GENES, CHROMOSOMES & CANCER 35:58 – 65 (2002)

Genomewide DNA Hypomethylation Is Associated

With Alterations on Chromosome 8 in Prostate
Carcinoma
Wolfgang A. Schulz,1,2* Jussi P. Elo,3,4 Andrea R. Florl,1 Sari Pennanen,3,4 Simon Santourlidis,1 Rainer Engers,2,5
Martin Buchardt,1 Hans-Helge Seifert,1 and Tapio Visakorpi3,4
1
Department of Urology, Heinrich-Heine-University, Düsseldorf, Germany
2
Biomedical Research Center, Heinrich-Heine-University, Düsseldorf, Germany
3
Laboratory of Cancer Genetics, Institute for Medical Technology, University of Tampere, Tampere, Finland
4
Tampere University Hospital, Tampere, Finland
5
Department of Pathology, Heinrich-Heine-University, Düsseldorf, Germany

To elucidate the relationship between genomewide DNA hypomethylation and chromosome instability, 55 prostate carcinoma
specimens were analyzed for extent of hypomethylation by Southern blot analysis of LINE-1 sequence methylation and for loss
or gain of chromosomal material by comparative genomic hybridization. Seventeen (31%) tumors showed strong hypo-
methylation of DNA, whereas four (7%) displayed slight hypomethylation and the rest of the tumors normal-level methylation.
Chromosomal aberrations were observed in 34 carcinomas. The most frequent chromosomal alterations were loss of 13q in
18 cases and aberrations in 8p (loss) or 8q (gain) in 16 cases. The presence of chromosomal loss or gain was significantly
associated with the presence of strong hypomethylation. A striking correlation (P ⫽ 0.00001) was observed between
aberrations on chromosome 8 and hypomethylation, whereas no association was seen between DNA hypomethylation and
loss of 13q. The association between DNA hypomethylation and the presence of metastases was statistically significant (P ⫽
0.044), and both chromosomal alterations and DNA hypomethylation tended to be more frequent in higher-stage tumors. In
conclusion, the data indicate that hypomethylation is associated with chromosomal instability in prostate cancer. Specifically,
a surprisingly strong association between alterations on chromosome 8 and genomewide hypomethylation was found. This
association suggests that DNA hypomethylation and alterations in chromosome 8 may be mechanistically linked to each other
in prostate carcinoma. © 2002 Wiley-Liss, Inc.

INTRODUCTION to this latter group. Although many specimens

In many human cancers, the overall level of
DNA cytosine methylation is decreased, even
though specific sites may be hypermethylated
(Baylin et al., 2000; Robertson, 2001). The de-
crease in overall DNA methylation affects different
types of repetitive sequences, such as retrotrans-
posons, endogenous retroviruses, and satellites,
which contain most of the methylcytosines in the
genome (Schulz, 1998; Ehrlich, 2000). In addition,
single-copy sequences may also become hypom-
ethylated (Cho et al., 2001; Takai et al., 2001).
Nevertheless, although DNA hypomethylation is
common in cancer and may not display much spec-
ificity with regard to sequence, it displays a remark-
able specificity with regard to cancer type. For
instance, DNA hypomethylation is regularly found
in bladder carcinoma, whereas the overall level of
DNA methylation is usually preserved in renal cell
carcinoma tissues (Florl et al., 1999). In other com-
mon cancers, hypomethylation is present in a frac-
tion of cases (Feinberg et al., 1988; Cravo et al.,
1996; Lin et al., 2001). Prostate carcinoma belongs
seem to retain an overall DNA methylation level
close to that of normal tissue, some show extensive
hypomethylation, with an up to 50% decrease in
overall methylation. Such extensive hypomethyla-
tion occurs mostly in metastatic carcinomas (Bed-
ford and van Helden, 1987, Santourlidis et al.,
1999), suggesting a relationship with tumor pro-
gression.
Causes and consequences of DNA hypomethy-
lation in human cancers are poorly understood.
Mouse cells engineered to contain very low levels
Supported by: Deutsche Forschungsgemeinschaft; Grant number:
Schu 604/7-2; Academy of Finland; Cancer Society of Finland;
Reino Lahtikari Foundation; Medical Research Fund of Tampere
University Hospital; Sigrid Juselius Foundation; Finnish Life and
Pension Insurance Companies.
*Correspondence to: Dr. Wolfgang A. Schulz, Urologische Klinik,
Heinrich-Heine-Universität, Moorenstrasse 5, 40225 Düsseldorf,
Germany. E-mail: wolfgang.schulz@uni-duesseldorf.de
Received 23 November 2001; Accepted 14 February 2002
DOI 10.1002/gcc.10092
Published online 15 April 2002 in
Wiley InterScience (www.interscience.wiley.com).

© 2002 Wiley-Liss, Inc.

 DNA METHYLATION/CHROMOSOMES IN PROSTATE CA
of DNA methylation show an increased mutation
rate, especially of deletions (Chen et al., 1998). In
humans suffering from immunodeficiency– centro-
meric instability–facial anomalies (ICF) syndrome,
which is caused by mutations in a DNA methyl-
transferase, DNA methylation of satellite se-
quences is severely impeded and chromosomes
that contain large satellite arrays tend to become
unstable (Xu et al., 1999). Likewise, demethylation
of specific satellite sequences is associated with an
increased frequency of loss of the corresponding
chromosome arms in several human cancers (Qu et
al., 1999a,b; Kanai et al., 2001). This mechanism
may also account for the association between the
extent of DNA hypomethylation and loss of chro-
mosome arm 9p in bladder carcinoma (Florl et al.,
2000). It is therefore likely that DNA hypomethy-
lation in human cancers promotes chromosomal
instability, especially loss of chromosome arms or
chromosomal fragments.
In the present study, we have investigated the
relationship between DNA hypomethylation and
chromosomal alterations in prostate carcinoma.
Specifically, we have attempted to determine
whether extensive DNA hypomethylation is asso-
ciated with chromosomal instability at large and/or
with the loss or gain of particular chromosomal
regions.
MATERIALS AND METHODS
Tumor Tissues
The material consisted of 50 specimens from
prostatectomies and 5 from transurethral resections
of locally recurrent tumors after androgen with-
drawal. The clinical staging and histological grad-
ing were performed according to the WHO classi-
fication. In summary, one tumor was staged as T1,
23 were staged as T2, 26 as T3, and 3 as T4; the
stage was unknown for two specimens. Four tu-
mors were histologically graded as G1, 33 as G2,
and 18 as G3, respectively. Lymph node and/or
distant metastases were present in 10 patients, in
38 patients no metastases were found, and for
seven patients no information on metastases was
available. Histopathological representativeness of
the specimens was verified by a trained pathologist.
DNA was extracted from the freshly frozen tumors
by standard methods.
Molecular and Cytogenetic Analyses
Aliquots from the same DNA were employed to
measure DNA hypomethylation as well as to detect
chromosomal losses or gains by comparative
59
genomic hybridization (CGH). To detect genome-
wide hypomethylation, 1 ␮g of DNA each was
extensively digested with the methylation-sensi-
tive restriction enzyme HpaII or its non-methyla-
tion-sensitive isoschizomer MspI, separated on aga-
rose gels, blotted, and hybridized with a 32P-
labeled specific LINE-1 probe. After HpaII
digestion, decreased methylation of LINE-1 se-
quences results in the appearance of new bands in
the 1.0- to 4.0-kb range on Southern blots, whose
intensities relative to the MspI signals, after correc-
tion for unequal loading of the two lanes, can be
used as a quantitative measure of genomewide
hypomethylation. With this method, background
hypomethylation is 0.5–1% in normal bladder and
approximately 2% in noncancerous prostate tissue
(Florl et al., 1999; Santourlidis et al., 1999).
CGH was done as previously described (Visa-
korpi et al., 1995). Briefly, DNA samples from
prostate tumors were labeled with FITC-dUTP
(DuPont, Boston, MA) and normal reference male
DNA with Texas Red-dUTP (DuPont) using nick
translation. Labeled DNAs were hybridized to nor-
mal male lymphocyte metaphase slides (Vysis,
Downers Grove, IL). After hybridization, the slides
were washed and counterstained. Five metaphases
from each hybridization were captured using a Pho-
tometrics ImagePoint CCD camera (Photometrics,
Tuscon, AZ) mounted on an Olympus BX50 epi-
fluorescence microscope (Olympus, Tokyo, Japan)
and IPLab Spectrum software program (Scanana-
lytics, Fairfax, VA). Relative DNA sequence copy
number changes were detected by analyzing the
fluorescence intensities of green (tumor) and red
(normal) signals along the lengths of all chromo-
somes in the metaphase spreads using the Quips
CGH analysis program (Vysis). Hybridizations of
FITC-labeled normal male DNA against Texas
Red–labeled normal female DNA, in each hybrid-
ization batch, were used as negative controls. The
mean green-to-red ratio and corresponding SD for
all autosomes remained between 0.85 and 1.15 in
these control hybridizations. Thus, chromosomal
regions with a mean ratio of 0.85 or less were
considered lost and those with a ratio 1.15 or more
gained in the prostate tumors. The Y chromosome
was excluded from CGH analysis.
Statistics
Statistical analysis of the data was performed by
chi-square and Kruskal–Wallis tests using software
from SPSS (Chicago, IL).
60 SCHULZ ET AL.
RESULTS
The data on DNA methylation analysis, CGH,
and clinical characteristics for each tumor are listed
in Table 1.
We have previously shown that overall DNA
methylation, as measured by decreased methyl-
ation of LINE-1 sequences, is usually slightly
lower in carcinoma-free prostates of elderly men
than in normal bladder or kidney (Santourlidis et
al., 1999), in which 0 –1% hypomethylation consti-
tutes the normal range. Therefore, we considered
up to 4% hypomethylation as indistinguishable
from normal. Seventeen (31%) prostate carcinomas
had pronounced (10% or more) DNA hypomethy-
lation, whereas 4 (7%) tumors showed slight (5–
9%) hypomethylation. The rest of the cases (62%)
showed methylation levels that could be consid-
ered normal (0 – 4%). DNA hypomethylation was
neither related to tumor T-stage nor histological
grade (Table 2). However, extensive hypomethy-
lation was observed in 7/11 (64%) cases with lymph
node or distant metastases (N⫹ or M⫹), but only
in 8/38 (21%) cases without detectable metastases
[P(␹2) ⫽ 0.044]. All five hormone-refractory locally
recurrent tumors displayed extensive hypomethy-
lation with some of the highest values observed
(Table 1). Nevertheless, the association between
DNA hypomethylation and the presence of metas-
tases remained statistically significant, if only tu-
mors untreated by androgen withdrawal were con-
sidered (P ⫽ 0.035).
Chromosomal alterations were detected by CGH
in 32 prostate carcinomas, whereas 23 specimens
did not show loss or gain of chromosomal regions.
Thirty tumors showed chromosome losses and 22
had gained chromosomes or chromosomal frag-
ments. The pattern of alterations in this series was
similar to that in other studies (Visakorpi et al.,
1995; Cher et al., 1996; Alers et al., 2000). The most
frequent losses involved chromosomes or chromo-
somal regions 13q (18 cases), 8p (13 cases), 6q (11
cases), 5q and 16q (8 cases each), 18q (5 cases), and
9p and 17p (4 cases each). The common minimal
deleted regions were 5q14 – q23, 6cen– q22, 8p12–
p21, 9p21–pter, 13q13– q21, 16q, 17p12–pter, and
18q12– qter. Almost all chromosomes were occa-
sionally affected by losses, notable exceptions be-
ing chromosomes 15 and 20 as well as 7q and 16p.
The most frequent chromosomal gains were ob-
served for chromosomes or chromosomal regions 8q
(10 cases), 7 and 19, 16p and 17q (5 cases), and the
telomeric region of chromosome 1p (3 cases). The
common minimal gained regions were 1p32–p36,
7p13– q22, 7q31– qter, 8q21, 8q23– qter, 16p13,
17q24 – qter, and 19 entirely. Gains of 8q occurred
concurrently with losses on 8p in 7/10 cases. Many
chromosomes or chromosome arms were never
gained, notably 5q, 6, 8p, 9p, 16q, and 18, which
were often affected by losses. The association be-
tween the presence of chromosomal gains and me-
tastases approached statistical significance (P ⫽
0.056), but otherwise neither the presence of chro-
mosomal alteration nor loss or gain was significantly
associated with the clinical or histopathological pa-
rameters tested (stage and grade). The association
of specific chromosomal alterations with histologi-
cal grade and clinical stage was investigated for the
two most frequent aberrations, losses on chromo-
some arm 13q and alterations of chromosome 8.
Neither alteration was significantly associated with
any clinical parameter. Of note, all androgen-inde-
pendent recurrent tumors harbored either losses on
chromosome arm 8p or gains of 8q or both.
As Table 1 demonstrates, DNA hypomethyla-
tion and overall chromosomal loss or gain as detect-
able by CGH tended to occur together. The me-
dian number of chromosomal alterations in prostate
cancers with pronounced hypomethylation (de-
fined as ⱖ10%) was 5, whereas in half of the 38
cases with normal methylation no alteration was
detected. However, no strict relationship was ob-
served. For example, pronounced DNA hypo-
methylation was found in four tumors (cases 1, 22,
32, and 33) without detectable chromosomal alter-
ations. There were also a number of cases (14, 17,
30, 41, and 42) displaying multiple chromosomal
losses and gains, but only slight or no changes in
DNA methylation. In spite of these exceptions,
most carcinomas with pronounced DNA hypo-
methylation displayed a high number of chromo-
somal alterations and vice versa. For example, 13 of
the 17 specimens with pronounced DNA hypo-
methylation contained between 4 and 14 chromo-
somal aberrations. Conversely, 13 of the 21 carci-
nomas with 4 or more chromosomal alterations
showed 10% or more DNA hypomethylation, and
only 5 had normal-level methylation. Overall, the
association between DNA hypomethylation and
chromosomal aberrations detectable by CGH (Fig.
1) was highly statistically significant [P(␹2) ⫽
0.0071; Kruskal–Wallis test: P ⫽ 0.0057]. For this
statistical analysis, data were divided into three
categories for DNA hypomethylation (defined as
above) and two categories (at least two losses
and/or gains vs. none or one) for chromosomal ab-
errations. Highly significant associations were also
observed between DNA hypomethylation and the
 DNA METHYLATION/CHROMOSOMES IN PROSTATE CA 61
TABLE 1. Clinical Characteristics, DNA Hypomethylation, and Chromosomal Alterations in Individual Prostate Carcinomas

Total
number
Case Tumor Tumor N/M % losses ⫹
number stage grade stage Hypomethylation Lossesa Gainsa gains

1 3 3 1/0 10 0 0 0
2 2 2 0/0 3 0 0 0
3 1 1 0/0 3 0 0 0
4 2 2 0/0 0 0 0 0
5 3 3 1/0 10 1p, 2q, 4q, 5q, 6q, 8p 8q 7
6 2 2 0/0 0 0 0 0
7 2 2 0/0 3 13q 16p 2
8 3 3 1/0 0 13q 0 1
9 4 2 1/0 0 0 0 0
10 4 3 0/0 0 0 0 0
11 3 3 0/0 1 0 0 0
12 3 2 0/0 4 13q 0 1
13 3 3 0/0 1 8p, 13q, 16q, 18 7q, 16p 6
14 2 2 0/0 0 0 0 0
15 3 3 0/0 0 5q, 6q, 8q, 9p, 11q, 12q, 13q 16pq, 17, 19 10
16 3 3 0/0 0 0 0 0
17 3 2 0/0 0 13q 0 1
18 2 2 0/0 0 0 0 0
19 3 3 1/0 19 0 0 0
20 3 2 0/0 0 0 0 0
21 2 3 0/0 4 0 0 0
22 3 3 0/0 15 1q, 2q, 5q, 6p, 8p 7, 8q 7
23 3 2 2/0 13 8p, 13q, 16q 8q 4
24 2 2 0/0 6 0 8q 1
25 2 2 0/0 5 7p, 11p, 11q, 17p 0 4
26 2 2 0/0 37 8p, 10p 8q, 17q 4
27 2 2 0/0 3 0 0 0
28 2 2 0/0 10 8p, 8q 16p, 17q 4
29 3 3 0/0 4 0 5p, 8q, 9q 3
30 3 3 1/0 2 2pq, 3pq, 4q, 6q, 12q, 13q 1p, 9q, 10q, 17, 19, 20q 12
31 3 2 0/0 13 6q, 12pq, 13q 1p 4
32 3 2 0/0 41 0 0 0
33 2 3 0/0 17 0 0 0
34 2 2 0/0 3 16q 0 1
35 3 1 0/0 4 13q 0 1
36 2 2 0/0 21 2p, 6q, 8p, 13q 22 5
37 3 1 0/0 7 5pq, 8p, 10q, 13 0 4
38 2 2 x/0 4 6q, 16q 19, 22 4
39 4 2 x/0 5 6q, 8p, 11q, 16q, 21q, 22 5p, 7p, 7q, 19p 10
40 2 2 x/0 35 13q, 14q, 16q, 18q, Xq 0 5
41 2 2 0/0 0 5q, 6q, 8p, 13q 7 5
42 2 2 0/0 0 0 0 0
43 2 2 0/0 0 6q, 9p, 13q 0 3
44 2 2 0/0 3 0 0 0
45 2 2 0/0 1 0 0 0
46 3 1 0/0 4 0 0 0
47 3 2 x/1 2 13q, 18q 19 3
48 xb 2 0/0 0 0 0 0
49 3 3 x/0 3 10 0 1
50 3 2 0/0 2 0 0 0
51c 2 3 2/2 30 5q, 8p, 10q, 13q, 14, 17p 1p, 1q, 8q, 13q, 17q 11
52c 3 2 x/1 12 17p, 19, 22q 8q 4
1p, 2q, 4p, 5pq, 6, 8p, 9pq,
53c 3 2 x/0 30 13q, 16q, 17p, 18q 8q, 16p, Xq 14
54c 3 3 x/1 30 2p, 8p, 16q 7, 8q 5
55c x 3 x/1 45 5q, 8p, 9p, 18q 7pq, X 6
a
Chromosomal arms affected by gains or losses.
b
x, not known.
c
Recurrent tumors after androgen-withdrawal treatment.
62 SCHULZ ET AL.

TABLE 2. DNA Hypomethylation and Chromosomal Alterations in Prostate Carcinomas*

Tumor stage Tumor grade NM status

T1/T2 T3/T4 G1 G2 G3 N⫹ or M⫹ N0 and M0

DNA hypomethylation
ⱖ10% 6 10 0 9 8 7 8
5–9% 2 2 1 3 0 0 2
⬍5% 16 17 3 21 10 4 28
Chromosome alterations
⬎1 10 14 1 15 9 7 25
ⱕ1 14 15 3 18 9 4 13
Median no. 0.5 1 0.5 1 2 4 0
Mean no. 2.1 3.3 1.3 2.5 3.8 4.6 1.8
Total number 24 29 4 33 18 11 38
*All figures refer to number of cases, with the exception of rows labeled “Median no.” and “Mean no.,” which refer to number of chromosomes.

Figure 1. Relationship between DNA

hypomethylation and chromosomal alter-
ations in prostate carcinoma. The extent
of DNA hypomethylation in individual
prostate carcinomas is shown for cases
with ⬎1 (top, 24 cases) or ⱕ1 (bottom,
31 cases) chromosomal alterations de-
tected by CGH; categories were selected
by the median value. The circled areas are
proportional to the number of cases
(1–12 per symbol). Recurrent tumors are
denoted by filled circles. The median ex-
tent of hypomethylation in each group is
denoted by t-bars.

Figure 2. Relationship between

DNA hypomethylation and specific
chromosomal alterations in prostate
carcinoma. The extent of DNA hy-
pomethylation in individual prostate
carcinomas is shown for cases with
alterations on chromosome 8 (top),
or with losses on 13q (bottom). p,
q, and pq denote loss of 8p, gain of
8q, or both, respectively. The circle
areas are proportional to the num-
ber of cases (1– 4 per symbol). Dot-
ted lines connect identical cases. Re-
current tumors are denoted by filled
circles. The median extent of hy-
pomethylation in each group is de-
noted by t-bars.

presence of either loss [P(␹2) ⫽ 0.012; Kruskal–
Wallis test: P ⫽ 0.012] or gain of at least one
chromosomal region [P(␹2) ⫽ 0.00021; Kruskal–
Wallis test: P ⫽ 0.00092].
The most frequent chromosomal alteration in
the present series of prostate carcinomas was loss of
chromosome arm 13q. It appeared to be unrelated
to DNA hypomethylation [P(␹2) ⫽ 0.94; Fig. 2].
DNA hypomethylation was clearly associated with
alterations on chromosome 8 (Fig. 2). Eleven spec-
imens of the 17 (65%) with loss of 8p and/or gain of
8q showed pronounced hypomethylation, and
three (18%) showed intermediate methylation. By
comparison, only 7 out of 18 (39%) tumors with 13q
loss displayed strong hypomethylation, and six
among these had alterations on chromosome 8 as
well. The association between alterations on chro-
mosome 8 and DNA hypomethylation showed the
highest significance of all tested associations in the
present study [loss: P(␹2) ⫽ 0.00001; Kruskal–Wal-
 DNA METHYLATION/CHROMOSOMES IN PROSTATE CA
lis test: P ⫽ 0.00013]. All seven specimens with
alterations on both arms displayed strong hypo-
methylation (Fig. 2). Likewise, all recurrent andro-
gen-independent tumors displayed alterations of
chromosome 8, as well as pronounced hypomethy-
lation.
DISCUSSION
The aim of the present study was to determine
the relationship between DNA hypomethylation
and chromosome instability in prostate carcinoma.
Prostate carcinoma seemed particularly suited, be-
cause DNA hypomethylation in this tumor differs
substantially between different cases. Many cases
show close to normal or only slightly decreased
overall DNA methylation levels, whereas hypo-
methylation is pronounced in some cases. For in-
stance, in this study 31% of the tumors showed
pronounced (ⱖ10%) hypomethylation. The distri-
bution of hypomethylation in prostate cancer is
very different from that, for example, in bladder
cancer, where DNA hypomethylation is present in
almost every specimen (Florl et al., 1999). We
found a highly significant correlation between
DNA hypomethylation and chromosomal instabil-
ity, which could, in theory, be explained by three
different arguments.
First, the relationship could be indirect, because
both alterations are related to tumor progression.
Indeed, both alterations tended to occur more fre-
quently in advanced stage tumors, particularly in
metastatic cases (N⫹ or M⫹), and in carcinomas
recurring after androgen deprivation therapy. How-
ever, their relationship toward each other was
much stronger than their association with any of
the clinical parameters of tumor progression such as
stage, grade, or even presence of metastases. This
suggests a direct relationship between hypomethy-
lation and chromosomal instability.
Second, DNA hypomethylation might facilitate
chromosomal instability, particularly chromosome
breaks, deletions, amplification, and illegitimate re-
combination through its effect on chromatin struc-
ture (Ehrlich, 2000; Robertson, 2001). Demethyl-
ation of repetitive sequences such as LINE-1
retrotransposons, which are otherwise strongly
methylated and maintained in an inactive chroma-
tin compartment, may be particularly important in
this regard. Comparative genomic hybridization
would detect such events, if they caused loss or
gain of sufficiently large chromosomal fragments in
a substantial fraction of tumor cells. The highly
significant association observed between either
gain or loss of chromosomal material and DNA
63
hypomethylation is consistent with the assumption
that DNA hypomethylation promotes chromosomal
instability. Because in some cases chromosomal insta-
bility was observed together with normal levels of
DNA methylation, DNA hypomethylation is obvi-
ously not the only factor promoting genomic insta-
bility.
The association between hypomethylation and
alterations in chromosome 8 was particularly evi-
dent. It has been suggested previously that associ-
ation of DNA hypomethylation with alterations on
chromosome 1, 9, or 16 in various cancers could be
attributable to hypomethylation of specific juxta-
centromeric satellite sequences on these chromo-
somes, leading to increased recombination be-
tween them (Qu et al., 1999a,b; Kanai et al., 2001;
Kimura et al., 2001; Wong et al., 2001). However,
chromosome 8 is not known to contain structural
features that would make it particularly prone to
destabilization by DNA demethylation. The mech-
anisms underlying the common changes affecting
this chromosome in prostate cancer are not known.
One important mechanism involves formation of an
isochromosome 8q with concomitant loss of 8p
(Macoska et al., 2000). Given that in the present
study concomitant loss of 8p and gain of 8q were
consistently associated with pronounced DNA hy-
pomethylation, it is conceivable that DNA hypo-
methylation facilitates this particular chromosomal
change.
Third, DNA hypomethylation might be a conse-
quence of chromosomal instability, either at large
or through alterations of specific chromosomes. Ob-
viously, our data argue against the first alternative.
If DNA hypomethylation was a consequence of
overall chromosomal instability, it ought to corre-
late to similar degrees with alterations of any chro-
mosome. However, no association was observed
between hypomethylation and the most frequent
chromosomal aberration, loss of 13q, whereas a
striking and highly significant correlation was re-
vealed with alterations of chromosome 8. In addi-
tion, the association of DNA hypomethylation with
chromosome 8 alterations, but not 13q loss, makes
it highly unlikely that the association between hy-
pomethylation and overall chromosomal instability
might result from both changes being more easily
detectable in samples containing a higher propor-
tion of carcinoma cells.
Multiple alterations on chromosome 8 are
thought to be important in the development and
progression of prostate carcinoma. The gain of 8q
has been shown to predict the course of disease
well, even better in combination with loss of 8p
64 SCHULZ ET AL.
(Jenkins et al., 1998; Sato et al., 1999; Alers et al.,
2000). Losses affecting 8p often involve the entire
arm, but at least two distinct regions can be distin-
guished. 8p21–22, which contains the prostate dif-
ferentiation gene NKX3A (He et al., 1997), is often
affected at early stages, whereas deletions in 8p23
are more often found in advanced-stage tumors.
Likewise, the entire arm is often involved in gains
of 8q, although isolated gain of smaller regions has
also been observed. The minimal commonly am-
plified regions according to CGH analyses are 8q21
and 8q23–24 (Visakorpi et al., 1995; Cher et al.,
1996; Nupponen et al., 1998). Because the ampli-
fication of the latter region often involves the MYC
gene, its activation may be crucial (Sato et al.,
1999). However, additional candidate genes on 8q,
such as EIF3S3, GC79, and PSCA, have been pro-
posed (Reiter et al., 1998; Nupponen et al., 1999;
Chang et al., 2000). The various changes on chro-
mosome 8 may each have quite different effects on
prostate cancer development, but because they of-
ten occur concomitantly, this is difficult to ascer-
tain. For this reason, we did not attempt to distin-
guish whether either the loss of 8p or the gain of 8q
might be associated with DNA hypomethylation. It
is quite possible that the relevant parameter for
progression of prostate carcinoma is in fact an im-
balance between genes located on chromosome 8
p- and q-arms, which respectively promote prolif-
eration and differentiation of prostate epithelial
cells. This imbalance may also disturb the mecha-
nisms that maintain DNA methylation levels dur-
ing cellular proliferation. Thus, it will be valuable
to determine the effects of presumed prostate can-
cer tumor-suppressor genes, such as NKX3A on 8p,
and of suspected oncogenes, such as MYC on 8q, on
DNA methylation levels in proliferating prostate
cells.
ACKNOWLEDGMENTS
We are most grateful to Dr. Reinhardt Willers,
Universitätsrechenzentrum Düsseldorf, for per-
forming the statistical analyses and to Dr. Marc
Cronauer for helpful discussions.
REFERENCES
Alers JC, Rochat J, Krijtenburg PJ, Hop WC, Kranse R, Rosenberg
C, Tanke HJ, Schröder FH, van Dekken H. 2000. Identification
of genetic markers for prostatic cancer progression. Lab Invest
80:931–942.
Baylin SB, Esteller M, Rountree MR, Bachman KE, Schuebel K,
Herman JG. 2001. Aberrant patterns of DNA methylation, chro-
matin formation and gene expression in cancer. Hum Mol Genet
10:687– 692.
Bedford MT, van Helden PD. 1987. Hypomethylation of DNA in
pathological conditions of the human prostate. Cancer Res 47:
5274 –5276.
Chang GT, Steenbeek M, Schippers E, Blok LJ, van Weerden WM,
van Alewijk DC, Eussen BH, van Steenbrugge GJ, Brinkmann
AO. 2000. Characterization of a zinc-finger protein and its associ-
ation with apoptosis in prostate cancer cells. J Natl Cancer Inst
92:1414 –1421.
Chen RZ, Petterson U, Beard C, Jackson-Grusby L, Jaenisch R.
1998. DNA hypomethylation leads to elevated mutation rates.
Nature 395:89 –93.
Cher ML, Bova GS, Moore DH, Small EJ, Carroll PR, Pin SS,
Epstein JI, Isaacs WB, Jensen RH. 1996. Genetic alterations in
untreated metastases and androgen-independent prostate cancer
detected by comparative genomic hybridization and allelotyping.
Cancer Res 56:3091–3102.
Cho M, Uemura H, Kim SC, Kawada Y, Yoshida K, Hirao Y, Konishi
N, Saga S, Yoshikawa K. 2001. Hypomethylation of the MN/CA9
promoter and upregulated MN/CA9 expression in human renal
cell carcinoma. Br J Cancer 85:563–567.
Cravo M, Pinto R, Fidalgo P, Chaves P, Gloria L, Nobre-Leitaor C,
Costa Mira F. 1996. Global DNA hypomethylation occurs in the
early stages of intestinal type gastric carcinoma. Gut 39:434 – 438.
Ehrlich M. 2000. DNA hypomethylation and cancer. In: Ehrlich M,
editor. DNA alterations in cancer. Natick, MA: Eaton. p 273–291.
Feinberg AP, Gehrke CW, Kuo KC, Ehrlich M. 1988. Reduced
genomic 5-methylcytosine content in human colonic neoplasia.
Cancer Res 48:1159 –1161.
Florl AR, Loewer R, Schmitz-Dräger BJ, Schulz WA. 1999. DNA
methylation and expression of L1 LINE and HERV-K provirus
sequences in urothelial and renal cell carcinoma. Br J Cancer
80:1312–1321.
Florl AR, Franke KH, Niederacher D, Gerharz CD, Seifert HH,
Schulz WA. 2000. DNA methylation and the mechanisms of
CDKN2A inactivation in transitional cell carcinoma of the urinary
bladder. Lab Invest 80:1513–1522.
He WW, Sciavolino PJ, Wing J, Augustus M, Hudson P, Meissner
PS, Curtis RT, Shell BK, Bostwick DG, Tindall DJ, Gelmann EP,
Abate-Shen C, Carter KC. 1997. A novel human prostate-specific,
androgen-regulated homeobox gene (NKX3.1) that maps to 8p21,
a region frequently deleted in prostate cancer. Genomics 43:69 –
77.
Jenkins R, Takahashi S, DeLacey K, Bergstrahl E, Lieber M. 1998.
Prognostic significance of allelic imbalance of chromosome arms
7q, 8p, 16q, and 18q in stage T3N0M0 prostate cancer. Genes
Chromosomes Cancer 21:131–143.
Kanai Y, Ushuima S, Kondo Y, Nakanishi Y, Hirohashi S. 2001.
DNA methyltransferase expression and DNA methylation of
CpG islands and peri-centromeric satellite regions in human colo-
rectal and stomach cancers. Int J Cancer 91:205–211.
Kimura F, Florl AR, Seifert HH, Louhelainen J, Maas S, Knowles
MA, Schulz WA. 2001. Destabilisation of chromosome 9 in tran-
sitional cell carcinoma of the urinary bladder. Br J Cancer 85:
1887–1893.
Lin CH, Hsieh SY, Sheen IS, Lee WC, Chen TC, Shyu WC, Liaw
YF. 2001. Genome-wide hypomethylation in hepatocellular car-
cinogenesis. Cancer Res 61:4238 – 4243.
Macoska JA, Beheshti B, Rhim JS, Hukku B, Lehr J, Pienta KJ,
Squire JA. 2000. Genetic characterization of immortalized human
prostate epithelial cell cultures: evidence for structural rearrange-
ments of chromosome 8 and i(8q) chromosome formation in pri-
mary tumor-derived cells. Cancer Genet Cytogenet 120:50 –57.
Nupponen NN, Kakkola L, Koivisto P, Visakorpi T. 1998. Genetic
alterations in hormone-refractory recurrent prostate carcinomas.
Am J Pathol 153:1481–1488.
Nupponen NN, Porkka K, Kakkola L, Tanner M, Persson K, Borg
Å, Isola J, Visakorpi T. 1999. Amplification and overexpression of
p40 subunit of eukaryotic translation initiation factor 3 in breast
and prostate cancer. Am J Pathol 154:1777–1783.
Qu G, Dubeau L, Narayan A, Yu MC, Ehrlich M. 1999a. Satellite
DNA hypomethylation vs. overall genomic hypomethylation in
ovarian epithelial tumors of different malignant potential. Mutat
Res 423:91–101.
Qu GZ, Grundy PE, Narayan A, Ehrlich M. 1999b. Frequent hy-
pomethylation in Wilms tumors of pericentromeric DNA in chro-
mosomes 1 and 16. Cancer Genet Cytogenet 109:34 –39.
Reiter RE, Gu Z, Watabe T, Thomas G, Szigeti K, Davis E, Wahl
M, Nisitani S, Yamashiro J, Le Beau MM, Loda M, Witte ON.
1998. Prostate stem cell antigen: a cell surface marker overex-
pressed in prostate cancer. Proc Natl Acad Sci USA 95:1735–1740.
Robertson KD. 2001. DNA methylation, methyltransferases, and
cancer. Oncogene 20:3139 –3155.
Santourlidis S, Florl A, Ackermann R, Wirtz HC, Schulz WA. 1999.
 DNA METHYLATION/CHROMOSOMES IN PROSTATE CA
High frequency of alterations in DNA methylation in adenocar-
cinoma of the prostate. Prostate 39:166 –174.
Sato K, Qian J, Slezak JM, Lieber MM, Bostwick DG, Bergstralh EJ,
Jenkins RB. 1999. Clinical significance of alterations of chromo-
some 8 in high-grade, advanced, nonmetastatic prostate carci-
noma. J Natl Cancer Inst 91:1574 –1580.
Schulz WA. 1998. DNA methylation in urological malignancies. Int
J Oncol 13:151–167.
Takai D, Gonzales FA, Tsai YC, Thayer MJ, Jones PA. 2001. Large
scale mapping of methylcytosines in CTCF-binding sites in the
human H19 promoter and aberrant hypomethylation in human
bladder cancer. Hum Mol Genet 10:2619 –2626.
65
Visakorpi T, Kallioniemi A, Syvänen A-C, Hyytinen E, Karhu R,
Tammela T, Isola J, Kallioniemi O-P. 1995. Genetic changes in
primary and recurrent prostate cancer by comparative genomic
hybridization. Cancer Res 55:342–347.
Wong N, Lam WC, Lai PB, Pang E, Lau WY, Johnson PJ. 2001.
Hypomethylation of chromosome 1 heterochromatin DNA corre-
lates with q-arm copy gain in human hepatocellular carcinoma.
Am J Pathol 159:465– 471.
Xu GL, Bestor TH, Bourc’his D, Hsieh CL, Tommerup N, Bugge
M, Hulten M, Qu X, Russo JJ, Viegas-Pequinot E. 1999. Chro-
mosome instability and immunodeficiency syndrome caused by
mutations in a DNA methyltransferase gene. Nature 402:187–191.