University of Southern California Dissertations and Theses

Molecular genetic alterations in breast cancer

(USC Thesis Other)

Molecular genetic alterations in breast cancer

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content IN F O R M A T IO N T O U S E R S
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter face, while others may be
from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back o f the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zed> Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M o lec u la r Genetic A lterations
in B r ea st Cancer
by
Jason Jerome Lukas
A Dissertation Presented to the
FACULTY o f the GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY
(Pathobiology)
August, 1998
© Jason Jerome Lukas
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI Number: 9919076
UMI Microform 9919076
Copyright 1999, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE GRADUATE SCHOOL
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
J a s o n J e r o m e L u k a s
under the direction of hi.s. Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
\raduate Studies
Date
DISSERTATION COMMITTEE
Chairperson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
D e d i c a t i o n :
To my father, for without his moral support, unfailing humor, and once in a lifetime
vacation packages, I would not have completed this dissertation.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A c k n o w l e d g e m e n t s
Time ripens all things. No man is born wise.
-Miguel de Cervantes, Don Quixote, Chapter 33.
Recollecting those individuals who must be acknowledged for their contribution
to my education at USC is a monumental task, especially because I need to go back to
the closing days of 1988 when I began working the lab of Michael Press. I thank first
Michael Press for his patient support, guidance and friendship as well as his belief in me
when I didn’t believe in myself. I also owe him my gratitude for hiring a college student
with little experience and a lot of lip. We both got more than we bargained for.
My sincere gratitude to the rest of my dissertation committee, Michael Stallcup,
Darryl Shibata, Timothy Triche and Fred Hall for their time and effort for providing
invaluable advice during this doctoral work. I must also thank those emeritus members
of the Press lab, Judy Udove and Marianne Friedmann, who showed me how fun
science can be and left me with the sense that the word ‘research’ resonates with
adventure. Those early experiences were strong incentives to start down the road to a
doctoral degree. It was my great pleasure to learn and work with you.
This thesis would not have been completed without the technical and moral
support of many people in the Press lab, as well as in other labs. Thanks particularly to
Mariana Keshmeshian, Ning Niu, Armine Arkelian, Angela Reles, Isabel! Schmitt and
Ivonne Villalobos for their hours of effort.
Finally, I thank the students who made the work in the lab as fun and enjoyable
as possible. I will always remember the elan of our singular laboratory environment.
Make it thy business to know thyself, which is the most difficult lesson
in the world.
Don Quixote, Chap. 52.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
L is t o f F ig u r e s.......................................................................................................................................... vi
l i s t o f T a b l e s .......................................................................................................................................... v n
a b s t r a c t .................................................................................................................................................. v m
AN INTRODUCTION TO BREAST CANCER..................................................... 1
HISTOLOGY OFTHE NORMAL AND ABNORMAL BREAST.................................................................. 4
The Normal Breast................................................................................................................... 4
Neoplastic Progression in the Abnormal Breast.................................................................................5
T h e G e n e tic s o f B r e a s t C a n c e r P r o g r e s s i o n ...........................................................................6
The p53 mediated DNA damage control pathway................................................................................7
The Monitor p53....................................................................................................................................7
p53 Upstream Activation....................................................................................................................10
Responses to p53 Activation: The Downstream Cascade................................................................1 0
The Role of p53 in Cell Cycle Control..............................................................................................10
G,S Transition.................................................................................................................................10
G,M transition.................................................................................................................................1 1
The Role of p53 in Centrosome Homeostasis................................................................................... II
The Role of p53 in Apoptosis........................................................................................................... 11
The Role of p53 in Genomic Stability..............................................................................................12
Nucleotide excision repair.............................................................................................................1 2
Homologous recombination.............................................................................................................. 15
Determining the Genetic Alterations on p53 and Related Genes in Breast Cancer................16
C o n c l u s io n s...............................................................................................................................................18
R e f e r e n c e s..................................................................................................................................................19
CHAPTER 1: P53 GENE MUTATIONS AND EXPRESSION IN
BREAST CARCINOMA IN SITU ......................................................................... 23
A b s t r a c t ....................................................................................................................................................2 3
In t r o d u c t io n ........................................................................................................................................... 2 4
M a t e r ia l s a n d M e t h o d s .................................................................................................................... 2 5
Tissue.................................................................................................................................... 25
Microdissection and DNA isolation.......................................................................................26
p53 primer design...................................................................................................................27
Single-strand Conformational Polymorphism (SSCP) and DNA Sequence Analysis 27
p53 Immunohistochemistry.................................................................................................................. 27
Statistical Analyses................................................................................................................ 28
R e s u l t s ........................................................................................................................................................2 8
p53 Mutations....................................................................................................................... 28
p53 Protein Expression.......................................................................................................... 32
Comparison o f p53 Mutations with Protein Overexpression................................................37
D is c u s s io n .................................................................................................................................................3 9
The structure o f p53: why mutations alter function.............................................................41
R K F V » R pNPp.^ 4 8
CHAPTER 2: WAF1/CIP1 GENE P O L ^ ^ O F ^ fflS M AND
EXPRESSION IN CARCINOMAS OF THE BREAST, OVARY AND
ENDOMETRIUM......................................................................................................... 52
A b s t r a c t ................................................................................................................................................... 5 2
In t r o d u c t io n ........................................................................................................................................... 5 4
M a t e r ia l s a n d M e t h o d s ....................................................................................................................5 5
Tissues....................................................................................................................................................... 55
DNA Isolation and PCR........................................................................................................56
Single-Strand Conformational Polymorphism Analysis.......................................................56
DNA sequencing....................................................................................................................................... 57
Restriction Analysis................................................................................................................................57
iv
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Immunohistochemistry.......................................................................................................... 58
Statistical Analyses................................................................................................................ 58
RESULTS........................................................................................................................................................5 9
DISCUSSION.................................................................................................................................................6 6
REFERENCES.................................................................................................................................................7 4
CHAPTER 3: ALTERNATIVE AND ABERRANT SPLICING OF THE
MDM2 ONCOGENE IN INVASIVE BREAST C A N C E R ............................. 77
ABSTRACT....................................................................................................................................................7 7
INTRODUCTION........................................................................................................................................... 7 9
M a t e r ia l s a n d Me t h o d s .................................................................................................................... 8 0
Tissues................................................................................................................................... 80
Total RNA isolation..............................................................................................................80
Reverse transcription..............................................................................................................80
mdm2 primer design and polymerase chain reaction (PCR)..................................................82
Analysis o f mutations in p53.................................................................................................82
mdm2 product cloning and DNA sequencing.........................................................................84
Immunohistochemistry.......................................................................................................... 84
p 5 3 .........................................................................................................................................................84
HER2/neu..............................................................................................................................................85
Epidermal Growth Factor receptor (EGFR)........................................................................................ 8 6
Estrogen and Progesterone receptors.................................................................................................8 6
DNA ploidy analysis............................................................................................................................8 6
Statistical Analyses............................................................................................................................. 8 6
Re s u l t s ........................................................................................................................................................8 7
Altered mdm2 mRNAs in tumor tissues................................................................................87
mdm2 mRNAs in benign tissues...........................................................................................95
PCR analysis ofmdm2 genomic sequence............................................................................ 96
Correlations o f altered mdm2 mRNAs with prognostic markers and clinical outcome 97
D is c u s s io n ...............................................................................................................................................1 0 0
Re f e r e n c e s...............................................................................................................................................1 0 4
CHAPTER 4: DIFFERENTIAL EXPRESSION OF GENES DURING
BREAST CANCER PROGRESSION................................................................. 107
A b s t r a c t ................................................................................................................................................. 10 7
In t r o d u c t io n ......................................................................................................................................... 108
M a t e r ia l s a n d M e t h o d s.................................................................................................................. 1 10
Tissue...................................................................................................................................110
Microdissection and Total RNA isolation............................................................................I l l
Differential mRNA Display (DD)........................................................................................Ill
Northern Analysis................................................................................................................ 112
Re s u l t s.......................................................................................................................................................113
D is c u s s io n .............................................................................................................................................. 1 2 4
p pFFRFNrpr S 1 9 9
SUMMARY AND FUTURE DliRECTIONSVr/!7/!/.7r/y!7/! ” .7r.y.V^Vy! l3 3
v
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L ist o f F ig u res
Figure 1-1: A Normal Breast........................................................................................................... 2
Figure 1-2: A model of breast carcinogenesis...........................................................................8
Figure 1-3: Some of the Multiple Pathways of Direct and Indirect p53 Action 13
Figure 1-1: SSCP analysis of exon 8 of p53 in DCIS.......................................................31
Figure 1-2: p53 Immunohistochemistry and DNA sequencing on a Single Tumor 33
Figure 1-3: p53 mutations in DCIS compared to the molecular structure of p53. 46
Figure 2-1. Single strand conformation polymorphism (SSCP) analysis of exon 2
from WAFl/CipI in endometrial carcinomas................................................................ 60
Figure 2-2. DNA Sequence Analysis of Breast Carcinomas in S itu............................... 6 2
Figure 2-3. p21W A F I/c,p l immunostaining in breast DCIS, invasive ovarian carcinoma
and endometrial carcinoma.................................................................................................... 63
Figure 2-4. Pair-wise differences in p21W A F I/C ip I immunostained cells of carcinoma
nuclei and of benign nuclei on a case-by-case basis...................................................68
Figure 3-1. Mdm2 alterations in invasive breast cancers and normal breast 88
Figure 3-2: mdm2 RT-PCR products in relation to a full length mdm2 cDNA 91
Figure 3-3: A comparison of representative segments of the abberantly spliced
mdm2 cDNAs..............................................................................................................................93
Figure 3-4: Logrank survival curves for breast cancer patients......................................9 8
Figure 4-1: Differential mRNA Display.................................................................................114
Figure 4-2: Northern Analysis of Differentially Displayed mRNAs........................... 121
vi
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L ist o f Tables
Table 1-1: p53 Alterations In Breast Carcinoma in situ Without Invasion...................36
Table 1-2: p53 Alterations in Breast Carcinoma in situ With Invasion.........................36
Table 1-3: p53 polymorphisms in Breast Carcinoma in situ ...............................................38
Table 1-4: Comparison of p53 .overexpression with p53 mutations in 58
cases...................................................................................................................................................38
Table 4-1: Characterization of cDNAs Corresponding to Genes Differentially
Expressed in Invasive Carcinoma, DCIS and Benign Hyperplasia........................... 116
Table 4-2: Known Sequences with Significant Homology to Genes Over
Expressed in Invasive Breast Cancer by Differential Display....................................117
Table 4-3: Known Sequences with Significant Homology to Genes Under
Expressed in Invasive Breast Carcinoma by Differential Display.........................118
Table 4-4: Sequences Over Expressed in Invasive Breast Carcinoma by
Confirmatory Northern Blots.................................................................................................. 123
Table 4-5: Sequences Under Expressed in Invasive Breast Carcinoma by
Confirmatory Northern Blots...................................................................................................123
vii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A bstr ac t
Current understanding of breast cancer genetics is based on invasive breast
disease studies. We set out to discern the genetic alterations in an early, pre-invasive
lesion of the breast, carcinoma in situ (CIS) to dissect the process of breast
carcinogenesis. Analysis of p53 in CIS both with and without invasive disease showed
that p53 mutations occur in both diseases and these mutations were the same. However,
higher p53 expression in CIS with invasive disease without mutations suggested
downstream gene alterations might cause overexpression of intact p53 through lack of
proper feedback. However, one such gene, WAF1, had no mutations in breast cancer.
However, mdm2 mRNAs exhibit alternative splicing in normal and breast
cancers, and aberrant splicing only in breast cancers. An alternative mdm2 splicing
product was associated with p53 mutations, while aberrant splicing products showed a
statistically significant correlation with p53 overexpression and decreased survival.
Thus, mdm2 alternative and aberrant splicing may account for overexpression of p53 in
the absence of p53 mutations and play a role in breast cancer etiology.
We also searched for genes having an impact on breast pathogenesis.
Differential mRNA display was used to analyze the mRNA expression patterns in
benign hyperplasia, ductal carcinoma in situ, and invasive breast disease from the same
tumor. Twenty genes differentially expressed genes were found with patterns of
expression suggestive of tumor suppressor genes and oncogenes. Overall, we reached a
number of important conclusions. 1). In breast cancer, p53 mutations occur prior to or
at the formation of CIS. 2). p53 mutations are the same in both CIS and invasive
disease, suggesting that a clonal relationship exists between these two different diseases.
viii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
3). mdm2 aberrant and alternative splicing correlated in statistical significant fashion
with p53 overexpression, in the absence of p53 mutations, indicating that mdm2
alterations may influence p53 overexpression. 4). mdm2 aberrant splicing showed a
statistically significant correlation with poor clinical outcome. Thus it appears that p53
and mdm2 alterations, as well as a host of known and unknown genes play a significant
role in breast cancer pathogenesis.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
An Introduction to Breast Cancer
Breast cancer is the most commonly diagnosed cancer among women in the
United States. Estimates suggest that in 1998, breast cancer will comprise 30% of
new cancer cases and 16% of the new deaths (1). Approximately one woman in 8
(12.5%) will receive a diagnosis of breast cancer in her lifetime. This represents
twice the risk of breast cancer as compared to 50 years ago.
The importance of understanding the genetic alterations in breast cancer is
underscored by differences in clinical outcome of women whose invasive breast
carcinoma show different observed genetic alterations (2-4). Currently, CIS
represents approximately 20-30% of new breast cancers (5). As mammography
comes into wider use and, consequently, the question of how to properly treat breast
CIS and pre-neoplastic lesions become increasingly important, understanding the
basic mechanisms of tumor development and progression are critical to rational,
successful treatment.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1-1: A Normal Breast
The normal breast consists of a large number of terminal duct lobular units which open
onto terminal ducts. From there, lactiferous ducts conduct milk to the nipple. Note
that in different areas of the breast, different diseases tend to develop. For instance,
ductal carcinoma in situ is thought to arise primarily form the terminal duct just outside
of the lobule.
2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Terminal duct
tobufar unit
Ductules or acini
Terminal duct
Lobule
Adipose tissue
Segm ental duct
Lactiferous duct
Lactiferous sinus
Nipple
Pagers cfeease
Nipple adenom a
Papillomas
Traumatic fat necrosis
Hyperplasia
Most carcinom as
Fibroadenoma )
Cysts j
3
^ p ^ u c e d with permission of the copyright owner. Further reproduction prohibited without permission
H isto lo g y o f t h e N o r m a l a n d A b n o r m a l B r e a s t
Embryologically, the breast is a modified skin appendage or sweat gland
beginning development in the sixth week of gestation and continuing slow growth
consisting mainly of branching of the mammary ducts through prepubertal life. In
the male, growth ceases at this time. In the female, before the onset of menstruation,
the growth rate of the intraductal stroma drastically increases; branching of the ducts
increases as well (6). During puberty, there is a large increase in breast size due
primarily to stromal growth, but also the terminal ducts give rise to numerous
saccular outpouchings that form the lobules. Non-lactating breast is composed of
15-25 lobes of tubulo-alveolar type gland divided into two major portions: the
terminal duct lobular unit, or lobule, consisting of the functional unit of the breast;
and the terminal duct, continuous with the lactiferous ducts, which ends in a
lactiferous sinus emptying independently at the nipple.
The Normal Breast
The outlet of the lactiferous ducts, the nipple, and the areola are covered with
a stratified squamous epithelium. The major breast ducts are lined by a layer of two
cell thick cuboidal epithelium. As the ducts ramify and decrease in size, so does the
epithelium, becoming a single layer of cuboidal cells, backed by flat myoepithelial
cells.
Only with the onset of pregnancy does the breast assume complete
morphological maturation and functional activity. Ovarian and placental hormones
stimulate growth and maturation of the breast. Numerous outpouchings form in the
4
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
lobule, giving it the appearance of a cluster of grapes. Lobular growth is prodigious:
at the end of pregnancy the composition of the breast is the opposite of the resting
breast, composed primarily of glandular elements. More importantly, the epithelial
cells terminally differentiate and begin to produce milk.
Neoplastic Progression in the Abnormal Breast.
Breast cancer is thought to arise from the terminal duct lobular unit (7) and is
thought to progress though a series of morphologic changes. These changes can be
characterized at the earliest stage by an overgrowth of the single layer of epithelial
cells to fill the duct or lobule with orderly cuboidal cells (mild hyperplasia). While
this does carry a slightly increased risk of breast cancer, moderate hyperplasia,
characterized by a more florid and multilayered appearance carries a 1.5-2 fold risk of
later cancer. Atypical hyperplasia, with a multilayering of disorderly cells with more
variable size and shape carries a 5 fold increased risk (8) (Figure 1-2). Ductal
carcinoma in situ (DCIS) and lobular carcinoma in situ (LCIS) represent an even
greater proliferation of anaplastic tumor cells which can fill a duct, but are contained
within the myoepithelial layer of the duct.
DCIS is generally sub classified as comedo or non-comedo types, which are
broad descriptions of individual tumor growth patterns. Comedo DCIS appears
more malignant cytologically, with large cell size, pleomorphic nuclei and numerous
mitoses. Non-comedo DCIS is sub classified into micropapillary, cribriform, solid;
in general, the non-comedo cells have a monotonous appearance, with small nuclei
and scant mitoses. Evidence indicates that comedo DCIS possesses a higher rate of
proliferation and is associated with a higher incidence of invasive disease than non
comedo DCIS (9). DCIS has an excellent prognosis; and though it is considered to
5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
be part of the progression of invasive disease, it is not known what proportion of
DCIS will advance to invade the surrounding tissue, nor in what time frame it will
progress. One 15 year study showed that invasive carcinoma develops in 28% of
women treated only with biopsy, usually in the same breast (6). Evidence suggests
that some, but not all, cases of DCIS progress to invasive cancer, and that the lesion
may be occult and clinically inactive for many years (10).
Invasive breast cancer is characterized by pleomorphic epithelial tumor cells
that penetrate the basement membrane and invade the surrounding tissue. Pure
invasive ductal carcinoma is the most commonly diagnosed histologic type (53%),
followed by invasive ductal carcinoma combined with other histologic types (22%).
Infiltrating lobular carcinoma represents approximately 10-15% of invasive breast
carcinomas. Medullary, colloid, Paget’s disease and other types together make up
less than 10% of the cancers diagnosed. Most tumors exhibit increases in dense
fibrous or collagenous stroma surrounding nests of tumor cells. The nuclei are
highly pleomorphic and the cells vary in size and shape and show frequent mitoses.
A number of factors influence the prognosis of those with breast cancer, such as size
of tumor and histologic type, as well as number of lymph nodes involved, hormone
receptor status, and presence of amplified or mutated oncogenes, such as p53.
T h e G en etics o f B r e a st C a n c e r P r o g r e ssio n
Because of regularity of diagosis of invasive breast cancer, as well as tumor
size and ease of procurement, the majority of research work is performed on invasive
breast cancer tissue. Studies reveal the multiple genetic alterations: from
amplification of the proto-oncogenes HER2Jneu, myc, epidermal growth factor
6
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
receptor, cyclin D l, and/or the fibroblast growth factor genes bek and fig, as well as
loss of tumor suppressor genes such as p53 and Rb (for a comprehensive review see
reference (11)). In comparision, far less work has been done to understand genetic
changes in ductal carcinoma in situ and earlier lesions. Chromosome 1 aneusomy
has been reported in CIS (12). Aneuploidy is also commonly described (13) as is
HER2/neu overexpression (14) and p53 overexpression and mutations. However,
no single genetic abnormality has been found in all cancers. Thus, the creation of
molecular tools to ensure a much earlier diagnosis and for developing ways to
reverse the initiation and subsequent steps of tumor development may be the most
effective means to control or prevent breast cancer.
The p53 mediated DNA damage control pathway
The Monitor p53
To remain viable, a cell must be able to respond to various forms of stress.
Damage to DNA, hypoxia, activation of oncogenes, and or certain viral infections are
distinct and separate cell stressors. While the cell conceivably could have evolved an
elaborate system of 4 distinct sensing pathways, p53 evolved to serve as a single
monitor to control responses to these cellular stresses (15). Normally, p53 protein is
maintained at a very low level, which accounts for the inability of immunologic
methods to detect the protein. During times of cellular stress, the amount of p53
rapidly increases, and is generally thought to be weakly immunopositive. It is
thought that this increase stems from an increase in protein half-life or possibly
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1-2: A model of breast carcinogenesis
This diagram o f the multistep model of breast carcinogenesis gives the pathologic
description below the photograph. Genes which are either known to be expressed or
altered in some way are designated above the photograph, at the location that the
alteration is thought to occur. Below the photographs are chromosomal alterations that
frequent these differemt morphologic entities. Note that this sequence of events, from
normal breast epithelium to invasive breast cancer, is not obligatory; certain gene
alterations, such as H E R l/neu overexpression in transgenic mice, have been found to
bypass certain histologies (arrow at top o f diagram.)
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
because the rate of protein translation is increased. p53 can either transactivate other
genes or operate independent of translation.
p53 Upstream Activation
In order to activate p53 function, the p53 protein must be signaled. These
signals are mediated by cell stresses such as double stranded DNA breaks from y or
ultraviolet radiation, hypoxia (16), and levels of ribonucleoside triphospates below a
certain threshold level (17). The ataxia-telangectasia gene has been implicated as an
upstream element of the p53 pathway and is thought to signal p53 via
phosphorylation (18) but because p53 protein can also bind to DNA ends or damaged
sites (19) no upstream components are yet absolutely identified.
Responses to p53 Activation: The Downstream Cascade
p53 is involved in several important aspects of cell homeostasis, such as cell
cycle arrest, apoptosis, control of genomic integrity and DNA repair. It regulates
these processes as a tetramer which specifically binds promoter sequences to
upregulate protein expression of genes such as WAFI (also known as p21, Cipl, or
CDKN1A), mdm2 (also known as hdm2), gadd45, cyclin G, Bax and Igf-Bp3. A
number of genes are downregulated, such as bcl2, c-jun and c-fos. A number of
components of the downstream DNA damage response pathway do not require
transcriptional regulation but are known to interact with p53.
The Role o f p53 in Cell Cycle Control
G,S Transition
After some forms of DNA damage, p53 is upregulated and transactivates
WAF1 (20). WAF1 is a universal inhibitor of the cyclins. WAF1 binds to 4 cyclin-
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cdk complexes: cyclin A-Cdk2, cyclin Al-Cdc2, cyclin Dl-Cdk4, and cyclin E-
Cdk2. It appears that one molecule of WAF1 in a complex is permissive for Cdk
kinase activity, while 2 molecules inhibits this activity (21). Inhibition of the Cdk
activity causes hypophosporylation of Rb, which then prevents the release of E2F
and blocks the G,S transition (Figure 1-3).
G M transition
Overexpression of wildtype p53 can inhibit entry into S phase. Furthermore,
use of mitotic spindle inhibitors in cells with wildtype p53 blocks progression of the
cell cycle in G2. In cells with mutant p53, DNA synthesis will be initiated,
increasing ploidy of the progeny cells (22). p53 appears to control the GjM
transition through a spindle checkpoint and possibly other protein interactions.
The Role o f pS3 in Centrosome Homeostasis
p53 also appears important in the regulation of the number of centrosomes in
a cell. p53 appears to copurify with centrosomes in some cell lines (23). p53 null
mice embryo fibroblasts produces abnormal numbers of centrosomes after a few
doublings in cell culture and produce spindles with 3 or 4 poles (24). Thus p53
appears to be an important part of the mechanism to control centrosome.
p53 appears to have numerous paths to effect cell cycle arrest. Presumably
each separate mechanism for arrest can be altered under different conditions to satisfy
the changing and varied cellular stressors.
The Role o f p53 in Apoptosis
Like p53’s broad spectrum of action in cell cycle arrest, several experiments
point to multiple genes downstream of p53 to cause apoptosis and although the exact
mechanisms are still unclear, p53 can induce apoptosis by transcription dependent
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and transcription independent means. In response to the same stimuli, p53 can
induce apoptosis in some cell types and cell cycle arrest in others. p53 upregulation
of WAF1 (25), bax (26), and possibly fas (27) has been linked to apoptosis, and the
downregulation of bc!2 (28) which normally inhibits the function of bax, further
promotes apoptosis. Overexpression of WAF1 can, in certain circumstances, block
apoptosis (29). Similarly, overexpression of bax and fas alone have not been shown
to induce apoptosis, implying the presence of downstream effectors that are required
for cell death. Very recent work (30) in colorectal carcinoma cell lines showed at
least 39 genes which are transcriptionally activated by p53 and which activate a
specific subset of redox related genes (p53 induced genes, or PIGs), which
collectively increase the cellular content of reactive oxygen species. This release
damages the mitochondria, and the subsequent release of calcium and protein then
stimulate caspases which then bring on apoptosis.
The Role o f p53 in Genomic Stability
Nucleotide excision repair
There has also been investigations of the transcription independent apoptosis
mechanism. p53 appears to interact with the proteins involved in nucleotide excision
repair complex transcription factor IIH ( iTllH). The TFIIH basal transcription
complex consists of ensemble of proteins which interact with damaged DNA in order
to repair it. The DNA helicases, excision repair cross-complementing 3 (ERCC3)
(also known as XPB due to it’s association with xeroderma pigmentosum) and
ERCC2 (or XPD), are integral members of the complex while the protein CSB
(named for it’s association with the Cockayne syndrome complementation group B)
is associated with TFUH though CSA (Cockayne syndrome group A). p53 binds
12
with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1-3: Some of the Multiple Pathways of Direct and Indirect p53 Action
A current schematic of the multiple upstream activators of p53 (above p53), two gene
products which play a role in controlling p53 activity (on either side of p53), and the
multiple downstream effectors o f p53. In general, an arrow indicates a postive effect,
while a flat line indicates an inhibitory effect. Along side the arrows are cofactors.
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
DNA damage
Hypoxia Oncogene Activation
(HPV-E6, SV40-T, Adeno El A)
p53 degradation
p73
^ p53 nuclear
inclusion?
mdm2
p53CP
pZllWBFl
Gadd45 R H bcl2
&
mitotic spinde
checkpoint
Apoptosis DNA Damage Repair
Arrest
B:Cdc2
Genomic Instability
redrawn from Jacks, Nature 381: 643; Agarwal, JBC 273: 1; und Wang J Cell Physiol 173: 247.
XPB at the helicase motif HI, a motif common among a large number of helicases,
such as XPB, XPD and CSB (ERCC6). XPB and XPD are implicated in the cancer-
prone repair syndrome xeroderma pigmentosum (31) while CSB is important in the
process of strand specific DNA repair and is associated with the Cockayne syndrome
(32). Furthermore, in the fibroblasts from patients with Li Fraumeni syndrome that
are heterozygous for p53, UV induced pyrimidine dimers, nucleotide excision repair
proceeded at a slower rate. These results reveal that p53 may play an important role
in modulating nucleotide excision repair.
Homologous recombination
p53 is also known to be involved in homologous recombination via
interactions with RAD51, a protein required for repair of double-strand breaks that
occur in mitosis or meiosis. RAD51 is the human homolog of RecA, a prokaryotic
protein which searches for homologous regions between two double-stranded DNA
molecules and promotes strand exchange. RecA is also involved in the
recombinatorial repair of double stranded DNA breaks and controls the DNA damage
response (33). Overexpressed wildtype p53 bound to and inhibited RAD51 mediated
transfer of complementary DNA strands. Interestingly, recently BRCA1 was found
in complexes with RAD51 (34) and also appears to enhance p53 dependent gene
expression of WAF1 and mdm2 by acting as a p53 coactivator (35). It has been
suggested that p53/BRCAl/RAD51 forms a temiary complex which senses DNA
damage. p53 then can monitor DNA damage and participate in numerous aspects of
repair as well.
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Determining the Genetic Alterations on p53 and Related Genes In
Breast Cancer
Because the functioning of p53 is important to many cellular processes, p53
appears to be inactivated in many human tumors. The majority of the alterations in
p53 are point mutations in critical areas of the gene, causing the expression of a
conformationally different, functionally defective protein. This defective protein is
thought to have an increased half-life compared to the wildtype protein, and is thus
detectable by immunohistochemistry. Many older studies have thus equated the
presence of p53 mutations with the detection of the p53 protein. More recent work
has shown that while overexpression of p53 does correlate with p53 mutations, there
is a distinct subset of cases that appear have overexpression but do not have
mutations (36, 37). We hypothesized that if p53 was altered in invasive breast
cancers, then p53 should be altered in CIS. A greater understanding of alterations in
DCIS would be beneficial in two ways. First, this knowledge would improve our
understanding of breast carcinogenesis by ascertaining if, in a certain subset of
tumors, DCIS is a precursor to invasion. Second, by determining if, in the process
of invasion, there was a difference in p53 overexpression, then these differences
could indicate other genes in the p53 DNA damage repair pathway or closely allied
with this pathway that may be altered during breast carcinogenesis.
Thus, Chapter 1 describes the analysis of the p53 tumor suppressor gene
DNA and protein in a cohort of ductal carcinomas in situ cases. We found that p53
was altered in DCIS and show that p53 is altered in a pattern very similar in both
DCIS and invasive disease. We found that a consistent pattern of positive p53
immunostaining in tumors without p53 mutations. This led us to investigate both a
downstream effector of p53, WAF1 (Chapter 2), for alterations in the gene and/or
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
protein expression, and a gene which down-regulated p53, known as mdm2. We
hypothesized that while p53 may be intact, an altered WAF1 could cause
overexpression of p53 through lack of proper feedback. No mutations in the WAF1
gene were found but we noted differences between WAF1 protein expression in
tumor cells and benign stromal cells varied in a fashion that appeared to be
characteristic of particular tumor types.
We continued the investigation by analyzing the mRNA of mdm2 (Chapter
3), a gene which is upregulated by p53, subsequently binds to and inactivates p53 as
well as directing p53 degradation. Alternative and aberrant mRNA splicing was
noted in both breast tumors and normal breast epithelium, while only alternative
splicing was found to occur in normal breast epithelium using an RT-PCR assay.
The presence of altered size mRNA RT-PCR products had a statistically significant
relationship with the presence of p53 overexpression as well as p53 mutations,
which suggests, unlike WAF1, that mdm2 alterations may cause a compensatory
upregulation of p53 protein in the absence of p53 mutations. Perhaps more
interesting is that mdm2 alterations correlate with p53 mutations, suggesting some
kind of causal relationship might exist between the two genes or, alternatively, that a
third gene, perhaps causing a mutator phenotype, might be affecting both genes.
Finally, expression of mdm2 aberrant RT-PCR products correlate in a statistically
significant fashion with poor outcome.
Breast cancer is considered to be the result of a variety of molecular genetic
alterations which lead to altered gene expression and altered cellular function. The
p53 pathway represents one known pathway which is altered in breast cancer. The
final component of these investigations (Chapter 4) was to identify new genes in
breast cancers which have alterations in expression and may contribute to the disease
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
process. Differential mRNA display was used to analyze the mRNA expression
patterns of three separate histology types: benign hyperplasia, ductal carcinoma in
situ, and invasive breast disease from the same tumor sample. We have found 20
genes that are differentially expressed. Overall, these studies show that p53 and
mdm2, as well as novel and known genes, appear to be important in breast cancer
pathogenesis.
C o n c l u sio n s
p53 is clearly a component in several biochemical pathways that are crucial to
the proper maintenance of cell homeostasis through DNA repair, cell cycle arrest and
apoptosis. Alterations in p53 can impair the cellular mechanisms to respond to stress
properly. Similarly, alterations of components either upstream or downstream of
p53 may be equivalent to an alteration of p53, and interfere with the proper function
of this pathway and similarly cause perturbations in the mechanisms of cellular
homeostasis. This work seeks to explore what alterations occur in the p53 pathway
in CIS and invasive disease and find new genes to more fully understand
carcinogenesis of the breast.
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R e f e r e n c e s
1. Landis, S., Murray, T., Bolden, S., and Wingo, P. Cancer Statistics, 1998,
CA Cancer J Clin. 48: 6-29, 1998.
2. Press, M. HER-2/neu Expression in Node-negative Breast Cancer Direct
Tissue Quantitation by Computerized Image Analysis and Association of
Overexpression with Increased Risk of Recurrent Disease., Cancer Res. 53:
4960-4970, 1993.
3. Slamon, D. Human breast cancer correlation of relapse and survival with
amplification of the HER2/neu oncogene., Science. 235: 177-182, 1987.
4. Slamon, D., Godolphin, W., Jones, L., Holt, J., Wong, S., Keith, D.,
Levin, W., Stuart, S., Udove, J., Ullrich, A., and Press, M. Studies of the
HER-2/neu proto-oncogene in human breast and ovarian cancer., Science.
244: 707-712, 1989.
5. Fonseca, R., Hartmann, L., Petersen, I., Donohue, J., Crotty, T., and
Gisvold, J. Ductal carcinoma in situ of the breast, Ann Intern Med. 127:
1013-1022, 1997.
6. Cotran, R., Kumar, V., and Robbin, S. Robbins Pathologic Basis of Disease,
4 edition edition: WB Saunders, 1989.
7. Azzopardi, J. Histological typing of breast tumours: World Health
Organization, 1981.
8. Page, D. L., Dupont, W. D., Rogers, L. W., Jensen, R. A., and Shuyler, P.
A. Continued local recurrence of carcinoma 15-25 years after a diagnosis of
low grade ductal carcinoma in situ of the breast treated only by biopsies,
Cancer. 76: 1197-2000, 1995.
9. Schnitt, S., Silen, W., Sadowsky, N., Connolly, J., and Harris, J. Ductal
carcinoma in situ (intraductal carcinoma) of the breast., N Engl J Med. 318:
898-903, 1988.
10. Harris, A. p53 expression in human breast cancer, Vol. 59, p. 69-88:
Academic Press, Inc., 1992.
11. Walker, R., Jones, J., Chappell, S., Walsh, T., and Shaw, J. Molecular
pathology of breast cancer and its application to clinical management, Cancer
Met Rev. 16: 5-27, 1997.
12. Harrison, M., Magee, H., O'Loughlin, J., Gorey, T„ and Dervan, P.
Chromosome 1 aneusomy, identified by interphase cytogenetics, in
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mammographicaUy detected ductal carcinoma in situ of the breast, J Pathol.
175: 303-309, 1995.
13. Leal, C. B., Schmitt, F. C„ Bento, M. J., Maia, N. C., and Lopes, C. S.
Ductal carcinoma in situ of the breast, Cancer. 75:2123-2131, 1995.
14. Ratcliffe, N., Wells, W., Wheeler, K., and Memoli, V. The combination of in
situ hybridization and immunohistochemical analysis: an evaluation of
Her2/neu expression in paraffin-embedded breast carcinomas and adjacent
normal-appearing breast epithelium., Mod Pathol. 12: 1247-1252, 1997.
15. Hall, D„ Meek, D., and Lane, D. p53: integrating the complexity, J Pathol.
180: 1-5, 1997.
16. Graeber, A., Osmanian, C., Jacks, T., Housman, D., Koch, C., Lowe, S.,
and Graccia, A. Hypoxia-mediated selection of cells with diminished apoptotic
potential in solid tumors., Nature. 379: 88-91, 1996.
17. Linke, S., Clarkin, K., DiLeonardo, A., Tsou, A., and Wahl, G. A
reversible, p53-dependent G0/G1 cell cycle arrest induced by ribonucleotide
depletion in the absensce of detectable DNA damage, Genes Dev. 10: 934-
937, 1996.
18. Kastan, M., Zhan, Q., el-Deiry, W., Carrier, F., Jacks, T., Walsh, W.,
Plunkett, B., Vogelstein, B., and Fomace, A. J. A mammalian cell cycle
checkpoint pathway utilizing p53 and GADD45 isdefective in ataxia-
telangiectasia.. Cell. 71: 587-597, 1992.
19. Lee, S., Elenbaas, B., Levine, A., and Griffith, J. p53 and its 14 kDa C-
terminal domain recognize primary DNA damage in the form of
insertion/deletion mismatches, Cell. 81: 1013-1020, 1995.
20. El Diery, W. S., Tokino, T„ Velculescu, V. E., Levy, D. B., Parsons, R.,
Trent, J. M., Lin, D., Mercer, W. E., and Kinzler, K. W. WAF1, a potential
mediator of p53 tumor suppression, Cell. 75: 817-825, 1993.
21. Zhang, H., Hannon, G., and Beach, D. p21-containing cyclin kinases exist in
both active and inactive states., Gene Dev. 8: 1750-1758, 1994.
22. Cross, S., Sanchez, C., Morgan, C., Schimke, M., Ramel, S., Idzerda, R.,
Rasking, W., and Reid, B. A p53 dependent mouse spindle checkpoint,
Science. 267: 1353-1356, 1995.
23. Brown, C., Doxsey, S., White, E., and Welch, W. Both viral (adenovirus
E1B) and cellular (hsp70, p53) components interact with centrosomes, J Cell
Physiol. 160: 47-60, 1994.
24. Fukasawa, K., Choi, T., Kuriyama, R., Rulong, S., and Vande Woude, G.
Abnormal centrosome amplification in the absence of p53, Science. 271:
1744-1747, 1996.
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
25. El Diery, W., Harper, J., O’ Connor, P., Veculescu, V., Canman, V.,
Jackman, J., Pietenpol, J., Burrell, M., Hill, D., Wang, Y., Wiman, K.,
Mercer, W., Kastan, M., Kohn, K., Kinzler, K., and Vogelstein, B. WAF1
is induced in p53-mediated G1 arrest and apoptosis, Cancer Res. 54: 1169-
1174, 1994.
26. Miyashita, T. and Reed, J. Tumor suppressor p53 is a direct transcriptional
activator of the human bax gene, Cell. 80: 293-299, 1995.
27. Owen-Schaub, L., Zhang, W., Cusack, J., Angelo, L., Santee, S., Fujiwara,
T., and Roth, J. Wild type human p53 and a temperature sensitive mutant
induce Fas/Apol expression, Mol Cell Biol. 15:, 1995.
28. Miyashita, T., Krajewski, S., Krajewska, M., Wang, H., Liebermann, D.,
Hofftnan, B., and Reed, J. Tumor suppressor p53 is a regulator of bcl-2 and
bax gene expression in vitro and in vivo, Oncogene. 9:, 1994.
29. Cannon, C., Golmer, T., Coutts, S., and Kastan, M. Growht factor
modulation of p53 mediated growth arrest versus apoptosis, Genes Dev. 9:,
1995.
30. Polyak, K., Xia, Y., Zweier, J., Kinzler, K., and Vogelstein, B. A model for
p53-induced apoptosis, Nature. 389: 300-305, 1997.
31. Wang, X.-W., Vermeulen, W., Coursen, J., Gibson, M., Lupoid, S.,
Forrester, K., Xu, G., Elmore, L., Yeh, H., Hoeijmakers, J., and Harris, C.
The XPB and XPD DNA helicase are components of the p53-mediated
apoptosis pathway, Gene Dev. 10: 1219-1232, 1996.
32. Troelstra, C., van Gool, A., de Wit, J., Vermeulen, W., Bootsma, D., and
Hoeijmakers, J. ERCC6, a member of a subfamily of putative helicases, is
involved in Cockayne’ s syndrome and preferential repair of active genes, Cell.
71: 939-953, 1992.
33. Sturzbecher, H.-W., Donzelmann, B., Henning, W., Knippschild, U., and
Buchhhop, S. p53 is linked directly to homologous recombination processes
via RAD51/RecA protein interation, EMBO J. 15: 1992-2002, 1996.
34. Scully, R„ Chen, J., Plug, A., Xiao, Y., Weaver, D., Feunteun, J., Ashley,
T., and Livingston, D. Association of BRCA1 with Rad5i in mitotic and
meiotic cells, Cell. 88:, 1997.
35. Ouchi, T., Monteiro, A., August, A., Aaronson, S., and Hanafusa, H.
BRCA1 regulates p53-dependent gene expression, Proc Natl Acad Sci, USA.
95: 2302-2306, 1998.
36. Battifora, H. p53 immunohistochemistry: a word of caution, Hum Pathol. 25:
435-437, 1994.
21
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37. Hall, P. and Lane, D. p53 in tumor pathology: can we trust
immunohistochemistry?, J Pathol. 172: 1-4, 1994.
22
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 1: p53 Gene Mutations and
Expression in Breast Carcinoma in Situ
A b s t r a c t
The p53 tumor suppressor gene is altered in approximately half of human
cancers. Although p53 mutations are common in invasive breast carcinoma, few have
been identified in breast carcinoma in situ (intraductal breast carcinomas). The current
study was undertaken to characterize p53 in a cohort of breast carcinoma in situ cases,
both with and without invasive disease. Fifty eight frozen breast biopsy samples were
used for these investigations. Twenty seven cases had only ductal carcinoma in situ
(CIS) and thirty one cases had evidence of both invasive and in situ carcinoma. DNA
sequence alterations in exons 2 through 11 of p53 were screened by the single-strand
conformational polymorphism (SSCP) technique. Exons with altered mobility were
sequenced. Among breast CIS cases without invasive disease, 22% had p53 mutations
and 7% had DNA sequence alterations of unknown significance. Analysis of breast CIS
with concurrent invasive disease demonstrated p53 mutations in 19% of cases and DNA
alteration of unknown significance in one case (3%). Each carcinoma containing a p53
mutation in the breast CIS component had the identical mutation in the invasive
component of the same tumor indicating a clonal relationship between the two tumor
components. p53 protein overexpression was identified in 22% of pure intraductal
breast carcinomas and in 35% of breast CIS with invasive disease. Comparison of
immunostaining and DNA sequence alterations showed a significant association between
overexpression and mutations (p=0.0037) in cases of CIS without invasion, and
similarly between overexpression and mutations in cases of CIS with invasion
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(p=0.007). p53 mutations and p53 overexpression were relatively common in
intraductal breast carcinomas but were not observed in adjacent normal breast lobules or
ducts in 9 cases available for DNA analysis. The frequency of p53 alterations when
comparing breast CIS with and without an invasive component indicated that p53
mutations usually occur prior to invasion during the progression of breast cancer, as is
observed for a number of other adult solid tumors.
I n t r o d u c t io n
Carcinoma of the breast is a multigenic disease which is generally thought to
progress through a stepwise process that involves both morphologic and genotypic
changes. It is thought that breast cancer develops from normal breast epithelium,
progressing through epithelial hyperplasia without cytological atypia, hyperplasia with
atypia, intraductal carcinoma or ductal carcinoma in situ (CIS), invasive cancer and,
finally, metastatic disease (1). Research on invasive breast cancer has identified many
altered genes; however, the specific time of occurrence and phenotypic effect of these
alterations during progression is uncertain. Epidemiological studies have linked certain
pre-invasive lesions to invasive breast cancer. Women diagnosed with ductal
hyperplasia or atypical ductal hyperplasia have a two-fold or a 4 to 5-fold increased risk,
respectively, of later invasive breast carcinoma relative to the unaffected population. A
diagnosis of breast ductal CIS indicates an approximately 10-fold increased risk of
future invasive breast carcinoma (2). If the formation of invasive breast carcinomas is
dependent on multiple genetic alterations, then morphologic phenotypes in breast
pathology may coincide with these alterations.
Most investigations of p53 in breast cancer have been conducted in primary
invasive breast carcinomas simply because of relatively easy availability of material.
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Approximately 22-58% of invasive breast cancers have p53 mutations or overexpression
(3-7). While the molecular genetic profile of p53 alterations has been well analyzed in
invasive breast carcinomas, most studies of breast CIS report only
immunohistochemical analysis of p53 expression using paraffin embedded tissues (8-
20). Few studies have evaluated p53 mutations in breast CIS. Overall, only six
mutations of p53 have been described among 99 breast CIS cases (6%) screened by a
variety of strategies (21-24). The small number of mutations identified suggests that
p53 mutations are infrequent in intraductal breast carcinomas or breast CIS. The higher
rate of p53 mutations/alterations in invasive disease has suggested to some that p53
mutations are linked primarily to invasive breast carcinoma and occur relatively late in
the disease process (1).
In the current study, p53 mutations were analyzed in exons 2 through 11 using
single stranded conformational polymorphism analysis followed by DNA sequence
analysis. p53 expression was analyzed by immunohistochemistry using frozen breast
CIS specimens to more completely characterize p53 alterations in breast carcinoma in
situ. The findings indicate that p53 is mutated in breast CIS as frequently as in invasive
breast carcinoma and that the invasive carcinomas are clonally derived from breast CIS
as indicated by maintenance of the same p53 sequence in intraductal and invasive disease
from the same case.
M a t e r ia l s a n d M e t h o d s
T issue
The use of human tissue in this study was reviewed and approved by the
University of Southern California Institutional Research Committee. Fifty-eight frozen
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
breast biopsy samples with CIS, stored at -186°C under liquid nitrogen and accessioned
sequentially between 1988 and 1996 by the USC Breast Tumor and Tissue Bank, were
used for this investigation. Twenty-seven cases had only carcinoma in situ and thirty-
one cases had evidence of both invasive and in situ carcinoma. Among the cases with
only CIS, 18 (66.5%) breast CIS were classified as comedocarcinoma, 3 (11%) as
cribriform ductal CIS, 2 (7.5%) as micropapillary ductal CIS, 2 (7.5%) as papillary
ductal CIS, and 2 (7.5%) as solid ductal CIS. In the cases with invasive disease and
CIS, 18 (58%) breast CIS were classified as comedocarcinoma, 8 (27%) as cribriform
ductal CIS, 1 (3%) as micropapillary ductal CIS, 2 (6.5%) as papillary ductal CIS, and
2 (6.5%) as solid ductal CIS. Frozen tissue sections, stained with hematoxylin-and-
eosin, were used to confirm the histological composition of the specimens. When
available, pathology reports were used to confirm the presence or absence of micro-
invasion. Those cases having micro-invasion were classified as CIS with invasive
disease. The invasive disease tumors were graded according to the BIoom-Richardson-
Scharf classification.
Microdissection and DNA isolation
The different tissue components of the breast specimens were separated
according to histomorphologic phenotypes with the assistance of a microscope by
microdissection. Ten to twenty frozen serial sections, 10 pm thick, were cut and fixed
in 95% ethanol. The initial section was stained with hematoxylin-and-eosin and
subsequent sections for microdissection were stained with ethyl green. Microdissection
was performed to separate benign breast epithelium, breast CIS and invasive carcinoma.
Care was taken in order to minimize contamination of epithelial cells with stromal cells.
After microdissection, the DNA of each component was extracted as previously
described (25). In three cases, paraffin embedded tissue was utilized in addition to
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
frozen tissue. These cases were sectioned and subjected to microdissection and DNA
extraction as described previously (25).
p53 primer design
The polymerase chain reaction (PCR) was used to amplify exons 2-11. Ten
different sets of l9-25mer oligonucleotide primers were designed using the genomic
sequence of p53 (Genbank accession #U94788 and X54156) (26). Primers were
designed to span each exon of the p53 open reading frame and sufficient bases of the
intronic sequence to ensure the splice donor and splice acceptor sites were included for
analysis (26).
Single-strand Conformational Polymorphism fSSCPi and DNA Sequence
A nalysis
The SSCP technique was initially used as a screen for DNA sequence alterations
in p53 exons 2 through 11 as described elsewhere (25). Exons with altered mobility
were analyzed for DNA sequence changes as described (25) using the CircumVent
Thermal Cycle Sequencing Kit (New England Biolabs, Beverly, MA) or the
ThermoSequenase Kit (Amersham, Arlington Heights, EL), according to the
manufacturer’s instructions.
p53 Immunohistochemistrv
p53 protein was identified in tissue using the peroxidase anti-peroxidase
technique. Frozen sections, 4pm thick, were incubated with anti-human p53 mouse
monoclonal antibody (DO-7 or Ab240, Dako Corporation, Carpinteria, CA or Zymed
Laboratories, South San Francisco, CA, respectively) as previously described (26).
The percentage of positively immunostained tumor cell nuclei was determined from the
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
number of nuclei containing immunoreaction product divided by the total number of
nuclei, both immunostained and unstained. A minimum of 100 tumor cells were scored.
Those breast tissues with p53 immunostaining in at least 10% of the cell nuclei were
considered to have p53 overexpression, while those with less than 10% p53
immunostained nuclei were considered to be within the normal range of p53 expression
(26).
Statistical Analyses
The association between p53 overexpression and p53 mutations was evaluated
using Fisher's exact test. The Fisher Exact test was also used to evaluate potential
associations between estrogen receptor (ER), progesterone receptor (PR), epidermal
growth factor receptor (EGFR) and the presence of p53 mutations. The Mantel-
Haentzel test was used to evaluate potential associations between HER-2/ne« receptor
immunostaining and the presence of p53 mutations and overexpression.
Re su lts
p53 M utations
SSCP screening identified altered mobility patterns in 25 of 58 cases for at least
one of the exons evaluated (Figure 1-1). Sequencing confirmed the presence of
mutations in 12 cases (Figure 1-2 and Tables 1-1, 1-2), DNA sequence alterations of
unknown significance in three cases and DNA polymorphisms in 8 cases. “DNA
mutations” refer to changes in the coding sequence giving rise to a different amino acid
sequence. In contrast, “DNA sequence alterations of unknown significance” refer to
changes in bases outside of the coding, or exon, DNA sequence, such as base changes
28
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
in the intron sequence or after the stop site. “DNA polymorphisms” refer to base
changes that are either silent and do not code for different amino acids, or base changes
which code for different amino acids but are not considered to create a defective p53
protein.
In 10 of 27 (37%) breast carcinoma in situ cases with no concurrent invasive
carcinoma there were six (22%) mutations, two (7.5%) alterations of unknown
significance (Table l-l) and two (7.5%) polymorphisms (Table 1-4). Four of 6
mutations, and both alterations of unknown significance, had comedocarcinoma
histology. Two mutations had papillary histology, while one polymorphism had
papillary histology and one had solid histology (Table 1-1). The DNA sequence
alterations were distributed between exons 4 and 11. One deletion of a G in codon 112
was predicted to result in a shift in the open reading frame with a Gly 112Ala and a stop
at codon 122 in exon 4. Four mutations, two Arg248Gln and two Arg249Gly, were
found in exon 7. One Ala276Pro mutation was identified in exon 8 and one
Gly325Stop mutation was identified in exon 9. A C->T polymorphism (CCG->CCA)
was found in the 4th exon at codon 36, a C->G polymorphism (CCC->CGC) at codon
72 and two identical T->A changes, at nucleotide 18717 in the p53 genomic sequence,
were noted in exon 11 after the termination codon (Table 1-1).
Four cases had sufficient DNA from benign epithelium for DNA sequence
analysis. No sample of benign epithelium had a p53 mutation when assessed by SSCP
and this was confirmed by DNA sequence analysis in two cases with a p53 mutation in
the carcinoma. However, one case with a DNA sequence alteration of unknown
significance, a T->A base change (at nucleotide 18717) after the termination site in exon
11, did contain the same sequence alteration in adjacent benign tissue as would be
expected for a DNA polymorphism.
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Among the 31 cases of breast CIS with invasive disease, 6 (19%) had
mutations, 1 had a DNA sequence alteration of unknown significance and 6 had DNA
polymorphisms. Four of the mutated cases had comedocarcinoma histology, one had a
solid histology, and one had a cribriform histology. The alteration of unknown
significance had a micropapillary histology. The 6 polymorphisms had 1 case with
comedocarcinoma, 2 cases with solid histology, and 3 cases with cribriform histology
(Table 1-2). The DNA alterations occurred between the 3rd intron and the 9th exon.
One alteration of unknown significance occurred in the 3rd intron, a 3 base
insertion beginning at base 11992. A Glnl65Stop mutation was noted in exon 5. Two
mutations were in exon 6, coding for a Leul94De and a Tyr205Ser change. Two
mutation occurred at the DNA binding codons, an Arg248Trp mutation was found in
exon 7 and an Ala276Pro change in exon 8. Sixteen bases were deleted in exon 9, a
loss of codons 323 to 328. Among these cases, there were a total of 6 polymorphisms:
2 in codon 72 of the 4th exon, a C->G transversion (Pro72Arg), and 4 A->C
transversions at codon 213 (Arg213Arg). Comparisons of the mutations in the CIS and
invasive components in all 5 cases with mutations and sufficient tissue available revealed
the same mutation in both components. A sixth case with a DNA mutation did not have
sufficient invasive material for analysis. In cases lacking p53 mutation in the breast CIS
component, no mutations were detected in the invasive component. As in the cases with
only CIS, no mutations of p53 were found in the benign epithelium in 2 cases with p53
mutations in both CIS and invasive disease (Table 1-2).
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1*1: SSCP analysis of exon 8 of p53 in DCIS.
Lanes marked wt (wildtype) show typical two band conformation, representing the two
complementary DNA strands of DNA. Lane marked MT (mutant) show the typical 4 band
conformation, with 2 bands migrating at the same rate as normal DNA, and 2 migrating
more slowly.
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p 53 Protein Expression.
In the entire cohort, 17 of the 58 cases (29%) showed p53 overexpression by
immunohistochemical staining in the carcinoma. Two cases had p53 overexpression
only in benign hyperplastic tissue. p53 protein was either exclusively or predominantly
nuclear (Figure 1-2). Cytoplasmic staining without nuclear staining was not identified.
In frozen tissue sections immunostaining is observed in nuclei of a low percentage of
normal, proliferatively active tissues. Therefore, we have used 10% immunostained
tumor cell nuclei as a value for separation of “normal expression” from
“overexpression” (26). The percentage of ceils with nuclear staining for p53 varied
from 2% to 90%. Only two cases had less than 10% of tumor cell nuclei positively
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1-2: p53 Immunohistochemistry and DNA sequencing on a Single Tumor
One representative case with sufficient invasive and in situ carcinoma as well as benign
ductal epithelium for DNA extraction were subjected to microdissection followed by DNA
extraction. Each separate morphology is represented in the top 3 photographs titled H&E,
showing normal breast epithelium (normal breast), breast ductal carcinoma in situ (Breast
CIS), and invasive breast carcinoma (invasive cancer). These different parts of the same
tumor were extracted and processed separately. Immunostaining for p53 using the mouse
monoclonal antibody D 07 is depicted in the second row of photographs (marked p53).
P53 overexpressing nuclei are identified by the brown diaminobenzidine
immunoprecipitate only in breast CIS and invasive disease. DNA sequencing (third row o f
photographs marked DNA sequencing) shows a truncating deletion in exon 9 of p53 only
in breast CIS and invasive disease. On the right of each photograph is the normal
sequence and on the left is the mutated sequence.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
immunostained. One of these had a p53 polymorphism and the other had wild-type
p53.
p53 overexpression was noted in 6 of 27 (22%) cases of breast CIS without
invasion (Table 1-1). The percent of nuclei immunostained ranged from 10% to 47%
and the average staining was 32.5%. Among the 6 cases with p53 overexpression, 5
were classified as comedocarcinoma and one as papillary CIS. Six cases had sufficient
benign epithelium for interpretation of immunostaining. Four cases with benign ductal
epithelium lacked p53 immunostaining, while two cases exhibited p53 immunostaining
in areas of benign ductal hyperplasia. Neither case with benign hyperplasia had a p53
mutation in the CIS portion. The amount of hyperplastic breasttissue available from
these cases was insufficient for DNA sequence analysis (Table 1-1). Eleven (35%) of
31 cases of CIS with invasive disease had p53 overexpression (Table 1-2). Nine cases
with CIS components had comedocarcinoma histology, one had a micropapillary
histology, and one had a cribriform histology. The percent of nuclei immunostained
ranged from 14% to 90% and the average staining among these samples was 41.5%.
Five cases with p53 overexpression in the CIS had no expression in the normal breast
epithelium. Only one case (C425) with both p53 immunostaining and p53 mutation had
sufficient benign DNA for sequencing. The benign tissue from this case did not have a
mutation in 3 separate areas sampled.
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Number Histology p53 Immunohistochemistry DNA Sequence Analysis
CIS nl epithelium Exon Codon type
C804 comedo 0% 14% hyp none no mutation
Cl 28 comedo 0% 20% hyp none no mutation
-
C723 solid 23% NA none no mutation
.
C2885* comedo 0% 0% 4 112delG nonsense
C953 comedo 59% 0% 7 Arg248Gln missense
Cl 022 comedo 34% NA 7 Arg248Gln missense
C693* papillary 0% NA 7 Arg249Gly missense
Cl 026 comedo 0% NA 7 Arg249Gly missense
Cl 046 comedo 10% 0% 9 Gly325Stop nonsense
C96* comedo 47% 0% 1 1 post stop1 AUS
C2545 comedo 40% NA 1 1 post stop1 AUS
* denotes cases with sufficient benign ductal epithelium for DNA analysis.
IHC = immunohistochemistry, NA = not available, AUS = alteration of unknown significance, hyp = hyperplasia, 1DC = invasive ductal carcinoma
'This AUS is located at nucleotide 18717, Genbank Accession #X54154
Table 1-2:
Breast CIS
p53 Alterations in Breast Carcinoma in situ With Invasion
Invasive Breast Carcinoma
Number Histology p53 immunohistochemistry DNA Sequence Analysis Histology p53
IHC
Exon Codon
CIS nl epithelium Exon Codon type
C2330 comedo M 0% none no mutation IdC 24%
•
C2798 comedo 25% NA none no mutation
-
IDC 30%
.
C2829 comedo 20% NA none no mutation
-
IDC 30%
_
C3124 comedo 60% NA none no mutation
-
IDC 70%
_ _
C1518 comedo 40% NA none no mutation
.
IDC 45%
.
C470 micropap 14% 0% 3rd intron cctctga->ccAtcCtTga AUS IDC NA
. -
C2929 solid 0% 0% 5 Glnl65Stop nonsense IDC 0% 5 165de!G
C1989* comedo 24% 0% 6 Leu 194H e
missense IDC 22% 6 Leul94Ue
C2325 comedo 53% 0% 6 Tyr205Ser missense IDC 50% 6 Tyr205Ser
C1752 cribriform 65% NA 7 Arg248Trp missense IDC 61% 7 Arg248Trp
C425* comedo 45% 0% 8 Ala276Pro missense NA NA NA NA
C2666 comedo 90% NA 9 del: codons 323-328 nonsense IDC 90% 9 del: codons 323-328
del=deletion
♦ denotes cases with sufficient
2 This AUS sequence starts at
benign ductal epithelium for DNA analysis.
nucleotide 11900 in the p53 genomic sequence (Genbank # X54154)
u >
Os
Comparison of p53 Mutations with Protein Overexpression
Overall, 11 of the 58 cases (19%) had both p53 overexpression and either
mutations or sequence alterations of undetermined significance, 37 had neither mutations
nor overexpression, 6 had overexpression without mutations and 4 had a p53 mutation
but lacked overexpression. Overexpression of p53 was highly correlated (p<0.0001,
Fisher’s Exact test) with the presence or absence of mutations / sequence alterations
(Table 1-4).
Among the 27 breast CIS cases without invasive disease, 5 had both p53
mutations or alterations of unknown significance and overexpression, 18 had neither
mutations or overexpression, 3 had p53 mutations but lacked overexpression, and 1 had
overexpression but lacked p53 mutations. Among the 31 breast CIS cases with
invasive carcinoma, 6 cases had both p53 mutations or DNA sequence alterations of
unknown significance and overexpression, 19 had neither p53 mutations nor
overexpression, 5 had p53 overexpression but lacked a p53 mutation, and 1 had a p53
mutation but lacked p53 overexpression. Comparison of immunostaining and DNA
sequence alterations revealed a significant association between overexpression and p53
alteration (p=0.0037) in cases of CIS without invasion, and similarly between
overexpression and DNA alterations in cases of O S with invasion (p=0.007).
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 1-3: p53 polymorphisms in Breast Carcinoma in situ
N um ber H isto lo g y Presence of
Invasion
DNA Sequence Analysis
E xon Codon sequence amino acid
C3540 papillary NO 4 36 CCG->CCA Ser->Ser
C2398 solid NO 4 72 CCC->CGC Pro->Arg
C l 895 cribriform YES 4 72 CCC->CGC Pro->Arg
C2286 solid YES 4 72 CCC->CGC Pro->Arg
C1518 cribriform YES 6 213 CGA->CGC Arg->Arg
C2340 solid YES 6 213 CGA->CGC Arg->Arg
C2775 comedo YES 6 213 CGA->CGC Arg->Arg
C2821 cribriform YES 6 213 CGA->CGC Arg->Arg
Table 1-4: Comparison of p53 overexpression with p53 mutations in 58
cases.
Positive Negative
Any 11* 4 15
Mutation
None 6 37 43
17 41 58
p<0.0001
* three alterations of unknown significance were included in this group.
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
D is c u s s io n
Breast carcinoma in situ (CIS) or intraductal breast carcinoma is the earliest
morphologically recognizable form of breast cancer. Breast CIS is confined to the
lumen of breast ducts and lobules without penetration of the basement membrane and,
therefore, without invasion of the breast stroma. However, cytological characteristics of
intraductal tumor cells are essentially indistinguishable from the cytological
characteristics of invasive breast carcinomas and intraductal carcinomas are considered
to represent an early stage in a continuum of breast neoplasia. Circumstantial evidence
supports this view. Breast CIS is associated with an increased risk of subsequent
development of invasive breast carcinoma. Overexpression of HER-2/nen (c-erb B-2)
and other oncoproteins, including p53 tumor suppressor protein product, known to be
frequently overexpressed in invasive breast cancer, are also frequently overexpressed in
breast CIS. Although overexpression of oncoproteins have been observed in breast
CIS, these are immunohistochemical staining studies of archival, paraffin-embedded
tissues which do not provide direct evidence of genetic alterations.
In previous reports, a total of 99 breast CIS cases have been screened for p53
mutations using constant denaturant gel electrophoresis (21), immunohistochemical
staining (24, 27) and/or SSCP (23, 24). Only six potential p53 gene mutations were
confirmed by DNA sequence analysis in these studies (21, 23, 24, 27). The p53 gene
mutations include five missense mutations: Arg202His (23), Met237He (23),
Gly248Asp (27), two Arg273His mutations (21, 24), and one frame-shift mutation, a
single-base deletion which is predicted to result in the introduction of a premature
termination site at codon 304 (23) (Figure 1-4). The frequency of confirmed p53
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mutations in the breast CIS cohorts screened was 3%, 4%, 7% and 11% (21, 23, 24,
27).
The present study shows that the frequency of p53 mutations in breast CIS
lacking invasive carcinoma (22%) and the frequency of p53 mutations in breast CIS
with invasive carcinoma (19%) was similar. There was also no evidence of an increase
in p53 overexpression during progression from CIS to invasive disease within the same
tumor (Table 1-2). Those breast tumors with both a CIS component and an invasive
component with wild-type p53 in the CIS component by DNA sequence analysis also
had wild-type p53 in the invasive component (data not shown). Furthermore, p53
mutations identified in a CIS component by DNA sequence analysis had the identical
mutation in the invasive component of the tumor confirming the clonal relationship
between CIS and invasive disease (Table 1-2). However, the frequency of p53
overexpression in breast CIS lacking invasion (22%) was not similar to the frequency of
p53 overexpression in breast CIS cases having invasion (35%). In the current study, no
increase in p53 mutations was observed between CIS and invasive disease nor was the
observed increase in p53 overexpression during progression statistically significant.
These results suggest that p53 mutations are relatively ‘early’ in progression of the
disease process, occurring primarily in intraductal breast carcinomas.
The estrogen receptor (ER), progesterone receptor (PR), epidermal growth
factor receptor (EGFR) and HER-2/new oncoprotein content had been assessed
separately in these tumors using immunohistochemistry and DNA ploidy in these tumors
was assessed using Feulgen staining and computerized image analysis (Lukas,
Dahmoush, Li, Tehrani, Petrosyan and Press, unpublished observations). Comparison
of p53 mutations and overexpression with these tumor markers revealed no associations
with ER, PR, EGFR or DNA ploidy. There was, however, a statistically significant
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
association between high levels of HER-2/new protein expression and p53 mutations
(p=0.019). This is probably related to the high proportion of comedocarcinomas
associated with both p53 alterations and HERUneu overexpression. High HER-2/neu
expression has been previously reported as showing a correlation with p53
overexpression in invasive breast carcinoma (28). However, we and others have
demonstrated a lack of association between HER2/neu overexpression and p53
mutations in endometrial carcinomas (29).
The structure of p53: whv mutations alter function
The crystal structure analysis of p53 protein provides new insight regarding
several important functional aspects of this molecule. The p53 core domain broadly
consists of a sheet structure which supports loops and helices that interact with DNA.
The support structure consists of two antiparallel B sheets which pack together to form a
sandwich with a hydrophobic core. This sandwich supports the LI, L2 and L3 loops,
as well as the HI and H2 helices which interact with DNA. These loops and helices
lack secondary structure, and thus stabilization is largely accomplished by the sharing of
a zinc ion and several interactions between side chains or the amino acid backbone (30-
32). Mutations which effect DNA binding as well as mutations which effect the
structure of the core domain as a whole can disable p53 function.
The p53 mutations reported here can be divided into 3 groups. Mutations were
found in sites which directly interact with DNA (Arg248Gln, Arg248Trp, Ala276Pro),
in sites which preserve p53 structural integrity (112delG, Glyl65Stop, Leul94De,
Tyr205Ser, Arg249Gly) and in sites outside the core domain (Gly325Stop, del323-328,
and alterations of unknown significance). Among the 12 mutations, 5 interfere directly
with p53/DNA binding, 6 interfere with internal stabilization of p53 and one codes for a
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
termination site in the 3’ oligomerization domain. Three alterations of unknown
significance were also outside the core domain and are discussed separately.
Three mutations were at Arg248 and, therefore, interfere with p53/DNA
binding. This codon, known to be critical to the anchoring of p53 in the minor groove
of DNA with 4 hydrogen bonds from 3 nitrogen atoms, is the most frequently mutated
codon. In this cohort, Arg248 is mutated to Arg248Gln and Arg248Trp, neither having
the number of nitrogen atoms necessary for bonding. Ala276 is important to p53/DNA
binding as the nitrogen of alanine binds to the phosphate group of DNA. The
Ala276Pro mutant does not have an nitrogen and will not bind DNA correctly. Each of
these incorrect amino acids cannot form bonds necessary for p53 stabilization and
function.
Six mutations effect the ability of p53 binding to DNA indirectly by altering the
structural integrity of the p53 molecule. Mutations in codon 112 and 165 create
premature stop codons, resulting in a truncated p53 protein. An amino acid change,
Leul94Ile at the junction of the L2 loop and the B sheet (Figure 4), probably indirectly
destabilizes the Arg248 binding (33). The Tyr205Ser mutation changes the central
codon in the S6 B strand. Here the B sheets are tightly packed together forming the
hydrophobic core (30). The loss of the tyrosine’s hydrophobic phenolic side chain may
alter the ability of these sheets to interact properly and, therefore, alter the protein
structure. Arg249 is a frequently altered codon which links the L2 loop to the L3 loop
with hydrogen bonds, stabilizing the Arg248 interaction in the minor groove. Two
cases have an Arg249Gly mutation and this change would destabilize the L3 interaction
in the minor groove. These mutations effect the overall structure of p53, and can have
an impact on the ability of p53 to bind DNA.
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Gly325 and del323-328 are in the oligomerization domain, 3’ to the DNA
binding ‘core’ of p53. This domain extends from codon 319 to 360, consisting of a
bend, a 6 strand, another short bend, and an a helix, and is required for p53 protein to
properly tetramerize (34, 35). A termination codon at Gly325, which corresponds to the
first bend in the domain, would create a truncated mutant that would abolish proper
oligomerization. Similarly, the del323-328 would also abolish proper oligomerization.
The Arg2l3 alteration does not code for an amino acid change. This codon is
the most frequently altered non-missense codon (36) and is reported to be a polymorphic
site (37). Noncoding DNA sequence alterations have been reported at high frequency in
sporadic breast tumors (38) and certain silent mutations at the wobble nucleotide in the
fibroblast growth factor receptor 2 gene do appear have phenotypic effects (39),
possibly through the opening of a cryptic splice site.
The alterations of unknown significance resist easy classification. One mutation
effects p53 outside of the core domain which nonetheless exhibits p53 overexpression.
This C-to-A alteration in the third intron is of undetermined significance because the
change does not appear to open any cryptic splice sites. However, these alterations do
appear to fall within the putative lariat branch point (40). Loss of this site could impair
the proper splicing out of this intron. The overexpression of p53 in this case (14% of
nuclei immunostained for p53 protein) argues that this alteration is significant. Two
cases with identical nucleotide 18717 T-to-A alterations after the termination codon, had
overexpression of p53 protein (47% and 40%), which similarly suggested this DNA
sequence alteration may effect the expression of p53. Recent work with the human a -
globin gene delineated a possible mechanism for p53 overexpression. In eukaryotes,
the 3’ untranslated region of mRNA is important as the repository for signals which
determine RNA localization, polyadenylation, translation initiation and RNA stability
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(41). In particular, the stability of a-globin mRNA is crucial to proper erythroid cell
differentiation, a-globin mRNA stability is specifically modulated by the binding of a
ribonucleoprotein complex to specific polycytosine runs in the 3’ untranslated region of
the gene (42). In p53, the site of the T -> A alteration is 5’ - CCTCCCT/ACCCC - 3 ’,
which is an exact match for the consensus binding site in a-globin mRNA (underlined).
Substitutions in the consensus binding site destabilizes the a-globin mRNA (43). The
evidence presented here suggests that a similar poly-C binding protein may also
influence the stability of p53 mRNA. Alterations such as the T->A substitution may
increase translation of p53 protein in cases with specific alterations at these consensus
sequences.
In conclusion, this study demonstrated genetic alterations of p53 as well as
overexpression of the p53 protein. A number of conclusions can be reached based on
this data. 1). Although prior studies indicated that there was an increasing percentage
of mutations inp53 in invasive disease when compared to CIS, our results show that
there is no change in p53 mutation rate. Because p53 mutations were in 22% of cases
of CIS alone and 19% of cases of CIS with invasive disease, and this is the same
percentage of p53 mutations in invasive breast cancers, the p53 gene is mutated prior to
or at the formation of the CIS lesion. 2). Identical mutations in both the CIS and
invasive lesions in the same tumor further support this notion that p53 mutations take
place prior to invasion and, furthermor, are evidence for a clonal relationship between
carcinoma in situ and invasive disease. 3). Although the results were not statistically
significant, the differences in the percentage of cells that were p53 immunoositive
between CIS alone and CIS with invasive disease suggest the possibility that the
transition between non-invasive and invasive breast disease is characterized by
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
alterations which indirectly effect expression of p53. Overall, this study shows that p53
mutations are important in pre-invasive breast disease.
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 1-3: p53 mutations in DCIS compared to the molecular structure of p53.
Top: The graph is a comparison of the number of mutations found the present study
(above the bar marked codon) and the number of mutations described in previous studies
(above the codon bar) and number of mutations described in previous studies (below the
codon bar) against the position of the mutations in the p53 protein. Mutations in red are
from cases of breast CIS with invasion; mutations in these cases were found in both DCIS
and invasive components. Mutations denoted in green are from cases of breast DCIS with
invasion. Grey regions in the codon bar are evolutionarily conserved regions. Bottom:
Mutation spectrum compared to the molecular structure of the p53 protein. In yellow,
regions of p53 that interact with DNA, while in green are regions that serve as support for
the DNA interacting domains.
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 1-3: p53 mutations compared to the structure of the p53 protein
3 1
o
♦ 3
< 0
s S
exon 4 I exon 5 I exon 6 I exon 7 exon 8 I exon 9 I exon 1 0 >
24*
t 2 1 «t 1 4 102
t — i — i — r*n— i — n
. .I'M . . ; — W - H
110 120 130 140 ISO 160 170 180 190 2 o I 210 220 230 1240 >250 260 27(
CO don (number)
270
ji
020 .
I
r*n— rpi— i — i — r -i I
280 290 3001 310 320 330 340 350
7 !
a}
1-
~ i 1 - - - - - -1 T TTI— I — T
I ? 4 Q l2 S
h >
Sequence specific DNA binding domain
1 ^ oligomerization
domain
-® r
= DNA interactive regions
= 'scaffolding' or structural regions
R eferences
1. Allred, D., O'Connell, P., and Fuqua, S. Biomarkers in early breast neoplasia, J
Cell Biochem. 17G: 125-131, 1993.
2. Page, D. L., Dupont, W. D., Rogers, L. W., Jensen, R. A., and Shuyler, P. A.
Continued local recurrence of carcinoma 15-25 years after a diagnosis of low
grade ductal carcinoma in situ of the breast treated only by biopsies, Cancer. 76:
1197-2000, 1995.
3. Eyfjord, J., Thorlacius, S., Steinarsdottir, M., Valgardsdottir, R.,
Ogmundsdottir, H., and Anamthawat-Jonsson, K. p53 abnormalities and
genomic instability in primary human breast carcinomas, Cancer Res. 55: 646-
651, 1995.
4. Harris, A. p53 expression in human breast cancer, Vol. 59, p. 69-88: Academic
Press, Inc., 1992.
5. Thorlacius, S., Borresen, A.-L., and Eyfjord, J. Somantic p53 mutations in
human breast carcinomas in an Icelandic population: a prognostic factor, Cancer
Res. 53: 1637-1641, 1993.
6. Harris, C. and Hollstein, M. Clinical implications of the p53 tumor suppressor
gene, New Engl J Med. 529: 1318-1327, 1993.
7. Barnes, D., Dublin, E., Fisher, C., Levinson, D., and Millis, R.
Immunohistochemical detection of p53 protein in mammary carcinoma: a
important independent indicator of prognosis?, Hum Pathol. 2 4 :468-476, 1993.
8. Allred, D., O'Connell, P., Fuqua, S., and Osborne, C. Immunohistochemical
studies of early breast cancer evolution, Breast Cancer Res Treat. 32: 13-18,
1994.
9. Barbareschi, M., Caffo, O., Doglioni, C., Fina, P., Marchetti, A., Buttitta, F.,
Leek, R., Morelli, L., Leonardi, E., Bevalacqua, G., Dalla Palma, P., and
Harris, A. p21W A FI immunohistochemical expression in breast carcinoma:
correlations with clinicopathological data, oestrogen receptor status, MIB1
expression, p53 gene and protein alteration and relapse-free survival, Brit J
Cancer. 74: 208-215, 1996.
10. Eriksson, E. T., Schimmelpenning, H., Aspenblad, U., Zetterberg, A., and
Auer, G. U. Immunohistochemical expression of the mutant p53 protein and
nuclear DNA content during the transition from benign to malignant breast
disease, Hum Pathol. 25: 1228-1233, 1994.
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
11. Bobrow, L. G., Happerfield, L. C., Gregory, W. M., and Millis, R. R. Ductal
carcinoma in situ: assessment of necrosis and nuclear morphology and their
association with biological markers, J Pathol. 176: 333-341, 1995.
12. Rajan, P. B., Scott, D. J., Perry, R. H., and Griffith, C. D. M. p53 protein
expression in ductal carcinoma in situ (DCIS) of the breast. Breast Cancer Res
Treat. 42: 283-290, 1997.
13. Siziopikou, K. P., Prioleau, J. E., Harris, J. R., and Schnitt, S. J. bcl-2
expression in the spectrum of preinvasive breast lesions, Cancer. 77: 499-506,
1996.
14. Leal, C., Schmitt, F., Bento, M., Maia, N., and Lopes, C. Ductal carcinoma in
situ: histologic categorization and its relationship to ploidy and
immunohistochemical expression of hormone receptors, p53, and c-erbB-2,
Cancer. 75: 2123-2131, 1995.
15. Perin, T., Canzonieri, V., Massarut, S., Bidoli, E., Rossi, C., Roncadin, M.,
and Carbone, A. Immunohistochemical evaluation of multiple biological markers
in ductal carcinoma in situ of the breast, Eur J Cancer. 32A: 1148-1155, 1996.
16. Poller, D. N., Hutchings, C. E., Galea, M., Bell, J. A., Nicholson, R. A.,
Elston, C. W., Blarney, R. W., and Ellis, I. O. p53 protein expression in human
breast carcinoma: relationship to expression of epidermal growth factor receptor,
c-erbB-2 protein overexpression, and oestrogen receptor, Brit J Cancer. 66: 583-
588, 1992.
17. Poller, D. N., Roberts, E. C., Bell, J. A., Elston, C. W., Blarney, R. W., and
Ellis, I. O. p53 protein expression in mammary ductal carcinoma in situ:
relationship to immunohistochemical expression of estrogen receptor and c-
eri?B2-protein, Hum Path. 24: 463-468, 1993.
18. Thor, A. D., Moore, D. H., Edgerton, S. M., Kawasaki, E. S., Reihsaus, E.,
Lynch, H. T., Marcus, J. N., Schwartz, L., Chen, L.-C., Mayall, B. H., and
Smith, H. S. Accumulation of p53 tumor suppressor gene protein: an
independent marker of prognosis in breast cancers, J Natl Cancer Inst. 84: 845-
855, 1992.
19. Walker, R. A., Dearing, S. J., Lane, D. P., and Varley, J. M. Expression of p53
protein in infiltrating and in situ breast carcinomas, J Pathol. 165: 203-211,
1991.
20. Zafrani, B., Leroyer, A., Fourquet, A., Laurent, M., Trophilme, D., Validire,
P., and Sastre-Garau, X. Mammographically-detected ductal in situ carcinoma of
the breast analyzed with a new classification. A study of 127 cases: correlation
with estrogen and progesterone receptors, p53 and c-erbB-2 proteins, and
proliferative activity, Sem Diag Pathol. 11: 208-214, 1994.
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
21. Chitemerere, M., Andersen, T. I., Holm, R., Karlsen, F., Borresen, A.-L., and
Nesland, J.-M. TP53 alteration in atypical ductal hyperplasia and ductal
carcinoma in situ of the breast, Breast Cancer Res Treat 41: 103-109, 1996.
22. Davidoff, A. M., Kerns, B.-J., Iglehart, J. D., and Marks, J. R. Maintenance of
p53 alterations throughout breast cancer progression, Cancer Res. 51: 2605-
2610, 1991.
23. Munn, K. E., Walker, R. A., Menasce, L., and Varley, J. M. Mutation of the
TP53 gene and allelic imbalance at chromosome 17pl3 in ductal carcinoma in
situ., Br. J. Cancer. 74: 1578-1585, 1996.
24. O'Malley, F. P., Vnencak-Jones, C. L., Dupont, W. D., Pari, F., Manning, S.,
and Page, D. L. p53 mutations are confined to the comedo type ductal carcinoma
in situ of the breast, Lab Invest. 71: 67-72, 1994.
25. Lukas, J., Groshen, S., Saffari, B., Niu, N., Reles, A., Wen, W.-H., Felix, J.,
Jones, L., Hall, F., and Press, M. WAFl/Cipl gene polymorphism and
expression in carcinomas of the breast, ovary and endometrium, Am J Pathol.
150: 167-175, 1997.
26. Wen, W.-H., Reles, A., Sullivan-Halley, J., Bernstein, L., Jones, L., El-
Naggar, A., Felix, J., Runnebaum, I., and Press, M. p53 mutations and
expression in ovarian cancers: correlation with overall survival, Int J Gyn One.
in press:, 1998.
27. Davidoff, A. M., Kerns, B.-J. M., Pence, J. C., Marks, J. R., and Iglehart, J.
D. p53 alterations in all stages of breast cancer, J Surg Oncol. 48: 260-267,
1991.
28. Martinazzi, M., Crivelli, F., Zampatti, C., and Martinazzi, S. Relationship
between p53 expression and other prognostic factors in human breast carcinoma.
An immunohistochemical study., Am J Clin Pathol. 100: 213-217, 1993.
29. Saffari, B., Runnebaum, I., Bernstein, L., Hong, D., Niu, N., Varkey, T.,
Grill, H.-G., Jones, L., El-Naggar, A., and Press, M. Association of p53 gene
mutations and expression with shortened overall survival in endometrial cancer,
Am J Pathol, submitted:, 1998.
30. Cho, Y., Gorina, S., Jeffrey, P., and Pavletich, N. Crystal structure of a p53
tumor suppressor-DNA complex: understanding tumorigenic mutations..
Science. 265: 346-355, 1994.
31. Arrowsmith, C. and Morin, P. New insights into p53 function from structural
studies, Oncogene. 12: 1379-1385, 1996.
32. Friend, S. p53: a glimpse at the puppet behind the shadow play, Science. 265:
334-335, 1994.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
33. Prives, C. How loops, beta sheets and alpha helices help us to understand p53,
Cell. 78: 543-546, 1994.
34. Clore, G., Omichinski, J., Sakaguchi, K., Zambrano, N., Sakamoto, H.,
Appella, E., and Gronenbom, A. High-resolution structure of the oligomerization
domain of p53 by multidimensional NMR, Science. 265: 386-391, 1994.
35. Clore, G., Ernst, J., Clubb, R., Omichinski, J., Poindexter Kennedy, W.,
Sakaguchi, K., Appella, E., and Gronenbom, A. Refined solution structure of
the oligomerization domain of the tumor suppressor p53, Nat Struct Biol. 2: 321-
333, 1995.
36. Harris, C. Structure and function of the p53 tumor suppressor gene: clues for
rational cancer therapeutic strategies, J Natl Cancer I. 88: 1442-1455, 1996.
37. Carbone, D., Chiba, I., and Mitsudomi, T. Polymorphism at codon 213 within
the p53 gene, Oncogene. 6: 1691-1692, 1991.
38. Glebov, O., McKenzie, K., White, C., and Sukumar, S. Frequent p53 gene
mutations and novel alleles in familial breast cancer, Cancer Res. 54: 3703-3709,
1994.
39. Reardon, W., Winter, R. M., Rutland, P., Pulleyn, L. J., Jones, B. M., and
Malcolm, S. Mutations in the fibroblast growth factor receptor 2 gene cause
Crouzon syndrome, Nat Genet. 8: 98-103, 1994.
40. Watson, J., Hopkins, N., Roberts, J., Steitz, J., and Weiner, A. Molecular
Biology of the Gene, Vol. 1: Benjamin/Cummings, 1987.
41. Jackson, R. Cytoplasmic regulation of mRNA function: the importance of the 3'
untranslated region, Cell. 74:9-14, 1993.
42. Kiledjian, M., Wang, X., and Liebhaber, S. Identification of two KH domain
proteins in the alpha-globin mRNP stability complex, EMBO J. 14: 4357-4364,
1995.
43. Wang, X., Kiledjian, M., Weiss, I., and Liebhaber, S. Detection and
characterization of a 3' untranslated region ribonucleoprotein complex associated
with human a-globin mRNA stability, Mol Cell Bio. 15: 1769-1777, 1995.
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 2: W AFl/Cipl Gene Polymorphism
and Expression in Carcinomas of the Breast,
Ovary and Endometrium
A b s t r a c t
The p53 gene is altered in approximately 50% of human cancers and is
considered to be important in the pathogenesis of these malignancies. p53 protein
product regulates the transition from G1 to S phase of the cell cycle and entry to the
DNA damage repair pathway. Since alterations in this pathway appear to be important
in a variety of human cancers, downstream effector proteins of p53 are potential sites for
somatic alterations. W AFl/Cipl, also known as WAF1, Cipl, sdil or CAP20, codes
for a 21 kD protein (p21W A F I/c,p l) which was recently described as a universal inhibitor
of cyclins and is thus critical in cell cycle control. Mutations in WAFl/Cipl are
potentially important in human malignancies because they could affect the control of the
cell cycle. To understand if mutations of WAFl/Cipl occur in cancer, we screened 53
cases of invasive breast carcinoma, 35 cases of ductal carcinoma in situ (CIS), 53
ovarian carcinomas, and 47 endometrial carcinomas in the second exon of WAFl/Cipl
(90% of the open reading frame). p21W A F I/c ,p| expression was characterized with
immunohistochemistry. Cells from the blood of 21 normal individuals were also
characterized using single-strand conformational polymorphism analysis, DNA
sequencing and restriction analysis. Single-strand conformational polymorphism
analysis demonstrated an altered mobility pattern for exon 2 in 12 invasive breast
cancers (22.6%), 5 ductal carcinomas in situ of the breast (14%), 8 invasive ovarian
carcinomas (15%), and 9 endometrial carcinomas (19%). In total, 209 samples were
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
screened and 39 cases (18.2%) had an altered codon 31. Each case with altered SSCP,
analyzed by DNA sequencing and/or restriction analysis, showed the same alteration of
codon 3 1, a C to A transversion encoding a change in amino acid sequence from serine
to arginine (31Ser -> 31Arg). DNA from the blood of 21 normal individuals showed
the same alteration in WAFl/Cipl in 4 cases (19%). Furthermore, paired normal tissue
was available for 3 of 20 breast carcinomas with the Ser31Arg transversion. Normal
DNA from all three cases showed the same 31Arg alteration as found in the tumor
tissue. These results indicate that codon 31 is a polymorphic site and that the serine to
arginine shift is a polymorphism. p21W A F I/G p l expression, identified by
immunohistochemistry, was found to vary in a pattern that depended both on the tissue
type and on the presence or absence of the codon 31 polymorphism. Using pair-wise
comparisons in breast CIS, we found higher protein expression in tumor nuclei as
compared to benign stromal cell nuclei (p = 0.002) or normal ductal epithelium (p =
0.005). Invasive breast cancer specimens showed a trend in p2 l W A F I/c ,p I
immunostaining similar to CIS but did not reach statistical significance (p = 0.12).
However, when cases with extensive desmoplastic reaction were excluded, a statistically
significant association (p = 0.019) similar to that in CIS was noted. In contrast to the
breast tumors, ovarian carcinomas exhibited significantly greater p2 l W A F I/C ip l expression
in the benign stromal (fibroblast) nuclei surrounding the tumor than in the carcinoma cell
nuclei (p = 0.016). Endometrial carcinoma revealed no difference in staining when
comparing benign tissue to carcinoma (p = 0.99); however, unlike breast and ovarian
carcinomas where there was no correlation between p2 i W A F ,/c ,p l expression and the
presence or absence of the alteration at the 31st codon, endometrial carcinomas showed
an increased percentage of immunopositive nuclei associated (p = 0.056) with 31Arg.
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
These results demonstrate tissue-specific expression patterns of WAFl/Cipl in different
tumors which appears to be characteristic of the tumor type.
I n t r o d u c t io n
The p53 tumor suppressor gene product is a transcription factor which is
expressed following DNA damage and induces transcription of downstream genes to
cause either GI arrest or apoptosis. The genes involved in the DNA damage response
pathway are not fully understood; however, one of these genes, referred to as
WAFl/Cipl, has been isolated and characterized (1-5). p53 directly induces
p21 wafi/cipi eX pression through a p53 binding element 2.4 kb upstream of the
WAFl/Cipl open reading frame. Expression of p2lW A F I/C ip l, and subsequent binding of
a cyclin-dependent kinase (CDK), a cyclin, and proliferating cell nuclear antigen
(PCNA), forms a quaternary complex, which is able to either activate or inhibit the CDK
kinase activity (6). A single p21W A F I/C ip l molecule stimulates phosphorylation of factors
facilitating the transition between G1 and S phase. At a greater concentration,
p21W A F I/c ,p l appears to inhibit the kinase reaction of the quaternary complex and halt the
cell cycle. p53 independent p2 l W A F I/c ,p l induction has also been described (7).
Inactivation of p2 iW A F I/c ,p I through alteration could result in unchecked growth and
destabilize the normal phenotype. This possibility was investigated in a series of human
cancers and selected normal tissues.
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
M a t e r ia l s a n d M e t h o d s
T issues
The use of human tissue in this study was reviewed and approved by the U.S.C.
Institutional Research Committee. Frozen tissue from 88 cases of breast cancer from the
U.S.C. Breast Tumor Bank were used in this study. Samples were stored at -80°C in a
Revco freezer. Fifty three cases were classified as strictly invasive disease, with
invasive ductal carcinomas making up 92% and invasive lobular carcinomas accounting
for 8% of the total. Thirteen cases had evidence of both invasive and in situ carcinoma
and 22 cases had solely ductal carcinoma in situ (CIS). Among cases with CIS
components, 57% were classified as comedocarcinoma, 19% papillary CIS, 11%
cribriform CIS, 8% CIS not otherwise specified, and 5% solid CIS. Frozen tissue
sections stained with hematoxylin and eosin were used to assess the histological
composition of the specimens. Normal tissue was also obtained from paraffin
embedded tissue blocks of our institutional archives in 2 of these cases and from frozen
tissue in 1 case.
Frozen tissue from 47 cases of endometrial cancer, including endometrioid
carcinomas (80%), malignant mullerian mixed tumors (14%), clear cell carcinomas
(3%), and serous papillary tumors (3%) were used in this study. Fifty-three epithelial
ovarian carcinomas were also characterized for WAFl/Cipl. The carcinomas included
papillary serous carcinomas (57%), endometrioid carcinomas (17%), poorly
differentiated adenocarcinomas (14%), clear cell carcinomas (6%), mucinous
carcinomas (4%), and mixed carcinomas (4%). Thirty milliliters of blood from a
random sample of 21 normal women (12 cases) and normal men (9 cases) were
analyzed.
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DNA Isolation and PCR
In the cases of breast CIS, a dissecting microscope was used to separate
different tissue components from 10 pm thick, frozen, tissue sections, fixed in 95%
ethanol, and stained with ethyl green. The different components, invasive cancer, ductal
carcinoma in situ, and benign tissues were separately subjected to DNA extraction.
Tissue was digested in a buffer of 10 mM Tris (pH 8.0), 100 mM NaCl, 25 mM HPT A,
0.5% SDS and 0.1 mg/ml proteinase K for 24 hours at 50°C. A mixture of
phenol/chloroform/isoamyl alcohol was used to extract the DNA, which was then
precipitated using ethanol. DNA was extracted from paraffin embedded tissue sections
(20 pm) in two cases of breast ductal CIS, which were then deparaffinized with xylene,
rehydrated through graded alcohols and digested with a proteinase K solution (100 mM
Tris, 4 mM EDTA, 0.5 mg/ml proteinase K for 24 hours at 58°C).
The 450 bp second exon of WAFl/Cipl was selected for analysis because it
codes for 90% of the protein product, including highly conserved sequences, such as
the putative zinc finger motif and part of the nuclear localization signal. The following
oligonucleotide primers were used: 5’-GGCGCCATCTCAGAACCGGC-3’ and 5’-
TGTCATGCTGGTCTGCCGCC-3’.
Single-Strand Conformational Polymorphism Analysis
Alterations in the second exon of WAFl/Cipl were screened by single-strand
conformational polymorphism (SSCP) technique for altered mobility. SSCP was
performed as previously described (8), with certain modifications (9). To increase the
autoradiographic signal of SSCP, two 3 5 S labeled nucleotides, dATP and dCTP
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Amersham) were incorporated with PCR. Separation of the PCR products was
performed with MDE (Mutation Detection Enhancement) gels (AT Biochem) according
to the manufacturer’s specifications. These gels offer improved separation of different
conformations by SSCP.
DNA sequencing
DNA sequencing was performed using the CircumVent Thermal Cycle
Sequencing Kit (New England Biolabs; Beverly, MA), according to the manufacturer’s
instructions. All cases with altered mobility by SSCP were reamplified from the original
genomic DNA and sequenced. The sequences were confirmed by a second round of
PCR amplification and DNA sequencing to rule out the possibility of PCR generated
artifacts.
Restriction A nalysis
Restriction analysis was also performed to confirm the sequence alterations.
Two restriction enzymes, Blp I and Bsm BI (New England Biolabs), distinguish the
normal sequence from an altered sequence at codon 31. Blp I recognizes the normal 5’ -
GCTNAGC - 3’ site and Bsm BI recognizes the sequence 5’ - GCAGAGN - 3’.
Samples (approximately 200 nanograms) of PCR-amplified DNA was digested
separately with each enzyme.
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Immunohistochemistrv
The peroxidase anti-peroxidase technique (10) was used to identify p2 i W A F I/C ip I
protein product in tissue sections. Frozen sections, 4pM thick, were cut and fixed in
picric acid/paraformaldehyde for 15 minutes. Initially, a mouse monoclonal antibody
(EA10; Oncogene Science, Uniondale, NY) and a rabbit polyclonal antibody (C-19;
Santa Cruz, Santa Cruz, CA.) were used for immunostaining of the WAFl/Cipl
protein. Preliminary results of C19 immunostaining showed background cytoplasmic,
membrane and extracellular stromal staining. Thereafter, only EA10 was used for
immunostaining of all of the carcinomas in this report. One hundred nuclei were
counted in each sample in random high power fields. Nuclear immunostaining was
considered positive. The positively immunostained percentage was derived from the
number of nuclei containing immunoreaction product divided by the total number of
nuclei, both stained and unstained. The primary antibody (EA10) was diluted to 5
pg/ml in 10% normal rabbit serum. The secondary antibody was diluted to 20 pg/ml in
10% normal rabbit serum. All cases were evaluated for the presence or absence of
p2j w A F i/c ip i immunostaining.
Statistical Analyses
Fisher’ s exact test was used to compare the differences in p21W A F I/C ip l staining in
cases with and without the codon 31 alteration. Paired t-test was used to compare the
difference in percent of p2 lW A F I /C 3 p l stained cells between tumor and benign populations.
All p values reported are 2 sided.
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R esu lts
An alteration in codon 31 of W AFl/Cipl, consisting of a C->A transversion which
codes for a 31Ser -> 31Arg shift, was observed in approximately 16% of all samples
analyzed by SSCP and DNA sequencing or restriction endonuclease digestion.
Seventeen of the 88 (19%) breast carcinomas had altered mobility by SSCP with a
similar pattern of migration in all cases (Figure 2-1). Twelve of 53 invasive breast
cancers (22.6%) and 5 of 35 ductal carcinoma in situ cases (14%) were identified as
altered by SSCP. Eleven of the 12 invasive breast cancers were ductal carcinomas and
one was a lobular carcinoma. Among the ductal CIS cases, at least one case of each
histologic type showed altered SSCP mobility. Similarly, 8 of 53 (15%) cases of
ovarian cancer and 9 of 47 cases (19%) of endometrial carcinomas had altered
WAFl/Cipl by SSCP (Figure 2-1). All cases, characterized by sequence and/or
restriction analysis, were heterozygous with the exception of 2 cases of invasive breast
cancer which were homozygous for the alteration (Figure 2-2). Five cases of breast
cancer, not altered by SSCP, were confirmed as having a normal DNA sequence. There
was no association between endometrial carcinoma or ovarian carcinoma histologic type
and the WAFl/Cipl alteration.
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2-1. Single strand conformation polymorphism (SSCP) analysis of exon 2 from
WAFl/Cipl in endometrial carcinomas.
Samples 1-3, 5 and 6 show a typical two band conformation, representing the
two complementary strands of DNA. Lanes 4 and 7 show the typical pattern observed
with four DNA bands, two migrating the same as normal DNA and two migrating more
slowly.
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The consistent nature of the alteration, observed at only a single codon, was
suggestive of a polymorphism. To confirm this possibility, normal tissue of 3 breast
cancers with altered W AFl/Cipl were sequenced and the 31Ser -> 31Arg transversion
was observed in all three samples. In addition, SSCP and restriction analysis was
performed on the DNA from peripheral blood cells of normal individuals. Four of 21
(19%) samples had an altered codon 31 in the WAFl/Cipl gene.
Expression of p21W A F 1 /C ip I was characterized in these tissues using
immunohistochemistry. One hundred nuclei were counted in tumor cells and benign
stromal elements in invasive breast cancer, ovarian and endometrial carcinomas. In
breast DCIS, one hundred nuclei from the DOS component, from normal ductal
epithelium and from benign stroma were scored. p2 i W A F I/c ,P I immunostaining was
identified exclusively in the nucleus of both tumor cells and normal cells. Comparisons
were made between the percentages of immunostained nuclei in carcinomatous and
benign cells in a pair-wise fashion for each case.
A substantial proportion of breast, ovarian and endometrial cancer specimens
showed immunostaining for p2 L W A F 1 /C ip l in either the tumor cell nuclei, benign stromal
cell nuclei or both. Among the cases of DCIS, p2 lW A F ,/c ‘ P | was expressed in at least
some nuclei in 33% and 35% of samples with normal ducts and benign stromal nuclei,
respectively, while p2 i W A F I/c ,p‘ immunostaining was noted in at least some
carcinomatous nuclei in 56% of DCIS samples. Among the invasive breast cancer
samples, p2 lW A F I/C |p‘ expression was observed in at least some stromal cell nuclei in
45% of cases, while p2 i W A F I/c ,p' expression was observed in at least some tumor cell
nuclei in 83% of invasive breast cancer cases.
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2-2. DNA Sequence Analysis of Breast Carcinomas in Situ.
The "Normal WAF1" gene contains an AGC codon 31, while the "Altered WAF1" gene
contains an AGA codon 31. The presence of both A and C nucleotides at the third
position of codon 31 in the "Altered WAF1" is consistent with either contamination by
normal stromal components or heterozygosity at this site. Since stromal cells were not
present within the duct lumen and DCIS cells were microdissected from the duct lumen, the
case shows heterozygosity at this site
Altered
WAF1
G A T C
Normal
WAF1
G A T C
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2-3. p2lW A F t /c ,p l immunostaining in breast DCIS, invasive ovarian carcinoma
and endometrial carcinoma.
Frozen tissue sections were immunostained by the peroxidase anti-peroxidase technique
with an anti-p2lW A F I /C ip l antibody, EA10.
A-C. Comparison of p21W A F 1 /c ,p l immunostaining of benign stroma (A), benign ductal
epithelium (B) and intraductal breast carcinoma cells (C) from a case of DCIS with
increased immunostaining in the nuclei of tumor cells relative to the nuclei of benign
epithelial and stromal cells. p2 i W A F ,/c ,< ’1 -containing nuclei are identified by brown
diaminobenzidine immunoprecipitate. Other nuclei are counterstained green by ethyl-
green histologic stain. (A. 1400X; B. 460X; C. 1400X).
D-E. Comparison of benign stromal cells (D) containing p2 iW A F I /c ,p l immunopositive
nuclei and carcinoma cells (E) without p21W A F I 'C ip l immunostaining from an ovarian
carcinoma. Tissue counterstained with ethyl green. (D, E. 1400X).
F-G. p2 iW A F I /c ,p ‘ immunostaining of an endometrial carcinoma with codon 31Arg
alteration of WAF1 showing immunopositive nuclei (F) compared with a low level of
p2 jW A F i/c ,P i jmmunostaining in an endometrial carcinoma with codon 31Ser (G). (1400X).
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In both breast carcinoma in situ and invasive breast cancer the overall percent of
nuclei immunostained was also higher in carcinoma nuclei than the percentage of
immunostained benign nuclei. When the percentages of cells immunostained were
averaged, there was a difference in staining between tumor and benign cells. On
average, in DCIS cases, 9% of carcinomatous nuclei were immunostained for
p21W A F ,/C lp I, whereas only 3% of normal ductal nuclei were immunostained and 4% of
benign stromal nuclei were immunostained forp21W A F ,/G p I. Samples of invasive breast
cancer showed a similar trend. Tumor cells had, on average, 10% of nuclei
immunostained while the surrounding benign tissue had a lower average staining of 5%.
Reactive desmoplasia, identified as broad sheets of dense connective tissue with
fibroblasts, was observed in 13 invasive breast carcinomas and contained a high
percentage of p21W A F I/c ,p l stained nuclei. When cases exhibiting extensive reactive
desmoplasia were excluded, tumors cells still had, on average, 10% of nuclei positive
for p21W A F 1 /c ip l, while the benign stromal elements expressed p21W A F ,/C ip l in only 2% of
cells. Using pair-wise comparisons, a statistically significant difference was observed
between p2 iW A F I/ap l immunostaining in DCIS nuclei and benign stromal (p = 0.002) or
epithelial cell nuclei (p = 0.005) (Figure 2-3, A, B and C). The findings for invasive
breast carcinoma showed the same trend but did not achieve statistical significance (p =
0.12). When the cases with marked reactive desmoplasia were excluded from statistical
analysis, a significant difference in immunostaining was noted between invasive
carcinoma and stromal elements (p = 0.019) similar to the difference noted in breast
DCIS. Extensive reactive fibrosis was found only in invasive breast cancers not in
breast DCIS.
In contrast to the breast cancers, the pattern of staining in ovarian cancer cases
showed higher p21W A F I/C ip l immunostaining in the benign tissue as compared to invasive
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
tumor tissue. Eighty percent of the ovarian cancer cases showed p21W A F ,/c,p I
immunostaining in either the carcinoma cell nuclei, the benign stromal cell nuclei or
both. There was a higher degree of p2 l W A F I/c ,p l expression in benign stromal nuclei
compared with tumor nuclei. On average, 13% of benign stromal cells expressed
p2 jW A F i/c ip i w|jjje oniy 7% 0f tumor nuclei immunostained positively for p21W A F 1 ,G p !. A
pair-wise comparison of tumor - benign differences demonstrated a higher percentage of
benign nuclei than carcinomatous nuclei exhibiting p 2 1 WAF,/CiPI immunostaining (p =
0.016) (Figure 2-3, D and E). Similar to the pattern of immunostaining in the cases of
invasive breast carcinomas with markedly reactive stromal fibrosis, a higher degree of
p2 iW A F !/c iP i eXpressjon was observed in the fibrous stroma surrounding the tumor.
In contrast to the other tumor types, pair-wise comparison of p2 lW A F I/C ip l
immunostaining in endometrial carcinomas revealed no differences in expression
between carcinoma cell nuclei and benign stromal nuclei (p = 0.99). However, when
p2 } W A F i/c .p i immunostaining was compared between cases with the codon 31Arg and
cases with the more common codon 31Ser, a higher percent of p2 lW A F 1 /c ,p l
immunostaining was observed with codon 31Arg (p = 0.056) (Figure 2-3, G and H).
The same comparison in the other tumor types revealed no significant associations.
D is c u s s io n
p21W A F I/c ,p I has been described as a mediator of p53 dependent tumor growth
suppression and as an inhibitor of G1 cyclin dependent kinases (CDKs) (11-13).
p21W A F I/c ,p l is induced by p53 expression and binds to the cyclin E-CDK2-PCNA
complex, inhibiting phosphylation of Rb by the complex and halting the cell cycle (14,
15). p21W A R /c‘ p l also plays a role in other cyclin-CDK-PCNA complexes (16).
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Independent of p53, p2 iW A F 1 /C iP ‘ expression is observed in cells undergoing terminal
differentiation in embryonic mice (17). In adult mouse tissues with or without
functional p53, p2 lW A F I/C ip I is expressed in the brain, fully differentiated columnar
epithelium of the gut, as well as the lungs, heart, and skeletal muscle (17). Thus,
p21w afi/ciP i appear to function as an inducible growth inhibitor in embryonic
differentiation as well as in the adult, and the role of p53 appears to be important only in
the adult at the G1 checkpoint.
Recent work has shown that the p2 lW A F l/C ip l protein functions both as a positive
and negative regulator of the CDKs through its binding in the quaternary complex (6).
A single p2 l W A F I/C iP ‘ molecule activates the complex, while multiple copies of
p21W A F I/C ip l inhibit the function of the complex. Zhang (6) speculates that the
association of p 2 lW A F 1 /c,P 1 with active kinase implies that there is another function for
p2 iw A F i/ciP i jn ^ cejj C yC je ^ a CDK assembly factor.
The second exon of WAFl/Cipl contains 90% of the open reading frame. It is
75% identical and 79% similar at the amino acid level between mouse and human. This
exon contains a putative zinc finger motif between amino acids 13 and 41 as well as a
potential nuclear localization signal between amino acids 140 and 163 (3). The
sequences between amino acids 13 and 56 are almost perfectly conserved between
mouse and human, and there is a strong homology between p2 l W A F I/C lP ' and p27K ip l
(18, 19) as well as p57K ip 2(20, 21) both recently described family members of the CDK
inhibitors. This conservation of the amino acid sequence suggests that this region is
important to the function of p21W A F 1 /C ip l.
Codon 31 is in the highly conserved putative zinc finger region. The amino acid
change from serine to arginine described here and by others (22, 23) is not a
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2-4. Pair-wise differences in p21W A F V C lp I immunostained cells of carcinoma nuclei
and of benign nuclei on a case-by-case basis.
Each data point is the value of the percent of p2l W A F I /C ip l immunostaining in benign nuclei
subtracted from the percent of nuclei immunostained in the carcinoma nuclei. The bold
face line indicates the mean value of data points for each tumor type. (A.) Comparison of
the values generated in two subgroups of invasive breast cancer cases (IBC) and two
subgroups of breast carcinoma in situ (CIS). The invasive carcinoma groups show
differences between invasive carcinoma cells and benign stroma for either all cases or cases
with limited reactive desmoplasia. The breast carcinoma in situ differences are for either
benign ductal epithelium or for benign stromal cells and DCIS immunostaining
percentages. (B.) Differences between percentage of p21 w a fi> c * p 1 containing nuclei
observed in benign stromal cells and carcinomas for both ovarian and endometrial tumors.
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
o f th e copyrightowner. Further reproduction prohibited without permission.
7 3
CD
■ o
o
C l
C
o
CD
Q.
" O
CD
CO
C/5
o'
o
WAF1 Immunostaining in Breast Tissue
Differences Between Tumor and Normal
50
40
£ 30
0
!§
.£ o
I * o c
§8
E «
(A
I I
e e
a o
c E
o il2
c c
© ©
& £
20
10
-10
-20
-30
-40
-50
a
A
AA
A
AA
A
A
A
A
A
&
A
AA
i t
A
A
A A
A
A
AA
IBC «• iftC »t
SUam* Sttem e |w jo ** Cl* y»
reactive 4eim ep< e»u| M •P'lnww"* tu o m e
0\
V O
Difference i n Percent Immunostained Cells
(Percent Tumor minus Percent Normal)
WAF1 Immunostaining in Ovarian and Endometrial Tissues
Differences Between Tumor and Normal
60
50 A
30
20
10
-10
-20
-30 1
-40
-50
-60
-70
-80
f t IlltllVt
Knoem«tit«l Tvmei
r t them*
conservative change. The serine at codon 31 is conserved between human p57K ip 2 and
p 2 | W AFI/Cipl
Comparison between the homologous region in p27Ic ip l and p2lW A F I/ap l reveal that while
this codon is not conserved, it is a neutral amino acid, threonine, that is substituted for
the serine. The alteration at codon 31 exchanges a serine, an uncharged polar amino
acid with a single hydroxyl sidechain, for an arginine, which is a basic, positively
charged amino acid with a seven-membered side chain. These observations suggest that
this change may create a phenotypic variant of the p2lW A F ,/C ip l protein (24-26). Recently
Chedid, et al (23) used a tumor suppression assay which revealed no functional
difference between the different proteins to show that the codon 31 alteration is a
polymorphism. However, their assay used MMTV promoter-driven constructs to
produce the two proteins. This assay would not reproduce in vivo protein levels as
neither the cells nor the experimental conditions can be considered equivalent to the
normal cellular environment (27).
The observed pattern of p 2 lW A F 1 /c ,p l expression in this study was different in
each tumor type. In the two cohorts of breast cancer tissue, there was a similar pattern
of p2iW A F I/c ,p l staining. Both in situ and invasive breast carcinoma cells showed an
increase in p 2 iW A F I/c ,p I staining in pair-wise comparisons of tumor and normal cell
immunostaining (Figure 2-4). Invasive breast carcinomas showed a trend (p = 0.12)
toward a higher percentage of p21W A F 1 /c ,p l staining in the tumor cells as compared to the
normal tissue. Exclusion of cases with reactive desmoplasia revealed a statistically
significant increase in staining in the invasive carcinoma as compared to benign stroma
(p = 0.019). Similarly, breast DCIS showed a statistically significant association
between high staining in tumor nuclei and lower expression in normal ductal epithelium
(p = 0.005) and benign stromal nuclei (p = 0.002). The pair-wise comparisons were
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
determined by subtracting the percentage of positive nuclei in benign tissue from the
percentage of immunopositive nuclei in tumor cells (Figure 2-4). Both invasive breast
carcinoma and breast DCIS had mean values of the differences which were above zero,
indicating greater staining in the tumor nuclei. Ovarian tissue exhibited the opposite
result. The mean of the pairwise differences was less than zero, indicating higher
p2 jW A Fi/opi immunostaining in the stromal elements. The mean of the pairwise
differences in endometrial tissues was -0.05; thus, there was essentially no difference in
staining between the endometrial carcinoma and stroma. Interestingly, we noted an
increasing percentage of nuclei expressing p21W A F ,/C ip l in breast cancer overall. While
33% of samples had p21W A F ,/C ip l expression in benign nuclei, 56% of samples had
immunostained nuclei in DCIS and 83% of samples showed evidence of p 2 lW A F 1 /c ,p I
expression in the nuclei in invasive disease. Conversely, a pair-wise comparison of
p2 j w A F i/cip i staining in ovarian carcinomas showed the opposite association (p = 0.016)
with a higher percentage of benign cell nuclei stained compared with carcinoma cells.
We found no significant differences overall between staining in either tumor or normal
tissue nuclei in endometrial carcinoma. However, a pair-wise comparison of
immunostaining in the presence of codon 31 Arg with immunostaining in the presence of
codon 3 ISer demonstrated increased immunostaining associated with codon 31 Arg (p =
0.056). No differences were observed between immunostaining in the presence of
31 Arg and immunostaining in the presence of 3 ISer in breast and ovarian cancers.
p53 expression was evaluated by immunohistochemistry and p53 mutations
were evaluated by SSCP and DNA sequencing in our laboratory for the endometrial
carcinoma cases (Saffari, et al., in press), ovarian carcinoma cases (Wen, in press)and
breast carcinoma in situ cases (Lukas, et al. in press). The Mantel-Haenszel test was
used to evaluate potential associations between the presence of p53 mutation and
71
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
WAFl/Cipl alteration, stratifying for the type of tumor. The scatterplots and F-test
associated with the partial correlation coefficient were used to evaluate the relationship
between P53 and p 2 lW A F ,/C ip l immunohistochemical results. There was no association
between WAFl/Cipl alteration and p53 mutation, nor was there any statistically
significant association between p2lW A F I/C ip l expression and P53 expression or mutations
in breast, ovarian or endometrial carcinomas. However, an inverse trend was noted in
immunostaining in both endometrium and in breast DCIS. In these tissues, scatterplots
suggested that high p21W A F 1 /C ip l immunostaining corresponds with low P53 staining.
High P53 immunostaining similarly tended to correlate with low p 2 lW A F I/c ,p l staining
but these results were not significant. Thus, in these carcinomas, p 2 lW A F 1 /c ,p I
expression did not appear to be solely controlled by p53 expression.
Expression of p 2 iW A F I/G p l in fibrous stroma of tumors is not expected as part of
a DNA-damage repair pathway since the DNA content of the connective tissue is
expected to be normal. Inflammatory fibrosis, frequently noted in invasive breast
carcinomas, is thought to be in great part caused by release of transforming growth
factor 6 (TGF-6). TGF-B is a multifunctional polypeptide which has an array of cellular
effects. It is reported to inhibit growth of normal breast epithelium in vivo, but it
stimulates fibroblast chemotaxis and fibrogenesis. The effects of TGF-B on cells are
variable. TGF-B was very recently found to upregulate p 2 lW A F 1 /c ,p l expression through
a TGF-B response element near the transcription initiation site in WAFl/Cip 1. This site
is physically and functionally separate from the p53 consensus sequence (28). It is
possible that TGF-B expression mediates the ingrowth of fibroblasts with concurrent
upregulation of p 2 lW A F 1 /c ,p I in the stroma in ovarian carcinomas and in invasive breast
cancers.
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In conclusion, WAFl/Cipl showed a non-conservative serine-to-arginine
substitution at codon 31 that was identified in approximately 18% of individuals and is
interpreted as a polymorphism. p2lW A F ,/C ip l expression was variable in different cancers
both with regard to expression in carcinoma cells and with regard to expression in
benign cells in the tumor. The differences between p 2 lW A F ,/C ip I expression in tumor
cells and benign stromal cells also varied in a fashion that appeared to be characteristic of
particular tumor types.
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R e f e r e n c e s
1. Xiong Y, Hannon GJ, Zhang H, Casso D, Kobayashi R, Beach D: p2 i W A F 1 /C ip l is a
universal inhibitor of cyclin kinases. Nature 1993, 366: 701-704
2. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ: The p21 Cdk-
interacting protein Cipl is a potent inhibitor of G1 cyclin-dependent kinases. Cell
1993, 75: 805-816
3. El-Diery WS, Tokino T, Velculescu VE, Levy DB, Parsons R, Trent JM, Lin D,
Mercer WE, Kinzler KW, Vogelstein B: WAF1, a potential mediator of p53
tumor suppression. Cell 1993, 75: 817-825
4. Nakanashi M, Robetorye RS, Adami GR, Pereira OM, Smith JR: Identification of
the active region of the DNA synthesis inhibitory gene p2 i W A F 1 /C IP 1 /S d l1 . EMBO J
1995, 14:555-563
5. El-Diery WS, Harper JW, O’Connor PM, Velculescu VE, Canman CE, Jackman J,
Pietenpol JA, Burrell M, Hill DE, Wang Y, Wiman KG, Mercer WE, Kastan
MB, Kohn KW, Elledge SJ, Kinzler KW, Vogelstein B: WAF1 is induced in
p53-mediated G1 arrest and apoptosis. Cancer Res 1994,54: 1169-1174
6. Zhang, H, Hannon GJ, Beach D: p21-containing cyclin kinases exist in both active
and inactive states. Genes & Development 1994, 8: 1750-1758
7. Michieli P, Chedid M, Lin D, Pierce JH, Mercer WE, Givol D: Induction of WAF1
by a p53 independent pathway. Cancer Res 1994, 54: 3391-3395
8. Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T: Detection of
polymorphisms of human DNA by gel electrophoresis as single-strand
conformation polymorphisms. Proc. Natl. Acad. Sci USA 1989, 86: 2766-2770
9. Michaud J, Brody LC, Steel G, Fontaine G, Martin LS, Valle D, Mitchell G:
Strand-separating conformational polymorphism analysis: efficacy of detection
of point mutations in the human ornithine 5-aminotransferase gene. Genomics
1992, 13: 389-394.
10. Press MF, Nousek-Goebl NA, King WJ, Herbst AL, Greene GL:
Immunohistochemical assessment of estrogen receptor distribution in the human
endometrium throughout the menstrual cycle. Lab Invest 1984, 51:495-502
11. Xiong Y, Zhang H, Beach D: D type cyclins associate with multiple protein
kinases and the DNA replication and repair factor PCNA. Cell 1992, 71: 505-
514
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
12. Zhang H, Xiong Y, Beach D: Proliferating cell nuclear antigen and p2iW A F 1 /C ip l are
components of multiple cell cycles kinase complexes. Mol Bio Cell 1993, 4:
897-906
13. Waga S, Hannon GJ, Beach D, Stillman B: The p2lW A F l/C ipI inhibitor of cyclin-
dependent kinases controls DNA replication by interaction with PCNA. Nature
1994, 369: 574-578
14. Li Y, Jenkins CW, Nichols MA, Xiong Y: Cell cycle expression and p53
W AF1/CIP1
regulation of the cyclin-dependent kinase inhibitor p21 Oncogene
1994, 9: 2261-2268
15. Deng C, Zhang P, Harper JW, Elledge SJ, Leder P: Mice Lacking p2lW A FI/ap,
Undergo Normal Development, but Are Defective in G1 Checkpoint Control.
Cell 1995, 82: 675-684
16. Halevy O, Novitch BG, Spicer DB, Skapek S, Rhee J, Hannon GJ, Beach D,
Lassar DB: Correlation of terminal cell cycle arrest of skeletal muscle with
induction of p21W A F 1 /C IP I byMyoD. Science 1995,267: 1018-1021
17. Parker SB, Eichele G, Zhang P, Rawls A, Sands AT, Bradley A, Olson EN,
Harper JW, Elledge SJ: p53-independent expression of p 2 iW A F I/Q pI in muscle
and other terminally differentiating cells. Science 1995, 267: 1024-1027
18. Polyak K, Lee M-H, Erdjument-Bromage H, Koff A, Roberts JM, Tempst P,
Massague J: Cloning of p27K ipI, a cyclin-dependent kinase inhibitor and a
potential mediator of extracellular antimitogenic signals. Cell 1994,78: 59-66
19. Ponce-Casteneda MV, Lee MH, Latres E, Polyak K, Lacombe L, Montgomery K,
Mathew S, Krauter K, Sheinfeld J, Massague J, Cordon-Cardo C: p27K ipl:
chromosomal mapping to 12p 12-12p 13.1 and absence of mutations in human
tumors. Cancer Res 1995,55: 1211-1214
20. Lee M-H, Reynisdottir I, Massague J: Cloning of p57K IP 2 , a cyclin-dependent
kinase inhibitor with unique domain structure and tissue distribution. Genes &
Development 1995,9: 639-649
21. Matsuoka S, Edwards MC, Bai C, Parker S, Zhang P, Baldini A, Harper JW,
Elledge SJ: p57iaf> 2 , a structurally distinct member of the p2iW A F I/C ip I Cdk
inhibitor family, is a candidate tumor suppressor gene. Genes & Development
1995, 9: 650-662
22. Shiohara M, el-Diery WS, Wada M, Nakamaki T, Takeuchi S, Yang R, Chen D-
L, Vogelstein B, Koeffler HP: Absence of WAF1 mutations in a variety of
human malignancies. Blood 1994, 84: 3781-3784
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23. Chedid M, Michieli P, Lengel C, Huppi K, Givol D: A single nucleotide at codon
31 (Ser/Arg) defines a polymorphism in a highly conserved region of the p53-
inducible gene WAF1/CIP1. Oncogene 1994,9: 3021-3024
24. Lewin, B: Genes V. Oxford, Oxford University Press, 1994, p 150.
25. Taioli E, Crofts F, Trachman J, Demopoulos R, Tonioli P, Garte SJ: A specific
affican-american CYP1A1 polymorphism is associated with adenocarcinoma of
the lung. Cancer Res 1995,455: 472-473
26. Nazar-Stewart V, Motulsky AG, Eaton DL, White E, Homung SK, Leng Z-T,
Stapleton P, Weiss NS: The glutathione s-tranferase p polymorphism as a
marker for susceptibility to lung carcinoma. Cancer Res 1993,53:2313-2318
27. El-Deiry WS, Tokino T, Waldman T, Oliner JD, Velculescu V, Burrell M, Hill
DE, Healy E, Rees JL, Hamilton SR, Kinzler KW, Vogelstein B: Topological
W AFl/CIPl
control of p21 expression in normal and neoplastic tissues. Cancer Res
1995, 55: 2910-2919
28. Datto MB, Li Y, Panus JF, Howe DJ, Xiong Y, Wang X-F: Transforming
growth factor-B induces the cyclin-dependent kinase inhibitor p21 through a p53-
independent mechanism. Proc. Natl. Acad. Sci. USA 1995,92: 5545-5549
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 3: Alternative and aberrant splicing of
the mdm2 oncogene in invasive breast cancer
A b st r a c t
The mdm2 oncogene regulates the activity of a number of genes critical to
genomic integrity. mdm2 can inhibit p53 mediated transactivation as well as control the
rapid degradation of p53 protein. Mdm2 also inhibits retinoblastoma protein (Rb) and
enhances E2F1 and DPI protein activity; the overall effect is to promote transition to S
phase. The mdm2 oncogene was originally cloned from a BALB/c mouse cell line,
where it was amplified and overexpressed. In humans, this gene is amplified in soft
tissue sarcomas and gliomas. While mdm2 is rarely amplified and is not mutated in
breast cancer, recent reports in normal and malignant breast epithelium reveal the
presence of truncated proteins. Other work suggested that alternative mdm2 mRNA
splicing defines these truncated proteins. Overall, these alterations may effect the nature
of the interactions of mdm2 with p53, Rb and E2F, and thus may effect cell cycle
control. This study was undertaken to characterize altered mdm2 splicing in a cohort of
38 invasive breast cancers and to establish if there are correlations with breast cancer
prognostic markers. Using primers spanning the full length of the mdm2 cDNA,
polymerase chain reaction (PCR) amplification on cDNAs produced a full length 1526
base pair RT-PCR product as well as products of other sizes. Thirty seven (97%) had a
full length 1526 base RT-PCR product. Eleven of these 37 samples (30%) also had
evidence of smaller RT-PCR products measuring 653, 281, 254, and/or 219 bases in
length. One sample did not have a 1526 base RT-PCR product but had only evidence of
281 and 219 bp products. Nine normal breast samples were also examined. Five
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
normal breast tissues had only the 1526 base products and 4 had 1526 and 653 base
PCR products. Sequence analysis of these products determined that the 653 bp product
is apparently produced by alternative splicing (defined as splicing at intron/exon
boundary consensus sites), while the 281, 254, and 219 RT-PCR products are
produced by aberrant splicing (defined as splicing at sites without proper splicing
consensus sequences, generally within exons). Immunohistochemistry was performed
on these cases using antibodies to HER2/ne« (c-erbB2), estrogen receptor (ER),
progesterone receptor (PR), epidermal growth factor receptor (EGFR) and p53. DNA
ploidy was also measured. Considering both 1526 and 653 bp RT-PCR products as
normal, expression of abberantly sized mRNAs (i.e. -200-300 base products) correlated
significantly with p53 overexpression (p=0.018) but only marginally with p53
mutations (p=0.071). There was also a statistically significant correlation between
expression of abberantly spliced products and the lack of progesterone receptor
(p=0.030). If only the 1526 bp RT-PCR product is considered normal, then expression
of alternative and aberrant products show a much stronger correlation with p53
mutations (p<0.001), a similar correlation with p53 overexpression (p=0.023) and a
similar correlation with lack of PR expression. These findings demonstrate the pattern
of mdm2 mRNA splicing in normal breast and show altered splicing of mdm2 in
invasive breast carcinomas an association between mdm2 altered splice products and
established markers of poor prognosis suggesting potential clinical utility. Furthermore,
these findings show that overexpression of p53 protein in the absence of p53 mutations
may be in part due to alterations in the mdm2 mRNA.
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
I n t r o d u c t io n
The p53 gene appears to be important in the pathogenesis of many types of
cancers include some invasive breast cancers. Initially, it was thought that
overexpression of p53 protein was caused by mutations in the p53 gene; however, since
a large review (1) showed that only 22% of tumors had a mutation in the DNA
sequence, this contention is the subject of some debate (2-4). p53 is a transactivator of
several genes such as WAFl/Cipl (5) and gaddAS (6) which are involved in the arrest of
the cell cycle at the G1 boundary and DNA damage repair, respectively. p53 regulates
its own expression through the mdm2 (7, 8) gene. These data suggest the possibility
that a gene downstream of p53 such as mdm2, if inactivated, might cause the
compensatory overproduction of p53 protein in the absence of mutations in the gene.
mdm2 is up-regulated by p53 and functions to downregulate p53 activity in two
ways: 1.) inhibiting p53 function by binding to the transactivation site of p53 (8) and
2.) promoting the rapid p53 degradation through a proteosome dependent mechanism
(9, 10). mdm2 is amplified in approximately 30% of soft tissue sarcomas (11, 12). In
human breast cancers, mdm2 amplification is a rare event, and no mutations have been
described (13). However, transgenic mice overexpressing mdm2 protein but lacking
functional p53 had deranged growth patterns in the breast ducts and lobules during
pregnancy or lactation (14). Ten to twenty percent of the mice developed mammary
carcinomas after long latency. Evidence of different sizes of mRNA transcripts was
noted when mdm2 was initially cloned (12) and more recently, altered mRNAs,
postulated to be the product of alternative splicing, were described in ovarian carcinomas
(15). These alternatively spliced mdm2 mRNAs were correlated with poor outcome.
Different mdm2 proteins have been noted in normal mammary epithelial cells and breast
cancer cell lines (16). In this study, mdm2 mRNA alterations were analyzed in a cohort
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of 38 invasive breast cancers and compared with clinical parameters in order to
understand what associations the alterations may have with established markers of poor
prognosis in breast cancer.
M a ter ia ls a n d M ethods
T issues
The use of human tissue in this study was reviewed and approved by the
University of Southern California Institutional Research Committee. Thirty eight frozen
invasive ductal breast carcinomas and nine cases of normal breast tissue, obtained from
storage at -186°C in a liquid nitrogen freezer by the USC Breast Tumor and Tissue
Bank, were used for these investigations. Frozen tissue sections stained with
hematoxylin and eosin were used to confirm the histological composition of the
specimens.
Total RNA isolation
Total RNA was extracted from 10-20 serially cut, 10 pm thick, frozen tissue
sections using TRIzol (Gibco/BRL, Gaithersburg, MD) according to the manufacturer’s
instructions. The RNA was precipitated using isopropanol, washed once with 70%
ethanol, and allowed to dry at room temperature. The RNA was dissolved in 30 pi of
RNAse free HjO and the optical density measured. The RNA was then checked for
integrity using northern blotting.
Reverse transcription
Mouse mammary leukemia virus reverse transcriptase (MMLV-RT, Gibco/BRL)
was used for reverse transcription to produce cDNAs according the following protocol.
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Two micrograms of total RNA was used in each reaction to ensure a representation of all
mRNAs. Total RNA was mixed with 0.5 pg of oligo dT,5 primer, brought up to a total
of 7 pi using RNAse free H2 0 , heated to 70°C for 10 minutes and afterward
immediately iced for 5 minutes. The RNA and oligo dTls (Promega, Madison, WI) was
added to 13.5 pi reverse transcriptase solution (40 units RNAguard (Pharmacia,
Piscataway, NJ), IX First Strand buffer (Gibco/BRL), 1.0 mM dNTPs (Promega),
0.125 mM MgCI2 (Promega), and incubated at 39°C for at least 1 hour. After cDNA
synthesis, the samples were stored frozen at -80°C.
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mdm2 primer design and polymerase chain reaction (PCR)
A nested PCR protocol was used to amplify the full length mdm2 cDNA as
described by Sigalas (15) with modifications detailed below. In brief, the full open
reading frame of mdm2 (1526 bases) was amplified using the following nested primer
sets. External primer pair Sense 5’-CTGGGGAGTCITGAGGGACC-3’ and
Antisense 5’-CAGGTTGTCTAAATTCCTAG-3\ Internal primer pair sense 5’-
CGCGAAAACCCCCGGGC AGGC AAATGTGC A-3 ’ and antisense 5’-
CTCTTATAGACAGGTCAACTAG-3’. PCR amplification was performed with 25 pi
reactions, using 1.6 mM MgCl2 , 40 pM of each primer, 1 mM dNTPs, 5 units of Taq
polymerase (Promega), and 100 ng of cDNA. Thirty cycles of 94°C (1 minute), 58°C
(1 minute), and 72°C (2 minutes) were performed in a Perkin Elmer 480 using mineral
oil overlay. After the external primers were used, a 2pl aliquot was transferred directly
to a new reaction tube for PCR using the internal primers. Reaction temperatures were
the same for both primer sets. The products of both reactions were run through 1.5%
agarose gels (SeaKem LE, FMC Bioproducts, Rockland, ME) with ethidium bromide
and visualized with ultraviolet transillumination.
PCR was also utilized to assess the mdm2 gene for large genomic deletions.
The primers spanned a 1569 bp regions stretching from the start of exon 5 into the 3’
untranslated region. Sense 5’-TTCTTTTTTATCTTGGCCAG-3\ Antisense 5’-
TCTC ATTT AAG AC AGAGT AG-3 ’.
Analysis of mutations in p53
In all 38 cases p53 was analyzed using the Mismatch Detect kit in conjunction
with the p53 cDNA screening module (Ambion, Austin, TX) according to the
manufacturer’s instructions. Briefly, cDNAs were amplified using a nested PCR
82
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
protocol, using external and internal primers specific to p53. These primers span the
evolutionarily conserved regions of p53 from codon 91 to codon 368. PCR
amplification was carried out using a 3 temperature protocol, utilizing 30 seconds at
94°C to denature, 30 seconds at 55°C to annealing primers, and 40 seconds at 72°C for
extension. Samples were amplified for thirty cycles, followed by 5 minutes at 72°C.
The internal primers were designed with a T7 RNA polymerase promotor on the 5’ end
of the sense promoter and a SP6 RNA polymerase promotor in the 5’ end of the
antisense primer. The appropriate RNA polymerase was added to the internal primer
PCR amplification product, and subsequently the RNA from either promoter was
hybridized to a complementary wild type p53 RNA. After hybridization, the samples
were digested using various RNAses, which cleave the RNA at sites of mismatched
base pairs. The digested samples were analyzed on 1.5% agarose gels (FMC).
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
mdm2 product cloning and DNA sequencing
Each case with products other than the full length (1526 bp) and 653 bp sizes
were either cloned using the TA cloning kit (Invitrogen, Carlsbad, CA) and sequenced,
or directly sequenced from the PCR products. For cloning and sequencing, cDNAs
were PCR amplified from the mdm2 internal primers to generate sufficient product. The
product was then purified with Qiagen PCR purification columns (Qiagen, Chatsworth,
CA) and ligated into the pCR 2.1 vector. After transformation, the plasmid was isolated
(Wizard plasmid minipreps, Promega) and sequenced using the ThermoSequenase kit
(Amersham, Arlington Heights, IL), according to the manufacturer’s instructions.
Approximately 250 ng of plasmid was used for sequencing. Alternatively, products
were gel purified using the Qiaquick Gel Extraction Kit (Qiagen) and directly sequenced
using the ThermoSequenase Kit.
Immunohistochemistrv
The peroxidase anti-peroxidase technique was used to identify various protein
products in tissue sections. Frozen sections, 4pm thick, were cut and fixed
appropriately and treated with primary antibody for 1 hour at room temperature.
Secondary and tertiary antibodies were applied for one half hour at room temperature.
Each slide was stained with ethyl green, a nuclear counterstain.
p5 3
p53 protein was localized with the anti-human p53 mouse monoclonal antibody
DO-7 (Dako Corporation, Carpinteria, CA) at a 1:100 dilution in 10% normal rabbit
serum in 31 cases. Nuclear staining was considered positive. In frozen tissue sections
immunostaining is observed in nuclei of a low percentage of normal, proliferatively
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
active tissues. Therefore, we have used 10% immunostained tumor cell nuclei as a
value for separation of “normal expression” from “overexpression” (17). Those breast
tissues with p53 immunostaining in at least 10% of the cell nuclei were considered to
have p53 overexpression, while those with less than 10% p53 immunostained nuclei
were considered to be within the normal range of p53 expression.
H ER2/neu
Immunostaining for the YSEKUneu oncoprotein was performed as described
(18) in 38 cases. In brief, rabbit anti-HER2/ne« was used at a 1:2000 dilution in 10%
normal goat serum. HER2/new expression was categorized as previously described
(18). Only membrane staining was considered positive.
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Epidermal Growth Factor receptor (EGFR)
Thirty four cases were immunostained for EGFR was performed using an
antibody from Amersham (RPN 513) and a 1:25 dilution in 10% normal rabbit serum.
Tissues were fixed in acetone for 15 minutes. Since membrane staining is common in
normal epithelial cells, only membrane staining in the carcinoma was considered
positive.
Estrogen and Progesterone receptors
Thirty eight cases were immunostained for estrogen and progesterone receptors
was performed as described previously (19) with the exception that the fixative used
was picric acid/paraformaldhyde. Immunostaining was described as a percentage of cell
nuclei with nuclear staining. For statistical purposes, all nuclear immunostaining was
considered positive.
DNA ploidy analysis
DNA ploidy analysis was carried out on all thirty eight cases using the CAS 200
workstation (Cell Analysis Systems) and DNA staining was performed according to the
manufacturer’s protocol (Cell Analysis Systems, Elmhurst, D L ) as describe
elsewhere(20). In brief, the frozen tissue sample is touched briefly to a prewarmed
slide, fixed with 10% formalin and stained with a modified Fuelgen stain.
Statistical Analyses
The association between mdm2 expression and the clinical parameters was
evaluated using Fisher's exact test and, where necessary, the Mantel Haenzel Chi-
Square test.
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R e su l t s
Altered mdm2 mRNAs in tumor tissues
PCR analysis of the reverse transcribed mRNA revealed the presence of the
expected full length (1526 base) product as well as other, smaller, products (Figure 3-
1). Thirty seven (97%) had a full length 1526 base RT-PCR product. Eleven of 37
samples (30%) also had evidence of smaller RT-PCR products which measured 653,
281, 254, and/or 219 bases in length. One sample did not have a 1526 base RT-PCR
product but had only evidence of 281 and 219 bp products. Six (16%) samples had
1526 and 653 base-pair products, 4 samples (10.5%) had 1526 and 219 bp products,
and one sample (2.5%) had 1526, 281 and 219 base products. Only one case had
smaller RT-PCR products, 281 and 219 bp in length, but lacked the normal 1526 bp
product.
The normal ductal or lobular epithelium from nine normal breast samples
(derived from reduction mammoplasties) were analyzed to clarify normal expression of
mdm2 in breast tissue. Five samples had 1526 base-pairs products alone and 4
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3-1. Mdm2 alterations in invasive breast cancers and normal breast.
Shown is an ethidium bromide stained 1% agarose gel with RT-PCR products
derived from normal breast epithelium and invasive breast cancers. Samples 1-18 are
invasive breast cancer; samples labeled N1-N4 are normal breast tissue. “L” indicates
the lanes loaded with the 1 kb DNA ladder (Gibco/BRL).
sample L I 2 3 4 5 6 7 8 9 10 1 1 12 1 3 14 15 16 1 7 1 8 L N1N2N3W
1.6 kb-
1.0 kb-
0.5 kb-
03 kb-
-1.6 kb
1.0 kb
-0.5 kb
-03 kb
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
samples had 1526 and 653 bp RT-PCR products. Compared to invasive breast
carcinomas, the 653 base product in normal breast epithelium was expressed at a low
level (Figure 3-1, lanes N1 and N4).
Sequence analysis was carried out to describe the exact nature of the altered
products. The 653 base fragment is an alternatively spliced fragment of mdm2,
consisting of exons 3 (exons 1 and 2 are noncoding) spliced to exon 12. Splicing takes
place at the intron/exon boundary splice sites as described in the mouse (21) (the human
genomic structure has not been characterized), and the truncated mRNA is in-frame.
This fragment was missing a large central portion of the mRNA (Figure 3-2, mdm2-653
bp), including 90% (81 of 90) amino acids of the 3’ end of the p53 binding domain, and
the entirety of the nuclear localization signal and acidic domain. The zinc and RING
finger domains remained intact. This product was identical to the 707 bp product
described previously called mdm2-B (15). (The only difference between the 653 bp
product described here and the 707 bp fragment characterized by Sigalas (15) is the
numbering. The 653 bp counted are only those between the start and stop codon, not
including the PCR primers. The 707 bp count included the PCR primers, which are
outside of the start and stop codons.)
The other mdm2 RT-PCR products, which consisted of two 281 bp products,
one 254 bp product, and five 219 bp products did not have similarly defined splicing
rules. The 281 bp product consisted of 204 bases of 5’ mdm2 open reading frame
sequence including exons 3 and 4 and 50 bases of exon 5 spliced into the 3’ 76 bases of
the exon 12 open reading frame. The pattern of splicing in these smaller fragments was
interesting because the splice donor and acceptor sites were in regions of exact sequence
homology. For instance, in a 281 base fragment, the 3’ end of exon 5 ends with a six
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
base sequence of AGAAGC and the same sequence is at the untranscribed end of the
downstream splice site in exon 12 (Figure 3-3, A).
This 6 base-pair sequence occurs at only 4 sites in mdm2. Thus the splice donor
sequence is 517-AGAAGC/aacaac-524 and 1708-agaagc/TAAAGA-1719. (Underlined
sequence indicates exact sequence match. Uppercase letters denote coding bases,
whereas lowercase letters indicate that these base are excised from the final mRNA.)
The 281 base-pair product was missing 43% of the p53 binding domain, the entire
nuclear localization signal, acidic domain, and zinc finger, as well as over half of the
RING finger domain (Figure 3-2, mdm2-281bp).
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3-2: mdm2 RT-PCR products in relation to a full length mdm2 cDNA.
The mdm2 full length cDNA with various protein domain motifs indicated is compared
with the exon structure of the mouse mdm2 gene (21) (mouse exon structure), which is
80.3% identical (12) to the human mdm2, particularly after the start site at base 312. Only
the RT-PCR products found in this study are represented. Superimposed on each RT-PCR
product schematic is the motif that remains in the truncated product. Dashed lines have
been dropped from the mouse exon structure as well as the human domain structure to
simply comparisons. The green domain is the putative p53 binding domain. The nuclear
localization signal is in black, while the acidic domain, which binds the L5 ribosomal
protein, is crosshatched. The DNA binding Zn finger domain is marked in red. The
RING finger, a putative RNA binding site, is marked with horizontal hatching. The yellow
box, in the RT-PCR product labelled mdm2-254bp, represents a insertion of 14 bases of
DNA which has no known homology.
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 3-2: mdm2 splice variants in normal breast and breast cancer
nuclear localization signal
I p53 binding site ^ acidic domain zinc finger
mdm2 Full Length human cDNA
amino acid:
i
1 3 260 289 333 H i
RING
finger 49,
mouse exon structure
mdm2-653 bp
mdm2-254 bp
mdm2-219 bp
vo
to
Figure 3-3: A comparison of representative segments of the abberantly spliced mdm2
cDNAs.
A and C) A diagram of two representative sequences that have high similarity and appear
to be hotspots for aberrant splicing. The CAPITAL letters refer to transcribed bases that
remain in the final mRNA products while the lower case letters refer to untranscribed
bases which are spliced out of the final mRNA product. The light blue box encloses
identical or highly similar sequences. The green box indicates the p53 binding site (see
Figure 2) and the horizontal hatching indicates the RING finger motif (Figure 2).
B) One clone with an insertion of a 14 base segment of DNA between the p53 binding
site (in green, see also Figure 2) and the RING finger domain (horizontal hatch). This
fragment does not have homology to known sequences in the NCBI databases, suggesting
that this is an untranscribed sequence of intron derived DNA.
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission.
Figure 3-3: Comparison of overlapping sequences in mdm2
A)mdm2-281 bp Q aacaac
I'AAACA
B) mdm2-254 bp
TGAAACTCCATCT
C) mdm2-219 bp
V O
i
The 254 base-pair product had a different splicing arrangement. In this single
case, the fragment contained the entirety of exons 3 and 4 and then was spliced into the
last 99 bases of exon 12. At the splice junction was a 14 base insertion,
GTGAAACTCCATCT, which had no known homology. The predicted final protein
product would be out of the original reading frame (Figure 3-2, mdm2-254 bp).
The 219 base-pair product corresponded to the sequence of exons 3 and 4 and
the first 16 bases of exon 5 spliced into the last 3’ 48 bases of exon 12. The splice site
varies around a 10 base site, TGGCCAGTAT. in the 5th exon and a 10 base site,
TGCCCAGTAT. in exon 12. Thus, the 5’ splice site is in a 10 base sequence
TCTTG/tggccagtat is the splice donor and a gccctg/TGCCCAGTAT the splice acceptor
at the 3’ site. This sequence occurs at only the 2 sites in mdm2 where aberrant splicing
occurs. In each of the 5 cases with the product used this region for splicing as well as
the 2 cases with 281 bp RT-PCR product. Thus, while the exact splice junctions were
either different from each otheror unknown, this landmark was in close proximity to all
splices (Figure 3-3, C).
In each of these smaller fragments, the majority of the mdm2 mRNA has been
excised. In general, over half of the p53 binding site was removed, as well as the
nuclear localization signal, the acidic domain, and a portion of the RING finger.
mdm2 mRNAs in benign tissues
PCR analysis of the reverse transcribed mRNAs from 9 sample of normal breast
tissue revealed the presence of the expected full length (1526 bp) RT-PCR product in
every case as well as a frill length and a 653 bp RT-PCR product (Figure 3-1, N1 and
N4) in 4 of 9 cases. The 653 bp product had a weaker intensity in normal epithelium
compared to breast carcinomas (Figure 3-1, lane 13 compared to lane Nl).
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PCR analysis of mdm2 genomic sequence
The truncated mdm2 mRNAs might possibly be caused by genomic deletions in
the mdm2 gene. In studies of yeast (22), a mutated DNA polymerase 5 (pol3-t) created
a mutator phenotype which deleted segments of DNA between direct repeats 3-7 bp
long, mimicking the results found in the mdm2 mRNAs. To investigate the possibility
that genomic deletions in mdm2 might give rise to trucated mRNAs, PCR was used to
amplify the genomic DNA across the direct repeats where the deletion takes place. No
deletions were noted in any of the 38 samples of invasive breast cancer (data not
shown).
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Correlations of altered mdm2 mRNAs with prognostic markers and
clinical outcome
The presence or absence of alternate and aberrant splicing products of the mdm2
gene was compared with HER2/«e« (c-erbB2), estrogen receptor (ER), progesterone
receptor (PR), epidermal growth factor receptor (EGFr) and p53 status. If only the
1526 full length base-pair product is considered normal, and the 653, 281,254, and 219
base fragments are considered abnormal, expression of abnormal mdm2 fragments was
correlated with low PR (p=0.036), p53 immunopositivity (p=0.023) and p53 mutations
(p<0.001). Low expression of ER was marginally significant (p=0.087). There was
no correlation with EGFR, HER2/ne« expression and DNA ploidy status.
When expression of the mdm2 653 bp product alternative splice product was
considered to be normal, then the correlations with prognostic markers changed. There
was still an association with low PR expression (p=0.030), and p53 immunopositivity
(p=0.018), but the correlations with low ER expression as well as p53 mutations was
not significant (p=0.395 and p=0.071, respectively).
Comparison of the presense or absense of altered mdm2 mRNAs to clinical
outcome showed a correlation between abnormal RT-PCR products and outcome.
Assessing the 653 bp RT-PCR products and RT-PCR products less than 300 bases as
abnormal, no correlation between outcome and the presense of absense of the altered
mdm2 mRNAs was noted. However, when only RT-PCR products less than 300 bp
(281, 254, 219 bp) were considered abnormal, a statistically significant correlation was
seen between expression of these RT-PCR products and poor clinical outcome
(p=0.0036) (Figure 3-4).
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 3-4: Logrank survival curves for breast cancer patients
Abnormal mdm2 RT-PCR products (281, 254 and 219 bp products) (dotted line) had a
lower overall survival time than did those patients with a normal (1526 and 653bp) mdm2
RT-PCR product (solid line).
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m
- p
o
D
• O
o
u
a
f i
n s
a
,o
m
in
ID
to
■ D
O
3
T >
O
L i
a
E -
p s
a
.a
CM
x >
c
i n
CM
ID
C O
CO
CM
in
n
C M
o
o
o
in
c\
o
in
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T i m e ( Y e a r s )
D is c u s s io n
Alternative and aberrant splicing of mRNA presents an important issue. Which
splice products are considered normal and which are abnormal? Alternative splicing,
when it occurs, usually make use of inherent exon-intron splice sites of single mRNA to
produce different mRNAs. It is common to find one or a number of exons excluded
from the final mRNA to form an alternately spliced product. Alternative splicing is a
common event in a number of different genes and appears to be a process by which a
single gene can produce different proteins with different functions. For instance, the
APC (adenosis polyposis coli) gene, a tumor suppressor gene altered early in colon
carcinogenesis, is commonly expressed both with and without exon 1 in the brain, heart
and skeletal muscle in humans and mice, producing novel amino terminal proteins (23).
Induction of terminal differentiation causes changes in APC isoforms. Thus, in
response to tissue dependent signals as well as signals to differentiate, APC is
alternatively spliced and creates different proteins (23).
Aberrant splicing, in contrast, is the splicing of mRNA which is misdirected and
does not occur at de facto splice sites. Both the tumor susceptibility gene 101 (TSG101)
and the fragile histidine triad gene (FH1T) mRNAs show evidence of alternative and
aberrant splicing. TSG101, a putative tumor suppressor gene which was originally
thought to be mutated in breast cancers (24, 25), was instead found to be subject to
aberrant splicing mimicking the original breast tumor ‘mutations’ (26, 27). These
aberrations were predominantly truncating alterations at no discernible splice junction.
Similarly, the FHTT gene, another putative tumor suppressor gene, shows evidence of
both alternative and aberrant splicing in breast (28), lung (29), and head and neck
tumors (30). In lung cancer, the splicing is characterized both by losses of individual
100
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
exons, loss of exons with insertions of unknown DNA sequences, and losses of whole
portions of mRNA without regard for proper splice junctions. However, only deletion
of exon 8 was noted in normal lung and bronchial cells, arguing against this loss being
specific to lung cancer (29). In these genes, alternative splicing is a relatively common
event not necessarily associated with malignancy.
Two recent studies of variant mdm2 proteins in breast cancer are in agreement
that multiple mdm2 proteins are produced, but conflict on the size of proteins normally
and abnormally expressed. One study of matched benign breast epithelium and breast
tumor found that a 57kD protein was more highly expressed in neoplastic tissues, which
suggested that mdm2 is altered during progression (31). In contrast, another study
showed that low molecular weight proteins predominated in benign epithelium (16).
Evidence presented here shows that mdm2 expression in breast cancer is characterized
by truncated mdm2 mRNAs less than 300 bases long. These truncations are not the
result of deletions of the genomic DNA. These results are similar to a study in ovarian
cancer and normal ovary which showed alternative and aberrant mRNA splicing
products only in ovarian carcinomas (15). However, our results reveal expression of
the alternatively spliced 653 bp RT-PCR product in normal tissues, whereas in the study
by Sigalas, et. al. (15) this 653 bp product was found only in ovarian carcinomas. In
the present cohort of invasive breast cancer and normal epithelium, expression of the
653 bp product was expressed more highly in the tumor tissue compared to normal
tissue. The TSG101 mRNA exhibits a similarly higher expression in a single
alternatively spliced product that occurs in both normal and tumor tissues. Therefore,
the transforming ability of this alternatively spliced mdm2 product may depend on the
relative stoichiometry of the product. Low levels appears to properly regulate the cell,
while high levels disregulate the cell. This type of concentration dependence was
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
described previously in WAFl/Cipl (32). Furthermore, there seems to be a relaxation
of the splicing fidelity. Results presented here show that mdm2 is alternatively spliced
in normal breast epithelium and is both alternatively and abberantly spliced in invasive
breast cancers. Both the pattern of altered splicing and amounts of altered products was
different between normal and neoplastic tissues.
The pattern of deletions between distant short direct repeats was very similar to
data from a DNA polymerase delta (pol 3) in Saccharomyces cerevisiae. In those
studies, a mutation of pol 3 (termed pol 3-t) caused an approximately 1000 fold increase
in 7-61 deletions between the repeat sequences. Initially, it was thought that these
deletions might be caused by genomic deletions in the DNA sequence of mdm2.
However, PCR across the expected site of loss showed that there was no deletion (data
not shown). It seems possible that a faulty RNA polymerase may play an important role
in the generation of these deletions.
The presence or absence of mdm2 alternate splicing is correlated with clinically
useful information. mdm2 was considered normal when presence of only a single 1526
base-pair RT-PCR product or a 1526 base product and a 653 base product were found.
The presence of estrogen receptor (ER), progesterone receptor (PR), HER2Ineu
oncoprotein, epidermal growth factor receptor (EGFr) were assessed with
immunohistochemistry, and DNA ploidy was assessed with computerized image
analysis. The presence of abnormal splice variants was correlated with lack of PR
(p=0.030), p53 immunopositivity (p=0.0l8) and marginally with p53 mutation
(p=0.071). Aberrant mRNAs were not associated with ER (p=0.395), HER2Jneu
(p=0.827), EGFR (p=0.3) or DNA ploidy (p=0.643).
Statistical analysis of this cohort assuming all splice alterations (other than the
full length mdm2 transcript) were abnormal provided an interesting comparison. While
102
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the association or lack of association between altered mdm2 mRNAs and ER, PR,
EGFR, HER2//ICH and DNA ploidy remained similar, there was a drastic change in the
correlation with p53 . mdm2 alterations were significantly correlated with p53
mutations (p<0.001) in the current analysis, in contrast to the marginally significant p-
value (p=0.071) in the previous analysis. This difference was because 6 of the 7
samples with a 653 bp RT-PCR product also had a mutant p53 gene. The correlation
with p53 overexpression remained significant (p=0.023). These data indicated that p53
mutations correlated most strongly with the presence of the 653 bp alternatively spliced
RT-PCR product. Some conclusions can be made based on these observations. First,
as we hypothesized, altered mdm2 RT-PCR products are associated with p53
overexpression, indicating that p53 overexpression may be caused in part by expression
of altered mdm2. Alteration of mdm2 and subsequent changes in the expression of this
mediator of p53 appears to play a significant role in the process of carcinogenesis as
well as disease outcome. Second, mdm2 alternative splicing is highly correlated with
the presence of p53 mutations, which is suggestive of a mechanism by which mutational
inactivation of p53 causes the expression of a modified mdm2 protein. This mechanism
assumes that the p53 mutation occurs prior to the alteration of mdm2. This is, however,
an open question. In any case, this may be a natural response pathway in order that
mdm2 can operate in some capacities by not bind and inactivate p53.
In conclusion, evidence presented here shows that mdm2 is spliced alternatively
in normal breast epithelium and alternatively and abberandy in breast cancer. The
presence of the abberandy spliced products correlates strongly with prognostic markers
and indicates that the aberrant splicing of mdm2 may have clinical significance. The
correlation between p53 overexpression and mdm2 splice alterations suggests that
alterations of mdm2 may in fact cause upregulation of p53 protein.
103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
R e f e r e n c e s
1. Greenblatt, M., Bennett, W., Hollstein, M., and Harris, C. Mutations in the p53
tumor suppressor gene: clues to cancer etiology and molecular pathogenesis,
Cancer Res. 54: 4855-4878, 1994.
2. Barbareschi, M., Caffo, O., Doglioni, C., Fina, P., Marchetti, A., Buttitta, F.,
Leek, R., Moreili, L., Leonardi, E., Bevalacqua, G., Dalla Palma, P., and
Harris, A. p21W A FI immunohistochemical expression in breast carcinoma:
correlations with clinicopathological data, oestrogen receptor status, MIB1
expression, p53 gene and protein alteration and relapse-free survival, Brit J
Cancer. 74: 208-215, 1996.
3. Martinazzi, M., Crivelli, F., Zampatti, C., and Martinazzi, S. Relationship
between p53 expression and other prognostic factors in human breast carcinoma.
An immunohistochemical study., Am J Clin Pathol. 100: 213-217, 1993.
4. Hall, D., Meek, D., and Lane, D. p53: integrating the complexity, J Pathol. 180:
1-5, 1997.
5. El Diery, W. S., Tokino, T., Velculescu, V. E., Levy, D. B., Parsons, R.,
Trent, J. M., Lin, D., Mercer, W. E., and Kinzler, K. W. WAF1, a potential
mediator of p53 tumor suppression, Cell. 75: 817-825, 1993.
6 . Kastan, M„ Zhan, Q., El Diery, W., Carrier, F., Jacks, T., Walsh, W .,
Plunkett, B., Vogelstein, B., and Fomace, A. J. A mammalian cell cycle
checkpoint pathway utilizing p53 and GADD45 is defective in ataxia-
telangiectasia, Cell. 71: 587-597, 1992.
7. Momand, J., Zambetti, G., Olson, D., George, D., and Levine, A. The mdm-2
oncogene product forms a complex with the p53 protein and inhibits p53-
mediated transactivation, Cell. 69: 1237-1245, 1992.
8 . Oliner, J., Pietenpol, J., Thiangalingam, S., Gyuris, J., Kinzler, K., and
Vogelstein, B. Oncoprotein MDM2 conceals the activation domain of tumour
suppressor p53, Nature. 362: 857-860, 1993.
9. Kubbatat, M., Jones, S., and Vousden, K. Regulation of p53 stability by
mdm2„ Nature. 387: 299-303, 1997.
10. Haupt, Y., Maya, R., Kazaz, A., and Oren, M. Mdm2 promotes the rapid
degradation of p53, Nature. 397: 296-299, 1997.
11. Cordon-Cardo, C., Latres, E., Drobnjak, M., Oliva, M., Pollack, D., Woodruff,
J., Marechal, W., Chen, J., Brennan, M., and Levine, A. Molecular
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
abnormalities of mdm2 and p53 genes in adult soft tissue sarcomas, Cancer Res.
54:, 1994.
12. Oliner, J., Kinzler, K., Meltzer, P., George, D., and Vogelstein, B.
Amplification of a gene encoding a p53-associated protein in human sarcomas,
Nature. 358: 80-83, 1992.
13. Marchetti, A., Buttitta, F., Girlando, S., Dalla Palma, P., Pelligrini, S., Fina,
P., Doglioni, C., Bevilacqua, G., and Barbareschi, M. mdm2 gene alterations
and mdm2 protein expression in breast carcinomas, J Pathol. 175:31-38, 1995.
14. Lundgren, K., Montes de Oca Luna, R., McNeill, Y., Emerick, E., Spencer, B.,
Barfield, C., Lozano, G., Rosenburg, M., and Finlay, C. Targeted expression of
mdm2 uncouples S phase from mitosis and inhibits mammary grand development
independent of p53., Genes & Dev. II: 714-725, 1997.
15. Sigalas, I., Calvert, A., Anderson, J., Neal, D., and Lunec, J. Alternatively
spliced mdm2 transcripts with loss of p53 binding domain sequences:
transforming ability and frequent detection in human cancer, Nature Med. 2: 912-
917, 1996.
16. Gudas, J., Nguyen, H., Klein, R., Katayose, D., Seth, P., and Cowan, K.
Differential expression of multiple mdm2 messenger RNAs and proteins in
normal and tumorigenic breast epithelial cells, Clin Cancer Res. /: 71-80, 1995.
17. Wen, W.-H., Reles, A., Sullivan-Halley, J., Bernstein, L., Jones, L., El-
Naggar, A., Felix, J., Runnebaum, I., and Press, M. p53 mutations and
expression in ovarian cancers: correlation with overall survival, Int J Gyn One.
in press:, 1998.
18. Slamon, D., Godolphin, W., Jones, L., Holt, J., Wong, S., Keith, D., Levin,
W., Stuart, S., Udove, J., Ullrich, A., and Press, M. Studies of the HER-2/nen
proto-oncogene in human breast and ovarian cancer., Science. 244: 707-712,
1989.
19. Press, M., Nousek-Goebl, N., King, W., Herbst, A., and Greene, G.
Immunohistochemical assessment of estrogen receptor distribution in the human
endometrium throughout the menstrual cycle, Lab Invest. 51:495-502, 1984.
20. Bacus, S., Bacus, J., Slamon, D., and Press, M. HER-2/neu oncogene
expression and DNA ploidy analysis in breast cancer, Arch Pathol Lab Med. 14:
164-169, 1990.
21. Montes de Oca Luna, R., Tabor, A., Eberspaecher, H., Hulboy, D., Worth, L.,
Colman, M., Finlay, C., and Lozano, G. The organization and expression of the
mdm2gene, Genomics. 33: 352-357, 1996.
22. Tran, H., Gordenin, D., and Resnick, M. The prevention of repeat-associated
deletions in sacchaormyces cerevisiae by mismatch repair depends on size and
origin of deletions, Genetics. 143: 1579-1587,1996.
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
23. Santoro, I. and Groden, J. Alternative splicing of the APC gene and its
association with terminal differentiation, Cancer Res. 57:488-494, 1997.
24. Li, L. and Cohen, S. tsglOl: a novel tumor susceptibility gene isolated by
controlled homozygous functional knockout of allelic loci in mammalian cells.,
Cell. 85: 319-329, 1996.
25. Li, L., Li, X., Francke, U., and Cohen, S. The TSG101 tumor susceptibility
gene is located in chromosome 11 band pl5 and is mutated in human breast
cancer., Cell. 88: 143-154, 1997.
26. Gayther, S., Barski, P., Batley, S., Li, L., de Foy, K., Cohen, S., Ponder, B.,
and Caldas, C. Aberrant splicing of the TSG101 and FHIT genes occurs
frequently in multiple malignancies and in normal tissues and mimics alterations
previously described in tumours, Oncogene. 15: 2119-2126, 1997.
27. Lee, M. and Feinburg, A. Aberrant splicing but not mutations of TSG101 in
human breast cancer, Cancer Res. 57: 3131-3134, 1997.
28. Negrini, M., Monaco, C., Vorechovsky, I., Ohta, M., Druck, T., Baffa, R.,
Huebner, K., and Croce, C. The FHIT gene at 3pl4.2 is abnormal in breast
carcinomas, Cancer Res. 56: 3173-3179, 1996.
29. Fong, K., Biesterveld, E., Virmani, A., Wistuba, I., Sekido, Y., Bader, S.,
Ahmadian, M., Ong, S., Rassool, F., Zimmerman, P., Giaccone, G., Gazdar,
A., and Minna, J. FHIT and FRA3B 3pl4.2 allele loss are common in lung
cancer and preneoplastic bronchial lesions are associated with cancer related
FHIT cDNA splicing abberations, Cancer Res. 57: 2256-2267, 1997.
30. Virgilio, L., Shuster, M., Gollin, S., Veronese, M., Ohta, M., Huebner, K .,
and Croce, C. FHIT gene alterations in head and neck squamous cell carcinomas,
Proc Natl Acad Sci, USA. 93: 9770-9775, 1996.
31. Bueso-Ramos, C., Manshouri, T., Haidar, M., Yung, Y., McCown, P.,
Ordonez, N., Glassman, A., Sneige, J., and Albitar, M. Abnormal expression of
mdm2 in breast carcinomas, Breast Cancer Res Treat. 37: 179-188, 1996.
32. Zhang, H., Hannon, G., and Beach, D. p21-containing cyclin kinases exist in
both active and inactive states., Gene Dev. 8: 1750-1758, 1994.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER 4: Differential Expression of Genes
During Breast Cancer Progression
A b s t r a c t
Malignancies are thought to develop as a result of molecular genetic alterations,
either inherited or acquired. Considerable evidence indicates that a malignant phenotype
is demonstrated only when the genes effecting this phenotype are continuously
expressed (oncogenes) or, alternatively, continuously interrupted (tumor suppressor
genes). Thus, it is expected that there is a set of genes whose expression is responsible
for the malignant phenotype in any given cancer. Comparison of gene expression in
benign and malignant cells should permit identification of those genes which are both
up-regulated and down-regulated in particular cancers.
Ductal carcinoma in situ (DCIS) of the breast is commonly associated with
invasive disease, and strong circumstantial evidence suggests that DCIS is a precursor to
invasive disease. Areas of advanced disease are frequently associated with hyperplasia
and adenosis, and it is thought that different phenotypic changes may be the result of
genotypic changes. The objectives of this study were to identify, isolate and
characterize genes whose expression is altered in the progression from benign
hyperplasia through DCIS to invasive breast disease and to determine if these genes play
a role in maintaining a malignant phenotype. We utilized one clinical specimen of breast
cancer containing of three separate histologic types: benign hyperplasia, ductal
carcinoma in situ, and invasive breast disease. Each histologically different lesion was
microdissected and the RNA extracted separately. This strategy allowed assessment of
107
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
incremental changes in mRNA expression using the differential mRNA display
technique (DD). One hundred different sequences differentially expressed by DD were
isolated and cloned. Northern blot analysis, using an panel of RNAs from normal
breast epithelium and invasive breast disease, were performed using each clone as a
probe to confirm differential expression. Fourteen clones had expression patterns which
matched the original DD expression. Five were known genes, and four had been
previously reported to be over- or under-expressed as described here. We also have
found 7 undescribed genes fragments which are differentially expressed and thus could
represent new and potentially important tumor suppressor genes or oncogenes.
I n t r o d u c t io n
Strong circumstantial evidence indicates ductal carcinoma in situ (DCIS) of the
breast is a precursor of invasive disease. This evidence includes the frequent finding of
DCIS associated with invasive carcinomas and the development of invasive carcinoma in
relation to biopsy findings originally reported as benign but on review found to be in
situ carcinoma (1). Furthermore, this model of DCIS as a pre-invasive malignant
change is consistent with the paradigm of carcinoma in situ in other organs, such as the
cervix.
Although the sequential steps in the neoplastic process are not well characterized
for any malignancy, the recent elucidation of the progression of colorectal cancer
indicates that a sequence of genetic alterations appears critical in colorectal
tumorigenesis. Morphologically, tumorigenesis of the colon appears to begin with
normal epithelium becoming hyperproliferative. Adenomatous growth (early,
intermediate, and late) may follow; as may invasive carcinoma and metastatic disease.
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Vogelstein and coworkers (2) found evidence of genetic changes at each step in this
progression. Thus, the phenotypic changes appear to correspond to genetic alterations.
The structure of the breast makes finding a similar series of morphologic
transitions in breast cancer difficult. However, in the tissue surrounding a focus of
invasive disease, there are generally areas of DCIS, epithelial hyperplasia and atypical
hyperplasia. Many studies reveal the genetic alterations found in invasive breast disease:
from amplification of the proto-oncogenes HER2/neu, myc, epidermal growth factor
receptor, cyclin D l, and/or the fibroblast growth factor genes bek and fig, as well as
loss of tumor suppressors genes such as p53 and Rb. Much less work has been
performed in DCIS and earlier lesions. We and others have recently shown that p53 is
mutated (3, 4), and over-expressed (5) in DCIS. P53 over-expression and p53
mutations have been noted in benign breast tissue (6). Frequent alteration in
chromosome 1 (7) and loss of heterozygosity in numerous loci in at least 11 % of cases
(8) are also described in DCIS. In vivo studies suggest that c-Ha-ras may play a role in
progression from fibrocystic disease to carcinoma in situ. However, to our knowledge,
no group is pursuing an investigation of the differences in gene expression between
breast DCIS cells and benign breast epithelium.
Early methods developed to identify and clone such over-expressed or under-
expressed genes were primarily based on the principle of subtractive hybridization (9).
Despite their usefulness, these methods analyze only a fraction of the overall changes in
gene expression, require large amounts of RNA and are lengthy and laborious. An
improved method, referred to as differential mRNA display (DD), was developed by
Liang and Pardee (10) using a gel-based technique that facilitates a rapid and extensive
analysis of differentially expressed mRNAs. This method utilizes PCR to display
mRNAs that are differentially expressed between different cell lines or tissues (10, 11).
109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Certain limitations of the original technique have been overcome such as the use of I-
base anchored oligo(dT)I 2 primer for increased representation of mRNAs and the use of
long composite primers to achieve reproducible patterns under more stringent PCR
conditions. Current differential mRNA display techniques produce highly consistent
and reproducible patterns and detects almost all mRNAs in a sample (12). This method
enables the systematic and nearly exhaustive definition of changes in gene expression of
cells with altered physiologic states permitting simultaneous identification of over
expressed and under-expressed genes in the same experiment.
A number of groups have used differential display to isolate and characterize
new genes from many tissue types. Using differential display to analyze the differences
between cultured cell lines from normal and invasive breast disease, Sager (13, 14)
isolated two candidate tumor suppressor genes expressed in normal cells: alpha-6
integrin, a component of the integrin receptor, and maspin, a serpin family protease. A
study for genes overexpressed in senescent human breast epithelial cells generated an
insulin growth factor-like binding protein, mac25 (15). This technique has been used to
isolate differentially expressed genes in endometrium (Saffari and Press, unpublished
observations), nerve (16), brain (17), colon (18) and prostate (19). Our work in breast
cancer has also generated some very promising results.
M aterials a nd M eth o ds
Tissue
The use of human tissue in this study was reviewed and approved by the
University of Southern California Institutional Research Committee. One fresh frozen
invasive ductal breast carcinoma (C2873) with a distinct areas of invasive carcinoma, in
110
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
situ comedocarcinoma and benign hyperplasia, stored at -186°C in a liquid nitrogen
freezer, was obtained from the USC Breast Tumor and Tissue Bank and used for this
study. A frozen tissue section stained with hematoxylin-and-eosin was used to confirm
the histological composition of the specimen.
Microdissection and Total RNA isolation
Microdissection was performed to separate benign breast hyperplastic
epithelium, breast CIS and invasive carcinoma. The different tissue components of the
breast specimen were microdissected with the assistance of a microscope. Each tissue
histology was microdissected out of single 20 micron thick tissue sections and the RNA
was extracted separately. The initial section was stained with hematoxylin-and-eosin
and subsequent sections for microdissection were stained with the nuclear stain ethyl
green. Approximately 60 sections were required to obtain 10 micrograms of RNA for
each tissue. Care was taken in order to minimize contamination of epithelial cells with
stromal cells. After microdissection, total RNA was extracted using TRIzol
(Gibco/BRL, Grand Island, NY) according to manufacturer’s instructions. The RNA
was precipitated using isopropanol, washed once with 70% ethanol, and allowed to dry
at room temperature. The RNA was rehydrated in 30 pi of RNAse free H2 0 and the
optical density measured.
Differential mRNA Display (DPI
DD was performed essentially as described (10, 20) with some modifications.
Two micrograms of total RNA was digested using DNAase I (Pharmacia, Piscataway,
NJ) to remove contaminating genomic DNA prior to reverse transcription. The total
RNA was then extracted using phenol/choroform/isoamyl alcohol (at a 50/49/1 ratio)
and RNA precipitated using ethanol.
I ll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cDNAs were prepared from each separate tissue type using MMLV reverse transcriptase
(Gibco/BRL) in concert with one of four degenerate oligo(dT) primers (dT,,VA,
dT,2 VC, dT1 2 VG, and dTI2 VT) (Operon Biotechnologies, Alameda, CA) where V is a
mixture of A, C and G (10). PCR was performed with the proper degenerate
oligo(dT),,VN and one of 26 arbitrary 10 base primers in the presence of 3 3 P-dATP
(DuPont, Boston, MA). Each arbitrary primer binds at a different site along the cDNA,
producing a series of random, but reproducible, PCR products. The samples were
separated side by side through 5.5% polyacrylamide gels by running at 120 watts for
approximately 2.5 hours, then dried and autoradiographed overnight using BioMax MR
autoradiography film (Kodak, Rochester, NY) (Figure 4-1). This technique generated
approximately 10,000 different bands representing distinct mRNA species.
The differentially expressed PCR products were identified, excised from the gel,
the DNA extracted and reamplified with the original degenerate oligo d(T)I2 VN and
arbitrary 10 base primer. The product was then purified with Qiagen PCR purification
columns (Qiagen, Chatsworth, CA) and ligated into pCR 2.0 vector (Invitrogen,
Carlsbad, CA). Inv-Fa’ bacteria (Invitrogen) were transformed and grown up on
selective media. Approximately 10-20 colonies were picked and PCR amplified to
assess for the proper size of insert using M13 primers specific to the vector. Colonies
with the appropriate sized insert had the plasmid isolated (Wizard plasmid minipreps,
Promega, Madison, WI) and sequenced using the Sequenase kit (Amersham, Arlington
Heights, EL), according to the manufacturer’s instructions. Approximately 250 ng of
plasmid was used for sequencing.
Northern Analysis
Northern analysis of each clone is an important step in confirming differential
expression. Northern hybridization using a panel of 15 invasive carcinomas and 4
112
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
samples of normal epithelium was performed to confirm that the partial cDNA clones
were differentially expressed. The northerns were assembled as follows. Invasive
breast cancers were assessed by hematoxilyn and eosin to assure there was a high tumor
to normal epithelium ratio prior to RNA extraction. A similar procedure was followed
with the tissues containing benign epithelium. Total RNA was extracted directly from
10-20 serially cut, 10 pm thick, frozen tissue sections using TRIzol (Gibco/BRL)
according to manufacturer’s instructions. The RNA was then separated on 1.0%
agarose gel containing formaldehyde, blotted onto on Duralon-UV membranes
(Stratagene, La Jolla, CA) and probed with P3 2 ATP labeled individual DD clones.
Hybridization was in QuikHyb solution (Stratagene) and washes were at high
stringency. Blots were exposed -70°C to BioMax MS film (Kodak), using BioMax MS
intensifying screens.
R e s u l t s.
In the invasive breast carcinoma - CIS - benign hyperplastic epithelium
specimen, we identified 231 differentially or aberrantly expressed genes. The majority,
130 of the 231 aberrantly expressed genes, were over-expressed in the carcinoma
relative to the benign hyperplastic epithelium. The other 101 of these 231 aberrantly
expressed genes were under-expressed in the carcinoma relative to the benign
hyperplastic epithelium. These bands were excised from the gels, and we were able to
reamplify 137 of 231 and clone 119 fragments. Sequencing of each sample revealed
100 individual partial gene fragments (Table 4-1). Database analysis of the partial
cDNAs revealed that 20 clones were known genes (Tables 4-2 and 4-3), eighteen
showed homology with partial cDNA clones whose sequence is deposited in Genbank
but have been otherwise not characterized, and that the majority, 66, had no significant
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4-1: Differential mRNA Display.
Polyacrylamide gel electrophoresis to display differentially expressed PCR products using
arbitrary 5’ -primer 1 with three different degenerate oligo-dT primers (dTl2 VC, dT,,VG,
dT,2 VT) as 3’ -primers. The majority of the cDNAs displayed show comparable levels of
expression in benign hyperplasia (H), DCIS (C), and invasive disease (I) mRNA. Examples
of cDNAs which are expressed differentially in benign hyperplasia, DCIS or in invasive
disease are identified (boxes in the illustration and Tables 2-5).
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Primer 1
VC VG VT
HCI HCI HCI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
Table 4-1
Characterization of cDNAs Corresponding to Genes Differentially Expressed in Invasive Breast Carcinoma,
DCIS and Benign Hyperplasia
Comparative Expression in No of Nucleotide Sequence Information
Invasive Breafct Carcinoma
cPNAs Qenbaok Database
Over Expressed 12 see Table 2
Over Expressed 15 uncharacterized cDNA clones
Over Expressed 38 no significant sequence homology, a novel gene
Under Expressed 8 see Table 3
Under Expressed 3 uncharacterized cDNA clones
Under Expressed 28 no significant sequence homology, a novel gene
Table 4-2
Known Sequences with Significant Homology to Genes Over Expressed in Invasive Breast Carcinoma by
Differential Display
C l o n e
Hbmologous Gene
JJL16 17-j} hydroxysteroid dehydrogenase
AJL5 Ferritin
FJL27 L5 ribosomal protein
FJL 28 DAP-5
BJL9-11 Follistatin
GJL8 ADP/ATP translocase
HJL8 a -2 macroglobulin
KJL11 branched acetyl transferase
GJL13 23 kD highly basic protein
JJL12 elongation factor 2 (EF2)
FJL25 human transposon L I.2
HJL11 translocon-associated protein
Funstkm /-Significance
estrone to E2 catalysis/ high E2= incr. BrCA risk
Intracellular iron storage / increased in colon CA
5S RNA component / complexes with MDM2
conveys resistance to IFN y induced cell death
FSH inhibitor
mitochondrial ATP exporter
protease inhibitor / high Gleason scores PrCA
developmental enzyme
related to LI3 ribosomal subunit
required for translocation in protein synthesis
jumping gene / hemophilia, insertion^ mutagenesis
ER transmembrane protein complex subunit
with permission o f th e copyright owner. Further reproduction prohibited without permission.
O
CD
Q _
Table 4-3
Known Sequences with Significant Homology to Genes Under Expressed in Invasive Breast Carcinoma by
Differential Display
ClPQS
DJL9-14
IJL1
EJL22-23
FJL5
CJL20
AJL1
GJL11-12
KJL2
H o m o l f l g Q u s - d e D e
1 apolipoprotein D
FTPG1 tumor suppressor
HL23 ribosomal protein
Rev interacting protein 1
adenylate kinase 3
OC-4/IGFBP3 protein
hRPB 33
T-cell receptor ft chain
ncancs
component of HDL / underexpressed in BrCA
protein tyrosine phospatase / altered in colon CA
ribosomal subunit/overexpressed in benign breast lesions
a member of the nucleoporins I interacts with HIV
phosphorylation of AMP
insulin-like growth factor binding protein 3
RNA polymerase II subunit / DNA NER repair
immunologic function / truncations in leukemias
0 0
sequence homology. These represent novel sequences not previously identified (Table
4-1).
Northern blots (having both invasive breast cancer and benign breast epithelial
RNAs) were performed on 100 clones to confirm or refute differential expression as
seen on DD. Fifty two showed no expression on the northern blots. Among these 52
genes, the majority were sequences which had no homology with known sequences.
Two known genes, the PTPG-l putative tumor suppressor gene and hRPB-33 also
exhibited no expression. Of the 48 samples which showed expression, 29 cases
exhibited approximately equal expression, while 19 clones had either over or under
expression when comparing the invasive breast cancers to the normal epithelia. Of these
19, 14 cases had expression patterns which corresponded to the pattern seen in the DD
gels (Figure 4-2, and Tables 4-4 and 4-5) and 5 had the opposite expression pattern.
These 14 cases contained 5 characterized genes. Three were overexpressed in
tumor: ferritin, DAP-5 and o^-macroglobuLin. Each of these upregulated genes had
been previously cited in the literature as upregulated in carcinoma. Ferritin is an
intracellular iron storage molecule, which exhibits growth enhancing properties in some
human cell lines. Northern analysis of ferritin showed high expression invasive breast
carcinoma RNA, expressed at a lower amount in a single case of CIS and not expressed
in any of 4 normal breast epithelial RNAs. Contrary to it’s name, Death Associated
Protein 5 (DAP-5) was found to protect HeLa cells from interferon gamma programmed
cell death. Expression of this gene on northern blots showed increased amounts of
mRNA in 10 of 13 invasive tumors and low or no expression in normal breast
epithelium. Similarly, 0,-macroglobulin is a protease inhibitor which regulates the
activity of many cytokines and growth factors. Northern analysis showed that each of
the invasive RNAs had increased expression of this gene compared to the normal
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
epithelial RNAs (data not shown). A putative tumor suppressor gene, BC-1 (Breast
Cancer-1) had no known homologies, and was very highly expressed (approximately
10X) compared to normal epithelium (Figure 4-2, D).
Two genes were expressed more highly in the normal breast epithelium:
apolipoprotein D and L17 ribosomal protein. Apolipoprotein D is a glycoprotein with a
role in binding lipids and steroids. Apolipoprotein D was very highly expressed in 4 of
5 normal breast epithelium cases, as compared to moderated expression in 3 of 13 cases
of invasive breast cancer (Figure 4-2, A). L17 ribosomal protein is a member of a large
group of ribosomal proteins which make up the ribosome, and is one that is
overexpressed in times of amino acid starvation. L17 was weakly expressed in normal
epithelium RNAs and not expressed in carcinoma (data not shown). We also found a
potential tumor suppressor gene. Called BREP-1 (BReast EPithelium-1), this gene is
currently uncharacterized EST and is expressed weakly in normal breast epithelium and
not at all in invasive breast cancers (Figure 4-2, C).
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4-2: Northern Analysis of Differentially Displayed mRNAs.
A) Ethidium bromide stained total RNA from a series of invasive breast cancers and
normal breast epithelium. 12.5 pg of total RNA was electrophoresed in 1% agarose
containing formaldehyde. Note equal loading and intactness of the 18S and 28S bands.
B) Northern Hybrization of Apolipoprotein D with Human Breast mRNAs. The
Apolipoprotein D recognizes a 1.1 kb message expressed at a high level in most in normal
breast epithelium (N 1 Br Epithelium and Nl. Epi) but is expressed at much lower levels or
not expressed in invasive breast carcinomas (Invasive Breast Carcinomas and Invasive
Breast CA).
C) Northern Hybridization of BREP-1 with Human Breast mRNAs. The BREP-1 probe
recognizes a 1.0 kb message that is expressed more highly in the invasive breast
carcinomas as compared to the normal epithelium.
D) Northern Hybridization of BC-1 with Human Breast mRNAs. The BC-1 fragment
recognizes a 1.6 kb message that is expressed more highly in the invasive breast
carcinomas as compared to the normal epithelium.
E) Northern Hybridization of Death Associated Protein-5 (DAP-5) with Human Breast
mRNAs. The DAP-5 fragment recognizes a 3.5 kb message that is expressed more highly
in the invasive breast carcinomas as compared to the normal epithelium.
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission o f th e copyright owner. Further reproduction prohibited without permission.
A. EtBr stained total RNA
I8S-
B. Apolipoprotein D (DJL9)
C. BREP-1 (CJL19)
D. BC-1 (AJL6)
E. Death Associated Protein 5
(FJL28)
to
to
Table 4-4
Sequences Over Expressed in Invasive Breast Carcinoma by Confirmatory
Northern Biots
Clone • Homologous .Gene
Function and Significance
AJL5 Ferritin cellular iron storage
AJL6 no homology-BCl unknown
BJL8 uncharacterized cDNA unknown
EJL15 no homology unknown
EJL19 uncharacterized cDNA unknown
FJL18 no homology unknown
FJL28 DAP-5 conveys resistance to IFN y induced cell death
GJL7 uncharacterized cDNA unknown
HJL8 a -2 macroglobulin protease inhibitor
HJL12 no homology unknown
Table 4-5
Sequences Under Expressed In Invasive Breast Carcinoma by Confirmatory
Northern Blots
C lo n e Hom olog om G sne function and significance
CJL19 uncharacterized cDNA-BREPI unknown
EJL22-23 L17 ribosomal protein ribosomal protein
DJL9-14 apolipoprotein D component of HDL
D iscu ssio n
Carcinoma of the breast is not a single morphologic entity but generally a
collection of different histologic tissue types. Microdissection and extraction of RNA
from the different morphologies allows the testing of the hypothesis that the
morphologic changes in breast cancer are directly related genetic changes. Because the
small size of the single case, differential display is well suited to reveal the differences in
gene expression between these different disease entities.
Analysis of the results of the differential display show that in both groups of
isolates that are identical or have strong homology to previously characterized genes,
there is evidence that validates the use of differential display. Among the potential
oncogenes, both ferritin and the L5 ribosomal protein have been previously described as
overexpressed in neoplastic tissues. Ferritin has been described as a possible growth
factor for HL-60 leukemia cells (21). The expression of the ribosomal subunit L5 is
also altered in colorectal carcinogenesis (22) and binds to the MDM-2 oncogene protein
(23). Two genes control hormone levels: follistatin, a binding protein that inhibits
follicle stimulating hormone, and 17-p hydroxysteroid dehydrogenase, which catalyzes
the conversion of the relatively inactive estrone to the highly active estrogen in the cell.
Overexpression of 17-p hydroxysteroid dehydrogenase could lead to increase estrogen
in the cell and increase cellular proliferation (24). Among the potential tumor suppressor
genes with lower expression in the carcinoma tissues, the putative tumor suppressor
gene, PTPG-1, codes for an intracellular protein tyrosine phosphatase which was
mutated in a human colonic carcinoma cell line (25). Apolipoprotein D is a component
of high density lipoprotein but also has growth regulatory functions. It is expressed in
124
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nonproliferating fibroblasts (26) and induction by rednoic acid is concomitant with
growth arrest and cell differentiation in human breast cancer cells (27).
In general, the expression of genes appear to conform to broad patterns
predicated on the tissue. For instance, in the benign hyperplasia, the pattern of gene
expression appears to relate to differentiation and the slowing of cell growth through the
expression of the PTPG-1 tumor suppressor gene, DNA repair proteins (hRBP 33) and
L17 ribosomal protein (also known as HL23). Genes which could not have been
predicted but have been previously reported as downregulated in tumors, such as
apolipoprotein D, are also illustrated. On the other hand, those transcripts over
expressed in the carcinoma were related to the robust growth properties of these tissues
such as ribosomal proteins, protein elongation factors, protein translocation factors were
also expressed.
Analysis of results from 100 confirmatory northerns proved differential
expression in 13 clones. Four had no homology to known genes and 4 were
uncharacterized cDNAs. Three known genes were found to overexpressed in the
carcinoma: ferritin, cl, macroglobulin, DAP-5; while two were underexpressed in these
tissues: apolipoprotein D and L17 ribosomal protein. Each of the potential oncogenes
have been previously described as upregulated in tumors and were genes that were
unexpected results from our screens. For instance, ferritin is the major intracellular iron
storage protein in many organisms. This protein was also recently described as a
growth factor human lung, leukemia and melanoma cell lines (21). In a large cohort of
colonic tissue, significantly greater amounts of ferritin were found in the carcinomas as
compared to normal epithelia (p<0.001) and greater ferritin expression was associated
with increased degree of dysplasia (p=0.039) (28). In the breast cell line MCF-10F
cells and their mortal progenitor, S-130 cells, subtractive hybridization, in revealed
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
increased expression of ferritin. Subsequent northern analysis and in situ hybridization
on breast cancer specimens showed increased ferritin protein in areas of ductal
hyperplasia, CIS and invasive disease. The authors postulated that increased iron may
account for the increased growth in the cells (29). Another transcript over expressed in
tumor, otj-macroglobulin, binds ferritin (30) and regulates the distribution and the
activity of many cytokines and growth factors such as transforming growth factor p,
tumor necrosis factor a, platelet derived growth factor, basic fibroblasts growth factor,
interleukin ip and others (31). Interestingly, (Xi-macroglobulin was initially described
as a protease inhibitor active against all types of proteases. While no reports suggest a
role for otj-macroglobulin in breast cancer, high levels of a,-macroglobulin correlated
significantly with high Gleason scores in prostate cancer. This suggests that a 2-
macroglobulin enhances growth either alone and/or by way of the growth factors it
carries. The death associated protein-5 (DAP-5) was a gene that we expected to see.
The fragment was 97% homologous to eukaryotic translation initiation factor 4G, part of
the mRNA cap binding complex. This complex of proteins is essential for proper
translation. Mild overexpression of DAP5 prevents HeLa cells from undergoing
apoptotic death, and thus can promote a tumorigenic phenotype (32).
Among those genes that we confirmed to be downregulated in the in situ and
invasive carcinoma, both apolipoprotein D (apoD) and L17 ribosomal protein were
initially surprising. ApoD is a glycoprotein component of high density lipoprotein in
human plasma with no apparent connection to cancer. Interestingly, apoD has no
sequence similarity with other apolipoproteins but has a high degree of homology to
plasma retinol binding proteins and is a member of the otj-microglobulin superfamily
(33). Most lecithin:cholesterol acetyltransferase is complexed with apoD suggesting that
apoD has a role in lipid binding and/or transport (34). In addition, apoD binds to
126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
progesterone and the steroid precursor pregnenlone and thus may also serve as a carrier
or chaperone for the steroid byproducts of cholesterol (35). The specific relationship of
apoD to breast cancer is supported by the fact that apoD is the major protein in cystic
breast disease fluid (35), and certain breast cancers synthesize apoD (36). Lower levels
of apoD are significantly associated with shorter relapse free and overall survival (37).
This data, combined with the evidence that apoD expression coincides with the
inhibition of cell growth in fibroblasts (26), breast cancer (27) and prostate cancer (38),
and this study, indicates that apoD may be a marker of cell differentiation and growth
arrest in breast and other tissues.
Another known gene, the L 17 ribosomal protein, is another innocuous case. In
mammals, there are 4 species of RNA and 70-90 different ribosomal proteins that are
associated with the RNA. The L17 protein is specifically upregulated and sequestered to
the nucleus during amino acid starvation (39). The broader implications of this
remained to be elucidated, yet this result accurately reflects the decreased nutrients
supply probably common to many tumors; furthermore, the central necrosis in the CIS
of the study case lends weight to this result.
DD is a laborious and robust technique that has been used extensively in the hunt
for novel genes. One of the well described problems is the high rate of false positive
clones noted here and in other studies (11, 20, 40). Since differential display was
performed on a single case, and the RNA was not represented on the confirmatory
northern blots, it is not possible to ascertain whether DD accurately defines the
expression of genes in that case. Thus, these individual DD variations may be
alterations that are specific to the case used and not wholly generalizable over a large
series of other individual tumors. Interestingly, at least two genes found in this
analysis, but not found to be expressed by northern blot, were also found using
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
different techniques in different tumor types (41). Therefore, given the number of genes
differentially displayed in a pattern corroborated by the literature, we believe that DD
provides valuable information regarding the genetic changes that define both the tumor
and normal phenotype. Still, the utility and robust nature of DD is clear from the
success in finding both known and unknown genes that may extend our understanding
of breast carcinogenesis.
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
References
1. Varley, J., Brammar, W., Lane, D., Swallow, J., Dolan, C., and Walker, R.
Loss of chromosome 17pl3 sequences and mutation of p53 in human breast
carcinomas., Oncogene. 6: 413-421, 1991.
2. Fearon, E. and Vogelstein, B. A genetic model for colorectal tumorigenesis, Cell.
61: 759-767, 1990.
3. Lukas, J., Niu, N., and Press, M. p53 mutation and overexpression in ductal
carcinoma in situ of the breast, submitted, 1998a.
4. O’ Malley, F. P., Vnencak-Jones, C. L., Dupont, W. D., Pari, F., Manning, S.,
and Page, D. L. p53 mutations are confined to the comedo type ductal carcinoma
in situ of the breast, Lab Invest. 71: 67-72, 1994.
5. Davidoff, A. M., Kerns, B.-J. M., Pence, J. C., Marks, J. R., and Iglehart, J.
D. p53 alterations in all stages of breast cancer, J Surg Oncol. 48: 260-267,
1991.
6 . Millikan, R., Hulka, B., Thor, A., Zhang, Y., Edgerton, S., Zhang, X., Pei,
H., He, M., Wold, L., Melton, L. J., Ballard, D., Conway, K., and Liu, E. T.
p53 mutations in benign breast tissue, J Clin Oncol. 13: 2293-2300, 1995.
7. Munn, K. E., Walker, R. A., Menasce, L., and Varley, J. M. Mutation of the
TP53 gene and allelic imbalance at chromosome 17pl3 in ductal carcinoma in
situ., Br. J. Cancer. 74: 1578-1585, 1996.
8. Stratton, M., Collins, N., Lakhani, S., and Sloane, J. Loss of heterozygosity in
ductal carcinoma in situ of the breast, J Pathol. 175: 195-201, 1995.
9. Hedrick, S., Cohen, D., Nielsen, E., and Davis, M. Isolation of cDNA clones
encoding T cell-specific membrane-associated proteins, Nature. 308: 149-153,
1984.
10. Liang, P. and Pardee, A. Differential display of eukaryotic messenger RNA by
means of the polymerase chain reaction, Science. 257: 967-971, 1992a.
11. Liang, P., Averboukh, L., Keyomarsi, K., Sager, R„ and Pardee, A.
Differential display and cloning of messenger RNAs from human breast cancer
versus mammary epithelial cells, Cancer Res. 5 2 :6966-6969, 1992b.
12. Prashar, Y. and Weissman, S. Analysis of differential gene expression by
display of 3’ end restriction fragments of cDNAs, Proc. Natl. Acad. Sci. USA.
93: 659-663, 1996.
129
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
13. Sager, R., Anisowicz, A., Neveu, M., Liang, P., and Sotiropoulou, G.
Identification by differential display of alpha 6 integrin as a candidate tumor
suppressor, FASEB J. 7: 964-970, 1993.
14. Zou, Z., Anisowicz, A., Hendrix, M., Thor, A., Neveu, M., Sheng, S., Rafidi,
K., Seftor, E., and Sager, R. Maspin, a serpin with tumor suppressing activity in
human mammary epithelial cells, Science. 263: 526-529, 1994.
15. Swisshelm, K., Ryan, K., Tsuchiya, K., and Sager, R. Enhanced expression of
an insulin growth factor-like binding protein (mac25) in senescent human
mammary epithelial cells and induced expression with retinoic acid., Proc Natl
Acad Sci USA. 92: 4472-4476, 1995.
16. Kiryu, S., Yao, G., Morita, N., Kato, H., and Kiyama, H. Nerve injury
enhances rat neuronal glutamate transporter expression: identification by
differential display PCR., J Neuro. 15: 7872-7878, 1995.
17. Inokuchi, K., Kato, A., Hiraia, K., Hishinuma, F., Inoue, M., and Ozawa, F.
Increase in activin beta A mRNA in rat hippocampus during long-term
potentiation, FEBS Letters. 3 8 2 :48-52, 1996.
18. Yeatman, T. and Mao, W. Identification of a differentially-expressed message
associated with colon cancer liver metastasis using an improved method of
differential display, Nucleic Acids Res. 23:, 1995.
19. Shen, R., Su, Z., Olsson, C., and Fisher, P. Identification of the human
prostatic carcinoma oncogene PTI-1 by rapid expression cloning and differential
RNA display, Proc Natl Acad Sci, USA. 92:, 1995.
20. Liang, P., Averboukh, L., and Pardee, A. Distribution and cloning of eukaryotic
mRNAs by means A dfferential display: refinements and optimization., Nucleic
Acids Res. 21: 3269-3275, 1993.
21. Kikyo, N., Hagiwara, K., Fujisawa, M., Kikyo, N., Yazaki, Y., and Okabe, T.
Purification of a cell growth factor from a human lung cancer cell line: its
relationship with ferritin, J Cell Physiol. 161: 106-110, 1994.
22. Frigerio, J., Dagom, J., and Iovanna, J. Cloning, sequencing and expression of
the L5, L21, L27a, L28, S5, S10 and S29 human ribosomal protein mRNAs,
Biochimica et Biophysica Acta. 1262: 64-68, 1995.
23. Marechal, V., Elenbaas, B., Piette, J., Nicolas, J., and Levine, A. The
ribosomal L5 protein is associated with mdm2 and mdm2-p53 complexes, Mol
Cell Biol. 14: 7414-7420, 1994.
24. Coldham, N. and James, V. A possible mechanism for increased breast cell
proliferation by progestins through increased reductive 17-beta hydroxysteroid
dehydrogenase activity, Int J Cancer. 45: 174-178, 1990.
25. Takekawa, M., Itoh, F., Hinoda, Y., Adachi, M., Ariyama, T., Inazawa, J.,
Imai, K., and Yachi, A. Chromosomal localization of the protein tyrosine
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
phosphatase G1 gene and characterization of the aberrant transcripts in human
colon cancer cells, FEBS Lett. 339: 222-228,1994.
26. Provost, P., Marcel, Y., Milne, R., Weech, P., and Rassart, E. Apolipoprotein
D transcription occurs specifically in nonproliferating quiescent and senescent
fibroblast cultures, FEBS Lett. 290: 139-141, 1991.
27. Lopez-Boado, Y., Tolivia, J., and Lobez-Otin, C. Apolipoprotein gene induction
by retinoic acid is concomitant with growth arrest and cell differentiation in
human breast cancer cells, J Biol Chem. 269: 26871-26878., 1994.
28. Yang, H., Hsu, J., Lee, J., Chan, S., Lin, X., and Chow, N. Adenoma-
carcinoma sequence: a reappraisal with immunohistochemical expression of
ferritin, J Surg Oncol. 60: 35-40, 1995.
29. Higgy, N., Salicioni, A., Russo, I., Zhang, P., and Russo, J. Differential
expression of human ferritin H chain gene in immortal human breast epithelial
MCF-10F cells, Mol Carcino. 20: 332-339, 1997.
30. Bellotti, V., Arosio, P., Cazzola, M., Cozzi, A., Levi, S., Meloni, F., and
Zoppone, E. Characteristics of a ferritin-binding protein present in human serum,
B rJ Haematol. 65: 489-493, 1987.
31. LaMarre, J., Wollenberg, G , Gonias, S., and Hayes, M. Cytokine binding and
clearance properties of proteinase-activated a 2-macroglobulins, Lab Invest. 65:
3-14, 1991.
32. Levy-Strumpf, N., Deiss, L., Berissi, H., and Kimchi, A. DAP-5, a novel
homolog of eukaryotic translation initiation factor 4G isolated as a putative
modulator of gamma interferon-induced programmed cell death, Mol Cell Biol.
17: 1615-1625, 1997.
33. Fielding, P. and Fielding, C. A cholesteryl ester transfer complex in human
plasma, Proc Natl Acad Sci. 77: 3327-3330, 1980.
34. Drayna, D., Fielding, C., McLean, J., Baer, B., Castro, G., Chen, E.,
Comstock, L., Henzel, W„ Kohr, W., Rhee, L., Wion, K., and Lawn, R.
Cloning and expression of human apolipoprotein D cDNA, J Bio Chem. 261:
16535-16529, 1986.
35. Simard, J., Dauvous, S., Haagensen, D., Levesque, C., Merand, Y., and
Labrie, F. Regulation of progesterone-binding breast cyst protein GCDFP-24
secretion by estrogens and androgens in human breast cancer cells: a new marker
of steroid action in breast cancer, Endocrinology. 126: 3223-3231, 1990.
36. Lopez-Boado, Y., Puente, X., Alvarez, S., Tolivia, J., Binderup, L., and
Lopez-Otin, C. Growth inhibition of human breast cancer cells by 1,25-
dihydroxyvitamin D3 is accompanied by induction of apolipoprotein D
expression, Cancer Res. 57: 4091-4097, 1997.
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
37. Diez-Itza, I., Vizoso, F., Merino, A., Sanchez, L., Tolivia, J., Fernandez, J.,
Ruibal, A., and Lopez-Otin, C. Expression and prognostic significance of
apolipoprotein D in breast cancer, Am J Pathol. 144: 310-320, 1994.
38. Simard, J., Veilleux, R., de Launoit, Y., Haagensen, D„ and Labrie, F.
Stimulation of apolipoprotein D secretion by steroids coincides with inhibition of
cell proliferation in human LNCaP prostate cancer cells., Cancer Res. 51: 4336-
4341, 1991.
39. Laine, R., Shay, N., and Kilburg, M. Nuclear retention of the induced mRNA
following amino acid-dependent trancriptional regulation of mammalian
ribosomal proteins L17 and S25, J Biol Chem. 269: 9693-9697, 1994.
40. Linskens, M., Feng, T ., Andrews, W., Enlow, B., Saati, S., Tonkin, L., Funk,
W., and Villeponteau, B. Cataloging altered gene expression in young and
senescent cells using enhanced differential display, Nucl Acids Res. 23: 3244-
3251, 1995.
41. Zhang, L., Zhou, W., Velculescu, V., Kern, S., Hruban, R., Hamilton, S.,
Vogelstein, B., and Kinzler, K. Gene expression in normal and cancer cells,
Science. 276: 1268-1272, 1997.
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Summary and Future Directions
In the course of the last 6 years, the study of p53 and the DNA damage repair
pathway in which it serves a central role has matured as a field. When this thesis work
was begun in 1992, the investigation of p53 was in a relative infancy. p53 was only at
that time being described as mutated in sporadic breast cancers, and no work had been
done in the pre-invasive breast lesions designed to more clearly understand the gene
alterations that play a role in breast carcinogenesis.
The initial impetus for this work was, in part, a response to a counter-intuitive
but widespread idea that p53 mutations cause an increase in p53 half life and thus the
protein becomes immunohistochemically detectable. p53 mutations and protein
overexpression had been reported in invasive breast carcinomas but there was not an
exact correlation between the two results. My specific interests were in the mechanisms
of breast cancer progression and to understand this, we began investigations into the
genetic alterations in a potential precursor lesion of invasive breast cancer, ductal
carcinoma in situ. (DCIS). These studies showed that p53 was mutated prior to or at the
DCIS stage. Furthermore, this alteration was conserved between DCIS and invasive
disease, evidence that these two distinct disease entities are clonally derived. This
correspondence had been assumed but was never rigorously proven. p53 mutations
correlated in a statistically significant way with p53 overexpression in both DCIS with
and without invasion, but we found more cases with p53 overexpression in the absence
of p53 mutations in those samples with both DCIS and invasive disease. We
hypothesized that a gene or genes downstream of p53, if altered and inactived, might
cause p53 upregulation to compensate for the lack of feedback.
We next analyzed the gene p21W A F l/C ip I, a universal inhibitor of the cyclins, for
DNA alterations and protein expression. We found that p 2 i W AF1/C|P' is not mutated in
DCIS or invasive breast cancer, ovarian cancer or endometrial cancers but did exhibit a
133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DNA polymorphism. WAFl/Cipl expression did not correlate with p53 expression.
However, p2lW A F 1 /c ,P 1 expression appears to vary in a pattern that was distinct with a
tumor type, indicating that WAFl/Cipl is probably controlled in a p53 dependent and
independent fashion.
Unlike W AFl/Cipl, mdm2 is thought to be upregulated solely by p53
expression, which in response modulates p53 function. Like WAF1, mdm2 also was
not mutated in breast cancer, but appeared to be alternatively spliced in some tissues.
We found that mdm2 was both alternatively and abberantly spliced in breast cancers but
only alternatively spliced in normal breast epithelium. These results support our
hypothesis by showing that overexpression of p53 correlated with mdm2 mRNA
alterations. This indicated that in the absence of p53 mutations, p53 overexpression
may be caused by alterations of other genes in the p53 DNA damage repair pathway;
specifically aberrant splicingof mdm2 mRNAs correlated with p53 overexpression
(p=0.018) but not p53 mutations (p=0.072). Thus, it appears that aberrant mdm2
expression leads to overexpression of p53. This study also showed that p53 mutations
existed in the majority of cases with an alteratively spliced mdm2, suggestive of a causal
relationship between p53 mutations and expression of a truncated mdm2 protein.
Finally, the correlation of mdm2 altered splicing with clinical outcome indicate the
potential clinical utility of mdm2, and that in the future, a more detailed assessment of
mdm2 may be important to understanding p53 function in both the basic and clinical
science settings. Taken together, this work indicates that alterations of mdm2 mRNA
contribute significantly to the pathogenesis of breast and other cancers, and it’s specific
contribution may be currently unappreciated because overexpression of p53 protein in
the absence of p53 mutations may ‘camoflage’ the mdm2 alteration. Alterations may
also define new subsets of breast as well as other cancers..
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The present work provides important information on the nature of the DOS
lesion and the relationship of this disease to invasive disease. The greater understanding
of this morphologic entity provides a possible means of access to an early window on
the evolution of this disease, particularly important because DOS is very amenable to
treatment. The greater understanding of both p53 and mdm2 alterations and their
importance to breast cancer pathogenesis and outcome will provide a potentially useful
means of treating early disease.
We also investigated the role of novel genes in breast carcinogenesis using
differential display. While the results are preliminary, the comparisons of different
tissues types yield an important array of genes which broadly define general functions of
the tissues and which specifically define novel genes which appear to be important in the
progression of cancer. At least two genes, BREP-l and BC-1, appear to be potentially
important in breast carcinogenesis. Identification and characterization of these new
cancer-related genes will permit assessment of the functional role of the encoded
proteins in normal breast epithelium and breast carcinomas. Understanding the role of
these novel genes in health and disease is expected to permit the development of new
approaches to diagnosing and treating this disease.
135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
& &
IM AG E EVALUATION
TEST TARGET (Q A -3)
/,
■ < ?
A
150mm
I I W I G E . In c
1653 E ast Main Street
R ochester. NY 14609 USA
Phone: 716/482-0300
Fax: 716/288-5989
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Tumor necrosis factor-alpha activation of human retinal pigment epithelial cells

PDF

Mechanisms of insulin resistance in skeletal muscle.

PDF

Structure-function studies of human macrophage colony-stimulating factor (m-csf, csf-1)

PDF

Mechanisms of acquisition of molecular genetic changes during tumor development and progression

PDF

Molecular genetic alterations in endometrial cancer.

PDF

Expression of class II PAX genes in pediatric sarcomas: Characterization of the human PAX3 promoter

PDF

Insight into multistep tumor progression by analysis ofp53,c-K-ras and microsatellite mutations and telomerase activity in human cancers

PDF

The HER-2/neu oncogene in human breast cancer: From molecular mechanisms to targeted therapies

PDF

Molecular genetic changes during ovarian epithelial tumor development and progression

PDF

Breakpoint cluster region (Bcr) and active Bcr related (Abr) proteins in critical negative regulation of phagocytic leukocytes

PDF

The clinical significance, detection and biological characteristics of occult metastasis in breast cancer

PDF

Hormonal regulation in ovarian cancer development

PDF

Regulation of cell cycle G(1) phase progression by the EWS-FLI1 fusion gene

PDF

The glucose-regulated proteins: Altered expression in human breast cancer cell lines and applications in cancer gene therapy in an in vivo mouse model system

PDF

Molecular mechanisms of androgen independence in prostate cancer

PDF

Cholesterol oxidation products cause vascular injury and induce sudanophilic lesion formation in animal models

PDF

Progesterone signaling in ovarian epithelial tumors

PDF

Expression of MMP-2 and MMP-9 and their inhibitors in neurodegenerative disorders

PDF

Physiological perturbations in pharmacokinetic animal studies

PDF

T regulatory cell based immunotherapy for the treatment of cancer

Asset Metadata

Creator Lukas, Jason Jerome (author)

Core Title Molecular genetic alterations in breast cancer

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Pathobiology

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, cell,health sciences, oncology,health sciences, pathology,OAI-PMH Harvest

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-397018

Unique identifier UC11353771

Identifier 9919076.pdf (filename),usctheses-c17-397018 (legacy record id)

Legacy Identifier 9919076.pdf

Dmrecord 397018

Document Type Dissertation

Rights Lukas, Jason Jerome

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA