| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Molecular Oncology, Markers, Clinical Correlates |
First Department of Pathology, Hamamatsu University School of Medicine, Hamamatsu, 431-3192 Japan
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-satellite probes for 17 chromosomes. Because
frequent necrosis, inflammatory and normal cell contamination of
stroma, and poor attachment of tumor cells affected the quality and
reliability of the observations even with a recently developed FISH
protocol (7)
, we used MW irradiation (9, 10, 11)
for centromere identification by in situ hybridization in
gastric cancer cells. This approach may circumvent the problems
mentioned above and facilitate an examination of pure cancerous
epithelial cells, and thus the accurate detection of chromosomal
centromeric instability. We were able to identify the earliest event in
the multistep process of chromosomal numerical abnormality in gastric
cancer and to propose the sequential alterations in each chromosome
according to the progression. |
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Histological Evaluation According to the Japanese Classification
System (12)
.
All surgical materials were routinely processed in paraffin-embedded,
H&E-stained sections. Histological classification was based on the
General Rules for Gastric Cancer Study in Japan (12)
. In
terms of histological subcategorization, tubular-type carcinoma was
classified as well (tub 1) and moderately (tub 2) differentiated types
according to the degree of glandular formation. Poorly differentiated
carcinoma was classified as solid type (por 1) and nonsolid type (por
2), which were further divided into tubular type with scirrhous
infiltration (tub-sci) mucocellular type with scirrhous infiltration
(muc-sci), and signet ring cell carcinoma (sig). The invasiveness of
the tumor in the submucosal layer in gastric wall was graded into
sm1(slight invasion to submucosal connective tissue), sm2 (invasion to
middle layer of the submucosa), and sm3 (invasion to the deep layer of
the submucosa). Details of this description are described elsewhere
(12)
.
FISH.
Fresh touch preparations of the gastric cancer tissue and normal
lymphocytes were placed onto the slide glass and subjected to FISH by a
modification of the method of Pinkel et al.
(13)
The samples were denatured for 2 min at 75°C in
70% formamide (Boehringer Mannheim, Mannheim, Germany), 2 x SSC
(pH 7.0) and dehydrated. A panel of 17 centromeric satellite DNA
probes (D1Z5, D2Z, D3Z1, D4Z1, D6Z1, D7Z1, D8Z2, D10Z1, D11Z1, D12Z3,
D15Z, D16Z2, D17Z1, D18Z1, D20Z1, DXZ1, and DYZ3) derived from
chromosomes 1, 2, 3, 4, 6, 7, 8, 10, 11, 12, 15, 16, 17, 18, 20, X, and
Y, respectively, was purchased from Oncor, Inc. (Gaithersburg, MD), and
all of the probes were labeled with digoxigenin-11-dUTP by nick
translation (14)
. Chromosomal and probe DNAs were
denatured simultaneously for 5 min at 75°C in the hybridization
mixture on the slide glass. Hybridization was performed at 37°C
overnight in a humidified chamber. Digoxigenin-labeled probes were
detected with an FITC digoxigenin system, and the nuclei were
counterstained with propidium iodide (1 µl/ml; Sigma, St. Louis, MO).
One hundred or more interphase nuclei of cancer cells were evaluated
for each probe with an Olympus epifluorescence microscope equipped with
a WIB filter (Olympus, Tokyo, Japan).
MW Treatment for FISH.
Because nonspecific signal or background noise presumably due to
crush and degeneration in the specimen can influence the FISH
evaluation, quality control of the procedure was performed. We applied
intermittent MW treatment to the slides in addition to the conventional
FISH protocol. After treatment with ethanol, a mixed solution of a
probe was dripped onto the slide, a coverslip was mounted, and then the
slide was treated in a temperature-controllable MW processor (MI-77;
Azumaya Co., Tokyo, Japan). The MW was set to beam irradiation in
intervals of 3 s on and 7 s off, at a frequency of 2.45 GHz
at 300 W of output power with the temperature sensor set to 38°C. To
compare the MW-irradiated and conventional FISH (described above), we
tested all 17 probes under the two conditions on a lymphocyte control.
Each probe was hybridized with MW irradiation for 5 min, 15 min, 30
min, 1 h, 3 h, 6 h, 12 h, and 18 h,
respectively, and the results were compared with those for lymphocytes
with nonirradiated hybridization. The numbers of nuclei identified by
each probe and its signal intensity were simultaneously compared
between these differently irradiated preparations. After this
standardization, the frequency of diploid cells after hybridization
with MW irradiation was compared with those without MW irradiation
according to time course (mean ± SD) after fluorescent staining.
Evaluation.
We used two indices for evaluation of numerical abnormalities of 17
chromosomes in 100 and more tumor cells in each case. We defined NCAI
as the percentage of the tumor cells that do not have normal chromosome
numbers (2 for autochromosome and chromosome X and 1 for chromosome Y).
Modal number was defined as the number of the chromosomes most
prevalent in tumor cells. The most frequent copy numbers (the modal
number) of individual chromosomes was considered to represent the
ploidy level of the tumor. We set the cutoff level at 20% of nuclei
with an abnormal number of signals among all of the tumor nuclei;
i.e., we call the cases with numerical abnormality for a
given particular probe when the NCAI for that probe is >20%. As for
chromosome Y in case 19, the modal number is still 1 (normal), but
>49% of tumor cells have abnormal numbers of chromosome Y. In case
23, the two signals for D15Z are still the modal in the tumor cells,
but 55% of the tumor cells are not diploid for chromosome 15.
Actually, amplification of chromosome 15 based on NCAI exists in case
23. Thus, these two indices differently represent the cases of abnormal
chromosomal numbers in the following description (15)
.
Statistical Analysis.
For each probe, statistical differences between the frequency of
chromosomal abnormalities and histopathological types, depth, or stage
categories were examined by 
2 analysis.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). We could also obtain a more distinct signal in irradiated cancer
cells than in nonirradiated ones (Fig. 2)
.
|
|
. In four early stage
cancer cases (up to depth sm2 level), numerical increases were limited
to chromosomes 1, 2, 4, and 20. All cases of early cancer had numerical
abnormalities in chromosomes 1 and 2 (Fig. 3)
. When cancer invaded to the lower submucosal layer (sm3) and the
proper muscle layer, additional abnormalities in centromere numbers
gradually appeared, although changes in chromosomes 6 and 18 occurred
in only one of nine cases (11%). Furthermore, chromosome 3 showed
disomy in all nine cases, in which the depth of invasion was limited to
the proper muscle layer of the gastric wall (Fig. 3)
. Loss of the Y
chromosome was found in only one of five male cases. In contrast,
abnormal amplification of chromosomes 1, 2, 4, 8, and 20 in the sm and
mp layer occurred frequently (Table 1)
. Polysomy, represented by modal number, occurred in these 5
chromosomes at a frequency of 89% (8 of 9), 78% (7 of 9), 56% (5 of
9), 56% (5 of 9), and 33% (3 of 9), respectively, and NCAI larger
than 20% was found in 100% (9 of 9), 100% (9 of 9), 89% (8 of 9),
56% (5 of 9), and 67% (6 of 9) of the cases studied. These values
were all significantly greater (P < 0.001) than the
corresponding values for each of the other chromosomes (Table 2)
.
|
|
|
|
. The prevalence of
amplification in chromosomes 1, 2, 4, and 8 occurs more frequently,
when the tumor invaded beyond the gastric wall (depth se and si) was
100% (11 of 11), 100% (11 of 11), 100% (11 of 11), and 72.7% (8 of
11), respectively, for the modal number and 100% (11 of 11), 100% (11
of 11), 100% (11 of 11), and 81.8% (9 of 11) of the cases for NCAI,
respectively. The total prevalence of the numerical abnormalities in
chromosomes 1, 2, 4, 8, and 20 was greater than that of all of the
other chromosomes, 3, 6, 7, 10, 11, 12, 15, 16, 17, 18, X, and Y
combined (Table 2
; P < 0.001). Abnormalities in
chromosomes 3, 6, and 18 were almost entirely restricted to advanced
stage gastric cancer (Fig. 3)
.
Sex Chromosome Involvement and Loss of the Other Chromosomes.
Thirteen (65%) of 20 male samples and 1 (25%) of 4 females samples
exhibited the gain of an X chromosome, whereas loss of X was not found
in any sample. Gain of a Y chromosome by modal number was found in 3 of
20 male cases (15%). Five cases (25%) were considered to have a gain
of Y chromosome based on NCAI. Loss of the Y chromosome was found in 4
samples (20%) by modal number and 5 cases (25%) by NCAI.
Significant loss in the other autosomal chromosomes was for chromosome 17 in 25% of tumor cells in case 12, in 37% of tumor cells for chromosome 18 in case 10, in 26% for chromosome 20 in case 17, and in 25% for chromosome 18 in case 19. Loss of chromosomes other than these were all <20% of tumor cells. There were less abnormalities in chromosomes in tumors of female patients, with no abnormalities being found in chromosomes 16 and 18 in female cases.
Pathological Subtype and Chromosomal Instability.
The prevalence of chromosomal numerical abnormality according to
pathological subtype is shown in Table 3
. The signet ring cell type (three cases) had the lowest rate of
chromosome number changes in terms of NCAI (32%), followed by muc-sci
(46%). The highest rate of alterations were seen in the tub-sci type
(92.6%) followed by tub 2 (77.6%). There was a significant difference
in chromosomal instability between the mucocellular (sig and muc-sci)
and tubular-type (tub 1, tub 2, por 1 and tub-sci) carcinomas in total
(39.0% versus 68.0%: P < 0.001; Table 3
).
Even in advanced stage, six samples of sig or muc-sci (cases 1, 2, 8,
16, 18, and 24) had relatively fewer numerical abnormalities, with
chromosomes 18 and Y being normal in all cases. Tub 2, por 1, and
tub-sci, which all have severe morphological atypism (cytologically as
well), showed a high frequency of abnormalities encompassing all of the
chromosomes investigated, with amplification being found preferentially
in chromosomes 1, 2, 8, and 20 in all 11 cases.
Clinical Stage and Prognosis and Chromosomal Instability.
Although we only studied a few cases with distant metastasis, severe
chromosomal abnormality was observed in these cases. In one who had
multiple liver metastasis at operation, numerical abnormalities were
found in all chromosomes, and in the other two cases who died due to
peritoneal dissemination a few months after the operation, only one or
two chromosomes with normal numbers could be found for either case. In
our results, both the prevalence and the numbers of numerical
abnormalities of the chromosomes increased as the stage progressed
(Table 4)
.
|
. On
the basis of these results, we propose that the instability of the
chromosomal centromeric copy number occurs in a multistep process for
m-sm 2, sm3-mp, and ss-si according to the progression of the cancer
(Figs. 3
and 4)
.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our series of 24 surgically resected gastric cancer samples, the quality and quantity of tumor cells based on cytology and histology were well controlled by the new protocol (11) . We were able to identify chromosomes susceptible to CIN in gastric cancer during the course of gastric cancer progression. Furthermore, we demonstrated a cumulative process of chromosomal centromeric abnormalities based on 17 currently available chromosome-specific probes. Fresh material taken from touch preparations of cancerous tissue was used for FISH analysis, and the cancer cell quantity and the difference between cytologically normal and cancer cells were monitored by using Papanicolaou and Giemsa staining of touch tissue as well as HE staining of touched sections. This procedure made it possible to detect abnormalities in chromosome number in tumor cells with greater accuracy.
Our stage-dependent study is the first to document the relationship
between depth of invasion in the gastric wall and chromosomal numerical
changes. Cancer in an early stage of invasion (up to depth sm 2) was
predominantly close to diploid; a DNA ploidy level of 3 or more was
seen in tub 2, por 1, and tub-sci, but not in the mucocellular type.
Stage-dependent analysis (Fig. 3)
revealed that abnormalities were
limited to chromosomes 1, 2, 4, and 20 in the less invasive stages of m
and sm1. When the cancer invaded to the sm3 level, abnormalities were
more widely seen in other chromosomes. However, in some of the highly
invasive cases (depth se), numerical abnormalities detected were
limited to only four or five chromosomes including 1, 2, and 4 in all
three cases. Therefore, we can conclude that chromosomes 1, 2, and 4
are the most susceptible to CIN in gastric carcinogenesis in both early
and advanced stages. This raises the possibilities that fewer
abnormalities may be an indicator of better prognosis.
On the basis of these observations, we depicted the scheme of
sequential chromosomal numerical abnormalities during the course of
progression of gastric cancer in Figs. 3
and 4
. The numerical increase
of centromeres of chromosomes 1 and 2 occurs in the earliest stage in
any subtypes of gastric cancer. As shown in this scheme, the biological
behavior of gastric cancer can be divided into those invading within
two-thirds3 of the submucosal layer and those invading to the deeper
layer of the gastric wall from the standpoint of CIN. In other words,
up to sm 2, there are few abnormalities in a limited chromosomes (1
, 2
, 4
, and 20) , whereas abnormalities occur with a higher frequency once
tumors start to penetrate through the submucosa (sm3), and then
abnormal amplification expands to other chromosomes. After increases in
chromosomes 1 and 2, increased numbers of chromosomes 4, 8, and 20
occur in a relatively earlier stage and at a high frequency. These may
be the most characteristic chromosomes of gastric cancer. In the next
stage, the abnormalities spread to chromosomes 7, 10, 11 12, 15, 16,
17, and X. The abnormalities in chromosome 3, 6, 18, and Y are found
only in a relatively late stage and are less likely to occur.
Chromosomes 3 and 18 seem to be strongly protected from centromeric numerical abnormality. Chromosome 3 showed virtually no abnormalities in nine cases limited to depth mp, which may be a useful marker of progression. Gain and loss of chromosome Y occur at an equal frequency. The loss of Y has previously been reported in conventional cytogenetic analysis of gastric cancers by Ochi et al. (5) and by Rodriguez et al. (16) in three of four and six of seven male tumor cases, respectively. Our results were consistent with their observations. Furthermore, van Dekkan et al. (1) and Rao et al. (2) reported loss of Y chromosome by FISH in five of eight and four of six cases, respectively. Because the stage of these tumors is not given in their reports, the significance of this high prevalence is hard to interpret. Rao et al. (2) also documented the loss of Y in three of three cases of esophageal cancer. In our data, loss of Y was found in 5 of 20 cases (25%), a lower rate than in previous reports. However, we are one of the first to show clear evidence of Y gain in three gastric cancer cases.
Chromosomal abnormalities according to histological subtypes are shown
in Table 3
. Subtypes tub 2, por 1, and tub-sci had a high frequency of
abnormalities for each chromosomes, whereas the mucocellular
type had a significantly lower number of abnormalities (Table 3)
. The
order of the sequential chromosomal abnormalities during the
progression seems to be common in both tubular and mucocellular
subtypes; however, the extent of aneuploidal cancer cells in each case
were quite different among subtypes. The extent of these abnormalities
possibly reflects the biological behavior of the cancer within these
subtypes. As shown in Table 3
, abnormality in any chromosomes was 32%
in sig, 46% in muc-sci, and in 50% in tub 1. In the tub-sci subtype,
most of the tumor cells (92.6%) had numerical aberrations, and this
profile is identical with that of scirrhous carcinoma.
The genetic events unique to gastric carcinogenesis remain unclear (6 , 17, 18, 19) . A high frequency of loss of heterozygosity for chromosome 1 has been observed. A tumor suppressor gene, p73, has also recently been discovered on the short arm of chromosome 1 (20) . CIN of chromosome 1 has recently been detected in breast (21 , 22) and bladder cancer (23) with FISH. Our data may add another type of human cancer to this list. It would be interesting to see the relationship between loss of heterozygosity and centromere amplification (24) .
Some specific genes have been reported to be associated with the pathogenesis of gastric cancer including c-met on chromosome 7p (24) , K-ras-2 on 12q (25) , and c-erbB2 (26, 27, 28) , E-cadherin (4) , and p53 on 17p (29, 30, 31) . Unexpectedly, FISH analysis showed no outstanding gain or loss in these chromosomes. In recent analyses of gastric cancer with CGH (32 , 33) , a considerably different profile of long and short arm abnormalities on the chromosomes has been documented. Because evaluation of the intensity in a chromosomal arm in CGH procedure is usually performed without assessment of numerical changes of that chromosome, our data would be complimentary to the CGH approach and should validate some of these studies.
Interestingly, we found that numerical abnormalities of chromosome 3 and 18 were rare during gastric carcinogenesis. Loss of the arms of these chromosomes is associated with lung (34) , gastric, and colon cancer (35) . More recently, CGH analysis suggests gain of 3p and 3q in gastric cancer (36) . The DCC gene, loss of which is assumed to be related to colon cancer progression, is also on chromosome 18. In these lines, we could presume that centromere stability and arm instability may complementarily exist in chromosomes 18 and 3, which is consistent with Lengauer’s hypothesis of different pathways of CIN and microsatellite instability (3 , 37) .
The mechanistic basis for CIN observed here is challenging. Lengauer et al. (3 , 37) advocate that centromeric instability represented by numerical centromeric abnormality can occur due to mechanisms different from microsatellite instability, which is caused by a mismatch repair deficiency. They further argue that the processes of carcinogenesis and cancer progression can arise through these two distinct pathways.
Conventional karyotyping of solid tumors is notoriously difficult. Even when tumor cells can be cultured, karyotypes may be modified by in vitro expansion of particular clones and artifactual chromosomal changes during the procedure; and these karyotypic changes obtained do not always reflect changes in vivo. The analysis of chromosomal abnormalities with FISH like ours can reveal naturally occurring alterations and cannot be simply compared with the previous karyotyping results.
In this study, we analyzed the chromosomal centromeric instability of 17 chromosomes and propose a stepwise process of CIN in gastric cancer progression for the first time. A chromosome-wide search for numerical abnormalities covering almost all of the chromosomes may provide new perspectives on genetic changes in gastric cancer. Our observations suggest that chromosomal gain is much more frequent than loss in terms of centromere numbers; thus, it could be postulated that oncogenic changes may contribute to gastric carcinogenesis more often than that by defective suppressor genes. However, some of these types of numerical abnormalities, on the other hand, may also occur as secondary changes during the course of carcinogenesis. Actually, the limitation of our argument is that the numbers of centromere cannot reflect the changes in chromosome arm. Intratumor heterogeneity cannot be assessed by the procedure in this study. We could not study chromosomes 5, 9, 13, 14, 19, 21, and 22 because of the lack of appropriate specific probes. Because there are a number of genes located on chromosomes 5, 9, and 13 that are presumed to be involved in human carcinogenesis, the status of the centromere gain or loss of these chromosomes will be of additional interest.
|
FOOTNOTES |
|---|
1 This work was supported by Uehara Memorial
Foundation, the Smoking Research Foundation, the Ministry of the Health
and Welfare (9-3), a Grant in Aid for Science Research (B) and the
priority areas from the Ministry of Education, Culture, Science and
Sports, Japan. 
2 To whom requests for reprints should be
addressed, at First Department of Pathology, Hamamatsu University
School of Medicine, 3600 Handa-cho, Hamamatsu 431-3192 Japan. Phone:
053-435-2220; Fax: 053-435-2225; E-mail: hsugimur{at}hama-med.ac.jp
3 The abbreviations used are: FISH, fluorescence
in situ hybridization; NCAI, numerical chromosomal
aberration index; CIN, chromosome instability or centrosome
instability; MW, microwave; CGH, comparative genomic hybridization.
Received 11/29/99; revised 5/11/00; accepted 5/15/00.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M Iwaizumi, K Shinmura, H Mori, H Yamada, M Suzuki, Y Kitayama, H Igarashi, T Nakamura, H Suzuki, Y Watanabe, et al. Human Sgo1 downregulation leads to chromosomal instability in colorectal cancer Gut, February 1, 2009; 58(2): 249 - 260. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Sugimura Detection of chromosome changes in pathology archives: an application of microwave-assisted fluorescence in situ hybridization to human carcinogenesis studies Carcinogenesis, April 1, 2008; 29(4): 681 - 687. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Yamashita, H. Igarashi, Y. Kitayama, T. Ozawa, S. Kiyose, H. Konno, T. Kazui, S. Ishikawa, H. Aburatani, F. Tanioka, et al. Chromosomal Numerical Abnormality Profiles of Gastrointestinal Stromal Tumors Jpn. J. Clin. Oncol., February 1, 2006; 36(2): 85 - 92. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Weise, T. Liehr, U. Claussen, and K.-J. Halbhuber Increased Efficiency of Fluorescence In Situ Hybridization (FISH) Using the Microwave J. Histochem. Cytochem., October 1, 2005; 53(10): 1301 - 1303. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Ohno, N. Terada, S.-i. Murata, R. Katoh, and S. Ohno Application of Cryotechniques with Freeze-substitution for the Immunohistochemical Demonstration of Intranuclear pCREB and Chromosome Territory J. Histochem. Cytochem., January 1, 2005; 53(1): 55 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Nakamura, J.-P. Song, J. Isogaki, Y. Kitayama, and H. Sugimura Multiple (Multicentric and Multifocal) Cancers in the Ipsilateral Breast with Different Histologies: Profiles of Chromosomal Numerical Abnormality Jpn. J. Clin. Oncol., September 1, 2003; 33(9): 463 - 469. [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 |
sm2.
Green, cancer cases up to depth m
3 or 
1