| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Biology and Genetics |
Departments of Genetics [S. M. K., M. E., S. H., M. R. T., R. A. L.] and Pathology [V. M. A.], Institute for Cancer Research and Department of Gynecological Oncology [J. K., C. G. T.], The Norwegian Radium Hospital, 0310 Oslo, Norway, and Division of Human Cancer Genetics, Comprehensive Cancer Center, Ohio State University, Columbus, Ohio 43210 [J. S., P. P.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
More than 300 benign mature teratomas of the ovary have been
investigated cytogenetically, but only 4% have shown abnormal
karyotypes, and only with numerical chromosome changes [reviewed in
Hoffner et al. (4)
]. Cytogenetic data on
malignant OGCTs are sparse; only 50 cases [25 pediatric cases
(patient 
15 years), 24 adult cases (patient > 15 years), and 1 case of unknown age] have been reported
previously (Refs. 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24
; Table 1
). Thirteen of these tumors were karyotypically normal, 17 had only
numerical changes, and 20 had structural clonal chromosomal
aberrations. The most frequently observed change was gain of material
from chromosome 12, most often because of extra copies of chromosome 12
and/or the occurrence of an isochromosome of the short arm of
chromosome 12, i(12p). Indeed, i(12p) has been the only recurrent
structural chromosomal rearrangement detected by karyotype analysis of
OGCTs (Table 1)
. To our knowledge, no allelotype studies on this tumor
type have been reported, and only one CGH study has been published
previously (7)
.
|
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). To obtain specimens of tumor parenchyma with minimal admixture of
stromal elements, H&E-stained sections from several blocks were
reevaluated by a pathologist. From each case, the block with the purest
tumor tissue and the least presence of normal cells (visually estimated
to be <25%) was selected for sampling.
|
DNA Extraction and Measurement.
DNA extraction of the 10 x 10- or 20 x 5-µm tumor sections was performed using a modification of the
procedure described by Miller et al. (27)
.
Briefly, deparaffinization of the tissue sections was performed four
times by xylol treatment at 55°C for 10 min. The tissue sections were
then repeatedly dehydrated (four times) in absolute ethanol at 55°C
for 10 min. After removal of the ethanol, the pellet was dried at
37°C and incubated overnight at 55°C in 1 ml of lysis buffer [50
mM Tris-HCl (pH 8.5), 1 mM
EDTA, and 0.5 Tween]. The next day, proteinase K was added. When the
solution looked clear, 10 µl of 10% SDS and 10–20 µl of RNase (10
mg/ml) were added, followed by incubation at 37°C for 1–3 h.
Precipitation of proteins was obtained by adding 300 µl of 6
M NaCl. After centrifugation, the DNA in the
supernatant was precipitated in cold absolute ethanol, dried, and
dissolved in 1x Tris-EDTA. DNA from peripheral blood
lymphocytes from a healthy female donor was used as reference DNA for
the OGCT. DNA from peripheral blood lymphocytes from a healthy male
donor and DNA from a previously CGH-analyzed malignant peripheral nerve
sheath tumor [case 347 in Lothe et al. (28)
]
showing known losses and low level gains were used as a negative and
positive control, respectively. The control and reference DNA was
extracted with phenol-chloroform, followed by ethanol precipitation.
The DNA concentration of each sample was measured in a 1 x 10-4 mg/ml Hoechst solution (Hoechst 33258) with a fluorometer (TKO 100; Hoefer Scientific Instruments, San Francisco, CA).
CGH.
The principal CGH procedure of Kallioniemi et al.
(25)
was used, with the modifications described by
el-Rifai et al. (29)
. Test and reference
DNAs were labeled with two different fluorochromes in a standard
nick-translation reaction, and a mixture of two fluorochrome-conjugated
nucleotides (New England Nuclear, Boston, MA; FITC-12-dCTP and
FITC-12-dUTP for tumor DNA and Texas Red-6-dCTP and Texas Red-6-dUTP
for normal DNA) was used to improve the labeling of GC-rich areas in
the genome. The amount of DNase I (Roche Molecular Biochemicals,
Mannheim, Germany) and DNA polymerase I (Life Technologies, Inc.,
Rockville, MD) and the amount of time used for nick translation were
adjusted for each sample to achieve DNA fragment lengths of 500-2000
bp. This was accomplished by evaluating the degree of genomic DNA
degradation in gel electrophoresis before nick translation, and the
size of the fragments was then confirmed by a second gel
electrophoresis after cutting. Equal amounts (1 µg) of labeled tumor
and reference DNA were mixed with 20 µg of Cot-1 DNA (Life
Technologies, Inc.), ethanol-precipitated, dried, and dissolved in
hybridization buffer (Vysis, Downers Grove, IL). After denaturation,
the DNA was hybridized to normal, denatured metaphase spreads and
incubated for 2–3 days at 37°C. Finally, the slides were washed and
mounted in Vectashield (Vector Laboratories, Burlingame, CA), an
antifade counterstain solution with DAPI.
Microscopy and Analysis.
Good metaphase spreads were selected and evaluated microscopically for
each fluorescent dye, and single-color images (FITC, Texas Red, and
DAPI) were sequentially acquired with a Cohu 4900 charge-coupled
device (12-bit gray scale) camera on a fluorescence microscope
with an automated filter wheel (Zeiss Axioplan, Oberkochen, Germany)
using Cytovision software and hardware (Applied Imaging, Newcastle,
United Kingdom). Chromosomes were identified based on their inverted
DAPI banding pattern. The background fluorescence was subtracted, and
the green and red signals of each metaphase were normalized to 1.0. The
tumor and reference fluorescence profiles for each chromosome were
obtained by integrating the corrected pixel values along slices
orthogonal to the chromosome axis. Thereafter, ratio profiles
(green:red fluorescence) were calculated for each chromosome, and data
from at least 13 representative copies of each chromosome (range,
13–23 representative copies) were combined to generate average ratio
profiles with 95% confidence intervals. CGH results were controlled by
performing two independent analyses for three tumors (9T, 12T, and 19T)
using the same type of labeling and for five tumors (18T, 33T, 34T,
48T, and 51T), using reverse labeling.
Thresholds.
Upper and lower threshold values of 1.25 and 0.75, respectively, were
used to interpret the gains and losses of DNA sequences. The
theoretical reasoning for the use of these cutoff values is that they
correspond to the gain or loss of one chromosome homologue in 50% of
the cells analyzed, given a diploid tumor. To detect the gain of one
homologue present in 50% of the cells of a triploid tumor, one would
choose thresholds at 1.17 and 0.83, which would inevitably increase the
risk of more false positives. Previous results by Baker et
al. (30)
have indicated that most ESTs and DGs are
triploid to tetraploid, but no information was given about the size of
the ploidy fractions in that study. Eight of the 25 tumors of our
series had been evaluated previously by DNA flow
cytometry4
: (a) 2 tumors were diploid (43T and 51T); (b) 2
tumors contained mainly diploid cells (44T and 45T); (c) 1
tumor (52T) had two-thirds diploid cells and one-third triploid cells;
(d) 2 tumors (33T and 40T) had mainly triploid to tetraploid
cells; and (e) about half of the cells in 1 tumor (49T) were
in the pentaploid to hexaploid range. The ploidy data therefore gave us
no reason to deviate from the relatively conservative thresholds of
1.25 and 0.75.
SRO.
At least two tumors were required to identify each of the two borders
of a SRO of chromosomal gain or loss, i.e., only two tumors
were required if the region of gain or loss was completely the same in
those two tumors. If the region of gain or loss did not exactly
overlap in the two tumors, one or two additional tumors were
necessary to define the SRO, ensuring that at least two tumors define
each border. Tumors showing a harlequin pattern, i.e., two
or more areas with gains or losses along a given chromosome arm, were
not accepted as informative with regard to defining SRO borders.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and Fig. 1
. The most common chromosomal imbalances are illustrated in Fig. 2
.
|
|
ESTs.
The four ESTs all revealed copy number changes (average, 6 copy number
changes/case; range, 3–12 copy number changes/case). Gain of 12p and
gain from 1q were seen in three of four ESTs, whereas gains from 3p,
11q, and Xp and loss from 18q were each found in two tumors.
ITs.
Compared with the DGs and ESTs, fewer changes were detected in the nine
ITs (average, 1.4 changes/case; range, 0–4 changes/case). No changes
were observed in four of the ITs (12T, 22T, 44T, and 46T). Gain of all
or parts of 1p, 16p, 19, and 22q was each detected in two of five ITs
with changes, whereas the other alterations were detected in one tumor
each (Table 1)
.
SRO in OGCT.
Among the 13 DGs and ESTs with gains from chromosome 12, 2 tumors (40T
and 49T; both DGs) showed gain of the entire chromosome 12, whereas 11
tumors (69%) showed gain of parts of chromosome 12 (9T, 14T, 15T, 16T,
19T, 33T, 35T, 43T, 48T, 52T, and 55T), with 12p and 12q23-ter as
minimal common regions of gain. Nine additional SROs were found,
corresponding to gains of 1p35-ter, 3p21, 9q33-ter, 11q12–13,
16p12-ter, 20q, 21q22, and 22q13 and loss of 13q21–22.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Very little is known about the genetic alterations behind germ cell
tumorigenesis in the ovaries. Since 1982, 20 reports describing
cytogenetic banding analysis of a total of 50 OGCTs have been
published (Table 1)
. Thirteen tumors revealed normal karyotypes,
whereas 37 had clonal chromosome aberrations. Among the tumors with
structural changes (n = 20), i(12p) was
present in DGs (4 of 4 tumors), ESTs (1 of 2 tumors), mixed GCTs (4 of
10 tumors), and in a metastasis with mature teratoma differentiation
from an IT, but not in any of the 3 pure ITs. However, gain of the
entire chromosome 12 was reported in two ITs (10
, 14)
. In
the present CGH study, gain of chromosome 12 or 12p sequences was found
in the majority of DGs and ESTs (70%). The observed gain of 12p may
reflect the presence of an i(12p) but might also be the result of other
alterations. Because some of the tumors with gain from chromosome 12
have gain of 12p only, whereas others have gain of 12q only (Fig. 1
),
it is reasonable to suppose that two target areas exist along this
chromosome. Gain of 12p was also reported in most DGs and ESTs in the
recent CGH study by Riopel et al. (7)
. In
summary, our findings and those of Riopel et al.
(7)
suggest that gain of 12p is equally frequent in adult
and pediatric cases: in our series, 9 of 11 tumors with 12p gain were
from patients older than 15 years, whereas 10 of 14 tumors with 12p
gain reported by Riopel et al. (7)
were from
patients younger than 15 years. Finally, none of these two CGH studies
has shown gain of 12p in IT.
Gains of chromosome 12 or i(12p) are not restricted to GCTs but have also been observed in different types of benign ovarian tumors (32, 33, 34, 35, 36) , as well as in carcinomas (37, 38, 39, 40) and in one mixed mesodermal tumor of the ovary (5) . The pathogenetically essential outcome of the gain of chromosome 12 and i(12p) remains unknown.
Other recurrent imbalances found in more than 30% of the DGs or in at
least two of the four ESTs studied were gain of all or parts of 1p, 1q,
3p, 6p, 7, 8, 11q, 12q, 15q, 17, 19, 20q, 21q, and 22q and losses from
13q and 18q. Gains from 1q, 3p, 6p, 7, 8, 15, 17, 20, 21, 22, and Xp
and losses from 13q and 18q have been described previously in more than
two ESTs, DGs, and/or mixed GCTs (Table 1
; Ref. 7
). To our
knowledge, the gains at 1p, 11q, and 12q and the gain of chromosome 19
that we have seen have not previously been associated with OGCTs and
suggest novel target areas in the development of these tumors. The most
frequent change in DGs and ESTs was gain of 12p. Gains from 12q and 21q
and loss of 13q were found in DGs at comparable frequencies, with
12q23-qter, 21q, and 13q21–22 being identified as SROs. The target
genes affected by these changes remain unknown.
Three DGs (4T, 9T, and 40T) were from patients who had gonadoblastoma together with DG. In tumor 4T, the only change was loss of copy number of chromosome X, probably reflecting the known constitutional XY karyotype of this patient. The tumor DNA (XY) was hybridized with normal female reference DNA (XX) onto normal (XX) metaphase chromosomes. Tumors 9T and 40T, on the other hand, showed several acquired genetic changes that in both tumors included gains of parts of chromosomes 3, 6, 7, 12, 15, 19, and 20 and losses from chromosomes 4, 5, and 13. None of the other DGs (zero of nine DGs) showed gain of 3p21 or loss of 5p, suggesting that these changes might be specifically associated with gonadoblastoma or DG originating from a gonadoblastoma. We do not know the constitutional genotypes of these cases, but the CGH data indicate an XX sex chromosome complement in case 9T and XY in case 40T.
Pure IT components showed fewer alterations than did the DGs and ESTs,
with one to four changes in five of the nine ITs examined. The number
of changes we detected in ITs seems to be higher than that seen in
previous CGH findings (7)
, where gain of chromosome 14 was
the only change in one of six ITs studied. On the other hand, the
imbalance profiles detected by us are in agreement with published
karyotypes showing aberrations in 11 of 20 ITs, 8 of which had
numerical changes only, and 3 of which had both structural and
numerical changes (Table 1)
. We did not find any changes that were
specific for the IT subgroup, although 16p gain seemed to be more
frequent in ITs than in other OGCTs.
A comparison of the processes of tumorigenesis for TGCT and OGCT reveals striking similarities as well as some interesting differences. An early genetic event in the genesis of TGCT is polyploidization of a dysplastic germ cell precursor, resulting in a near-tetraploid carcinoma in situ (41) . Nonrandom losses and gains of chromosome regions accompany the further evolution of the carcinoma in situ into a seminoma and then into a nonseminoma, which typically are hypertriploid and hypotriploid lesions, respectively. An isochromosome for 12p is very common in primary TGCTs, and CGH studies have also revealed recurrent gains of all or parts of chromosomes and chromosome arms 1q, 7, 8, 12, and 21 as well as losses from 13q (42, 43, 44, 45) . With the exception of gonadoblastoma, precursor lesions such as carcinoma in situ have not been found in OGCTs in patients with abnormal sexual development. However, ploidy measurements by image analysis have revealed that most DGs and ESTs have cell populations in the triploid to tetraploid range (30) . In contrast, the ITs of females are usually diploid, although aneuploidy has been detected in some grade 3 lesions (30 , 46) . The presence of i(12p) in both male and female GCTs and the similarities in other, presumably secondary, chromosomal imbalances (+7, +8, +12, +21, and -13) and ploidy indicate that these tumors evolve through some of the same pathogenetic mechanisms in both sexes. However, it seems that ITs develop through a different pathway than that of other malignant OGCTs, inasmuch as these tumors typically are diploid and do not have i(12p) or other imbalances of chromosome 12.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported by grants from the Norwegian Cancer
Society (to S. H., M. R. T., and R. A. L.) and in part by Grant
P30 CA16058 from the National Cancer Institute, Bethesda, Maryland (to
J. S. and P. P.). S. M. K. is a research fellow of the Norwegian
Cancer Society. 
2 To whom requests for reprints should be
addressed, at Department of Genetics, Institute for Cancer Research,
The Norwegian Radium Hospital, Montebello, 0310 Oslo, Norway.
3 The abbreviations used are: OGCT, ovarian germ
cell tumor; GCT, germ cell tumor; CGH, comparative genomic
hybridization; DG, dysgerminoma; EST, endodermal sinus tumor; IT,
immature teratoma; SRO, smallest region of overlap; DAPI,
4',6-diamidino-2-phenylindole; TGCT, testicular germ cell tumor.
4 V. M. Abeler, J. Kærn, and C. G. Tropé,
unpublished observations.
Received 11/30/99. Accepted 4/ 4/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. D. Palmer, N. L. Barbosa-Morais, E. L. Gooding, B. Muralidhar, C. M. Thornton, M. R. Pett, I. Roberts, D. T. Schneider, N. Thorne, S. Tavare, et al. Pediatric Malignant Germ Cell Tumors Show Characteristic Transcriptome Profiles Cancer Res., June 1, 2008; 68(11): 4239 - 4247. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
