Download Genetic Changes in Primary and Recurrent Prostate Cancer by

[CANCERRESEARCH
55, 342-347,January15,19951
Genetic Changes in Primary and Recurrent Prostate Cancer by Comparative
Genomic Hybridization'
Tapio Visakorpi,2 Anne H. Kallioniemi, Ann-Christine Syvänen,Eija R. Hyytinen, Ritva Karhu, Teuvo Tammela,
Jorma
J- Isola, and Offi-P.
Kalliomemi
Laboratory of Cancer Genetics, Department of Clinical Chemistry fT. V., A. H. K., E. R. H., R. K., 0-P. K.J, and Division of Urology, Department of Surgery IT. T.J, Tampere
University Hospital, P. 0. Box 2000. FIN-33521 Tampere; Department of Human Molecular Genetics, National Public Health Institute, Helsinki (A-C. S.); and Department of
Biomedical Sciences, University of Tampere. Tampere If. J. I.]. Finland
ABSTRACT
of the known TSGs and oncogenes in prostate cancer development are
also poorly known. For example, ras oncogene, which is commonly
Genetic changes leading to the development of prostate cancer and
factors that underlie the clinical progression of the disease are poorly
characterized.
Here, we used comparative
genomic hybridization
(CGH)
to screen for DNA sequence copy number changes along all chromosomes
in 31 primary and 9 recurrent uncultured prostate carcinomas. The aim
of the study was to identify those chromosome regions that contain genes
important for the development of prostate cancer and to identify genetic
markers of tumor progression. CGH analysis indicated that 74% of
primary prostate carcinomas showed DNA sequence copy number
changes. Losses were 5 times more common than gains and most often
involved
Sp (32%),
13q (32%),
6q (22%),
16q (19%),
lSq (19%),
and 9p
(16%). Allelic loss studies with 5 polymorphic microsatellite markers for
4 different chromosomes were done from 13 samples and showed a 76%
concordance with CGH results. In local recurrences that developed during
endocrine therapy, there were significantly more gains (P < 0.001) and
losses (P < 0.05) of DNA sequences than in primary tumors, with gains of
Sq (found in 89% of recurrences versus 6% of primary tumors), X (56%
versus 0%), and 7 (56% versus 10%), as well as loss of Sp (78% versus
32%), being particularly
often involved. In conclusion,
our CGH results
indicate that losses of several chromosomal regions are common genetic
changes in primary tumors, suggesting that deletional inactivation of
putative
tumor
underlie
development
genetic
suppressor
changes
genes
in these chromosomal
of prostate
cancer.
seen in recurrent
tumors
sites is likely
to
the pattern
of
Furthermore,
with the frequent
gains of 7, Sq,
and X suggests that the progression of prostate cancer and development of
hormone-independent
chromosome
prostate
growth
aberrations
may
may
have
have
a distinct
diagnostic
genetic
utility
basis.
These
as markers
of
Genetic changes underlying the development and progression of
prostate cancer are poorly known. Classical cytogenetic studies are
to carry out in prostate
cancer because
13q, 16q, and 18q (2—7). However,
LOH studies
to extensive analyses of a single chromosome
with only
the vast majority
it possible
to survey
the entire genome
1—3probes/chromosome
of the genome
unexamined.
are typically
fluorescence intensity ratio measurements along all chromosomes,
indicate those regions of the genome that were either over- or under
represented in the tumor genome. The utility of CGH is based on the
concept that regions with increased copy number reveal chromosome
sites that may contain dominant oncogenes, whereas regions with
decreased copy number may be putative TSG loci (11). Thus, in a
single hybridization, CGH allows screening of all chromosomal sites
that are likely to contain genes with an important role in tumor
development.
We took advantage of the potential of CGH to screen for losses and
gains of DNA sequences in 31 primary uncultured prostate carcino
mas. The aim was to identify
those
chromosomal
regions
that are
often involved in copy number aberrations and may thus contain
genes implicated in the development of prostate cancer. In the second
genetic
aberrations
in the primary
tumors
were
AND METHODS
2. That of recurrent tumors was: grade II, 4; and grade Ill, 5. The primary
tumor specimens were prostatectomy specimens (20 cases), transurethral re
limited
or analysis of all
section specimens (7 cases), or Tm-Cut needle biopsy specimens (4 cases), all
arm, thus leaving
The specific
taken prior to the administration of any hormonal therapy, whereas recurrent
prostate carcinomas were all transurethral resection specimens taken from
roles of any
I Supported by the Reino Lahtikari Foundation, Paulo's Foundation, the Sigrid Juselius
Foundation, The Academy of Finland, the Finnish Cancer Society, and the Finnish Cancer
Institute.
for reprints
should
be addressed,
at Laboratory
of Cancer
Genet
irs, National Center for Human Genome Research, NIH, 9000 Rockville Pike, Building
49, Room 4C20, Bethesda, MD 20892.
used
are:
LOH,
loss
of
heterozygosity;
TSG,
tumor
therapy (orchiectomy
6 cases, luteinizing
hormone-releasing hormone agonist 2 cases, estrogen 1 case). The recurrent
tumors came from patients with symptoms of urethral obstruction, indicating
18 U.S.C. Section 1734 solely to indicate this fact.
abbreviations
be
by
distribution (19) ofprimary tumors was: grade I, 10; grade II, 19; and grade III,
The costs of publication of this article were defrayed in part by the payment of page
3 The
differences
as evidenced
The material consisted of 31 primary and 9 recurrent uncultured prostate
carcinomas. Six benign prostate hyperplasia samples were also evaluated. The
ThM stagedistribution(18) of theprimaryprostatecarcinomaswas:T1N@M@,
1; T2N@JM@J,
14; T2N@M@,
2; T3N0MIJ,3; T3N@M@,
3; T4NXM@J,
2;
1;
T2N1M,J,1; T2NXM1,1; T3N@M1,2; and unknown, 1. The histological grade
charges. This article must therefore be hereby marked advertisement in accordance with
requests
in
for gains and losses of
metaphase
chromosomes.
In such a hybridization,
tween the binding of the labeled DNA sequences,
patients who received only endocrine
whom
infrequently
DNA sequences (11—17).CGH is based on the simultaneous hybrid
ization of differentially labeled tumor and normal DNA to normal
MATERIALS
Received 8/12/94; accepted 11/10/94.
2 To
relatively
of the preferential
growth of nonmalignant cells. In many cases only a normal karyotype
has been found (1). The aberrations that have been reported most often
include deletions of 7q, lOq, and 8 as well as gains of chromosome 7.
Occasionally, double minute chromosomes have been reported (1).
LOH3 studies, which are thought to highlight chromosomal sites
harboring mutated tumor suppressor genes, have implicated 8p, lOq,
chromosomes
makes
is involved
compared with those detected in 9 recurrent prostate tumors. The aim
of this part of the study was to identify genetic changes that underlie
clinical tumor progression.
INTRODUCTION
@
in human malignancies,
cancer (8, 9). Furthermore, factors that determine the prog
patients with prostate cancer are poorly known, and genetic
of tumor progression are urgently required (10).
is a newly developed molecular cytogenetic method that
part of the study,
cancer progression.
very difficult
affected
prostate
nosis of
markers
CGH
local progression of the disease despite ongoing therapy.
Five-@xmsections were cut from freshly frozen tumor blocks embedded in
Tissue-Tek
(Miles, Inc., Diagnostic Division, Elkhart, IN) and stained with
hematoxylin and eosin to ensure the histological representativeness of the
samples. Some specimens were trimmed by cutting normal tissue away with a
scalpel. High-molecular-weight tumor DNA was isolated either from homog
enized tumor specimens or from 200-sam frozen sections of the tumors using
suppressor
standard protocols. DNA also was isolated from the peripheral blood of 13
gene; CGH, comparative genomic hybridization; ThM, tumor-nodes-metastasis.
342
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
GENETIC CHANGES IN PROSTATIC CARCINOMA
patients (for LOH studies) and from normal male donors (reference DNA for
CGH).
Analysis and Interpretation
of the LOH Studies. Aliquots (3 j@l)of the
fluorescent PCR products were denatured in 50% formamide containing blue
gated DNAS, as described previously (1 1, 20). Briefly, DNA samples from the
dextran and analyzed on 6% denaturing polyacrylamide gels using an auto
matic DNA sequencer (ALF; Pharmacia Biotech AB). The relative quantities
of the PCR products were determined with the aid of ALF DNA Fragment
Manager program V1.1 (Pharmacia Biotech AB). When the ratio between the
tumors were labeled with FITC-dUTP (DuPont, Boston, MA), and normal
two alleles amplified from a tumor sample differed significantly
Comparative Genomic Hybridization
Hybridization.
CGH was performed
using directly
fluorochrome-conju
(2—5-fold)
male DNA was labeled with Texas red-dUTP (DuPont) using nick translation.
from that obtained in the leukocyte sample, it was interpreted as a sign of LOH.
Labeled tumor and normal DNAS (400 ng each) together with 10 g.@g
of
unlabeled Cot-i DNA (Gibco BRL, Gaithersburg, MD) in 10 @l
of hybridiza
tion mixture [50% formamide, 10% dextran sulfate, 2X SSC (1X SSC is 0.15
Statistical
M NaCl-0.015
M sodium
citrate,
pH
7)]
were
denatured
at 70°C
for
5 mm
and
applied on normal lymphocyte metaphase preparations. Prior to hybridization,
Analysis
The statistical significance between the primary and recurrent tumors in the
total number of genetic aberrations and the frequencies of selected changes
amide solution (70% formamide, 2X SSC, pH 7) and dehydrated in a series of
was calculated with the nonparametric Kruskal-Wallis test and 2-tailed
Fisher's exact test, respectively. The ic statistic was used to analyze the level
70, 85, and 100% ethanols;
of agreement between CGH and LOH results.
the metaphase preparations
were denatured at 72—74°C
for 3 mm in a form
this was followed
by proteinase
K (0.1 mg/ml in
20 m@iTris-HCI, 2 mr@iCaC12,pH 7.5) treatment at room temperature and
dehydration once again as described above. The hybridization was done at
37°Cfor 48 h. After hybridization, the slides were washed three times in 50%
formamide/2X SSC (pH 7), twice in 2X SSC, and once in 0.1 X SSC at 45°C
followed by 2X SSC and 0.1 MNaH2PO4-0.1MNa2HPO4-0.1%NP4O(jH 8),
and distilled water at room temperature for 10 mm each. After air drying, the
slides were counterstained
with 4',6-diamidino-2-phenylindole,
0.1 @xWml,in
an antifade solution.
Digital Image Analysis. Three single-color images (matching 4',6-dia
midino-2-phenylindole,
FITC, and Texas red fluorescence) were collected
from each metaphase spread using a Nikon SA epifluorescence microscope
(Nikon Corp., Tokyo, Japan) and a Xillix charge-coupled-devicecamera
(Xilli.x Technologies
workstation
Corp., Vancouver, BC, Canada) interfaced to a Sun LX
(Sun Microsystems
to six three-color
Computer
Corp., Mountain
digital images were collected
View,
CA). Four
from each hybridization.
Relative DNA sequence copy number changes were detected by analyzing the
hybridization intensities of tumor and normal DNAS along the length of all
chromosomes in the metaphase spreads, as described earlier (20). The absolute
fluorescence intensities were normalized so that the average green:red ratio of
all chromosome
objects in each metaphase was 1.0. The fmal results were
plotted as a series of green:red ratio profiles and corresponding SDs for each
human chromosome from pter to qter.
Interpretation
of CGH results followed
previously
described
protocols
(20).
Hybridizations of FITC-labeled normal female DNA with Texas red-labeled
normal male DNA were used as negative controls. The mean green:red ratio
and the corresponding SD for all autosomes remained between 0.9 and 1.1 in
these control hybridizations.
Chromosomal
regions where the mean ratio and
the corresponding SD were less than 0.85 were therefore considered lost, and
regions where the mean and the corresponding SD were greater than 1.15
gained in the tumor genome. The entire Y chromosome was excluded from
analysis. Hybridizations of DNA from the MCF-7 breast cancer cell line
against normal female DNA were used as additional positive controls in each
hybridization batch.
Loss of Heterozygosity
Oligonucleotides. Primers for amplification of the microsatellite loci
D6S283
(6q16.3—q21),
q22), and D16S413
D8S265
(8p23.l),
(16q) were synthesized
D8S282
(8p22),
according
D13S153
(13q14—
to the sequences
given
in the Genethon catalogue. The 5' ends of the upstream (AC strand) D6S283,
D13S153, and D16S422 primers and the downstream (OT strand) D8S265 and
D8S282 primerswere fluorescencelabeledduring the synthesis using the
FluorePrime reagent (Pharmacia Biotech AB, Uppsala, Sweden).
PCR. Twenty ng of DNA isolated from patients' leukocytes or tumor
samples were amplified separately with each set of primers. The PCR mixtures
contained 50 pmol of both primers, the four deoxynucleotide triphosphates at
0.2 mr@tconcentration, and 1.25 units of DynaZyme DNA polymerase
(Finnzymes Oy, Espoo, Finland) in 50 @tlof buffer supplied with the enzyme.
RESULTS
Overview
perplasias
of Genetic Changes.
showed
None of the benign prostate hy
any gains or losses of DNA sequences
by CGH. Six
(19%) of the primary prostate cancers showed relative DNA sequence
gains, and 23 (74%) showed losses at 1 or more chromosomal sites
(Fig. 1). Eight tumors (26%) had no copy number alterations. On
average, there were 2.9 (range, 0—12)aberrations per primary tumor:
0.5 gain (range, 0—4)and 2.4 deletions (range, 0—9).Fig. 2 shows an
example of green:red fluorescence ratio profiles of a primary prostate
carcinoma analyzed by CGH.
All recurrent prostate cancers showed both relative gains and losses
of DNA sequences (Fig. 3). The total numbers of aberrations per
tumor (mean, 7.8; range, 4—15),as well as gains (mean, 2.2; range,
1—4)and losses (mean, 5.6; range, 3—12)were significantly higher in
the recurrences than in primary tumors (Fig. 4A). Significance values
for these differences were P < 0.01 for all aberrations, P < 0.001 for
gains, and P < 0.05 for losses.
Losses and Gains. Chromosome arms that were lost most fre
quently in primary prostate cancers were 8p (32% of the cases), 13q
(32%), 6q (22%), 16q (19%), 18q (19%), and 9p (16%). The minimal
overlapping regions of loss in each chromosome were 8p12—pter,
13q21—31,ócen—q21,l6cen—q23, 18q22—qter, and 9p23—pter.Gain
of the entire long arm of chromosome 8 was found in two (6%) cases.
Chromosome arms that were lost most frequently in recurrent
prostate cancers were 8p (78%), 13q (56%), 16q (56%), 6q (44%), and
5q (44%), and the minimal regions 8p21—pter, l3cen—q21, 16q22—
qter, 6q13—q21,and 5q14—q23. Gain of 8q was seen in 8 of 9 (89%)
recurrent prostate cancers and usually affected the entire arm. Gains of
chromosome 7 (minimal common region 7pl3) and X (Xpll—q13 and
Xq23—qter) were both seen in 56% of cases. Fig. 4, B and C,
illustrates the main differences in the frequencies of genetic changes
in primary and recurrent prostate cancer.
Taking primary and recurrent tumors together, 10 cases had a gain
at 8q. In one tumor, gain was limited to 8q24, while nine were gains
of the entire long arm. Seven of these tumors also had loss of 8p.
Comparison between CGH and LOH Results. Detection of
losses by CGH was compared with data from LOH studies using five
polymorphic microsatellite markers (1)65283, D8S265, D8S282,
D13S153, and D16S413) for four different chromosome arms. In
total, 37 comparisons were done in 5 loci. A 76% concordance was
found between CGH and allelic loss results (Table 1). The ic coeffi
cient was 0.507. In 13% of cases, LOH was found without loss by
CGH; in another 11% of cases, CGH showed loss of DNA sequences,
but no LOH was detected.
343
The PCR was initiated by heating the samples at 95°Cfor 3 mm, followed by
addition of enzyme at 80°C.Twenty-five PCR cycles of 1 min at 95°C,1 mm
at 58°C(markers D6S283, D8S282, and D16S422) or at 54°C(markers
D8S265 and D13S153),
and 1 min at 72°Cwere carried out in a programmable
heat block (FTC 100; Mi Research, Inc., Watertown, MA).
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
@
I
@R.
__
GENETIC CHANGES IN PROSTATIC CARCINOMA
I
@
i
I
I
i@
Fig.1. Summaryof all gainsand lossesof
@
i@k 1@ I P I@I I
DNA sequences observed in 31 primary pros
tate carcinomas by CGH. Gains are shown on
8
6
I
9
theleftsideof thechromosomeideogramsand
losses on the right. Chromosome Y was cx
cluded from analysis.
4
16
17
19
@t
20
18
1
14
I@I1
•
[email protected]. .
I@@1
11
21
@I
22
Y
15
i@_
.iiiiii
I. UI1@I@
12
5
@lI I II Ii @iII
13
@
11
2
@
@
10
.
@l
I
11
. @..
t@•
4@:.:::;:..::.:::::::::.:.::.@
5@:.@...:;;..:::;:::::.::::@::::::::
@
L@I@@IJ
I
I
1
1 I@I@
N
I
@@JI
@ 6b@
@:@:
.;..::
7fT::T..!....:@r::.:::I.r..
. :@
.
Et@
I
1
@I@II 1
1
iUi
II
I
•I@
1
ii
@ 9HH
‘
@I10@H. ii -@.
@..
12@@
@
L
@I
@1 II
@i@i@ii
Ill_I_Il_I
•i
iii
Fig. 2. Mean grcen:red ratio profiles for all
chromosomes (except Y) from pter to qtcr oh
tamed from CGH analysis of a prhnary pros
tate cancer. ----,
the baseline value (1.0)
representing the mean grecn:red ratio for the
entire sample;
ratios 03 and 1.5. changes
in the grecn:redratioprofileindicatelossesat
6cen—q22,8p, 11q14—q23, 16q, and 22q; and
gains at Sp, 7p, 7q32—qter,
and 19p. Bars, SD.
@ 13H•iI.iI ]5@i@J@9
@TIi•@@•i
@I__uI_I1
ioH—@ 17L
@
@
@
t-I_--@II_I1
@I•I
I 18@
LUE .1 ••@@I
tI.ft
I
19 I
I@ I
.
21@'@22@
I 20 ___
.
________
V __
•@FIJ
@i i
@i.
i
_________
344
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
OENFI1C cHANOE5 IN PROSTATIC CARCINOMA
@
@II
@II
‘1@k
uli ‘i―
!@‘
I@
Fig. 3. Summary of all gains and losses of
Il
DNAsequencesobservedin 9 recurrentpros
tate carcinomas by CGH. Gains arc shown on
theleftsideof thechromosomeideogramsand
losses on the right. Chromosome Y was cx
cluded from analysis
@
ho'
I Ioii@
Ii 1@
II
19
16
13
14
17
18
20
21
@I
22
Y
iO@
x
15
be common mechanisms of LOH (27). Furthermore, CGH can detect
only physical losses affecting regions larger than 10 mega-base pairs.
In four cases (11%), a loss was clearly detected by CGH, but no LOH
was found. The reason for this inconsistency remains unknown, but it
is possible that in these cases the microsatellite markers, which have
been only genetically mapped, were localized outside the region of
loss.
The majority of prostate cancer patients have disease that is no
longer curable at the time of diagnosis. However, approximately 70%
of these patients will respond to androgen ablation therapy. Although
endocrine treatment is initially effective, the cancer cells later become
androgen independent and the disease progresses despite the therapy
(28). There are very few data on the genetic events that determine the
malignant potential of prostate cancer and its response to endocrine
therapy. Knowledge of the mechanisms underlying hormone-indepen
dent growth are important because at present, there are no effective
therapies for hormone-resistant prostate cancers. We thought that the
analysis of genetic changes in local recurrences that arise during
hormonal therapy would highlight those aberrations causing
aggressive clinical behavior.
The nine cases of androgen-resistant recurrent prostate carcinomas
came from patients who had originally received endocrine therapy but
developed clinical signs oflocal tumor progression. We found that the
total number of genetic changes per tumor was almost 3 times higher
in recurrences than in primary tumors. Whereas gains and amplifica
tions were uncommon in primary tumors, all recurrences showed
gains of at least one chromosomal region. In particular, gain of 8q,
either alone or in association with 8p loss, was found 8 of 9 recurrent
prostate cancers. This suggests that the long arm of chromosome 8
may harbor a gene(s) involved in the progression of prostate cancer
and its evolution toward hormone independence. The myc oncogene is
located at 8q24 and has been shown to be overexpressed in poorly
differentiated prostate cancers (29, 30). Because the entire long arm of
chromosome 8 was usually present at an increased copy number, it is
likely that other genes instead of or in addition to myc are involved.
Gains of chromosomes 7 and X also were found in more than
one-half of the recurrent prostate carcinomas but very infrequently in
primary tumors. In two tumors, the entire chromosome 7 was gained,
while three tumors showed partial gains with the minimal overlapping
region at ‘7pl3.Fluorescence in situ hybridization has shown that
DISCUSSION
This study represents a genome-wide survey of DNA sequence
copy number changes in prostate cancer using CGH. We found that
74% of the primary carcinomas showed gains and/or losses of DNA
sequences, which is a significantly higher number than seen by
cytogenetic studies. This reflects the power of CGH in revealing
aberrations across the genome in uncultured cells. In contrast, none of
the benign prostatic hyperplasias showed any genetic alterations by
CGH. In primary tumors, losses predominated over gains with a ratio
of 5:1. The complete absence of high-level amplification and the low
overall frequency of gains in primary prostate cancer are striking as
compared to the extensive amplifications seen, for example, in breast
cancer (13). This suggests that inactivation of putative recessive TSGs
in several chromosomal sites is especially important in prostate cancer
development.
The most commonly lost chromosomal regions in primary prostate
cancer were 8p, 13q, 6q, 16q, 18q, and 9p. Several studies have shown
loss of heterozygosities at 8p, 13q, 16q, and 18q in prostate cancer
(2—7),but 6q and 9p losses have not been reported previously in
prostate cancer. In one previous study, possible LOH at 6q and 9p was
studied but not found. However, only a single probe per chromosome
arm was used (7). According to CGH, the critical region at 6q was
6cen—q21,indicating that this region may harbor a TSG important in
the development of prostate cancer. Previously, LOH of the same
region was reported in melanoma and in ovarian and breast cancers
(21—23).Furthermore, transfection of a normal human chromosome 6
suppresses the tumorigenicity of both breast and melanoma cancer
cell lines (24, 25). Taken together, these results support the presence
of a TSG in 6q that may be involved in several tumor types, including
prostate cancer. The putative TSG at 9p also remains unknown.
Recently, a new TSG, MTS1, was identified at 9p21 (26). Whether
this gene is also involved in prostate cancer remains to be determined.
According to CGH, the minimal deleted region was 9p23—pter,
which
suggests that the target TSG for 9p loss in prostate cancer may reside
distal to MTSJ.
The overall concordance between CGH and LOH results was 76%.
In five cases (13%), LOH was detected but there was no loss by CGH.
CGH is sensitive only to physical losses of DNA sequences and not to
losses of specific alleles. Mitotic recombinations and losses followed
by duplication of the remaining allele cannot be seen by CGH but may
345
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
GENETIC CHANGES IN PROSTATIC CARCINOMA
@
trisomy 7 is common in clinically high-stage prostate cancer as well
as in progression specimens (3 1). Recently, aneusomy of chromosome
7 was shown
to be associated
with poor prognosis
in prostate
differentchromosome
Table I Comparison oflosses found by CGH with WH measurement of 4
arms
D65283; D16S422)in
8q. D85265, D8S282; 13q, D135153; 16q,
cancersLOHNo
13 prostate
cancer
(32).Chromosome
7 containsmanycandidate
genes,suchasEGFR,
TotalNoloss
PA!], RAF, and MDRJ, that may participate in the progression of
prostate cancer. Gains of chromosome X were also variable. It is
interesting that two recurrent tumors showed high-level amplification,
one at Xpl l—q13and another at Xq23—qter. According to the Ge
nome Data Base, the Xpl l—q13region contains many possible target
A
0
E
21Loss
16Total
**
I#1
6
0
*
C
0
.c
U
4
0
0
z
2
Total
B
U)
0
Gains
Losses
100
80
**
60
**
40
C
0
U
0
37
regions that may harbor important genes for prostate cancer tumori
genesis and progression. Losses found by CGH in primary tumors
involving 6q (minimal overlapping region, 6cen—q21)and 9p (9p23—
pter) suggest two new regions that may contain prostate cancer TSGs
in addition to the previously reported TSG loci 8p, 13q, 16q, and l8q.
Gains of DNA sequences at 7 (7p13), 8q (8q24—qter),and X (Xpl 1—
q13 and Xq23—qter)appear important for prostate cancer progression.
Further studies with specific probes are required to narrow down the
critical regions in each chromosome and to identify the genes
0
a.
5
12
17
somal regions may underlie the progression of prostate cancer.
In conclusion, these CGH results highlight several chromosomal
E
0
DI
0
16
4
20
Loss
genes such as AR, ARAFI, ELKJ, IL2RG, PGKJ, PGKIPJ, PHKAJ,
TFE3, TIMPJ, ZNF2J, and ZNF8J. Gains involving the same region
have been found previously in about 35% of osteosarcomas by CGH,4
suggesting that this chromosomal region contains a dominantly acting
oncogene involved in several tumor types. The other amplified region
at X23—qter contains genes such as HPRT, LJCAM, MCF2, and
MPPJ as a possible target gene. Further studies with specific probes
to these two regions and individual candidate genes are in progress.
Losses in the recurrent tumors in general involved the same regions as
in primary tumors, but their overall frequency was higher. However,
the frequency of Sq losses was over 7 times higher in recurrent tumors
than in primary tumors. Adenomatous polyposis coli TSG is localized
to 5q21. LOH of the adenomatous polyposis coli region has been
found in 20-30% of advanced prostate carcinomas (5, 6). These results
indicate that an increased overall number of genetic changes and, specif
ically, gains and amplifications of certain chromosomes and chromo
10
8
loss
involved.
20
ACKNOWLEDGMENTS
0
7p
8q
We thank Sari Pennanen and Ritva Timonen for their technical assistance.
X
REFERENCES
C
1. Sandberg, A. A. Chromosomal abnormalities and related events in prostate cancer.
Hum. Pathol., 23: 368—380,1992.
100
2. Carter, B. S., Ewing, C. M., Ward, W. S., Treiger, B. F., Aalders, T. W., Schalken,
3. A., Epstein. J. I., and Isaacs, W. B. Alldic loss of chromosomes 16q and l0q in
U)
0
*
80
human prostate cancer. Proc. Nail. Aced. Sci. USA, 87: 8751—8755,1990.
3. Bergerheim, U. S. R., Kunimi, K., Collins, V. P., and Ekman, P. Deletion mapping of
E
chromosomes 8, 10, and 16 in human prostatic carcinoma. Genes Chromosomes &
60
Cancer, 3: 215—220,1991.
4. Bova, 0. S., Carter, B. S., Bussemakers, M. J. G., Emi, M., Fujiwara, Y., Kyprianou,
*
0
0
DI
0
C
0
U
4)
a.
N., Jacobs, S. C., Robinson, J. C., Epstein, J. I., Walsh, P. C., and lasses, W. B.
Homozygous deletion and frequent allelic loss of chromosome 8p22 loci in human
prostate cancer. Cancer Res., 53: 3869—3873,1993.
40
5. Phillips, S. M. A., Morton, D. 0., Lee, S. .1., Wallace, D. M. A., and Neoptelemos,
J. P. Loss of heterozygosityof the retinoblastomaand adenomatouspolyposis
susceptibility gene loci and in chromosomes lOp, lOq and 16q in human prostate
20@
cancer. Br. J. Urol., 73: 390—395,1994.
6. Brewster, S. F., Browne, S., and Brown, K. W. Somatic allelic loss at the DCC, APC.
nm23-HI and p53 tumor suppressor gene loci in human prostatic carcinoma. J. Urol.,
151: 1073—1077,1994.
5q
6q
8p
13q
7. Kunimi, K., Bergerheim, U. S. P., Larsson, I-L., Ekman, P.. and Collins, V. P.
Allelotyping of human prostatic adenocarcinoma. Genomics, II: 530—536,1991.
16q
Fig. 4. Comparison of genetic aberrations between primary (0) and recurrent ()
prostate cancer, showing mean number of genetic changes per tumor (A) as well as
frequency of gains (B) and losses (C) at selected chromosome arms. °,P < 0.05; **,
P < 0.01; °°°,
P < 0.001; Fisher's exact test.
8. Carter, B. S., Epstein, J. I., and Isaacs, W. B. Ras gene mutations in human prostate
cancer. Cancer Res., 50: 6830—6832, 1990.
4M.Tarkkanen,
personal
communication.
346
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
GENETICCHANGESIN PROSTATICCARCINOMA
9. Gummerlock, P. H., Poonamallee, U. R., Meyers, F. J., and de Vere White, R. W.
Activated ras alleles in human carcinoma of prostate are rare. Cancer Res., 51:
1632—1637,
1991.
10. Visakorpi, T., Kallioniemi, 0-P., Koivula, T., and Isola, J. Review of new prognostic
factors in prostatic carcinoma. Eur. Urol., 24: 438—449,1993.
11. Kallioniemi, A., Kallioniemi, 0-P., Sudar, D., Rutovitz, D., Gray, J. W., Waldman,
F., and Pinkel, D. Comparative genomic hybridization for molecular cytogenetic
analysis of solid tumors. Science (Washington DC), 258: 818—821,1992.
12. du Manoir, S., Speicher, M. R., Joos, S., SchrOck, E., Lichter, P., and Cremer, T.
Detection of complete and partial chromosome gains and losses by comparative
genomic in situ hybridization. Hum. Genet., 90: 590—610, 1993.
13. Kallioniemi, A., Kallioniemi, 0-P., Piper, J., Tanner, M., Stokke T., Chen, L., Smith,
H. S., Pinkel, D., Gray, J. W., and Waldman, F. M. Detection and mapping of
amplified DNA sequences in breast cancer by comparative genomic hybridization.
Proc. Nail. Acad. Sci. USA, 91: 2156—2160, 1994.
14. SchrOck, E., Thiel, 0., Lozanova, T., du Manoir, S., Meffert, M-C., Jauch, A.,
18.
19.
20.
and Croce, C. M. Suppression of tumorigenicity of breast cancer cells by microcell
mediated chromosome transfer: studies on chromosomes 6 and 11. Cancer Res., 54:
1331—1336,
1994.
25. Trent, J. M., Stanbridge, E. J., McBride, H. L., Meese, E. U., Casey, 0., Araujo, D. E.,
568—571,
1990.
26. Kamb, A., Weaver-Feldhaus,
lignant gliomas reveals multiple amplification sites and nonrandom chromosomal
gains and losses. Am. J. Pathol., 144: 1203—1218,1994.
15. Ried, T., Petersen, I., Holtgreve-Grez, H., Speicher, M. R., SchrOck, E., du Manoir,
17.
melanoma. Cancer Res., 51: 5449—5453, 1991.
24. Negrini, M., Sabbioni, S., Possati, L., Rattan, S., Corallini, A., Barbanti-Brodano, 0.,
Witkowski, C. M., and Nagle, R. B. Tumorigenicity in human melanoma cell lines
controlled by introduction of human chromosome 6. Science (Washington DC), 247:
Speicher, M. R., Ntlrnberg, P., Vogel, S., JSnisch, W., Doms-Keller, H., Ried, T.,
Witkowski, R., and Cremer, T. Comparative genomic hybridization of human ma
16.
21. Devilee, P., van Vliet, M., van Sloun, P., Dijkshoorn, K., Hermans, J., Pearson, P. L.,
and Cornelisse, C. J. Allelotype of human breast carcinoma: a second major site of
loss of heterozygosity is on chromosome 6q. Oncogene. 6: 1705—1711.
1991.
22. Foulkes, W. D., Ragoussis, J., Stamp, G. W. H., Allan, G. J., and Trowsdale, J.
Frequent loss of heterozygosity on chromosome 6 in human ovarian carcinoma. Br.
J. Cancer, 67: 551—559,1993.
23. Millikin, D., Meese, E., Vogelstein, B., Witkowski, C., and Trent, J. Loss of
heterozygosity for loci on the long arm of chromosome 6 in human malignant
J., Liu, Q., Harshman, K., Stockert, E., Day. III., R. S.,
Johnson, B. E., and Skolnick, M. H. A cell cycle regulator potentially involved in
genesis of many tumor types. Science (Washington DC), 264: 436—440,1994.
27. Cavanee, W. K. Accumulation of genetic defects during astrocytoma progression.
S., and Cremer, T. Mapping of multiple DNA gains and losses in primary small cell
Cancer (Phila.), 70: 1788—1793,1992.
lung carcinomas by comparative genomic hybridization. Cancer Res., 54: 1801—1806,
28. Stearns, M. E., and McGarvey, T. Biology of disease. Prostate cancer: therapeutics,
1994.
diagnostic, and basic studies. Lab. Invest., 67: 540—552,1992.
Joos, S., Scherthan, H., Speicher, M. R., Schiegel, J., Cremer, T., and Lichter, P.
29. Bunyan, R., Sawczuk, I. S., Benson, M. C., Siegal, J. D., and Olsson, C. A. Enhanced
Detection of amplified DNA sequences by reverse chromosome painting using
expression of the c-myc proto-oncogene in high-grade human prostate cancers.
genomic tumor DNA as probe. Hum. Genet., 90: 584—589,1993.
Prostate, 11: 327—337,1987.
Speicher, M. R., Prescher, G., du Manoir, S., Jauch, A., Horsthemke, B., Bornfeld, N.,
30. Hemming, W. H., Hamel, A., MacDonald, R., Ramsey, E., Peuigrew, N. M.,
Becher, R., and Cremer, T. Chromosomal gains and losses in uveal melanomas
Johnston, B., Dodd, J. 0., and Matusik, R. J. Expression of the c-myc proto-oncogene
detected by comparative genomic hybridization. Cancer Res., 54: 3817—3823,
in human prostatic carcinoma and benign hyperplasia. Cancer Res., 46: 1535—1538,
1994.
1986.
Union Intemationale Contre Ic Cancer TNM Classification of Malignant Tumours,
31. Bandyk, M. 0., Thao, L., Troncoso, P., Pisters, L. L., Palmer, J. L., von Eschenbach,
Ed. 4. Berlin: Springer-Verlag, 1987.
A. C., Chung, L. W. K., and Liang, J. C. Trisomy 7: a potential marker of human
Mostofi, F. K. Histological Typing of Prostate Turnouts. Geneva: World Health
prostate cancer progression. Genes Chromosomes & Cancer, 9: 19—27,1994.
Organization, 1980.
32. Alcaraz, A., Takahashi, S., Brown, J. A., Herath, J. F., Bergsralh, E. J., Larso-Keller,
Kallioniemi, 0-P., Kallioniemi, A., Piper, J., Isola, J., Waldman, F. M., Gray, J. W.,
and Pinkel, D. Optimizing comparative genomic hybridization for analysis of DNA
J. J., Lieber, M. M., and Jenkins, R. B. Aneuploidy and aneusomy of chromosome 7
detected by fluorescence in situ hybridization are markers of poor prognosis in
sequence copy number changes in solid tumors. Genes Chrom. Cancer, 10: 231—243,
prostate cancer. Cancer Rex., 54: 3998—4002, 1994.
1994.
347
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.
Genetic Changes in Primary and Recurrent Prostate Cancer by
Comparative Genomic Hybridization
Tapio Visakorpi, Anne H. Kallioniemi, Ann-Christine Syvänen, et al.
Cancer Res 1995;55:342-347.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/55/2/342
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on April 28, 2017. © 1995 American Association for Cancer Research.