Loss of heterozygosity at chromosome 11 in breast cancer: Association of prognostic factors with genetic alterations

Abstract

We examined DNA from 116 female and four male breast cancer patients for loss of heterozygosity (LOH). DNA was analysed by polymerase chain reaction using ten microsatellite markers on chromosome 11. Three distinct regions of LOH were identified: 11p15.5, 11q13 and 11q22-qter with a LOH frequency of 19, 23 and 37-43% respectively. The marker D11S969 showing the highest frequency of LOH (43%) is located at the 11q24.1-q25 region. No previous molecular genetic studies have shown frequent LOH at the region telomeric to q23 on chromosome 11. Southern analysis revealed that LOH at 11q13 was due to amplification, whereas LOH at 11q22qter was due to deletion. LOH at 11p15.5 was associated with paucity of hormone receptor proteins, high S-phase and positive node status. An association was found between LOH at 11q13 and positive node status. LOH at the 11q22-qter region correlated with a high S-phase fraction. A significant association was found between LOH at 11p15 and chromosome regions 17q21 (the BRCA1 region) and 3p.

Images

Figure 1

Figure 4

Full text

Brish
JmalS
d
Cancer
(1995)
72,
696-701
9
©
1995
Stockton
Press
AM
rghts
reserved
0007-0920/95
$12.00
Loss
of
heterozygosity
at
chromosome
11
in
breast
cancer:
association
of
prognostic
factors
with
genetic
alterations
J
Gudmundsson,
RB
Barkardottir,
G
Eiriksdottir,
T
Baldursson,
A
Arason,
V
Egilsson
and
S
Ingvarsson
Department
of
Pathologyv
Division
of
Cell
Biology,
University
and
National
Hospital
of
Iceland,
Box
1465
IS-121
Revkjavik,
Iceland.
S_mnuary
We
examined
DNA
from
116
female
and
four
male
breast
cancer
patients
for
loss
of
heterozygosity
(LOH).
DNA
was
analysed
by
polymerase
chain
reaction
using
ten
microsatellite
markers
on
chromosome
11.
Three
distinct
regions
of
LOH
were
identified:
lIpl5.5,
11q13
and
lq22
-qter
with
a
LOH
frequency
of
19,
23
and
37-43%
respectively.
The
marker
DllS%9
showing
the
highest
frequency
of
LOH
(43%)
is
located
at
the
II
q24.1
-q25
region.
No
previous
molecular
genetic
studies
have
shown
frequent
LOH
at
the
region
telomeric
to
q23
on
chromosome
11.
Southern
analysis
revealed
that
LOH
at
1q13
was
due
to
amplification,
whereas
LOH
at
Iq22-qter
was
due
to
deletion.
LOH
at
llpl5.5
was
associated
with
paucity
of
hormone
receptor
proteins,
high
S-phase
and
positive
node
status.
An
association
was
found
between
LOH
at
I1q13
and
positive
node
status.
LOH
at
the
1q22
-qter
region
correlated
with
a
high
S-phase
fraction.
A
significant
association
was
found
between
LOH
at
llpl5
and
chromosome
regions
17q21
(the
BRCAI
region)
and
3p.
Keywords
breast
cancer;
chromosome
11;
loss
of
heterozygosity
The
majority
(approximately
90%)
of
breast
cancer
cases
are
considered
to
be
sporadic.
Multiple
genetic
alterations
accumulating
in
cells
result
in
alterations
of
normal
growth
control.
Characterisation
of
the
genes
that
play
a
role
in
this
tumongenic
process
is
a
necessary
step
towards
understand-
ing
it.
Mapping
the
chromosomal
regions
that
are
altered
in
breast
cancer
cells
has
proven
to
be
a
powerful
way
of
locating
these
genes.
Deletion
and
gene
activation
are
known
to
be
the
most
frequent
genetic
changes
in
breast
cancer
cells.
Chromosomal
translocations
may
also
be
involved
in
the
development
of
breast cancer,
as
suggested
by
Lindblom
et
al.
(1994),
who
showed
that
the
constitutional
llq;22q
trans-
location
predisposes
to
breast
cancer.
Chromosomal
regions
that
are
known
to
be
amplified
in
breast
cancer
are
8q
(Escot
et
al.,
1986),
1
1q13
(Varley
et
al.,
1988),
17q
(Yokota
et
al.,
1986)
and
20q
(Kalliomemi
et
al.,
1994).
Regions
with
fre-
quent
LOH
in
breast
cancer
are
lp
(Genuardi
et
al.,
1989),
lq
(Chen
et
al..
1989),
3p
(Eiriksdottir
et
al.,
1995),
6q
(Devilee
et
al.,
1991),
I
lpl5
(Ali
et
al.,
1987),
13q
(Lundberg
et
al..
1987),
16q
(Sato
et
al.,
1990),
17p
(Mackay
et
al.,
1988),
17q
and
18q
(Cropp
et
al.,
1990).
Chromosome
11
has
been
shown
to
possess
the
Wilms'
tumour
1
gene
(WTI)
on
the
p-arm
(Madden
et
al.,
1991).
The
MEN-I
locus
has
been
mapped
to
the
1lql3
region
(Larsson
et
al.,
1988)
and
the
ataxia
telangiectasia
(AT)
genes
have
been
mapped
to
the
1
lq22-23
region
(Gatti
et
al.,
1988).
Epidemiological
studies
suggest
that
heterozygous
AT
carriers
may
be
predisposed
to
cancer
(Swift
et
al.,
1991).
The
relative
risk
for
breast
cancer
has
been
estimated
to
be
5-fold
greater
in
women
carrying
the
AT
gene(s)
than
in
the
normal
population.
Cytological
and
LOH
studies
have
des-
cribed
aberrations
on
chromosome
1
lq22-q23
in
breast
cancer
(Ferti-Passantonopoulo
et
al.,
1991;
Carter
et
al.,
1994).
Studies
on
cancer
cell
lines
(e.g.
MCF-7)
have
shown
that
chromosome
11
suppresses
tumorigenicity
when
injected
into
cells
lacking
a
normal
chromosome
11
(Negrini
et
al.,
1994).
The
long
arm
of
chromosome
11
suppresses tumorigenicity
of
HeLa
cells
(Misra
and
Srivatsan,
1989),
suggesting
a
tumour-suppressor
gene
on
llq.
In
this
study
we
have
used
a
panel
of
polymorphic
mic-
Correspondence:
J
Gudmundsson
Received
30
November
1994;
revised
30
March
1995;
accepted
13
April
1995
rosatellite
markers
to
identify
and
investigate
regions
show-
ing
aberration
on
chromosome
11.
Materials
awd
methods
Samnples
Primary
breast
carcinoma
tissue
was
obtained
on
the
day
of
surgery,
immediately
frozen,
and
stored
at
-
80C.
Peripheral
blood
leucocytes
were
the
source
of
normal
DNA.
Salting
out
procedure
(Miller
et
al.,
1988)
and
phenol
extraction
methods
were
used
to
obtain
DNA
from
whole
blood
and
tumour
samples
respectively.
The
ratio
of
tumour
vs
normal
cells
in
the
samples
was
evaluated
by
histological
examina-
tion.
Tumours
with
scores
of
tumour
cells
<55%
were
ex-
cluded
from
the
study.
The
choice
of
cut-off
level
for
tumour
cell
fraction
in
the
samples
was
based
upon
results
from
our
studies
(A
Arason,
unpublished
results)
and
by
Gruis
et
al.
(1993),
who
demonstrated
by
titration
experiments
that
LOH
can
be
detected
in
samples
with
as
much
as
60%
normal
DNA
contamination.
PCR
analysis
of
microsatellites
Microsatellite
markers
used
for
LOH
analysis
of
chromosome
11
are
listed
in
Table
I.
The
microsatellite
markers
used
for
chromosome
17
were:
TP53
for
the
17pl3
region
(the
p53
gene)
and
THRA,
D17S800,
D17S855
and
D17S579
for
the
17q21
region
(containing
the
BRCAI
gene)
(T
Baldursson
et
al.,
manuscript
in
preparation).
The
markers
used
for
chromosome
3p
are:
D3S726,
D3S1211,
RIK,
PH3H2,
D3S1029, D3S1076,
D3S1067,
D3S1233,
D3S1217,
D3S1210
and
D3S1101
(Eiriksdottir
et
al.,
1995).
PCR
was
done
with
50
ng
of
genomic
DNA
in
25
iLl
volumes
using
DynaZyme
DNA
polymerase
(Finnzymes
Oy)
at
0.5
units
per
reaction
and
the
buffer
supplied
with
the
polymerase.
Primers
were
labelled
with
[y-ATP-32P]
(Amer-
sham)
using
T4-polynucleotide
kinase
(Amersham).
Samples
were
subjected
to
35
cycles
of
amplification,
consisting
of
50
s
at
94°C,
40
s
at
55'C
and
40
s
at
72°C,
followed
by
final
extension
for
10
min
at
72°C.
PCR
products
were
separated
on
6.5%
acrylamide
sequencing
gels
and
exposed
for
visualisation
by
autoradiography
on
Dupont
Cronex-4
film.
Autoradiograms
were
inspected
visually.
Any
absence
or

ciwo
se
11
in
bha
canoer
J
Gudmundsson
et
al
m
697
Table
I
Information
about
microsatellite
markers
and
results
from
PCR
analysis
NVo.
of
pairs
No.
of
pairs
Distance
No.
of
pairs
Informative
with
allelic
(cM)
Locus
Locationa
Tipe
Referemce
examined
(%0)
imbalance
(%)
Gyapay
et
al.
(1994)
Gyapay
et
al.
(1994)
Gyapay
et
al.
(1994)
Polymeropolus
et
al.
(1989)
Brown
et
al.
(1990)
Litt
et
al.
(1989)
Gyapay
et
al.
(1994)
Gyapay
et
al.
(1994)
Gyapay
et
al.
(1994)
Gyapay
et
al.
(1994)
'According
to
the
Genome
data
base.
significant
decrease
in
the
intensity
of
one
allele
relative
to
the
other
was
considered
LOH
(see
Figure
1).
Southern
blot
analjsis
Aliquots
of
3-10ig
of
genomic
DNA
were
digested
over-
night
with
a
suitable
restnrction
enzyme
according
to
the
manufacturer's
procedure,
loaded
onto
0.8%
agarose
gels
and
electrophoresed
at
35-55
V
overnight,
transferred
to
a
Hybond
nylon
membrane
(Amersham)
according
to
standard
protocols
(Sambrook
et
al.,
1989).
The
RFLP
probes
used
in
this
study
were
SS6,
FGF3
(llql3.3);
STMY1,
MMP3
(1lq22.3);
MCT
128.1,
D1IS144
(1lq22.3-q23);
and
phi
2-
11-2.2,
Dl1S34
(1lq23-qter).
A
probe
for
the
MOS
gene,
HM2A
(8ql
1),
was
used
as
an
internal
control
for
a
normal
copy
number
of
alleles.
The
probes
were
labelled
using
a
Megaprime
DNA
labelling
kit
(Amersham).
Hybridisation
was
carried
out
overnight
at
65°C
and
the
filters
were
washed
at
65°C,
2
x
15min
with
2
x
SSC,
and
2
x
15min
with
2
x
SSC/0.
1%
sodium
dodecyl
sulphate
(SDS),
followed
by
a
2
x
25
min
stringency
wash
with
0.2
x
SSC.
1175
1187
984
1151
N
T
N
T
N
T
N
T
147
bp
145
bp
-
141
bp
139
bp
-
135bp
Figre
I
Results
from
PCR
analysis
with
microsatellite
marker
DlIS912
of
four
samples,
all
exhibiting
LOH.
N.
normal
DNA;
T.
tumour
DNA.
Statistical
methods
Chi-square
analysis
was
used
to
test
for
association
between
the
genetic
events
examined
and
the
clinicopathological
parameters
of
the
patients.
The
chnicopathological
charac-
teristics
were
categorised
as
follows:
oestrogen
receptors
(ER):
negative
(S
10
fmol
mg-1
protein)
or
positive
(>
10
fmol
mg-'
protein);
progesterone
receptors
(PgR):
negative
(
S
25
fmol
mg-'
protein)
or
positive
(>
25
fmol
mg-'
protein);
histological
type:
ductal
or
lobular;
lymph
node
status:
negative
or
positive;
tumour
size:
2
cm
or
>2
cm;
age:
<50
years
or
>
50
years;
S-phase
fraction,
<
7%
or
>
7%;
and
ploidy:
diploid
or
non-diploid.
All
patients
were
checked
for
family
history
of
breast
cancer.
The
family
coefficient
was
categonsed
as
follows:
those
who
had
at
least
one
first-degree
or
second-degree
relative
with
breast
cancer
or
those
who
had
no
known
relative
with
breast
cancer.
Others
were
not
included
in
the
calcula-
tions.
The
chi-square
test
was
also
used
to
assess
the
relationship
between
LOH
at
chromosome
I
lq
and
LOH
at
chromosomes
17
and
3p.
Results
PCR
analysis
We
screened
116
female
and
four
male
primary
breast
tumours
for
LOH
with
ten
polymorphic
markers
on
chromosome
11,
seven
of
them
located
at
the
q-arm
and
three
located
at
the
p-arm.
Fifty-three
(45%)
of
the
116
tumours
showed
LOH
with
at
least
one
of
the
ten
markers.
Seven
tumours
(6%)
had
LOH
only
at
Ilp,
33
tumours
(28%)
had
LOH
only
at
llq,
and
13
tumours
(11%)
had
LOH
at
both
1
lq
and
lp,
five
of
which
had
lost
the
whole
chromosome.
Three
of
the
four
male
breast
tumours
had
LOH
at
1
lq
and
two
of
them
also
at
lp.
The
frequency
of
LOH
for
each
of
the
ten
markers
ranged
from
7-43%,
being
highest
at
the
most
telomeric
part
of
chromosome
llq
at
43%
(DIIS969)
and
at
the
llq22
region
at
37%
(DIIS35
and
DlIS927).
The
lowest
frequency
of
LOH
was
observed
at
the
most
proximal
region
on
lip
(DlIS903).
Figure
2
shows
PCR
results
from
nine
of
the
tumours
analysed.
Tumours
561,
593
and
1200
are
examples
where
only
the
most
telomeric
region
(1
lq24-qter)
was
lost
(marker
Dl1IS925
or
more
distal
markers).
Tumours
842,
1030
and
567
only
showed
aberrations
proximal
to
the
DlIS925
locus
at
chromosome
llq,
and
the
LOH
in
these
tumours
included
the
1
1q22
region.
Tumours
1071,
1201
and
1121
only
had
LOH
at
the
lIp
region.
Southern
analysis
Southern
analysis
was
carried
out
to
distinguish
between
amplification
and
deletions
at
chromosome
1llq.
DNA
from
17
of
the
46
tumours
that
showed
LOH
at
lq
with
mic-
rosatellites
was
available
for
Southern
hybridisation.
We
pro-
bed
for
four
different
loci:
one
at
1q13
(FGF3)
and
three
at
llq22-qter
(MMP3,
DlIS144
and
DIIS34).
We
were
not
Dl
lS922
Dl
lS907
D
1
S903
FGF3([NT2)
Dl
lS527
D1I
S35
D
1I
S927
Dl
I
S925
DII
S912
DIIS%9
lIp15.5
llpl3-ql3
lIpl3-ql3
I
lql3.3
I
Iql3.5
1
1q22
I
1q22
I
lq22.3
-q24
I
lq25
1
q24.1
-q25
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
AC
repeat
103
106
95
113
110
97
87
105
101
79
90
(87)
88
(83)
74
(78)
84
(74)
103
(94)
76
(78)
75
(86)
87
(83)
88
(87)
58
(73)
17
(19)
11(12)
5
(7)
19
(23)
26
(25)
28
(37)
28
(37)
27
(31)
31
(35)
25
(43)

Chromosome
11
in
bren
cancer
9
J
Gudmundsson
et
al
698
able
to
detect
a
single
amplification
with
the
three
probes
that
we
used
at
1
1q22-qter.
However.
Southern
analyses
proved
that
eight
of
the
ten
tumours,
available
for
Southern
analyses
and
showing
LOH
with
the
FGF3
microsatellite
marker,
were
amplified.
The
amplification
in
each
sample
was
estimated
by
titration
to
range
from
2-
to
about
20-fold
(data
not
shown).
Figure
3
shows
the
pattern
of
LOH
detected
by
PCR
vs
the
results
of
the
Southern
analysis
for
the
eight
tumours
having
amplification
at
1
Iqi
3.3
and
the
two
tumours
that
had
no
amplification.
Tumours
1120,
795,
1110.
549,
975.