MMP-9/RECK Imbalance: A Mechanism Associated with High-Grade Cervical Lesions and Genital Infection by Human Papillomavirus and Chlamydia trachomatis

Cervical cancer is etiologically associated with infection with high-oncogenic risk human papillomavirus (HPV; 1). This tumor has the capacity to metastasize to distant tissues and organs, a fact strongly associated with failure of cervical cancer treatment. The evolution of metastasis is a multistep process and its initial stages can be identified in the preneoplasic phase (2). Transformation of a noninvasive cervical neoplasia into an invasive cervical carcinoma requires the migration of epithelial cells through the subjacent extracellular matrix (ECM), a process in which the matrix metalloproteinases (MMP) play a critical role by degrading ECM components (3, 4).

MMPs are zinc-dependent endopeptidases involved in various physiologic and pathologic processes, including tissue remodeling, development, and regulation of inflammation (5, 6). During carcinogenesis, these enzymes promote changes in the stroma and microenvironment contributing to tumor progression (5–7). Furthermore, MMPs are also dependent on local equilibrium with its physiologic inhibitors, such as the RECK protein (Reversion-inducing Cysteine-rich protein with Kazal motifs), which regulates MMP-2 and MMP-9 activities (4, 7).

Overexpression of MMP-2 and MMP-9 has been shown to occur in invasive cervical carcinoma and high-grade cervical lesions (3, 8, 9). We have previously shown that HPV16-positive cell lines express higher levels of MMP-2, MT1-MMP, and TIMP-2 than HPV-negative cell lines. MMP-9 was expressed at very low levels in both HPV-negative and HPV-positive cell lines (10). To analyze MMPs status in a system that reproduces in vivo conditions, we have previously developed organotypic epithelial cultures with keratinocytes expressing HPV 16 oncoproteins. We have demonstrated that HPV 16 E7 induces pro–MMP-9 activity, and that E6 and E7 coexpression downregulates RECK protein levels (11). This latter study also demonstrated that, even in small number of clinical samples, there is an inverse correlation between RECK expression and degree of cervical lesion. Furthermore, previous work from our group demonstrated the association between high-grade cervical lesions and high expression of MMP-9 in conventional Pap smears, especially when the patient had concomitant genital infection (12).

Clinical studies have shown an association between high-grade cervical intraepithelial neoplasia (CIN) and Chlamydia trachomatis infection (13–16). Considering that concomitant genital infections have been assumed to be risk factors for HPV persistence (14), we hypothesized that infection-induced cervical inflammation were capable of influencing the carcinogenic process driven by HPV. One can consider the role of MMPs in inflammatory process, because the involvement of these enzymes has already been demonstrated in organotypic cultures of fallopian tube infected in vitro by C. trachomatis (17) and in the pathology of ocular trachoma (18). However, within the context of cervical carcinogenesis, this possible relationship has not yet been elucidated. It is biologically plausible that chronic infection by C. trachomatis causes an unresolved cervical inflammation that may lead to tissue damage, involving an interplay between HPV, MMPs, and their inhibitors. Therefore, in the present study, we evaluated the role of MMP-9 and RECK expression in cervical lesions and genital infections.

This cross-sectional study included 332 women attending the CAISM Women's Hospital (Centro de Atenção Integral à Saúde da Mulher), Family Planning of Human Reproduction Unit, in the School of Medical Sciences, State University of Campinas (UNICAMP), from August 2013 to August 2014. The case group was composed of 136 women that underwent large loop excision of the transformation zone (LLETZ) due to high-grade cytologic abnormalities in referral Pap smears. This group had a confirmed histologic diagnosis of CIN grade 2 (CIN2; n = 31), CIN grade 3 (CIN3; n = 80), or squamous cell carcinoma (SCC; n = 25). The control group was composed of 196 women who presented at least two previous normal Pap smears (n = 162), cervicitis (n = 9), or CIN grade 1 (CIN1; n = 25) diagnosed by LLETZ. No cases of adenocarcinoma were included. CIN1 cases were included in the control group because low-grade cervical lesions have more similar characteristics to those of a HPV productive benign infection than of a transforming infection (19). Morphologic diagnosis of biopsies obtained by LLETZ were assessed according to the International Histological Classification of Tumors (World Health Organization; ref. 20).

This study was approved by the Ethics in Research Committee of the School of Pharmaceutical Sciences of the University of São Paulo (USP; ref. 411.692/2013). After signing the informed consent form, women were submitted to a specular examination for the collection of cervical samples, which were preserved partly in Surepath vials (BD Surepath, Pap test) for liquid-based cytology (LBC), RNA and DNA extraction, and partly in sterile saline medium for cervical infections diagnosis. We also collected a cervical smear to perform a GRAM-stained bacterioscopy.

An 8-mL fraction of the cervical sample collected in the Surepath vial was centrifuged twice with the addition of 4 mL of gradient wash buffer. The resulting cellular pellet was suspended in 2 mL of Surepath medium. From this cellular suspension, 200 μL were used to prepare the LBC slides, 200 μL were used for DNA extraction (for HPV genotyping) and 200 μL for RNA extraction (for analysis of gene expression).

LBC slides were used to determine the expression of MMP-9 and RECK by immunocytochemistry analysis. To prepare the LBC slides, we used the automated PrepStain system (TriPath Imaging, Inc.). As a positive control of MMP-9 labeling, we used the HT1080 cell line derived from human fibrossarcoma (ATCC), which was chosen because it is known to be abundant in MMP-9. As a positive control for RECK labeling, we used primary human fibroblasts extracted from human foreskin at the Clinical Cytopathology Laboratory of the School of Pharmaceutical Sciences, USP.

To evaluate the expression of MMP-9, the primary antibody anti–MMP-9 (MAB 13415; Chemicon-Millipore) was used at a 1:150 dilution in PBS solution. To evaluate RECK expression, we used the primary antibody anti-RECK (BD 611512; Biosciences) also at the 1:150 dilution in PBS. Incubation with primary antibodies was performed overnight in a wet chamber at 4°C. Incubation with secondary antibody and staining with DAB chromogen was performed using the Envision + Systems/HRPP Commercial Kit (K4007/DAKO Cytomation). After the immunocytochemical reaction, we proceeded to nuclei staining with hematoxylin. Immunocytochemical analysis was performed on 321 LBC slides for MMP-9 protein and on 327 LBC slides for RECK protein.

To analyze the immunocytochemistry slides, we adapted the criteria described by Li and colleagues (9). We performed a blind test in which the observer was not informed of the diagnosis of any cervical lesions. The slides were analyzed at ×400 magnification using a light microscope. Cells were counted and classified as “positive” or “negative” according to DAB chromogen staining: brown stained cells were considered as “positive” regardless of the intensity of the color (pale-yellow, brown, or yellow-brown), and blue stained cells (hematoxylin only) were considered “negative.” Counting was performed in two distant random fields, avoiding the specimen edges. In case of discrepant scores between these two fields, a third field was counted. Slides grading was performed as follows: score 1–0 to 25% of DAB-stained squamous cells; score 2–26 to 50% of DAB-stained squamous cells; score 3–51 to 75% of DAB-stained squamous cells; score 4—more than 75% of DAB-stained squamous cells. Both atypical and typical squamous cells were considered in the counting.

At first, the analysis was performed by two independent observers (P.C. Pennacchi and M.G. Discacciati) and, in case of disagreement, a third observer (M. Zonta) was consulted. Scores 1 and 2 were considered as negative or mild expression, and scores 3 and 4 were considered as moderate or accentuated expression.

Total RNA was extracted from cervical samples of 4 women from the control group (no neoplasia/CIN1) and from 11 women in the case group (CIN2/CIN3/SCC) using TRizol reagent (Invitrogen). RNA samples were quantified by UV light spectrophotometer at 260 nm (ND-1000–NanoDrop) and RNA quality was verified by the A₂₆₀/A₂₈₀ ratio. cDNA was synthesized from 1 μg RNA using the Reverse Transcription c-DNA High-capacity Kit (4388813, AB-Applied Biosystems). Real-time quantitative PCR (RT-qPCR) was carried out using TaqMan-specific primers for the genes of interest (MMP-9: Hs00234579-m1, RECK: Hs01019179-m1; Applied Biosystems). The assay was performed using ABI Prism 7500 Sequence Detection System equipment (Applied Biosystems), under the following conditions: Activation of the AmpEraseUNG enzyme at 50°C for 2 minutes, followed by initial denaturation at 95°C for 10 minutes, 40 cycles of 15 seconds at 95°C to end the denaturation process, and finally, annealing and extension of the primers and probe at 60°C for 1 minute. The level of gene expression was calculated in relation to the endogenous controls expression (β-actin: Hs01060665-g1 and GAPDH: Hs02758991-g1; Applied Biosystems) using the cycle threshold (C_t) method.

DNA was extracted from 320 cervicovaginal samples collected in Surepath vial. The cellular suspension was centrifuged and resuspended in 200 μL of digestion solution (50 mmol/L Tris-HCL, 1 mmol/L EDTA, and 0.5% Triton X-100) + 200 μL TEP buffer [200 μg/mL of proteinase K in Tris-EDTA (TE)]. After 24 hours of digestion at 55°C, the nucleic acids were purified with Phenol:Chloroform:Isoamyl Alcohol (25:24:1) followed by precipitation with 3 mol/L sodium acetate plus 2.5 volumes of absolute ethanol, and washing with 70% ethanol. The DNA pellet was resuspended in 50 μL of water.

For HPV genotyping, we used the Linear Array Genotyping HPV Kit (Roche Molecular Systems Inc.) according to the manufacturer's protocol. This test uses target-DNA amplification by PCR and nucleic acids hybridization, detecting 37 types of low- and high-risk HPVs: HPV 6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 45, 51, 52, 53, 54, 55, 56, 58, 59, 61, 62, 64, 66, 67, 68, 69, 70, 71, 72, 73, 81, 82, 83, 84, IS39, and CP6108. Twelve women were excluded from HPV genotyping because of inadequate internal controls (β-globin).

Multiplex PCR (M-PCR) was performed for the diagnosis of seven concomitant genital infections in cervical samples in accordance with a previously standardized and validated method (21). The diagnosed concomitant genital infections were: C. trachomatis, Herpes virus simplex 1 and 2 (HSV1 and HSV2), Neisseria gonorrhoeae, Mycoplasma hominis, Trichomonas vaginalis, and Treponema pallidum. This analysis was performed at the Clinical Cytology Laboratory of the State University of Maringá (Laboratório de Citologia Clínica da Universidade Estadual de Maringá, UEM, Paraná, Brazil), using endocervical samples collected in sterile saline solution. DNA extraction of these samples was performed using the PureLink RNA-DNA Viral Purification Mini Kit (12280-150; Invitrogen) following the manufacturer's instructions. Briefly, primers were characterized by compatible melting temperatures and yielded amplicons with sizes easily separable by PAGE. The protocol consisted of a reaction mixture of 25 μL containing 2.5 mmol/L of each dNTP, 0.6 mmol/L of MgCl₂, 25 mmol/L of the specific primer, 5 μL of extracted DNA (50 ng of total sample) and 1 U of Platinum Taq DNA polymerase (Invitrogen). The PCR conditions comprised of 35 amplification cycles of denaturation for 10 minutes at 94°C, annealing for 1 minute at 62°C, extension for 1 minute at 72°C, and final extension for 10 minutes at 72°C (Thermal cycler; Biosystem). The M-PCR products were analyzed in 8% polyacrylamide gels (Fig. 1). Positive controls for all genital infections were derived from positive clinical samples diagnosed by reference methods, including culture and/or single PCR (sPCR; ref. 22). Human β-globin–specific primers GH20/PC04 were used as internal controls for amplification and DNA integrity. The diagnosis of infections by multiplex PCR was not determined in 47 samples because of inadequate DNA internal control (β-globin).

The GRAM-stained slides were analyzed using the criteria described by Nugent and colleagues (23), which applies a scoring system for the presence of bacterial morphotypes of Gardnerella vaginalis, Lactobacillus sp, and Mobiluncus sp. A score of 0 to 3 consisted of normal microbiota, a score of 4 to 6 consisted of moderately disturbed microbiota, and a score equal or greater than 7 consisted of bacterial vaginosis (BV).

The association between cervical lesions, genital infections, MMP-9, and RECK expression was evaluated by OR with a 95% confidence interval (CI). The social and demographic characteristics were evaluated by the χ² test. Multiple logistic regression analysis was used to calculate the ORs adjusted for the control variables, such as age, age at first sexual intercourse, ethnicity, parity, smoking, and genital infections. The statistical analyzes described above were performed using the software SAS version 9.1.3 (SAS Institute). The RT-qPCR gene expression of MMP-9 and RECK data was evaluated by the Student t test, using GraphPad Prism version 5.0 (GraphPad Software). Significance level was defined at 5% for all statistical analysis.

In the present study, the mean age of women was 35.4 ± 10 years in the case group (CIN2/CIN3/SCC) and 35.8 ± 8 years in the control group (no neoplasia/CIN1), with no statistical difference between the two groups. The profile of social and demographic characteristics in the study groups is shown in Table 1. We found a statistically significant higher frequency of smoking women in the case group when compared with the women in the control group (37% vs. 21%, respectively; P = 0.001). We also found a higher frequency of women having had first sexual intercourse at age below 15 years in the study group (44%) when compared with the women in the control group (30%), and this difference was statistically significant (P = 0.008).

We were able to gather data on a large spectrum of diagnosed genital infections from 300 women, by GRAM and M-PCR techniques. The most prevalent genital infection was BV (23%), followed by C. trachomatis (11%), Trichomonas vaginalis (3%), and Candida spp (3%). Neisseria gonorrhoeae was detected in 4 women (1.3%), whereas HSV1, HSV2, and Treponema pallidum were identified in 2 women each (0.9%). One case of Treponema pallidum was associated with an infection by C. trachomatis.

The overall frequency of HPV was 52% (166/320). Of the 166 positive cases, 32% presented multiple HPV infections and 68% presented a single infection. We detected 32 different types of HPV. The most prevalent type detected was HPV16 (34%), followed by HPV31 (8%), HPV52 (7%), HPV33 (6.5%), HPV58 (6%), HPV35 (5.5%), and HPV62 (4%). HPV18 represented 2.6% of the cases. In the case group (CIN2/CIN3/SCC), the frequency of HPV was 89%, whereas in the control group (no neoplasia/CIN1), the frequency was 24%. Regarding the oncogenesis risk of HPV, 91% of the HPV-positive cases were high oncogenic risk types, 7% were low oncogenic risk types, and 2% were probably oncogenic.

With regard to the phylogenetic distribution of HPV, infection by alpha-9 species (which includes HPV types 16, 31, 33, 35, 52, and 58) occurred in 75% of the cases, and infection by alpha-7 species (HPVs 18, 39, 45, 56, 59, 66, and 68) occurred in 9% of the cases, whereas a combination of these two species was found in 5% of all cases. Other HPV species were inexpressively detected such as alpha-5, alpha-6, alpha-11, and alpha-15.

The relationship of the HPV status with the MMP-9 and RECK expression is shown in Table 2, whereas the relationship between HPV and other infectious agents is represented in Table 3. We demonstrated that the presence of HPV, especially HPV16 and other alpha-9 HPV types, was significantly associated with increased expression of MMP-9 and low expression of RECK (Table 2). Women infected by C. trachomatis were more likely to have a moderate/accentuated MMP-9 expression, which was about 2.6 times higher than the cases negative for C. trachomatis, with statistically significant results even after adjusting for the presence of HPV (OR, 2.6; 95% CI, 1.1–6.7). Furthermore, women with C. trachomatis had a higher prevalence of HPV multiple infection (OR, 2.7; 95% CI, 0.8–9.0). We also found a statistically significant association between C. trachomatis and high-grade cervical disease (CIN2/CIN3/SCC) when compared with the control group, that is, no neoplasia/CIN1 cases (OR, 3.7; 95% CI, 1.1–4.9). However, this association was not seen for BV or other genital infections in general (Table 3).

When we considered only women infected by high-risk HPV, we found that frequency of moderate/accentuated MMP-9 expression in women with concomitant infection by C. trachomatis was higher (17%) when compared with negative-mild MMP-9 expression (6%), although this result was not statistically significant (OR, 3.3; 95% CI, 0.7–15.8). Moreover, frequency of moderate/accentuated RECK expression in C. trachomatis–positive samples is lower (7%) when compared with negative-mild RECK expression (16%), but this difference was also not significant (OR, 0.4; 95% CI, 0.7–1.7; data not shown).

One important finding reported in Table 4 is that women from the case group (CIN2/CIN3/SCC) had a greater frequency of moderate/accentuated expression of MMP-9 than the women in the control group (no neoplasia/CIN1), and this association was statistically significant (OR, 4.2; 95% CI, 2.2–7.8). In addition, when evaluating the expression of MMP-9 for each type of neoplasic cervical lesion in comparison with women without cervical neoplasia, we observed an increase in risk of moderate/accentuated expression in higher grades of CIN. Table 4 also shows that there is a strong association between CIN3 and moderate/accentuated expression of MMP-9 (OR, 5.4; 95% CI, 2.4–12.4). When considering RECK expression this relation is inverted: Lower RECK expression is associated with higher grades of CIN. These results were also statistically significant after a multivariate analysis adjusted for the study variables, such as age, age at first sexual intercourse, ethnicity, parity, smoking, and genital infections.

Figure 2 represents the immunocytochemical staining for MMP-9 and RECK proteins, showing that there is a clear increase in the intensity and number of cells stained for MMP-9 in CIN3 lesions (Fig. 2B) when compared with normal cervical smears (Fig. 2A). Moreover, we observed a decrease in the expression of RECK in CIN3 lesions (Fig. 2E) when compared with normal cervical smears (Fig. 2D).

Figure 2C and F shows a graphic representation of the MMP-9 and RECK mRNA expression evaluated by RT-qPCR. We observed that, women in the case group (CIN2/CIN3/SCC) presented a higher MMP-9 mRNA expression than women in the control group (no neoplasia/CIN1), but this result was not statistically significant (P = 0.2). On the other hand, there was a statistically significant lower RECK mRNA expression in the case group when compared with the control group (P = 0.03). These results are consistent with the ones we found in the immunocytochemical analyzes.

In this study, we presented MMP-9 and RECK expression analysis in LBC cervical samples of women with CIN and cervical SCC, concomitant with other genital infections. We found that higher expression of MMP-9 and lower expression of RECK is strongly associated with high-grade cervical lesions and with the presence of alpha-9 HPVs, which includes HPV 16. In addition, we demonstrated, for the first time to the best of our knowledge, that C. trachomatis infection is associated with MMP-9/RECK imbalance. This may constitute a possible mechanism by which inflammation contributes to HPV-mediated carcinogenesis. Furthermore, our results are in accordance with data from the literature that have already described the relation between C. trachomatis infection and high-grade cervical lesions (13–16).

In a previous study from our group, da Silva Cardeal and colleagues (10) discussed that there is no correlation between RECK expression and low MMP-9 activity in HPV16-positive (SiHa and CAsKi) and HPV16-negative (C33A) human cervical carcinoma cell lines. In that model, the authors hypothesized that TIMP-2, but not RECK, is involved in MMP inhibition. However, in the present study, data from clinical samples evidenced an imbalance between MMP and RECK, which has already been shown in other human cancers (24–26) and in organotypic epithelial cultures with HPV-infected keratinocytes (11).

Our clinical data are in agreement with the study from Wang and colleagues (27), which demonstrated that higher expression of RECK and lower expression of MMP-9 and MMP-2 constitutes a positive prognostic survival factor in patients with invasive cervical carcinoma. Ghosh and colleagues (28) investigated the gelatinolytic activity of matrix MMP-2 and MMP-9 in preinvasive and invasive carcinoma of the uterine cervix and concluded that both MMP-2 and MMP-9 have a role in cancer progression and remodeling of the ectocervix. Furthermore, a study conducted in cervical biopsies showed a decrease in RECK expression in preneoplasic and neoplasic lesions (11).

Results from the present study show that women infected by alpha-9 HPVs, which includes HPV16, are more likely to have higher expression of MMP-9 and lower expression of RECK when compared with women infected by alpha-7 HPV or other HPV species. In this context, Cardeal and colleagues (11) indicated that HPV16 E7 expression is associated with increased pro–MMP-9 activity in the epithelial component of organotypic cultures, whereas E6 and E7 HPV16 oncoproteins coexpression downregulates RECK and TIMP-2 levels in organotypic cultures. Moreover, clinical studies have demonstrated that women infected by alpha-9 HPV have higher risk of CIN3 development or CIN2 clinical progression, when compared with women infected by alpha-7 HPV (29, 30).

Altogether, we can conclude that the analysis of MMP-9 expression in women infected by HPV could be an important tool to be used in clinical practice. Herein, we simultaneously investigated MMP-9, RECK, and HPV in LBC samples, which is a simple screening method that allows morphologic and molecular analyzes in the same sample. Not all women infected by HPV will present a clinically important lesion or will develop cervical carcinoma (19); therefore, it is necessary to search for new biomarkers that could positively affect the efficacy and screening outcome of the Pap test.

It is also important to consider other factors that might interfere in the regulation of such biomarkers; therefore, in this study, we analyzed the effect of concomitant genital infections on the protein expression of MMP-9 and RECK. The presence of genital infection is increasingly considered as a risk factor for the persistence of HPV infection and progression of the associated lesions, and the mechanisms of these associations have been the object of various studies (31–34). In conventional Pap smears, Matheus and colleagues (12) showed a high expression of MMP-9 in the presence of genital infections. Herein, we found a significant association between C. trachomatis and high-grade cervical lesions (CIN2/CIN3/SCC; OR, 3.7; 95% CI, 1.3–11.3), which has already been amply demonstrated in the literature (13–16). In addition, we found that women with C. trachomatis have a 2.7 higher chance of presenting multiple infections than a single infection by HPV, although this result was not statistically significant in this sample size (OR, 2.7; 95% CI, 0.8–9.0). Similarly, Paba and colleagues (32) demonstrated that the risk of a woman presenting an infection due to two concomitant types of HPV is 7.2 times higher in the presence of C. trachomatis. One large longitudinal study also showed that prior infection by C. trachomatis is the greatest risk factor for the persistence of HPV DNA (33). Furthermore, Jensel and colleagues (16) concluded that repeated infections by C. trachomatis increase the risk of CIN3 development among women with prevalent and persistent high-risk HPV infection.

The possible causal factors of the association between C. trachomatis, HPV, and cervical neoplasia are poorly understood. The inflammatory response, cellular hypertrophy, and squamous metaplasia caused by C. trachomatis infection are thought as possible mechanisms of the association between C. trachomatis and cervical neoplasia, because these events promote a cellular turnover that provides an advantage for the carcinogenic process induced by HPV (34). Thus, because MMPs are important mediators of tissue remodeling and inflammation, it is to be expected that they act in the inflammatory process caused by C. trachomatis. We found that women infected by C. trachomatis had higher MMP-9 expression in relation to women negative for C. trachomatis. An earlier study on ocular trachoma induced by C. trachomatis showed an increase in inflammatory factors and MMPs, including MMP-9 (18). More recently, another study demonstrated an increase in production of MMP-9 and MMP-2 in organotypic cultures of the fallopian tube infected in vitro by C. trachomatis (17). However, as far as we know, this is the first study demonstrating the effect of C. trachomatis infection on MMP-9/RECK expression in the context of cervical carcinogenesis.

In addition to genital infections, behavior and physiologic characteristics such as smoking, parity, oral contraceptives, alcohol, and age at first sexual intercourse have been described as factors that influence acquisition of HPV and development of neoplastic lesion (35–38). In this context, we found a higher frequency of smoking patients in the study group when compared with the control group. In fact, a large longitudinal study have reported smoking as a risk factor for CIN3 and invasive cervical carcinoma (38). The induction of epigenetic changes could explain this finding, as it has been shown that smoking increases the risk of methylation of the tumor-suppressor gene CDKN2A (p16), an epigenetic inactivation strongly associated with the pathogenesis of malignancies (39).

In this study, the frequency of women having had first sexual intercourse at age below 15 years was significantly higher in the study group (44%) when compared with the control group (30%). Our data are in agreement with the literature, once sexual precocity is known to increase susceptibility to the first HPV infection, as well as exposure to the carcinogenic effects of HPV, which is associated with higher risk of CIN3 (35, 36, 40). On the other hand, CIN1 is associated to recent acquisition of HPV infection, regardless of the age of first sexual intercourse (36).

In conclusion, our results show that higher expression of MMP-9 and lower expression of RECK is associated with high-grade CIN. Moreover, we show that MMP-9/RECK imbalance is significantly associated with genital infection by C. trachomatis and infection by alpha-9 HPV, which includes HPV16. Genital infection by C. trachomatis is also associated with high-grade CIN and SCC, suggesting that infection by C. trachomatis could act as a cofactor in HPV-mediated cervical carcinogenesis, and that the MMP-9/RECK imbalance during cervical inflammation might play an important role in this context.

No potential conflicts of interest were disclosed.

Conception and design: M.G. Discacciati, L.C. Zeferino, M.E.L. Consolaro, L.L. Villa, S.S. Maria-Engler

Development of methodology: M.G. Discacciati, F. Gimenes, M.C. Costa, M. Zonta, L. Termini, M.E.L. Consolaro

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M.G. Discacciati, P.C. Pennacchi, F. Faião-Flores, L.C. Zeferino, S.M. Derchain, J.C. Teixeira, M.C. Costa, S.S. Maria-Engler

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M.G. Discacciati, F. Gimenes, P.C. Pennacchi, F. Faião-Flores, J.C. Teixeira, M. Zonta, L. Termini, E. Boccardo, A. Longatto-Filho, L.L. Villa, S.S. Maria-Engler

Writing, review, and/or revision of the manuscript: M.G. Discacciati, F. Gimenes, S.M. Derchain, J.C. Teixeira, M.C. Costa, L. Termini, E. Boccardo, A. Longatto-Filho, M.E.L. Consolaro, L.L. Villa, S.S. Maria-Engler

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): M.G. Discacciati, J.C. Teixeira, S.S. Maria-Engler

Study supervision: M.G. Discacciati, S.S. Maria-Engler

Other (financial support): L.L. Villa

The authors gratefully acknowledge the collaboration of Douglas Munoz and Liliana Lucci Andrade for morphologic analysis of LETZ biopsies, and Markus Traue and Luiz Bahamondes for their assistance to the women included in this study.

This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), numbers 2008/03232-1 (to L.L. Villa) and 2012/09746-2 (to M.G. Discacciati and S.S. Maria-Engler) and National Counsel of Technological and Scientific Development pharmaceutical innovation (CNPQ-INCT-if; to S.S. Maria-Engler).

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.