MicroRNA-mediated extracellular matrix remodeling in squamous cell carcinoma of the oral cavity.

BACKGROUND
We evaluated microRNAs and extracellular matrix component profiles in squamous cell carcinoma of the oral cavity (OSCC) in comparison to healthy mucosa.


METHODS
Retrospective study investigating 64 microRNAs related to oncogenic process and to constituents of the extracellular matrix. We also performed immunohistochemical assays for molecules involved in the same biological processes.


RESULTS
High expression of miR-21-5p (p < 0.001) and miR-106-5p (p < 0.001) and low expression of miR-320a (p = 0.001) and miR-222-3p (p = 0.001) were predictors of malignancy. Individually, miR-21-5p exhibited the best statistical performance (area under the curve = 0.972; 95% confidence interval: 0.911-1.000) in the differentiation between tumor tissue and healthy mucosa. Moreover, tumor sample showed increased expression of MMP-2, MMP-9, α-laminin, and β-laminin in tumor-related fibroblasts and lower continuity of type IV collagen in the basement membrane.


CONCLUSION
The present study demonstrates the biological effects of microRNAs on the carcinogenesis of OSCC as well as the intense modification of the tumor microenvironment.


Introduction
Malignant neoplasms of the oral cavity are a major public health problem, given their increased incidence in the world and their unchanged mortality rate, despite advances in therapeutic modalities (1)(2)(3). The most common histological type is squamous cell carcinoma (SCC), which accounts for more than 90% of cases and whose main risk factor is smoking and its association with alcohol abuse (4)(5)(6)(7). The presence of lymph node metastases is one of the main factors contributing to worsening prognoses of patients with SCC of the oral cavity, and the greater the infiltration of the neoplasia is, as measured by tumor thickness or depth of invasion, the greater the contact of neoplastic cells with blood and lymphatic vessels, with consequent development of metastases (8)(9)(10)(11)(12)(13).
Among the factors that determine the highest invasion rate of a neoplasm is its interaction with the basement membrane and the extracellular matrix (ECM), making the environment more conducive to remodeled cell growth (14)(15)(16)(17). Thought in the past as only a support structure composed of fibrillar proteins (collagen and elastin), proteoglycans and structural glycoproteins, today, the ECM has been shown to be extremely important for tumor progression, due to the degradation of its components and angiogenesis, in addition to its role in proliferation, neoplasm migration and apoptosis (18)(19)(20).
Several genetic and epigenetic factors, including microRNAs (miRNAs), have been implicated in the interaction between neoplasias and the ECM, thus allowing for its remodeling and tumor invasion (21). miRNAs are small non-coding RNA molecules composed of 19 to 25 nucleotides that act by binding to the untranslated 3' region of messenger RNA (mRNA), resulting in degradation or translation inhibition. miRNAs act by negatively regulating gene expression at the posttranscription level (22). Previous studies have shown that the expression of miRNAs is dysregulated in several types of cancer, as more than 50% of the genes that regulate miRNA expression are found in fragile sites of chromosomes, which are susceptible to deletions or rearrangements. This dysregulation of miRNA expression may influence cell differentiation and proliferation and, therefore, miRNAs can act by themselves as oncogenes or tumor suppressor genes (23)(24)(25)(26).
Although several studies are being developed to determine the influence of miRNAs on the progression of SCC of the oral cavity, it is still difficult to identify biomarkers with good accuracy for predicting malignancy. The aim of this study is to evaluate the role of extracellular matrix remodeling, as well as the involvement of microRNAs, in the process of oral cavity SCC carcinogenesis.

Methods
This study was approved by the Institutional Review Board under protocol number (CAAE: 32884214.5.0000.0065). The informed consent was obtained from all the patients, except from those who have of lost ambulatory follow-up and the dead, which was dispensed as it is a retrospective study, with analysis of paraffined material. Paulo -ICESP), Sao Paulo, Brazil, were used. Tissues of oral mucosa preserved in paraffin blocks obtained from 11 patients who underwent surgeries due to benign oral cavity diseases were also included, constituting the control group.
Demographic, clinical and anatomopathological data were obtained from electronic medical records.
Tumor thickness measurements were performed using slides stained with hematoxylin and eosin by an independent pathologist. Representative areas with SCC were selected for tissue microarray (TMA) construction and immunohistochemical analysis of the chosen patient blocks.

MicroRNA detection technique
The level of microRNA expression was assessed by means of quantitative reverse transcription polymerase chain reaction (RT-qPCR). The RNA contained in 10 5μm sections of paraffin-embedded tissue was extracted using a MagMAX ® FFPE RNA Ultra Kit (Applied Biosystems ® , Foster City, Ca, USA) according to the manufacturer's protocol. RT-qPCR was performed using the TaqMan ® Low Density Array (TLDA; Thermo Fisher Scientific ® , Waltham, Ma, USA) system according to the manufacturer's protocol. RNA was reverse transcribed and polyadenylated according to the provided protocol, generating cDNA. All amplifications were performed in triplicate, and the cycle threshold (CT) values were determined with the Cloud platform (Thermo Fisher Scientific ® , Waltham, Ma, USA). miRNAs with CT values > 38 were considered indeterminate and discarded. Only miRNAs detected in at least 68% of the samples were considered for statistical analysis. Normalization between samples was performed using the Quantil method using Expander software, and miR-let-7i was used to normalize the expression values of the remaining miRNAs. A list of primers for the miRNAs studied is provided in Table 5  .
Regarding the Ki-67 antigen, its evaluation was quantitatively performed with the count of the first 500 tumor cells detected, counting from the first section evaluated, excluding areas of inflammation, fibrosis, necrosis and poor fixation. After counting 500 cells, the percentage of stained cells was obtained, regardless of the staining intensity (27).
Immunoreactivity to CD34 and VEGFR3 enabled us to calculate the microvessel density (MVD), obtained by the arithmetic mean of the number of vessels present in four sections from the same patient. All stained endothelial cells and cell clusters were counted as microvessels. For analysis, counts were performed at 400x magnification (28).

Statistical analysis
The values obtained by the study of each continuous parametric distribution variable were organized and described using the mean and standard deviation (SD) or standard error (SE). For categorical variables, absolute and relative frequencies were used. The cutoff values for quantitative variables were determined by receiver operating characteristic (ROC) curve analysis, and those for qualitative variables were determined by category associations. The Mann-Whitney test was used to compare quantitative variables between two sample populations. Additionally, variables with P-value < 0.10 in the univariate analysis were subjected to multivariate analyses, which were performed through linear regression models, for the selection of independent quantitative variables.

Group characteristics
Twenty-six patients diagnosed with SCC of the oral cavity were included in the study. Of these, 18 were men (69.2%), with a mean age of 62.4 years (SD of 8.9 years).
Eleven specimens of healthy mucosa obtained from surgeries involving benign lesions of the oral cavity were included, composing the "control" group of the study. Of these, nine were men (81.8%), with a mean age of 54 years (SD of 9.2 years), and six were still smokers (54.5%). At the time of the anatomopathological analysis of the surgical excision product, 15 patients (57%) had lymph node metastasis, 11 with extracapsular extension.
The mean follow-up period was 21 months. There were seven deaths (26.9%) due to the disease during the study period. The demographic and clinical variables of the study are shown in Table 1. and miR-34a-5p were observed in the tumor tissue compared to the healthy mucosa in the univariate analysis. The complete results are provided in Table 2.

Analysis of the expression of immunohistochemical markers in the normal mucosa and SCC of the oral cavity
Regarding the immunohistochemical markers, there was greater expression of p53, EGFR, metalloproteinase-2 (MMP-2), laminin beta, Ki-67 and CD34 in the tumor cells than in the healthy mucosa. Furthermore, increased expression of MMP-2, metalloproteinase-9 (MMP-9), laminin alpha and laminin beta in tumor-associated fibroblasts and lower continuity of type IV collagen in the basement membrane were observed. The complete data are provided in Table 3. These results demonstrate higher rates of cell proliferation, neoangiogenesis and extracellular matrix degradation in tumor specimens, including tumor-associated fibroblasts.

Comparison between differential microRNA expression and expression profile of immunohistochemical markers in SCC of the oral cavity
The microRNA expression levels were compared to various immunohistochemical markers that were classified based on staining intensity (high or low). There were several significant associations, and only these are shown in Table 4, demonstrating the interaction between microRNAs and various biological events in oral cavity SCC: significantly higher expression levels of let-7d-5p and miR17-5p, an increase in Bcl-2, lower expression of miR-378a-3p and higher expression of MMP-2 in peritumoral fibroblasts; higher expression of miR-29a-3p and higher MMP-2 levels in tumor cells; higher expression of miR-485-5p and higher p53 levels, higher expression of miR-200b-3p and increased rates of laminin beta in tumor-associated fibroblasts; higher expression of miR-29a-3p and higher rates of laminin beta in neoplastic cells; higher expression of miR-376c-3p accompanied by more heparanase 2 in tumor cells; higher expression of miR-155-5p and higher rates of laminin alpha in tumor-related fibroblasts; higher miR-494-3p expression and lower continuity of type IV collagen in the basement membrane; and higher miR-203a-3p expression and higher decorin rates, higher miR-27a-3p expression and lower miR-9-5p expression with higher Ki-67 levels, higher miR-99a-5p expression in high CD34-density microvessels and lower expression of miR-221-3p associated with a higher density of microvessels stained with VEGFR3.

Discussion
The The findings of the present study are vast and should be further explored. As limitations, we note that the sample size used was small, even for an experimental model.
We believe that microRNAs identified as differentially expressed between superficial and thick tumors should be validated in larger cohorts. However, the study was pioneer in this field and it is one of the biggest in number of microRNAs evaluated. Moreover, it was the first one to demonstrate the association of microRNAs in the ECM remodeling in oral cavity SCC patients.
In the present study, high miR-21-5p and miR-106-5p expression and low miR-