Matrix metalloproteinase expression in keratocystic odontogenic tumors and primary cells

Abstract

Keratocystic odontogenic tumors (KCOTs) are locally invasive, rapidly proliferating cystic lesions of the jaw. The bone-invasive nature of these tumors has been previously associated with the expression of matrix metalloproteinases (MMPs), which degrade the extracellular matrix. The purpose of this study was to assess the expression and activity of MMPs in primary KCOT cells and tumor tissue.

Methods

Four independently established KCOT primary cell populations were grown in Dulbecco’s modified Eagle medium supplemented with 10% FBS and antibiotics. Primary cells were analyzed by qRT-PCR and immunohistochemistry (IHC), and for secretion of active MMPs. Primary tumor sections were analyzed by IHC.

Results

Of the 18 human MMPs examined, 9 were consistently expressed in primary KCOT cells. MMP-2 and MMP-14 were highly expressed in all KCOT populations, while MMP-1, 3, 11, 12, 16, 17, and 19 were moderately expressed. MMP-3, 11, 12, 16, 17 and 19 were shown to be expressed in KCOTs for the first time. No significant differences in MMPS profiles were found between syndromic (KCOT-3) and non-syndromic cell populations (KCOT-1/2/4). Protein expression of MMP-1, 11, 12, 14 and 16 was confirmed in each KCOT cell populations by IHC. KCOT-3 cells secreted active MMP-2 as determined by a gel zymography assay. Expression of MMP-1, 2, 3, 11, 12, 14, and 16 was confirmed in matching primary KCOT tumor sections representing syndromic and non-syndromic KCOTs.

Conclusion

KCOT primary cell populations and tumors express a wide range of MMPs, which likely play a role in the bone-invasive nature of these tumors.

Introduction

Keratocystic odontogenic tumors (KCOTs) are locally invasive, rapidly proliferating cystic lesions of the jaw (1,2). They are invasive within the maxilla or mandible and are reported to recur at the primary site depending on the form of surgical treatment used (3–60%) (3). KCOTs can occur as non-syndromic, sporadic cases or associated with Nevoid basal cell carcinoma syndrome (NBCCS, also known as Gorlin’s syndrome). Both syndromic and non-syndromic KCOTs can have mutations within PTCH1/2, repressing receptors for hedgehog signaling, which may play a role in the development and progression of KCOTs (1).

Matrix metalloproteinases (MMPs) have been previously reported to be expressed and play a role in the invasive nature of KCOTs and other odontogenic tumors. MMPs are enzymes known to regulate the tumor microenvironment, including degradation of the extracellular matrix and components of the basement membrane (4). These enzymes have been linked with the metastatic and invasive potential of many human tumors and cancers. MMP-2, 9, 14 and others have been correlated with poor prognosis and outcomes in human cancers including prostate, breast, and colorectal cancer. Due to the tendency of KCOTs to invade the bone, studies have examined the expression of MMPs in primary tumors. However, due to the lack of established cellular models, these studies have been limited to immunostaining of paraffin-embedded tissues. In this study, we examined the expression of 18 human MMPs in primary syndromic and non-syndromic KCOT cell populations established in our laboratory as well the primary tumor (1,2). We identified several MMPs not previously reported expressed by KCOTs and confirmed the expression of various MMPs in primary KCOT tumor tissue.

Materials and methods

Quantitative real-time PCR (qRT-PCR)

Total RNA was isolated using the Qiagen RNeasy Mini Kit (Valencia, CA) according to manufacturer’s instructions. Isolated RNA was converted to cDNA using the Bio-Rad iScript cDNA Synthesis Kit (Herculus, CA). qRT-PCR reactions were performed using the RT² SYBR Green/Rox qPCR master mix (SABiosciences, Frederick, MD) and relative transcriptional levels of selected genes was detected using the ABI Prism 7500 Sequence Detection System (Applied Biosystems, Carlsbad, CA). The primers for MMP-3, 12 and 16 were obtained from RT² quantitative PCR primer assays (SABiosciences). Primers for MMP-1, 2, 11, 14, 17 and 19 were synthesized by Invitrogen (Carlsbad, CA; Supplementary Table 1). Cycle threshold (Ct) values corresponding to transcriptional levels were obtained and normalized to the GAPDH housekeeping gene to determine the delta Ct (dCt) value.

Immunohistochemistry (IHC)

Paraffin-embedded tumor tissue sections (5 μm) were cut, deparaffinized and rehydrated. Citrate antigen retrieval was performed for 10 min at 97 °C (pH 6.0). Samples were blocked in 3% hydrogen peroxide block for 15 min and incubated with specific primary antibodies overnight at 4 °C. Horseradish peroxidase polymer conjugated antibody was applied for 10 min and diaminobenzidine tetrachloride was used as the chromagen (Invitrogen) and hematoxylin for counterstain. For immunocytochemistry (ICC), KCOT cells isolated from primary tumors as previously described, were grown in 4-well chamber slides until 70% confluent then fixed with 4% formaldehyde (1,2). Staining was conducted as previously described for tissue sections. Antibodies used were MMP-1, 3, and 16 (Millipore, Billerica, MA); MMP-2 and 11 (Thermo Scientific, Lafayette, CO); MMP-12 (R&D Systems, Minneapolis, MN) and MMP-14 (Abcam, Cambridge, MA). Negative controls were incubated with PBST and secondary antibody without primary antibody. Samples were imaged with a Nikon Eclipse TE2000-E inverted microscope (Nikon Instruments, Melville, NY).

Gelatin zymography

Cells were cultured in 500 μl of serum-free media and supernatant was harvested after 24 h. Samples were electrophoresed under non-reducing conditions using 10% Tris-Gly gels containing 0.1% gelatin (Invitrogen). Gels were washed in 2.5% Triton X-100, incubated overnight in 50 mM Tris incubation buffer, then stained with SimpleBlue SafeStain (Invitrogen).

Results

MMP expression and activity in KCOT cell populations

Due to the invasive potential of KCOTs, the expression of MMPs was examined. The expression of 18 human MMPs was assessed in KCOT-1, 2, 3, and 4 primary cells. MMP-1, 2, 3, 11, 12, 14, 16, 17, and 19 were consistently expressed in all of the tumor cell populations (Figure 1A). MMP-2 and 14 demonstrated the highest mRNA expression levels. There was no significant difference in expression between non-syndromic KCOTs (KCOT-1, 2, and 4) and the syndromic KCOT-3. The protein expression of MMP-1, 11, 12, 14 and 16 were confirmed in each cell population using ICC (Figure 1B). Consistent with the high mRNA levels, MMP-14 showed strong staining and MMP-12 had the lowest mRNA levels and staining in KCOT-4 cells. MMP-1 expression was diffuse throughout the cell, whereas MMP-11, 14 and 16 were primarily expressed in the cytoplasm or membrane. Activity of MMP-2, which was expressed in all of the KCOT cell populations, was determined by gelatin zymography. KCOT-3 cells secreted pro and active MMP-2 as seen by the clearance zones in stained gels at approximately 72 kDa band, pro-MMP-2, and 68 kDa, active MMP-2 (Figure 1C).

Figure 1.

Matrix Metalloproteinase (MMP) expression and activity in primary KCOT cell populations. (A) Gene expression levels of MMPs in the four KCOT cell populations as determined by qRT-PCR using GAPDH expression to normalize the dataset. All experiments were performed in triplicate and repeated in duplicate, the error bars indicate SD. (B) Immunohistochemical staining of KCOT cells with primary antibodies to MMP-1, 11, 12, 14, and 16 (10×). Negative control shows staining without primary antibody. Insets are magnified 200%. (C) Pro- and active MMP-2 secreted into the conditioned media (24 h) from KCOT-3 cells detected using a gelatin zymography assay.

MMP expression in KCOT samples

Tumor tissue sections from the same tumors used to establish the primary cell populations were used to confirm expression of MMPs and to determine localization of MMPs within the tumor. Two tumors representative of non-syndromic KCOTs (KCOT-1 and 4) and one representative of a NBCCS-associated KCOT (KCOT-3) were chosen for this analysis. MMP-1, 2, 3, 11, 12, 14, and 16 were all expressed in all three primary tumors (Figure 2). MMPs showed either diffuse staining throughout the tumor or were localized along the parakeratin layer of the epithelial cells. Some MMPs are found only in the epithelial portion of the tumors, while others are also expressed in the tumor-associated stroma. For example, MMP-3 appears to be in KCOT-1 and KCOT-3 tumor and stroma; whereas in KCOT-4 it is primarily expressed in the tumor along the parakeratin layer. In general, KCOT-3, from a NBCSS patient, appeared to express lower levels of MMP-1, 2, 12 and 14.

Figure 2.

Expression of MMPs in KCOT tumors. Paraffin-embedded tumor sections representing non-syndromic (KCOT-1, KCOT-4) and syndromic (KCOT-3) KCOTs were immunostained for MMP-1, 2, 3, 11, 12, 14, and 16 (10×). Insets are magnified by 200%. The negative control was stained without primary antibodies. All sections were counterstained with hematoxylin.

Discussion

Odontogenic tumors originate from cellular remnants of tooth development and encompass a diverse group of tumors, of which the most common are KCOTs and ameloblastomas (2). KCOTs are benign, locally aggressive intra-osseous lesions of the jaw associated with a high rate of recurrence (3–60%) depending on the surgical treatment employed. A recent study reported on a higher prevalence in the posterior mandible, a slightly higher prevalence in males (1.2:1), and a 32% recurrence rate following curettage. KCOTs have a distinct histologic appearance with palisading of the basal epithelial cells and a corrugated, parakeratin stratified squamous epithelial layer (5). Limited studies have reported the presence of MMPs in KCOTs examining only a few human MMPs, and rarely is enzyme activity documented.

Numerous studies have described MMPs as mediators of the tumor microenvironment contributing to the development and progression of a diverse selection of human tumors and cancers, including odontogenic tumors (4,6). Roles attributed to MMPs in tumor development include regulation of tumor growth by mediating the release of stimulating growth factors into the tumor microenviroment (e.g. insulin-like growth factors, epidermal growth factor receptor ligands; MMP-1, 2, 3, 7, 9, 11, 19), promoting angiogenesis by degrading the extracellular matrix and increasing availability to vascular endothelial growth factor (MMP-1, 2, 7, 9, and 14) and supporting cell migration and invasion by promoting epithelial-to-mesenchymal transition allowing cells to decrease cell-cell adhesion (MMP-2, 3, 9, 13, 14). Of these MMPs important in tumor growth and progression, KCOT cells expressed MMP-1, 2, 3, 11, 14, and 19. We confirmed the activity of MMP-2 secreted by KCOT-3 cells into conditioned media. Active MMP-2 has not previously been reported in KCOTs.

Earlier studies have identified the expression of MMP-1, 2, 9, 13, and 26 in some KCOTs. An immunohistochemical study showed varying expression of MMP-1, 7, and 26 in syndromic (76%, 67%, and 62%, respectively) versus non-syndromic KCOTs (15%, 40%, and 35%) (6). We showed mRNA and protein expression of MMP-1 in each of the KCOT cell populations and the primary tumors examined. In sharp contrast, MMP-7 and MMP-26 were unable to be detected in our KCOT populations (data not shown). MMP-2 and 9 proteins were detected by immunostaining in KCOT and calcifying cystic odontogenic tumors, and MMP-9 was detected in radicular cysts, dentigerous cysts, KCOTs, and ameloblastomas (7,8). In one study, MMP-9 was more frequently expressed in the tumor-associated stroma of KCOTs (60%) then the tumor epithelium (10%). We were unable to detect MMP-9 in the KCOT cell populations we tested, which were derived from the epithelial tumor cells (data not shown) (1,2). MMP-13 immunostaining has previous been reported in 13 of 26 non-syndromic KCOTs and 11 of 13 syndromic cases (9). In our samples, MMP-13 was only expressed in the syndromic population (KCOT-3) and one non-sydromic population (KCOT-4) at very low mRNA levels (data not shown). Similar to our findings, Wahlgren et al. reported MMP-2 expression in 100% (10 of 10) of KCOTs examined (5). Interestingly, MMP-2 is activated in part by the membrane-type MMP-14, which has not previously been reported in KCOTs (10). KCOT-3 cells express both MMP-2 and 14, and secrete active MMP-2.

Conclusions

KCOT primary cell populations express a wide variety of MMPs, including many that have not been previously reported in KCOTs (MMP-3, 11, 12, 14, 16, 17, and 19). Primary tumors representative of syndromic and non-syndromic KCOTs were also shown to express many MMPs, which likely plays a role in the bone-invasive nature of these tumors.

Supplementary Material

Supp Table

Supplementary Table 1: Quantitative Real-Time MMP qRT-PCR primers