Zinc deficiency potentiates induction and progression of lingual and
esophageal tumors in p53-deficient mice
Louise Y.Y.Fong
1,, Yubao Jiang
2and John L.Farber
2Department of Molecular Virology, Immunology and Medical Genetics, 1Comprehensive Cancer Center, Ohio State University, Columbus, OH 43210, USA and2Department of Pathology, Anatomy and Cell Biology, Thomas Jefferson University, Philadelphia, PA 19107, USA
To whom correspondence should be addressed at: Comprehensive Cancer Center, Ohio State University, Room 388A, Tzagournis Medical Research Facility, 420 W. 12th Avenue, Columbus, OH 43210, USA.
Tel:+1 614 688 5914; Fax: +1 614 688 4020; Email: Louise.Fong@osumc.edu
Upper aerodigestive tract (UADT) cancer, including oral
and esophageal cancer, is an important cause of cancer
deaths worldwide. Patients with UADT cancer are
freq-uently zinc deficient (ZD) and show a loss of function of
the pivotal tumor suppressor gene p53. The present study
examined whether zinc deficiency in collaboration with p53
insufficiency (p53+/) promotes lingual and esophageal
tumorigenesis in mice exposed to low doses of the
carcino-gen 4-nitroquinoline 1-oxide. In wild-type mice, ZD
signi-ficantly increased the incidence of lingual and esophageal
tumors from 0% in zinc sufficient (ZS) ZS:p53+/+ mice to
40%. On the p53+/ background, ZD:p53+/ mice had
significantly greater tumor incidence and multiplicity than
ZS:p53+/ and ZD:p53+/+ mice, with a high frequency
of progression to malignancy. Sixty-nine and 31% of
ZD:p53+/ lingual and esophageal tumors, respectively,
were squamous cell carcinoma versus 19 and 0% of
ZS:p53+/ tumors (tongue, P ¼ 0.003; esophagus, P ¼
0.005). Immunohistochemical analysis revealed that the
increased cellular proliferation observed in preneoplastic
lingual and esophageal lesions, as well as invasive
carcino-mas, was accompanied by overexpression of cytokeratin 14,
cyclooxygenase-2 and metallothionein. In summary, a new
UADT cancer model is developed in ZD:p53+/ mouse that
recapitulates aspects of the human cancer and provides
opportunities to probe the genetic changes intrinsic to
UADT carcinogenesis and to test strategies for prevention
and reversal of this deadly cancer.
Introduction
Upper aerodigestive tract (UADT) cancer, including
esopha-geal and tongue tumors, is an important cause of morbidity and
mortality worldwide (1). The incidence of UADT cancer is
increasing worldwide, including that in young adults and those
without the known risk factors of tobacco and alcohol use (2).
The prognosis of esophageal cancer is poor, with a 5-year
survival rate of only 10%. The survival with oral cancer,
the major site being the tongue, is equally dismal. Patients
with oral cancer have a high mortality rate, because of field
cancerization effects that result in second primary tumors,
particularly in the esophagus (3,4). In addition, patients with
oral cancer are frequently zinc deficient (ZD) (5,6), a condition
associated with an increased risk for esophageal squamous cell
cancer (ESCC) (7,8). Abnet et al. (9) established a direct
connection between zinc deficiency and human ESCC, by
using X-ray fluorescence spectroscopy to measure zinc,
copper, iron, nickel and sulfur in esophageal biopsy samples
obtained from residents in a high ESCC incidence area in
China. Subjects were matched on baseline histology and
followed for 16 years: 90% of subjects in the highest zinc
quartile versus 65% in the lowest quartile were cancer-free
for 16 years. No associations with ESCC cancer risk were
found for any of the other elements studied. These findings
in humans are consistent with our conclusion from rodent
model studies that zinc deficiency promotes esophageal
cancer (10–12).
We have developed ZD rodent cancer models and found that
zinc deficiency creates a precancerous condition in the rat
UADT by causing unrestrained cell proliferation (11–13)
and extensive changes in gene expression, including
upregu-lated expression of cyclooxygenase-2 (COX-2), keratin 14
(KRT14) and metallothionein-1 (MT-1) (13,14).
Overexpres-sion of COX-2 has been reported in a variety of human
prema-lignant and maprema-lignant lesions, including UADT cancer
(15–17). MT protein is overexpressed in human squamous
cell carcinoma (SCC) of the esophagus (18) and tongue (19).
KRT14 is a biomarker of human and rodent esophageal
car-cinogenesis (20,21). A ZD diet in rats accelerates
carcinogen-esis in the esophagus and forestomach that results from a single
exposure to the carcinogen of N-nitrosomethylbenzylamine
(NMBA) (22,23) and at multiple sites in the UADT with
con-tinuous exposure to the carcinogen 4-nitroquinoline 1-oxide
(NQO) (13). On the other hand, zinc replenishment rapidly
reverses cell proliferation, stimulates apoptosis, corrects
abnormal gene expression in esophageal epithelium and
inhib-its tumorigenesis (13,14,24).
The p53 tumor suppressor protein plays a pivotal role in
preventing uncontrolled cellular proliferation. A loss of
func-tion of the TP53 tumor suppressor gene is demonstrated in over
50% of all human cancers, including oral and esophageal
cancer (25). Our previous work shows that zinc deficiency
modulates the enhanced genetic susceptibility to esophageal
cancer by NMBA in p53-deficient mice (21).
The mouse tongue is not sensitive to chemical
carcinogen-esis. Mice exposed to 10 p.p.m. NQO in the drinking water for
50 weeks (26) or 20 p.p.m. for 8 weeks and killed at week 24
(27) did not develop any lesions in the tongue. NQO at high
concentrations of 50 and 100 p.p.m. in the drinking water,
however, induced both esophageal and tongue tumors in
mice, with progression to malignancy (27). In the present
Abbreviations:COX, cyclooxygenase; DAB, 3,30-diaminobenzidine tetrahydrochloride; H&E, hematoxylin and eosin; KRT14, keratin 14; MT, metallothionein; NQO, 4-nitroquinoline 1-oxide; NMBA, N-nitrosomethyl-benzylamine; PCNA, proliferating cell nuclear antigen; SCC, squamous cell carcinoma; UADT, upper aerodigestive tract; ZD, zinc deficient; ZS, zinc sufficient.
study, we tested the hypothesis that zinc deficiency increases
cellular proliferation in the murine tongue, as it does in the rat
tongue (13), and that zinc deficiency in collaboration with p53
insufficiency (p53+/) promotes lingual and esophageal
tumorigenesis in ZD:p53+/ mice exposed to low doses of
NQO in the drinking water (26,27). In addition, we examined
whether an association exists between increased cellular
proli-feration and overexpression of the tumor markers KRT14,
COX-2 and MT in premalignant and malignant lesions during
UADT carcinogenesis.
Materials and methods
Chemicals and diets
NQO was from Wako Chemicals, USA (Richmond, VA). Custom-formulated, egg white-based ZD and zinc sufficient (ZS) diets containing 1.5 and 70 p.p.m. zinc, respectively, were prepared by Teklad (Madison, WI). The deficient diet is nutritionally complete and is identical to ZS diet except for the concentration of elemental zinc (21).
P53-deficient mice
Breeder pairs, homozygous C57BL/6J-Trp53 males and heterozygous C57BL/ 6J-Trp53 females that have been backcrossed for 11 generations, were pur-chased from The Jackson Laboratory (Bar Harbor, ME) to generate p53/ and p53+/ mice for this study. The p53+/ and p53/ offsprings were differentiated by genotyping of tail DNA using a PCR-based method (28). Wild-type C57BL/6J controls were obtained from The Jackson Laboratory. Experimental design
This study was approved by the Institutional Laboratory Animal Care of and Use Committee of the Thomas Jefferson University, Philadelphia, PA and conducted under National Institutes of Health guidelines. The experiment was conducted in three batches when the animals became available from our breeding colonies. Each batch contained all treatment groups listed below. Four-week-old mice were housed 3–5 to a polycarbonate cage with a stainless-steel wire floor. The animals had free access to deionized drinking water. The mice were randomized into two dietary groups and were fed ad libitum a ZD or control ZS diet, forming six experimental groups: ZD:p53/ (n ¼ 15), ZS:p53/ (n ¼ 14), ZD:p53+/ (n ¼ 16), ZS:p53+/ (n ¼ 26), ZD:p53+/+ (n ¼ 14) and ZS:p53+/+ (n ¼ 12). After 3 weeks, all animals were switched to drinking water containing 20–10 p.p.m. NQO for 21 weeks: 20 p.p.m. for the first 3 weeks and 10 p.p.m. for the remaining 18 weeks. NQO was freshly prepared every week and administered in light-shielded water bottles. The animals were monitored daily for signs of ill health. Moribund mice were killed and autopsied. For the analysis of tumor incidence, all mice were killed at week 21.
Tumor analysis
After anesthetization with isoflurane (Abbott Laboratories, North Chicago, IL), the animals were subjected to complete autopsies, with particular attention to tongue, esophagus and forestomach. Tumors>0.5 mm were mapped and coun-ted. Tongue, esophagus and forestomach were fixed in phosphate-buffered formalin and embedded in paraffin and 4-mm sections were cut. Hematoxylin and eosin (H&E) staining for histopathology and immunohistochemical analyses of the biomarkers were performed on sections from all animals. Immunohistochemical detection of cell proliferation
Esophagus and tongue sections were deparaffinized and rehydrated in a graded alcohol series and then incubated with mouse anti-proliferating cell nuclear antigen (anti-PCNA) monoclonal antibodies (Santa Cruz Biotechnology Inc., Santa Cruz, CA) at a 1:250 dilution overnight at 37C in a humidified chamber, followed by incubation with biotinylated goat anti-mouse antibodies (DakoCytomation, Carpinteria, CA) at 1:750 dilution and streptavidin horseradish peroxidase (DakoCytomation). PCNA staining was visualized by incubation with the chromogen 3-amino-9-ethylcarbazole (DakoCytoma-tion). Cells with a red reaction product in the nucleus were defined as positive for PCNA.
Immunohistochemical analysis of KRT14, COX-2 and MT protein expression Following deparaffinization and rehydration in a graded alcohol series, esophageal and lingual sections were heated in citrate buffer (0.01 M, pH 6.0) in a microwave oven (85–90C, 3· 5 min) and non-specific binding sites were then blocked with goat or rabbit serum. Sections were incubated overnight at 37C in a humidified chamber with one of the following primary
antibodies: mouse anti-KRT14 monoclonal antiserum (Clone LL002, Novocastra Lab., Newcastle upon Tyne, UK) at a 1:100 dilution, rabbit anti-COX-2 polyclonal antiserum (Caymen Chemical, Ann Harbor, MI) at a 1:50 dilution or mouse anti-MT monoclonal antiserum (Clone E9; DakoCyto-mation) at a 1:50 dilution. Sections were then incubated with the appropriate biotinylated secondary antibodies (DakoCytomation) and streptavidin horse-radish peroxidase (DakoCytomation). Staining of individual proteins was visualized by incubating sections with 3,30-diaminobenzidine tetrahydro-chloride (DAB) and lightly counterstaining them with hematoxylin. Zinc determination
The testes were removed from male and hair from female animals at necropsy. Samples of testis or hair were dried to constant weight at 90C and ashed in a furnace. Ashed samples were dissolved in 0.1 N HCl, and the zinc content determined by atomic absorption spectrometry as described (11). Zinc content was expressed asmg/g dry weight of testis or hair. The well-described overt signs of zinc deficiency for rats (11), including foci of alopecia, skin lesions and retarded growth, were not evident in ZD:WT mice, which had similar body weights at endpoint (data not shown). Zinc content in the testis (male) and hair (female), however, was significantly lower in ZD than ZS mice, regardless of genotype. For example, zinc content of testis and hair was significantly lower in ZD:p53+/ than ZS:p53+/ mice (testis, 113 ± 11 versus 153 ± 8 mg/g; hair, 129 ± 11 versus 192 ± 16mg/g, P < 0.001), a result consistent with previous studies (21).
Statistical analysis
Data on tumor multiplicity were analyzed by two-way analysis of variance (ANOVA), followed by Tukey’s highest significant difference test. Tumor incidence differences were analyzed by Fisher’s exact test, two-tailed. All statistical tests were two-sided and were considered statistically significant at P< 0.05.
Results
UADT tumorigenesis in ZD:p53+/ mice
In order to determine whether the combined deficiency of zinc
and p53 promotes UADT carcinogenesis, p53/, p53+/
and p53+/+ mice on ZD or ZS diet were given for
21 weeks drinking water containing 20 to 10 p.p.m. NQO.
While all heterozygous p53+/ and wild-type p53+/+ mice
survived for 21 weeks, only 2 of 15 (13%) nullizygous
ZD:p53/ mice and 10 of 14 (71%) ZS:p53/ and were
alive after 15 weeks (Table I). Among the heterozygous and
wild-type mice, only 2 of 16 ZD:p53+/ mice had spontaneous
(non-UADT) splenic lymphomas at week 21. In contrast,
ZD:p53/ mice succumbed to the rapid development of
spontaneous tumors (29), in addition to induced lingual and
esophageal lesions (Table I). These NQO-induced lesions
appeared to occur earlier in ZD:p53/ than in ZS:p53/
mice, a finding in agreement with our previous studies with
NMBA-induced forestomach tumors in ZD:p53/ mice (21).
Leukoplakia, the most common oral intraepithelial
neo-plasia in humans and a precursor of SCC, were consistently
found on the ZD:p53/ tongue between 2 and 8 weeks after
NQO exposure. By week 13, ZD:p53/ mice typically
harbored tumors at multiple sites in the UADT. For example,
ZD:p53/ mouse no. 15 showed multiple tumors in tongue,
esophagus, forestomach and hard palate at week 21 (Figure 1A,
lower panel, forestomach tumor data not shown).
Histo-pathological examination revealed progression to malignancy
in ZD:p53/ tongue and esophagus at week 11 (Figure 2B
and D) but mostly hyperplasia in ZS:p53/ tongue and
esophagus at week 12 (Figure 2A and C). By 20 weeks,
however, malignant changes in tongues became evident in
ZS:p53/ mice (data not shown). These data demonstrate
that in the absence of both p53 alleles, zinc deficiency
greatly accelerates the induction and progression of UADT
tumors in mice.
Wild-type ZS mice did not develop lingual or esophageal
lesions (0%) after exposure for 21 weeks to low levels of
NQO, a result consistent with previous reports (26,27). Zinc
deficiency, however, significantly enhanced the development
of tumors at multiples sites in wild-type mice, with a tumor
incidence of 50, 36 and 43% in tongue, esophagus and
forestomach, respectively (Figure 1B; ZD:p53+/+ versus
ZS:p53+/+: tongue, P ¼ 0.006; esophagus, P ¼ 0.04).
These results demonstrate that in wild-type mice, zinc
deficiency increases the incidence of NQO-induced UADT
tumors as it does in rats (13).
Haploinsufficiency for p53 alone enhanced the induction
of lingual and esophageal tumors in ZS mice, resulting in a
significantly higher incidence of lingual (62%) and esophageal
(38%) tumors in ZS:p53+/ than wild-type ZS:p53+/+ mice
(tongue, P
¼ 0.0003; esophagus, P ¼ 0.02; Figure 1B).
Haploinsufficiency for p53 alone, however, had no effects
on tumor progression (Figure 1B–D). In contrast,
combina-tion of p53 haploinsufficiency and zinc deficiency led to a
significantly greater tumor incidence/multiplicity and
fre-quency of tumor progression, as compared with heterozygous
ZS:p53+/ or wild-type ZD:p53+/+ mice (Figure 1B–D).
Fifty-six percent of ZD:p53+/ mice harbored tumors in all
three sites (Figure 1A, upper panel), as compared with 23% of
ZS:p53+/ (P ¼ 0.047), 14% of ZD:p53+/+ (P ¼ 0.026) and
none of the ZS:p53+/+ mice (P ¼ 0.0028). In addition,
ZD:p53+/ mice had a higher incidence of SCC, with 69%
of lingual and 31% of esophageal tumors presenting as SCC
versus 19% of lingual and 0% of esophageal tumors from
ZS:p53+/ mice (tongue, P ¼ 0.003; esophagus, P ¼ 0.005;
Figure 1C). These results demonstrate that zinc deficiency
or p53 haploinsufficiency acting alone increases UADT
tumors induction by low levels of NQO and acting together
greatly enhances tumor induction with rapid progression to
malignancy.
Cell proliferation and marker gene expression in tongue and
esophagus of NQO-treated wild-type mice
Histopathological examination revealed that esophagus and
tongue from wild-type ZS:p53+/+ mice at week 21 showed
a moderately thickened epithelium, with PCNA-positive nuclei
largely restricted to the basal cell layer (esophagus, Figure 3C
and G) and in focal hyperplastic lesions (tongue, Figure 3A and
E). In contrast, similarly treated ZD:p53+/+ esophagi were
highly proliferative, with PCNA-positive nuclei in many
cell layers and in focal hyperplastic lesions (Figure 3D and
H). ZD:p53+/+ lingual epithelia were typically hyperplastic,
showing dysplastic changes or early carcinoma in situ
(Figure 3B), with abundant PCNA-positive nuclei in dysplastic
areas (Figure 3F).
Immunohistochemistry was then used to determine whether
these early precancerous lesions in the ZD:p53+/+ mice are
accompanied by overexpression of the tumor markers KRT14,
COX-2 and MT. ZS:p53+/+ esophagus and tongue displayed
Table I. Tumor types in p53/ mice on ZD or control ZS diet after continuous exposure to 20–10 p.p.m. NQO in the drinking water Case Sex TTS (weeks) Age (weeks) Gross anatomy
Tongue Esophagus Forestomach Other sites
ZD
1 F 2.1 8 Leukoplakia Thick Thick Spleen
2 M 2.6 9 Leukoplakia Thick Thick –
3 M 4.0 10 Leukoplakia Thick Fused Ts, 6 –
4 M 4.1 10 Leukoplakia Thick 6 Ts, 0.5–2 –
5 F 4.7 11 2 Ts, 0.5 mm Thick Thick –
6 M 5.9 12 Leukoplakia Thick 2 Ts, 1 mm –
7 M 7.4 13 Leukoplakia Thick Fused Ts, 3 mm –
8 M 7.6 14 Leukoplakia Thick Thick
9 M 10.9 17 3 Ts, 0.5 mm Thick Thick Spleen
10 M 10.9 17 3 Ts, 0.5–1 mm Thick Thick Thymus, kidney
11 F 10.9 17 2 Ts, 1 mm Thick Thick Thymus
12 M 13.1 19 4 Ts, 0.5–1 mm 1 T, 1 mm Thick Spleen
13 F 14.7 22 2 Ts, 2 and 1 mm Very thick Fused tumors Spleen
14 M 20.9 27 5 Ts, 1–1.5 mm 3 Ts, 1–2.5 mm Fused tumors – 15 M 20.9 27 6 Ts, 1–3 mm 9 Ts, 1–3 mm 1 T, 2.5 mm Upper palate ZS 16 F 6.9 13 – – – Thymus 17 F 12 18 – – Thymus 18 M 12.1 18 Leukoplakia – – Thymus 19 M 12.1 18 Leukcoplakia – – 20 F 15.0 21 1 fused T, 7· 3 mm 1 T, 1.5 mm – – 21 M 15.7 22 7 Ts, 0.5–1 mm 3 Ts, all 1 mm – – 22 M 15.9 22 2 Ts, 1 mm 6 Ts, 1–2 mm – Spleen, thymus 23 M 17.9 24 5 Ts, 0.5–1.5 mm 5 Ts, 0.5–1 mm – Spleen 24 F 18.1 24 3 Ts, 0.5 mm 1 T, 1 mm – – 25 F 18.9 25 1 T, 3 mm 1 lg. T, 4 mm – Spleen 26 F 19.4 25 1 T, 3· 2 mm 1 T, 1 mm – Kidney 27 M 19.7 26 2 Ts, 1–2 mm 1 T, 1 mm – Subcutis, liver 28 M 21.0 27 4 Ts, 0.5–1.5 mm 8 Ts, 0.5–1.5 mm – Spleen 29 M 21.0 27 2 Ts, 1 and 3 mm 6 Ts, 1–2 mm 4 Ts, 0.5–1 mm Thymus
Four-week-old mice were given ZD or ZS diet and deionized water for 2 weeks before switching to drinking water containing 20–10 p.p.m. NQO, 20 p.p.m. for 3 weeks and 10 p.p.m. for the remaining 18 weeks. T, tumors; TTS, time to sacrifice (weeks) after exposure to NQO; moribund animals were killed.
moderate immunostaining of KRT14 and MT, mostly confined
to the basal cell layers (KRT14: esophagus, Figure 3K, tongue,
Figure 3I; MT: esophagus, Figure 3S, tongue, Figure 3Q).
Invariably, COX-2 staining was weak and diffuse in both
tissues of these mice (esophagus, Figure 3O, tongue,
Figure 3M). In contrast, early precancerous ZD:p53+/+ lesions
showed strong expression of all three markers. For example,
intense cytoplasmic staining of KRT14 was found in
hyper-plastic esophagus (Figure 3L) and dyshyper-plastic tongue with early
cancerous changes (Figure 3J) and strong to intense nuclear
and cytoplasmic expression of MT occurred in proliferative
esophagus (Figure 3T) and in lingual focal hyperplastic lesions
(Figure 3R). These same lesions also displayed moderate to
strong cytoplasmic COX-2 expression (esophagus, Figure 3P;
tongue, Figure 3N). These results in NQO-treated ZD
wild-type mice are in agreement with our findings in precancerous
ZD rat esophagus and tongue (13,14).
Cell proliferation and marker gene expression in lingual SCC
and ESCC of ZD:p53+/ mice
After 21 weeks, ZS:p53+/ esophagus and tongue showed
mostly basal cell hyperplasia, focal cell hyperplasia and
papillomas, with a very low incidence of tumor progression
only in the tongue (Figure 1D). An example of a ZS:p53+/
tongue showing dysplasia with microinvasion is presented
in Figure 4A. In combination, p53 insufficiency and zinc
deficiency led to high incidence of SCC in both tongue and
esophagus in ZD:p53+/ mice (Figure 1D). Examples of
invasive lingual SCC and ESCC are shown in Figure 4F
and K. ZD:p53+/ carcinomas were highly proliferative
with abundant PCNA-positive nuclei in the tumors and in
adjacent hyperplastic epithelia (lingual SCC, Figure 4G;
ESCC, Figure 4L). A high correlation was evident between
the spatial localization of KRT14, COX-2 or MT and that of
PCNA in serial lingual sections of early cancerous lesions
of ZS:p53+/ and in lingual SCC and ESCC of ZD:p53+/
mice (lingual early cancerous lesion, compare Figure 4C–E
with Figure 4B; lingual SCC, compare Figure 4H–J with
Figure 4G; ESCC, compare Figure 4M–O with Figure 4L).
Together, these results show an association between the cell
proliferation marker PCNA and tumor markers KRT14,
COX-2
or
MT
in
hyperplasia
(Figure
3),
dysplasia
(Figure 4) and invasive carcinomas (Figure 4). They provide
evidence that all three markers, MT, in particular, are highly
Fig. 1.Exquisite susceptibility to UADT carcinogenesis in ZD:p53+/ mice. The mice were treated with NQO at 20 p.p.m. in the drinking water for 3 weeks and then 10 p.p.m. for another 18 weeks. (A) Gross anatomy at week 21: upper panel, a representative ZD:p53+/ mouse shows multiple tumors in tongue and esophagus and a large fused tumor in forestomach. Lower panel, a ZD:p53/ mouse that survived 21 weeks of NQO treatment exhibits multiple large tumors in tongue, numerous tumors throughout the esophagus, large tumors in upper palate (oral cavity), as well as tumors in forestomach (forestomach not shown). (B–D) Tumor incidence, tumor multiplicity and carcinoma incidence in ZD:p53+/ (n ¼ 16), ZS:p53+/ (n ¼ 26),
ZD:p53+/+ (n ¼ 14) and ZS:p53+/+ (n ¼ 12) mice at week 21. (B) Tumor incidence (no. of mice with tumors/total no. of mice, %): Tongue, *ZD:p53+/ versus ZS:p53+/, P ¼ 0.007; **ZD:p53+/+ versus ZS:p53+/+ (0%), P ¼ 0.006; ZS:p53+/ versus ZS:p53+/+ (0%), P ¼ 0.0003; ZD:p53+/ versus ZD:p53+/+, P ¼ 0.0016. Esophagus,†
ZD:p53+/ versus ZS:p53+/, P ¼ 0.029;††
ZD:p53+/+ versus ZS:p53+/+ (0%), P ¼ 0.04; ZS:p53+/ versus ZS:p53+/+ (0%), P ¼ 0.02. (C) Tumor multiplicity (no. of tumors/site, mean ± SD): Tongue,ZD:p53+/ versus ZS:p53+/, P < 0.01; ZD:p53+/ versus ZD:p53+/+, P < 0.01; ZD:p53+/+ versus ZS:p53+/+ (0%), P < 0.05; ZS:p53+/ versus ZS:p53+/+ (0%), P < 0.01. Esophagus, ZD:p53+/ versus ZS:p53+/, P < 0.01; ZD:p53+/ versus ZD:p53+/+, P < 0.01. Forestomach, ZD:p53+/ versus ZS:p53+/, P < 0.01; ZD:p53+/ versus ZD:p53+/+, P < 0.01. (D) Carcinoma incidence (no. of mice with carcinomas/total number of mice, %): Tongue,ZD:p53+/ versus ZS:p53+/, P ¼ 0.003; Esophagus, ZD:p53+/ versus ZS:p53+/ (0%), P ¼ 0.005. Tumor incidence was performed by two-tailed Fisher’s exact test. Tumor multiplicity was analyzed by two-way ANOVA, followed by Tukey’s highest significant difference test. All statistical tests were two-sided.
relevant intermediate and endpoint biomarkers in UADT
carcinogenesis.
Discussion
The data presented above show that with combined deficiency
of zinc and p53, the mouse tongue and esophagus become
exquisitely sensitive to the carcinogenicity of low
concentra-tions of NQO (Figure 1). In addition, the results document that
zinc deficiency and haploinsufficiency acting alone increases
the yield of NQO-induced UADT benign tumors, whereas the
combined action drives tumor progression to malignancy.
Although UADT tumor induction and progression occurred
earlier in nullizygous ZD:p53/ mice than heterozygous
ZD:p53+/ mice, the nullizygous mouse is not a feasible
cancer model for UADT cancer because it succumbs to
induced and spontaneous tumorigenesis at an early age
(Table I). In contrast, Ide et al. (31) reported that p53+/
and p53/ mice, on a mixed genetic background of
C3H/HeN, C57BL/6 and CBA, were tumor-free with exposure
to drinking water containing 10 p.p.m. NQO for 50 and
20 weeks, respectively. It is possible that p53-deficient mice
on a C57BL/6 background, as in the case of the present study,
are more sensitive to the tumorigenic activity of NQO than
those on a mixed genetic background.
Complex genetic mouse models provide a useful
frame-work for understanding the interrelationship of p53 deficiency
and specific genes in oral–esophageal cancer development,
for example, the collaboration of p53 deficiency and
over-expression of cyclin D1 (30), or p53 deficiency and xeroderma
pigmentosum group A (XPA) gene deficiency (31). The
L2D1
+/p53+/ model that overexpresses cyclin D1 in the
oral–esophageal epithelium of p53-deficient mice develops
invasive oral–esophageal cancer, whereas control cyclin D1
L2D1
+mice exhibit only oral–esophageal dysplasia (30).
XPA/p53+/ mice generated by crossing XPA/ mice
with p53+/ mice demonstrate that p53 haploinsufficiency
greatly accelerates the onset of lingual tumors in XPA/
mice with continuous exposure to 10 p.p.m. NQO in the
drinking water (31). In either model, the remaining p53 allele
was not lost during tumor formation (30–33) and p53
deficiency acting alone did not bring about progression to
malignancy. These genetic models support the concept that
p53 plays an important role in oral–esophageal epithelial
carcinogenesis.
The growing list of genes in human cancers with aberrant
expression, however, points toward a more complex view of
the carcinogenesis process. In this regard, our UADT cancer
model in ZD:p53+/ mouse has obvious benefits. By
render-ing the p53+/ mice nutritionally deficient in zinc, this model
mimics aspects of human UADT cancers. Both zinc deficiency
(5–9) and a loss of function of the tumor suppressor gene TP53
(25) are associated with human UADT carcinogenesis. Zinc
deficiency causes substantial cell proliferation in the squamous
epithelium of the UADT, as well as extensive changes in gene
expression (13,14). On the other hand, loss of p53 function
results in genetic instability and increased cell cycle
progres-sion (34,35). Thus, the grave biological consequences of
com-bining zinc deficiency and p53 deficiency are rapid UADT
tumor induction and progression to malignancy (Figure 1D,
Table I).
The sequence of histopathologic events of human oral and
ESCC is well characterized and proceeds from hyperplasia to
focal hyperplastic lesions, dysplasia and finally invasive
carcinoma. Although many molecular alterations have been
reported in UADT carcinogenesis, intermediate biomarkers
that accurately predict the risk of developing cancer, prognosis
and effect of therapeutic treatment have not been clearly
Fig. 2. Histopathology of tongue and esophagus from ZD:p53/ and ZS:p53/ mice. The mice were treated with NQO in the drinking water at 20 p.p.m. for 3 weeks and then 10 p.p.m. for another 18 weeks. Representative H&E-stained sections are shown. At week 12, ZS:p53/ tongue and esophagus showed a highly hyperplastic epithelium (A, tongue; C, esophagus). At week 11, a ZD:p53/ tongue showed evidence of early carcinoma in situ (B) and a ZD:p53/ esophagus, SCC (D). Scale bars: 50 mm (A and C); 25 mm (B and D).
defined. In human cancers, MT overexpression is often
posi-tively correlated with the metastatic and proliferative activities
of ESCC (18) and a worse prognosis for oral SCC (36). KRT14
is an indicator of tumor progression in human ESCC (20) and
oral dysplasia-SCC sequence (37,38). COX-2 expression is
correlated with the proliferation activity in esophageal
dys-plasia (39) and with early stage carcinogenesis in the oral
dysplasia-carcinoma sequence (40).
Recent data from our laboratory (13,14) show that (i) the
hyperplastic esophagus and tongue of ZD rats overexpressed
COX-2 protein and mRNA; (ii) the zinc sensitive gene MT-1
and tumor marker KRT14 are upregulated
>6-fold in
hyper-plastic ZD rat esophagus; and (iii) upon zinc replenishment, the
overexpression of all three markers, as well as the hyperplastic
phenotype of the esophagus were rapidly reduced. The present
study examined whether the sequence of histopathologic
events in UADT carcinogenesis after 21 weeks of NQO
treatment is accompanied by overexpression of these three
biomarkers, KRT14, COX-2 and MT. Our results
demon-strate in a mouse UADT cancer model that expression of
bio-markers KRT14, MT and COX-2 in lingual and esophageal
carcinogenesis is correlated with cell proliferation and
the histopathologic hyperplasia-dysplasia-carcinoma sequence
(Figures 3 and 4).
In summary, we have developed an in vivo UADT cancer
model in ZD:p53+/ mouse that reproduces features of human
oral–esophageal cancer. This model will be useful to study the
molecular mechanisms of UADT carcinogenesis and to test
novel strategies for prevention and reversal of this deadly
cancer.
Fig. 3.Spatial localization of PCNA, KRT14, COX-2 and MT in premalignant lingual and esophageal lesions in ZD wild-type mice. The mice were treated with 20 p.p.m. of NQO for 3 weeks and 10 p.p.m. for another 18 weeks. Representative serial H&E and immunohistochemically-stained sections from ZD and ZS wild-type mice are shown. H&E-stained sections showed a moderately thickened lingual (A) and esophageal (C) epithelium from ZS mice and a hyperplastic tongue with dysplasia (B and inset) and a proliferative esophageal epithelium with focal hyperplastic lesions from ZD mice. PCNA, E–H; KRT14, I–L; COX-2, M–P; and MT, Q–T. Scale bars: 50mm (A, B); 25 mm (C–T).
Acknowledgements
This work was supported by grant no. 04B106-REN from the American Institute for Cancer Research (LYYF). We thank Mr Karl Smalley, Kimmel Cancer Institute, Thomas Jefferson University for help with statistical analysis of data.
Conflict of Interest Statement: None declared.
References
1. Parkin,D.M., Bray,F., Ferlay,J. and Pisani,P. (2001) Estimating the world cancer burden: Globocan 2000. Int. J. Cancer, 94, 153–156.
2. Moore,S.R., Johnson,N.W., Pierce,A.M. and Wilson,D.F. (2000) The epidemiology of tongue cancer: a review of global incidence. Oral Dis., 6, 75–84.
3. Slaughter,D.P., Southwick,H.W. and Smejkal,W. (1953) Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer, 6, 963–968.
4. Makuuchi,H., Machimura,T., Shimada,H. et al. (1996) Endoscopic screening for esophageal cancer in 788 patients with head and neck cancers. Tokai J. Exp. Clin. Med., 21, 139–145.
5. Doerr,T.D., Prasad,A.S., Marks,S.C., Beck,F.W., Shamsa,F.H., Penny,H.S. and Mathog,R.H. (1997) Zinc deficiency in head and neck cancer patients. J. Am. Coll. Nutr., 16, 418–422.
6. Kleier,C., Werkmeister,R. and Joos,U. (1998) Zinc and vitamin A deficiency in diseases of the mouth mucosa. Mund Kiefer Gesichtschir, 2, 320–325.
7. Yang,C.S. (1980) Research on esophageal cancer in China: a review. Cancer Res., 40, 2633–2644.
8. van Rensburg,S.J. (1981) Epidemiologic and dietary evidence for a specific nutritional predisposition to esophageal cancer. J. Natl Cancer Inst., 67, 243–251.
9. Abnet,C.C., Lai,B., Qiao,Y.L., Vogt,S., Luo,X.M., Taylor,P.R., Dong,Z.W., Mark,S.D. and Dawsey,S.M. (2005) Zinc concentration in esophageal biopsy specimens measured by x-ray fluorescence and esophageal cancer risk. J. Natl Cancer Inst., 97, 301–306.
10. Fong,L.Y., Sivak,A. and Newberne,P.M. (1978) Zinc deficiency and methylbenzylnitrosamine-induced esophageal cancer in rats. J. Natl Cancer Inst., 61, 145–150.
Fig. 4. Spatial localization of KRT14, COX-2, MT and PCNA in ZS:p53+/ lingual dysplasia showing microinvasion and ZD:p53+/ lingual SCC and ESCC. The mice were treated with 20 p.p.m. of NQO for 3 weeks and then 10 p.p.m. for another 18 weeks. Representative serial H&E and
immunohistochemically-stained sections are shown: (A–E) ZS:p53+/ lingual dysplasia, (F–J) ZD:p53+/ lingual SCC, and (K–O) ZD:p53+/ ESCC. H&E-stained sections showed lingual dysplasia (A), invasive lingual SCC (F) and ESCC (K). Immunohistochemical staining for PCNA (B, G and L), KRT14 (C, H and M), COX-2 (D, I and N) and MT (E, J and O). Scale bars: 50mm.
11. Fong,L.Y., Li,J.X., Farber,J.L. and Magee,P.N. (1996) Cell proliferation and esophageal carcinogenesis in the zinc-deficient rat. Carcinogenesis, 17, 1841–1848.
12. Fong,L.Y. and Magee,P.N. (1999) Dietary zinc deficiency enhances esophageal cell proliferation and N-nitrosomethylbenzylamine (NMBA)-induced esophageal tumor incidence in C57BL/6 mouse. Cancer Lett., 143, 63–69.
13. Fong,L.Y., Zhang,L., Jiang,Y. and Farber,J.L. (2005) Dietary zinc modulation of COX-2 expression and lingual and esophageal carcino-genesis in rats. J. Natl Cancer Inst., 97, 40–50.
14. Liu,C.G., Zhang,L., Jiang,Y., Chatterjee,D., Croce,C.M., Huebner,K. and Fong,L.Y. (2005) Modulation of gene expression in precancerous rat esophagus by dietary zinc deficit and replenishment. Cancer Res., 65, 7790–7799.
15. Chan,G., Boyle,J.O., Yang,E.K. et al. (1999) Cyclooxygenase-2 expression is up-regulated in squamous cell carcinoma of the head and neck. Cancer Res., 59, 991–994.
16. Zimmermann,K.C., Sarbia,M., Weber,A.A., Borchard,F., Gabbert,H.E. and Schror,K. (1999) Cyclooxygenase-2 expression in human esophageal carcinoma. Cancer Res., 59, 198–204.
17. Maaser,K., Daubler,P., Barthel,B., Heine,B., von Lampe,B., Stein,H., Hoffmeister,B., Scherer,H. and Scherubl,H. (2003) Oesophageal squa-mous cell neoplasia in head and neck cancer patients: upregulation of COX-2 during carcinogenesis. Br. J. Cancer, 88, 1217–1222.
18. Hishikawa,Y., Koji,T., Dhar,D.K., Kinugasa,S., Yamaguchi,M. and Nagasue,N. (1999) Metallothionein expression correlates with metastatic and proliferative potential in squamous cell carcinoma of the oesophagus. Br. J. Cancer, 81, 712–720.
19. Sundelin,K., Jadner,M., Norberg-Spaak,L., Davidsson,A. and Hellquist,H.B. (1997) Metallothionein and Fas (CD95) are expressed in squamous cell carcinoma of the tongue. Eur. J. Cancer, 33, 1860–1864. 20. Cintorino,M., Tripod,S.A., Santopietro,R., Antonio,P., Lutfi,A., Chang,F.,
Syrjanen,S., Shen,Q., Tosi,P. and Syrjanen,K. (2001) Cytokeratin expression patterns as an indicator of tumour progression in oesophageal squamous cell carcinoma. Anticancer Res., 21, 4195–4201. 21. Fong,L.Y., Ishii,H., Nguyen,V.T., Vecchione,A., Farber,J.L., Croce,C.M.
and Huebner,K. (2003) p53 deficiency accelerates induction and progression of esophageal and forestomach tumors in zinc-deficient mice. Cancer Res., 63, 186–195.
22. Fong,L.Y., Lau,K.M., Huebner,K. and Magee,P.N. (1997) Induction of esophageal tumors in zinc-deficient rats by single low doses of N-nitrosomethylbenzylamine (NMBA): analysis of cell proliferation, and mutations in H-ras and p53 genes. Carcinogenesis, 18, 1477–1484. 23. Fong,L.Y., Nguyen,V.T., Farber,J.L., Huebner,K. and Magee,P.N. (2000)
Early deregulation of the the p16ink4a-cyclin D1/cyclin-dependent kinase 4-retinoblastoma pathway in cell proliferation-driven esophageal tumorigenesis in zinc-deficient rats. Cancer Res., 60, 4589–4595. 24. Fong,L.Y., Nguyen,V.T. and Farber,J.L. (2001) Esophageal cancer
prevention in zinc-deficient rats: rapid induction of apoptosis by replenishing zinc. J. Natl Cancer Inst., 93, 1525–1533.
25. Greenblatt,M.S., Bennett,W.P., Hollstein,M. and Harris,C.C. (1994) Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res., 54, 4855–4878.
26. Ide,F., Oda,H., Nakatsuru,Y., Kusama,K., Sakashita,H., Tanaka,K. and Ishikawa,T. (2001) Xeroderma pigmentosum group A gene action as a
protection factor against 4-nitroquinoline 1-oxide-induced tongue car-cinogenesis. Carcinogenesis, 22, 567–572.
27. Tang,X.H., Knudsen,B., Bemis,D., Tickoo,S. and Gudas,L.J. (2004) Oral cavity and esophageal carcinogenesis modeled in carcinogen-treated mice. Clin. Cancer Res., 10, 301–313.
28. Jacks,T., Remington,L., Williams,B.O., Schmitt,E.M., Halachmi,S., Bronson,R.T. and Weinberg,R.A. (1994) Tumor spectrum analysis in p53-mutant mice. Curr. Biol., 4, 1–7.
29. Donehower,L.A, Harvey,M., Slagle,B.L., McArthur,M.J., Montgomery,C.A.,Jr, Butel,J.S. and Bradley,A. (1992) Mice deficient for p53 are developmentally normal but susceptible to spontaneous tumours. Nature, 356, 215–221.
30. Opitz,O.G., Harada,H., Suliman,Y., Rhoades,B., Sharpless,N.E., Kent,R., Kopelovich,L., Nakagawa,H. and Rustgi,A.K. (2002) A mouse model of human oral-esophageal cancer. J. Clin. Invest., 110, 761–769.
31. Ide,F., Kitada,M., Sakashita,H., Kusama,K., Tanaka,K. and Ishikawa,T. (2003) p53 haploinsufficiency profoundly accelerates the onset of tongue tumors in mice lacking the xeroderma pigmentosum group A gene. Am. J. Pathol., 163, 1729–1733.
32. Venkatachalam,S., Shi,Y.P., Jones,S.N., Vogel,H., Bradley,A., Pinkel,D. and Donehower,L.A. (1998) Retention of wild-type p53 in tumors from p53 heterozygous mice: reduction of p53 dosage can promote cancer formation. EMBO J., 17, 4657–4667.
33. Jerry,D.J., Butel,J.S., Donehower,L.A., Paulson,E.J., Cochran,C., Wiseman,R.W. and Medina,D. (1994) Infrequent p53 mutations in 7,12-dimethylbenz[a]anthracene-induced mammary tumors in BALB/c and p53 hemizygous mice. Mol. Carcinog., 3, 175–183.
34. Levine,A.J. (1997) p53, the cellular gatekeeper for growth and division. Cell, 88, 323–331.
35. Lakin,N.D. and Jackson,S.P. (1999) Regulation of p53 in response to DNA damage. Oncogene, 18, 7644–7655.
36. Cardoso,S.V., Barbosa,H.M., Candellori,I.M., Loyola,A.M. and Aguiar,M.C. (2002) Prognostic impact of metallothionein on oral squa-mous cell carcinoma. Virchows Arch., 441, 174–178.
37. Su,L., Morgan,P.R. and Lane,E.B. (1996) Keratin 14 and 19 expression in normal, dysplastic and malignant oral epithelia. A study using in situ hybridization and immunohistochemistry. J. Oral. Pathol. Med., 25, 293–301.
38. Ohkura,S., Kondoh,N., Hada,A., Arai,M., Yamazaki,Y., Sindoh,M., Takahashi,M., Matsumoto,I. and Yamamoto,M. (2005) Differential expression of the keratin-4, -13, -14, -17 and transglutaminase 3 genes during the development of oral squamous cell carcinoma from leuko-plakia. Oral Oncol., 41, 607–613.
39. Shamma,A., Yamamoto,H., Doki,Y. et al. (2000) Up-regulation of cyclooxygenase-2 in squamous carcinogenesis of the esophagus. Clin. Cancer Res., 6, 1229–1238.
40. Shibata,M., Kodani,I., Osaki,M., Araki,K., Adachi,H., Ryoke,K. and Ito,H. (2005) Cyclo-oxygenase-1 and -2 expression in human oral mucosa, dysplasias and squamous cell carcinomas and their pathological sig-nificance. Oral Oncol., 41, 304–312.
Received January 10, 2006; revised February 28, 2006; accepted March 7, 2006