expected
three exon 8 lymphoid isoform. Thus the pattern of CD45 expression is
heterogeneous among macrophage subclones of the same tumor.

The subclones of tumor 4 can function effectively in antigen presentation.
In order to test the ability of the tumor 4 subclones to present antigen, an assay
system required an antigen-specific T helper cell line that could be stimulated to
produce interleukin 2 (IL-2) upon presentation. For these experiments, the putative
antigen-presenting cells were co-cultivated with a T cell hybridoma specific for
chicken ovalbumin (cOVA) and restricted for l—Ad (the haplotype for BABL/c),
DO.11.10/54.4. In the presence of cOVA, authentic macrophages such as those

in a peritoneal exudate stimulate the hybridoma to produce lL-2 (Figure 7A). IL-2

129

production was assayed by the application of media supernatants from
co-cultivations to an IL—2 dependent cell line, CTLL-2. All of the macrophage-like
tumor 4 subclones displayed antigen presentation capacities comparable to
peritoneal exudates (Figure 7A). Furthermore, all of the tumor 4 subclones
showed antigen presentation capacities dramatically greater than either the
parental R2 pre-B cell line or tumor 1, a B cell tumor derived from R2 (Figure 7A).
lL-2 production was dependent on the presence of cOVA during co-cultivation of
presenting cells with cells of the helper T cell hybridoma. Supematants produced
in the absence of cOVA were analyzed for all the cell lines and the values for lL—2
production were found to be near zero (data not shown). These control values
were subtracted from those determined for supernatants produced in the presence
of cOVA to generate the data presented in Figure 7A, B and C. lL—2 production
was also dependent on the presence of T cell hybridoma cells. Supematants
produced by incubations of putative presenting cells with cOVA in the absence of
T cell hybridoma cells had no detectable lL—2 (data not shown). The ability to
present antigen was not stimulated by exposure to LPS for any of these cell lines
(data not shown). A macrophage-like outgrth from tumor 5 (also derived from
R2) and the cells of a macrophage-like tumor derived from the pre-B cell line R1
(9) showed levels of antigen presentation similar to those observed for the tumor
4 subclones (Figure 7B). The ability to present antigen may therefore be a
common phenomenon among v-Ha-ras-transformed B lymphoid cells that acquire
macrophage-like characteristics.

Authentic antigen presentation should be MHC—restricted, so all of the

putative antigen-presenting cells were also co-cultivated with a T cell hybridoma

130
specific for cOVA and restricted for l—Aq, 30023-244. As exemplified by subclone

4 of tumor 4 (T 4.4) and the macrophage tumor derived from R1 (R1T), the antigen

presentation observed is MHC-restricted (Figure 7C).

l-A Expresslon

Antigen presentation to T cells requires la expression and the observation
of l—Ad-restricted presentation (Figure 70) indicates that these "lineage switch"
macrophages express l—Ad. In order to assess whether the acquisition of
presentation capacity correlated with acquisition of la expression, in particular
l—Ad, we performed flow cytometry with FITC-conjugated anti-mouse l-Ad on the
macrophage cell lines and their pre—B cell precursors. While both R1 and R2
(pre-B cells) displayed no detectable l—Ad, the macrophage cell lines represented

by R1T and T44 showed a low expression of l—Ad (Figure 8).

131

 

Figure 6. RT—PCR analysis of CD45. RT—PCR was performed on the polyA+
RNAs of R2, the six subclones of tumor 4 and P388D1 (a myeloid control). The
products were electrophoresed through 2% agarose and stained with ethidium
bromide. (C) 3-exon plasmid control; (1—6) T4 subclones; (M) 123 bp ladder; (Mp)
P388D. PCR products smaller than the 0 exon product may represent an RNA
species lacking an additional exon (Chang and Esselman, unpublished results).

132

 

 

 

 

 

 

 

 

 

 

 

 

'. _. - . 1
.\ o :93 B W
; 0 Ti : ~ Q
i o T41 E .. ‘. $5111
3 v ‘1“: 3 H“ LJ
g v T4 3 E 9
.1 a T4 4 3 p
7: I T4 -~ '? -
'5 A T4 0 E:
‘2 '3 e
g E -
L"! 3)
l‘ f‘ 4
l 1' T V l 3 . 1 1 v .
O 2 4 6 3 'C ‘; ’4 ‘6. '5 ;I‘ 2: G 2 4 6 8 “O '2 34 i6 ‘8 23
4
No. of presenting cells/culture ix 10 i No. of presenting cells/culture (x 10‘)

 

( v
6 .. \
V
'AT ”-31
v RlT ll-A“)
0 14411-4 l
o T14 ll-qu

 

 
 

 

"-2 produced (units/ml)
(A

 

 

 

 

I i . 1 I T Tr T
O l 2 3 4 5 5 7 8 9‘ 10
No. of presenting cells/culture ix ii) i

Figure 7. Antigen presentation. (A) Antigen presentation assays were performed on R2, the tumor
4 subclones (4.1—4.6), tumor 1 (a B cell tumor derived from R2),

and peritoneal exudates (er). (B) Antigen presentation assays were performed on R2, a
macrophage outgrth of tumor 5 (TS; derived from R2 in a different animal), R1 (another pre-B
cell line transformed by v-Ha-ras). and R1T (a macrophage tumor derived from R1). (C) Antigen
presentation assays were performed on a tumor 4 subclone (T 4.4) and R1T with DO-11.10/54.4
(l-A'd-restricted) or 30023-244 (l-Aq-restricted). Each point represents an average value obtained
from a dilution series for each presentation supernatant, testing the response of CTLL-2 cells to
lL—2 in those superantants. The results shown are representative of at least two experiments with

each cell line.

133

Allliiiflliifi 4’

Joacogufioo

>332; 0... .823 P... 328.2093 .6 33a of ~¢ ace 2. .8; 43.: v2.2.3 2.3.6.: 33: canon—580-92“. Jobcoo o no Canoe:

u_.o..u<-_._sc.

33.24.3090»: 5:. 3033923: 2». one «:2 .3. .2. co Eaton mo: F528;“. so:

 

 

 

 

 

mozmomNEODJu
F. _ _ P . . mi _
PE pm —
P _ p b n _
«.3. «I _

 

£339.98 31. .o v.53...

SLNDOO

134
DISCUSSION:

This study demonstrates the capacity of several macrophage-like tumor cell
lines derived from v—Ha—ras-transformed pre—B cell lines to present antigen with
MHC-restriction. This finding establishes that cells having undergone "lineage
switching" can perform a function normally associated with a fully differentiated
macrophage or B cell. While numerous examples exist of B lymphomas with the
capacity to present antigen (Chesnut et al., 1982; Walker et al., 1982), neither the
pre—B cell precursors of the macrophage-like cell lines nor B cell lymphoma cell
lines derived from those precursors could present antigen. Thus the capacity to
present antigen appears to correlate with the differentiation of these cells along the
macrophage lineage. Indeed, two cell lines with the most dramatic level of antigen
presentation had lost expression of the B cell isoform of CD45 and displayed a
pattern of CD45 more typical of a myeloid cell (subclones 1 and 5, Figure 6; T4.1
and T45, Figure 7A). Perhaps, loss of the B cell isoform of CD45 is indicative of
further maturation along the myeloid lineage. it may be worthwhile to investigate
the role of CD45 in macrophage function. The fact that similar antigen
presentation abilities were found in macrophage derivatives of two completely
independent cell lines (R1 and R2) suggests the generality of this phenomenon.

Since it is well established that lL-1 along with antigen presentation is an
important co-activator of T cells, it is surprising that the inducibility of cytokine
release by LPS (Table 1) does not correlate with the effectiveness of antigen
presentation by the T4 subclones (Figure 7A). Apparently the low levels of lL—1
that some of these macrophages are capable of elaborating is sufficient for T cell

activation. The observation that two of the best lines for antigen presentation 0' 4.1

135
and T45) have a weak response to LPS suggests that LPS-induced cytokine

release may not be an adequate measure in itself for evaluating macrophage
function.

The "lineage switch" macrophages reported here express a low level of la
(Figure 8). This is consistent with the previous report of Davidson et al. (1988).
The precursor pre—B cells lack detectable la. Perhaps la expression is the critical
property determining the capacity to present antigen among these cells. Certainly,
la expression is necessary for antigen presentation, but its sufficiency for antigen
presentation among the cell lines we have studied will require further
experimentation.

The six macrophage-like subclones of tumor 4, while possessing a common
rearranged kappa allele, displayed a variety of kappa light chain gene
rearrangements at their other kappa allele. Compared to their parental cell line,
these cells have progressed along the B as well as the monocyte/macrophage
lineage. The varying rearrangements of one kappa allele suggests rearrangement
subsequent to macrophage conversion and that at least certain elements of
lymphoid and macrophage differentiation programs are not mutually exclusive.
The fact that the R2 cell line can also generate a lymphoma (T 1, Figure 7A) that
expresses both mu and kappa chains (data not shown) demonstrates the potential
of this cell line to differentiate quite far along either the lymphoid or macrophage
pathways. The relationship between lymphoid and macrophage differentiation
revealed in these cells differs somewhat from that seen in cases of "lineage switch“
previously reported. Klinken et al. (1988) found “lineage switch“ macrophages at

both the pre-B and B cell stages of immunoglobulin rearrangement. However, they

136

did not find macrophages that had progressed in their immunoglobulin
rearrangement compared to their lymphoid cell precursors, as we have. Davidson
et al. (1988), examining v—Ha—ras-transformants similar to those reported here,
could induce those cells to differentiate into either lymphoid or macrophage cells
upon exposure to LPS. The lymphoid derivatives they reported did not progress
beyond the pre-B cell stage, while we have identified an immunoglobulin producing
tumor derived from a pre-B cell line that also gave rise to a macrophage tumor.
Perhaps the more complex environment provided during tumor challenge allowed
the cells described here to more fully develop along the lymphoid lineage when
that pathway was selected. At any rate, the v—Ha—ras-transformed pre-B cells
described here seem truly bipotentiai.

The ability of these cells which undergo an apparent "lineage switch" to
perform a fully differentiated function presents the possibility that they may
represent an unusual but normal subset of hematopoietic cells rather than an
oddity induced by transformation. The existence of both lymphoid and
macrophage characteristics in a cell fully capable of antigen presentation suggests
greater plasticity in hematopoietic lineage commitment than conventionally thought

to be the case.

MATERIALS AND METHODS:

Cell Lines. R1 and R2 are v—Ha—ras-transformed murine pre—B cell lines
described in Schwartz et al. (1986a). Tumors derived from R1 and R2 were
generated as described in Schwartz et al. (19863,b) in syngeneic BALB/c mice and

in BALB/c athymic nude mice. Briefly, cells were washed twice in RPMI 1640 and

137

were then resuspended in the same at 8x106 cells per ml. Five week old mice
were injected intraperitoneally with 0.25 ml of the cellular suspension. Tumor 4, in
particular, was isolated from an inguinal lymph node at 74 days post injection.
Tumor cell lines were readily produced from explanted tumors by dispersal and
transfer to feeder cultures of adherent bone marrow cells (Whitlock et al., 1983).
All of these cell lines were cultured over feeder cells in RPMI1640 supplemented
with 5% fetal calf serum and 5x10'5 M 2-mercaptoethanol.

The subclones of tumor 4 were generated from single colonies grown in soft
agar medium as described by Whitlock et al. (1983). The T cell hybridoma,
DO—11.10/54.4, was a generous gift of Drs. Philippa Marrack and John Kappler
(University of Colorado, Denver) (White et al., 1983). This hybridoma is specific
for chicken ovalbumin in the context of l—Ad and crossreacts weakly with chicken
ovalbumin in the context of l—Ab. 30023-244, another T cell hybridoma, was also
a gift of Drs. Marrack and Kappler. This hybridoma is specific for chicken
ovalbumin in the context of either l-Aq or l—E. CTLL-2 is a T cell line responsive
to lL—2 and was obtained from the ATCC. All of these cell lines were cultured in
RPMI1640 supplemented with 10% fetal calf serum and 5x10"5 M
2-mercaptoethanol in the absence of any feeder cells.

Peritoneal exudates containing macrophages were produced from BALB/c
mice treated 1 week previously with a 0.5 ml intraperitoneal injection of pristane.
Nucleic Acid Analysis. Cytoplasmic RNA was isolated from actively growing cells
by a sodium dodecyl sulfate-urea procedure as described by Schwartz et al.
(1981). Poly A+ RNA was selected by oligo—dT cellulose chromatography (Rave

et al., 1979). RNA was denatured, electrophoresed in a formaldehyde-1% agarose

138
gel (15), and transferred to Nytran (Schleicher and Schuell) (Thomas, 1980).

High molecular weight DNA was isolated from nuclei collected in the
preceding RNA isolation procedure as described in Schwartz et al. (1986a). DNA
was digested with restriction enzymes as noted in the figure legends,
electrophoresed through 0.8% agarose and transferred to Nytran (Southern, 1975).

Hybridization probes were prepared by nick translation (Rigby et al., 1979)
through the incorporation of [a-32P] dATP (3000 Ci/mmol; lCN). The v—Ha—ras
probe was the replicative form of phage M13mp10 containing a 0.46 kb EcoRI
fragment corresponding to v—Ha—ras encoding sequences (Ellis et al., 1980). The
env probe was a 0.8 kb BamHI fragment from the env region of Friend murlne
leukemia virus and is specific for the env sequences of murine ecotropic
retroviruses (Silver and Kozak, 1986). The rmyc probe was the 4.7 kb genomic
Hindill fragment of murine emyc (Stanton et al., 1984). The murine c—myb probe
was a cloned 2.4 kb cDNA (a generous gift of Dr. Timothy Bender, University of
Virginia, Charlottesville). The fms probe was a cloned 2.7 kb CIal-BamHl fragment
of the McDonough strain of feline sarcoma virus (Donner et al., 1982). The murine
BZ-microglobulin probe was a cloned 0.5 kb cDNA (Parnes et al., 1981). The rat
glyceraldehyde-S-phosphate dehydrogenase (GAPDH) probe was a cloned 1.3 kb
cDNA (Fort et al., 1985). The murine kappa light chain probe was the replicative
form of phage M13mp10 containing a genomic 0.48 kb Hpal-Bglll fragment
extending from a point about 50 base pairs within the 5’ terminus of the kappa light
chain constant region gene to the poly A addition site (Seidman and Leder, 1978).
The murlne mu heavy chain probe was a cloned cDNA (p 12) which extends from

CH2 to the 3’-untranslated region of the secreted form of mu mRNA (Rogers et al.,

139

1980). All hybridizations were performed under aqueous conditions in 5 x SSC at
65° C and washed to a stringency of 0.1 x SSC at 65° C.

Reverse Transcription-Palymerase Chain Reaction (RT-PCR). RT-PCR was
performed according to the procedure of Chang et al. (1989; 1991) using poly A+
RNA as substrate. The primers were a sense primer specific to exon 2
(GCCCTTCTGGACACAGAAGT; base positions 167—186) and an anti-sense primer
specific to exon 9 (AATTCACAGTAATGTTCCCAAACAT; base positions 764- 740)
of the cDNA of murine CD45 (Thomas et al., 1987). cDNA was prepared by
incubating 1 ug of poly A+ RNA for 60 min at 37° C with 200 units of MoMuLV
reverse transcriptase in a 20 ul reaction volume containing 50 mM Tris-HCl (pH
8.3), 75 mM KCI, 3 mM MgCl2, 5 mM DTT, 100 149/ml BSA, 40 units RNasin, 500
uM dNTP and 200 ng of anti-sense primer. A 5 pl aliquot was used directly for
PCR amplification in a 50 pl reaction volume containing 50 mM KCl, 10 mM
Tris-HCl (pH 9.3), 3 mM MgCl2, 0.1% w/v gelatin, 500 uM dNTP, 400 ng of sense
and anti-sense primers and 2.5 units of Taq polymerase. PCR was performed in
a DNA Thermal Cycler (Perkin-Elmer-Cetus, Inc.) for 24 cycles. Each cycle
consisted of 40 s at 94° C for denaturation, 15 s at 55° C for annealing and 30 s
at 72° C for elongation. The first cycle was preceded by a 5 min incubation at
94° C and the last cycle followed by a 4 min incubation at 72° C.

Cytological Analyses. Cells were cytocentrifuged onto a microscope slide and
allowed to air dry overnight. The cells were then incubated with either rat
anti-B220 (monoclonal 14.8) or rat anti-MAC—1 (Boehringer Mannheim). Goat
anti-rat immunoglobulin-horseradish peroxidase (Boehringer Mannheim) was used

in a secondary incubation for detection. The presence of a-naphthyl acetate

140
esterase was determined by cytochemical staining (Yam et al., 1971) with a Sigma

research kit. Nonspecific phagocytosis of latex beads was assayed by the method
of Raschke et al. (Raschke et al., 1978).

Antigen Presentation. Assays for antigen presentation were performed in a
manner similar to that described by Marrack et al. (1989). Briefly, the cell lines to
be assayed for antigen presentation were titrated into 200 pl microcultures
containing 105 cells of either the T cell hybridoma DO-11.10/54.4 or 30023-244,
both of which produces lL—2 in response to the presentation of chicken ovalbumin
(cOVA) in the context of l—Ad or i—Aq, respectively. These assays were carried
out in RPMI 1640 supplemented with 10% fetal calf serum, 5x10'5 M
2-mercaptoethanol and, where required, cOVA at 1 mg/ml. After 24 hours,
incubation supernatants from these cultures were assayed for IL-2 using CTLL-2,
an lL—2—dependent cytotoxic T cell line. Two-fold serial dilutions of supernatants
were added to 5x103 CTLL—2 cells in 100 pl microcultures and incubated for 48
hours at 37° C. MT‘I’ (Sigma), a substrate for production of a colored product
indicative of cell survival (Mosmann, 1983), was added at 0.5 mg/ml and the
cultures incubated for an additional 4 hours at 37° C. Acid-isopropanol (40 mM
HCI) was then added to dissolve the MIT formazan reaction product. The optical
density of each well was quantitated by an ELISA reader at a wavelength of 540
nm. The specific activity of IL-2 in the supernatants was determined by
comparison to a standard curve produced through the use of purified recombinant
lL—2 (Cetus lnc.).

Flow Cytometry. Cells were stained in PBS, 2% FCS with either FlTC-conjugated

monoclonal antibody AMS-32.1 (anti-mouse l-Ad) (Phar Mingen) or

141

FlTC-conjugated mouse lgGZb, K (Phar Mingen) as an isotype-matched control.
Cells were then fixed in PBS, 2% FCS, 0.5% formaldehyde and stored at 4° C until
analysis. Flow cytometry was performed using an Ortho Diagnostics

Cytofluorograph 50-H.

ACKNOWLEDGMENTS:

This work was supported by Public Health Service grants CA45360 (ROS)
and GM35774 (WJE) and by a grant from the Elsa U. Pardee Foundation (RCS).
We are indebted to Alfred Ayala for the cytokine assays. The authors thank Donna
Paulnock for helpful discussions, and Susan Conrad and Michele Fluck for

thoughtful comments on this manuscript.

‘ 142
REFERENCES:

1. Avlv, H. and Leder, P. (1975) Purification of biologically active globulin messenger RNA by
chromatography on oligothymidylic acid-cellulose. Proc. Natl. Acad. Sci. USA 69, 1408-1412.

2. Ayala, A., Perrin, M.M., Wagner, MA. and Chaudry, l.H. (1990a) Enhanced susceptibility to
sepsis following simple hemorrhage: depression of Fc and 03b receptor-mediated phagocytosis.
Arch. Surg. 125:70-75.

3. Ayala, A., Perrin, M.M., Meldrum, D.R., Enel, W. and Chaudry, l.H. (1990b) Hemorrhage induces
an increase in serum TNF which is not associated with elevated levels of endotoxln. Cytoklne 3,
170-175.

4. Borzillo, G.V., Ashmun, RA and Sherr, C.J. (1990) Macrophage lineage switching of murine
early pre-B lymphoid cells expressing transduced fms genes. Mol. Cell. Biol. 10, 2703-2714.

5. Boyd, AW. and Schrader, J.W. (1982) Derivation of macrophage-like lines from the pre—B
lymphoma ABLS 8.1 using 5-azacytldine. Nature 297, 691-693.

6. Chang, H.L, Zaroukian, M.H. and Esselman, W.J. (1989) T200 alternate exon use in murlne
lymphoid cells determined by reverse transcription-polymerase chain reaction. J. lmmunol. 143,
315—321.

7. Chang, H.L, Lefrancois, L., Zaroukian, M.H. and Esselman, W.J. (1991) Developmental
expression of CD45 alternate axons in murlne T cells. Evidence of additional exon use. J. Immunol.
147, 1687-1693.

8. Chen, S.-C., Redenius, D. and Schwartz, RC. (1991) Tumorigenesis of a v-Ha-ras—expresslng
pre-B cell line selects for c-myc activation. Biochem. Biophys. Res. Comm. 178, 1343-1350.

9. Chesnut, R.W., Colon, SM. and Grey, HM. (1982) Antigen presentation by normal 8 cells, B
cell tumors and macrophages: functional and biochemical comparison. J. Immunol. 128,
1764-1768.

10. Coffman, R.L, and Weissman, LL. (1981) A monoclonal antibody that recognizes B cells and
8 cell precursors in mice. J. Exp. Med. 153, 269-279.

11. Davidson, W.F., Pierce, J.H., Rudikoff, S. and Morse, H.C. (1988) Relationships between B cell
and myeloid differentiation. J. Exp. Med. 168. 389-407.

12. Donner, L, Fedele, LA., Garon, C.F., Anderson, SJ. and Sherr, G.J. (1982) McDonough feline
sarcoma virus: characterization of the moiecularly cloned provirus and its feline oncogene (v-fms).
J. Virol. 41, 489-500.

13. Ellis, R.W., De Feo, D., Maryak, J.M., Young, HA, Shih, T.Y.. Chang, E.H., Lowy, DR. and
Scolnick, EM. (1980) Dual evolutionary origin for the rat genomic genetic sequences of Harvey
murlne sarcoma virus. J. Virol. 36, 408-420.

14. Fort, P., Marty, L, Piechaczyk, M., El Salrouty, S., Dani, C., Jeanteur, J. and Blanchard, J.M.
1985. Various rat adult tissues express only one major mRNA species from the
glyceraldehyde-S-phosphate dehydrogenase multlgenlc family. Nucl. Acids Res. 13, 1431-1442.

15. Gonda, T.J., and Metcalf, D. (1984) Expression of myb, myc and fos proto—oncogenes during
the differentiation of murlne myeloid cells. nature 310, 249-251.

143

16. Greaves, M.F., Chan, LC., Furley, A.J.W., Watt, SM. and Molgard, H.W. (1986) Lineage
promiscuity in hemapoletlc differentiation and leukemia. Blood 67, 1—11.

17. Hanecak, R., Zovich, D.G., Pattengale, PK and Fan, H. (19%) Differentiation In vino of a
leukemia virus-induced B-cell lymphoma into macrophages. Mol. Cell. Biol. 9, 2264-2268.

18. Holmes, K.L., Pierce, J.H., Davidson, W.F. and Morse, H.C. (1986) Murine hematopoietic cells
with pre-B or pre— B/myeloid characteristics are generated by in vitro transformation with
retroviruses containing fes, ras. abl, and src oncogenes. J. Exp. Med. 164, 443-457.

19. Huitner, L, Szots, H., Welle, M., Van Snick. J., Moeiler, J. and Dormer, P. (1989) Mouse bone
marrow-derived interleukin 3-dependent mast cells and autonomous sublines produce interieukin
6. Immunology 67, 408-413.

20. Kemp, D.J.. Harris, AW. and Adams. J.M. (1980) Transcripts of the immunoglobulin 04 gene
vary in structure and Splicing during lymphoid development. Proc. Natl. Acad. Sci. USA 77,
7400-7404.

21. Klinken, S.P., Alexander, W.S. and Adams, J.M. (1988) Hemapoletlc lineage switch: v—raf
oncogene converts 51-myc transgenic B cells into macrophages. Cell 53, 857-867.

22. Marrack, P., McCormack, J. and Kappler, J. (1989) Presentation of antigen, foreign major
histocompatibility complex proteins and self by thymus cortical epithelium. Nature 338, 503-505.

23. McCulloch, EA. (1983) Stem cells in normal and leukemic hemopolesls. Blood 62, 1-13.

24. Mosmann, T. (1983) Rapid colorimetric assay for cellular growth and survival: application to
proliferation and cytotoxicity assays. J. Immunol. Meth. 65. 55-63.

25. Parnes, J.R., Velan, B., Felsenfeld, A., Ramanathan, L, Ferrini, U., Appella, E. and Seidman, J.G.
(1981) Mouse Bz-mlcroglobulln cDNA clones: a screening procedure for cDNA clones
corresponding to rare mRNAs. Proc. Natl. Acad. Sci. USA 78, 2253—2257.

26. Ralph, S.J., Thomas, M.L., Morton, CC. and Trowbrldge, LS. (1987) Stmctural variants of
human T200 glycoprotein (leukocyte common antigen). EMBO J. 6, 1251-1257.

27. Raschke, W.G., Baird, 8., Ralph. P. and Nakoinz, l. (1978) Functional macrophage cell lines
transformed by Abelson leukemia virus. Cell 15, 261-267.

28. Rave. N., Ckvenjakou. R. and Blodtker. H. (1979) Identification of procollagen mRNAs
transferred to DBM paper from formaldehyde agarose gels. Nucl. Acids Res. 6, 3559-3567.

29. Rigby, P.W., Dieckmann, M., Rhodes, C. and Berg, P. (1979) Labeling deoxyrlbonuclelc acid
to high specific activity in vitro by nick translation with DNA polymerase I. J. Mol. Biol. 113,
237-251.

30. Rogers, J., Early, P., Carter, C., Calame, K, Bond, M., Hood, L and Wall, R. (1980) Two
mRNAs with different 3’ ends encode membrane-bound and secreted forms of immunoglobulin a
chain. Cell 20, 303-312.

31. Saga, Y., Tung, J.S., Shen, F.W. and Boyse, EA. (1987) Altematlve use of 5’ exons In the
specification of Ly—5 isoforms distinguishing hematopoietic cell lineages. Proc. Natl. Acad. Sci. USA
84. 5364-5368.

144

32. Schwartz, R.G., Sonenshein, G.E., Bothweli, A and Gefter, ML (1981) Multiple expression of
lg k-chain encoding RNA species in murlne plasmacytoma cells. J. Immunol. 126, 2104-2108.

33. Schwartz, R.C., Stanton, L.W., Riley, S.G., Marcu, KB. and Witte, O.N. (1986a) Synergism of
v-myc and v-Ha-ras in the in vitro neoplastic progression of murlne lymphoid cells. Mol. Cell.
Biol. 6. 3221—3231.

34. Schwartz, R.G., Stanton, L.W., Marcu, KB. and Witte. O.N. (1986b) An In vino model for tumor
progression in murlne lymphoid cells. Curr. Top. Microbiol. Immunol. 132, 75-80.

35. Seidman, J.G. and Leder, P. (1978) The arrangement and rearrangement of antibody genes.
Nature 276, 790-795.

36. Sheng-Ong, G.LC., Holmes. KL and Morse, H.C. (1987) Phorbol ester-induced growth arrest
of murlne myelomonocytic leukemic cells with virus disrupted myb locus is not accompanied by
decreased myc and myb expression. Proc. Natl. Acad. Sci. USA 84, 199—203.

37. Silver, J., and Kozak, C. (1986) Common proviral integration region on mouse chromosome
7 In lymphomas and myelogenous leukemias induced by Friend murlne leukemia virus. J. Virol.
57. 526-533.

38. Southern, EM. (1975) Detection of specific sequences among DNA fragments separated by
gel electrophoresis. J. Mol. Biol. 98, 502—517.

39. Springer, T., Galfre, G., Secher, 0.8. and Milsteln, C. (1979) Mac-1: a macrophage
differentiation antigen identified by a monoclonal antibody. Eur. J. Immunol. 9, 301-306.

40. Stanton, LW., Fahrfander, P.D., Tesser, PM. and Marcu, KB. (1984) Nucleotide sequence
comparison of normal and translocated murine c-myc genes. Nature 310, 423-425.

41. Streuli, M., Hall, L.R.. Saga, Y., Schlossman, SF. and Salto, H. (1987) Differential usage of
three exons generates at least five different mRNAs encoding human leukocyte common antigens.
J. Exp. Med. 166, 1548-1566.

42. Thomas, M.L, Reynolds, P.J., Chain, A., Ben-Neriah, Y., and Trowbrldge, LS. (1987) B-cell
variant of mouse T200 (Ly-5): evidence for altematlve mRNA splicing. Proc. Natl. Acad. Sci. USA
84. 5360-5363.

43. Thomas, P. (1980) Hybridization of denatured RNA and small DNA fragments transferred to
nitrocellulose. Proc. Natl. Acad. Sci. USA 77, 5201-5205.

44. Walker, E., Warner, N.L., Chesnut, R., Kappler, J. and Marrack, P. (1982) Antigen-specific, I
region-restricted interactions in vitro between tumor cell lines and T cell hybridomas. J. Immunol.
128, 2164-2169.

45. White, J., Hasklns, KM., Marrack, P. and Kappler, J. (1983) Use of l region-restricted,
antigen-specific T cell hybridomas to produce idlotyplcally specific anti-receptor antibodies. J.
lmmunol. 130,1033-1037.

46. Whitlock, C.A., Ziegler, S.F., Treiman, LJ., Stafford, J.l. and Witte, ON. (1983) Differentiation
of cloned populations of immature B cells after transformation with Abelson murlne leukemia virus.
Cell 32, 903-911.

47. Yam, L.T., Li, CY. and Crosby. W.H. (1971) Cytochemical identification of monocytes and
granulocytes. Am. J. Clin. Path. 55, 283-290

APPENDIX 8
Introduction:

This appendix presents “data not shown“ pertinent to the manuscript that
comprises Chapter 2. These include data that demonstrated (i) that the tumors
derived from R2 possess a more transformed phenotype than the R2 cell line; 0i)
that alternative mechanisms of c-myc gene activation may occur in tumors derived
from R2; (iii) the capability of the pre-B cells to undergo further differentiation
during the process of tumorigenesis. A list of probes of oncogenes, tumor
suppressor genes, growth factor genes, and sites of frequent viral integration used
to screen for additional genetic lesions in the tumor cells is also included (Table

2).

Materials and Methods:

Cell culture. Cells were cultured under conditions described in Chen et al.
(1991)(Chapter 2). Recombinant IL—7, a generous gift of Dr. Steven Gillis
(lmmunex Corp., Seattle), was used in the culture of cells in the absence of feeder
layers.

Tumor challenges. Cells were washed twice in RPMI 1640 and were then
resuspended in the same at 8x106 cells per ml. BALB/c mice, 4 weeks old, were
injected intraperitoneally with either 0.25 ml or 0.5 ml of the cellular suspension.
Animals were observed for a maximum of 10 weeks postinjection. Animals were
sacrificed and autopsied when they become moribund or at 10 weeks. Latency

in these experiments refers to the time until the animals became moribund.

145

146

Nucleic acid analyses. All hybridization analyses were performed as described
in Chen et al. (1991)(Chapter 2). The kappa probe is a 3.4 kb Hindill restriction
fragment derived from a plasmid (pcK), which contains a genomic fragment
corresponding to kappa constant region sequences (Seidman and Cedar, 1978).
Nuclear run-on experiments. Nuclear run-on experiments were performed
essentially as described in Stewart et al. (1987) with the following modifications.
Labeling was accomplished through the incorporation of (a -32P)UTP (600
Ci/mmol; Amersham). Isolated RNA was partially hydrolyzed in NaOH in order to
allow differential analysis of initiation and elongation of c—myc transcription. RNA
was resuspended in 250 pl of 20 mM HEPES (pH 7.5), 5 mM EDTA. 62.5 ul of 1M
sodium hydroxide was added and the mixture was incubated on ice for 10 minutes.
The reaction was quenched with 125 pl of 1M HEPES (free acid; pH 5.5).
Hybridizations were carried out in 50% formamide, 5 x SSC, 5x Denhardt’s
solution, 0.1% SDS. 50 mM sodium phosphate (pH 6.8), 250149/ ml single-stranded
salmon sperm DNA, 5% dextran sulfate for at least 40 hours at 42° C.
Hybridizations were washed to a stringency of 0.1 x SSC at 55° C. Single-stranded
M13 probes were a generous gift of Dr. Alain Nepveu (Ludwig Institute for Cancer
Research, Montreal). The exon 1 c-myc probe contained the 0.5 kb BamHl-Bglll
fragment of the murine crmyc gene. The exon 2 and 3 probe contain a 3.4 kb
BamHl-Hindlll fragment of the murine crmyc gene. The antisense control
contained the 145 bp HaellI-Hindlll fragment at the 5’ end of the murine c—myc
gene. The rGAPDH probe contained the 1.3 kb cDNA of rGAPDH.

mRNA turnover assay. R2 cells and tumor cell lines were cultured as described

in Chen et al. (1991)(Chapter 2), and maintained in an actively growing state by

147

addition of fresh media 24 hours prior to harvest. Cells were harvested, and
counted with trypan blue staining to assess their viability. 1.5 x 10" cells were
resuspended in 60 ml standard medium supplemented with Actinomycin D at a
concentration of 50 micrograms per ml, and then split into three portions. The first
20 ml were immediately subjected to cytoplasmic RNA preparation by methods
described in Chen et al. (1991)(Chapter 2), while the other two portions were
incubated for either 1 /2 hour or 1 1 /2 hours after the addition of Actinomycin D.
RNA of R2 and the tumor cell lines isolated at each time point was analyzed by a
Northern blotting procedure as described in Chen et al. (1991)(Chapter 2).
Northern blots were sequentially hybridized to probes for c-myc and rGAPDH. The
stability of c-myc RNA in R2 and the tumor cell lines was assessed by their relative

decrease in abundance over time by the method described in the table legend.

Results:
The tumors possess a more transformed phenotype than their parental cell
line.

As described in the Chapter 2, the infrequent occurrence and latency of
tumors derived from the R2 cell line raised the question of whether tumor
progression required the acquisition of genetic lesions in addition to v-Ha-ras
expression. Other explanations for the weak tumorigenicity of the R2 cell line are
plausible. First, it is possible that R2 cells are normally able to grow into tumors
if they do not acquire other genetic lesions, which may somehow stimulate the
host immune system to eliminate or inhibit their growth. This scenario is unlikely,

since Schwartz et al. (1986b) found R2 to be weakly tumorigenic in athymic nude

148

mice. Second, the infrequent occurrence may relate to the variation of epigenetic
factors among hosts, and the long latency may only reflect the in vivo slow growth
properties of R2 cells. In order to ascertain whether the tumor cells had gained
growth properties consistent with tumor progression, the growth properties of early
passage tumor cell lines were compared to the R2 parental line. Two growth
properties were tested: the ability to grow in vitro independent of adherent feeder
cells and the ability to generate tumors in syngeneic mice.

R2 had previously been found to be dependent on an adherent cellular
feeder layer for growth (Schwartz et al., 1986a). All of the tumor cell lines grew
well on feeder layers. In order to test whether the tumors had gained some level
of growth factor independence, the tumor cell lines and R2 were cultured in the
absence of feeder layers. While R2 was incapable of growth in the absence of a
feeder layer, all of the tumor cell lines exhibited growth (Figure 1). This is most
consistently the case at higher cell density.

Since lL-7 has been reported to be required for the growth of pre-B cells
(Namen et al., 1988), it was of interest to test the lL-7 responsiveness of R2 and
the tumor cell lines derived from it. R2 (Figure 1) and other v-Ha-ras-transformed
pre-B cell lines (data not shown) are both dependent on and dramatically
responsive to lL-7 for growth in the absence of a feeder layer. Several of the
tumor cell lines, although independent of cellular feeder layers and lL-7 for growth,
retained some responsiveness to lL-7 (Figure 1).

The abilities of the tumor cell lines to form tumors in syngeneic BALB/c mice

were compared to that of the parental R2 cell line. Tumor cell lines 1, 2, 3, 4 and

149

 

 

/f

 

 

 

 

 

 

(ILLS

 

-./////‘

.”*.._._._. "—

 

 

 

 

 

 

‘03

M
‘4
5.1
.

*m-U

*omm OIL?

 

Figure 1. Growth of the R2 and seven tumor cell lines (1-7) in the presence and
absence of lL-7. Cells were removed from feeder layers and resuspended in fresh
medium with and without 10 units/ml IL-7. Either 5x10a or 5x104 cells were plated
in suspension in 1 ml. On day 4, viable cells were counted in the presence of
trypan blue. The graph plots the average total cells of duplicate cultures.

150

Table 1. Tumor challenges in BALB/c mice.

 

Cell Line Frequency (animals) Average Latency Till Moribund (days)

 

2x106 cells“ 4x106 cells‘ 2x106 cells‘ 4x106 cells“
R2 1 /4 1 /4 81 70
1 1 /5 5/5 44 25
2 2/5 5/5 41 24
3 1 /4 5/5 44 27
4 4/4 ND. 22 ND.
5 ND. 3/5 ND. 50
6 ND. 1 /5 ND. 22
7 2/2 3/4 16 45

 

N.D. = Not determined
‘Two inocula were used.

7 were dramatically more tumorigenic than R2 (Table 1). Tumor cell lines 5 and
6 had lower frequencies of tumor induction than the other tumor cell lines. This
is consistent with their lack of growth in the absence of a feeder layer in
experiments of longer duration than those presented in Figure 1 (data not shown).
All of the mice challenged with tumor cell lines did have a shorter latency till
moribund than did those challenged with R2 (T able 1).

The data on feeder layer independence and tumorigenicity demonstrate that
the tumor cell lines have acquired transformed growth potential which exceeds that
of their parental cell line.

To examine the molecular events that underlie the transformed phenotypes of
tumor cells, we have used a set of probes of oncogenes, tumor suppressor genes,
growth factor genes, and sites of frequent viral integration (Table 2) to test for their

involvement. Gene rearrangement or altered expression of these genes or sites

151

.co_mmocaxo

.anx..nx ..uom..om ....nc.=.n=

.9505? .3 38.35.39.

<zaL0m

._:Eamu_zn

noc_eaxo an: «

.>¢ouu.>¢

Loan» >.cou

..¢oum ..¢
b

636: o... .8533 9.0... m..3uo> «5 so...) 5.... 2.3.8 .3 «35.39. 2.53.5.0

.303 505.8: co 6339.98

.933 <5. .8 5232.3 0.538.023.53. «33.3., 5...: 3o... 53.
ucueoucoccao.

.9533 ..o 303 50533 :0 tenuous“.

.n
<zo.o

"comuo_>o.nn<

.n Lao:u_.04

caufip...o uo mxmcounx
moo..... .3 5523.6
.1 .3395

32.... so 2695..
.o.z 6:6.6666

.o.z uca.oaou
.32.... 63:53.
38...... .3355

.a cuscom_m
.3 cgdommw
.: .65..

.: o.o..oas.
.x 0....485.
.66 amaz<
.o as..:
.x 6.66
gene.
xcoLL
uu.<
Loncmm
..s.<

.u Lump—m
.:.¢ Lacuna..o.

.x.¢ Luuugtoa

.o cass.m

.o ouu_3

OUSLOwflu

.a.h

0

6.9.3 mm .8 c3833.... .855 of E 3302.. 3 >2. 35 355

o_u:n
.636
6625

«came

-n¢o¢
opm.q
can
-n¢oa
-nama

.4..~Lxmua
umsma.c moo
n_aumo:.mn
oc>moua
oc>moua
n.u:n
Nwmama
magma
-n¢ma
n—uaa
npuan
mmnmmq
m.~.nau:a

.ouuo>

n=-_umm ax n.o
.uma.n= ax o.«

.c ax ~.o

_¢._umm ax c.—

.xm ox no.o
_¢-m: ax o."
n: ax m.n
n: nx o.~

__:>¢-_¢ ax N.—

.¢ 6. mm.o
.3 a. o.n
.a a. 4.~

.:¢ 6. L.~
.g a. 5.8

.ocx ax ~.n

.36. a. m.o

.um a. m.o

.oUz-_uma nx 0.9

__a>< as mm.o

.u ox o.~
.m nx e.~
_¢ ax m.o

_:m-_u.u nx N.N

.a 93 n.p
.m 93 ~.p
__:>a nx 0.9

_¢._nx ax o.~

>¢..=a
.a
.xu..cax
.cax
>¢..¢

.cax .>¢..=m..¢

AucoEmacevonocm

33:33 «>333 05 Lo» 9.29.9.5 E

.zu
.um..=a
.zo

.¢

.zm..a
<3... .5...
<2“

.a..=m
.s..=m
.c..=m
<za..¢
<2a..¢..=m
<2¢..¢..=o
<2¢..¢

>¢

.u..=m
.a..=m

.zu

oe>~c
n m

m.~
c.s
o.o~A
o.o~
o.-
o—A
oémA
«.m—
m.P~

o.o~A
o.o~w
o.o~A
c.opx
c.o~A
o.o~A

o.—P
o.oa
c.-
bo.o~A
no.o~a
o.¢~A
o.c—A
o.-A
c.o~A
o.opx
o.-A

“savoucom

p-_;a
—._>o
P-u>n
~.E_q
—-e_o
¢-_>.e
n-_>.e
~-_>.e
—-_>.E

xae
ommm
u< .mmo.uu=x
nmo
ax
u_x
h-.—
—-mmu
m-.—
mmuzu
woc-u
n>E-u
u>E.z
msw->
xu.
xuc
o.m.>
.no.>

acme

vow: 9.0: =5 «39:. .~ 033

152

was examined by Southern and Northern blotting analyses. Tumor 4 showed
altered expression of several genes, and was described in detail in Bretz et al.
(1992)(Appendix A). The c-myc gene was aberrantly expressed in four of the
seven tumor cell lines tested (two of these also showing retroviral integration at c-
myc) as described in Chen et al. (1991)(Chapter 2). No gene rearrangement was
found in about a 20 kb region surrounding most of the other oncogenes or
frequent viral integration sites tested in the tumor cells. No altered expression of
other tested genes was detected in the tumor cells, with the exception of Tumor

4 (see Chapter 4).

Studies of mechanisms of elevated c-myc expression

We have observed a 3-fold increased level of c-myc mRNA in tumors 1 and
3, presumably as a result of the proviral insertion, and 4-fold elevated levels of
c-myc mRNA in tumors 6 and 7 in the absence of any obvious genetic alteration
(Figure 6 in Chapter 2). The c-myc mRNAs were all approximately 2.4kb in length,
suggesting normal promoter usage and an unaltered RNA structure (Figure 6 in
Chapter 2). RNAse protection studies confirmed normal promoter usage (data not
shown).

Nuclear run-on experiments were performed in order to examine (i) whether
the elevated c-myc mRNA levels reflected increased transcription and (Ii) the role
of transcriptional attenuation (Bentley and Groudine, 1986) in modulating c-myc
expression in these tumors. Transcriptional initiation of c-myc was evaluated by
hybridization of labeled RNA to a probe for exon 1 of c-myc, while elongation of

transcription was evaluated by hybridization to a probe for exons 2 and 3 of c-myc.

153

R2
MYC EXON I

MYC EXON 2&3

ANTISENSE
rGAPDH

 

TRANSCRIPTION
LEVELS

MYC EXONI 1.0 3.1 3.1 1.6 1.0

MYC EXON 28:3 1.0 2.0 1.5 2.0 1.2

Figure 2. Nuclear run-on transcription. Labeled nuclear run-on products from nz
and tumors 1, 3, 6 and 7 were hybridized to probes for c-myc exon 1, c-myc exons
2 and 3, antisense upstream of c-myc and rGAPDH. Relative levels of transcription
were determined by densitometry and normalized to rGAPDH.

154

Densitometric values were normalized to the amount of hybridization detected to
a probe for rGAPDH. Background hybridization was evaluated with a probe for
anti-sense transcripts upstream of the c-myc locus. In comparison to R2, tumors
1 and 3 showed about 3-fold increased transcriptional initiation and about 1.5 to
2-fold increased elongation (Figure 2). Since these values average the activity of
the retrovirally activated allele(s) with the normal c-myc allele, they are likely to
underestimate the degree to which the presence of the MoMuLV LTR affects c-myc
transcription. The elevated transcription detected in this experiment is consistent
with the 2 to 3-fold elevated steady state levels of c-myc mRNA observed by
Northern analysis (Figure 6 in Chapter 2). Since initiation of transcription increases
more than elongation, increased initiation rather than decreased attenuation is the
probable mechanism by which retroviral integration increases c-myc mRNA
expression. Tumor 6 showed about 1 .5-fold increased initiation and about 2-fold
increased elongation (Figure 2). This is consistent with the 3-fold increased steady
state levels of c—myc mRNA observed for tumor 6 in Figure 6(of Chapter 2). On
the other hand, tumor 7 did not show significant increases in either parameter.
The data offer no evidence for a transcriptional mechanism leading to the elevated
c-myc mRNA levels observed for tumor 7 (Figure 6 in Chapter 2).
Determinations of the half-life of c-myc mRNA in the presence of
actinomycin D did not reveal significant differences between R2 and tumors 1, 3,
6, and 7 (Table 3). R2 and tumor cell lines were expanded and maintained in an
actively growing state, and then treated with 5 microgram/ml Actinomycin D.
Portions of cells were then harvested at three time points: 0 hour, 1 /2 hour, and

1 1/2 hour after adding Actinomycin D. RNAs of each cell line were isolated at the

155

three time points and subjected to Northern hybridization analyses with a c-myc-
specific probe and a rGAPDH gene probe. Densitometric measurements were
taken for both c-myc and rGAPDH specific signals on the resulting autoradiogram.
Since the level of rGAPDH remained relatively unchanged among R2 and tumor cell
lines during the 1 1 /2 hour time period, we normalized the level of c-myc mRNA
in each time point to that of rGAPDH at the same time point under the assumption
that the degradation rate of rGAPDH mRNA is the same in R2 and tumor cell lines.
The stability of c-myc mRNA was determined by the assessment of the relative
half-life assigned to R2 and each tumor cell line by the method described in the
table legend. The relative T 1 /2 of c-myc mRNA is 32 minutes, whereas those of
tumor cell lines ranged from 28 to 40 minutes. Therefore, we concluded that
alterations in the stability of c-myc mRNA did not play a role in the increased c-myc

mRNA levels observed in these tumor cells.

Table 3. Relative stability of c-myc mRNA in R2 and tumors

 

cells relative T 1 /2 (min)*

 

n2 32(29,34)
T1 40(35,4e)
T3 36(31,42)
T6 28(26,29)
T7 38(35,41)

 

*These values were determined by the following calculation procedure. The signal
of c-myc was normalized to rGAPDH for a loading control, and then the resulting
values were plotted on a semilog graph (log of hybridization intensity versus time
(minutes)). The relative T 1 /2 was determined by the time required for the loss of
half of the original hybridization intensity.

156

Observatlon of light chain rearrangement In tumor cell llnes

One other interesting observation obtained in these studies was the
detection of immunoglobulin light chain rearrangements in most of the tumors
(Figure 3). As mentioned in Chapter 2, all tumor cell lines possess the same
immunoglobulin (lg) heavy chain rearrangements as that of the parental R2 cell line
(Figure 4). A southern blot hybridization analysis with a probe specific to the J
region of lg heavy chain revealed hybridizations to a 6.0 kb germ line fragment
(Figure 4, lane 1) on EcoRl digested liver DNA and to one or two rearranged
fragments (7.0 kb and 3.5 kb, Figure 4, lane 2-9) on DNA of the R2 and tumor
cells. DNA from the original tumor sample (a portion of which may consist normal
cells) of tumor 6 was used in the analysis, therefore, a germ line fragment was also
observed (lane 8). When the same DNAs were digested with BamHI, and probed
with a DNA fragment from the constant region of the kappa chain, all the tumor cell
lines showed dramatically different banding patterns (Fig. 3, lanes 2, 4, 5, 7, and
8) from the germ line fragment observed in the R2 cell line and liver (Fig. 3, lanes
1 and 9, respectively). Tumor cell lines 2 and 5 (Fig. 3, lanes 3 and 6,
respectively) both retained the germ line fragment, but a portion of these cells may
already have undergone kappa chain rearrangements as judged by the presence
of some minor hybridized bands. No lambda light chain gene rearrangement was
found in the R2 and tumor cell lines (data not shown). Immunopreoipitation
experiments with both anti-mu and anti-kappa chain anti-sera revealed that the
rearrangements successfully produced authentic immunoglobulin heavy and light

chain proteins(data not shown).

157

KAPPA
32.1.2fi34 56 7L

1 ;: ."'-'
235 j 5;
c

9.4 >, .

6.6 b, In»

4.4 ,-

2.0>

Figure 3. Kappa chain gene rearrangement. BamHl digested R2 (lane 1), liver
(lane 9), and tumor cell DNAs (lane 2-8) were subjected to gel electrophoresis
through 0.8 % agarose, blotted to nytran paper, and hybridized with a nick-
translated probe of a kappa-chain-specific DNA fragment. Size markers are the
positions of an ethidium bromide-stained Hindlll digest of bacteriophage lambda

and are denoted in kilobases.

158

1 2 3 4 5 6 7 8 9
23>!” * A.
9.4»
65> v " T "" u

'41»

23>
2.0V

Figure 4. Mu chain gene rearrangement. EcoRl digested liver (lane 1), R2 (lane
2), and tumor DNAs (lane 3-9) were electrophoresed through a 0.8 % agarose gel.
The blot was probed for the J region of mu heavy chain gene. Size markers are
the positions of an ethidium bromide- stained Hindlll digest of bacteriophage
lambda and are denoted in kilobases.

159

Discussion:

The data presented here confirm that the R2 tumors had acquired a more
transformed phenotype as evidenced by their high tumorigenicity and their growth
factor independence. The fact that these tumors were not uniformly tumorigenic
(i.e. did not cause tumors in all the tumor-challenged mice) suggested the
possibility of tumor regression through a variety of mechanisms. These may
include stimulation of host immune system perhaps by generating a recognizable
tumor specific antigen, or growth inhibition by some intrinsic factors that evolved
in tumor cells during the course of tumor challenge. The intermediate transformed
phenotypes of these tumors may provide tools for studying tumor progression.

The regulatory mechanisms responsible for increased levels of c-myc mRNA
probably occur at both transcriptional and posttranscriptional levels as described
by Klein and Klein (1985). In the case of tumors 1 and 3, MoMuLV had integrated
immediately 5’ to c-myc in a reverse transcriptional orientation, suggesting
enhancer activation. Nuclear run-on experiments showed that these tumors with
MoMuLV integration near c-myc displayed increased levels of transcription
consistent with the increased steady state level of c-myc mRNA. The nature of the
events leading to elevated levels of c-myc mRNA in the other two tumors is
unclear. Tumor 6 displayed increased transcription, but no gross structural
alterations were observed on either the DNA or RNA levels. The attenuation of
transcription in the first exon of c-myc was also unaffected. Tumor 7 showed no
increase in transcription and, as in the case of tumor 6, there were no structural
abnormalities observed in either the c-myo gene or its transcription products.

Southern blot analyses of genomic restriction fragments extending approximately

160

15 kb upstream and 30 kb downstream of c-myc detected no rearrangements in
either tumor 6 or 7 (data not shown) suggesting that retroviral integration is unlikely
to play a direct role in the activation of c-myc in these tumors. On the other hand,
MoMuLV integrations at MIvi-1 and MIvi—4 can affect c-myc over a long distance
(Lazo et al., 1990). However, we have not detected rearrangements at these sites
(data not shown). The increased c-myc mRNA level in tumor 7 may be due to
posttranscriptional mechanisms such as increased maturation rate of poly A“ RNA
and increased transport rate through the nuclear membrane. Studies of mRNA
half-life show that these mechanisms are not involved.

The mechanism by which c-myc activation promotes tumorigenesis is
unclear. Expression of v-myc has been reported to abrogate the dependence of
lymphoid and myeloid cell lines on lL-2 and lL-3 for growth (Payne et al., 1982).
Since all the tumors with and without c—myc activation show reduced dependence
on lL-7, it is difficult to conclude that a similar causal relationship exists between
c-myc activation and alleviation of IL—7 dependence.

The fact that pre-B cells were able to continue to undergo light chain
rearrangements and to successfully produce kappa chains suggests that
introduction of the v-Ha-ras oncogene does not block B cell differentiation. It also
implies that B cell lymphomas may result from a gradual acquisition of genetic
lesions by a precursor B cell that can progress along the differentiation pathway.
Similar results had been shown in Eli-myc transgenic mice, in which tumors of
both pre-B and B cells were obtained (Adams et al., 1985; Harris et al., 1988). On
the other hand, the majority of tumors induced by v-abl seem to be frozen at a

pre-B stage (Sklar et al., 1974; Sklar et al.,1975; Weimann, 1976). Interestingly,

161

long term cultures (as long as the period of tumor challenge) of the R2 cell line did
not show evidence of light chain rearrangement in vitro (data not shown). It seems
that some in vivo factors may be important for light chain rearrangement in
addition to a successful heavy chain rearrangement. In vitro triggering of light
chain rearrangements of R2 cells may help to identify these factors, and provide
more insights on B cell differentiation.

Although no gross gene aberrations in genes other than c-myc were
detected in the R2-derived tumor cells, we can not exclude their roles in the
additive transformed phenotypes of these cells. Mutations of other types such as
point mutation, small deletion, or viral integration in regions outside of the tested
regions may have occurred. Further investigations will be required to elucidate the

secondary events in the tumor progression of v-Ha-ras transformed pre-B cells.

162

References:

Adams J.M., A.W. Harris, C.A. Pinkert, LM. Corcoran, W.S. Alexander, S. Cory, R.D. Palmiter, and
R.L Brinster. (1985). The c-myc oncogene driven by immunoglobulin enhancers induce lymphold
malignancy in transgenic mice. Nature 318, 533-538.

Bentley D.L, and MA. Groudine. (1986). A block to elongation is largely responsible for decreased
transcription of c-myc in differentiated HL-60. Nature 321, 702-704.

Bretz J. S.-C. Chen, D. Redenius, H.-L Chang, W.J. Esselman, and RC. Schwartz (1992). Uneage
switch macrophages can present antigen. Develop. lmmunol. in press.

Chen S.-C., D. Redenius, and RC. Schwartz. (1991). Tumorigenesis of a v-Ha-ras-expresslng pre-B
cell line selects for c-myc activation. Biochem. Biophy. Res. Comm. 178, 1343-1350.

Cuypers H.T., G. Selten, W. Quint, M. Zijlstra, E.R. Maandag, W. Boelens, P. van Wezenbeek, C.
Mellef, and A. Bems. (1984). Murine leukemia virus-induced T cell lymphomagenesis: integration of
proviruses in a distinct chromosomal reglon. Cell 37, 141-150.

Graham M., J.M. Adams, and 8. Cory. (1985). Murine T lymphomas wlth retroviral inserts In the
chromosomal 15 locus for plasmacytoma variant translocations. Nature 314, 740-743.

Lazo P.A. , J.S. Lee, and RN. Tschilis. (1990). Long-distance activation of the myc protooncogene
by provirus insertion in MIvi-1 or MIvi-4 in rat T-cell lymphoma. Proc. Natl. Acad. Sci. USA 87,
170-173.

Mucenski M.L., B.A. Taylor, J.M. lhle, J.N. Hartley, H.C. Morse, Ill, N.A. Jenkins, and N.G. Copeland.
(1988a). Identification of a common ecotropic viral integration site evi-1 in the DNA of AKXD murlne
myeloid tumors. Mol. Cell. Biol. 8, 301-308.

Namen A.E., A.E. Schmierer, C.J. March, R.W. Overell, LS. Park, D.L Urdal, and D.Y. Mochizukl.
(1988). B cell precursor growth-promoting actlvlty: purification and characterization of a growth
factor active on lymphocyte precursors. J. Exp. Med. 107, 988-1002.

Payne G.S., J.M. Adams, and HE. Varmus. (1982). Multiple arrangements of viral DNA and an
activated host oncogene in bursal lymphomas. Nature 295, 209-213.

Schwartz R.C., LW. Stanton, S.C. Riley, KB. Marcu, and ON. Witte. (1986a).
Synergism of v-myc and v-Ha-ras in the In vitro neoplastic progresslon of murlne lymphoid cells.
Mol. Cell. Biol. 6, 3221-3231.

Schwartz R.C., LW. Santon, S.C. Riley, KB. Marcu, and ON. Witte. (1986b). An in vitro model for
tumor progression in murlne lymphoid cells. Curr. Topics Micro. Immunol. 132. 75-80.

Sklar M.D., B.J. White, and WP. Rowe. (1974). Initiation of oncogenic transformation of mouse
lymphocytes in vitro by Abelson Leukemia Virus. Proc. Natl. Acad. Sci. USA 71, 4077-4081.

Stewart C.J., M. Ito, and SE. Conrad. (1987). Evidence for transcriptional and post-transcriptional
control of the cellular thymidine kinase gene. Mol. Cell. Biol. 7, 1156-1163.

Tsichlis P N., G. Strauss. and EH. Liu. (19833). A common region for proviral DNA integration in
MoMuLV-induced rat thymic lymphomas. Nature 302, 445-449.

Tsichlis P.N., P.G. Strauss, and CA. Kozak. (1984). Cellular DNA region involved in induction of
thymic lymphoma(MIvi-Z) maps to chromosome 15. Mol. Cell. Biol. 4, 997-1000.

Summary and Discussion

The goals of my thesis study were to search for oncogenes that were
capable of cooperating with the v-Ha-ras oncogene in transforming murine B cells,
and to examine the differentiation status of tumor cells. Two model systems were
used. The first model system tested for genes that could cooperate v-Ha-ras in
transforming murine B cells in vivo, whereas the second model system examined
this issue in vitro.

In the first model system, we found that tumors had acquired more
malignant growth properties by comparing the growth characteristics of R2, a v-Hr-
ras-expressing pre-B cell line, and tumor cells derived from it. The tumor cell lines
exhibited greater tumorigenicity and shorter latency in tumor development than the
parental R2 cell line. In addition, tumor cell lines had acquired growth factor
independence. Two major molecular events have been shown to be associated
with these transformed phenotypes: an incgased number of virus-associated
restriction fragments on southern blot analysis, and alteration of c-myc expression.

First, studies on the viral integration patterns of R2 cells and tumors
suggested that 6/7 tumors were an outgrth of a single R2 subclone, which
contained unique viral related fragments. These unique virus-associated fragments
could have resulted either from recombination of the original integration sites of the
parental cell line, or from new viral integrations, or both. Oncogene activations had
been found previously through each mechanism. We found two tumors that had
activated c-myc by viral integration.

Many, if not all, proviruses near c-myc have sustained deletions of the viral

genome, probably as a result of homologous recombination between the two

163

164
two LTRs. Sometimes little more than a solitary LTR is left intact (Robinson and

Gagnon, 1986). A mechanistic explanation of the requirement for these deletions
was given by Cullen et al. (1984), who showed that transcription starting in the 3’
LTR of an intact provirus is quenched by transcription driven by the 5’ LTR. By
artificial termination of transcription within the provirus, or by the deletions as found
in tumors, the activity of the 3' LTR becomes sufficient to act as a strong promoter
(Fujisawa et al., 1985) and thus to activate oncogenes. The c-myc rearrangement
in two of the R2 tumor cell lines resembles the above observation in that the
genome of MoMuLV integrated in the c-myc gene in these two tumors has
undergone a further rearrangement (data not shown). This viral genome
rearrangement was assessed by a similar approach to that described in Figure 4
(Chen et al., 1991), with the exception that restriction mapping utilized enzymes
that cut inside the viral genome rather than in the LTRs. However, we do not know
the significance of this viral genome rearrangement, since the activation of c-myc
gene in this case is presumably through an enhancer activation rather than a
promoter activation.

The reintegration of MoMuLV and avian leukosis virus (ALV) into the cellular
genome of infected cells has been well documented in the tumors they generate,
and are thought to be causative for tumor formation through the activation of
nearby cellular oncogenes (Jaenisch, 1976; Payne et al., 1982). Studies on
transgenic mice carrying a single copy of MoMuLV DNA sequences has shown an
increase to two MoMuLV specific DNA copies per haploid mouse genome in
preleukemic tissues, and a further increase to 3-4 copies in leukemic tissues

(Jaenisch, 1979). This amplification of MoMuLV is seen in target organs but not

165

in nontarget organs, and thus appears to be related to leukemic transformation.
At least one tumor specific virus-related fragment (bone-1) in our studies
belonged to the aforementioned class (reintegration), since DNA of the parental R2
cell line showed a germ line pattern at this locus. The nature of other tumor
specific virus-related fragments is unknown. We were unable to detect any
transcriptional activity over a 20 kb region in the vicinity of the bonc-1 locus.
However, sequence homology of bonc-1 was found to genomes of many
mammalian species and chicken. The high conservation of bone-1 suggests that
this region has an important function. On the other hand, activation of
protooncogenes through distal provirus integrations (270 kb) has been observed
(Lazo et al., 1990), and it is possible that a putative oncogene is located beyond
the 20 kb region studied. Whether this putative gene would indeed be able to
cooperate with v-Ha-ras gene in course of tumor progression, and whether it would
have B cell specificity in the induction of neoplasia remains to be tested.
Although I was not able to definitively identify a novel oncogene during the
period of my thesis study, the fact that MoMuLV was highly mobilized in infected
cells provides a method to identify new protooncogenes involved in hematopoietic
diseases. In retrospect, modifications of the insertional mutagen (MoMuLV) might
have been added to facilitate the identification of viral integration sites. These
modifications might include the addition of a bacterial supF gene or some other
bacterial indicator gene to the MoMuLV sequences. Tagging with a supF gene
would allow the construction of an integration library in a amber mutant 1. phage
that could contain only clones with sequences of supF integrated proviruses and

their flanking cellular loci as described by Shih et al. (1988).

166

The second molecular event found among the tumor cell lines derived from
the R2 cell line was the increased steady state level of c-myc mRNA in four of the
seven tumors. Studies of c-myc activation have revealed that the c-myc gene can
be deregulated by several mechanisms: increased transcription, increased
elongation, and perhaps increased transport through nuclear membranes (for
review see Klein and Klein, 1985). Our findings are consistent with varied
mechanisms for deregulation of c-myc. Two of the four tumors studied here had
suffered a MoMuLV integration at about 150 bp 5’ of the first exon of the c-myo
gene in a reverse transcriptional orientation, presumably driving the expression of
the c-myc gene through enhancer activation and increased transcription. The
exact mechanism of c-myc activation of the other two tumors is unknown, although
one tumor appears to be transcriptionally activated.

Our finding of MoMuLV integration into the c-myc gene in B cell lymphomas
is novel. MoMuLV integration in the c-myc locus has been commonly found in
murine thymomas (Selten et al., 1984). Feline leukemia virus integration near c-
myc has been observed to occur only in T cell lymphoma (Neil et al., 1984).
Mucenski et al. (1987) reported that viral integration in the c-myc locus occurs only
in murine T cell lymphomas from studies on a panel of B cell and T cell
lymphomas from AKRD mice. Our finding suggests that the tissue tropism of
MoMuLV (and perhaps other MLVs) in naturally occurring tumors O.e., thymomas)
may not be attributable to tissue specific integration sites. The observation that
viral integrations into pim-1, originally identified as a common T cell integration site,
were also detected in pre-B, and B cell lymphomas supports this idea (Mucenski

et al., 1987). Although some virus integration sites such as mlvi-1, mlvi-2, fis-1,

167

and pvt-1 have been shown only found in T cell lymphomas (Mucenski et al.1987),
the significance of these loci in tumorigenesis is not clear since no gene products
have been found.

C-myc gene alterations have been associated with many B cell neoplasias
and have been proven to be able to initiate or promote B cell malignancy in
transgenic mice (Adams and Cory, 1991). Cooperation of the c-myc gene with the
v-Ha-ras gene in transformation has been shown in both fibroblasts (Land at al.,
1983a) and B lymphoid cells (Schwartz et al., 1986a). The occurrence of c-myc
alteration in tumor challenges with a v-Ha-ras-expressing pre-B cell suggests the
validity of this model system in identifying genes that are associated with the in
vivo tumor progression of B cell neoplasia.

Besides retroviral integration and c-myc activation, we did not detect other
gene rearrangements in tumors when probes of other known oncogenes, tumor
suppressor genes, growth factor genes, and flanking genes of frequent viral
integration sites (as listed in Appendix B) were used in southern hybridization
analyses. Another approach applicable to the search for non-virally related
secondary events involved in tumor progression of these tumors is the transfection
of tumor DNA into indicator cells. If this approach were taken, care would have
to be taken in choosing a appropriate indicator cell, since the tumor cells all
posses a v-Ha-ras gene which is known to be dominant in focus formation assays
in NlH3T3 cells.

Two other interesting features were observed among the v-Ha-ras tumors:
a continuation of B cell differentiation and a lineage switch to macrophage-like

cells. As discussed in appendix B, most tumors derived from the pre-B cell line,

168

R2, had undergone kappa chain rearrangements and produced kappa chain
proteins. However, we did not observe detectable light chain rearrangements in
R2 cells that had been cultured in vitro for a similar or longer period of time as that
required for tumorigenesis. This discrepancy was particularly interesting to us,
since it is generally thought that light chain rearrangements will be triggered once
a successful heavy chain rearrangement has been generated in the same cell.
These results suggested that additional factors are required for light chain
rearrangements besides the production of an authentic heavy chain protein. This
factor was absent in the Whitlock-WItte culture system but present in mice. A
similar lack of in vitro light chain rearrangement was also found in another v-Ha-ras
transformed cell line, R1, which was generated by the same procedure as that of
R2. Tumors derived from R1 cells also exhibited light chain rearrangement (data
not shown), suggesting the generality of this phenomenon.

On the other hand, it is also possible that some inhibitory factors may be
present in the Whitlock-Witte culture system to prevent differentiation of these pre-B
cell lines. A similar phenomenon has been demonstrated by Alt et al. (1981) on
cell lines derived after in vitro infection of bone marrow or fetal liver cells with
Abelson murlne leukemia virus (A-MuLV). Most, if not all, of the cell lines were pre-
8 cells and they rarely went through light chain rearrangement, even after a long
period of cultivation in vitro. In contrast, tumors induced by A-MuLV often display
light chain rearrangements (Sklar et al., 1975; Weimann, 1976). However, the fact
that certain subclones of pre-B cell lines established in the Whitlock and Witte are
capable of undergone light chain rearrangements in vitro (Denis and Wm, 1986)

argues against this possibility. Perhaps, this in vitro inhibition of light chain

169

rearrangements in both v-Ha-ras and v-abl expressing pre-B cells are associated
with these oncogene expression. An in vivo factor may bypass this inhibitory
effect of oncogene expression in tumors carrying the same oncogenes. At any
rate, these v-Ha-ras transformed cell lines provide a good model system to search

for factors that may interfere or participate in the differentiation process of B cells.

Another interesting feature of these studies was the discovery of a lineage
switch of certain B cell tumor lines to macrophage-like cells. Although other
groups have reported a similar phenomenon (Klinken et al., 1988; Davidson et al.,
1988; Borzillo et al., 1990), the studies presented here uniquely establish the
functionality of these lineage switch macrophages: LPS-induced cytokine release
and the ability to present antigen. These macrophage cell lines may be the first
macrophage lines ever reported to present antigens in vitro (John Cohen, U.
Colorado, personal communication). Therefore, they are potentially useful in
research areas such as of B cell activation, clonal expansion of T cells, and other
cell-cell interactions that require macrophages with an antigen presenting capacity.
They and their antecedent cells (the R1 and R2 cell lines) may prove useful in
studying the lineage determination of lymphoid and myeloid cells.

A second model system which involved the introduction of oncogenes into
v-Ha-ras transformed cells or the co-infection of bone marrow cells with
retroviruses carrying a v-Ha-ras oncogene and another oncogene of interest
provided a convenient way to test the oncogenic potentials of various oncogenes.
With this model, we have shown that expression of lL-7 was not sufficient to cause

a fully tumorigenic phenotype in a v-Ha-ras expressing cell line. These findings

170

were consistent with what had been reported by Young et al. (1991) and Overell
et al. (1991). Both groups reported that IL-7 can participate in the tumorigenesis
of murine B cell neoplasia, with the requirement of other secondary events. The
insufficiency of lL-7 expression to confer tumorigenicity may imply the existence of
a tightly regulated mechanism of lL-7 induced growth in vivo.

Another important finding from this model system was the discrepancy
between the highly tumorigenic phenotype of cells transformed by a co-infection
of bone marrow with v-Ha-ras and myc, or v-Ha-ras and lL-7 and the non-
tumorigenic phenotype of cells sequentially infected with the same two
oncogenes. The clonality of the outgrowing cell lines in the co-infection of bone-
marrow and their high frequency and short latency for generation of tumors
contrast sharply with the properties of v-Ha-ras cell lines superinfected with v-myc
or lL-7 viruses. The clonal outgrowths of co-infections of bone marrow may
represent the dominance of a rare and more highly transformed cell over many
other cells carrying one or both oncogenes in this heterogeneous population.
While lL-7 and v-Ha-ras expression may contribute to the transformed phenotype,
the cooperation of other rare oncogenic events is critical to the tumorigenic
phenotype. The more homogeneous population in the sequential infection of a
single cell line does not as easily allow selection for other rare events. Therefore,
interpretation of experiments of this sort should be made carefully. To ensure that
the synergistic effect of two oncogenes in any particular system is indeed derived
from the two oncogenes and not other "co-selected events“, the two oncogenes
should be sequentially introduced into a cell line rather than a heterogeneous

population. Cells with an intermediate transformed phenotype derived from the

171

these experiments can be used as targets for testing the involvement of other
oncogenic events.

in summary, results from both of our model systems for B cell neoplasia
suggested that multiple events (at least three) were required for the induction of
a fully malignant B cell phenotype. Both the c-myc gene and lL-7 gene were
capable of cooperating with the v-Ha-ras oncogene in transforming murine B cells.
However, to achieve a fully tumorigenic phenotype, a tertiary event(s) was required
in both instances. The two models systems provide the opportunity to identify this
tertiary event (3) as well as other independent steps involved in the tumor
progression of B cell neoplasia. They also provide good tools to study B cell
differentiation and lineage determination of myeloid and lymphoid cells.

In the future, if particular gene alterations are found to be restricted to the
development of murine B cell neoplasia, and a correlation is found between murine
and human B cell neoplasia, detection of early-stage neoplastic cells may be
possible. Detection protocols may operate through the identification of mutant
(quantitatively or qualitatively) gene products secreted into the blood or other body
fluids, or through the detection of antibodies to the mutant gene products. In
addition, the presence of genetic alterations in tumors may provide a molecular
tool for improved prognostic evaluation of patients, as is now possible with
colorectal cancer (Vogelstein et al., 1989; Kern et al., 1989). Finally, the
identification of mutant gene products in tumors may provide targets for new

chemotherapeutic agents.

   

"‘illllilllllllli