Immunoexpression of cleaved caspase-3 shows lower apoptotic area indices in lip carcinomas than in intraoral cancer

ABSTRACT Objective This study aimed to evaluate apoptosis by assessing cleaved caspase-3 immunoexpression in hyperplastic, potentially malignant disorder (PMD), and malignant tumors in intraoral and lower lip sites. Material and Methods A retrospective study using paraffin blocks with tissues from patients with inflammatory fibrous hyperplasia (IFH), actinic cheilitis, oral leukoplakia, lower lip and intraoral squamous cell carcinoma (SCC) was performed. The tissues were evaluated by immunohistochemical analysis with anti-cleaved caspase-3 antibody. Apoptotic area index was then correlated with lesion type. Results From 120 lesions assessed, 55 (46%) were cleaved caspase-3-positive. The SCC samples (n=40) had the highest apoptotic area indices (n=35; 87.5%). Significant differences were detected between SCCs and PMDs (p=0.0003), as well as SCCs and IFHs (p=0.001), regarding caspase-3 immunopositivity. Carcinomas of the lower lip had lower apoptotic area indices than intraoral cancer (p=0.0015). Conclusions Cleaved caspase-3 immunoexpression showed differences in oral SCCs and PMDs and demonstrated a distinct role of apoptosis in carcinogenesis of intraoral and lower lip cancer. In future, the expression of cleaved caspase-3 with other target molecules in oral cancer may be helpful in delineating the prognosis and treatment of these tumors.


INTRODUCTION
Oral cancer is the 11 th most common cancer in the world 6 . In Brazil, data from the Brazilian National Cancer Institute indicate that oral cancer the twelfth among females, with a total of 15,490 new cases expected in 2016. More than 90% of oral cancer cases are squamous cell carcinoma (SCC) 12 . Early detection of high-risk premalignancy can decrease morbidity and mortality associated with oral cancer 22 . Oral leukoplakia (OL) and actinic cheilitis (AC) are premalignant lesions that show histological diversity and are associated with an overall increased risk for the development of invasive oral SCC 7,22,30 .
Risk factors for the development of intraoral and extraoral lesions are distinct. Strong associations between intraoral SCC and alcohol consumption and tobacco smoking are well established. In contrast, cancers of the lip vermilion are strongly associated with chronic sun exposure, although they are also related to the cigarette placement site on the lip 2,17,19,29 .
Caspases are a family of cysteine proteases involved in signaling and execution of apoptotic cell death. They are synthesized as pro-enzymes and activated by proteolytic cleavage, and may cleave other caspases as part of the apoptotic signaling cascade. Caspase-3 is a major effector caspase. Its proenzyme is expressed in a wide range of tissues, including those with a high tissue turnover, such as intestinal epithelium and epidermis. Active caspase-3 staining was considered a reliable method of apoptotic index scoring 8 .
Investigations of apoptotic cells levels in oral SCC are contradictory. Some studies showing elevated 8,16,18 expression for oral SCC when compared with normal mucosa and others have suggested that apoptosis decreases as histological abnormality increases 24 .
Considering the different etiopathogeneses involved in the potentially malignant disorder (PMD) and SCC of intraoral and lower lip lesions, we hypothesized that a distinct apoptotic index was associated with these pathologies. Therefore, the purpose of this study was to characterize and compare cleaved caspase-3 immunoexpression malignant tumors.

Study population
with biopsy specimens were analyzed (n=120; 20 OL cases, 16 with and 4 without dysplasia; 20 AC cases, 15 with and 5 without dysplasia; 40 oral SCC cases: 20 intraoral SCCs and 20 lower lip SCCs; 40 intraoral IFH and 20 lower lip IFH). These blocks of Pathology at Antonio Pedro University Hospital (Federal Fluminense University). Hematoxylineosin-stained slides were reviewed, and initial This study was approved by the institutional ethics committee (CMM/HUAP 33/06).

Immunohistochemistry
Immunohistochemistry was performed on serial silane-pretreated glass slides (Sigma Chemicals, St. Louis, MO, USA), using a streptavidin-biotin immersed in 3% hydrogen peroxide for 30 min to block endogenous peroxidase activity and heated (94-96°C) for 30 min in 1 mM citric acid buffer (pH 6.0) for antigen retrieval (Target antigen retrieval solution, DAKOCytomation, Carpinteria, CA, USA). Sections from each block were incubated in 1:600 diluted solution of rabbit polyclonal antibody to cleaved caspase-3 protein, for 16-18 h at 4°C, as primary antibody (Asp 175, Cell Signaling Technology, Danvers, MA, USA). Peroxidase (DAB + Liquid DAB substrate, Chromogen system, DAKOCytomation, Carpinteria, CA, USA). Sections were counterstained with Harris' hematoxylin, dehydrated, cleared in xylene, and mounted. The primary antibody was omitted and replaced by an antibody diluent (Antibody diluent with background reducing components solution, DAKOCytomation, Carpinteria, CA, USA) for negative control, resulting in immunonegative slides in all instances. Positive controls comprised oral lichen planus sections, since several apoptotic keratinocytes are usually present in this lesion.

Quantitative Immunohistochemical Analysis
used to score the immunohistochemistry in the distinct lesions and sites evaluated. The positive slides were quantitatively evaluated by apoptotic area index.
of immunostaining by Nikon Microscope Eclipse E400 (Nikon, Tokyo, Tokyo, Japan) and Evolution TM MP Color 5.0 Mega-pixel Camera Kit (Media (20×/0.4 Plan Achromat; objective lens; total area: 705.630 μm 2 ), selected from hot-spot areas, were acquired per slide. Image analysis was performed with Image-Pro Plus 4.5 software (Media Cybernetics, Silver Spring, MD, USA).
United discrimination plane was used to segment the images (24-bit). This method, developed by one of the authors (VB) 3 from different images to determine the positivity discrimination plane, minimizing a possible visual variation in detection of immunostained areas over time when an interactive discrimination plane is used. To measure the total area, the discrimination plane was set at 0-255 in all RGB channels. Because immunohistochemical signals from epithelial tissue had to be separated from stromal signals, all lesion areas were manually outlined in all microscopic positivity was expressed as an area index (positive area/total area) instead of a labeling index because it was not possible to distinguish whether a positive cell.

Statistical methods
Shapiro-Wilk test was used to assess data normality. Kruskal-Wallis nonparametric test (H) and Mann-Whitney test (U) were used to compare apoptotic area indices within each group (intraoral -IFH, OL, and SCC; lower lip -IFH, AC, and SCC) and between groups. Binomial test and Dixon Q test were used in PMD quantitative result assessments.
When grouping lesions according to their location, Kruskal-Wallis test showed statistical differences between indices of the lower lip group lesions (IFH, AC, SCC; H=6.514; d.f.=2; p=0.039). When compared with AC (n=9), SCCs (n=15) had higher apoptotic area indices and Mann-Whitney between these groups (U=28; p=0.018). It was not possible to compare lower lip SCC or AC with IFH. In lower lip group, the occurrence of only one positive case for IFH failed to bring evidence of possible differences between AC and SCC. In the intraoral group (IFH, OL, SCC) Kruskal-Wallis test indicated statistical differences between indices of these lesion groups (H=12.144; d.f.=2; p=0.002). SCCs (n=20) had higher apoptotic area indices than IFH and OL, and Mann-Whitney test showed and OL (n=6; U=17; p=0.007), and between SCC and IFH (U=4; p=0.002). No statistical differences were detected between area indices of OL and IFH lesions (U=11; p=0.914).
In the PMD group, all intraoral leukoplakia lesions staining positive with anti-cleaved caspase-3 antibody were characterized by epithelial dysplasia.
noted when comparing area indices for intraoral leukoplakias with dysplasia (n=6) and AC with dysplasia (n=6; 0.0001 0.00007). The apoptotic area index of AC without epithelial dysplasia was higher than in AC epithelial dysplasia cases (U=1; p=0.036).

DISCUSSION
Apoptosis is a key aspect in the pathogenesis of many diseases, e.g., cancer, and in the response of neoplastic cells to systemic therapy 9,25,28 . Several studies in this area often compared normal and/or dysplastic epithelium with carcinoma 13,20,24,27 . IFH epithelium was treated as a control in our study normal epithelium. Areas of apparently benign epithelium adjacent to the malignant tissue were used in some studies as a normal control group 1 . However, this approach also has limitations because 13 , which asserts the occurrence of early genetic changes in histologically normal cells surrounding malignant tumors.
In our study, oral SCC had a higher apoptotic area index than other types of lesions (p<0.05) and IFH epithelium had the lowest apoptotic area index. Hague, et al. 8 (2004) observed that the proportion of active caspase-positive cells in oral SCCs was markedly higher than in normal oral increases during oral carcinogenesis. It has been also shown that in normal epithelium, cells positive for cleaved caspase-3 were rare. Our results concur statistical difference between apoptotic area indices of premalignant and IFH groups, however, cleaved caspase-3-positive specimens predominated in premalignant lesions when compared with IFH. Similarly, other studies have reported a progressive increase of apoptotic cells, from normal epithelium state, through epithelial dysplasia, until oral SCC development 16,18 .
In contrast, some studies suggested that the number of apoptotic cells decreased in oral lesions, as a result of increased molecular abnormalities in epithelial cells and cancer 20,24 . Tanda, et al. 27 the number of apoptotic cells in OL compared with normal oral mucosa. Considering the differences in apoptosis detection methods, analyses modes, and study models used in the investigations, inter-study 21 .
Our results showed differences between intraoral and lip SCCs. Cleaved caspase-3 levels were with lower lip SCCs. Lip SCCs have the best 5-year survival rates (over 90%) of all mouth SCCs, demonstrating their distinctive biological behavior 29 . Alterations in p53 expression suggest that this proteins are involved in lower lip carcinogenesis 15 . TP53 mutation caused by UV irradiation decreases the number of apoptotic keratinocytes in skin cancer 14 . Keratinocytes with wild type TP53 are usually eliminated by apoptosis, via p53, and keratinocytes with UV-induced TP53 mutation should be more susceptible to the tumor-promoting effects of UV radiation. These cells should survive, with an increased risk of becoming malignant 14 . Thus, it is important to emphasize the differences in molecular carcinogenesis of oral and lip SCCs, identifying proteins that may be employed as prognostic markers.
Molecular markers and immunoregulatory events predicting the malignant potential of AC are not well established and remain to be elucidated 7 . According to our data, the apoptotic area index of AC without epithelial dysplasia was higher than in cases with epithelial dysplasia. This too can be explained by UV radiation-caused TP53 mutation for cases without epithelial dysplasia, inhibiting the induction of apoptosis, as previously discussed. In contrast, all OL cases positive for cleaved caspase-3 protein showed epithelium dysplasia, probably suggesting a different molecular mechanism depending on etiological factors involved in premalignant lesion formation.
Several studies reported the potential use of activated caspase-3 as a biomarker to predict tumor responses to treatments 11 , in association with other prognostic variables (e.g., vascular invasion, lymph node metastasis, advanced clinical stage, and size of tumor) 4 , suggesting that cleaved caspase-3 could be used as a potential factor to predict tumor progression and poor prognosis in several cancer types 10 . In the current study, evaluation of cleaved caspase-3 levels in tissue biopsies showed differences in apoptotic area indices between pathologies and locations, which should be further studies.

CONCLUSIONS
In summary, the present study showed differences in the immunoexpression of cleaved caspase-3 in oral SCCs and PMDs. In addition, it showed a distinct role of apoptosis in carcinogenesis of intraoral and lower lip cancer, which may be helpful, in the future, in delineating the prognosis and treatment of these tumors. Further studies, correlating the expression of cleaved caspase-3 with other target molecules in oral cancer, should lead to the establishment of an effective prognostic immunohistochemical biomarker panel for oral carcinogenesis.