Introduction

Matrix metalloproteinases (MMPs) are a superfamily of proteolytic enzymes capable of degrading extracellular matrix and basement membrane (Stamenkovic et al. 2000; Curran and Murray 1999). Overexpression of MMPs in multiple carcinomas was found to be important in relation to tumor invasion and metastasis. Recent studies have also indicated that MMPs are involved in early tumorigenesis. MMP-2 (gelatinase A), after being processed into its active form (IV collagenase) (Woessner 1991), primarily hydrolyzes type IV collagen, the major structural component of basement membrane (Stamenkovic 2000). In addition, active MMP-2 can degrade insulin-like growth factor binding proteins and release insulin-like growth factors, which modulate proliferation and apoptosis of cells (Egeblad and Werb 2002). Therefore, MMP-2 activities may be highly involved in oncogenesis and tumor progression.

The MMP-2 promoter contains sequences for the binding of AP-2, p53, Sp1, and Sp3 (Qin et al. 1999). Price et al. (2001) identified a −1306C→T polymorphism in the MMP-2 promoter. This base transition is located in the CCACC box of the Sp1 binding site and eliminates promoter activity. It is likely that the −1306CC genotype may be associated with a high transcription level and enzyme activity of MMP-2, and eventually could affect individual susceptibility to neoplasms. Studies continue to demonstrate that the −1306CC genotype can double or quadruple risk factors for lung, stomach, colorectal, oral cavity, and breast cancers (Yu et al. 2002; Miao et al. 2003; Xu et al. 2006; Lin et al. 2004; Zhou et al. 2004).

In Mexico, cervical uterine cancer is the second most frequent malignant neoplasia in women, surpassed only by breast cancer. The incidence rate of cervical intraepithelial neoplasia and invasive cervical uterine cancer was 163.7 and 9.8 cases per 100,000 women over 14 years of age in 2006, showing the great impact cervical neoplasia has on public health in Mexico. About 80–90% of cervical uterine cancers are squamous cell carcinomas. The only factor directly linked to the development of cervical uterine cancer is chronic infection from high-risk human papilloma viruses (HPVs). HPV DNA has been shown to be found in nearly 100% of affected patients (Franceschi 2005; Prescrire Int. 2007) Nevertheless, it is still not precisely known which genetic factors cause an HPV-infected woman to develop a neoplasm. The MMP-2 gene could be related to oncogenesis and/or dissemination, since numerous studies have shown that MMP-2 expression is gradually increased according to the degree of cervical intraepithelial neoplasia and cervical uterine cancer (Ahmed et al. 2004; Gaiotto et al. 2004; Nasr et al. 2005). Nevertheless, no MMP-2 genetic studies have been conducted in women with cervical uterine cancer, a fact that underscores the need for such studies.

The present case-control study aims to investigate the contribution of MMP-2 −1306C→T gene polymorphism toward the risk of cervical intraepithelial neoplasia and cervical uterine cancer.

Materials and Methods

Study Population

In the case-control association study, 54 women patients presenting with invasive squamous cell cervical cancer (SCCA), 100 with cervical intraepithelial neoplasia, and 126 nonhospitalized women with normal cervical cytology agreed to participate in the study. Participants were recruited from a single public hospital center, where patients received antineoplasic treatment or follow-up and the control group received routine cervical cytology from April 2005 to July 2006. Patients and control were unrelated Mexican mestizo subjects from the state of Colima, Mexico. All the SCCA cases were classified histopathologically (squamous cell carcinoma) according to the WHO classification of cervical tumors (Tavassoeli and Devilee 2003), and precancerous lesions of the cervix were graded using the CIN (cervical intraepithelial neoplasia) classification (54, 18, and 28 patients with CIN I, II, and III, respectively). Age, sex, education, menarche, first sexual intercourse, number of gestations and children, menopausal status, family history, and cigarette smoking history were obtained from a questionnaire administered by an interviewer. Informed consent was obtained from all participants. Women were identified as smokers if they had a lifetime history of smoking more than 100 cigarettes. Desquamation cells from the buccal inside cheek were taken to extract DNA. The institutional review board of the University of Colima School of Medicine and the Colima State Health Department approved this study.

Polymorphism Analysis

Desquamated buccal cells were placed directly in 400 μl of lysis buffer (Tris–HCl at 100 mM pH 8.8, EDTA at 5 mM, SDS at 0.5%, NaCl at 100 mM), to which 10 μl SDS at 10% and 45 μl K proteinase at 10 mg/ml were then added, and the mixture was incubated for 3 h at 55°C. Phenol–chloroform extraction and DNA precipitation with absolute ethanol were immediately carried out. The DNA was resuspended in 100 μl water and stored at −20°C until analyzed. Polymerase chain reactions (PCR) for the MMP-2 gene promoter polymorphic region were carried out in a total volume of 25 μl, as previously described (Delgado-Enciso et al. 2008). The solution contained 50 ng genomic DNA, 0.1 μM each primer, 1 × Taq polymerase buffer (1.4 mM MgCl2), 0.2 mM dNTP, and 0.25 U AmpliTaq DNA polymerase (Perkin Elmer, Foster City, CA). The previously reported primers (Delgado-Enciso et al. 2008) were designed based on the sequence of the Homo sapiens MMP2 promoter (Genbank, HSA298926), forward 5′-ccctgtgccccacctttttcagat-3′ and backward 5′-agctgagacctgaagagctaaacagct-3′. The backward primer incorporates the PvuII cut sequence when the −1306C allele is amplified. PCR was performed in a programmable thermal cycler (Mastercycler Personal, Eppendorf AG, Germany). The cycling conditions for the MMP-2 gene PvuII polymorphism were set as follows: one cycle at 94°C for 3 min; 35 cycles at 94°C for 30 s, 63.5°C for 40 s, and 72°C for 40 s; and one final extension cycle at 72°C for 5 min. The 150 bp PCR product was mixed with 2 U PvuII (Invitrogen, San Diego, CA) and the reaction buffer, according to the manufacturer’s instructions. Two fragments measuring 125 bp and 25 bp were present if the product was digestible (−1306C allele). The digestion product was loaded into a 3% agarose gel containing ethidium bromide for electrophoresis. The polymorphism was categorized as a CC homozygote (cuttable), TT homozygote (uncuttable), or TC heterozygote. Genotyping of 5% of random individuals was confirmed by sequencing.

Statistical Analysis

Hardy–Weinberg equilibrium was evaluated by comparing observed and expected genotypes, using the χ2 analysis. The Fisher exact test was used to examine differences between controls and cases. The associations between environmental factors/MMP-2 genotype and risk of disease genesis were estimated by odds ratio (OR) and associated 95% confidence interval (CI) using the Epi Info, Version 6, computer-assisted program (Centers for Disease Control and Prevention, Atlanta, GA). Statistical significance was reached when P < 0.05.

Results

The mean age was 41.8 (SD 9.1) and 43.5 (SD 14.5) years in the controls and cases (SCCA/CIN), respectively. There was no significant difference in this variable. Other characteristics of the population studied are summarized in Table 1. Early age of first coitus (≤19 years) and multigestation (≥4) were associated with SCCA and CIN. The crude OR was 4.1 (95% CI: 1.87–9.47, P = 0.0001) for SCCA and 3.2 (95% CI: 1.74–5.91, P = 0.00005) for intraepithelial neoplasia in subjects with early age of first coitus, and 4.8 (95% CI 2.25–10.39, P = 0.000006) for SCCA and 2.01 (95% CI 1.16–3.64, P = 0.007) for intraepithelial neoplasia in women with multigestation, compared with women experiencing first coitus at ≥20 years of age or with ≤3 gestations.

Table 1 Distribution of selected characteristics by control-case status
Full size table

Figure 1 shows some genotyping results after gel electrophoresis and ethidium bromide staining.

Fig. 1
figure 1

Genotyping of cervical cancer patients. Left panel: the three possible genotypic patterns (TT, CT, CC). Right panel: representative samples. Electrophoresis in 3% agarose gel. bp, base pairs

Full size image

The allele frequencies for MMP-2 −1306C and −1306T were 79% and 21% in the control group, 90% and 10% in cervical uterine cancer patients (P = 0.01), and 83% and 17% in cervical intraepithelial neoplasia patients (P = 0.28). The distribution of MMP-2 genotypes (Table 2) in the control group was in Hardy–Weinberg equilibrium. As the TT genotype was rare (none of the SCCA cases and 2.4% of the control group), it was combined with the CT genotype as the variant group for subsequent estimation of risk analysis. The crude OR in subjects carrying the CC genotype was 2.57 (95% CI: 1.15–5.86, P = 0.01) for SCCA and 1.61 (95% CI 0.89–2.93, P = 0.09) for all cases of cervical intraepithelial neoplasia. After stratifying the sample according to CIN grade, the CC genotype showed no association with any of the three grades (data not shown).

Table 2 Distribution of MMP-2 −1306 genotypes by control-case status
Full size table

The effects of the MMP-2 polymorphism were additionally examined with stratification by age of first coitus. It was found that increased risk of SCCA associated with the CC genotype was more pronounced in women who experienced first coitus at an early age (OR 3.9), compared with women with CT/TT genotypes who had also experienced sexual relations at an early age (Table 3). In concordance with the previous results, the CC genotype was shown to be associated with development of intraepithelial neoplasia in women who had first coitus at 19 years of age or younger (OR 2.67) (Table 3).

Table 3 Risk of cervical cancer or neoplasia related to −1306 MMP-2 genotypes by age of first coitus
Full size table

Discussion

A high degree of MMP-2 expression has been demonstrated in cervical uterine cancer and premalignant lesions; however, there are no previous reports to determine the role MMP-2 genetic polymorphisms play in this neoplasia. The present study found that the −1306CC genotype increases the probability for invasive SCCA. The OR is similar to those previously reported for the majority of neoplasms in which this polymorphism has been studied (OR 2, approximately) (Yu et al. 2002; Miao et al. 2003; Xu et al. 2006; Zhou et al. 2004). MMP-2 plays an important role in multiple-stage carcinogenesis. A number of studies have shown that MMP-2 is overexpressed in various cancer tissues, and its involvement in tumor initiation, invasion, angiogenesis, and metastasis was critical (Egeblad and Werb 2002). This gives much relevance to the study of genetic markers that could affect MMP-2 expression, since they accompany the individual from birth and may be affecting his or her susceptibility to neoplasia as well as influencing cancer phenotypes. Within the MMP-2 genetic polymorphisms, −1306C→T is of great interest because it alters the expression level in vitro (Price et al. 2001) and has been associated with lung, stomach, colorectal, oral cavity, and breast cancer development (Yu et al. 2002; Miao et al. 2003; Lin et al. 2004; Zhou et al. 2004; Xu et al. 2006; Delgado-Enciso et al. 2008). Reports on these findings, however, are scarce and have been carried out mainly on Chinese populations. Studies in other populations have produced conflicting results, such as the Elander et al. (2006) report in which no association was found between the MMP2 promoter polymorphisms and colorectal cancer in a Swedish population.

In this study, the risk related to CC genotype for the development of cervical premalignant lesions or SCCA was observed to be increased in women who had first coitus at an early age. This is in concordance with a possible interaction between these factors on the pathogenesis of cervical neoplasia. It has been demonstrated that women who begin their sexual activity at an early age are more susceptible to cervical HPV infection and thus have a greater HPV prevalence and risk of cervical neoplasia (Kahn et al. 2002; Collins et al. 2005). It is probable that the risk increase related to the CC genotype in women whose first coitus was at an early age reflects a direct interaction between HPV infection and CC genotype. The present study, however, did not determine the HPV infection, making further studies necessary to assess the hypothesis.

On the other hand, it was seen that the −1306 CC genotype alone did not significantly modify the risk of cervical intraepithelial neoplasia, suggesting that this polymorphism is greatly relevant only in the invasive process. The CC genotype, however, is associated with the development of premalignant lesions only in women who had first coitus at an early age.

The results clearly demonstrate the association between SCCA and the CC genotype. Nevertheless, because of the small sample size, the pronounced risk of the combination of CC genotype with early first coitus is not a definitive conclusion, and it warrants further analysis with larger samples. Further studies also need to be carried out on the interaction between the MMP-2 −1306 polymorphism and other factors such as early age diagnosis, tabachism, or multigestation.

In conclusion, the results indicate that from the genetic level, MMP-2 may be implicated in the genesis of cervical intraepithelial neoplasia and squamous cell cervical cancer, with the −1306 polymorphism being of great importance. The CC genotype increases the risk of developing cervical neoplasia, and this risk is probably accentuated in women affected by other environmental factors. This is the first report to associate MMP-2 genetically with cervical intraepithelial neoplasia or cervical uterine cancer. Further studies on this topic need to be carried out in other populations.