| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Kimmel Cancer Center, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania 19107 [L. Y. Y. F., R. M., K. H.], and Division of Gastroenterology, Department of Genetics, Cancer Center, Abramson Family Cancer Research Institute, University of Pennsylvania, Philadelphia, Pennsylvania 19104 [H. N., A. K. R.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
ZS:WT. The high rate of cell proliferation was accompanied by overexpression of cell cycle progression and tumorigenesis biomarkers, including proliferating cell nuclear antigen, cyclin D1, cyclin-dependent kinase 4, p53, cytokeratin 14, epidermal growth factor receptor, and by a reduced rate of apoptosis. ZD substantially increased forestomach tumor incidence in TG mice: 85% of ZD:TG versus 14% of ZS:TG mice had forestomach tumors (P < 0.001), with progression to malignancy occurring only in ZD:TG tumors. Additionally, 14% of ZD:TG mice developed esophageal tumors and esophageal intestinal metaplasia at 77 days. Thus, cyclin D1 overexpression, in cooperation with ZD, decontrols cell proliferation, ensuring cell expansion, a prerequisite for cancer development. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The sequence of histopathological changes in ESCC development typically involves hyperplasia, mild to severe dysplasia, carcinoma in situ, and, finally, invasive carcinoma. Genetic steps that accompany such changes frequently include mutation of the TP53 tumor suppressor gene, disruption of cell cycle control in G1 involving the p16ink4a-cyclin D1/Cdk4-Rb pathway, activation of oncogenes such as EGFR, and inactivation of several tumor suppressor genes (7) . Notably, cyclin D1 overexpression because of gene amplification is a critical genetic alteration in ESCC (8 , 9) and in Barrett’s esophagus, a premalignant condition of esophageal adenocarcinoma (10) . Overexpression of cyclin D1 has been reported in esophageal papillomas and carcinomas isolated from rats at end point after multiple exposures to NMBA (11 , 12) , providing evidence that overexpression of cyclin D1 is also fundamental to the process of esophageal carcinogenesis in animal models.
Nakagawa et al. (13) developed a TG mouse in which the EBV ED-L2 promoter targets the transgene to the stratified squamous epithelium of tongue, esophagus, and forestomach, resulting in a dysplastic phenotype associated with increased cell proliferation (14) . However, TG mice, exposed to multiple doses of NMBA, showed increased severity of esophageal dysplasia after 12–15 months, indicating that cyclin D1 overexpression was not sufficient to elicit a tumorigenic response to induction by NMBA (15) , the most widely used agent for esophageal tumor induction in rodents (16) .
Our ZD-NMBA rat esophageal cancer model is an important tool to study molecular mechanism(s) underlying ESCC cancer development and prevention (17, 18, 19, 20, 21, 22, 23) . In this model, tumorigenesis is driven by increased cell proliferation, created in the esophageal epithelium by dietary ZD (18) ; thus, a single nontumorigenic NMBA dose (24) elicits an 80–95% tumor incidence in ZD esophagus after 12 weeks (19, 20, 21, 22, 23) . Tumor initiation (20) and reversal (23) is extremely rapid in ZD esophagus, allowing quasi-synchronous cellular responses to such processes to be precisely delineated. For example, 24 h after NMBA treatment, there was an expansion in the size of focal hyperplastic lesions in ZD esophageal epithelium, with abundant PCNA-positive cells. Adjacent sections displayed concurrent increases in cyclin D1, Cdk4, Rb expression, and reduced p16ink4a staining (20) , demonstrating a link between deregulation of the p16ink4a-cyclin D1/Cdk4-Rb pathway that regulates G1-S progression and rapid initiation of esophageal tumors.
ZD mice also develop tumors readily in response to NMBA (25 , 26) , with a lower incidence of esophageal than forestomach tumors; the rodent forestomach is considered as an extension of the lower esophagus (25) . Given that cyclin D1 overexpression (13, 14, 15) and dietary ZD (25 , 26) in mice are both connected with a high rate of esophageal/forestomach cell proliferation, we postulated that in combination, they may additionally deregulate cell cycle progression, thereby predisposing TG mice fed a ZD diet to NMBA-induced carcinogenesis.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Transgenic Mice.
The generation of TG mice was described previously (13)
. Breeding of TG males from a single founder line to B6 females generated a total of 65 TG mice for the tumorigenesis study. TG offspring were analyzed for presence of the transgene by genotyping of tail DNA using a PCR amplification method (13)
. WT B6 male mice were obtained from Jackson Laboratory (Bar Harbor, ME).
Experimental Design.
This study was approved by the Thomas Jefferson University Institutional Animal Care and Use Committee and conducted under NIH guidelines. Briefly, 4–5-week-old mice were randomized into two dietary groups and were fed ad libitum with a ZD or ZS diet and given free access to deionized drinking water, thus forming four experimental groups, ZD:TG, ZS:TG, ZD:WT, and ZS:WT. After 5 weeks, 7 mice from each group were sacrificed to determine the effect of dietary ZD on the extent of cell proliferation in the esophagus/forestomach. The remaining mice (10–14/group) received a single NMBA dose at 2 mg/kg body weight. At 25 days after NMBA treatment, a moribund ZD:TG mouse was autopsied and was found to harbor large fused tumors in the forestomach. Thus, we sacrificed half the transgenic mice at 25 days, and the remaining at 77 days for end point tumor incidence analysis.
Zinc Determination.
At autopsy, testis from male mice and hair from female mice were dried to constant weights at 90°C and ashed in a furnace. Ashed samples were dissolved in 0.1 N HCl and zinc levels determined by atomic spectrometry (26)
, using a Perkin-Elmer Atomic Absorption Spectrometer Analyst 100 (Perkin-Elmer, Norwalk, CT). Zinc content was expressed as µg/g dry weight. Zinc content in the testis (male) and hair (female) was significantly lower in ZD than ZS mice, independent of genotype or NMBA treatment. For example, in NMBA-untreated mice, testis and hair zinc content were significantly lower in ZD:TG versus ZS:TG mice (testis, 115 ± 10 versus 151 ± 6 µg/g; hair, 132 ± 12 and 189 ± 15 µg/g, P < 0.001), in agreement with previous studies (26)
.
Tumor Analysis.
At sacrifice, the animals were subjected to complete autopsies. Whole stomach and esophagus were excised and opened longitudinally. Tumors > 0.5 mm in diameter in the esophagus/forestomach were counted.
Isolation of Forestomach Tissue.
Whole esophagus/forestomachs were fixed in buffered formalin and embedded in paraffin. Cross-sections of the forestomach, including SCJ, were cut 4-µm thick. Serial sections were prepared and stained with H&E for histopathology or left unstained for immunohistochemical studies.
Cell Proliferation Determination by PCNA Immunohistochemistry.
Monoclonal mouse anti-PCNA (Santa Cruz Biotechnology, Santa Cruz, CA) was used at 1:500 dilution, followed by incubations with biotinylated goat antimouse antibody and streptavidin horseradish peroxidase. PCNA was localized by incubation with 3-amino-9-ethylcarbazole-substrate-chromogen system (Dako Corp.) and a light hematoxylin counterstain. Cells with red reaction product in the nucleus were considered positive for the presence of PCNA.
Immunohistochemical Detection of Cyclin D1, Cdk4, p53, Cytokeratin, EGFR, Bax, and Bcl-2.
After deparaffinization and rehydration in graded alcohols, sections were heated in citrate buffer (0.01 M, pH 6.0) in a microwave oven (85–90°C, 3 x 5 min) before nonspecific binding sites were blocked with goat/rabbit serum. Sections were incubated overnight at 37°C in a humidified chamber with respective primary antibodies: rabbit anticyclin D1 polyclonal antibody (Lab Vision Corp., Fremont, CA) at 1:100 dilution; rabbit anti-Cdk4 polyclonal antibody (Santa Cruz Biotechnology) at 1:150 dilution; rabbit anti-p53 polyclonal antibody (Novocastra Lab., Newcastle upon Tyne, United Kingdom) at 1:250; mouse anticytokeratin 14 monoclonal antiserum (Clone LL002; Novocastra Lab.) at a 1:40 dilution, goat anti-EGFR polyclonal antibody (Santa Cruz Biotechnology) at 1:60 dilution; rabbit anti-Bcl-2 polyclonal antiserum (Santa Cruz Biotechnology) at a 1:3000 dilution, and rabbit anti-Bax polyclonal antiserum (Santa Cruz Biotechnology) at a 1:800 dilution. Incubation with appropriate biotinylated secondary antibodies followed. Slides were then incubated with streptavidin horseradish peroxidase, and expression of individual proteins was localized by incubation with DAB and a light hematoxylin counterstain. The cyclin D1 monoclonal antibody used has no cross-reactivity with cyclin D2 or D3, and Cdk4 polyclonal antibody has no cross-reactivity with other Cdks.
Bcl-2:Bax Ratio.
The Bcl-2:Bax ratio was obtained by dividing the immunoreactive score of Bcl-2 by that of Bax. An immunoreactive score was calculated by multiplying the grade of percentage of positive cells by that of intensity of staining (27)
. The percentage of Bax- or Bcl-2-positive cells was graded semiquantitatively as follows: 0 = 0–5%; 1 = 6–25%; 2 = 26–50%; 3 = 51–75%; and 4 = 76–100%; the intensity of staining was graded as follows: 0 = none; 1 = weak; 2 = moderate; and 3 = intense. When heterogeneous staining intensities were found within one forestomach section, each component was graded independently, and the results were summed. For example, a specimen that contained 50% of cells with moderate intensity (2 x 2 = 4), 25% of cells with intense immunostaining (1 x 3 = 3), and 25% of cells with no staining (1 x 0 = 0) would receive an immunoreactive score 7 (4 + 3 + 0 = 7).
Apoptosis Detection.
Apoptosis was assessed by the ISOL method that uses T4 DNA ligase to specifically join DNase I-type ends from genomic DNA in the sample to biotin-labeled hairpin oligonucleotide probes (28
, 29) . These probes specifically and sensitively detect double-strand breaks in apoptotic cells. Although conventional in situ detection techniques such as terminal deoxynucleotidyl transferase-mediated nick end labeling are useful in detecting internucleosomal DNA cleavage, they do not differentiate Dnase I-type cleavage, which results from the activation of apoptotic endonucleaseas. Thus, the ISOL method is more selective in avoiding labeling of randomly damaged DNA.
ISOL Assay.
The double-strand breaks in DNA in apoptotic cells were detected using an ApopTag ISOL kit (Serological Corp., Norcross, GA). Briefly, the sections were deparaffinized, rehydrated in a graded alcohol series, and incubated with proteinase K. Endogenous peroxidase in the sections was inhibited with 3% hydrogen peroxide. The slides were then incubated with T4 DNA ligase enzyme to catalyze blunt end ligation of biotinylated Oligo B (blunt end oligo) and fragmented double-strand DNA (16°C–20°C for 17 h). Next, the slides were incubated with streptavidin peroxidase, and DNA fragmentation was detected by staining with DAB. Finally, the sections were counterstained with methyl green. Sections from rat mammary gland in which extensive apoptosis occurs served as a positive control. Negative controls omitted T4 DNA ligase enzyme.
Statistical Analysis.
Data on cell proliferation were analyzed by one-way ANOVA with the SAS statistical computer program (30)
. Tumor incidence differences were analyzed by two-tailed Fisher’s exact test (31)
. All statistical tests were two-sided and were considered statistically significant at P < 0.05.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. By contrast, transgenic ZS:TG mice on ZS diet, showed thin forestomach epithelia and slightly thickened SCJs (Fig. 1A, a)
, resembling ZS:WT forestomachs (data not shown; Ref. 26
), whereas, ZD:WT forestomachs exhibited thickened mucosa (data not shown; Ref. 26
). The forestomach/esophagus thickening was greatest in ZD:TG mice, followed by ZD:WT and then ZS:TGZS:WT. These results indicate that cyclin D1 overexpression in TG mice, together with dietary ZD, produced greater thickening in forestomach than observed with either condition alone, suggesting the combination led to unbridled cell proliferation.
|
shows that as early as 25 days after a single NMBA dose, 85% (11 of 13) of ZD:TG mice developed tumors in forestomach/SCJ, with 46% showing fused tumors, often occupying the entire forestomach (Fig. 1A, d)
and 39% with large individual tumors (1.5–2.5 mm in diameter). Conversely, 86% (12 of 14) of ZS:TG mice were tumor-free (Fig. 1A, c)
, and 14% (2 of 14) had very small forestomach tumors (<1 mm in diameter, data not shown), a result at variance with a previous study in which NMBA-exposed TG mice without zinc restriction developed dysplasia without cancer (15)
. This variation is probably attributable to a difference in the route of administration of NMBA, s.c. versus intragastric administration in this study. There was no increase in tumor incidence in either group when the end point was extended to 77 days (Table 1)
. At both 25 and 77 days, ZD:TG mice had significantly higher tumor incidence than ZS:TG mice (P < 0.001), indicating that ZD increases the incidence of tumors in transgenic mice overexpressing cyclin D1. On the other hand, a single NMBA dose induced very small forestomach tumors in 20% (2 of 10) of ZD:WT at 77 days but 0% (0 of 10) of ZS:WT mice; the difference was not statistically significant. The lack of effect of ZD on forestomach tumorigenesis in WT mice is attributable to the single rather than multiple NMBA doses (25)
. Notably, there was a significant difference in forestomach tumor incidence between transgenic mice on a ZD diet (ZD:TG versus ZD:WT, P = 0.003) but not those on a ZS diet (Table 1)
. These results suggest that the accelerated and enhanced NMBA-induced forestomach tumorigenesis observed in ZD:TG mice was driven by joint efforts of ZD and cyclin D1 overexpression.
|
, and squamous cell carcinomas (Fig. 1B, c and e)
. Conversely, ZS:TG forestomach exhibited mostly hyperplasia, with infrequent occurrence of dysplasia and papillomas (data not shown). These results are consistent with the tumorigenesis data presented for these two groups of animals in Table 1
. In addition, ZD:TG mice had significantly higher incidence of forestomach squamous cell carcinomas than ZS:TG animals at both end points [4 of 13 (31%) versus 0 of 14 (0%) at 25 days, P = 0.04; and 5 of 14 (36%) versus 0 of 10 (0%) at 77 days, P = 0.05]. These data showed that in the ZD condition, tumor progression to malignancy was very rapid in TG mice.
Esophageal Lesions in NMBA-treated ZD:TG Mice.
Macroscopic esophageal lesions were not detected in ZD:WT mice after a single NMBA dose, a result consistent with our previous reports that, unlike the rat, mouse esophagus is less sensitive to NMBA carcinogenicity than forestomach (25
, 26)
. However, small esophageal tumors were detected in 8% (1 of 13) and 14% (2 of 14) of ZD:TG mice at 25 and 77 days, respectively, but not in ZS:TG mice (P > 0.05). H&E-stained sections of ZD:TG esophagi typically showed the occurrence of a thickened epithelium with focal hyperplastic lesions (Fig. 1B, d)
. An example of an early papilloma is shown in Fig. 1B, b
. In addition, esophageal intestinal metaplasia was detected in 2 ZD:TG mice, 1 at 25 days (Fig. 1B, f)
and the other 77 days (data not shown). By contrast, esophagi from ZS:TG mice at both end points exhibited none of these lesions, and ZS:TG esophageal epithelia were typically 2–5 cells thick with mild to moderate basal cell hyperplasia (data not shown).
Unchecked Cell Proliferation in Untreated ZD:TG Forestomach.
We determined whether a combination of cyclin D1 overexpression in TG mice and ZD resulted in greater abnormalities in cell cycle-related proteins than observed with either condition alone. We accomplished this by immunohistochemical analyses to compare the expression of relevant biomarkers, including PCNA, cyclin D1, Cdk4, p53, EGFR, and cytokeratin 14 [an important biomarker of cell differentiation in esophageal carcinogenesis (26
, 32)
], in the forestomach of ZS:TG versus ZD:TG mice and in control ZS:WT and ZD:WT animals. We present detailed results from TG forestomachs only, except in the case of EGFR, where expression patterns have not been reported for ZD:WT forestomach/esophagus.
Fig. 2a
shows that ZS:TG forestomach epithelium was covered by a thin keratin layer and typically showed 2–5 cell thick mucosa with mild folds and basal cell hyperplasia, a phenotype in between that of a ZS:WT and ZD:WT mouse (data not shown; Ref. 32
). In stark contrast, ZD:TG epithelium, covered by a thick keratin layer, was very thick with deep mucosal down-growth and severe focal hyperplastic lesions (Fig. 2b)
, greater than that observed in ZD:WT counterparts (data not shown; Ref. 32
). Expression of PCNA, an endogenous cell proliferation marker, was mainly in the basal cell layer in ZS:TG forestomach (Fig. 2c)
but was abundant and detected in many cell layers in the highly hyperplastic ZD:TG epithelium (Fig. 2d)
. As expected, expression of cyclin D1 and its catalytic partner, Cdk4, was strong in both ZS:TG and ZD:TG forestomachs but was scattered in the former (cyclin D1, Fig. 2e
; Cdk4, Fig. 2g
) and abundant in many cell layers in the latter (cyclin D1, Fig. 2f
; Cdk4, Fig. 2h
). Expression of p53, a key protein in cell cycle checkpoint, was very weak in ZS:TG (Fig. 2i)
but moderate and plentiful in ZD:TG forestomach (Fig. 2j)
. Cytoplasmic expression of cytokeratin 14, a biomarker in human (33)
and mouse esophageal tumorigenesis (26)
, although strongly detected in both ZS:TG and ZD:TG forestomach, was restricted to the basal cell layer of the former (ZS:TG, Fig. 2k
) but present in the many cell layers of the latter (ZD:TG, Fig. 2l
). In general, the level of expression in each instance was highest in ZD:TG forestomach, followed by ZD:WT and then ZS:TG>ZS:WT.
|
. A slight increase in EGFR expression was detected in the basal cells of ZS:TG esophagi (Fig. 3c)
, consistent with results reported by Mueller et al. (14)
. By contrast, proliferative ZD:WT and ZD:TG esophagi displayed strong EGFR staining in many cell layers, with intensity stronger in ZD:TG (Fig. 3g)
than ZD:WT (Fig. 3e)
. A similar staining pattern was found in the forestomach: EGFR expression was strong and intense in many cell layers of the proliferative ZD:TG forestomach (Fig. 3h)
, followed by ZD:WT (Fig. 3f)
and then ZS:TG (Fig. 3d)
and weakest in ZS:WT forestomach (Fig. 3b)
.
|
. PCNA-LI (%) of cells in S phase, a measure of cellular proliferation, is significantly higher in ZD:WT than ZS:WT tissues (forestomach, 37 ± 2.9 versus 29 ± 2.5, P < 0.001; esophagus, 35 ± 2.7 versus 28 ± 3.4, P < 0.01) and in ZD:TG than ZS:TG tissues (forestomach, 45 ± 7.6 versus 32 ± 4.4, P < 0.001; esophagus, 40 ± 2.2 versus 30 ± 3.7, P < 0.01). In addition, forestomach and esophagus cell proliferation was significantly higher in ZD:TG than ZD:WT mice (forestomach and esophagus, P < 0.01, Fig. 4
), demonstrating that the highest rate of cell proliferation was achieved when cyclin D1 overexpression in TG mice was combined with ZD.
|
Reduced Apoptosis in Untreated ZD:TG Forestomach.
To determine whether apoptosis was reduced in the highly proliferative, untreated ZD:TG forestomach relative to its ZS:TG counterpart, the ApopTag ISOL kit (Serological Corp.) that specifically stains DNA fragmentation or double-strand breaks in apoptotic cells was used. As shown in Fig. 5
, ZD:TG epithelium displayed sparse occurrence of ISOL-positive nuclei, mainly in the outermost cell layers (Fig. 5d)
. Conversely, ZS:TG forestomachs regularly showed ISOL-positive nuclei in basal and suprabasal cell layers, as shown in Fig. 5a
. Bcl-2, an antiapoptotic protein, was strongly expressed in the proliferative ZD:TG epithelium (Fig. 5e)
but moderately in ZS:TG epithelium (Fig. 5b)
. On the contrary, Bax, a proapoptotic protein, was weakly expressed in the hyperplastic ZD:TG epithelium but moderately in ZS:TG epithelium. A semiquantitative immunostaining analysis of ZD:TG forestomach produced immunoreactive scores of 3.8 for Bcl-2 and 2.0 for Bax expression, resulting in a Bcl-2/Bax ratio of 1.9 (Table 2)
. Corresponding ZS:TG forestomach had a score of 2.4 and 2.5 for Bcl-2 and Bax, respectively, producing a Bcl-2:Bax ratio of 1. Thus, an inverse relationship between Bcl-2 and Bax, favoring cell proliferation, was established in ZD:TG but not in ZS:TG forestomach, a result consistent with the higher proliferative rate observed in the former. Additionally, at 25 days after NMBA treatment, the Bcl-2:Bax ratio was increased 
2-fold in ZD:TG but remained nearly unchanged for ZS:TG forestomach, in line with the high tumor incidence detected in ZD:TG versus ZS:TG mice (Table 1)
.
|
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, Fig. 1
). Without the added proliferation provided by ZD, the combination of cyclin D1 overexpression and NMBA was insufficient to initiate esophageal tumorigenesis or mount a vigorous tumorigenic response in the forestomach. The occurrence of forestomach tumors in ZD:TG mice was significantly different from the corresponding ZS:TG group (P < 0.001, 25 and 77 days) and from ZD:WT mice (P = 0.003, 77 days, Table 1
), affirming that collaboration between cyclin D1 overexpression and ZD are required for rapid tumor initiation by NMBA. Previous reports of carcinogenesis studies with mice overexpressing cyclin D1 indicate cellular and tissue context is important. In tissues such as esophagus (15 , 38) and skin keratinocytes (39 , 40) , additional factors are required to promote cancer, whereas in liver (41) and breast (42) , cyclin D1 alone is sufficient to elicit a tumorigenic response. For example, combined cyclin D1 overexpression and multiple exposures to NMBA resulted only in increased cell proliferation and severity of esophageal and forestomach squamous dysplasia after 15 months, without cancer development (15) , whereas cooperation between cyclin D1 overexpression and p53 deficiency led to histological evidence of severe oral-esophageal squamous epithelial dysplasia and cancer within 6 months of age (38) . Collectively, these studies indicate that the level of cyclin D1 expression is not a rate-limiting factor, but rather, for cyclin D1 overexpression to be tumorigenic, either on its own or in combination with other agents, a high rate of cell proliferation must be sustained (38 , 40, 41, 42) .
This study showed a high rate of forestomach cell proliferation in untreated mice overexpressing cyclin D1 on a ZD diet, greater than that occurred with either condition alone (Fig. 4)
. This high rate of proliferation was accompanied by overexpression of several key biomarkers in cell cycle progression (Fig. 2)
. For example, the hyperplastic ZD:TG epithelium displayed abundant PCNA-positive nuclei in many cell layers concurrently with overexpression of cyclin D1, Cdk4, p53, and cytokeratin 14 (Fig. 2
, right panels). Conversely, ZS:TG epithelium with PCNA-positive nuclei limited to basal and directly suprabasal cell layers showed moderate expression of cyclin D1, Cdk4, and cytokeratin 14 but weak expression of p53 in basal cell layers (Fig. 2
, left panels). In general, expression of PCNA and that of other biomarkers was greatest in ZD:TG followed by ZD:WT and then ZS:TG>ZS:WT mice, affirming an association between unbridled cell proliferation and deregulation of cell cycle progression.
The EGFR signaling pathway contributes to a number of processes important to tumor development, including cell proliferation, apoptosis, angiogenesis, and metastasis (43)
. EGFR overexpression occurs frequently in foci of human esophageal squamous dysplasia, a precancerous lesion of ESCC (44)
and organotypic culture of primary human esophageal squamous epithelial cells overexpressing EGFR showed epithelial cell proliferation and hyperplasia, with migration of cells from the basal compartment to the suprabasal compartment (45)
. We have now demonstrated for the first time that the increased cell proliferation induced in esophagus/forestomach by ZD is associated with EGFR overexpression (esophagus, Fig. 3e
versus Fig. 3a
; forestomach, Fig. 3f
versus Fig. 3b
), a result that helps to explain the phenomenon of sustained cell expansion in ZD target tissues. Strikingly, ZD mice overexpressing cyclin D1 exhibited an even more intense and extensive EGFR overexpression than ZD:WT mice, extending from basal to the outermost cell layers of the hyperplastic target tissues (esophagus, Fig. 3g
versus Fig. 3e
; forestomach, Fig. 3h
versus Fig. 3f
). By contrast, EGFR staining in ZS:TG tissues was only slightly increased compared with ZS:WT and was confined to basal cell layers (esophagus, Fig. 3c
versus Fig. 3a
; forestomach, Fig. 3d
versus Fig. 3b
), consistent with a previous report by Mueller et al. (14)
. Thus, the combination of cyclin D1 overexpression and ZD led to overexpression of EGFR in ZD:TG forestomach, resulting in a hyperplastic and dysplastic microenvironment, conducive to rapid tumor initiation by NMBA. In this regard, EGFR signaling provides a survival signal in murine skin tumor development by activating an antiapoptotic pathway that inhibits keratinocyte differentiation, thereby keeping basal cells in a proliferative state (46)
.
An antiapoptotic pathway was induced in the proliferative ZD:TG forestomach epithelium. Compared with untreated ZS:TG forestomach, the rate of apoptosis was very low in ZD:TG forestomach. For example, in ZS:TG forestomach, apoptotic cells were regularly found in the basal and suprabasal cell layers (Fig. 5a), whereas in the hyperplastic ZD:TG forestomach epithelium, apoptotic cells were typically scattered and scarce (Fig. 5d), indicating that apoptosis was decreased. Accordingly, the immunoreactive score for Bcl-2, an antiapoptotic protein, was significantly higher in untreated ZD:TG than ZS:TG forestomach (3.8 versus 2.4), whereas that for Bax, a proapoptotic protein, was lower in ZD:TG than ZS:TG forestomach, translating to a high Bcl-2:Bax immunoreactive ratio for ZD:TG forestomach (ZD:TG versus ZS:TG, 1.9 versus 1), favorable to cell proliferation. At 25 days after NMBA treatment, tumor-bearing ZD:TG forestomach showed an additional increase in the Bcl-2:Bax immunoreactive ratio (3.5) whereas that for ZS:TG forestomach remained at the same level as in untreated forestomach. These data point to a microenvironment in which cell proliferation was preferred via inhibition of apoptosis, providing a platform for rapid tumor initiation by NMBA.
We have, for some time, been developing the mouse model of esophageal/forestomach cancer by manipulating expression of esophageal cancer-relevant proteins such as p53 and Fhit, important in activation of intrinsic apoptotic pathways, so that their loss greatly increased esophageal tumor susceptibility (26 , 47) .
Absence of the TP53 tumor suppressor gene rendered ZD mice exquisitely susceptible to NMBA-induced carcinogenesis (26) . The rapid rate of tumor induction/progression in ZD:p53-/- mice after a single NMBA dose was accompanied by an increase in the rate of cell proliferation and a decrease in apoptosis. Conversely, overexpression of AZ arrested NMBA-induced forestomach tumorigenesis in ZD mice (32) . AZ is a multifunctional regulator of polyamine metabolism that inhibits ornithine decarboxylase activity and restricts polyamine levels (48, 49, 50) , thereby inhibiting cell proliferation (51) . The inhibition of tumor development in ZD mice by AZ overexpression was associated with suppression of cell proliferation and stimulation of apoptosis (32) . Thus, in a background of high cellular proliferation provided by ZD that is known to accelerate tumorigenesis, AZ overexpression has the opposite effect of p53 loss.
It is likely that crossing recombinant mice so that overexpression of cyclin D1, underexpression of p53, and/or Fhit in the same mouse strain, in addition to allowing dissection of pathways to esophageal cancer, will generate a truer model of human esophageal cancer for prevention and therapy studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by American Institute for Cancer Research Grant 02A025-REN (to L. Y. Y. F.), the Italian-American Cancer Foundation Fellowship (to R. M.), NIH Grants DK3377 (to A. K. R.) and NIH P01 DE12467 (to A. K. R.), and by National Cancer Institute Grants CA77738 (to K. H.) and CA53036. 
2 To whom requests for reprints should be addressed, at Thomas Jefferson University, Jefferson Medical College, Kimmel Cancer Institute, 233 South 10th Street, Philadelphia, PA 19107-6799. Phone: (215) 503-4656; Fax: (215) 923-3528.
3 The abbreviations used are: ESCC, esophageal squamous cell carcinoma; TG, cyclin D1 transgenic; WT, wild-type; ZD, zinc deficient; ZS, zinc sufficient; NMBA, N-nitrosomethylbenzylamine; Cdk, cyclin-dependent kinase; Rb, retinoblastoma; B6, C57BL/6; SCJ, squamocolumnar junction between fore- and hindstomach; PCNA, proliferating cell nuclear antigen; EGFR, epidermal growth factor receptor; ISOL, in situ oligo ligation method; LI, labeling index; AZ, antizyme; DAB, 3,3'-diaminobenzidine tetrahydrochloride.
Received 3/14/03. Accepted 5/ 6/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-Difluoromethylornithine inhibits N-nitrosomethylbenzylamine-induced esophageal carcinogenesis in zinc-deficient rats: effects on esophageal cell proliferation and apoptosis. Cancer Res., 58: 5380-5388, 1998.-Difluoromethylornithine induction of apoptosis: a mechanism which reverses pre-established cell proliferation and cancer initiation in esophageal carcinogenesis in zinc-deficient rats. Cancer Epidemiol. Biomark. Prev., 10: 189-198, 2001.
and their receptor in the human oesophagus. Histochem. J., 29: 745-758, 1997.[Medline]
This article has been cited by other articles:
|
|
 |
|
  C.-G. Liu, L. Zhang, Y. Jiang, D. Chatterjee, C. M. Croce, K. Huebner, and L. Y.Y. Fong Modulation of Gene Expression in Precancerous Rat Esophagus by Dietary Zinc Deficit and Replenishment Cancer Res., September 1, 2005; 65(17): 7790 - 7799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Y. Y. Fong, L. Zhang, Y. Jiang, and J. L. Farber Dietary Zinc Modulation of COX-2 Expression and Lingual and Esophageal Carcinogenesis in Rats J Natl Cancer Inst, January 5, 2005; 97(1): 40 - 50. [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 |
