| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
From the Department of Pathology* and Division of Cell and Tissue Culture, Kitasato University School of Medicine, Kanagawa; and the Department of Carcinogenesis, The Cancer Institute, Japanese Foundation for Cancer Research, Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
The functional consequences of ß-catenin expression during tumorigenesis are complex. For example, ß-catenin may exert oncogenic effects by activation of transcription from the gene promoters for cyclin D1 and c-myc, positive regulators of cell proliferation acting at the G1 to S phase transition.7-9 However, overexpressed ß-catenin can also induce accumulation of p53 that protects against oncogenic effects.10,11 Moreover, induction of PML (promyelocytic leukemia) expression by ß-catenin suppresses the tumorigenicity of renal carcinoma cells, indicating a novel pathway for catenin-mediated growth inhibition.12
We previously demonstrated that ß-catenin abnormalities are common in endometrioid type endometrial carcinomas (Em Cas) with areas of squamous differentiation (SqD).13 Interestingly, nuclear accumulation was more frequent in SqD areas, pointing to a possible role of ß-catenin during trans-differentiation from glandular to squamoid features. To test this hypothesis, we here investigated the expression of ß-catenin, as well as cyclin D1, p53, p21WAF1, and PML, in Em Cas in vivo. In addition, the functional role of ß-catenin-TCF4 mediated transcription was also examined in Em Ca cells in vitro.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Histological findings were reviewed for 234 hysterectomy specimens of endometrioid type Em Cas in the case records of Kitasato University Hospital for the period from 1988 to 2002, according to the criteria of the World Health Organization histological classification (1994). Areas of SqD within tumors were subdivided into two categories, morules and squamous metaplasia (SqM) foci, as described previously.13,14
Briefly, the former were defined as areas consisting of spindle- to cuboidal-shaped cells forming growth sheets (Figure 1A)
and the latter were designated as foci consisting of cells with abundant eosinophilic cytoplasm and clear intercellular bridging (Figure 2A)
. The tumors with SqD areas investigated comprised 78 grade 1 and 2 grade 2 lesions, but no grade 3 carcinomas, including 45 (all grade 1) cases with morules and 35 (33 grade one 1 and 2 grade 2) with SqM foci. Some cases overlapped with those used in our previous study.13
Eighteen grade 1 carcinomas without SqD areas were also used as controls. All tissues were routinely fixed in 10% formalin and processed for embedding in paraffin wax.
|
|
Immunohistochemistry was performed using a combination of the microwave oven heating and standard streptavidin-biotin-peroxidase complex (LSAB kit; DAKO, Copenhagen, Denmark) methods. Briefly, slides were heated in 10 mmol/L citrate buffer (pH 6.0) for two 10-minute cycles using a microwave oven and then incubated overnight at 4°C with anti-ß-catenin mouse monoclonal (Transduction Laboratories, Lexington, KY), anti-cyclin D1 mouse monoclonal (DAKO), anti-p53 mouse monoclonal (DO7; Novocastra Lab. Ltd., Newcastle, UK), anti-p21WAF1 mouse monoclonal (WAF1; Calbiochem, Cambridge, MA), anti-human Ki-67 antigen rabbit polyclonal (DAKO), or anti-PML rabbit polyclonal (Chemicom, Temecula, CA) antibodies.
To determine labeling indices (LIs) for ß-catenin, cyclin D1, p53, p21WAF1, and Ki-67 nuclear immunoreactivity, immunopositive nuclei were counted for at least 1000 tumor cells in five randomly selected fields of glandular carcinoma components (surrounding carcinoma, Sur-Ca) and at least 700 cells in all of the SqD areas within tumors in each case, using high-power (x40 objective and x10 ocular) magnification. LIs were then calculated as numbers per 100 cells. Scoring of the PML immunoreactivity was also performed according to the method described previously.14,15 Briefly, the percentage of immunopositive cells in the total tumor cell population was subdivided into five categories as follows; 0, all negative; 1, <30% positive cells; 2, 30 to 50%; 3, 50 to 70%; 4, >70%. The immunointensity was also subclassified into four groups in comparison with infiltrating lymphocytes and stromal cells as positive internal controls, as follows: 0, negative; 1+, weak; 2+, moderate; 3+, strong. The internal controls were set as 3+. Immunoreactivity scores for both SqD and the Sur-Ca lesions for each case were produced by multiplication of the two values.
Polymerase Chain Reaction and Sequencing
Genomic DNA was extracted from cultured cells and 4-µm-thick paraffin wax sections of clinical samples, using proteinase K/phenol-chloroform methods. Exon 3 of the ß-catenin gene was amplified and sequenced as described previously.12,16
Cell Line and Plasmids
The Ishikawa Em Ca cell line17
was maintained in Eagle’s minimal essential medium with 10% bovine calf serum. The ß-catenin plasmid containing a deleted Ser-45 (ß-cat
S45) was kindly provided by Dr. Y. Nakamura (Institute of Medical Science, University of Tokyo, Tokyo, Japan). The cyclin D1 (–962CD1-luc) and the PML (PML-S-luc) promoter reporter plasmids were from Dr. O. Tetsu (University of California, San Francisco, CA) and Dr. A. Ben-Ze’ev (The Weizmann Institute of Science, Rehovot, Israel), respectively. The p14ARF (–3704 p14ARF-luc) promoter reporter construct bearing a 3407-bp fragment of p14ARF promoter region created from the p14ARF promoter gene (provided by Dr. K. Robertson, The University of Southern California, Los Angeles, CA) was subcloned into pGL2-basic (Promega, Madison, WI). TCF4 lacking a 30-amino acid ß-catenin binding site in its N-terminus (TCF4N30) plasmid (provided by Dr. S. Hirohashi, National Cancer Center Research Institute, Tokyo, Japan) was introduced into pcDNA3.1+ (Invitrogen, Tokyo, Japan). The p21 promoter reporter (p21-luc) and wild-type (wt) p53 expression (pCMVp53wt) plasmids were from Dr. C. Prives (Columbia University, New York, NY) and the PML expression (pSG5-PML) construct was from Dr. T. Sternsdorf (The Salk Institute for Biological Studies, California). Top-FLASH and Fop-FLASH (Upstate Biotechnology, Lake Placid, NY) were also applied.
Establishment of Cells with Stable and Inducible Mutant ß-Catenin Expression
To establish cells stably expressing mutant ß-catenin, the ß-catS45 construct and empty pcDNA3.1+ plasmids for comparison were transfected into Ishikawa cells using LipofectAMINE PLUS (Invitrogen). After 
2 weeks of culture in the presence of 1 mg/ml of Geneticin (Life Technologies, Inc., Grand Island, NY), the colonies stably expressing ß-catS45 were screened by reverse transcriptase-polymerase chain reaction assay using a combination of a forward primer lacking codon 45 (5'-CCACTACCACAGCTCCTCT-3' at exon 3) and a reverse primer (5'-TGAGCTCGAGTCATTGCATAC-3' at exon 4), and were confirmed by sequencing analysis.
A stable clone, capable of producing mutant ß-catenin, was also established using a tetracycline-regulated expression system (Invitrogen). Ishikawa cells were double-transfected sequentially with regulatory pcDNA6/TR and either responsive pcDNA4/TO-ß-catS45 or empty pcDNA4/TO plasmids using LipofectAMINE PLUS (Invitrogen). After selection of colonies with 5 µg/ml of blasticidin and 250 µg/ml of zeocin, those with induced expression of ß-catS45 in the presence of 1 µg/ml of tetracycline were screened and isolated in a similar manner.
Transient Transfection and Luciferase Reporter Assay
Cells were plated to form 60 to 80% confluent cultures in 24-well dishes. All transfection experiments were performed using the LipofectAMINE PLUS method (Invitrogen), in duplicate or triplicate, in accordance with the manufacturer’s instructions. The pRL-TK plasmid (Promega) was co-transfected to normalize for transfection efficiency and the total amount of transfected plasmid was made equal by addition of empty pcDNA3.1+ vector (Invitrogen). Luciferase activity was assayed 24 hours after transfection, using a dual-luciferase reporter assay system (Promega).
Western Blot Assays
Total cellular proteins were isolated using RIPA buffer [50 mmol/L Tris-HCl (pH7.2), 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate] containing protease inhibitor (phenylmethyl sulfonyl fluoride). Aliquots of 1 to 20 µg of total proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to a membrane, and probed with anti-cyclin D1 mouse monoclonal (sc-8396; Santa Cruz Biotechnology, Santa Cruz, CA), anti-p14ARF rabbit polyclonal (sc-8340, Santa Cruz Biotechnology), anti-p53 mouse monoclonal (DO7, Novocastra Lab. Ltd), anti-p21WAF1 mouse monoclonal (WAF1, Calbiochem), anti-PML rabbit polyclonal (Chemicon), and anti-ß-actin mouse monoclonal (Sigma Chemical Co., St. Louis, MO) antibodies, coupled with the ECL detection system (Amersham Pharmacia Biotech, Tokyo, Japan).
Immunofluorescence
Cells were grown to form 50 to 60% confluent cultures on glass slides. The monolayers were fixed with 3.7% paraformaldehyde for 10 minutes, permeabilized in 0.2% Triton X-100 for 10 minutes, then incubated with anti-ß-catenin mouse monoclonal antibody (Transduction Laboratories) for 1 hour at room temperature. Forty hours after transient transfection of ß-catS45 plasmid into Ishikawa cells or 1 µg/ml of tetracycline treatment for ß-catS45-inducible clones, cells were also stained with both anti-ß-catenin monoclonal (Transduction Laboratories) and anti-p53 polyclonal (CM1, Novocastra Lab. Ltd.) antibodies. The secondary antibodies were fluorescein isothiocyanate-labeled rabbit anti-mouse IgG and rhodamine-labeled goat anti-rabbit IgG (both from Molecular Probes, Leiden, The Netherlands).
Senescence-Associated ß-Galactosidase (SA-ß-gal) Assay
The ß-catS45-inducible and mock clones were grown on glass slides, and 2, 4, or 5 days after tetracycline treatment were stained for SA-ß-gal activity, in triplicate, as described previously.18
At least 500 cells were scored to determine the percentages of blue-stained cells.
Statistics
Comparative data were analyzed using the Mann-Whitney U-test, the Pearson’s correlation coefficient, and the chi-square test. The cutoff for statistical significance was set as P < 0.05.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Average ages of patients with Em Cas featuring areas of morules and SqM were 41.2 ± 10.5 years (mean ± SD) and 54.9 ± 11.1 years, respectively, the difference being significant (P < 0.0001) with a value of 56.9 ± 6.1 for tumors without SqD areas. Examples of immunohistochemical findings for ß-catenin, cyclin D1, p53, p21WAF1, Ki-67, and PML in Em Ca with morule and SqM are illustrated in Figures 1 and 2
.
Average LIs for both nuclear ß-catenin and cyclin D1 were significantly higher in morules than SqM and the Sur-Ca lesions. Significant stepwise decreases from morules, through SqM, to the Sur-Ca lesions were observed in LIs for both p53 and p21WAF1, but not Ki-67. PML immunoreactivity was significantly lower in morules than in Sur-Ca lesions, whereas SqM areas demonstrated significant increase. There were significant differences in LIs for nuclear ß-catenin, p53, and p21WAF1 between glandular components of tumors with and without SqD areas (Figure 3)
. Significantly positive correlations of LI values for nuclear ß-catenin, as well as cyclin D1, p53, and p21WAF1, were observed between SqD areas and the Sur-Ca lesions (Figure 4)
, while such associations were not evident for Ki-67 and PML (data not shown).
|
|
. None of the tumors without SqD showed any mutations in exon 3 of the gene. In morules and SqM areas, no significant differences in LI values for the markers investigated were observed when cases were subdivided into two groups on the basis of the ß-catenin gene status, with the exception of nuclear ß-catenin LI in SqM areas (Figure 5A)
. Similar findings were also noted in the Sur-Ca lesions (Figure 5B)
.
|
|
, nuclear ß-catenin LI values were positively correlated with LIs for cyclin D1, p53, and p21WAF1, and inversely with Ki-67 LIs and PML scores overall. Similar associations with cyclin D1 LIs and PML scores were also observed in SqD but not Sur-Ca categories. Positive correlations among LIs of p53, p21WAF1, and cyclin D1 were evident in all three categories, with some exceptions. There was no correlation of PML scores with either p53 or p21WAF1 LIs in any of the three categories (data not shown).
|
Based on sequence analysis of ß-catenin and p53 genes, Ishikawa cells with wild types of both genes were selected from five Em Ca cell lines, including Hec6, Hec88, Hec108, and Hec265 cells. The Ish-bcat#7 clone constitutively expressed the ß-catS45 mRNA, resulting in activation of ß-catenin-TCF transcription, as assessed with reference to TOP and FOP reporter constructs (Figure 6, A and B)
. On immunofluorescence analysis, strong nuclear ß-catenin accumulation was detected in 30% of cells of the stable clone, whereas immunoreactivity was limited to cell membranes in the mock case. Some Ish-bcat#7 cells with nuclear ß-catenin expression demonstrated increase in size, and were more spread and flattened and multinucleated but not squamoid in appearance (Figure 6C)
. In addition, increased expression of cyclin D1, p14ARF, p53, and p21WAF1 but not PML was evident in Ish-bcat#7 cells (Figure 6D)
.
|
In the presence of tetracycline, the Ish-tet#8 clone exhibited ß-catS45 mRNA expression, with activation of the TOP-reporter construct and elevated expression levels of cyclin D1, p14ARF, p53, and p21WAF1 but not PML, in a time-dependent manner (Figure 7
; A to C). Cells with inducible nuclear ß-catenin showed overexpression of endogenous p53, in line with the results of transient transfection of ß-catS45 into the parent cells (Figure 7D)
.
|
. Cells with squamoid features were never observed.
|
Transient transfection of ß-catS45 constructs caused a dose-dependent increase in reporter activity of the –3407 p14ARF-luc containing one perfect TCF binding site (position –2686, AACAATG) and two imperfect sites (position –686, CACAAAG and –579, AGCAATG A), the transcriptional activity being inhibited by co-transfection of the TCF4N30 plasmid with dose dependence (Figure 9, A and B)
, in line with the case for cyclin D1 reporter activity (Figure 9, C and D)
. Changes in PML promoter activity on transfection of ß-catS45 were relatively minor, whereas p21WAF1 promoter activity induced by wt p53 was suppressed by co-transfection of PML in a dose-dependent manner (Figure 9, E and F)
.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
; 2) most tumors investigated were categorized into the grade 1 endometrioid type, and p53 gene mutations are relatively rare, in contrast to the grade 3 or nonendometrioid types;19-21
3) most Sur-Ca lesions showed low p53 LI values (<10%), and high LIs (>30%) are typical (90%) of mutated cases.22
We demonstrated here that high levels of nuclear ß-catenin expression can be detected even in morules of tumors without gene mutations, in line with findings that gene mutations are not sufficient by themselves to cause nuclear accumulation.23
Given that most cases were early onset (average age, 41.2 ± 10.5 years) in the premenopausal phase, it is possible that endogenous progesterone may affect the regulation of nuclear ß-catenin expression, because the nuclear accumulation was closely linked with gene mutations in SqM areas within tumors in older patients (54.9 ± 11.1 years) (Figure 4A)
. Moreover, our previous results demonstrated a stabilization of nuclear and cytoplasmic ß-catenin by progesterone in Em Ca in vivo and in vitro.24
Other possible factors include alterations in ß-catenin degradation-related genes, such as APC and Axin, but such gene mutations are considered to be rare in Em Cas.23,25
It remains to be elucidated why expression levels of nuclear ß-catenin were relatively low in the Sur-Cas around morule lesions.
There are various lines of evidence that nuclear ß-catenin is involved in dedifferentiation or trans-differentiation during development or progression of gastric and colorectal carcinomas.26-28 In ß-catenin transgenic mice, stabilization of ß-catenin through the deletion of exon 3 results in development of squamous metaplasia of the mammary epithelium.29,30 These data prompted us to investigate the association between overexpression of ß-catenin and induction of SqD in Em Ca cells. Although Ishikawa cells overexpressing nuclear ß-catenin showed a senescence-like appearance, there were no squamoid elements, indicating that other factors, presumably including the tumor environment consisting of stromal cells and extracellular matrix, are also necessary for SqD. This conclusion is supported by the finding that specific signals from the microenvironment regulate intracellular ß-catenin distribution and subsequently the process of dedifferentiation and redifferentiation of colorectal tumor cells.28 One interesting result in our previous study was an increase in the percentage of SqD areas within Em Cas in patients receiving progesterone therapy.31
It was recently shown that deregulated ß-catenin elicits ARF-mediated p53-dependent growth arrest.11
Our in vitro results suggest that overexpression of ß-catenin might result in activated transcription from the p14ARF promoter in a TCF4-dependent manner (Figure 9, A and B)
, this being sufficient to induce p53 in nuclei (Figure 7D)
. Positive correlations among nuclear ß-catenin, p53, and p21WAF1 expression were also evident in vivo (Table 2)
. It is therefore conceivable that nuclear ß-catenin can activate the p53-p21WAF1 pathway, probably through the p14ARF-Mdm2-p53 regulatory loop,32
so that nuclear accumulation may be an initial signal for trans-differentiation toward the squamoid phenotype in Em Ca cells (Figure 10)
. This conclusion is supported by three lines of evidence. The first is that nuclear accumulation can be detected even in small squamoid nests consisting of a few cells, as described previously;13
the second is that expression levels of ß-catenin, p53, and p21WAF1 significantly differ between tumors with and without SqD areas (Figure 3)
; and the third is that there were significant correlations of ß-catenin and its target gene expression between SqD and the Sur-Ca lesions (Figure 4)
. With regard to the relative levels of nuclear ß-catenin and p53 in SqM areas, one possible explanation is the existence of an autoregulatory feedback loop between ß-catenin and p53, because p53 also induces ß-catenin degradation in response to growth arrest.33
Further studies of these points are clearly warranted.
|
In 293T cells, ß-catenin can activate transcription from the PML promoter, independent of TCF/LEF status, suggesting that it may operate by interacting with the basic transcriptional machinery.12
In our results, appreciable increase in PML promoter activity was not conferred by transfection of ß-catenin in Ishikawa cells (Figure 9E)
, in line with findings of the protein levels (Figures 6D and 7C)
, and an inverse correlation between nuclear ß-catenin and PML expression was evident in SqD areas (Table 2)
.
PML plays an important role in the p53-regulatory pathway for apoptosis and cell growth suppression without direct binding to p53-responsive elements of target genes, such as p21WAF1, Bax, and GADD45.36
Given the high levels of PML, p53, and p21WAF1 expression in SqM areas (Figure 3)
, an increase in p21 promoter activity was expected on co-transfection of wt p53 and PML, but our reporter assay demonstrated that PML caused suppression of p53-dependent p21 transcriptional activity in Ishikawa cells (Figure 9F)
. Similar findings have also been reported for MG63, human osteosarcoma cells.37
Thus, there are promoter-specific and cell-type-specific aspects to the functional cooperation between p53 and PML.
In conclusion, our findings imply that in Em Cas, nuclear ß-catenin can simultaneously induce activation of the p53-p21WAF1 pathway and overexpression of cyclin D1, leading to suppression of cell proliferation or induction of cell senescence. However, overexpression of ß-catenin alone is not sufficient for development of the squamoid phenotype of Em Ca cells, suggesting that nuclear accumulation is an initial signal for the trans-differentiation.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported, in part, by research grants from the Kanagawa Academy of Science and Technology and grants from the Kanzawa Medical Research Foundation.
Accepted for publication January 28, 2004.
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N Horree, P J van Diest, P van der Groep, D M D S Sie-Go, and A P M Heintz Progressive derailment of cell cycle regulators in endometrial carcinogenesis J. Clin. Pathol., January 1, 2008; 61(1): 36 - 42. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Saegusa, M. Hashimura, T. Kuwata, M. Hamano, Y. Wani, and I. Okayasu A functional role of Cdx2 in {beta}-catenin signaling during transdifferentiation in endometrial carcinomas Carcinogenesis, September 1, 2007; 28(9): 1885 - 1892. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
