Matrix metalloproteinase-2 (MMP-2) plays important roles in cancer development and aggression. Our previous studies revealed a strong association between the MMP-2 −1306C/T polymorphism and risk of several cancers. A novel −735C/T polymorphism in MMP-2 promoter has been identified but the function is undefined. This study examined our hypothesis that these two polymorphisms might have functional relevance and impact on risk of esophageal squamous cell carcinoma in the context of haplotype. Genotypes and haplotypes were analyzed in 527 cases and 777 controls and odds ratios (ORs) and 95% confidence intervals (CIs) were estimated by logistic regression. The function of the polymorphisms was examined by electrophoretic mobility shift assays, luciferase gene expression assays, and reverse transcriptase-PCR analyses. It was found that the −735C→T transition disrupts an Sp1 site and displays a lower promoter activity. The C-1306–C-735 haplotype had 7-fold increased luciferase expression and 3.7-fold increased MMP-2 mRNA levels in esophageal tissues compared with the T-1306–T-735 haplotype. A case-control analysis revealed a 1.52-fold (95% CI = 1.17–1.96) or 1.30-fold (95% CI = 1.04–1.63) excess risk of developing esophageal squamous cell carcinoma for the −1306CC or −735CC genotype carriers compared with noncarriers, respectively. A greater association was observed between elevated risk of developing esophageal squamous cell carcinoma and C-1306 or C-735 allele containing haplotypes, with the risk being highest for the C-1306–C-735 haplotype compared with the T-1306–T-735 haplotype (OR = 6.53; 95% CI = 2.78–15.33). The C-1306–C-735 haplotype was also associated with increased risk for distant metastasis of esophageal squamous cell carcinoma (OR = 3.34; 95% CI = 1.16–9.63). These findings suggest that the C-1306–C-735 haplotype in the MMP-2 promoter contributes to risk of the occurrence and metastasis of esophageal squamous cell carcinoma by increasing expression of MMP-2.

Accumulating evidence has shown that matrix metalloproteinases (MMPs) play a critical role in cancer development and aggression because of their capability to degrade or break down both extracellular matrix and basement membrane, the important physical barriers in preventing against expanding growth and migration of cancer cells (1, 2, 3). Moreover, recent studies have demonstrated that MMPs also have many other substrates involving in several steps of cancer development such as apoptosis, cell proliferation, angiogenesis, and immune surveillance (4). MMP-2, a member of MMPs, predominantly degrades gelatin and type IV collagen, the major structural component of basement membrane, and thus has been considered as an important factor in cancer invasion and metastasis (5, 6). Nevertheless, this enzyme also has activity toward a spectrum of functional molecules such as growth factor-binding proteins and growth factor receptors. For instance, MMP-2 can cleave insulin-like growth factor-binding proteins and release insulin-like growth factors (7), which are well known to have a strong effect on stimulating cell proliferation and inhibiting apoptosis. MMP-2 can also target fibroblast growth factor receptor 1, yielding active soluble ectodomain of the receptor (8), which has the potential to modulate the mitogenic and angiogenic activities of fibroblast growth factor. These findings suggest that MMP-2 contributes not only to cancer progression but also to cancer development (4).

MMP-2 is constitutively expressed by a large number of cell types and overexpressed in a wide variety of human cancers, including esophageal squamous cell carcinoma (9, 10, 11). Furthermore, the production of this proteinase in tumors is made by not only cancer cells but also normal stromal cells (9, 10, 11), suggesting that the overexpression of MMP-2 is probably due to transcriptional changes but not gene amplification or activating mutations. Because the human MMP-2 promoter contains a number of cis-acting regulatory elements, the constitutive and induced expression of this proteinase is likely to be subject to regulation by transcription factors (12, 13). Several single nucleotide polymorphisms (SNPs) in the MMP-2 promoter region have been identified (14). Among them, a C to T transition located at nucleotide −1306 abolishes a Sp1-binding site and consequently diminishes promoter activity. Transient transfection experiments showed that reporter gene expression driven by the C allelic MMP-2 promoter was significantly greater than reporter gene expression driven by the T allelic counterpart both in epithelial cells and in macrophages, indicating the functional significance of this polymorphism (14). Recently, another C to T transition located at nucleotide −735 in the promoter region of MMP-2 has been identified (15). but the functional significance is undefined. Bioinformatic analysis suggests that the −735C/T polymorphism might also disrupt a consensus sequence for Sp1-binding site, implying that this polymorphism might have the potential to alter MMP-2 transcription.

We have previously shown that the functional −1306C/T polymorphism in MMP-2 is associated with susceptibility to cancers of the lung, gastric cardia, and breast (16, 17, 18), suggesting that the MMP-2 polymorphism might be a general risk factor for common cancers. On the basis of these findings, we additionally hypothesized that the −735C/T polymorphism in MMP-2 might also have impact on individual susceptibility to cancer. Furthermore, because the −735C/T site is close to the −1306C/T site in the MMP-2 promoter, these two SNPs might be in linkage disequilibrium and act in an interaction manner. To test these hypotheses, we have examined the functional relevance of the −735C/T polymorphism, alone, and in combination with −1306C/T polymorphisms in the context of haplotypes. Moreover, we conducted a case-control study to investigate the relationship between the genotypes and haplotypes of these polymorphisms in MMP-2 promoter and risk of the occurrence and metastasis of esophageal squamous cell carcinoma.

Subjects for the Case-Control Study.

This study recruited 527 patients with esophageal squamous cell carcinoma and 777 healthy controls. All subjects were ethnic Han Chinese. Patients were consecutively recruited from January 1997 to November 2001 at the Cancer Hospital, Chinese Academy of Medical Sciences (Beijing, China). All patients with histopathologically confirmed esophageal squamous cell carcinoma were enrolled, yielding 92% response rate. A portion of cases (n = 240) was enrolled in our previous molecular epidemiologic studies of esophageal squamous cell carcinoma (19). The pathological stage of esophageal squamous cell carcinoma at the time of diagnosis was classified by senior pathologists of the hospital on the basis of postoperative histopathological examination or biopsy according to the Tumor-Node-Metastasis classification (20). Healthy controls were cancer-free individuals living in Beijing region, and they were selected from a community cancer-screening program for early detection of cancer conducted during the same period as the cases were collected (17). These controls were randomly selected from a pool of 2800 individuals based on a physical examination, and the response rate was 96%. The selection criteria included no individual history of cancer and frequency matched to esophageal squamous cell carcinoma cases on sex and age (±5 years). At recruitment, informed consent was obtained from each subject, and each participant was then interviewed to collect detailed information on demographic characteristics and lifetime history of tobacco use. This study was approved by the Institutional Review Board of the Chinese Academy of Medical Sciences Cancer Institute.

MMP-2 Genotyping.

Genomic DNA from controls and most of cases was isolated from blood samples. Approximate 30% DNA samples from cases were isolated from surgically resected normal tissues adjacent to the tumor of esophageal squamous cell carcinoma patients. MMP-2 genotypes at the −1306C/T site were determined by PCR-based denaturing high-performance liquid chromatography analysis as described previously (16, 17). MMP-2 genotypes at the −735C/T site were analyzed by PCR-based restriction fragment length polymorphism methods (15). The primers used for amplifying DNA containing the −735C/T site were 5′-ATAGGGTAAACCTCCCCACATT-3′ and 5′-GGTAAAATGAGGCTGAGACCTG-3′, which produce a 300-bp fragment. Digestion of the PCR product with HinfI (New England Biolabs, Beverly, MA) at 37°C overnight produced one fragment of 300 bp for the CC genotype; three fragments of 300, 254, and 46 bp for the CT genotype; and two fragments of 254 and 46 bp for the TT genotype. The genotypes identified by HinfI digestion were confirmed by DNA sequencing, and sequences were compared with the published MMP-2 promoter sequences (12). All of the genotyping was performed with blinding to case/control status, and blinded quality control samples were inserted to validate genotypes. Concordance for blinded samples was 100%.

Statistical Analysis.

χ2 test was used to compare the distribution of MMP-2 genotypes and haplotypes between cases and controls and between metastatic and nonmetastatic cases. Lighter or heavier smokers were categorized by the approximate 50th percentile pack-year value among controls, i.e., <26 or ≥26 pack-years [(cigarettes per day ÷ 20) × (years smoked)]. Hardy-Weinberg equilibrium was tested by a goodness-of-fit χ2 test. Haplotype frequencies and linkage disequilibrium coefficient were estimated using PHASE (21) and EH (EH-plus) software (22), respectively. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated by unconditional logistic regression as a measure of association with risk of the development and metastasis of esophageal squamous cell carcinoma. All ORs were adjusted for age, sex, and pack-years smoked. Statistical analyses were performed with Statistical Analysis System software (version 6.12; SAS Institute, Cary, NC).

Electrophoretic Mobility Shift Assay.

Synthetic double-stranded oligonucleotides 5′-GAGAATGCGGACCCTCCTGGGAGT-3′ and 5′-GAGAATGCGGACTCTCCTGGGAGT-3′corresponding to the −735C or −735T sequence from the MMP-2 promoter region were labeled with [γ-32P]ATP (Amersham Pharmacia Biotech, Piscataway, NJ) according to the Gel-Shift Assay Kit protocol (Promega, Madison, WI). For each gel shift reaction, an aliquot of labeled oligonucleotide (>30,000 cpm) was incubated with 10 μg of nuclear extract from HEK293 cells for 20 minutes at room temperature in a 20-μL reaction mixture containing 40 mmol/L HEPES (pH 7.9), 12.5% glycerol, 1.0 mmol/L EDTA, 2 μg poly(deoxyinosinic-deoxycytidylic acid), 0.1 μg of BSA, and 0.2 μg of sheared salmon sperm. For competition experiments, a 50 to 100-fold molar excess of unlabeled −735C or −735T probe, a Sp1 consensus-binding site (5′-ATTCGATCGGGGCGGGGCGAGC-3′), a Sp1-mutant consensus site (5′-ATTCGATCGGTTCGGGGCGAGC-3′), or a nonspecific oligonucleotide (5′-GAGCGCATACTGACTATCGGAGAC-3′) was preincubated for 10 minutes at room temperature with the nuclear extracts before the addition of the labeled probe. For supershift experiments, antibody against Sp1 and rabbit IgG (Upstate Biotechnology, Inc., Lake Placid, NY) was used. The antibody or rabbit IgG (2 μg) was incubated with nuclear extracts at 4°C for 30 minutes, followed by an additional incubation for 20 minutes at room temperature with a labeled Sp1 consensus-binding site or −735C probe. Samples were run on a nondenaturing 6% polyacrylamide gel in 0.5× Tris-borate EDTA, at 300 V for 1 to 2 hours. The gels were dried and visualized by autoradiography using Kodak X-Omat film at −80°C.

Construction of Reporter Plasmids.

A PCR-based approach was used to generate a construct encompassing −1691 to +10 of human MMP-2 promoter as described previously (14). The primers used for amplification of this DNA fragment were 5′-GTCAAAGCTTAAAACTGACTCTGGAAAGTCA-3′ and 5′-GTCAGGTACCTCTGGATGCAGCGGAAACAAG-3′, which contain 5′-HindIII and 3′-KpnI cloning sites (italicized sequences), respectively. To ensure high fidelity amplification, PfuTurbo polymerase (Promega) was used in the PCR. The PCR product was digested with HindIII and KpnI and ligated into an appropriately digested pGL3-Basic vector (Promega). The resulting construct was designated as pT–T because sequence analysis demonstrated that it contains a T at both −1306 and −735 polymorphic sites. This pT–T product was subsequently used as a template to generate other three constructs containing a C at −1306 and/or −735 positions by using the site-specific mutagenesis, respectively, which were denominated as pC–T, pT–C, and pC–C. All constructs used in this study were restriction mapped and sequenced to confirm their authenticity.

Transient Transfections and Luciferase Assays.

HEK293 cells were grown in RPMI 1640 supplemented with 10% (v/v) heat-inactivated FCS, 2 mmol/L l-glutamine, 100 units/mL penicillin, and 100 units/mL streptomycin at 37°C and 5% CO2 in a humidified incubator. For transient transfection experiments, 5 × 104 cells were plated in 10-mm 24-multiwell plates and grown to 60 to 70% confluence. Transfection was carried out using LipofectAMINE Reagent (Life Technologies, Inc., Rockville, MD) according to the manufacturer’s protocol. Cells were cotransfected with 0.5 μg of reporter plasmid and 0.1 μg of pRL-SV40 (Luciferase Assay System; Promega); the latter was used to standardize transfection efficiency. Luciferase activity was determined according to the manufacturer’s protocol using a luciferase assay system (Promega). Briefly, cells were scraped into lysis reagent, transferred to microfuge tubes and centrifuged for 30 seconds at 12,000 × g. Luciferase activity was measured using a manual luminometer (Turner Designs; TD 20/20) by mixing 100 μL of luciferase assay reagent with 20 μL of 1:10-diluted cell lysate, and the value for each sample was recorded three times at 10-second intervals. For each plasmid construct, three independent transfection experiments were performed, and each was done in triplicate. Results were expressed as a ratio of luciferase activity to pRL-SV40. Differences were determined by t test, and P of <0.01 was considered significant.

Reverse Transcriptase-PCR Analysis.

Fifty-two normal esophageal tissues adjacent to the tumors were obtained from surgically removed specimens of individual patients. The normal tissues sampled at least 5 cm away from the margin of the tumor were immediately placed in liquid nitrogen and then stored at −80°C before analysis. Total RNA was isolated from tissues using the Trizol Reagent (Life Technologies, Inc.). An aliquot of total RNA (2 μg) from each specimen was reverse transcribed into single-strand cDNA using oligo (dT)15 primer and Superscript II (Life Technologies, Inc.). Each single-strand cDNA was diluted for subsequent PCR amplification of MMP-2 and β-actin, with the latter being used as an internal quantitative control. The primers used for amplification of MMP-2 were 5′-TTCAAGGACCGGTTCATTTGGCGGACTGTG-3′ and 5′-TTCCAAACTTCACGCTCTTCAGACTTTGGTT-3′, which produce a 493-bp fragment, and for β-actin were 5′-CGTCTGGACCTGGCTGGCCGGGACC-3′ and 5′-CTAGAAGCATTTGCGGTGGACGATG-3′, which generate a 601-bp fragment. PCR was carried out under the following conditions: an initial denaturing step of 5 minutes at 95°C, followed by 30 cycles of 40 seconds at 94°C, 40 seconds at 60°C, and 40 seconds at 72°C, and a final elongation step of 7 minutes at 72°C. PCR products were separated and visualized in 1.5% agarose gel containing ethidium bromide and quantified by using a UVP GDS-8000 image analysis system (UVP, Inc., Upland, CA). The relative density of the MMP-2 band was calculated based on the density of the β-actin band in each sample.

Disruption of Sp1-Binding Site in the MMP-2 Promoter by the −735T SNP.

By using the AliBaba 2 software,1 we found that the −735C/T polymorphism may alter a consensus sequence for Sp1-binding site (CCCTCC→CTCTCC). Electrophoretic mobility shift assays were thus designed to examine whether the Sp1 consensus sequence is abolished by the presence of a T at the −735 site. As shown in Fig. 1, a clear DNA-protein complex was detected with the −735C probe (Fig. 1, Lane 2) but not the −735T probe (Fig. 1, Lane 9) in the assays. To determine the sequence specificity of this DNA-protein complex, competition experiments were performed. The band was competed by 50- and 100-fold excess of unlabeled −735C probe (Fig. 1, Lanes 3 and 4) but not by the same concentrations of unlabeled −735T probe (Fig. 1, Lanes 5 and 6). The specificity of this band was additionally confirmed by the addition of 100-fold excess of nonspecific competitor (Fig. 1, Lane 7). These assays clearly demonstrated the ability of the −735C allele, but not the −735T allele, to bind specifically the nuclear protein. We next performed electrophoretic mobility shift assays using above described −735C probe and an Sp1 consensus probe, and the results are shown in Fig. 2,A. The labeled Sp1 consensus probe formed a specific DNA-protein complex (Fig. 2,A, Lane 2), which was identical to that observed with the −735C probe (Lane 9) and was eliminated by excess unlabeled probe (Fig. 2,A, Lanes 3, 10, and 14, respectively). Lanes 6 and 13 in Fig. 2,A demonstrate the specificity of the DNA-protein complex by competition experiments with 100-fold excess of unlabeled mutated Sp1 consensus probe. Furthermore, the ability of the −735C but not the −735T allele to compete specifically for Sp1 binding was confirmed by additional competition experiments using 100-fold excess of unlabeled −735T probe (Fig. 2,A, Lanes 4 and 12), 50- or 100-fold excess of unlabeled −735C probe (Fig. 2,A, Lanes 5 and 7), and 100-fold excess of unlabeled Sp1 consensus probe (Lane 11). Supershift assays were then performed to confirm Sp1 binding using the Sp1 consensus probe (Fig. 2,B, Lanes 1–4) or −735C probe (Fig. 2,B, Lanes 5–8) in either the absence (Fig. 2,B, Lanes 1 and 5) or presence of antibody against Sp1 or rabbit IgG (Fig. 2,B, Lanes 2, 3, 6, and 7, respectively). The DNA-protein complex was successfully supershifted with the anti-Sp1 antibody (Fig. 2,B, Lanes 2 and 6) but not the rabbit IgG (Fig. 2 B, Lanes 3 and 7). Taken together, these results clearly demonstrate that nucleotide −735C/T in the MMP-2 promoter is within a Sp1-binding sequence and the C to T substitution disrupts the binding site.

Effects of MMP-2 −735C/T and −1306C/T SNPs on Transcriptional Activity.

To directly determine the allele-specific effects of Sp1 binding on native promoter activity, four luciferase reporter gene constructs were generated by PCR, spanning −1691 to +10 of the MMP-2 promoter region, with either a T or C at the −735 and −1306 polymorphic sites (Fig. 3,A), and they were used to transfect transiently HEK293 cells. As shown in Fig. 3,B, reporter gene expression driven by the −735C allelic MMP-2 promoter was ∼3-fold greater than that driven by the −735T allelic counterpart (3.54 ± 0.71 versus 1.24 ± 0.38, P < 0.0001), indicating the biological significance of this SNP on promoter activity via recruitment of Sp1 transcriptional factor. The combined effects of the −735C/T and −1306C/T SNPs on transcriptional activity were further assessed. Interestingly, the expression levels of reporter gene driven by the pT–T, pT–C, pC–T, and pC–C alleles increased gradually, with the values being 1.24 ± 0.38, 3.54 ± 0.71, 5.28 ± 0.84, and 8.79 ± 1.01, respectively (Fig. 3,B). These results indicate a strong synergic effect between −735T and −1306T SNPs in the context of haplotype on reducing transcriptional activity of MMP-2. The effect of these two SNPs on MMP-2 transcriptional activity was additionally examined by reverse transcriptase-PCR analysis of MMP-2 mRNA in individual esophageal tissues (Fig. 4). MMP-2 mRNA levels were significantly lower in individuals with the −1306CT or TT genotype than that in those with the −1306CC genotype [0.04 ± 0.04 (n = 15) versus 0.20 ± 0.26 (n = 37) arbitrary units; P = 0.022]. For the −735C/T polymorphism, the mean level of the mRNA was also lower in individuals with the −735CT or TT genotype compared with the −735CC genotype, although the difference did not reach statistical significance [0.08 ± 0.06 (n = 19) versus 0.20 ± 0.28 (n = 33); P = 0.078]. We also compared MMP-2 mRNA levels in function of haplotypes, and individuals with the MMP-2 C-1306–C-735 haplotype had significantly higher mRNA levels in their esophageal tissues compared with those with the haplotype consisting of at least one T allele at the −1306 or −735 site [0.26 ± 0.32 (n = 23) versus 0.07 ± 0.06 (n = 29); P = 0.003].

Genotypes and Risk of Esophageal Squamous Cell Carcinoma.

Having the functional consequences of these two SNPs, we next conducted a case-control study to examine the association of them with risk of esophageal squamous cell carcinoma. There were no significant differences between cases and controls in sex distribution (75 versus 72% males). The mean age (±SD, years) for cases and controls were 58.3 ± 9.7 and 57.6 ± 7.6, respectively (P = 0.187). More smokers were presented in cases compared with controls (60.3 versus 47.4%, P < 0.0001). Of the 486 patients who had detailed pathological data, 41 (8.4%) patients had stage I esophageal squamous cell carcinoma, 190 (39.1%) had stage II esophageal squamous cell carcinoma, and 239 (49.2%) had stage III esophageal squamous cell carcinoma, whereas only 16 (3.3%) patients had stage IV disease with distant metastasis. The genotyping results are shown in Table 1. The allele frequencies for −1306C and −735C were 0.836 and 0.749 in controls, compared with 0.882 and 0.783 in cases, respectively. The observed genotype frequencies of both −1306C/T and −735C/T sites in controls conformed to the Hardy-Weinberg equilibrium (P = 0.367 and 0.836, respectively). The frequencies for the −1306CC, CT, and TT genotypes in cases differed significantly from those in controls (χ2 = 11.45, P = 0.003, df = 2), with the CC homozygotes being higher in cases than in controls (77.6 versus 69.4%, P = 0.001). The difference in genotype frequencies at the −735C/T site between cases and controls was borderline significant (χ2 = 5.75, P = 0.056, df = 2). Because of functional significance of the heterozygous genotypes (−1306CT and −735CT), they were combined with the respective homozygous −1306TT or −735TT genotypes for risk estimation (Table 1). Multivariate logistic regression analysis showed that the −1306CC carriers had a 1.5-fold excess risk for developing esophageal squamous cell carcinoma compared with noncarriers (adjusted OR = 1.52; 95% CI = 1.17–1.96). Similarly, the −735CC carriers also had increased risk for esophageal squamous cell carcinoma compared with noncarriers (adjusted OR = 1.30; 95% CI = 1.04–1.63). In the stratification analysis, age, sex, and smoking had no effect on the risk of esophageal squamous cell carcinoma related to the MMP-2 genotypes (data not shown).

Haplotypes and Risk of Esophageal Squamous Cell Carcinoma.

The interactions of multiple SNPs within a haplotype may have impact on biological phenotype (23). We thus analyzed the effect on risk of esophageal squamous cell carcinoma by these two SNPs in the MMP-2 promoter in the context of haplotypes. The haplotype frequencies were computed from unphased genotypes using PHASE software and the results are presented in Table 2. We observed a significant difference in haplotype frequencies between cases and controls (χ2 = 26.63, P < 0.0001, df = 3). Compared with the T-1306–T-735 haplotype, each of the other haplotype containing at least one −1306C or −735C allele was associated with increased risk of esophageal squamous cell carcinoma. The adjusted ORs of esophageal squamous cell carcinoma for the T-1306–C-735, C-1306–T-735, and C-1306–C-735 haplotypes were OR = 5.19 (95% CI = 2.14–12.59), OR = 6.04 (95% CI = 2.52–14.48), and OR = 6.53 (95% CI = 2.78–15.33), respectively (trend test, P < 0.001). A linkage disequilibrium of these two SNPs was observed. The χ2 test of statistical significance for a two-locus disequilibrium gave a test statistic value of 18.2 (D′ = 0.68) for the cases, 11.0 (D′ = 0.33) for the controls, and 22.6 (D′ = 0.40) for all subjects. The linkage disequilibrium was statistically significant (P < 0.001).

Haplotypes and Disease Status of Esophageal Squamous Cell Carcinoma.

The association between the MMP-2 −1306C/T and −735C/T polymorphisms and esophageal squamous cell carcinoma disease stage at the time of diagnosis was additionally evaluated. We did not observe any significant association between the disease stage and −735C/T polymorphism alone or in combination with −1306C/T polymorphism (data not shown). However, the advanced esophageal squamous cell carcinoma stage appeared to be associated with the haplotype, including these two polymorphisms. The frequency of the C-1306–C-735 haplotype in patients with stage IV esophageal squamous cell carcinoma was 87.5% (28 of 32), which was significantly higher than that in patients with stage I–III tumor (67.1%, 631 of 940; χ2 = 4.99, P = 0.026). Multivariate logistic regression analysis showed that patients carrying the MMP-2 promoter C-1306–C-735 haplotype had >3-fold increased risk for developing distant metastasis of esophageal squamous cell carcinoma, compared with other haplotypes consisting of at least one −1306T or −735T allele (adjusted OR = 3.34; 95% CI = 1.16–9.63).

Esophageal squamous cell carcinoma is one of the most common cancers in the world. The prognosis of this cancer is poor with the overall 5-year survival rate being ∼20% (24). Tobacco smoke, alcohol, nutritional deficiency, and exposure to certain chemical carcinogens are thought to be risk factors for esophageal squamous cell carcinoma. However, not all exposed individuals develop the disease, suggesting that genetic factors may play important roles in esophageal carcinogenesis (25). We sought to identify genetic factors that confer individual susceptibility to the cancer. We analyzed 527 esophageal squamous cell carcinoma patients and 777 controls for the −1306C/T and −735C/T polymorphisms in the MMP-2 promoter and observed that these two polymorphisms were associated with increased risk for developing the cancer. The −1306CC and −735CC genotypes appear to have strong interaction within a haplotype to influence the cancer risk; carriers of the C-1306–C-735 haplotype had >6-fold increased risk for developing esophageal squamous cell carcinoma compared with noncarriers. These extended results are consistent with our previous findings showing that the −1306C/T polymorphism may be genetic risk factor for cancers of the lung, gastric cardia, and breast (16, 17, 18). Moreover, we also observed a significant association between the MMP-2 haplotype and distant metastasis of esophageal squamous cell carcinoma. Patients with the C-1306–C-735 haplotype were at >3-fold increased risk for having distant metastasis of the cancer at the time of diagnosis.

Our functional analysis suggested that the association of MMP-2 haplotypes with increased risk of the occurrence and metastasis of esophageal squamous cell carcinoma might be attributed to gain-of-function of this gene resulting from the promoter SNPs. We found that the −735C/T SNP locates in a core recognition sequence of Sp1 in the MMP-2 promoter region. Through electrophoretic mobility shift assays, the DNA-Sp1 complex was detected as binding to the −735C allele but not the −735T allele. Competition assays combined with supershift analysis additionally confirmed that the protein binding to this region is Sp1. The potential cis-acting regulatory elements in the MMP-2 promoter region have been extensively investigated by Qin et al. (13). However, they did not find an Sp1 site around nucleotide −735 in their study. Although the reason for this difference between their study and ours is not known, it is most likely that the MMP-2 promoter they used as starting material in their study was the −735TT genotype in which the Sp1 site was not present. For the −1306C/T polymorphism, a similar result showing disruption of an Sp1 promoter site by the C→T transition has been shown previously by Price et al. (14). Because Sp1 is a ubiquitously expressed transcriptional factor that regulates a variety of gene in a constitutive or inducible manner (13, 26, 27, 28), it is clear that sequence variations that destroy the Sp1-binding sites such as the MMP-2 −735C/T and −1306C/T polymorphisms may alter the level and specificity of gene transcription. Our luciferase assays and MMP-2 mRNA analysis in esophageal tissues indeed demonstrated a significant difference in transcriptional activity between the −735C and −735T alleles. More importantly, we observed a remarkably elevated transcriptional activity when the −1306C and −735C alleles (C-1306–C-735 haplotype) are concomitantly presented in the MMP-2 promoter, indicating an interaction between these two SNPs within a haplotype. Taken together, theses data strongly suggest that the presence of Sp1-binding sequences in the −1306C and −735C alleles enhance MMP-2 transcription, which in turn produces higher levels of MMP-2 protein in C-1306–C-735 carriers than in noncarriers. Because MMP-2 plays important roles in all stages of cancer initiation and development (4), it would be expected that individuals who carry the C-1306–C-735 haplotype and therefore have elevated expression of this enzyme over lifetime may be at highest susceptibility to carcinogenesis.

The association between high level of constitutive expression of MMP-2 and susceptibility to tumor formation has been tested in several studies with genetically modified animals. It was found that when induced by carcinogenic stimulus, mice that lack the Mmp2 or Mmp9 gene developed fewer tumors than mice having these genes (29). Cancer cells injected via vein were found to be less capable of colonizing the lungs of Mmp2-knockout mice than the lungs of wild-type mice (30). Conversely, transgenic mice that overexpress membrane type MMP-1, a known activator of pro-MMP-2, were at increased susceptibility to mammary tumor formation and metastasis (31). These data strongly support our observation that the genetic polymorphisms resulting in high expression of MMP-2 over a lifetime may increase cancer susceptibility. In addition, functional polymorphisms in some other MMP genes have also been reported to be associated with risk of certain cancers. For instance, a single guanine insertion polymorphism in the MMP-1 promoter region (1G to 2G), which creates an Ets binding site and enhance transcriptional activity, has been associated with increased risk for lung (32), colorectal (33, 34), endometrial (35), and ovarian cancer (36). A single adenosine insertion polymorphism in the MMP-3 promoter (6A allele), which has half the transcriptional activity of the 5A allele, has been linked to reduced susceptibility to breast cancer (37).

Local overexpression of MMP-2 has been shown to be related to lymph node metastasis and poor prognosis of certain cancers, including esophageal squamous cell carcinoma (5, 6, 10, 11, 38). Several studies has suggested that genetic polymorphism in the promoter of MMP-1 (1G/2G) or MMP-3 (5A/6A), which alters the transcription activity of the genes, may influence invasiveness or metastasis of some types of cancer (33, 37, 39). In the present study, we did not find any significant association between pathological stage of esophageal squamous cell carcinoma and MMP-2 polymorphisms at either −1306C/T or −735C/T site. These findings are consistent with our previous observation in gastric cardia cancer study showing that −1306C/T polymorphism was not associated with lymph node metastasis (17). However, in the present study, we did observe a statistical significant association between distant metastasis of esophageal squamous cell carcinoma and the two-site C-1306–C-735 haplotype of the −1306C/T and −735C/T polymorphisms. Again, these findings are in agreement with functional analysis of the polymorphisms and may reflect the fact that haplotype may be more precise and powerful than genotype as genetic marker for risk estimate. Additional examinations of larger patient series with more detailed clinic outcome especially the survival rate are warranted. In addition, it would be interesting to investigate haplotypes tagging more SNPs in MMP-2 or other related genes as genetic risk factors for the occurrence and progression of esophageal squamous cell carcinoma and other cancers.

Our case-control study has some drawbacks. Because patients were recruited from only one hospital and controls were from a cancer-screening program, the study subjects may not be representative of the general population. However, because we used incident cases and recruited a large number of subjects, our results are unlikely to be attributable to selection bias. Nevertheless, it would be important to confirm these findings in multicenter case-control studies or in population-based prospective studies.

In summary, our study identified a novel Sp1 regulatory element in the promoter of MMP-2 and suggests that the haplotype for the −1306C/T and −735C/T polymorphisms in the MMP-2 promoter is a genetic susceptibility factor for the occurrence and metastasis of esophageal squamous cell carcinoma in a Chinese population. These molecular epidemiologic findings are consistent with the results obtained from functional analysis. Because MMP-2 is overexpressed in many cancer types, additional studies on other types of common cancers would be warranted in different ethnic populations.

Fig. 1.

Electrophoretic mobility shift assays with 32P-labeled 24-bp oligonucleotides containing the −735C or −735T allele and nuclear extracts from HEK293 cells. Lanes 1 and 8 show the mobilities of the labeled oligonucleotides without nuclear extracts; Lanes 2 and 9 present the mobilities of the labeled oligonucleotides with nuclear extracts in the absence of competitor; Lanes 37 and 10–14 present the mobilities of the labeled oligonucleotides with nuclear extracts in the presence of various unlabeled competitors as indicated below the bottom autoradiograph. The arrow localizes a major oligonucleotide-nuclear protein complex.

Fig. 1.

Electrophoretic mobility shift assays with 32P-labeled 24-bp oligonucleotides containing the −735C or −735T allele and nuclear extracts from HEK293 cells. Lanes 1 and 8 show the mobilities of the labeled oligonucleotides without nuclear extracts; Lanes 2 and 9 present the mobilities of the labeled oligonucleotides with nuclear extracts in the absence of competitor; Lanes 37 and 10–14 present the mobilities of the labeled oligonucleotides with nuclear extracts in the presence of various unlabeled competitors as indicated below the bottom autoradiograph. The arrow localizes a major oligonucleotide-nuclear protein complex.

Close modal
Fig. 2.

A, electrophoretic mobility shift assays with 32P-labeled oligonucleotides containing Sp1 consensus site (Lanes 1–7) or the −735C allele (Lanes 8–14) and nuclear extracts from HEK293 cells. Specificity of nuclear protein binding is demonstrated by competition with cold excess oligonucleotides containing Sp1 consensus site (Lanes 3 and 11) versus mutated Sp1 consensus site (Lanes 6 and 13) or the −735T allele (Lanes 4 and 12) versus −735C allele (Lanes 5, 7, 10, and 14). The arrow localizes a major oligonucleotide-nuclear protein complex. B, electrophoretic mobility shift assays with 32P-labeled oligonucleotides containing Sp1 consensus site (Lanes 1–3) or the −735C allele (Lanes 5–7) and nuclear extracts from HEK293 cells in the absence or presence of anti-Sp1 antibody. Antibody depletion analysis was carried out by preincubating the nuclear extracts with anti-Sp1 antibody (Lanes 2 and 6) or rabbit IgG (Lanes 3 and 7). Labeled oligonucleotides incubated without the nuclear extracts were included as negative controls (Lanes 4 and 8). The arrows localize a major oligonucleotide-nuclear protein complex (I) and the supershifted band of the protein complex by anti-Sp1 antibody (SS).

Fig. 2.

A, electrophoretic mobility shift assays with 32P-labeled oligonucleotides containing Sp1 consensus site (Lanes 1–7) or the −735C allele (Lanes 8–14) and nuclear extracts from HEK293 cells. Specificity of nuclear protein binding is demonstrated by competition with cold excess oligonucleotides containing Sp1 consensus site (Lanes 3 and 11) versus mutated Sp1 consensus site (Lanes 6 and 13) or the −735T allele (Lanes 4 and 12) versus −735C allele (Lanes 5, 7, 10, and 14). The arrow localizes a major oligonucleotide-nuclear protein complex. B, electrophoretic mobility shift assays with 32P-labeled oligonucleotides containing Sp1 consensus site (Lanes 1–3) or the −735C allele (Lanes 5–7) and nuclear extracts from HEK293 cells in the absence or presence of anti-Sp1 antibody. Antibody depletion analysis was carried out by preincubating the nuclear extracts with anti-Sp1 antibody (Lanes 2 and 6) or rabbit IgG (Lanes 3 and 7). Labeled oligonucleotides incubated without the nuclear extracts were included as negative controls (Lanes 4 and 8). The arrows localize a major oligonucleotide-nuclear protein complex (I) and the supershifted band of the protein complex by anti-Sp1 antibody (SS).

Close modal
Fig. 3.

Transient reporter gene expression assays with constructs containing full-length MMP-2 promoter. A, schematic of reporter gene constructs having a full-length MMP2 promoter with the only difference between the four constructs being a T or C at the −1306 and −735 polymorphic sites. B, luciferase expression of the four constructs in HEK293 cells cotransfected with pRL-SV40 to standardize transfection efficiency. Luciferase levels of pGL-3 Basic and pRL-SV40 were determined in triplicate and standardized for transfection efficiency. Fold increase was measured by defining the activity of the empty pGL-3 Basic vector as 1. Data shown are the means fold increase ± SD from three independent transfection experiments, each done in triplicate. ∗∗, P < 0.0001 compared with the pT–T construct.

Fig. 3.

Transient reporter gene expression assays with constructs containing full-length MMP-2 promoter. A, schematic of reporter gene constructs having a full-length MMP2 promoter with the only difference between the four constructs being a T or C at the −1306 and −735 polymorphic sites. B, luciferase expression of the four constructs in HEK293 cells cotransfected with pRL-SV40 to standardize transfection efficiency. Luciferase levels of pGL-3 Basic and pRL-SV40 were determined in triplicate and standardized for transfection efficiency. Fold increase was measured by defining the activity of the empty pGL-3 Basic vector as 1. Data shown are the means fold increase ± SD from three independent transfection experiments, each done in triplicate. ∗∗, P < 0.0001 compared with the pT–T construct.

Close modal
Fig. 4.

Representative gel picture showing reverse transcriptase-PCR analysis of MMP-2 mRNA in esophageal tissues from individuals with different MMP-2 promoter genotypes. The expected length of PCR products was 601 bp for β-actin and 493 bp for MMP-2. Lane M, DNA size marker; Lanes 1–11, individual samples. Individual’s genotype designation: Lanes 1–3, 7, 9, and 10, −735CC/−1306CC; Lanes 4 and 11, −735TT/−1306CT; and Lanes 5 and 8, −735CT/−1306TT.

Fig. 4.

Representative gel picture showing reverse transcriptase-PCR analysis of MMP-2 mRNA in esophageal tissues from individuals with different MMP-2 promoter genotypes. The expected length of PCR products was 601 bp for β-actin and 493 bp for MMP-2. Lane M, DNA size marker; Lanes 1–11, individual samples. Individual’s genotype designation: Lanes 1–3, 7, 9, and 10, −735CC/−1306CC; Lanes 4 and 11, −735TT/−1306CT; and Lanes 5 and 8, −735CT/−1306TT.

Close modal

Grant support: Beijing Municipal Commission of Science and Technology Grant H0209-20030130, State Key Basic Research Program Grant G1998051204, and National Natural Science Foundation Grant 39825122.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Note: C. Yu and Y. Zhou contributed equally to this work.

Requests for reprints: Dongxin Lin, Department of Etiology and Carcinogenesis, Cancer Institute, Chinese Academy of Medical Sciences, Beijing 100021, China. Phone: 86-10-677-22460; Fax: 86-10-677-13359; E-mail: dlin@public.bta.net.cn

1

Internet address: http://www.gene-regulation.com/pub/programs.html#alibaba2.

Table 1

Genotype and allele frequencies of MMP-2 among cases and controls and their contributions to the risk of esophageal squamous cell carcinoma

GenotypeCases (n = 527)Controls (n = 777)OR* (95% CI)
No.(%)No.(%)
MMP-2 −1306C/T      
TT (1.1) 18 (2.3) 1.00 (reference) 
CT 112 (21.3) 220 (28.3)  
CC 409 (77.6) 539 (69.4) 1.52 (1.17–1.96) 
C allele frequency 0.882  0.836   
MMP-2 −735C/T      
TT 25 (4.7) 39 (5.0) 1.00 (reference) 
CT 179 (34.0) 313 (40.3)  
CC 323 (61.3) 425 (54.7) 1.30 (1.04–1.63) 
C allele frequency 0.783  0.749   
GenotypeCases (n = 527)Controls (n = 777)OR* (95% CI)
No.(%)No.(%)
MMP-2 −1306C/T      
TT (1.1) 18 (2.3) 1.00 (reference) 
CT 112 (21.3) 220 (28.3)  
CC 409 (77.6) 539 (69.4) 1.52 (1.17–1.96) 
C allele frequency 0.882  0.836   
MMP-2 −735C/T      
TT 25 (4.7) 39 (5.0) 1.00 (reference) 
CT 179 (34.0) 313 (40.3)  
CC 323 (61.3) 425 (54.7) 1.30 (1.04–1.63) 
C allele frequency 0.783  0.749   
*

ORs and 95% CIs were calculated by unconditional logistic regression with the MMP-2–variant genotypes (CT or TT) as the reference group and adjusting for sex, age, smoking status, and other genotype where it was appropriate.

Table 2

Risk estimates for extended MMP-2 haplotypes in esophageal squamous cell carcinoma cases and controls

HaplotypeCases (n = 527)Controls (n = 777)aOR (95% CI)*P
No. of chromosomes (%)No. of chromosomes (%)
T-1306–T-735 6 (0.6) 52 (3.3) 1.00 (reference)  
T-1306–C-735 118 (11.2) 204 (13.1) 5.19 (2.14–12.59) < 0.0001 
C-1306–T-735 223 (21.2) 339 (21.8) 6.04 (2.52–14.48) < 0.0001 
C-1306–C-735 707 (67.1) 959 (61.7) 6.53 (2.78–15.33) < 0.0001 
HaplotypeCases (n = 527)Controls (n = 777)aOR (95% CI)*P
No. of chromosomes (%)No. of chromosomes (%)
T-1306–T-735 6 (0.6) 52 (3.3) 1.00 (reference)  
T-1306–C-735 118 (11.2) 204 (13.1) 5.19 (2.14–12.59) < 0.0001 
C-1306–T-735 223 (21.2) 339 (21.8) 6.04 (2.52–14.48) < 0.0001 
C-1306–C-735 707 (67.1) 959 (61.7) 6.53 (2.78–15.33) < 0.0001 
*

Adjusted for sex, age, and smoking status.

P <0.0001, test for trend.

1
Werb Z ECM and cell surface proteolysis: regulating cellular ecology.
Cell
1997
;
91
:
439
-42.
2
Chambers AF, Matrissian LM Changing views of the role of matrix metalloproteinase in metastasis.
J Natl Cancer Inst (Bethesda)
1997
;
89
:
1260
-70.
3
Curran S, Murray GI Matrix metalloproteinases in tumour invasion and metastasis.
J Pathol
1999
;
189
:
300
-8.
4
Egeblad M, Werb Z New functions for the matrix metalloproteinases in cancer progression.
Nat Rev Cancer
2002
;
2
:
161
-74.
5
Stamenkovic I Matrix metalloproteinases in tumor invasion and metastasis.
Semin Cancer Biol
2000
;
10
:
415
-33.
6
Nagase H, Woessner JF, Jr. Matrix metalloproteinase.
J Biol Chem
1999
;
274
:
21491
-4.
7
Sternlicht MD, Werb Z Matrix metalloproteinase Kreis T Vale R eds. .
Guidebook to the extracellular matrix, anchor and adhesion proteins
1999
p. 503
-603. Oxford University Press New York
8
Levi E, Fridman R, Miao H-Q, Ma Y-C, Yayon A, Voldavsky I Matrix metalloproteinase 2 releases active soluble ectodomain of fibroblast growth factor receptor 1.
Proc Natl Acad Sci USA
1996
;
93
:
7069
-74.
9
Samantaray S, Sharma R, Chattopadhyaya TK, Gupta SD, Ralhan R Increased expression of MMP-2 and MMP-9 in esophageal squamous cell carcinoma.
J Cancer Res Clin Oncol
2004
;
130
:
37
-44.
10
Yamashita K, Upadhay S, Mimori K, Inoue H, Mori M Clinical significance of secreted protein acidic and rich in cystein in esophageal carcinoma and its relation to carcinoma progression.
Cancer (Phila.)
2003
;
97
:
2412
-9.
11
Koyama H, Iwata H, Kuwabara Y, Iwase H, Kobayashi S, Fujii Y Gelatinolytic activity of matrix of matrix metalloproteinase-2 and -9 in oesophageal carcinoma: a study using in situ zymography.
Eur J Cancer
2000
;
36
:
2164
-70.
12
Bian J, Sun Y Transcriptional activation by p53 of the human type IV collagenase (gelatinase A or matrix metalloproteinase 2) promoter.
Mol Cell Biol
1977
;
17
:
6330
-8.
13
Qin H, Sun Y, Benveniste EN The transcription factors Sp1, Sp3, and AP-2 are required for constitutive matrix metalloproteinase-2 gene expression in astroglioma cells.
J Biol Chem
1999
;
274
:
29130
-7.
14
Price SJ, Greaves DR, Watkins H Identification of novel, functional genetic variants in the human matrix metalloproteinase-2 gene: role of Sp1 in allele-specific transcriptional regulation.
J Biol Chem
2000
;
276
:
7549
-58.
15
Vasku V, Vasku A, Tschoplova S, Izakovicova Holla L, Semradova V, Vacha J Genotype association of C(-735)T polymorphism in matrix metalloproteinase 2 gene with G(8002)A endothelin 1 gene with plaque psoriasis.
Dermatology
2002
;
204
:
262
-5.
16
Yu C, Pan K, Xing D, et al Correlation between a single nucleotide polymorphism in the matrix metalloproteinase-2 promoter and risk of lung cancer.
Cancer Res
2002
;
62
:
6430
-3.
17
Miao X, Yu C, Tan W, et al A functional polymorphism in the matrix metalloproteinase-2 gene promoter (-1306C/T) is associated with risk of development but not metastasis of gastric cardia adenocarcinoma.
Cancer Res
2002
;
63
:
3987
-90.
18
Zhou Y, Yu C, Miao X, et al Substantial reduction in risk of breast cancer associated with genetic polymorphisms in the promoters of the matrix metalloproteinase-2 and tissue inhibitor of metalloproteinase-2 genes.
Carcinogenesis (Lond)
2003
;
25
:
399
-404.
19
Song C, Xing D, Tan W, Wei Q, Lin D Methylenetetrahydrofolate reductase polymorphisms increase risk of esophageal squamous cell carcinoma in a Chinese population.
Cancer Res
2001
;
61
:
3272
-5.
20
Sobin LH, Wittekind C .
UICC TNM classification of malignant tumors
5th ed. 
1997
Wiley-Leiss New York
21
Stephens M, Smith NJ, Donnelly P A new statistical method for haplotype reconstruction from population data.
Am J Hum Genet
2001
;
68
:
978
-89.
22
Ott J Predicting the range of linkage disequilibrium.
Proc Natl Acad Sci USA
2000
;
97
:
2
-3.
23
Drysdale C, McGraw D, Stack C, et al Complex promoter and coding region beta 2-adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness.
Proc Natl Acad Sci USA
2000
;
97
:
10483
-8.
24
Liu JF, Wang QZ, Hou J Surgical treatment for cancer of the oesophagus and gastric cardia in Hebei, China.
Br J Surg
2004
;
91
:
90
-8.
25
Xing D, Tan W, Lin D Genetic polymorphisms and susceptibility to esophageal cancer among Chinese population (Review).
Oncol Rep
2003
;
10
:
1615
-23.
26
Kadonaga JT, Carner KR, Masiarz FR, Tjian R Isolation of cDNA encoding transcription factor Sp1 and functional analysis of the DNA binding domain.
Cell
1987
;
51
:
1079
-90.
27
Turner J, Crossley M Mammalian Kruppel-like transcription factors: more than just a pretty finger.
Trends Biochem Sci
1999
;
24
:
236
-40.
28
Margana RK, Boggaram V Functional analysis of surfactant protein B (SP-B) promoter Sp1, Sp3, TTF-1, and HNF-3alpha transcription factors are necessary for lung cell-specific activation of SP-B gene transcription.
J Biol Chem
1997
;
272
:
3083
-90.
29
Itoh T, Tanioka M, Yoshida H, Yoshioka T, Nishimoto H, Itohara S Reduced angiogenesis and tumor progression in gelatinase A-deficient mice.
Cancer Res
1998
;
58
:
1048
-51.
30
Bergers G, Brekken R, McMahon G, et al Matrix metalloproteinase-9 triggers the angiogenic switch during carcinogenesis.
Nat Cell Biol
2000
;
2
:
737
-44.
31
Ha HY, Moon HB, Nam MS, et al Overexpression of membrane-type matrix metalloproteinase-1 gene induces mammary gland abnormalities and adenocarcinoma in transgenic mice.
Cancer Res
2001
;
61
:
984
-90.
32
Zhu Y, Spitz MR, Lei L, Mills GB, Wu X A single nucleotide polymorphism in the matrix metalloproteinase-1 promoter enhances lung cancer susceptibility.
Cancer Res
2001
;
61
:
7825
-9.
33
Ghilardi G, Biondi ML, Mangoni J, et al Matrix metalloproteinase-1 promoter polymorphism 1G/2G is correlated with colorectal cancer invasiveness.
Clin Cancer Res
2001
;
7
:
2344
-6.
34
Hinoda Y, Okayama N, Takano N, et al Association of functional polymorphisms of matrix metalloproteinase (MMP)-1 and MMP-3 gene with colorectal cancer.
Int J Cancer
2002
;
102
:
526
-9.
35
Nishioka Y, Kobayashi K, Sagae S, et al A single nucleotide polymorphism in the matrix metalloproteinase-1 promoter in endometrial carcinomas.
Jpn J Cancer Res
2000
;
91
:
612
-5.
36
Kanamori Y, Matsushima M, Minaguchi T, et al Correlation between expression of the matrix metalloproteinase-1 gene in ovarian cancer and insertion/deletion polymorphism in its promoter region.
Cancer Res
1999
;
59
:
4225
-7.
37
Ghilardi G, Biondi ML, Caputo M, et al A single nucleotide polymorphism in the matrix metalloproteinase-3 promoter enhances breast cancer susceptibility.
Clin Cancer Res
2002
;
8
:
3820
-3.
38
Shima I, Sasaguri Y, Kusukawa J, et al Production of matrix metalloproteinase-2 and metalloproteinase-3 related to malignant behavior of esophageal carcinoma: a clinicopathological study.
Cancer (Phila.)
1992
;
70
:
2747
-53.
39
Ye S, Dhillon S, Turner SJ, et al Invasiveness of cutaneous malignant melanoma is influenced by matrix metalloproteinase 1 gene polymorphism.
Cancer Res
2001
;
61
:
1296
-8.