Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Gastroenterology. Author manuscript; available in PMC 2010 March 1.
Published in final edited form as:
PMCID: PMC2650087

Zinc Replenishment Reverses Overexpression of the Proinflammatory Mediator S100A8 and Esophageal Preneoplasia in the Rat


Background & Aims

Zinc-deficiency is implicated in the pathogenesis of human esophageal cancer. In the rat esophagus, it induces cell proliferation, modulates genetic expression, and enhances carcinogenesis. Zinc-replenishment reverses proliferation and inhibits carcinogenesis. The zinc-deficient rat model allows the identification of biological differences affected by zinc during early esophageal carcinogenesis.


We evaluated gene expression profiles of esophageal epithelia from zinc-deficient and replenished rats versus sufficient rats using Affymetrix Rat Genome GeneChip. We characterized the role of the top-upregulated gene S100A8 in esophageal hyperplasia/reversal and in chemically-induced esophageal carcinogenesis in zinc-modulated animals by immunohistochemistry and real-time quantitative polymerase chain reaction.


The hyperplastic deficient esophagus has a distinct expression signature with the proinflammation-gene S100A8 and S100A9 upregulated 57- and 5-fold. “Response to external stimulus” comprising S100A8 was the only significantly overrepresented biological pathway among the upregulated genes. Zinc-replenishment rapidly restored to control levels the expression of S100A8/A9 and 27 other genes and reversed the hyperplastic phenotype. With its receptor RAGE, co-localization and overexpression of S100A8 protein occurred in the deficient esophagus that overexpressed NF-κB p65 and COX-2 protein. Zinc-replenishment but not by a COX-2 inhibitor reduced the overexpression of these 4 proteins. Additionally, esophageal S100A8/A9 mRNA levels were directly associated with the diverse tumorigenic outcome in zinc-deficient and zinc-replenished rats.


In vivo zinc regulates S100A8 expression and modulates the link between S100A8-RAGE interaction and downstream NF-κB/COX-2 signaling. The finding that zinc regulates an inflammatory pathway in esophageal carcinogenesis may lead to prevention and therapy for this cancer.


Esophageal cancer, including esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma (EAC), is the sixth most common cancer in the world.1 ESCC is the predominant histological subtype. Because of an absence of early symptoms, patients with ESCC are frequently diagnosed at an advanced stage of the disease and, consequently have a poor prognosis. Despite advances in treatment protocols that combine surgery with radiotherapy and chemotherapy, the overall 5-year survival rate remains at ~10% for ESCC worldwide. In order to better prevent and treat this deadly cancer, it is critical to understand its biology and to discover novel early biomarkers for chemopreventive and therapeutic regimens.

Epidemiologic studies have implicated dietary zinc-deficiency (ZD) in the etiology of ESCC.2, 3 Abnet et al.4 provided the strongest evidence of an association between dietary ZD and esophageal cancer in humans by establishing an inverse relationship between zinc concentration in biopsy samples from a high ESCC incidence area (Linxian, China) and subsequent risk of developing cancer.

We have developed a ZD rat tumor model that reproduces this feature of human ESCC. By causing unregulated cell proliferation and extensive genetic changes in the squamous epithelium, dietary ZD creates a protumorigenic microenvironment in the rat esophagus and tongue. As a result, ZD rats are exquisitely sensitive to N-nitrosomethylbenzylamine (NMBA)-induced esophageal5, 6 and 4-nitroquinoline 1-oxide (NQO)-induced lingual carcinogenesis.7 Zinc replenishment (ZR) rapidly re-establishes a normal phenotype by reversing cell proliferation, restoring normal gene expression, stimulating apoptosis and, thereby, inhibiting carcinogenesis.7-9

DNA microarray analysis provides a powerful tool to understand how zinc might modify genetic events in esophageal carcinogenesis. Using Bioarray chips with ~8,000 genes we previously showed that 33 genes were differentially expressed in hyperplastic ZD versus control zinc-sufficient (ZS) esophagus, including upregulation of the zinc-sensitive gene metallothionein-1.10 Although this pattern of gene expression is different from those reported for small intestine and liver in which ZD does not cause proliferative lesions,11, 12 the mechanism by which ZD induces esophageal hyperplasia is not known. We hypothesized that global transcriptome profiling using an expanded genome array (Affymetrix) that analyzes >30,000 transcripts and variants from ~28,000 rat genes would identify key biologic differences affected by nutritional ZD during early esophageal carcinogenesis. Here, we evaluated gene expression profiles of esophageal epithelia from zinc-modulated rats and focused on the role of the proinflammation gene S100A8 (S100 calcium binding protein A8) in early esophageal carcinogenesis and its reversal.

Materials and methods


Weanling male Sprague-Dawley rats were from Taconic Laboratory (Germantown, NY). Custom-formulated ZD and ZS diets (Harlan Teklad, Madison, WI) were identical except for the zinc content.5

Experimental design

This study was approved by The Ohio State University Institutional Animal Care and Use Committee. Male weanling rats were fed a ZD diet ad libitum (n=24) or pair-fed a control ZS diet (n = 12). After 6 weeks ZD rats evidenced increased cell proliferation in the esophagus, as assessed by increased expression of proliferating cell nuclear antigen PCNA.5 Zinc gluconate (1.0 mg elemental zinc) in saline was then administered intragastrically to 12 ZD rats, which were immediately switched to the ZS diet to form the ZR group. At 48 hours after replenishment, all animals (ZD, ZR, and ZS; n = 12/group) were killed. Whole esophagi (n = 4/group) were fixed in buffered formalin and embedded in paraffin. Esophageal epithelia prepared by using a blade (under microscopic guidance) to remove the submucosal and muscularis layers were snap-frozen in liquid nitrogen and stored at -80°C for RNA (n = 4/group) and protein (n = 4/group) preparation6, 7, 10.

RNA preparation

Total RNA was prepared using TRIZOL reagent (Invitrogen, Carlsbad, CA). All RNA samples displayed intact ribosomal RNA 18S and 28S bands as determined by the Agilent 2100 Bioanalyzer.

Gene expression profiling

Microarray analysis was conducted on individual RNA samples from ZS, ZD, and ZR rats (n = 4/group) using Rat Genome 230 2.0 GeneChip (Affymetrix, Santa Clara, CA), containing more than 31,000 probe sets that analyze >30,000 transcripts and variants from ~28,000 rat genes. Five micrograms of total RNA was reverse transcribed into cDNA followed by in vitro transcription and labeling to produce biotin-labeled cRNA (IVT Labeling Kit). The labeled cRNA was hybridized to the array, followed staining with streptavidin-phycoerythrin. The distribution of fluorescent material on the oligonucleotide probes sets was obtained using GeneArray Scanner.

Expression data analysis

The Significance Analysis of Microarrays software (SAM, version 3.0) was used to identify differentially expressed mRNAs following normalization of array data by the BRBArrayTools software (Biometric Research Branch, National Cancer Institute). Based on the False Discovery Rate, the cutoff for significance was set so that the estimated q-value was between zero and 5%.13 The data were submitted to Array Express (Accession number: ETABM-504).

Pathway analysis of expression data

The Gene Ontology database was used to categorize genes into functional groups. The Expression Analysis Systematic Explorer (EASE) software that uses Fisher's exact test statistics (EASE score) for pathway analysis was employed to determine the probability of observing the number of genes within a list of interest versus the number of gene in each category on the array.14

Real-time quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR)

To confirm the differential expression for selected genes, qRT-PCR was performed on the same RNA samples used for array analyses, using pre-designed probes, GAPDH as the normalizer, and the comparative Ct method. The analysis used the ABI Prism 7900HT Sequence detection system (Applied Biosystems, Foster City, CA).

Immunoblotting analysis

Immunoblotting analysis was performed on all dietary groups (n=4/group) as described.7 Membranes were probed with goat polyclonal against S100A8, goat polyclonal against PTGES, and mouse monoclonal against SERPINB3 (Santa Cruz Biotechnology, Santa Cruz, CA), chicken IgY against DNMT3A (Abcam, Cambridge, MA), mouse monoclonal against PPP2R1a (Upatate), or mouse monoclonal against CSRP3 (Genway, San Diego, CA). GAPDH (Calbiochem, San Diego, CA) was used as a loading control. Band intensities were evaluated by GS800 calibrated densitometer (Bio-Rad, Hercules, CA).

S100A8 and RAGE immunohistochemistry (IHC)

Following deparaffinization and antigen retrieval, esophageal sections were incubated with mouse anti-S100A8 monoclonal antiserum (BMA Biomedicals, Augst, Switzerland) or mouse anti-RAGE monoclonal antiserum (R&D Systems, Minneapolis, MN) followed by incubation with appropriate biotinylated secondary antibodies and streptavidin horseradish peroxidase.7 Protein was localized by incubation with 3-amino-9-ethylcarbazole substrate-chromogen (AEC) (Dako Corporation, Carpinteria, CA).

The percentage of cells positive for S100A8 was determined as follows: 0 = 0%-5%, 1 = 6%-25%, 2 = 26%-50%, 3 = 51%-75%, and 4 = 76%-100%. The intensity of S100A8 staining was graded as follows: 0 = none, 1 = weak, 2 = moderate, and 3 = intense. An immunoreactive score was calculated by multiplying the grade of percentage of positive cells by the grade of intensity of staining.

Esophageal S100A8/A9 mRNA expression during carcinogenesis

S100A8/A9 mRNA expression was determined by qRT-PCR in esophagi from 6 ZD and 6 ZS rats at 0 hour (before NMBA dosing) and from zinc-modulated rats at 5 and 15 weeks after NMBA treatment in a tumorigenesis experiment. Here, 120 ZD and 60 ZS rats were given their respective diets for 6 weeks to establish increased esophageal proliferation in ZD rats (0 hour). All rats were then administered an intragastric dose of NMBA (2 mg/kg body weight). Immediately thereafter, 60 ZD rats were switched to a ZS diet to form the ZR group while the remaining ZD and ZS rats continued on their diets. Intragastric NMBA was administered again at week 6 and 12. At 0 hour and at various times after the first dose, 8 rats per group were killed for esophageal tissue isolation. All remaining rats were killed at 15 weeks for tumor incidence analysis.

Serum zinc measurement

Serum zinc analysis was by atomic absorption spectrometry.5

Statistical Analysis

Data on serum zinc and mRNA levels, tumor multiplicity were analyzed by one-way analysis of variance (ANOVA). Tumor incidence differences were by Fisher's exact test. Statistical tests were two-sided and were considered statistically significant at P <.05.


Zinc replenishment reverses the hyperplastic esophageal phenotype

The effect of zinc treatments on esophageal histopathology and serum zinc levels was first determined. After 6 weeks of experimental diets, ZD rat esophageal epithelia were highly hyperplastic, as compared with control ZS mucosa (typically 2 to 3 cells thick). At 48 hours after intragastric zinc, ZR mucosa had thinned to a thickness of 3 to 5 cells (Figure 1D). Serum zinc levels (μg/100 ml) in ZD rats were one-third those in control ZS rats (41±8 versus 112±12; P<.001). Forty-eight hours after ZR, serum zinc contents in ZR rats rose from a ZD level of 41±8 to 188±36 (P<.001).

Figure 1
Expression profiling of zinc-modulated rat esophageal mucosa. (A) Dendrogram illustrating clustering of 30,000 transcripts in ZD, ZR, and ZS esophagi. (B) 103 genes in ZD versus ZS esophagi (fold-change ≥ 4). (C) Scatterplot of ZD versus ZS esophagi. ...

A distinct gene expression signature in ZD esophagus

A hierarchical clustering analysis of 30,000 transcripts showed that the gene expression patterns between ZD and ZS esophageal epithelia were distinct (Figure 1A). Using a cutoff point of q value < 30% and ≥ 2-fold difference, 2305 dysregulated probe sets were identified in the hyperplastic ZD versus ZS esophagus (1060 up and 1245 downregulated; Supplementary Table 1). The present data included 8 (5 up: CA2, SCD1, ENO1, CNNB, and Lypla2; and 3 down: FETUB, STS, and GPLD1) of the 33 genes identified in the previous microarray analysis by the Rat 1 Bioarray chip.10

By filtering the data using a cutoff of q value < 5.0% and ≥ 4-fold differences in expression levels, a group of 103 differentially expressed genes were identified in ZD esophagus (61 up and 42 down, Table 1), thereby revealing a distinct and novel molecular signature as compared with ZS esophagus (Figure 1B). Three of these genes (2 up- CA2 and SCD1, and 1 downregulated FETUB) were previously reported in ZD esophagus,10 the remaining 100 genes represent new biomarkers associated with esophageal hyperplasia.

Table 1
Gene expression signature in precancerous zinc-deficient rat esophagus

The most upregulated genes were S100A8 (up 57-fold), SERPINB3, IFIT2, TMPRSS11D, IFIT3, and SECTM1 with fold-changes ranging from 57- to 13-fold; and the most downregulated genes were P2RX2, SLC15A1, GOS2, TRDN, KBTBD10, and CABC1 with fold-changes ranging from -12.5 to -7.7-fold (Figure 1C, scatter plot). At 48 hours after ZR, the altered expression of 14 genes, including P2R2X2, SLC15A1, KLK6, PRSS35, SFTPC, DMNT3A, GDF15, CFTR, AREG, IFIT3, SCGB1A1, PTGES, GADD45A, and SERPINB3 returned to control levels of ZD/ZS ~1.0 (Figure 2A). The abnormal expression levels of another 15 genes, including the top-upregulated S100A8 (Figure 2A) and S100A9, RHOV, HPGD, OCM, IVL, FETUB, FXYD4, IFIT2, PLAT, G0S2, MYL2, ACTC1, IGSF4A, CPT1A (data not shown) were reverted to a fold-change level of ZD/ZS ~2.0. Importantly, these data show that 29 of 103 genes (28%) are highly regulated by zinc nutrition and the altered and corrected expression is associated with induction and reversal of esophageal hyperplasia.

Figure 2
Validation of array data. (A) Array data showing that ZR restores the altered expressions of 14 genes in ZD esophagi to ZD/ZS ~1, and S100A8 to ZD/ZS < 2 levels. (B) qRT-PCR validation for 7 genes. (C) Immunoblots (n = 3/group) for 6 genes and ...

Identification of overrepresented pathways

EASE pathway analysis program14 was used to identify biological pathways overrepresented among these 103 genes (Table 2). Among the upregulated genes, “response to external stimulus” pathway comprising S100A8 and 6 genes (CDKN1A, GJA1, MX2, PPP2R1A, SECTM1, and SERPINB3) was the only significantly overrepresented pathway (EASE score =.02), a novel result that supports S100A8 as a relevant marker in ZD-induced esophageal hyperplasia. Among the downregulated genes, EASE uncovered an intriguing correlation (EASE score <.001) between the genes associated with muscle contraction/development pathway (10 genes: ATP2A1, CKM, CRYAB, MYBPC1, MYH8, MYL3, MYH6, SMPX, TNNT3, and TRDN) and organogenesis pathway (7 genes: CRYAB, MYBPC1, MYH6, MYH8, MYL3, PCP4, and TNNT3) and esophageal hyperplasia. This finding is consistent with a microarray analysis of mouse skin carcinogenesis in which EASE indicated a correlation between muscle-associated genes and skin differentiation.15

Table 2
EASE analysis of functional pathways among the 103 differentially expressed genes in zinc-deficient rat esophagus

Validation of expression data

The real-time qRT-PCR data verified the expression changes on the arrays, although with larger or smaller magnitudes (Figure 2B). As an example, qRT-PCR data for S100A8 in ZD and ZR esophagi were increased 77- and 12-fold above ZS, as compared with the array results (57- and 2.4-fold increase). qRT-PCR data for S100A9 (not shown in Figure 2B) in ZD esophagus was 3.1-fold above ZS as compared with the array data of 5-fold.

The protein products of the upregulated genes S100A8, SERPINB3, DNMT3A, PTGES, and PPP2R1A were 14.7-, 15.5-, 3.5, 4.2, and 2-fold higher in ZD than ZS esophagi, respectively. The protein product of the downregulated gene CSRP3 was 0.23-fold of ZS controls (Figure 2 C-D). ZR significantly reduced the abnormal protein expression of S100A8, SERPINB3, and PTGES, a result indicating that the zinc-modulating effects observed at the transcript level were reflected at the protein level for these genes.

In vivo link between S100A8-RAGE interaction and NF-κB p65 - COX-2 protein expression

We focused our study on the top-upregulated gene S100A8 (up 57-fold) because of its role in inflammation and cancer. Interaction of S100A8/S100A9 ligands with their receptors triggers intracellular signaling cascades, including NF-κB signaling.16 S100A9 (up 5-fold) was not selected to complement IHC studies for S100A8, because an appropriate antibody was not available.

We determined whether a link occurs between S100A8 overexpression and downstream NF-κB signaling in the earliest stage of esophageal carcinogenesis. S100A8 and its putative receptor RAGE (receptor for advanced glycation end products) protein expression were analyzed by IHC in near serial sections from archived esophageal tissues that had documented co-overexpression and co-reduction of COX-2 (cyclooxygenase-2) protein7 and the transcription factor NF-κB p658 under identical ZD and ZR conditions; NF-κB controls the expression of the inducible enzyme COX-2. IHC was performed on esophageal tissue samples from 11 ZD, 10 ZR, and 6 control ZS rats.

ZS esophageal epithelia showed sporadic and weak cytoplasmic staining of S100A8 (Figure 3Q). In contrast, the hyperplastic ZD esophageal epithelia displayed frequent and intense cytoplasmic staining with notable spatial patterns. As examples, intense S100A8 staining was observed in many cell layers throughout the entire esophageal section (Figure 3A and E [higher magnification]) and in bands that alternated with zones of no expression (Figure 3I). As early as 8 hours after zinc treatment, S100A8 expression in the still proliferative ZR esophagus showed a substantial reduction in staining intensity and number of cells stained (Figure 4A3 versus 3A). By 48 hours S100A8 expression was weak and diffuse (Figure 3M) in the thinned ZR esophageal epithelia. The semi-quantitative mean immunoreactive scores for ZS, ZD, and ZR esophageal epithelia were 1.3 ± 0.4, 7.8 ± 0.6, and 3.4 ± 0.7, respectively, translating to a ~6-fold increase in ZD versus ZS esophagi and a 2.7-fold increase in ZR versus ZS esophagi. These data are consistent with those of immunoblot analysis in which the 14.7-fold increase of S100A8 expression in ZD esophagi was reduced to 2.5-fold that of ZS esophagi following ZR (Figure 2C and D).

Figure 3
Spatial and temporal localization of S100A8 and RAGE protein (AEC, red) in near serial sections of archived esophageal tissues from zinc-modulated rats that overexpressed NF-κB p65 and COX-2 (3,3′-diaminobenzidine tetrahydrochloride, brown). ...
Figure 4
(A) Effects of celecoxib and zinc on S100A8 and RAGE protein expression. S100A8 and RAGE protein (AEC, red) immunostaining was intense in hyperplastic ZD esophagi at 8 hours after celecoxib treatment (A1, A2) but sporadic/weak after zinc treatment (A3, ...

RAGE the putative receptor for S100A8 displayed a spatial pattern similar to S100A8. Parallel to S100A8 (Figure 3Q), RAGE protein expression was very low in ZS esophageal epithelia (Figure 3R). Under ZD conditions, RAGE was co-overexpressed with S100A8 in near serial esophageal tissue sections (Figure 3B, F, and J versus A, E, and I). Upon ZR, RAGE expression was concomitantly co-reduced with S100A8 (8 hours after ZR, Figure 4A4 and A3; 48 hours after ZR, Figure 3N and M). These data showed that co-overexpression of S100A8 and RAGE protein occurs during early esophageal carcinogenesis and is reduced upon ZR. Thus, in vivo zinc regulates RAGE protein expression in the same manner as S100A8.

Importantly, our IHC data revealed that the spatial and temporal localization of S100A8 and RAGE protein expression mirrored that of NF-κB p65 and COX-2.7, 8 For the purpose of comparison, our published IHC data on NF-κB p65 and COX-2 expression are presented (NF-κB p65, Figure 3C, G, K, O, and S; COX-2, Figure 3D, H, L, P, and T). In normal ZS esophagus, all 4 proteins showed very low levels of expression (Figure 3, Q-T). Under ZD conditions all 4 proteins were overexpressed. They exhibited a strikingly similar spatial pattern in hyperplastic esophageal epithelia (Figure 3, A-D, E-H, and I-L). Within hours after ZR, the expression of all 4 proteins was concomitantly reduced (Figure 3, M-P). These results demonstrate an in vivo link between S100A8-RAGE co-overexpression and downstream NF-κB-COX-2 signaling that is modulated by zinc nutrition.

Effect of COX-2 inhibitor on S100A8/RAGE protein expression in ZD esophagus

The ability was determined of the COX-2 inhibitor celecoxib to inhibit S100A8/RAGE overexpression events upstream of COX-2. IHC was performed on archived esophageal samples from ZD rats that were treated with an oral dose of celecoxib.7 Within hours ZR reduced S100A8 and RAGE overexpression in the still hyperplastic esophagus (Figure 4A3 and A4), celecoxib had little effect in curbing S100A8 or RAGE overexpression in hyperplastic ZD esophagus (Figure 4A1 and A2). The result provides an explanation for the inefficacy of celecoxib in UADT cancer prevention in ZD animals.8

S100A8/A9 mRNA expression levels correlate with esophageal tumor outcome

The influence of dietary zinc on esophageal S100A8/A9 mRNA expression was further defined by qRT-PCR analysis in a tumorigenesis study (Figure 4B). Consistent with the array data (Table 1), hyperplastic ZD esophagi had significantly higher S100A8 (P<.001) and S100A9 (P<.01) mRNA levels than ZS esophagi. Importantly, S100A8/A9 mRNA overexpression in ZD esophagi was sustained during esophageal carcinogenesis (5 and 15 weeks after NMBA dosing) and was associated with a high tumorigenic outcome in ZD versus ZS rats (tumor incidence, 100% versus 16.6%, P=7.8×10-6; multiplicity, 11 ± 3.8 versus 0.5 ± 0.3, P<.001). Replenishing zinc led to reduced S100A8/A9 mRNA levels at both time points and a low tumorigenic outcome in ZR versus non-replenished ZD rats (incidence, 100% versus 28.9%, P=5.0×;10-5; multiplicity, 11 ± 3.8 versus 0.6 ± 0.4, P<.001). Thus, esophageal S100A8/A9 mRNA levels were regulated by zinc and were directly associated with the risk of developing tumors in zinc-modulated rats.

S100A8 protein expression in human and mouse ESCC

Finally, we investigated whether expression of the marker S100A8 identified in the hyperplastic ZD rat esophageal mucosa was overexpressed in human and mouse esophageal ESCC. IHC was performed on 16 human ESCC and 5 mouse ESCC from NQO-treated ZD:p53+/- mice.17

In human ESCC, cytoplasmic S100A8 immunostaining was observed in all 16 samples with varying degree of intensities. As examples: 5 showed moderate staining (+2; Figure 5C), 8 had strong staining in tumor tissues (+3; Figure 5A and B), and 3 had weak staining (+1; data not shown). Additionally, strong S100A8 immunostaining was detected in tumor stroma (Figure 5A-C; arrows). Adjacent normal esophageal epithelium showed very weak staining of S100A8 (Figure 5D).

Figure 5
S100A8 protein expression in human (A-D) and mouse (E-F) ESCC. Panels A-C show moderate to intense S100A8 immunostaining (AEC, red) in tumor tissues and stroma (arrows), and D shows very weak staining in adjacent near-normal esophageal mucosa in an ESCC ...

ESCC from NQO-treated, zinc-deficient p53+/- mice showed strong cytoplasmic immunostaining of S100A8 protein in invasive tumor areas (Figure 5E and F). By contrast, carcinoma free control zinc-sufficient p53+/- esophagus displayed weak staining (data not shown).


The present expression profiling analysis shows that the hyperplastic ZD rat esophagus has a distinct and novel molecular signature (Figure 1B). Importantly, the proinflammation gene S100A8 (up 57-fold, q value = .0%) and its heterodimer S100A9 (up 5-fold, q value = .4%) and 29 other genes were highly zinc-responsive. Upon ZR their altered expression rapidly reverted to control levels, accompanied by the reversal of the hyperplastic phenotype. S100A8 together with 6 other genes are members of the only significantly overrepresented “response to external stimulus” biological pathway (EASE score = .02, Table 2) among the upregulated genes. This result suggests that S100A8 is a relevant marker belonging to a causal pathway that drives esophageal cell proliferation rather than simply an epiphenomenon of this process or of nutritional zinc deficit.

S100A8 and S100A9 encode the S100 family member of calcium binding proteins that also bind zinc. Originally discovered as immunogenic protein expressed and secreted by neutrophils, S100A8/A9 have emerged as important mediators in inflammation. As a result, they may play a key role in inflammation-associated cancers.16 Indeed, S100A8/A9 are overexpressed in a variety of human cancers,16 including skin SCC18 and Barrett's esophagus, a precancerous condition of EAC.19 In ESCC conflicting cDNA microarray data have been published. S100A8 downregulation was reported in a Chinese population,20 whereas upregulation was noted in an Indian population.21

Here in a tumorigenesis experiment (Figure 4B), we provide evidence to show that a ZD diet induces and maintains S100A8/A9 mRNA overexpression before and during NMBA-induced esophageal carcinogenesis. Switching to a ZS diet during carcinogenesis leads to substantially reduced S100A8/A9 mRNA levels in ZR animals. Importantly, the diverse tumorigenic outcome in ZD and ZR animals are directly associated with their elevated and reduced S100A8/A9 mRNA levels during carcinogenesis. The data establish for the first time that zinc regulates S100A8/A9 mRNA expression in vivo in esophageal epithelial cells, thereby influencing the risk for developing cancer.

RAGE is the putative receptor of these two proteins. In vitro engagement of RAGE by S100A8/A9 activates NF-κB signaling in various cells and tumor cells.22, 23 In vivo in a multistage mouse skin carcinogenesis model, Gebhardt et al.24 provided direct genetic evidence that S100A8/A9 binds to RAGE, and RAGE signaling mediates sustained skin inflammation and promotes tumor development. Blockade of RAGE suppresses tumor growth and metastasis.24, 25 Here we showed that RAGE was co-overexpressed with S100A8 in the hyperplastic ZD esophageal epithelia that evidenced overexpression of NF-κB p65 and COX-2 proteins. ZR reduced the overexpression of all four proteins, thereby suppressing RAGE/S100A8 signaling, and reversed esophageal hyperplasia (Figure 3). These data demonstrate in vivo that zinc regulates the proinflammatory mediator S100A8 and modulates its interaction with RAGE and the downstream NF-κB-COX-2 signaling pathway, providing the first evidence for an inflammation-modulating role of zinc in early esophageal carcinogenesis and its reversal. Consistent with this finding, NF-κB is recognized as a link between inflammation and cancer development/progression. The IκB kinase/NF-κB activation pathway is a target for cancer prevention.26 In vivo zinc supplementation in humans led to downregulation of inflammatory cytokines and inhibition of induced NF-κB activation.27 In vitro zinc supplementation also inhibited NF-κB activation and suppressed the tumorigenic potential of human prostate cancer cells.28 Our data show that the protumorigenic and antitumorigenic effects of zinc may well operate through activation and suppression of S100A8-RAGE interaction and downstream NF-κB signaling.

Other classes of genes common among the upregulated genes include those associated with immune responses, apoptosis, cell proliferation and regulation of transcription (Table 1). Many of these genes are zinc-responsive and their roles in esophageal carcinogenesis are unclear. As examples, the highly upregulated gene SERPINB3 is overexpressed in several human cancers, including head and neck SCC.29 The CDKN1A gene that encodes the cyclin-dependent kinase inhibitor p21 has divergent roles in tumorigenesis. CDKN1A overexpression correlates with chromosomal instability and is an adverse prognostic predictor for patients with ESCC.30 Another upregulated gene, PTGES, the eicosanoid pathway enzyme that converts PGH2 to PGE2, is elevated in various types of cancer, and is often associated with increased expression of COX-2.31 Importantly, a similar association between PTGES and COX-2 overexpression is found in the ZD esophagus.

The finding here that the de novo methyltransferase gene DMNT3A is zinc-responsive suggests that DNA methylation32 may be a mechanism for transcriptional gene silencing of candidate genes in ZD esophagus. Thus, the downregulated gene G0S2 is epigenetically silenced through promoter hypermethylation in primary head and neck SCC development.33 Loss of the structural protein UPK1B expression in bladder carcinoma is attributed to methylation of a CpG island within the UPK1B promoter34 and hypermethylation of TPM1 gene is a mechanism for TPM1 silencing in breast cancer cell lines.35 Together, this study has identified relevant genetic changes, induced by zinc-deficiency, that may act to produce the protumorigenic microenvironment in the ZD esophagus that favors carcinogenesis.

The role of zinc-deficiency as a cause of disease and as a determinant factor in the progression of disease is gaining attention.36 This concept is based on the many biological functions of zinc. Zinc is an essential nutrient. Zinc plays an important role in the immune system. There are more than 300 zinc-containing enzymes. Zinc ions are key structural components of nearly 2000 proteins that are mainly nuclear transcription factors, which contribute to control of cell proliferation, differentiation, and apoptosis through regulation of gene expression.37 It is now established that zinc deficiency induces oxidative stress and activates/inhibits oxidant-sensitive transcription factors that can affect cell function, proliferation and survival, thereby predisposing to disease.38 Interestingly, S100A8 was strongly induced in mouse keratinocytes in response to UV-induced oxidative stress.39 It is not known, however, whether oxidative stress induced by zinc deficiency would lead to S100A8 induction in the ZD esophagus. Given the recent interest in the idea of an association of inflammation and the genesis and perpetuation of cancer,40, 41 the finding here that zinc regulates a key inflammation pathway in esophageal carcinogenesis provides a new understanding of the role of zinc in esophageal cancer initiation, progression, and prevention.

Supplementary Material



Funding: National Institutes of Health CA118560 to Louise YY Fong

Abbreviations used in this paper

S100 calcium binding protein A8
receptor for advanced glycation end products
upper aerodigestive tract cancer
zinc- deficiency
nuclear factor-κB
squamous cell carcinoma
proliferating cell nuclear antigen
Expression Analysis Systematic Explorer


“No conflicts of interest exist”


1. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55:74–108. [PubMed]
2. van Rensburg SJ. Epidemiologic and dietary evidence for a specific nutritional predisposition to esophageal cancer. J Natl Cancer Inst. 1981;67:243–51. [PubMed]
3. Yang CS. Research on esophageal cancer in China: a review. Cancer Res. 1980;40:2633–44. [PubMed]
4. Abnet CC, Lai B, Qiao YL, Vogt S, Luo XM, Taylor PR, Dong ZW, Mark SD, Dawsey SM. Zinc concentration in esophageal biopsy specimens measured by xray fluorescence and esophageal cancer risk. J Natl Cancer Inst. 2005;97:301–6. [PubMed]
5. Fong LY, Li JX, Farber JL, Magee PN. Cell proliferation and esophageal carcinogenesis in the zinc-deficient rat. Carcinogenesis. 1996;17:1841–8. [PubMed]
6. Fong LY, Nguyen VT, Farber JL, Huebner K, Magee PN. 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. 2000;60:4589–95. [PubMed]
7. Fong LY, Zhang L, Jiang Y, Farber JL. Dietary zinc modulation of COX-2 expression and lingual and esophageal carcinogenesis in rats. J Natl Cancer Inst. 2005;97:40–50. [PubMed]
8. Fong LY, Jiang Y, Riley M, Liu X, Smalley KJ, Guttridge DC, Farber JL. Prevention of upper aerodigestive tract cancer in zinc-deficient rodents: inefficacy of genetic or pharmacological disruption of COX-2. Int J Cancer. 2008;122:978–89. [PMC free article] [PubMed]
9. Fong LY, Nguyen VT, Farber JL. Esophageal cancer prevention in zinc-deficient rats: rapid induction of apoptosis by replenishing zinc. J Natl Cancer Inst. 2001;93:1525–33. [PubMed]
10. Liu CG, Zhang L, Jiang Y, Chatterjee D, Croce CM, Huebner K, Fong LY. Modulation of gene expression in precancerous rat esophagus by dietary zinc deficit and replenishment. Cancer Res. 2005;65:7790–9. [PubMed]
11. Blanchard RK, Moore JB, Green CL, Cousins RJ. Modulation of intestinal gene expression by dietary zinc status: effectiveness of cDNA arrays for expression profiling of a single nutrient deficiency. Proc Natl Acad Sci U S A. 2001;98:13507–13. [PubMed]
12. tom Dieck H, Doring F, Roth HP, Daniel H. Changes in rat hepatic gene expression in response to zinc deficiency as assessed by DNA arrays. J Nutr. 2003;133:1004–10. [PubMed]
13. Storey JD, Tibshirani R. Statistical methods for identifying differentially expressed genes in DNA microarrays. Methods Mol Biol. 2003;224:149–57. [PubMed]
14. Dennis G, Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA. DAVID: Database for Annotation, Visualization, and Integrated Discovery. Genome Biol. 2003;4:P3. [PubMed]
15. Ridd K, Zhang SD, Edwards RE, Davies R, Greaves P, Wolfreys A, Smith AG, Gant TW. Association of gene expression with sequential proliferation, differentiation and tumor formation in murine skin. Carcinogenesis. 2006;27:1556–66. [PubMed]
16. Gebhardt C, Nemeth J, Angel P, Hess J. S100A8 and S100A9 in inflammation and cancer. Biochem Pharmacol. 2006;72:1622–31. [PubMed]
17. Fong LY, Jiang Y, Farber JL. Zinc deficiency potentiates induction and progression of lingual and esophageal tumors in p53-deficient mice. Carcinogenesis. 2006;27:1489–96. [PubMed]
18. Hummerich L, Muller R, Hess J, Kokocinski F, Hahn M, Furstenberger G, Mauch C, Lichter P, Angel P. Identification of novel tumour-associated genes differentially expressed in the process of squamous cell cancer development. Oncogene. 2006;25:111–21. [PubMed]
19. Bax DA, Siersema PD, Haringsma J, Kuipers EJ, Vos AJ, Van Dekken H, Van Vliet AH, Kusters JG. High-grade dysplasia in Barrett's esophagus is associated with increased expression of calgranulin A and B. Scand J Gastroenterol. 2007;42:902–10. [PubMed]
20. Luo A, Kong J, Hu G, Liew CC, Xiong M, Wang X, Ji J, Wang T, Zhi H, Wu M, Liu Z. Discovery of Ca2+-relevant and differentiation-associated genes downregulated in esophageal squamous cell carcinoma using cDNA microarray. Oncogene. 2004;23:1291–9. [PubMed]
21. Kumar A, Chatopadhyay T, Raziuddin M, Ralhan R. Discovery of deregulation of zinc homeostasis and its associated genes in esophageal squamous cell carcinoma using cDNA microarray. Int J Cancer. 2007;120:230–42. [PubMed]
22. Hermani A, De Servi B, Medunjanin S, Tessier PA, Mayer D. S100A8 and S100A9 activate MAP kinase and NF-kappaB signaling pathways and trigger translocation of RAGE in human prostate cancer cells. Exp Cell Res. 2006;312:184–97. [PubMed]
23. Ghavami S, Rashedi I, Dattilo BM, Eshraghi M, Chazin WJ, Hashemi M, Wesselborg S, Kerkhoff C, Los M. S100A8/A9 at low concentration promotes tumor cell growth via RAGE ligation and MAP kinase-dependent pathway. J Leukoc Biol. 2008 [PMC free article] [PubMed]
24. Gebhardt C, Riehl A, Durchdewald M, Nemeth J, Furstenberger G, Muller-Decker K, Enk A, Arnold B, Bierhaus A, Nawroth PP, Hess J, Angel P. RAGE signaling sustains inflammation and promotes tumor development. J Exp Med. 2008;205:275–85. [PMC free article] [PubMed]
25. Taguchi A, Blood DC, del Toro G, Canet A, Lee DC, Qu W, Tanji N, Lu Y, Lalla E, Fu C, Hofmann MA, Kislinger T, Ingram M, Lu A, Tanaka H, Hori O, Ogawa S, Stern DM, Schmidt AM. Blockade of RAGE-amphoterin signalling suppresses tumour growth and metastases. Nature. 2000;405:354–60. [PubMed]
26. Greten FR, Karin M. The IKK/NF-kappaB activation pathway-a target for prevention and treatment of cancer. Cancer Lett. 2004;206:193–9. [PubMed]
27. Prasad AS, Bao B, Beck FW, Kucuk O, Sarkar FH. Antioxidant effect of zinc in humans. Free Radical Biology & Medicine. 2004;37:1182–90. [PubMed]
28. Uzzo RG, Crispen PL, Golovine K, Makhov P, Horwitz EM, Kolenko VM. Diverse effects of zinc on NF-kappaB and AP-1 transcription factors: implications for prostate cancer progression. Carcinogenesis. 2006;27:1980–90. [PubMed]
29. Stenman J, Hedstrom J, Grenman R, Leivo I, Finne P, Palotie A, Orpana A. Relative levels of SCCA2 and SCCA1 mRNA in primary tumors predicts recurrent disease in squamous cell cancer of the head and neck. Int J Cancer. 2001;95:39–43. [PubMed]
30. Goan YG, Hsu HK, Chang HC, Chou YP, Chiang KH, Cheng JT. Deregulated p21(WAF1) overexpression impacts survival of surgically resected esophageal squamous cell carcinoma patients. Ann Thorac Surg. 2005;80:1007–16. [PubMed]
31. Yoshimatsu K, Golijanin D, Paty PB, Soslow RA, Jakobsson PJ, DeLellis RA, Subbaramaiah K, Dannenberg AJ. Inducible microsomal prostaglandin E synthase is overexpressed in colorectal adenomas and cancer. Clin Cancer Res. 2001;7:3971–6. [PubMed]
32. Baylin SB, Herman JG, Graff JR, Vertino PM, Issa JP. Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res. 1998;72:141–96. [PubMed]
33. Tokumaru Y, Yamashita K, Osada M, Nomoto S, Sun DI, Xiao Y, Hoque MO, Westra WH, Califano JA, Sidransky D. Inverse correlation between cyclin A1 hypermethylation and p53 mutation in head and neck cancer identified by reversal of epigenetic silencing. Cancer Res. 2004;64:5982–7. [PubMed]
34. Varga AE, Leonardos L, Jackson P, Marreiros A, Cowled PA. Methylation of a CpG island within the uroplakin Ib promoter: a possible mechanism for loss of uroplakin Ib expression in bladder carcinoma. Neoplasia. 2004;6:128–35. [PMC free article] [PubMed]
35. Bharadwaj S, Prasad GL. Tropomyosin-1, a novel suppressor of cellular transformation is downregulated by promoter methylation in cancer cells. Cancer Lett. 2002;183:205–13. [PubMed]
36. Maret W, Sandstead HH. Zinc requirements and the risks and benefits of zinc supplementation. J Trace Elem Med Biol. 2006;20:3–18. [PubMed]
37. Berg JM, Shi Y. The galvanization of biology: a growing appreciation for the roles of zinc. Science. 1996;271:1081–5. [PubMed]
38. Herbein G, Varin A, Fulop T. NF-kappaB, AP-1, Zinc-deficiency and aging. Biogerontology. 2006;7:409–19. [PubMed]
39. Grimbaldeston MA, Geczy CL, Tedla N, Finlay-Jones JJ, Hart PH. S100A8 induction in keratinocytes by ultraviolet A irradiation is dependent on reactive oxygen intermediates. J Invest Dermatol. 2003;121:1168–74. [PubMed]
40. Perwez Hussain S, Harris CC. Inflammation and cancer: an ancient link with novel potentials. Int J Cancer. 2007;121:2373–80. [PubMed]
41. Balkwill F, Charles KA, Mantovani A. Smoldering and polarized inflammation in the initiation and promotion of malignant disease. Cancer Cell. 2005;7:211–7. [PubMed]