Role of estrogen receptor β in colonic epithelium

  1. Osamu Wada-Hiraike*,†,
  2. Otabek Imamov*,
  3. Haruko Hiraike†,
  4. Kjell Hultenby‡,
  5. Thomas Schwend*,
  6. Yoko Omoto*,
  7. Margaret Warner*, and
  8. Jan-Åke Gustafsson*,§
  1. *Department of Biosciences and Nutrition and
  2. ‡Clinical Research Centre, Karolinska Institute, Novum, S-141 86 Huddinge, Sweden; and
  3. †Department of Obstetrics and Gynecology, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan

  1. Contributed by Jan-Åke Gustafsson, December 29, 2005

 
Next Section

Abstract

Several papers report that the colon is one of the tissues regulated by estrogen receptor (ER)β. To better understand the physiological role of ERβ in colonic tissue, we have compared morphology, proliferation, and differentiation of colonic epithelium in ERβ−/− mice and WT littermates. BrdUrd labeling revealed that the number of proliferating cells was higher in ERβ−/− mice and that the migration of labeled cells toward the luminal surface was faster in ERβ−/− mice than in WT littermates. Additionally, in the absence of ERβ, there was a decrease in apoptosis, which was measured by immunohistochemical staining of cleaved caspase-3. The state of differentiation of the colonic epithelial cells was studied by using epithelial markers. In ERβ−/− mice, there was a significant decrease in the expression of the differentiation marker cytokeratin (CK)20 and in the cellular adhesion molecules α-catenin (an adherens junction protein) and plectin (a hemidesmosomal protein). These changes were also evident by electron microscopy as abnormalities in tight junctions and in the number and shape of desmosomes in ERβ−/− mice. These findings suggest a role for ERβ in the organization and architectural maintenance of the colon. Furthermore, our results indicate that the rapidly proliferating cells of the colonic epithelium in ERβ−/− mice are lost by increased shedding and not by increased apoptosis. In this way, hyperproliferative cells that lack ERβ do not form hyperplastic lesions and do not accumulate in the superficial epithelium.

Targeted disruption of estrogen receptor (ER)β in mice (1) has revealed roles for ERβ in many tissues and organs, including the ovary, uterus, mammary gland, ventral prostate, salivary gland, immune system, and central nervous system (2–6). In some tissues, both ERα and ERβ are expressed, and the specific function of each receptor is difficult to evaluate, especially in cases in which the two receptors oppose each other’s functions. The mature ventral prostate epithelium expresses ERβ, not ERα (6), and in ventral prostates of ERβ−/− mice there are reports of hyperplastic foci, epithelial hyperproliferation, decreased apoptosis, and accumulation of incompletely differentiated cells in an intermediate cellular pool (2). These results suggest a possible direct influence of estrogens on prostatic epithelium mediated by ERβ, and evidence for such a role has come from the antiproliferative effects of ERβ-selective agonists on rodent prostate (7).

Estrogens are used for the reduction of postmenopausal symptoms and for preservation of bone mineral density. Recent epidemiological studies suggest that combined estrogen and progestogen hormone replacement therapy reduces the incidence of colorectal cancer (CRC) in postmenopausal women (8–11). Another epidemiological study shows that the withdrawal of estrogen may increase the risk of microsatellite instability-positive CRC (12). Therefore, hormone replacement therapy has a significant role in reduction of CRC, and estrogens seem to have a protective effect against CRC through ER.

CRC is one of the very common malignancies in industrialized countries, and 875,000 or more people are diagnosed with CRC annually (13). The prognosis for patients with CRC is heavily dependent on the stage at diagnosis, and almost all patients require surgical resection. The 5-year survival rate is >90% for Dukes’ stage A but only 5% for Dukes’ stage D. Therefore, early diagnosis and prevention against CRC is extremely important. CRC is thought to develop in a sequence from aberrant crypt proliferation or benign hyperplasia to benign adenoma and then, in most cases, to adenocarcinoma (14). It is not known where in the progression of this disease estrogen plays its beneficial role, and a detailed analysis of the protective role of hormone replacement therapy and the function of ER against CRC is not yet available.

ERβ plays a multifaceted role in the functional differentiation of various epithelial and nonepithelial cell types. Previous studies have revealed that ERβ is the predominant ER expressed in colonic tissues (15, 16), that the ERβ gene is methylated in 90% of colon cancer tissues (17), and that expression of ERβ is selectively lost in human malignant colon tissue (18). Additionally, an ERβ-selective agonist has been reported to be an effective treatment in animal models of inflammatory bowel disease (19). To understand the role of ERβ in the regulation of normal colonic tissue development, we have compared the colon of WT mice with that of ERβ−/− mice with respect to morphological characteristics and markers of proliferation, apoptosis, adhesion, and differentiation.

Previous SectionNext Section

Results

ERβ−/− Colon Phenotype.

To examine the effects of the loss of ERβ on colon morphology, colon sections from ERβ−/− mice and their WT littermates were stained with hematoxylin/eosin and Alcian blue. In 4-month-old ERβ−/− mice, there was more lymphoid infiltration below the epithelium, especially in proximal colon, than in WT mice (Fig. 1 A and B). However, the total number of lymphoid aggregates, also known as gut-associated lymphoid tissue, was not affected by loss of ERβ. By 12 months of age, the lymphoid infiltration in the ERβ−/− mouse colon was much less severe. Alcian blue staining revealed that in ERβ−/− mice (Fig. 1 D and F), the location of mucin was very different from that in WT mice (Fig. 1 C and E), especially in the distal colon. In the upper crypts and sometimes at the top of the crypts in ERβ−/− mouse colon, mucin appeared in large globs (Fig. 1 D and F). Other histological features such as crypt length, crypt area, nuclear density, number of cells per crypt, and occurrence of aberrant crypt foci in colonic epithelium were not changed by loss of ERβ.

Fig. 1.
View larger version:
Fig. 1.

Morphological comparison between WT and ERβ−/− littermates and the expression of ERβ in WT mice. (A and B) Hematoxylin/eosin-stained sections of the colon from 4-month-old WT (A) and ERβ−/− (B) littermates show that the proximal part of the ERβ−/− colon exhibits subepithelial infiltration of lymphoid tissue. (C–F) Alcian blue-stained sections of the colon from 4-month-old WT (C and E) and ERβ−/− mice. D and F show that the acidic mucin arrangement in ERβ−/− mice is not as well organized as it is in WT littermates. Intermediate (C and D) and distal (E and F) parts of colon are shown. (G–I) The expression of ERβ in WT mice was confirmed with ERβ-specific antibodies. Positively stained cells (brown) are predominantly located in the superficial epithelium. Proximal (G), intermediate (H), and distal (I) parts of colon are shown.


Expression of ERβ, evaluated by immunohistochemical studies with a specific ERβ antibody, revealed that in WT mice, the expression of ERβ was predominantly restricted to epithelial cells at the luminal surface of colonic mucosa (Fig. 1 G–I).

Measurement of Epithelial Cellular Proliferation and Migration.

In the present study, we injected the thymidine analog BrdUrd to label S-phase cells. At least 10 intact, histologically normal, and well oriented crypts from each animal were evaluated. For each unit, BrdUrd-stained cells and the most apical cells from the lumen of colonic mucosa were counted. We found that the number of BrdUrd-labeled cells in ERβ−/− mice was 1.6-fold higher than in WT littermates (Fig. 2 A). To evaluate whether cellular migration rates were altered in ERβ−/− mice, we measured the distance traveled by the labeled cells upward from the base of the crypts. In ERβ−/− mice, BrdUrd-labeled cells had migrated much closer to the luminal surface than they did in WT littermates (Fig. 2 B). These data indicate that in ERβ−/− mice, more cells were in S phase and that cells were shifting more rapidly toward the top of the crypt. However, despite the rapid proliferation and upward movement of cells, there was no cellular accumulation in the superficial epithelium.

Fig. 2.
View larger version:
Fig. 2.

Incorporation of BrdUrd into DNA of the colonic cells in 12-month-old WT and ERβ−/− littermates. Labeled cells are present throughout the crypt. (A) The number of BrdUrd-positive cells was counted, and the data are shown in the histogram. (B) The rate of migration of BrdUrd-positive cells from the base to the luminal surface of the crypt was evaluated and expressed as the number of unlabeled cells preceding the first BrdUrd-labeled cell on the side of the crypt. In ERβ−/− mice, BrdUrd-stained cells are closer to the lumen of the colon. (C) Representative BrdUrd stainings of WT and ERβ−/− mouse crypts.


Apoptotic Activity in ERβ−/− and WT Mice.

To examine apoptosis in the colonic epithelium, an antibody specific for cleaved caspase-3 was used to detect activation of caspase-3 in situ in WT and ERβ−/− mice. Apoptotic cells were found only at the luminal surface of the colonic epithelium in WT mice. In ERβ−/− mice, there were very few positive signals in the epithelium (Fig. 3). In both genotypes, positive immunohistochemical staining at the apex of the crypts confirmed that fully differentiated cells are lost through apoptosis. No apoptotic cells were detected in the lower parts of the crypts where cells were proliferating and moving toward the luminal surface.

Fig. 3.
View larger version:
Fig. 3.

Expression of cleaved caspase-3 in the colon of WT and ERβ−/− mice. Apoptosis in the colon of WT and ERβ−/− littermates was studied by observing the expression of cleaved caspase-3. Apoptotic signals are seen in the superficial epithelial cells. In both WT (A and B) and ERβ−/− (C and D) mice, there are immunohistochemically stained cells only at the top of the crypt. The expression of cleaved caspase-3 is significantly lower in ERβ−/− mice than in WT littermates.


Differentiation of the Colonic Epithelium in ERβ−/− and WT Littermates.

The pattern of differentiation of the colonic epithelium in the absence of ERβ signaling was studied by immunohistochemical staining and Western blotting. In accordance with previous literature, in WT mouse colon, cytokeratin (CK)20 expression was essentially absent from cells at the base of the crypt. Expression of CK20 was found in the more differentiated suprabasal regions, mainly on the upper surface and in scattered cells on the surface (20). In ERβ−/− mice, there was an overall decrease in CK20 expression (Fig. 4 A–D). In contrast, expression of CK7, which is preferentially expressed in cells at the base of the crypts, was similar in WT and ERβ−/− mice, as judged from immunohistochemical staining (data not shown). Western blotting was used to confirm the reduction in CK20 expression. Expression of CK20 in ERβ−/− mouse colon was significantly decreased in both nuclear and cytosolic fractions (Fig. 4 E).

Fig. 4.
View larger version:
Fig. 4.

Expression of CK20 in the colon of WT and ERβ−/− mice. The differentiation marker CK20 is confined to luminal cells in the colon of WT (A and B) and ERβ−/− (C and D) mice. In ERβ−/− mice, expression of CK20 is lower than it is in WT mice. (E) Western blotting of colonic nuclear and cytosolic fractions confirmed the immunohistochemical observation of a decreased expression of CK20 in ERβ−/− mice.


The Adherens Junction of Colonic Epithelium in ERβ−/− Mice.

E-cadherin uses its cytoplasmic domain to couple to catenins and the actin cytoskeleton to form a complex known as the adherens junction complex. These catenins are β-catenin, p120 (γ-catenin), and α-catenin. α-Catenin provides a direct link to the actin filament system by binding to the other catenins. Expression of catenins in the colonic epithelium in the presence or absence of ERβ was studied by immunohistochemical staining and Western blotting. The analysis revealed no difference in expression of E-cadherin (Fig. 5 A–C) or β-catenin (Fig. 5 A, F, and G) except for increased cytoplasmic β-catenin in ERβ−/− mice. However, there was a substantial decrease in α-catenin expression in the membrane fraction of ERβ−/− mouse epithelium. No difference was detected in the nuclear fraction (Fig. 5 A). Immunohistochemistry revealed a clearly evident decrease in the membrane staining of α-catenin in ERβ−/− mice (Fig. 5 D and E).

Fig. 5.
View larger version:
Fig. 5.

Expression of adherens junction molecules in the colon of WT and ERβ−/− littermates. (A) Western blotting of colonic nuclear, cytosolic, and membrane fractions. Levels of E-cadherin and β-catenin were comparable in WT and ERβ−/− mice except for the slight increase in β-catenin in the cytosolic fraction in ERβ−/− mice. Expression of α-catenin was decreased in the membrane fraction of ERβ−/− mouse colon. (B–G) The expression of these adherens junction molecules was confirmed by immunohistochemical staining. (B and C) E-cadherin. (D and E) α-Catenin. (F and G) β-Catenin. (B, D, and F) WT. (C, E, and G) ERβ−/−. α-Catenin expression in ERβ−/− mouse colon (E) is considerably lower than it is in WT littermates (D).


The Hemidesmosome Protein Plectin in ERβ−/− Mice.

To evaluate whether there were overall differences in protein expression in the colon of WT and ERβ−/− mice, we examined protein profiles in various subcellular fractions of the colon epithelium by using SDS/PAGE. We found that three proteins with molecular masses greater than 250 kDa were decreased in ERβ−/− mouse colon. By using a proteomic method (MALDI-TOF MS), we confirmed that these proteins were plectin and α- and β-spectrin (data not shown). The most versatile cytoskeletal linker protein known to date is plectin. We confirmed the expression of plectin by immunofluorescence. Specific cytoplasmic and perinuclear signals were observed at the top of the crypts in WT mice. In ERβ−/− mouse colon, overall staining was weak, and perinuclear staining was markedly lower than that seen in WT littermates (Fig. 6).

Fig. 6.
View larger version:
Fig. 6.

Expression of hemidesmosomal protein plectin. The expression of plectin was examined by immunofluorescence. There is prominent expression of plectin (bright staining) in the cytoplasm and perinuclear area in WT mice (Left). Staining in ERβ−/− mice is much reduced (Right).


The Desmosome and Tight Junction of Colonic Epithelium in ERβ−/− Mice.

EM studies were done to investigate ultrastructural differences between WT and ERβ−/− mouse colons. Morphologically, the polarity of epithelial cells and the number of crypts were normal in WT mice (Fig. 7 A). The epithelial cells in ERβ−/− mouse colon appeared more tightly packed (Fig. 7 B). In sections perpendicular to the plane of the epithelium, normal tight junctions were clearly visible in WT mouse colon (Fig. 7 C). The junctions were located at the apical surface between all of the luminal epithelial cells and were all the same length. In ERβ−/− mice, tight junctions were very different from those of WT littermates: In many fields that were examined, there was an abrupt loss of tight junctions between cells at the apical side of colonic epithelium (Fig. 7 D). EM also revealed abnormal desmosomes in ERβ−/− mouse colon (Fig. 7 F). In ERβ−/− mice, a common finding was that the cytoplasmic sides of the desmosomes were not well matched between adjacent cells. At the same time, there were fewer desmosomes in ERβ−/− (Fig. 7 F) mice than in WT littermates (Fig. 7 E).

Fig. 7.
View larger version:
Fig. 7.

Tight junctions and desmosomes in colonic epithelium examined by EM. Lower magnification of the colonic epithelium in 4-month-old WT (A) and ERβ−/− (B) littermates shows the tightly packed colonic epithelium in ERβ−/− mice. Tight junctions are morphologically normal in WT mice, and there is a clear border between cells (C). However, tight junctions are not visible between most cells in ERβ−/− mice (D). In ERβ−/− mice (F), there were fewer desmosomes than in WT mice (E), and they were of irregular, asymmetrical shape.


Previous SectionNext Section

Discussion

Here we present evidence for a physiological role of ERβ in colonic tissue, where it is expressed in the superficial epithelium. Contrary to expectations, morphological changes in ERβ−/− mice were minimal. Crypt shape, number of cells per crypt, gut-associated lymphoid aggregation, and aberrant crypt foci were comparable to those of WT littermates. We found a remarkable amount of subcellular lymphoid aggregation in ERβ−/− mouse colon. This finding may be attributed to the fact that the hematopoietic status of ERβ−/− mice was abnormal because these mice have been shown to have myelogenous hyperplasia in the bone marrow (5). We previously showed that by 1.5 years of age, ERβ−/− mice exhibit a myeloproliferative disease resembling human chronic myeloid leukemia with lymphoid blast crisis (5). Considering that ERβ is important for the maintenance of hematopoietic stem cell quiescence, infiltration of lymphatic and eosinophilic compartments into the colon is not unexpected. What was unexpected was the reduction of infiltration with age. In the human population, the age of inflammatory bowel disease onset is relatively early, i.e., childhood or adolescence (21). The transient feature of subcellular lymphoid aggregation in ERβ−/− mice may be related to the age and hormonal status of the mice. Further investigations into inflammatory bowel disease in the absence of ERβ signaling are clearly required.

We observed a disorganization of mucin localization in ERβ−/− mice. Interestingly, in cancer, alteration of mucin expression and secretion occurs during malignant transformation (22, 23). Even though the disorganized mucin localization in ERβ−/− mice may be unrelated to tumor formation, it may help to explain the rapid colonic cellular proliferation in ERβ−/− mice discussed below.

Cells of the colon are continuously renewed through a process initiated by stem cell division. The daughter cells produced differentiate and migrate from the bottom to the top of the crypt, lose the capacity to divide, and are shed within several days. Immunohistochemical staining showed prominent expression of ERβ in colonic superficial epithelium, prompting us to suggest that ERβ has a significant role in differentiation. In the hyperproliferative state, a progressive accumulation of somatic mutations occurs in the stem cells, which have a long residence period in the mucosa. This cumulative damage may contribute to the development of a malignant phenotype that possesses a growth advantage over other stem cells within the crypt. In the present study, we found that BrdUrd-labeled cells were increased 1.6-fold in ERβ−/− mice. We have previously shown that in the ventral prostate epithelium, in the absence of ERβ, the percentage of proliferating cells increases 3.6-fold (2). This difference in epithelial cellular proliferation rate between prostatic and colonic tissue in ERβ−/− mice may be attributed to the fact that colonic tissue growth is much faster than that of ventral prostate. The appearance of labeled cells closer to the luminal surface in colons of ERβ−/− mice indicates that proliferating cells shift toward the top of the crypt more rapidly than in WT littermates.

The pattern of differentiation of the colonic epithelium in the absence of ERβ signaling was studied by immunohistochemical staining and Western blotting. CK20 expression in ERβ−/− mice was decreased as judged from both Western blotting and immunohistochemistry. A previous study on human colon adenocarcinoma showed that decline in ERβ (16) and CK20 (24, 25) expression paralleled the loss of differentiation of the colon cancer. Another study showed a correlation between loss of ERβ expression and type of Dukes’ stage (26). ERβ is now believed to have a protective role against colon cancer development (17, 18). The present study shows that loss of ERβ is correlated with loss of differentiation. At the luminal surface of the colon, senescent epithelial cells are normally eliminated by apoptosis or detachment. The mechanisms involved in the detachment include an increased repulsive force or a decreased adhesion between cells and the extracellular matrix at the site of shedding. Increased proliferation and decreased apoptosis are the main features of neoplasia. There were very few apoptotic cells in the colonic epithelium of ERβ−/− mice. In our current model, deficiency of ERβ leads to hyperproliferation, loss of differentiation, and decreased apoptosis in the epithelium of colon. However, surprisingly, no tumor formation was observed in ERβ−/− mice. Our hypothesis to explain this paradox is that the cells at the top of the crypt are detaching rapidly in ERβ−/− mice, resulting in the normal appearance of colonic crypt.

To test this hypothesis, we examined the adhesion molecules in ERβ−/− mice. The destabilization of cellular integrity and cytoarchitecture observed as a consequence of α-catenin and plectin decrease and malpositioning, especially of the epithelium of the colonic tissue, provides us with compelling evidence for this hypothesis. α-Catenin is required to mediate the formation of adherens junctions in epithelial cells and is a key protein in the formation of the radial cables that are necessary for stabilizing adherens junctions, sealing membranes, and assembling epithelial sheets (27). Plectin acts as an integrator and stabilizer of cellular architecture (28). It links cytoskeletal filaments to each other and to other organelles in the interior of the cytoplasm and attaches cytoskeletal filament networks to the plasma membrane. EM findings confirmed the dysfunction in the epithelium of the colonic tissue. Instead of standard features characteristic of tight junctions, the lateral surface contacts between epithelial cells of ERβ−/− mice were irregular and of abnormal shape. Desmosomes are crucial for maintenance of normal adhesion function of epithelial cells. In ERβ−/− mice, there were fewer desmosomes, and those that were present were of irregular, asymmetrical shape. Given these changes in adherens junctions and hemidesmosomes, it is clear that adhesion of epithelial cells in ERβ−/− mice is impaired. Failure of epithelial cells to adhere to each other might be the cause of the rapid loss of cells from the luminal surface and the increased cell turnover at the base of the crypts.

Estrogen has diverse biological effects, and many of these result from a direct interaction of estrogen with a nuclear receptor that activates the expression of genes encoding proteins with important biological functions. Diet, lifestyle, and other nongenetic factors such as gut flora (29) are thought to have a strong impact on CRC risk. Among them, phytoestrogens are believed to play a role as oncoprotective agents and to lower the risk of CRC in rodent models. Lignans, plant precursors of the phytoestrogens enterolactone and enterodiol, or lignan-rich food can inhibit colon cancer development in animal models (30). Enterolactone is a weak ER agonist (31). Because ERβ is the only ER in the colonic epithelium, there is reason to speculate that phytoestrogens exert their protective effect by interacting with ERβ.

The present study shows that ERβ seems to be essential for maintenance of cellular homeostasis and for driving cellular differentiation in the colon. Further investigations are needed to elucidate the molecular mechanisms that underlie the signaling of ERβ in normal cellular growth, thereby evaluating ERβ as a possible target for modalities of prevention and medical treatment of CRC.

Previous SectionNext Section

Materials and Methods

Animals.

Animals were used in accordance with the guidelines for care and use of experimental animals issued by Stockholm’s Södra Djurförsöksetiska Nämnd. Mice were fed a soy-free diet and allowed to drink tap water ad libitum. Mice were bred from heterozygous mice. Genotyping using PCR was performed on DNA isolated from the tails of 2-week-old mice as described elsewhere. Mice were killed by cervical dislocation. For immunohistochemical studies, colons were removed, rinsed in ice-cold PBS, slit open longitudinally, and fixed flat between wet filter papers overnight in 4% paraformaldehyde. The colons were routinely embedded in paraffin wax, and 4-μm sections were mounted on organosilane-coated slides. For Western blotting, enterocytes were isolated by lightly scraping the mucosal surface with the edge of a glass slide. This material was then washed in cold PBS containing proteinase inhibitor mixture (Roche Diagnostics, Mannheim, Germany) and centrifuged at low speed. The resulting pellets were snap-frozen and kept at −80°C until they were analyzed. The efficacy of isolating crypt-epithelial cells was confirmed by light microscopy. No fibroblasts or muscle cells were observed in the preparation.

Chemicals and Antibodies.

BrdUrd was purchased from Roche. Chicken polyclonal antibody anti-ERβ 503 IgY was produced in our laboratory. Rabbit antibodies were anti-cleaved caspase-3 (cleaved at Asp-175; Cell Signaling Technology, Beverly, MA, catalog no. 5A1); anti-α-catenin (Santa Cruz Biotechnology, catalog no. sc-7894), and anti-β-catenin (Invitrogen). Mouse polyclonal antibodies anti-E-cadherin (catalog no. 610181), anti-plectin (catalog no. 611348), and mouse monoclonal antibody anti-BrdUrd (catalog no. 3D4) were purchased from BD Biosciences Pharmingen. Mouse monoclonal antibodies anti-CK20 (catalog no. RCK105) and anti-CK7 (catalog no. SPM140) were purchased from Abcam, Ltd. (Cambridge, U.K.). Biotinylated goat anti-rabbit IgG, anti-mouse IgG, and avidin–biotin complex kits were from Vector Laboratories.

Treatment with BrdUrd.

For measurement of proliferation and migration, eight mice were treated i.p. with BrdUrd at 30 mg/kg every 12 h for 3 days (total of six injections). They were killed 12 h after the last injection.

Histology and Immunohistochemistry.

The presence of aberrant crypt foci in the epithelium was investigated after staining the colonic mucosa with 0.2% methylene blue dissolved in 70% ethanol. Hematoxylin/eosin staining was used for histological evaluation under the light microscope. Alcian blue staining (pH 2.5) was used for the evaluation of mucin production. Methylene blue and Alcian blue 8GX were purchased from Sigma/Aldrich. For immunohistochemistry, 15 WT and 15 ERβ−/− 4- to 12-month-old mice were used. Paraffin sections (4 μm) were dewaxed in xylene and rehydrated through graded ethanol to water. Antigens were retrieved by boiling in 10 mM citrate buffer (pH 6.0) for 30 min. The cooled sections were incubated in methanol containing 0.3% H2O2 for 30 min to quench endogenous peroxidase. To block the nonspecific binding, sections were incubated in PBS containing 3% BSA and 0.5% Nonidet P-40 for 10 min at room temperature (RT). Sections were then incubated with the following antisera: anti-BrdUrd (1:800), anti-cleaved caspase-3 (1:50), anti-α-catenin (1:200), anti-β-catenin (1:800), anti-CK20 (1:50), anti-CK7 (1:50), and anti-E-cadherin (1:200) in 3% BSA overnight at 4°C. Negative controls were incubated with only 3% BSA without primary antibody. The avidin–biotin complex method was used to visualize the signal, according to the supplier’s manual (Vector Laboratories). The sections were incubated in appropriate biotinylated Ig solution (1:200) for 1 h at RT, followed by washing with PBS and incubation in avidin–biotin-horseradish peroxidase for 1 h. After washing in PBS, sections were developed with 3,3′-diaminobenzidine tetrahydrochloride substrate (DAKO), lightly counterstained with Mayer’s hematoxylin, dehydrated through ethanol series and xylene, and mounted. For immunofluorescence detection, sections were incubated with anti-plectin IgG (1:100) for 1 h at RT. Secondary antibody was Cy-3-conjugated anti-mouse IgG (1:100). After incubation for 1 h, the sections were dipped in DAPI solution (0.1 μg/ml in PBS) for 20 sec at RT to delineate nuclei and mounted in Vectashield antifading medium (Vector Laboratories).

Western Blotting.

Nuclear, membrane, or cytosolic fractions were separated on 4–20% Tris/glycine SDS gels (Invitrogen) and transferred onto a poly(vinylidene difluoride) membrane. The membrane was blocked with blocking buffer (10% nonfat milk in Tris-buffered saline, pH 7.5, containing 0.1% Nonidet P-40), followed by incubation with the primary antibody at RT for 1 h. The membrane was washed three times with Tris-buffered saline. Signals were detected by using a peroxidase-conjugated secondary antibody, and enhanced chemiluminescence detection (Amersham Pharmacia Biosciences) was performed according to the supplier’s recommendation.

EM.

Colons were dissected, and small pieces were cut and immediately fixed in 2% glutaraldehyde/0.5% paraformaldehyde in 0.1 M sodium cacodylate buffer (caco)/0.1 M sucrose/3 mM CaCl2 (pH 7.4) overnight. Specimens were rinsed in 0.15 M caco and postfixed by incubation for 2 h in 2% osmium tetroxide in 0.07 M caco containing 3 mM CaCl2. The specimens were dehydrated in an ascending series of alcohol into acetone and embedded in LX-112 epoxy resin (Ladd Research Industries, Burlington, VT). Semithin sections (0.5 μm) were placed on glass slides, stained with toluidine blue, and examined in a light microscope. Ultrathin sections were cut and contrasted with uranyl acetate followed by lead citrate and examined in a Tecnai 10 (FEI, Eindhoven, The Netherlands) transmission electron microscope at 80 kV.

Previous SectionNext Section

Acknowledgments

This study was supported by the Ministry of Education, Science, and Culture of Japan; a Schering grant from the Japan Menopausal Society (to O.W.-H.); and grants from the European Union Network of Excellence CASCADE, the Swedish Cancer Fund, and KaroBio AB.

Previous SectionNext Section

Footnotes

  • §To whom correspondence should be addressed. E-mail: jan-ake.gustafsson{at}mednut.ki.se

  • Author contributions: O.W.-H., Y.O., M.W., and J.-A.G. designed research; O.W.-H., K.H., T.S., and J.-A.G. performed research; O.W.-H., O.I., and M.W. analyzed data; and O.W.-H., H.H., M.W., and J.-A.G. wrote the paper.

  • Conflict of interest statement: No conflicts declared.

  • Abbreviations:
    CK,
    cytokeratin;
    CRC,
    colorectal cancer;
    ER,
    estrogen receptor;
    RT,
    room temperature.
Previous Section
 

References

  1. ↵
    1. Kuiper G. G.,
    2. Enmark E.,
    3. Pelto-Huikko M.,
    4. Nilsson S.,
    5. Gustafsson J.-Å.
    (1996) Proc. Natl. Acad. Sci. USA 93:5925–5930, pmid:8650195.
    Abstract/FREE Full Text
  2. ↵
    1. Imamov O.,
    2. Morani A.,
    3. Shim G. J.,
    4. Omoto Y.,
    5. Thulin-Andersson C.,
    6. Warner M.,
    7. Gustafsson J.-Å.
    (2004) Proc. Natl. Acad. Sci. USA 101:9375–9380, pmid:15187231.
    Abstract/FREE Full Text
  3. ↵
    1. Nilsson S.,
    2. Mäkelä S.,
    3. Treuter E.,
    4. Tujague M.,
    5. Thomsen J.,
    6. Andersson G.,
    7. Enmark E.,
    8. Pettersson K.,
    9. Warner M.,
    10. Gustafsson J.-Å.
    (2001) Physiol. Rev 81:1535–1565, pmid:11581496.
    Abstract/FREE Full Text
  4. ↵
    1. Förster C.,
    2. Mäkelä S.,
    3. Wärri A.,
    4. Kietz S.,
    5. Becker D.,
    6. Hultenby K.,
    7. Warner M.,
    8. Gustafsson J.-Å.
    (2002) Proc. Natl. Acad. Sci. USA 99:15578–15583, pmid:12438700.
    Abstract/FREE Full Text
  5. ↵
    1. Shim G. J.,
    2. Wang L.,
    3. Andersson S.,
    4. Nagy N.,
    5. Kis L. L.,
    6. Zhang Q.,
    7. Mäkelä S.,
    8. Warner M.,
    9. Gustafsson J.-Å.
    (2003) Proc. Natl. Acad. Sci. USA 100:6694–6699, pmid:12740446.
    Abstract/FREE Full Text
  6. ↵
    1. Weihua Z.,
    2. Mäkelä S.,
    3. Andersson L. C.,
    4. Salmi S.,
    5. Saji S.,
    6. Webster J. I.,
    7. Jensen E. V.,
    8. Nilsson S.,
    9. Warner M.,
    10. Gustafsson J.-Å.
    (2001) Proc. Natl. Acad. Sci. USA 98:6330–6335, pmid:11371645.
    Abstract/FREE Full Text
  7. ↵
    1. Neubauer B. L.,
    2. McNulty A. M.,
    3. Chedid M.,
    4. Chen K.,
    5. Goode R. L.,
    6. Johnson M. A.,
    7. Jones C. D.,
    8. Krishnan V.,
    9. Lynch R.,
    10. Osborne H. E.,
    11. Graff J. R.
    (2003) Cancer Res 63:6056–6062, pmid:14522935.
    Abstract/FREE Full Text
  8. ↵
    1. Rossouw J. E.,
    2. Anderson G. L.,
    3. Prentice R. L.,
    4. LaCroix A. Z.,
    5. Kooperberg C.,
    6. Stefanick M. L.,
    7. Jackson R. D.,
    8. Beresford S. A.,
    9. Howard B. V.,
    10. Johnson K. C.,
    11. et al.
    (2002) J. Am. Med. Assoc 288:321–333.
    Abstract/FREE Full Text
  9. ↵
    1. Nelson H. D.,
    2. Humphrey L. L.,
    3. Nygren P.,
    4. Teutsch S. M.,
    5. Allan J. D.
    (2002) J. Am. Med. Assoc 288:872–881.
    Abstract/FREE Full Text
  10. ↵
    1. Farquhar C. M.,
    2. Marjoribanks J.,
    3. Lethaby A.,
    4. Lamberts Q.,
    5. Suckling J. A.
    (2005) Cochrane Database Syst. Rev, CD004143, pmid:16034922.
  11. ↵
    1. Chlebowski R. T.,
    2. Wactawski-Wende J.,
    3. Ritenbaugh C.,
    4. Hubbell F. A.,
    5. Ascensao J.,
    6. Rodabough R. J.,
    7. Rosenberg C. A.,
    8. Taylor V. M.,
    9. Harris R.,
    10. Chen C.,
    11. et al.
    (2004) N. Engl. J. Med 350:991–1004, pmid:14999111.
    Abstract/FREE Full Text
  12. ↵
    1. Slattery M. L.,
    2. Potter J. D.,
    3. Curtin K.,
    4. Edwards S.,
    5. Ma K. N.,
    6. Anderson K.,
    7. Schaffer D.,
    8. Samowitz W. S.
    (2001) Cancer Res 61:126–130, pmid:11196149.
    Abstract/FREE Full Text
  13. ↵
    1. Potter J. D.
    (1999) J. Natl. Cancer Inst 91:916–932, pmid:10359544.
    Abstract/FREE Full Text
  14. ↵
    1. Vogelstein B.,
    2. Fearon E. R.,
    3. Hamilton S. R.,
    4. Kern S. E.,
    5. Preisinger A. C.,
    6. Leppert M.,
    7. Nakamura Y.,
    8. White R.,
    9. Smits A. M.,
    10. Bos J. L.
    (1988) N. Engl. J. Med 319:525–532, pmid:2841597.
    Abstract
  15. ↵
    1. Campbell-Thompson M.,
    2. Lynch I. J.,
    3. Bhardwaj B.
    (2001) Cancer Res 61:632–640, pmid:11212261.
    Abstract/FREE Full Text
  16. ↵
    1. Konstantinopoulos P. A.,
    2. Kominea A.,
    3. Vandoros G.,
    4. Sykiotis G. P.,
    5. Andricopoulos P.,
    6. Varakis I.,
    7. Sotiropoulou-Bonikou G.,
    8. Papavassiliou A. G.
    (2003) Eur. J. Cancer 39:1251–1258, pmid:12763213.
    CrossRefMedlineISI
  17. ↵
    1. Issa J. P.,
    2. Ottaviano Y. L.,
    3. Celano P.,
    4. Hamilton S. R.,
    5. Davidson N. E.,
    6. Baylin S. B.
    (1994) Nat. Genet 7:536–540, pmid:7951326.
    CrossRefMedlineISI
  18. ↵
    1. Foley E. F.,
    2. Jazaeri A. A.,
    3. Shupnik M. A.,
    4. Jazaeri O.,
    5. Rice L. W.
    (2000) Cancer Res 60:245–248, pmid:10667568.
    Abstract/FREE Full Text
  19. ↵
    1. Harris H. A.,
    2. Albert L. M.,
    3. Leathurby Y.,
    4. Malamas M. S.,
    5. Mewshaw R. E.,
    6. Miller C. P.,
    7. Kharode Y. P.,
    8. Marzolf J.,
    9. Komm B. S.,
    10. Winneker R. C.,
    11. et al.
    (2003) Endocrinology 144:4241–4249, pmid:14500559.
    Abstract/FREE Full Text
  20. ↵
    1. Zhou Q.,
    2. Toivola D. M.,
    3. Feng N.,
    4. Greenberg H. B.,
    5. Franke W. W.,
    6. Omary M. B.
    (2003) Mol. Biol. Cell 14:2959–2971, pmid:12857878.
    Abstract/FREE Full Text
  21. ↵
    1. Oliva-Hemker M.,
    2. Fiocchi C.
    (2002) Inflamm. Bowel Dis 8:112–128, pmid:11854610.
    CrossRefMedlineISI
  22. ↵
    1. Aksoy N.,
    2. Corfield A. P.,
    3. Sheehan J. K.
    (2000) Clin. Biochem 33:167–173, pmid:10913513.
    CrossRefMedlineISI
  23. ↵
    1. Corfield A. P.,
    2. Clamp J. R.,
    3. Casey A. D.,
    4. Paraskeva C.
    (1990) Int. J. Cancer 46:1059–1065, pmid:2249893.
    MedlineISI
  24. ↵
    1. Wildi S.,
    2. Kleeff J.,
    3. Maruyama H.,
    4. Maurer C. A.,
    5. Friess H.,
    6. Buchler M. W.,
    7. Lander A. D.,
    8. Korc M.
    (1999) Clin. Cancer Res 5:2840–2847, pmid:10537351.
    Abstract/FREE Full Text
  25. ↵
    1. Cross H. S.,
    2. Bajna E.,
    3. Bises G.,
    4. Genser D.,
    5. Kallay E.,
    6. Potzi R.,
    7. Wenzl E.,
    8. Wrba F.,
    9. Roka R.,
    10. Peterlik M.
    (1996) Anticancer Res 16:2333–2337, pmid:8694565.
    MedlineISI
  26. ↵
    1. Jassam N.,
    2. Bell S. M.,