european urology 49 (2006) 169–175
available at www.sciencedirect.com
journal homepage: www.europeanurology.com
From Lab to Clinic
Relationship of NKX3.1 and MYC Gene Copy Number Ratio
and DNA Hypomethylation to Prostate Carcinoma Stage
Roland Kindich a, Andrea R. Florl b, Jörn Kamradt a, Jan Lehmann a, Mirko Müller b,
Bernd Wullich a, Wolfgang A. Schulz b,c,*
a
Clinic of Urology and Pediatric Urology, University of the Saarland, Homburg, Germany
Department of Urology, Heinrich Heine University, Moorenstr. 5; 40225 Düsseldorf, Germany
c
Center for Biological and Medical Research, Heinrich Heine University, Düsseldorf, Germany
b
Article info
Abstract
Article history:
Accepted September 8, 2005
Published online ahead of
print on November 15, 2005
Objective: High stage prostate cancers have been reported to frequently
harbor chromosome 8 alterations and hypomethylation of LINE-1 retrotransposons. The potential of these parameters for molecular staging of
prostate carcinoma was investigated.
Methods: High molecular weight DNA was extracted from 63 carcinoma
tissues (22 pT2, 38 pT3, 3 pT4). Chromosome 8 alterations were followed
by determining the ratio of NKX3.1 (at 8p21) to MYC (at 8q24) gene copy
numbers (NKX3.1:MYC ratio) using a new real-time PCR technique. LINE1 hypomethylation was quantified by Southern blot analysis.
Results: In 42 carcinomas NKX3.1 copy numbers were altered, with
decreases in 32 cases. Copy numbers of MYC were increased in 38 cases
and diminished in four. The NKX3.1:MYC ratio was altered in 45 specimens, with a decrease in all but two. NKX3.1 loss was associated with
tumor stage ( p < 0.03) and MYC gain with Gleason score ( p < 0.03). The
NKX3.1:MYC ratio was highly significantly associated with tumor stage
( p < 0.002), displaying 66% sensitivity and 87% specificity. LINE-1 hypomethylation was related ( p < 0.004) to tumor stage, but exhibited lower
sensitivity (59%) and specificity (77%).
Conclusion: A straightforward PCR technique detecting chromosome 8
alterations might be useful to predict which prostate cancers are organconfined while determination of hypomethylation appears to be somewhat less well suited.
# 2005 Elsevier B.V. All rights reserved.
Keywords:
Prostate cancer
Real-time quantitative PCR
Chromosome 8
L1 Retrotransposon
DNA Methylation
* Corresponding author. Tel. +49 211 81 18966; Fax: +49 211 81 15846.
E-mail address: wolfgang.schulz@uni-duesseldorf.de (W.A. Schulz).
0302-2838/$ – see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2005.09.012
170
1.
european urology 49 (2006) 169–175
Introduction
The most frequently altered chromosome in prostate carcinoma may be chromosome 8 [1]. In general,
sequences on 8p tend to be lost, whereas sequences
on 8q tend to be gained. These changes are however
not uniform and in individual carcinomas various
regions on 8p and 8q may be affected individually or
concomitantly. Among the most consistently gained
or lost genes, respectively, are MYC at 8q24.21 and
NKX3.1 at 8p21.2.
A considerable number of studies employing
cytogenetic techniques or loss of heterozygosity
analyses have concordantly shown that alterations
of chromosome 8 are associated with a more
aggressive behavior of prostate cancer, reflected
by higher tumor stage, higher grades, or higher rates
of metastasis and recurrences found in diverse
studies [2–11]. Most studies report an association of
such properties with 8q gain, while in others a
correlation with 8p loss has been found. In fact, the
two changes are in many cases linked mechanistically [12,13]. For instance, formation of an 8q
isochromosome or deletion of 8p distally from
recurrent breakpoint regions at 8p12 or 8p21 with
subsequent duplication of the remaining chromosome segment lead to concurrent gain of 8q and loss
of 8p sequences. The individual effects of 8p loss and
8q gain are therefore difficult to discern. For this
reason, one might hypothesize that the most robust
parameter associated with the overall clinical
behavior of prostate carcinoma might be the ratio
of 8p to 8q which reflects both changes. Accordingly,
we surmised that a simple and rapid measurement
of this ratio might be useful for ‘molecular staging’ of
this cancer and as a prognostic parameter. As most
losses on 8p include NKX3.1 as a target and most
gains on 8q include MYC, we have recently developed a relatively simple and straightforward, but
highly standardized PCR-based method to determine the relative copy numbers of these two genes
as an easily measurable parameter for numerical
chromosome 8 alterations [14].
Alterations on chromosome 8 in prostate carcinoma were also found to correlate with diminished
methylation of repetitive LINE-1 sequences [15].
LINE-1 sequences are the most frequent retrotransposons in humans constituting approximately
18% of the genome. They are highly methylated in
normal somatic tissues including prostate. In cancer
cells certain single-copy sequences become hypermethylated, while the methylation of repetitive
sequences, especially LINE-1 retrotransposons, is
often diminished. In effect, this leads to a decrease
in the methylcytosine content of cancer cells termed
genome-wide (‘global’) hypomethylation (reviewed
in [16]), which can be measured directly or through
its parallel effect on LINE-1 methylation. Pronounced global hypomethylation is associated with
chromosomal instability in several human cancer
types [16]. In prostate carcinoma, hypermethylation
of a number of genes including GSTP1 takes place
early during the development of most tumors
[17,18], while LINE-1 hypomethylation is found only
in a subset of cancers, usually those with a higher
stage [17]. Particularly pronounced hypomethylation is observed in specimen from metastatic and
androgen-independent cases [15]. Therefore, LINE-1
hypomethylation behaves similar to chromosome 8
alterations, begging the question whether the two
changes are causally related.
We have therefore investigated the relationship
of the NKX3.1 and MYC gene copy numbers and their
ratio as well as of LINE-1 hypomethylation to
prostate carcinoma stage and to Gleason grade.
2.
Methods
2.1.
Patients and tissues
The prostate carcinoma samples used were a subset of a
previously described series [17] that was selected by the
criteria of availability of greater DNA amounts and a complete
standardized patient follow-up. Except for one T4 tumor,
specimens were obtained by radical prostatectomy, macroscopically dissected by a specially qualified pathologist with
histological control of adjacent sections as described in detail
elsewhere [17], rapidly frozen and stored until extraction of
DNA and RNA. By the 1997 TNM classification, in the present
subseries, 22 carcinomas were staged as pT2, 38 as pT3, and 3
tumors as T4. With respect to Gleason sum, 17 were graded as
<7, 27 as 7, and 19 as >7. Lymph node metastases were present
in 13 cases. The individual tumors are listed in Table 1.
2.2.
DNA extraction and PCR analysis
DNA was extracted by a standard procedure using the Blood
and Cell Culture DNA Midi Kit (Qiagen, Hilden, Germany) with
extended proteinase K digestion. Quantitative real-time PCR
for NKX3.1 and MYC was performed as described14 using the
LightCyclerTM system (Roche Diagnostics, Mannheim, Germany) with the FastStart DNA Master SYBR Green I LightCycler
Kit (Roche Diagnostics). Relative gene copy number was
calculated by a modification of the 2 DDCT method from the
real-time PCR efficiencies (E) determined for each individual
run, and the CP deviations of the target and reference gene in a
test sample vs. a control according to the equation: E [(CPtargetCPref)test
(CPtarget-CPref)control]
. For the present study, the method
was recalibrated using a new set of leukocyte DNA samples
from 8 healthy individuals. Mean (2 SD) gene copy number
values in the normal DNA were 2.10 0.42 for NKX3.1,
1.76 0.27 for MYC, and 1.19 0.21 for the NKX3.1:MYC ratio,
171
european urology 49 (2006) 169–175
Table 1 – Tumor samples studied and overview of the results
Tumor No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
Sample No.
P8
P10
P14
P26
P28
P36
P42
P43
P44
P47
P49
P51
P52
P65
P74
P80
P85
P87
P91
P97
P99
P107
P109
P117
P119
P123
P133
P135
P137
P139
P141
P143
P145
P147
P153
P155
P161
P163
P165
P167
P169
P171
P173
P175
P177
P183
P185
P193
P197
P201
P203
P205
P209
P211
P213
P219
P223
P225
P230
P236
P238
P243
P247
Patient
Age
pT
66
68
74
71
68
73
74
58
57
53
62
59
56
62
53
66
68
62
61
72
68
59
66
68
62
57
72
65
73
65
68
66
71
70
75
55
63
66
71
67
72
61
63
73
68
67
56
68
63
66
73
73
71
67
60
64
72
62
67
74
61
46
55
4
3c
2c
3b
3a
3c
3c
3a
2c
3c
3a
2c
3c
3b
3a
3a
3a
3a
3b
3a
2b
3a
3a
3b
3b
2a
2b
2a
2b
3b
2b
3b
4
2b
3a
3a
2b
3a
3a
3a
3a
2b
2b
2b
3c
3a
2b
2b
3b
3a
2a
3a
3a
3b
2a
4
2b
3b
2a
3a
2a
2a
3c
pN
2
1
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
1
0
0
0
1
0
1
0
1
0
0
1
0
0
1
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
Gleason
Score
10
7
7
7
7
7
8
7
7
7
7
7
9
7
5
8
5
8
9
7
5
7
9
5
9
5
7
5
8
9
4
8
7
5
5
5
5
5
7
9
7
5
7
8
9
6
5
8
8
8
7
7
7
9
7
7
7
6
7
7
6
8
7
% Hypomet.
30
13
0
0
15
1
0
0
0
0
19
0
10
9
13
2
13
6
17
2
0
3
0
11
1
1
0
0
1
1
2
17
16
0
2
5
15
2
49
46
3
8
3
28
3
11
14
0
19
1
0
5
8
6
8
21
4
4
0
4
0
4
3
Gene Copy Number/Cell
NKX
MYC
NKX3.1/MYC
1.05
0.94
1.69
1.38
2.21
0.84
2.06
1.80
2.16
1.70
1.00
1.75
1.01
1.51
1.39
0.79
1.00
2.26
1.91
1.00
1.66
1.01
1.23
3.38
1.19
1.06
1.76
1.26
2.78
1.25
1.65
1.48
1.60
2.82
1.89
1.51
1.51
0.93
1.15
2.58
0.96
1.68
1.98
1.67
1.40
1.64
2.26
2.05
2.32
1.98
2.93
2.63
1.54
2.97
1.24
1.64
2.48
1.04
2.60
2.95
1.87
2.33
2.69
2.35
2.83
1.83
2.98
3.43
2.04
2.19
2.21
1.64
1.92
3.28
2.04
2.33
2.08
3.32
1.49
0.64
1.67
2.04
2.12
2.10
1.76
1.87
2.25
1.72
0.91
2.18
1.14
2.18
1.68
1.87
1.92
1.59
2.57
1.73
1.76
1.65
2.70
2.52
3.35
2.12
1.55
1.62
1.47
2.68
2.26
2.09
2.25
2.25
2.45
2.43
2.12
2.61
2.95
2.08
1.99
2.84
1.51
3.91
2.16
2.01
2.65
2.64
0.45
0.33
0.93
0.46
0.65
0.41
0.94
0.81
1.32
0.88
0.31
0.86
0.43
0.72
0.42
0.53
1.56
1.35
0.94
0.47
0.79
0.57
0.66
1.50
0.69
1.16
0.81
1.11
1.27
0.75
0.88
0.77
1.01
1.10
1.09
0.86
0.92
0.34
0.46
0.77
0.45
1.08
1.22
1.14
0.52
0.72
1.08
0.91
1.03
0.81
1.21
1.24
0.59
1.01
0.60
0.82
0.87
0.69
0.67
1.37
0.93
0.88
1.02
172
european urology 49 (2006) 169–175
respectively. LINE-1 hypomethylation was determined quantitatively by Southern blot analysis as described [15,17]. In
keeping with previous reports, normal methylation was
defined as <4% hypomethylation, moderate hypomethylation
as 4–12% and pronounced hypomethylation as >12%.
2.3.
Statistical procedures
The non-parametric Mann-Whitney-U test was applied to
compare samples with continuous variables (% hypomethylation, NKX3.1:MYC ratio) grouped by nominal data (e.g. pT2 vs.
pT3/4, Gleason <7 vs. 7, MYC gain yes/no) as variables did not
comply with Gaussian distribution according to the ShapiroWilk method. As a measure of association for nominal data
Fisher’s exact test was performed. All p-values were based on
two-sided tests and the threshold to accept statistical
significance was set at the alpha level 0.05. Analyses were
performed with the statistical software package SPSS version
10.0 (SPSS Inc., Chicago, USA).
3.
Results
Overall, 63 prostate carcinoma tissues were investigated. In 42 (67%) specimens, NKX3.1 copy numbers were significantly altered (see Methods section
for cut-offs used), with a decrease indicative of gene
loss in 32 (51%) cases. Gene copy numbers of MYC
were altered also (incidentally) in 42 (67%) specimens, with a gain in 38 (60%) cases. The ratio of
NKX3.1 to MYC gene copy number (NKX3.1:MYC ratio
for short) was significantly altered in 45 (71%)
specimens, with a decrease in all but two of them.
Loss of NKX3.1 correlated significantly with tumor
stage ( p < 0.03 for loss). By comparison, MYC gain
was rather associated with Gleason grading, showing a highly significant relationship with the prime
Gleason grade ( p < 0.003) and a weaker relationship
to Gleason sum ( p < 0.03 for Gleason 7). Importantly, the NKX3.1:MYC ratio was strongly correlated
to tumor stage. Specifically, a highly significant
( p < 0.002) discrimination between organ-confined
cancers (pT2) and cancers extending beyond the
prostate (pT3/4) was obtained (Fig. 1). Remarkably,
the 15 cancers with the lowest ratios, but also the 2
cancers with the highest ratios, were all staged as
pT3 or T4 (Fig. 1). At the optimal cut-off value of 0.85
determined by a receiver operating characteristic
(ROC) curve, the NKX3.1:MYC ratio displayed a 66%
sensitivity and a 87% specificity in identifying
cancers with a stage exceeding pT2.
Among the 63 specimens, 32 (51%) did not show
significant hypomethylation of LINE-1 sequences
compared to normal prostate tissue. Fifteen (24%)
carcinomas had moderate hypomethylation and in
16 (25%) specimens hypomethylation was pronounced. Percent LINE-1 hypomethylation was
Fig. 1 – Relationship of NKX3.1:MYC gene copy number ratio
to prostate carcinoma stage. Box plots of NKX3.1:MYC gene
copy number ratio as determined by real-time
quantitative PCR for prostate carcinomas staged as pT2
(left) or pT3 or T4 (right). Mean W 2 SD for normal
samples are indicated by the horizontal lines.
highly significantly related to tumor stage (pT2 vs.
pT3/4; p < 0.004), but this value did not discriminate
advanced stage cancers quite as well as the
NKX3.1:MYC ratio. At a cut-off value of >3%
(corresponding to none vs. moderate or pronounced
hypomethylation) a sensitivity of 59% and a specificity of 77% was achieved.
Interestingly, neither copy numbers of MYC nor
NKX3.1 appeared related to hypomethylation. The
NKX3.1:MYC ratio was likewise not significantly
related to LINE-1 hypomethylation as illustrated
by Fig. 2. The figure also illustrates that the
combined use of the two markers does not improve
the distinction between cancers staged as pT2 vs.
pT3 and pT4.
In a multivariate analysis using a generalized
linear model with a logit link function the predictive
value for capsular penetration of the NKX3.1:MYC
ratio was compared to the additional clinical
parameters Gleason grade, Gleason sum, preoperative PSA value, and the result of digital rectal
examination. Only the NKX3.1:MYC ratio remained
significant at p = 0.03. LINE-1 hypomethylation did
not retain a significant predictive value in multivariate analysis.
Neither of the parameters measured (MYC and
NKX3.1 copy numbers and their ratio, LINE-1
hypomethylation) was significantly related to biochemical recurrence during the follow-up period so
far (average 32 months).
4.
Discussion
Previous studies have concomitantly demonstrated
a relationship between alterations on chromosome
8 and clinical parameters of prostate carcinoma
european urology 49 (2006) 169–175
173
Fig. 2 – Relationship of NKX3.1:MYC gene copy number ratio to LINE-1 hypomethylation. (A) The NKX3.1:MYC gene copy
number ratio as determined by real-time quantitative PCR is plotted against the extent of LINE-1 hypomethylation as
determined by Southern blot analysis. Tumors staged as pT2 are indicated by circles and tumors as pT3 or T4 are indicated
by diamonds. The hatched rectangle indicates the area in which normal sample would be located. There is no correlation
between gene copy number changes and hypomethylation. The figure illustrates however how higher stage carcinomas
tend to present with more pronounced hypomethylation as well as a decreased ratio. Also note that even an optimal
combination of hypomethylation and NKX3.1:MYC ratio values (indicated by the dotted line) does not improve sensitivity
compared to the NKX3.1:MYC ratio alone: Even though one T2 carcinoma remains in the area below the line, 15 T3
carcinomas would be wrongly classified. (B) An example of LINE-1 hypomethylation analysis: DNA is cut with the
methylation-sensitive restriction enzyme HpaII (H) or its insensitive isoschizomer MspI (M), run on an agarose gel, blotted
and hybridized with a LINE-1 specific probe. In normal prostate tissue (PN), almost no low molecular weight bands are
observed in the HpaII lane, whereas bands appear at the indicated sizes in prostate carcinoma that can be quantified as
indicated at the bottom of the figure.
such as tumor stage and Gleason score, as well as
tumor recurrences after radical prostatectomy [2–
11]. Remarkably, associations were found with
losses on 8p in some studies [4,7,8,11] and with
gains at 8q in others [5,9,10]. The biological basis of
these findings is still not understood. A plausible,
but unproven hypothesis is that the progression rate
of prostate cancer depends on a balance between
several genes located on 8p and 8q. Regarding
prostate carcinoma diagnostics the robust association between chromosome 8 alterations and tumor
progression raises the interesting prospect that
measurements of surrogate parameters that are
regularly associated with chromosome 8 alterations
could be employed to gain information on an
individual prostate carcinoma. We have therefore
investigated two parameters related to chromosome
8 alterations. The NKX3.1:MYC copy number ratio is
obviously related to chromosome 8 alterations,
because the two genes are located on different arms
in regions regularly affected by losses and gains,
respectively. LINE-1 hypomethylation has been
reported to be highly significantly associated with
chromosome 8 alterations [15], but this association
could well be indirect (see below).
In the present study, both parameters showed
strong correlations to tumor stage. Interestingly,
neither of them was significantly associated with
Gleason score, although gains in MYC were, in
accord with previous studies [9,10]. Correct prediction of prostate carcinoma stage prior to surgery is
currently imperfect, even though it would be
desirable for an optimal choice of therapy. According to the results with the present series, the
NKX3.1:MYC ratio measured by a relatively simple
quantitative PCR assay appeared to be particularly
useful for the distinction between carcinomas
confined to the organ and those extending beyond
it. The ratio remained a significant predictor of
capsular penetration in multivariate analysis, suggesting that it may provide additional information to
the clinical parameters. If this result can be
corroborated in further, ideally prospective studies
on distinct and larger tumor series, the determination of this ratio might become a useful addition to
nomograms currently in use. An advantage of this
174
european urology 49 (2006) 169–175
assay would be its speed and requirement for
minimal amounts of DNA. It could therefore be
applied to biopsy samples containing a suitable
fraction of carcinoma cells prior to surgery aiding in
refinement of the therapeutic strategy.
In prostate cancer, most losses of 8p decrease the
relative NKX3.1 copy number [19] and most gains of
8q increase the relative MYC gene copy number [10].
Thus, the majority of alterations at 8p or 8q alone or
combined ought to result in a change in their ratio.
In contrast, it is unclear whether and how LINE-1
hypomethylation and chromosome 8 alterations are
mechanistically linked. Both have been found to be
common in very advanced cases of prostate cancer,
leading to a highly significant statistical association
[15]. The present series contained a much lower
fraction of androgen-refractory and/or metastatic
cases and no significant association between chromosome 8 gene copy number changes and LINE-1
hypomethylation was observed. This finding argues
that the two alterations are not mechanistically
linked to each other, but instead are both characteristic of very advanced prostate carcinomas. At
any rate, LINE-1 hypomethylation did not distinguish tumor stages as well as the NKX3.1:MYC ratio.
The use of hypomethylation for molecular staging of
prostate cancer is also hampered by the lack of a
simple PCR-based assay for its detection. The
technique used in the present study is sensitive
and quantitative, but requires moderate amounts of
high molecular weight DNA, which cannot always
be obtained from biopsies. As very recently new
techniques for the detection of DNA hypomethylation have been proposed [20,21], it will be interesting
to explore their potential in the context of prostate
carcinoma.
In conclusion, our analysis suggests that both the
determination of the NKX3.1:MYC copy number ratio
and the detection of LINE-1 DNA hypomethylation
might be helpful to predict prostate carcinoma
stage. The gene copy number ratio appears to be
more useful, being technically more practical and
exhibiting a higher sensitivity and specificity for the
distinction between organ-confined and advanced
stage cancers. We therefore suggest that this
technique should be tested more widely to explore
and develop its potential in the clinic.
Acknowledgements
We are most grateful to Ms. Christiane Hader for
excellent technical assistance. The study was
supported by the Deutsche Krebshilfe (grant numbers 70-2936 Wu I and 70-3193 Schu 1).
References
[1] Dong JT. Chromosomal deletions and tumor suppressor
genes in prostate cancer. Cancer Metast Rev 2001;20:173–
93.
[2] Takahashi S, Alcaraz A, Brown JA, et al. Aneusomies of
chromosomes 8 and Y detected by fluorescence in situ
hybridization are prognostic markers for pathological
stage C (pT3N0M0) prostate carcinoma. Clin Cancer Res
1996;2:137–45.
[3] Alers JC, Krijtenburg PJ, Rosenberg C, et al. Interphase
cytogenetics of prostatic tumor progression: specific chromosomal abnormalities are involved in metastasis to the
bone. Lab Invest 1997;77:437–48.
[4] Jenkins R, Takahashi S, DeLacey K, et al. Prognostic significance of allelic imbalances of chromosome arms 7q,
8p, 16q, and 18q in stage T3N0M0 prostate cancer. Genes
Chromosomes Cancer 1998;21:131–43.
[5] Sato K, Qian J, Slezak JM, et al. Clinical significance of
alterations of chromosome 8 in high-grade, advanced,
nonmetastatic prostate carcinoma. J Natl Cancer Inst
1999;91:1574–80.
[6] Alers JC, Rochat J, Krijtenburg PJ, et al. Identification of
genetic markers for prostatic cancer progression. Lab
Invest 2000;80:931–41.
[7] Macoska JA, Trybus TM, Wojno KJ. 8p22 loss concurrent
with 8c gain is associated with poor outcome in prostate
cancer. Urology 2000;55:776–82.
[8] Matsuyama H, Pan Y, Oba K, et al. Deletions on chromosome 8p22 may predict disease progression as well as
pathological staging in prostate cancer. Clin Cancer Res
2001;7:3139–43.
[9] Steiner T, Junker K, Burkhardt F, et al. Gain in chromosome 8q correlates with early progression in hormonal
treated prostate cancer. Eur Urol 2002;41:167–71.
[10] Tsuchiya N, Slezak JM, Lieber MM, et al. Clinical significance of alterations of chromosome 8 detected by fluorescence in situ hybridization analysis in pathologic
organ-confined prostate cancer. Genes Chromosomes
Cancer 2002;34:363–71.
[11] Matsuyama H, Pan Y, Yoshihiro S, et al. Clinical significance of chromosome 8p, 10q, and 16q deletions in prostate cancer. Prostate 2003;54:103–11.
[12] Macoska JA, Beheshti B, Rhim JS, et al. Genetic characterization of immortalized human prostate epithelial cell
cultures. Evidence for structural rearrangements of
chromosome 8 and i(8q) chromosome formation in
primary tumor-derived cells. Cancer Genet Cytogenet
2000;120:50–7.
[13] Birnbaum D, Adélaide J, Popovici C, et al. Chromosome
arm 8p and cancer: a fragile hypothesis. Lancet Oncol
2003;4:639–42.
[14] Kindich R, Florl AR, Jung V, et al. Application of a modified
real-time PCR technique for relative gene copy number
quantification to the determination of the relationship
between NKX3.1 loss and MYC gain in prostate cancer.
Clin Chem 2005;51:649–52.
[15] Schulz WA, Elo JP, Florl AR, et al. Genome-wide DNA
hypomethylation is associated with alterations on chro-
european urology 49 (2006) 169–175
[16]
[17]
[18]
[19]
[20]
[21]
mosome 8 in prostate carcinoma. Genes Chromosomes
Cancer 2002;35:58–65.
Ehrlich M. DNA methylation in cancer: too much, but also
too little. Oncogene 2002;21:5400–13.
Florl AR, Steinhoff C, Müller M, et al. Coordinate hypermethylation at specific sites in prostate carcinoma precedes LINE-1 hypomethylation. Brit J Cancer 2004;91:
985–94.
Yegnasubramanian S, Kowalski J, Gonzalgo ML, et al.
Hypermethylation of CpG islands in primary and metastatic human prostate cancer. Cancer Res 2004;64: 1975–86.
Asatiani E, Huang WX, Wang A, et al. Deletion, methylation, and expression of the NKX3.1 suppressor gene in
primary human prostate cancer. Cancer Res 2005;
65:1164–73.
Chalitchagorn K, Shuangshoti S, Hourpai N, et al. Distinctive pattern of LINE-1 methylation level in normal tissues
and the association with carcinogenesis. Oncogene
2004;23:8841–6.
Yang AS, Estecio MR, Doshi K, et al. A simple method for
estimating global DNA methylation using bisulfite PCR of
repetitive DNA elements. Nucleic Acids Res 2004;32:e38.
Editorial Comment
Henk van der Poel, Amsterdam, The Netherlands
h.vd.poel@nki.nl
The central role of the MYC oncogene in prostate
cancer is proven by several genetic experiments
[1,2]. Its combination with other potential oncogenes such as ras and Pim1 already clarified that its
role should be seen in the context of other (potential)
oncogenes and tumor suppressor genes.
The loss of the NKX3.1 gene, a homeobox gene, is
an early event in prostate cancer development. Loss
of NKX3.1 function was recently shown to be
involved in the oxidative stress response [3] and
thus the loss of NKX3.1 function fits the hypothesis
that prostate cancer is related to inflammatory
abnormalities [4]. Both loss of NKX3.1 and overexpression of MYC have been described in presumed
premalignant lesions such as prostate intraepithelial neoplasia and low grade cancer.
Here, Kindich et al. also suggest that gene copy
loss of NKX3.1 was associated with Gleason score
whereas MYC gene amplification increased with
tumor grade. In this relatively small population the
authors claim that the ratio of NKX3.1 and MYC
175
gene copies had additional predictive value to
known clinical parameters such as Gleason
score, DRE findings, and preoperative PSA with
respect to the discrimination between pT2 and pT3
tumors.
These data would be interesting for staging if the
results are reproducible on prostate biopsies. Until
then, these findings will not change urological
practice. More importantly, these data do provide
ground for new hypotheses on prostate cancer
progression. Unfortunately, the authors fail to
discuss such hypotheses. It remains to be proven
that gene copy number is associated with protein
levels but assuming that such a relation is present in
these prostate cancer, one can hypothesize that a
more aggressive phenotype, i.e. extracapsular
extension or nodal metastases, seem dependent
on responses to oxidative stress as is suggested by
the correlation with a (slightly) lower NKX3.1 copy
number (Table 1). If oxidative responses are important in prostate cancer progression, this may
provide entries for therapeutic exploitation. Moreover, the higher copy numbers of MYC in pT3 tumors
suggests that these tumors may depend on the
dedifferentional properties of MYC for extraprostatic growth.
Data from analyses like presented by Kindich
et al. should be seen in the light of available
mechanistic information rather than make authors
suggest that these markers may help daily urology,
yet.
References
[1] Ellwood-Yen K, Graeber TG, Wongvipat J, Iruela-Arispe ML,
Zhang J, Matusik R, et al. Myc-driven murine prostate
cancer shares molecular features with human prostate
tumors. Cancer Cell 2003;4:223–38.
[2] Williams K, Fernandez S, Stien X, Ishii K, Love HD, Lau YF,
et al. Unopposed c-MYC expression in benign prostatic
epithelium causes a cancer phenotype. Prostate 2005;
63:369–84.
[3] Ouyang X, DeWeese TL, Nelson WG, Abate-Shen C. Loss-offunction of Nkx3.1 promotes increased oxidative damage
in prostate carcinogenesis. Cancer Res 2005;65:6773–9.
[4] Nelson WG, De Marzo AM, DeWeese TL, Isaacs WB.
The role of inflammation in the pathogenesis of prostate
cancer. J Urol 2004;172:S6–11.