Matrix Metalloproteinase-1 Promoter Polymorphism 1G/2G Is Correlated with Colorectal Cancer Invasiveness

CRC² is the second most common cancer in males and the third most common in females in Italy.

An orderly sequence of nonrandom events leads to the development of CRC through the transformation of epithelial tissue into premalignant adenomatous polyp and eventually into invasive carcinoma (1).

The outcome depends on the degree of bowel wall invasion and on metastatic spread to regional lymph nodes or distant organs.

Disease staging takes into account the histopathological evaluation of the surgical specimen that includes locoregional and interaortocaval nodes.

Basal membrane and extracellular matrix represent two physical barriers to malignant invasion: their degradation by MMPs plays a key role in tumor progression and metastatic spread (2).

Among proteolytic enzymes, the MMP family has a substrate-specific degradation activity; specifically, MMP-3 (stromelysin-1) acts on type IV collagen that forms the basal membrane, and MMP-1 (interstitial collagenase) acts on fibrillar collagen and gelatin; MMP-3 has an additional activity activating pro-MMP-1 to MMP-1.

A prognostic value of MMP expression in tumor tissue has been shown (3); the relevance of MMP-1 in both CRC and esophageal cancer has been reported by Murray et al.(4, 5), and the relevance of MMP-3 in both CRC and breast cancer has also been observed (6, 7). The metastatic process is associated with local overexpression of MMPs (2).

Heppner et al.(8) have shown in breast cancer that the expression of many metalloproteinases in tumor tissue may represent a tumor-induced host response.

Rutter et al.(9) have reported recently that the insertion of a G nucleotide at −1607 bp in the nucleotide sequence of the MMP-1 gene promoter generates a new 5′-GGA-3′ sequence that corresponds to a core recognition sequence of the binding site for members of the Ets family of transcription factors (10, 11). The 2G homozygous polymorphism of the promoters results in an increased transcription activity in melanoma cell lines and in normal fibroblasts compared with 1G homozygotes and controls (9).

Another two reports have demonstrated the correlation between the 2G allele and ovarian carcinoma and endometrial carcinoma, respectively, both in their homozygous and heterozygous forms (12, 13).

Similarly, the insertion/deletion mechanism of an A nucleotide at −1171 bp in the MMP-3 gene promoter sequence results in a polymorphism (5A/6A) in which the transcriptional activity of the 5A homozygous is approximately double that of the 6A homozygous (14). In addition to its proteolytic activity, MMP-3 releases several cell surface molecules such as E-cadherin, a known contributor to cancer development (15).

The aim of this study was to investigate possible correlations between MMP-1 and MMP-3 promoter polymorphisms and CRC clinical phenotypes and, specifically, if genetic analysis is capable of identifying a subgroup of CRC patients with a disease which is more aggressive or prone to metastasize.

MMP-1 and MMP-3 gene promoter sequences were obtained form peripheral blood samples of 60 patients with CRC of different stages who underwent surgery and of 164 sex- and age-matched healthy subjects (control group). CRC patients were 40 males and 20 females; the median age was 68 years (range: 32–88).

CRC patients were grouped according to Duke’s classification modified by Astler and Coller, on the basis of the postoperative histopathological evaluation. The group assignment was than reevaluated at the end of the follow-up period that ranged from 6 to 30 months (median: 21 months); the 60 patients with CRC were assigned to two subgroups according to the presence (M+) or absence of detectable metastasis (M−) at the time of diagnosis and at the end of the follow-up.

The INSTA GENE (Bio-Rad) commercial kit was used for DNA extraction from whole blood collected using K₃EDTA as anticoagulant.

The PCR reaction for MMP-1 and MMP-3 was carried out in a total volume of 25 μl with 5 ml of extracted genomic DNA; 100 μm dATP, dGTP, dTTP, and dCTP; 1.5 mm MgCl₂; and 1 unit of Taq polymerase; the two primers, forward and reverse, were each at a concentration of 80 nm. The primers were designed with the Primer Express software. The MMP-1 primer sequence is: forward primer 5′-CCCTCTTGAACTCACATGTTATG-3′; reverse primer 5′-ACTTTCCTCCCCTTATGGATTCC-3′. For MMP-3: forward primer 5′-TCCTCATATCAATGTGGCCAAA-3′; reverse primer 5′-CGGCACCTGGCCTAAAGAC-3′.

The PCR reaction starts with 5-min incubation at 94°C to activate the enzyme, followed by 35 cycles of 20 s at 94°C, 20 s at 55°C, and 30 s at 72°C.

The amplification was verified on an agarose gel (2%) followed directly by sequencing with an automatic sequencer in fluorescent DNA capillary electrophoresis (ABI Prism 310; Applied Biosystems).

Statistical analysis was performed by χ² test with the Sigmastat for Windows Version 2.03 software (SPSS, Inc.). Yates correction for continuity was used, and OR was calculated when required.

The polymorphism distribution in the controls was as expected, according to the Hardy-Weinberger principle.

In CRC patients, the MMP-1 allelic variation was significantly different compared with controls (P = 0.018; OR, 2.21; 95% CI, 1.17–4.16), with the distribution curve shifted to the right (greater frequency of 2G homozygosity).

In CRC patients, no difference in MMP-3 allelic variant distribution compared with controls was noted.

The relevant figures are summarized in Table 1.

Subgroups have been compared with controls and among allelic variants.

MMP-3 promoter allelic variant distribution has shown no statistically significant differences between subgroup M+ (presence of metastases) versus controls, nor between subgroups M+ versus M− (absence of metastases) or M− versus controls.

As to MMP-1 promoter polymorphism, a strong correlation between 2G homozygosity and the M+ subgroup (presence of metastases) has been observed, both at the time of diagnosis (OR, 4.73; P = 0.0082) and at the end of the follow-up period considered (OR, 3.90; P = 0.0131; Table 2).

The frequency of the 2G allele showed no statistical difference between controls and metastasis-free (M−) patients (P = 0.52).

During follow-up, six deaths were recorded which were directly related to malignancy; 3 patients were homozygous for 2G, and 3 were heterozygous (1G/2G) for MMP-1 promoter.

A correlation between the transcription-enhancing insertion of a single G nucleotide in the MMP-1 gene promoter region and MMP-1 overexpression has been shown previously in ovarian cancer tumor tissue (12). Similarly, an association between local MMP-1 overexpression and unfavorable prognosis has been reported for both CRC and esophageal cancer (4, 5).

The elevated frequency of 2G homozygosity in melanoma and breast cancer cell lines has been considered responsible of a more aggressive behavior by virtue of higher MMP-1 levels present in tumor tissues (9, 12).

The attempt to correlate MMP-1 gene promoter insertion/deletion polymorphisms with CRC clinical stage and prognosis shifts the focus from tumor-related factors to genetically predetermined host response mechanisms that relate to cancer invasion susceptibility.

Our results point out a significant difference in MMP-1 allelic variants distribution with a 2.21 exceeding risk of CRC for 2G homozygous patients.

The subsequent analysis of initial staging and clinical outcome seems to suggest differences in tumor biological behavior which are associated with host polymorphism. The frequency of metastases at the time of diagnosis is 4.7-fold higher in 2G homozygous patients. At the end of the follow-up, the percentage of metastases in 2G homozygous patients was twice (2G/2G: 65%; 1G/1G + 1G/2G: 32%).

The increased aggressiveness of the tumor associated with the host 2G allele is paralleled by the observation that the frequency of the 2G allele is substantially higher in the M+ subgroup relative to both the control group and the M− subgroup. The lack of a statistically significant difference in the frequency of the 2G allele between controls and metastasis-free CRC patients seems to support the relevance of host-dependent factors in the limitation of CRC invasiveness.

MMP-3 is known to lyse basal membrane collagen and to induce the synthesis of other MMPs, among which MMP-1. It may play a role both in local invasiveness and metastatic spread; the latter in fact involves the ability of neoplastic cells to cross the basal membrane of both the epithelium and the vascular endothelium. MMP-3 overexpression by some tumor types (16) is consistent with this hypothesis. The insertion of an A in the MMP-3 gene promoter sequence halves its transcription activity. It is conceivable the higher transcription activity associated with the 5A allele may enhance tumor invasiveness.

Our study demonstrates the lack of correlation between CRC, its clinical stage, the presence of metastases, and the 5A/6A polymorphism of the MMP-3 promoter, questioning the role of host 5A allele as a relevant factor inducing the well-known local MMP-3 overexpression. This feature seems to differentiate CRC patients from other patients with breast, lung, and ovarian cancer (15, 17).

In conclusion, our study seems to suggest that the presence of the 2G allele in the MMP-1 gene promoter sequence may be a facilitating factor for cancer growth and metastasis in CRC patients. This activity might be mediated by host MMP-1 overexpression induced by the tumor, which seems to decrease the ability of CRC-affected individuals to limit, both locally and systemically, tumor cell invasiveness.

We may hypothesize that correcting for other potentially relevant variables (social status, quality of health care services, and environmental and nutritional variables), the presence of the 2G allele is associated with increased tumor growth and invasiveness, and thus, the higher frequency of the 2G allele in advanced-stage CRC patients compared with more localized cases may not be explained by a delay in diagnosis but, rather, may represent a true additional risk factor.