Experimental and Molecular Pathology 97 (2014) 332–337
Contents lists available at ScienceDirect
Experimental and Molecular Pathology journal homepage: www.elsevier.com/locate/yexmp
Hedgehog signaling pathway mediates tongue tumorigenesis in wild-type mice but not in Gal3-deficient mice Débora de Oliveira Santos a, Adriano Mota Loyola b, Sérgio Vitorino Cardoso b, Roger Chammas c, Fu-Tong Liu d, Paulo Rogério de Faria a,⁎ a
Uberlândia Federal University, Biomedical Science Institute, Department of Morphology, Uberlândia, Brazil Uberlândia Federal University, School of Dentistry, Uberlândia, Brazil São Paulo University, School of Medicine, São Paulo, Brazil d Department of Dermatology, University of California, Davis, School of Medicine, Sacramento, CA, USA b c
a r t i c l e
i n f o
Article history: Received 22 August 2014 Accepted 12 September 2014 Available online 16 September 2014 Keywords: Oral carcinogenesis Galectin-3 Tongue Mice Sonic hedgehog Wnt signaling
a b s t r a c t Oral squamous cell carcinoma (OSCC) is one of the most aggressive cancers of the oral cavity and an important cause of death worldwide. Currently, there are limited clinical tools aiding clinicians to establish its early diagnosis, and genetic and epigenetic events leading to the pathogenesis of OSCC remain unsolved. The use of carcinogen-induced knocked out mouse models would help to improve its early detection and also determine the role of proteins such as galectin-3 (Gal3) in this process. Here we used a mouse model of oral carcinogenesis employing two mouse genotypes: wild-type (Gal3+/+) and galectin-3-deficient mice (Gal3−/−) challenged by the carcinogen 4NQO for 16 weeks. After induction, the expression of Wnt1, Wnt3A, Shh and Gli3 proteins in tongue samples was evaluated using an immunohistochemistry approach. All samples of dysplasia and carcinoma were negative for Wnt1. Wnt3A expression was detected in both Gal3+/+ and Gal3−/− mice, at similar levels. Wnt3A expression did not predict tongue tumorigenesis in either genotype. Dysplastic- and carcinomaexpressing Shh was statistically significantly higher in Gal3+/+ mice than Gal3−/− mice (p b 0.0001), and was associated with tongue tumorigenesis only in the former. Gli3 expression decreased and increased from dysplasia to carcinoma in Gal3+/+ and Gal3−/− mice, respectively, although the difference was not significant. The results suggest that activated Wnt signaling is present in both mice, and that the Hh signaling pathway might play a role in tongue carcinoma development in Gal3+/+ mice. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Oral squamous cell carcinoma (OSCC) accounts for more than 90% of all malignant tumors arising in the oral epithelium surface and is a leading cause of cancer death worldwide (da Silva et al., 2011; Sant'ana et al., 2011). Although the complete mechanism responsible for its development has not yet been elucidated, it is known that genetic and epigenetic events affecting keratinocyte DNA may lead to OSCC formation. In this regard, the 4NQO-induced oral carcinogenesis model has been the goldstandard approach to identifying early changes in genes and their respective proteins as well as dysregulated signaling pathways that might be involved in the pathogenesis of OSCC (Kanojia and Vaidya, 2006; Reibel, 2003). Galectins are carbohydrate-binding proteins that play crucial roles in different tissues and cells, with galectin-3 (Gal3) being a central ⁎ Corresponding author at: Universidade Federal de Uberlândia, Instituto de Ciências Biomédicas, Avenida Pará 1720, Bloco 2B, Laboratório de Histologia, sala 2B-256, CEP: 38405-320, Uberlândia, MG, Brazil. E-mail address: [email protected] (P.R. de Faria).
http://dx.doi.org/10.1016/j.yexmp.2014.09.018 0014-4800/© 2014 Elsevier Inc. All rights reserved.
member of this family (Liu and Rabinovich, 2005; Markowska et al., 2010). Gal3, a protein of 31 kDa, has been strongly linked to tumorigenesis in many tissues, but its role in OSCC remains to be determined (Sant'ana et al., 2011; Takenaka et al., 2004). To shed light on this matter, our group recently published an in vivo study showing that the incidence of tongue carcinoma did not differ significantly between wild-type (Gal3+/+) and Gal3-deficient mice (Gal3−/−), indicating that Gal3 does not seem to play any role in driving oral tumor formation (de Faria et al., 2011). The Wnt signaling pathway has been implicated in a variety of normal cellular processes, including embryogenesis and differentiation (Giles et al., 2003). Studies have shown that aberrant Wnt signaling activation promotes tumorigenesis and may also be associated with OSCC development (Giles et al., 2003; Perez-Sayans et al., 2012). Following Wnt activation, beta-catenin is stabilized in the cytoplasm and then ferried to the nucleus via Gal3 mediation whereby Wnt target genes are upregulated (Sant'ana et al., 2011; Shimura et al., 2004; Song et al., 2009). Otherwise, glycogen synthase kinase-3beta (GSK-3b)-mediated beta-catenin phosphorylation leads it to be degraded via the ubiquitin–proteosome complex (Hagen and Vidal-Puig, 2002). We investigated this discovery
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
regarding the Gal3–beta-catenin interaction, and hence the interaction with the Wnt signaling pathway, using the same mouse model (Sant'ana et al., 2011). In this study, we found a significant increase in non-membranous beta-catenin expression from dysplasia to carcinoma when compared with membranous expression in both Gal3+/+ and Gal3−/− mice, suggesting that Wnt signaling seems to be activated even in the absence of Gal3 (Sant'ana et al., 2011). The Hedgehog signaling pathway (Hh) also plays a pivotal role in embryonic development (Beachy et al., 2004). In mammals, three Hh homologous genes are recognized: Sonic (Shh), Indian (Ihh), and Desert hedgehog (Dhh) (Mimeault and Batra, 2010). After binding of these proteins to the membrane receptor Patched and inducing Gli protein activation, a group of cell cycle-regulating genes are transcribed (Ruiz i Altaba, 1999; Tang et al., 2007; Xuan et al., 2006). As in the Wnt signaling pathway, GSK-3beta regulates this pathway by inhibiting the Gli family of proteins and it has been shown that this inhibition occurs mainly via a Sufu–GSK-3beta–Gli interaction, and in part due to a Sufu–Gal3 interaction (Briscoe and Therond, 2013; Kise et al., 2009; Mill et al., 2005; Paces-Fessy et al., 2004). Moreover, a recent report showed that augmented expression of inactive GSK-3beta was linked to tongue tumorigenesis in Gal3+/+ mice but not in Gal3 −/− mice (Mendonca et al., 2012). These findings prompted us to address whether the Hh signaling pathway might be differentially activated in Gal3+/+ mice compared to Gal3−/− mice in this setting. So, the aim of the current study was to investigate whether the Wnt and Hh signaling pathways influence tongue carcinogenesis in Gal3+/+ and Gal3−/− mice. The results suggest that the Hh signaling pathway plays an essential role in driving tongue tumorigenesis in Gal3+/+ mice and also confirm our previous report that activation of the Wnt signaling pathway occurs even in the absence of Gal3 and seems to be mediated by the Wnt3 signal. 2. Material and methods 2.1. Experimental protocol To study the expression of Wnt-1, Wnt-3A, Shh, and Gli-3 proteins in dysplasias and carcinomas developed experimentally in mouse tongue, a standardized protocol for the 4NQO-induced experimental mouse tongue carcinogenesis model was employed. For this, 38 Gal-3 −/− male mice aged 6 weeks and weighing 21–23 g, kindly bred and supplied by Hsu's group (Hsu et al., 2000), and 36 age-matched Gal3+/+ mice (control group) were challenged with 4NQO after an acclimatization period of 2 weeks. All animals were housed in the Center for Bioterism and Animal Experimentation at the Universidade Federal de Uberlândia (UFU), and maintained in controlled conditions of humidity, temperature, and a 12-h light–dark cycle. The animal study was managed following an animal protocol approved by the Committee on Animal Experimentation of the UFU. The carcinogen 4NQO was prepared based on the protocol established by Tang et al. (2004), with some modifications. In short, 4NQO was diluted in filtered water to a final concentration of 100 μg/ml. Every week a fresh 4NQO solution was prepared and given to the mice in their drinking water ad libitum. For all groups, 4NQO administration terminated after 16 weeks of carcinogen intake. After this period of induction, groups of five Gal3+/+ and Gal3−/− mice were killed at specific time points, as follows: immediately after the end of treatment (at week 16), 4 weeks later (at week 20), 8 weeks later (at week 24), 12 weeks later (at week 28), and 16 weeks later (at week 32). Before killing the mice by cervical dislocation, all of them were anesthetized with xylazine and ketamine via intraperitoneal injection. After that, their tongues were removed, fixed in buffered formalin (4%) for 24 h, routinely processed, and embedded in paraffin. The first fragment of each paraffin block was cut into 5-μm thick sections using a microtome and stained with hematoxylin and eosin (HE) for histopathological examination focusing on the detection of dysplasia (mild, moderate and severe) and carcinoma. To this end, both
333
tongue-developing lesions were diagnosed in accordance with criteria reported by Lumerman et al. (1995) and Cardesa et al. (2005), respectively. All lingual tissue samples were analyzed independently by three oral pathologists (PRF, SVC, and AML) and cases with controversial diagnoses were decided by consensus. In the meantime, serial sections of 3-μm thickness were taken from the same paraffin blocks with the aim of studying the expression of each antibody using a standardized immunohistochemical tool. 2.2. Immunohistochemistry The streptavidin–biotin peroxidase method was used to evaluate the expression of Wnt-1, Wnt-3A, Shh, and Gli-3 proteins in dysplasias and carcinomas in each group of mice. All information about the antibodies employed in the present study and their respective dilutions is depicted in Table 1. Initially, the fragments were mounted on glass slides previously treated with 3-aminopropyltrietoxy silane (Sigma, Chemical Co., St. Louis, USA), dewaxed in xylene and dehydrated in decreasing ethanol solutions. After this, the tissue sections were immersed in a solution of ammonium hydroxide (10%) and ethanol (95%) for 15 min to remove the formolic pigment and then treated with H2O2 (10 V) to block endogenous peroxidase activity. Next, these tissue sections were boiled in 1 mM EDTA solution, pH 8.0, with three cycles of 5 min each in a microwave, for antigen retrieval. Then, the slides were washed in Tris–HCl buffer solution (20 mM, pH 7.8) and incubated with the primary antibodies overnight at 4 °C. Following incubation, the fragments were washed in Tris–HCl buffer solution and then incubated with the biotinylated secondary antibody and streptavidin conjugated to horseradish peroxidase for 30 min each (LSAB, Dako, Carpinteria, CA, USA). The antibody reaction was detected using the diaminobenzidine chromogen and the tissue samples then counterstained with Harris's hematoxylin. A human breast carcinoma sample was used as a positive control for Wnt-1 and Wnt-3A, and the stomach and testis for Shh and Gli-3, respectively. Omission of primary antibody was used as the negative control. 2.3. Immunohistochemical evaluation A semi-quantitative analysis was used to evaluate the immunoreactivity of each antibody employed in the present study. To this end, both the intensity and percentage of positive cells were considered in the analysis (Sinicrope et al., 1995). Immunostaining intensity was scored in three categories: weak (1), moderate (2), and intense (3). The percentage of positive cells was classified in four categories: 0 (b5%), 1 (5–25%), 2 (26–50%), 3 (51–75%), and 4 (N75%). To determine the mean score/lesion value, the percentage and immunostaining intensity of positive cells were established for each lesion's field as mentioned above, and then the values multiplied to obtain a preliminary score/ field. Lastly, the sum of each field's score was divided by the number of fields took from the lesion to give the mean score/lesion value. All lesions were independently evaluated by two observers (DOS and PRF), and those ones were differentially scored were reanalyzed together until achieving a consensus. Moreover, due to the small size
Table 1 Antibodies used in immunohistochemical analysis. Antibody Wnt-1 Wnt-3A Gli-3 Shh a b
Source a
Santa Cruz Miliporeb Santa Cruza Santa Cruza
Catalog number
Dilution
6280 09-162 20688 9024
1:200 1:200 1:50 1:100
Santa Cruz Biotechnology, Santa Cruz, CA, USA. Millipore Corporation, Billerica, MA, USA.
334
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
of the lesions, especially dysplastic ones, all fields obtained/lesion were analyzed. 2.4. Statistical analysis Statistical analyses were conducted using GraphPad Prism 6.00 software (GraphPad Software, San Diego, USA). The D'Agostino–Pearson normality test was applied for each test. The Kruskal Wallis test followed by the Dunn post-test was used to compare the expression of each protein in mild, moderate, severe dysplasia and carcinoma from Gal-3+/+ and Gal3−/− mice. To investigate the role of these proteins during 4NQO-induced tongue carcinogenesis, differences in their expression between dysplasia (as a group) and carcinoma were evaluated using the Mann–Whitney U test. Statistical significance was determined at p b 0.05. 3. Results The findings on 4NQO-induced tongue carcinogenesis in Gal3+/+ and Gal3 −/− mice have previously been published elsewhere (de Faria et al., 2011). The immunohistochemical staining of Wnt1 was negative in all dysplasias and carcinomas diagnosed in the tongue of Gal3 +/+ and Gal3−/− mice. The frequency of dysplasias and carcinomas showing positivity for Wnt-3A was 80% and 73.3% respectively in Gal3 +/+ mice and, 69.5% and 66.7% in Gal3 −/− mice respectively. Independently from lesion subtype and mouse genotype, the intensity of Wnt3A expression ranged from mild to strong and was observed in the nucleus, cytoplasm and membrane (Fig. 1A). Regardless of lesion subtype, the mean positivity was always higher in Gal3+/+ than Gal3 −/− mice, although the difference was not significant (Fig. 1B). Likewise, no
significant difference was found in the mean Wnt3A expression during 4NQO-driven tongue transformation in each group of mice (Fig. 1C). In terms of Shh expression, 96.4% and 54.3% of dysplastic lesions and 100% and 40.7% of carcinomas diagnosed in 4NQO-treated Gal3 +/+ and Gal3−/− mice, respectively, were positive for this protein. Overall, the intensity of Shh expression in dysplasias and carcinomas was similar in both genotypes, ranging from mild to strong. Likewise, a cytoplasmic pattern of immunoreactivity was observed in both lesions independently of mouse genotype, with some lesions showing nuclear and rarely membrane expression (Fig. 2A). The mean Shh expression showed a steady increment during 4NQO-induced tongue carcinogenesis in Gal3 +/+ mice, with a significant difference between mild dysplasia and carcinoma (p b 0.001), and between moderate dysplasia and carcinoma (p b 0.01). The mean Shh expression increased to a much lesser extent from mild to severe dysplasia in Gal3 −/− mice, although the difference was not significant. Next, comparison of all 4NQO-induced, Shh-expressing lesions revealed a significant difference between Gal3 +/+ and Gal3 −/− mice (p b 0.0001) (Fig. 2B). Finally, the enhancement of Shh expression during 4NQO-driven tongue tumorigenesis was statistically significant in Gal3+/+ mice (p = 0.0037) but not in Gal3−/− mice (p = 0.74) (Fig. 2C). In terms of Gli-3, the frequency of dysplastic lesions and carcinomas exhibiting positivity was 93.6% and 64.7% in Gal3+/+ mice, and 84.3% and 73.9% in Gal3−/− mice, respectively. Despite more frequently observing Gli3-positive lesions in the latter group, this difference was not statistically significant. As observed for Shh protein, the intensity of Gl3 immunostaining varied from mild to strong in all lesions studied independently of mouse genotype, and with respect to its immunolocalization, it was predominantly seen in the cytoplasm, less frequently in the nucleus and rarely in the plasma membrane (Fig. 3A). The mean Gli3 immunoreactivity showed a gradual reduction from mild dysplasia to
Fig. 1. Wnt-3A expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Wnt-3A intensity in each lesion analyzed (Kruskal–Wallis test, p N 0.05). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in each group of mice (Mann–Whitney test, p N 0.05).
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
335
Fig. 2. Shh expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Shh intensity in each lesion analyzed (Kruskal–Wallis test, p b 0.001). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in Gal3+/+ mice (Mann–Whitney test, p b 0.01) and Gal3−/− mice (Mann–Whitney test, p N 0.05).
carcinoma in Gal3+/+ mice and a slight increment from mild dysplasia to carcinoma in Gal3−/− mice, although the difference was not significant (Fig. 3B). Unlike Shh expression, Gli3 expression during 4NQOinduced tongue tumorigenesis was not significant in either mouse genotype (Fig. 3C). 4. Discussion In this study, we analyzed the potential role of some key regulatory proteins of two important signaling pathways during tongue carcinogenesis in mice that have been implicated in tumorigenesis in many tissues. To sum up, our findings revealed similar mean Wnt3A expression in the setting of tongue malignant transformation in both groups of mice, confirming a previously published report showing aberrant Wnt signaling pathway activity from dysplasia to carcinoma in Gal3 +/+ and Gal3−/− mice. Together, these results indicate that Wnt signaling pathway activation is not dependent on Gal3, particularly during tongue tumorigenesis, at least in mice. At the same time, it was found that Shh protein, but not Gli3 protein, was strongly associated with tongue carcinoma in Gal3+/+ mice, suggesting that Shh protein, the hallmark of the activated Hh pathway, is essential for tongue tumor formation in these mice and might be regulated at a certain level by Gal3. The relationship between the Wnt signaling pathway and Gal3 protein was first described by Shimura et al. (2004), who showed that this lectin acts as a chaperone by helping beta-catenin translocate into the nucleus to enhance Wnt target gene transcription, including that of CCND1 and MYC (Shimura et al., 2004, 2005). Its critical role is currently understood as involving the promotion of tumor development. Giving more insight into this matter, Abdel-Aziz et al. (2008) found a significantly higher incidence of lung tumors in Gal3 +/+ than Gal3 −/− mice and concluded that the absence of Gal3 restrained
lung carcinogenesis by inhibiting the Wnt signaling pathway. In contrast, Eude-Le Parco et al. (2009) showed no influence of Gal3 in the development of breast and colon tumors or in the occurrence of metastasis in LGALS3-silenced mice. Similarly, our previous report found no difference in the incidence of tongue tumor between Gal3 +/+ and Gal3−/− mice, which in part might be associated with equal and persistent Wnt signaling pathway activation in both genotypes (de Faria et al., 2011; Sant'ana et al., 2011). This observation was confirmed herein, because Wnt3 expression, an important activator of this pathway, was similarly expressed in all dysplasia and carcinoma samples from both genotypes. No study using 4NQO-induced tongue tumorigenesis model has evaluated the Wnt3A protein in Gal3−/− mice. In human oral leukoplakia samples, for instance, Wnt3 expression was detected in dysplasia-developing leukoplakias, consistent with the Wnt signalinginduced malignancy (Ishida et al., 2007). Unexpectedly, we found no Wnt1 expression in any lesions investigated despite all control samples being positive for this protein. This finding is at odds with the study by Fracalossi et al. (2010), who identified Wnt1 expression in the majority of dysplasias and carcinomas from rats aged 12–20 weeks, with no difference between the two lesions. Taken as a whole, these findings allow us to conclude that tongue tumorigenesis might be driven by the Wnt signaling pathway through Wnt3 mediation, even in the absence of Gal3. Regarding the Hh signaling pathway players, it was found that Shh immunoreactivity was significantly higher in Gal3+/+ than Gal3−/− mice in both dysplastic and carcinomatous lesions. Furthermore, Gal3+/+developing carcinomas exhibited much higher Shh expression than dysplasias and this finding was strikingly associated with tongue tumorigenesis. Similar results have been observed in gastric and cervical carcinomas when compared with their respective premalignant counterparts (Wang et al., 2006; Xuan et al., 2006). Additionally, recent data have demonstrated that Shh overexpression might be involved in
336
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
Fig. 3. Gli3 expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Gli3 intensity in each lesion analyzed (Kruskal–Wallis test, p N 0.05). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in Gal3+/+ and Gal3−/− mice (Mann–Whitney test, p N 0.05).
small cell carcinoma of lung, adenocarcinoma of the digestive tract, and more recently OSCC (Berman et al., 2003; Wang et al., 2012; Watkins et al., 2003). In a study using OSCC cell lines, it was shown that the activated Hh signaling pathway induced augmented 50-bromo-20deoxyuridine (BrdU) incorporation by tumoral cells and then an increment in the cellular proliferation index (Nishimaki et al., 2004). Also, an established OSCC xenograft mouse model study revealed that nude mice transfected with Shh-inhibited cell lines presented reduced tumor growth when compared with control siRNA-transfected and parental cells (Honami et al., 2012). As a whole, these findings suggest that dysregulated Hh signaling pathway activation might be triggered during the transition from dysplasia to carcinoma in the tongue epithelial lining of Gal3+/+ mice, probably augmenting cellular proliferation. Another protein investigated in the present article was Gli3. This is a member of the Gli family of transcription factors, which together with Gli1 and Gli2, is responsible for triggering the transcription of genes under Hh signaling-dependent activity (Schneider et al., 2010). On the other hand, in the absence of Hh signaling, these proteins are phosphorylated by PKA, GSK-3b and CKI in a sequential manner, which, in turn, targets them for splitting by the ubiquitin–proteosome complex, therefore interrupting their transcriptional activities (Schneider et al., 2010; Tempe et al., 2006). Moreover, the occurrence of such inhibition is also seen in the presence of Hh signaling activity but in a GSK-3bindependent manner (Kise et al., 2009; Tempe et al., 2006). In the present study the level of Gli3 expression showed a slight reduction from dysplasia to carcinoma in Gal3+/+ mice, while an inverse pattern of expression was observed in Gal3 −/− mice. However, the difference was not significant. Furthermore, no association with tongue tumorigenesis was noted in either genotype. At first glance, this finding seems to be contradictory because significant overexpression of Shh
was detected in Gal3 +/+ mouse lesions, which might indicate increased Hh signaling pathway upregulation in this group when compared with Gal3 −/− mice. However, Gli3 expression is not always accompanied by uncontrolled Hh signaling activity and one recent study showed higher Gli3 expression in healthy mucosa than in head and neck carcinoma samples (Dimitrova et al., 2013). It is well known that Gli3 contains C-terminal activation and N-terminal repressor domains, which allows it to have a dual role in the Hh pathway under influence of the Hh ligand and its proteolytic processing after Sufu– GSK-3b-mediated phosphorylation (Amakye et al., 2013; Briscoe and Therond, 2013; Kise et al., 2009). Given a recent report revealing high inactive GSK-3b expression in Gal3+/+ mouse lesions, one might speculate that it could have influenced Gli3 expression (Mendonca et al., 2012). However, as a predominantly cytoplasmic rather than nuclear pattern of expression was observed in all samples studied, this suggests an additional control mechanism regulating the level of Gli3 in this setting. A recent study showed that Sufu, a well-known negative regulator of Hh signaling, is needed to efficiently promote the Gli3–GSK-3b interaction and then its phosphorylation and partial processing (Kise et al., 2009). Another study observed that Sufu, which also tethers Gli proteins in the cytoplasm, interacts physically with Gal3, preventing it from accumulating in the nucleus (Paces-Fessy et al., 2004). Whether the Sufu–Gal3 interaction is essential in the regulation of Gli protein activities, including Gli3, and whether it also occurs in the context of tongue carcinogenesis in these mice remains an open question of great importance for future research. In conclusion, our results show that Shh protein is frequently expressed in Gal3+/+ mouse lesions and is strongly associated with tongue tumorigenesis, establishing that the Hh signaling pathway plays an essential role in this context. Additionally, we confirmed our
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
previous report that Wnt signaling activation is not influenced by the Gal3 protein and seems to occur in both genotypes during malignant transformation through the Wnt3A signal. However, further study will be required to determine how Gal3 itself might lead to Hh signaling activation in the setting of tongue malignant transformation in Gal3+/+ mice. Conflict of interest statement All authors disclose any financial and personal relationships with other people or organizations that could inappropriately influence (bias) their work. We denied any potential conflicts of interest include employment, consultancies, stock ownership, honoraria, paid expert testimony, patent applications/registrations, and grants or other funding. Acknowledgments The authors would like to acknowledge the Research Supporting Foundation of Minas Gerais (FAPEMIG-CDS-APQ-00397-09) for the financial support. References Abdel-Aziz, H.O., et al., 2008. Targeted disruption of the galectin-3 gene results in decreased susceptibility to NNK-induced lung tumorigenesis: an oligonucleotide microarray study. J. Cancer Res. Clin. Oncol. 134, 777–788. Amakye, D., et al., 2013. Unraveling the therapeutic potential of the Hedgehog pathway in cancer. Nat. Med. 19, 1410–1422. Beachy, P.A., et al., 2004. Tissue repair and stem cell renewal in carcinogenesis. Nature 432, 324–331. Berman, D.M., et al., 2003. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature 425, 846–851. Briscoe, J., Therond, P.P., 2013. The mechanisms of Hedgehog signalling and its roles in development and disease. Nat. Rev. Mol. Cell Biol. 14, 416–429. Cardesa, A., et al., 2005. Tumors of the hypopharynx, larynx and trachea. In: Barnes, L., et al. (Eds.), World Health Organization Classification of Tumors. Pathology and Genetics of Head and Neck Tumors. IARC, Lyon, pp. 118–121. da Silva, S.D., et al., 2011. Advances and applications of oral cancer basic research. Oral Oncol. 47, 783–791. de Faria, P.R., et al., 2011. Absence of galectin-3 does not affect the development of experimental tongue carcinomas in mice. Exp. Mol. Pathol. 90, 189–193. Dimitrova, K., et al., 2013. Overexpression of the Hedgehog signalling pathway in head and neck squamous cell carcinoma. Onkologie 36, 279–286. Eude-Le Parco, I., et al., 2009. Genetic assessment of the importance of galectin-3 in cancer initiation, progression, and dissemination in mice. Glycobiology 19, 68–75. Fracalossi, A.C., et al., 2010. Wnt/beta-catenin signalling pathway following rat tongue carcinogenesis induced by 4-nitroquinoline 1-oxide. Exp. Mol. Pathol. 88, 176–183. Giles, R.H., et al., 2003. Caught up in a Wnt storm: Wnt signaling in cancer. Biochim. Biophys. Acta 1653, 1–24. Hagen, T., Vidal-Puig, A., 2002. Characterisation of the phosphorylation of beta-catenin at the GSK-3 priming site Ser45. Biochem. Biophys. Res. Commun. 294, 324–328. Honami, T., et al., 2012. Sonic hedgehog signaling promotes growth of oral squamous cell carcinoma cells associated with bone destruction. Oral Oncol. 48, 49–55. Hsu, D.K., et al., 2000. Targeted disruption of the galectin-3 gene results in attenuated peritoneal inflammatory responses. Am. J. Pathol. 156, 1073–1083. Ishida, K., et al., 2007. Nuclear localization of beta-catenin involved in precancerous change in oral leukoplakia. Mol. Cancer 6, 62.
337
Kanojia, D., Vaidya, M.M., 2006. 4-Nitroquinoline-1-oxide induced experimental oral carcinogenesis. Oral Oncol. 42, 655–667. Kise, Y., et al., 2009. Sufu recruits GSK3beta for efficient processing of Gli3. Biochem. Biophys. Res. Commun. 387, 569–574. Liu, F.T., Rabinovich, G.A., 2005. Galectins as modulators of tumour progression. Nat. Rev. Cancer 5, 29–41. Lumerman, H., et al., 1995. Oral epithelial dysplasia and the development of invasive squamous cell carcinoma. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 79, 321–329. Markowska, A.I., et al., 2010. Galectin-3 is an important mediator of VEGF- and bFGFmediated angiogenic response. J. Exp. Med. 207, 1981–1993. Mendonca, D.F., et al., 2012. The inactive form of glycogen synthase kinase-3beta is associated with the development of carcinomas in galectin-3 wild-type mice, but not in galectin-3-deficient mice. Int. J. Clin. Exp. Pathol. 5, 547–554. Mill, P., et al., 2005. Shh controls epithelial proliferation via independent pathways that converge on N-Myc. Dev. Cell 9, 293–303. Mimeault, M., Batra, S.K., 2010. Frequent deregulations in the hedgehog signaling network and cross-talks with the epidermal growth factor receptor pathway involved in cancer progression and targeted therapies. Pharmacol. Rev. 62, 497–524. Nishimaki, H., et al., 2004. A role of activated Sonic hedgehog signaling for the cellular proliferation of oral squamous cell carcinoma cell line. Biochem. Biophys. Res. Commun. 314, 313–320. Paces-Fessy, M., et al., 2004. The negative regulator of Gli, suppressor of fused (Sufu), interacts with SAP18, galectin3 and other nuclear proteins. Biochem. J. 378, 353–362. Perez-Sayans, M., et al., 2012. The role of the adenomatous polyposis coli (APC) in oral squamous cell carcinoma. Oral Oncol. 48, 56–60. Reibel, J., 2003. Prognosis of oral pre-malignant lesions: significance of clinical, histopathological, and molecular biological characteristics. Crit. Rev. Oral Biol. Med. 14, 47–62. Ruiz i Altaba, A., 1999. Gli proteins and Hedgehog signaling: development and cancer. Trends Genet. 15, 418–425. Sant'ana, J.M., et al., 2011. Activation of the Wnt/beta-catenin signaling pathway during oral carcinogenesis process is not influenced by the absence of galectin-3 in mice. Anticancer Res. 31, 2805–2811. Schneider, F.T., et al., 2010. Sonic hedgehog acts as a negative regulator of {beta}-catenin signaling in the adult tongue epithelium. Am. J. Pathol. 177, 404–414. Shimura, T., et al., 2004. Galectin-3, a novel binding partner of beta-catenin. Cancer Res. 64, 6363–6367. Shimura, T., et al., 2005. Implication of galectin-3 in Wnt signaling. Cancer Res. 65, 3535–3537. Sinicrope, F.A., et al., 1995. bcl-2 and p53 oncoprotein expression during colorectal tumorigenesis. Cancer Res. 55, 237–241. Song, S., et al., 2009. Galectin-3 mediates nuclear beta-catenin accumulation and Wnt signaling in human colon cancer cells by regulation of glycogen synthase kinase3beta activity. Cancer Res. 69, 1343–1349. Takenaka, Y., et al., 2004. Galectin-3 and metastasis. Glycoconj. J. 19, 543–549. Tang, X.H., et al., 2004. Oral cavity and esophageal carcinogenesis modeled in carcinogentreated mice. Clin. Cancer Res. 10, 301–313. Tang, J.Y., et al., 2007. Novel Hedgehog pathway targets against basal cell carcinoma. Toxicol. Appl. Pharmacol. 224, 257–264. Tempe, D., et al., 2006. Multisite protein kinase A and glycogen synthase kinase 3beta phosphorylation leads to Gli3 ubiquitination by SCFbetaTrCP. Mol. Cell. Biol. 26, 4316–4326. Wang, L.H., et al., 2006. Increased expression of sonic hedgehog and altered methylation of its promoter region in gastric cancer and its related lesions. Mod. Pathol. 19, 675–683. Wang, Y.F., et al., 2012. Expression of hedgehog signaling molecules as a prognostic indicator of oral squamous cell carcinoma. Head Neck 34, 1556–1561. Watkins, D.N., et al., 2003. Hedgehog signaling: progenitor phenotype in small-cell lung cancer. Cell Cycle 2, 196–198. Xuan, Y.H., et al., 2006. Enhanced expression of hedgehog signaling molecules in squamous cell carcinoma of uterine cervix and its precursor lesions. Mod. Pathol. 19, 1139–1147.
Contents lists available at ScienceDirect
Experimental and Molecular Pathology journal homepage: www.elsevier.com/locate/yexmp
Hedgehog signaling pathway mediates tongue tumorigenesis in wild-type mice but not in Gal3-deficient mice Débora de Oliveira Santos a, Adriano Mota Loyola b, Sérgio Vitorino Cardoso b, Roger Chammas c, Fu-Tong Liu d, Paulo Rogério de Faria a,⁎ a
Uberlândia Federal University, Biomedical Science Institute, Department of Morphology, Uberlândia, Brazil Uberlândia Federal University, School of Dentistry, Uberlândia, Brazil São Paulo University, School of Medicine, São Paulo, Brazil d Department of Dermatology, University of California, Davis, School of Medicine, Sacramento, CA, USA b c
a r t i c l e
i n f o
Article history: Received 22 August 2014 Accepted 12 September 2014 Available online 16 September 2014 Keywords: Oral carcinogenesis Galectin-3 Tongue Mice Sonic hedgehog Wnt signaling
a b s t r a c t Oral squamous cell carcinoma (OSCC) is one of the most aggressive cancers of the oral cavity and an important cause of death worldwide. Currently, there are limited clinical tools aiding clinicians to establish its early diagnosis, and genetic and epigenetic events leading to the pathogenesis of OSCC remain unsolved. The use of carcinogen-induced knocked out mouse models would help to improve its early detection and also determine the role of proteins such as galectin-3 (Gal3) in this process. Here we used a mouse model of oral carcinogenesis employing two mouse genotypes: wild-type (Gal3+/+) and galectin-3-deficient mice (Gal3−/−) challenged by the carcinogen 4NQO for 16 weeks. After induction, the expression of Wnt1, Wnt3A, Shh and Gli3 proteins in tongue samples was evaluated using an immunohistochemistry approach. All samples of dysplasia and carcinoma were negative for Wnt1. Wnt3A expression was detected in both Gal3+/+ and Gal3−/− mice, at similar levels. Wnt3A expression did not predict tongue tumorigenesis in either genotype. Dysplastic- and carcinomaexpressing Shh was statistically significantly higher in Gal3+/+ mice than Gal3−/− mice (p b 0.0001), and was associated with tongue tumorigenesis only in the former. Gli3 expression decreased and increased from dysplasia to carcinoma in Gal3+/+ and Gal3−/− mice, respectively, although the difference was not significant. The results suggest that activated Wnt signaling is present in both mice, and that the Hh signaling pathway might play a role in tongue carcinoma development in Gal3+/+ mice. © 2014 Elsevier Inc. All rights reserved.
1. Introduction Oral squamous cell carcinoma (OSCC) accounts for more than 90% of all malignant tumors arising in the oral epithelium surface and is a leading cause of cancer death worldwide (da Silva et al., 2011; Sant'ana et al., 2011). Although the complete mechanism responsible for its development has not yet been elucidated, it is known that genetic and epigenetic events affecting keratinocyte DNA may lead to OSCC formation. In this regard, the 4NQO-induced oral carcinogenesis model has been the goldstandard approach to identifying early changes in genes and their respective proteins as well as dysregulated signaling pathways that might be involved in the pathogenesis of OSCC (Kanojia and Vaidya, 2006; Reibel, 2003). Galectins are carbohydrate-binding proteins that play crucial roles in different tissues and cells, with galectin-3 (Gal3) being a central ⁎ Corresponding author at: Universidade Federal de Uberlândia, Instituto de Ciências Biomédicas, Avenida Pará 1720, Bloco 2B, Laboratório de Histologia, sala 2B-256, CEP: 38405-320, Uberlândia, MG, Brazil. E-mail address: [email protected] (P.R. de Faria).
http://dx.doi.org/10.1016/j.yexmp.2014.09.018 0014-4800/© 2014 Elsevier Inc. All rights reserved.
member of this family (Liu and Rabinovich, 2005; Markowska et al., 2010). Gal3, a protein of 31 kDa, has been strongly linked to tumorigenesis in many tissues, but its role in OSCC remains to be determined (Sant'ana et al., 2011; Takenaka et al., 2004). To shed light on this matter, our group recently published an in vivo study showing that the incidence of tongue carcinoma did not differ significantly between wild-type (Gal3+/+) and Gal3-deficient mice (Gal3−/−), indicating that Gal3 does not seem to play any role in driving oral tumor formation (de Faria et al., 2011). The Wnt signaling pathway has been implicated in a variety of normal cellular processes, including embryogenesis and differentiation (Giles et al., 2003). Studies have shown that aberrant Wnt signaling activation promotes tumorigenesis and may also be associated with OSCC development (Giles et al., 2003; Perez-Sayans et al., 2012). Following Wnt activation, beta-catenin is stabilized in the cytoplasm and then ferried to the nucleus via Gal3 mediation whereby Wnt target genes are upregulated (Sant'ana et al., 2011; Shimura et al., 2004; Song et al., 2009). Otherwise, glycogen synthase kinase-3beta (GSK-3b)-mediated beta-catenin phosphorylation leads it to be degraded via the ubiquitin–proteosome complex (Hagen and Vidal-Puig, 2002). We investigated this discovery
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
regarding the Gal3–beta-catenin interaction, and hence the interaction with the Wnt signaling pathway, using the same mouse model (Sant'ana et al., 2011). In this study, we found a significant increase in non-membranous beta-catenin expression from dysplasia to carcinoma when compared with membranous expression in both Gal3+/+ and Gal3−/− mice, suggesting that Wnt signaling seems to be activated even in the absence of Gal3 (Sant'ana et al., 2011). The Hedgehog signaling pathway (Hh) also plays a pivotal role in embryonic development (Beachy et al., 2004). In mammals, three Hh homologous genes are recognized: Sonic (Shh), Indian (Ihh), and Desert hedgehog (Dhh) (Mimeault and Batra, 2010). After binding of these proteins to the membrane receptor Patched and inducing Gli protein activation, a group of cell cycle-regulating genes are transcribed (Ruiz i Altaba, 1999; Tang et al., 2007; Xuan et al., 2006). As in the Wnt signaling pathway, GSK-3beta regulates this pathway by inhibiting the Gli family of proteins and it has been shown that this inhibition occurs mainly via a Sufu–GSK-3beta–Gli interaction, and in part due to a Sufu–Gal3 interaction (Briscoe and Therond, 2013; Kise et al., 2009; Mill et al., 2005; Paces-Fessy et al., 2004). Moreover, a recent report showed that augmented expression of inactive GSK-3beta was linked to tongue tumorigenesis in Gal3+/+ mice but not in Gal3 −/− mice (Mendonca et al., 2012). These findings prompted us to address whether the Hh signaling pathway might be differentially activated in Gal3+/+ mice compared to Gal3−/− mice in this setting. So, the aim of the current study was to investigate whether the Wnt and Hh signaling pathways influence tongue carcinogenesis in Gal3+/+ and Gal3−/− mice. The results suggest that the Hh signaling pathway plays an essential role in driving tongue tumorigenesis in Gal3+/+ mice and also confirm our previous report that activation of the Wnt signaling pathway occurs even in the absence of Gal3 and seems to be mediated by the Wnt3 signal. 2. Material and methods 2.1. Experimental protocol To study the expression of Wnt-1, Wnt-3A, Shh, and Gli-3 proteins in dysplasias and carcinomas developed experimentally in mouse tongue, a standardized protocol for the 4NQO-induced experimental mouse tongue carcinogenesis model was employed. For this, 38 Gal-3 −/− male mice aged 6 weeks and weighing 21–23 g, kindly bred and supplied by Hsu's group (Hsu et al., 2000), and 36 age-matched Gal3+/+ mice (control group) were challenged with 4NQO after an acclimatization period of 2 weeks. All animals were housed in the Center for Bioterism and Animal Experimentation at the Universidade Federal de Uberlândia (UFU), and maintained in controlled conditions of humidity, temperature, and a 12-h light–dark cycle. The animal study was managed following an animal protocol approved by the Committee on Animal Experimentation of the UFU. The carcinogen 4NQO was prepared based on the protocol established by Tang et al. (2004), with some modifications. In short, 4NQO was diluted in filtered water to a final concentration of 100 μg/ml. Every week a fresh 4NQO solution was prepared and given to the mice in their drinking water ad libitum. For all groups, 4NQO administration terminated after 16 weeks of carcinogen intake. After this period of induction, groups of five Gal3+/+ and Gal3−/− mice were killed at specific time points, as follows: immediately after the end of treatment (at week 16), 4 weeks later (at week 20), 8 weeks later (at week 24), 12 weeks later (at week 28), and 16 weeks later (at week 32). Before killing the mice by cervical dislocation, all of them were anesthetized with xylazine and ketamine via intraperitoneal injection. After that, their tongues were removed, fixed in buffered formalin (4%) for 24 h, routinely processed, and embedded in paraffin. The first fragment of each paraffin block was cut into 5-μm thick sections using a microtome and stained with hematoxylin and eosin (HE) for histopathological examination focusing on the detection of dysplasia (mild, moderate and severe) and carcinoma. To this end, both
333
tongue-developing lesions were diagnosed in accordance with criteria reported by Lumerman et al. (1995) and Cardesa et al. (2005), respectively. All lingual tissue samples were analyzed independently by three oral pathologists (PRF, SVC, and AML) and cases with controversial diagnoses were decided by consensus. In the meantime, serial sections of 3-μm thickness were taken from the same paraffin blocks with the aim of studying the expression of each antibody using a standardized immunohistochemical tool. 2.2. Immunohistochemistry The streptavidin–biotin peroxidase method was used to evaluate the expression of Wnt-1, Wnt-3A, Shh, and Gli-3 proteins in dysplasias and carcinomas in each group of mice. All information about the antibodies employed in the present study and their respective dilutions is depicted in Table 1. Initially, the fragments were mounted on glass slides previously treated with 3-aminopropyltrietoxy silane (Sigma, Chemical Co., St. Louis, USA), dewaxed in xylene and dehydrated in decreasing ethanol solutions. After this, the tissue sections were immersed in a solution of ammonium hydroxide (10%) and ethanol (95%) for 15 min to remove the formolic pigment and then treated with H2O2 (10 V) to block endogenous peroxidase activity. Next, these tissue sections were boiled in 1 mM EDTA solution, pH 8.0, with three cycles of 5 min each in a microwave, for antigen retrieval. Then, the slides were washed in Tris–HCl buffer solution (20 mM, pH 7.8) and incubated with the primary antibodies overnight at 4 °C. Following incubation, the fragments were washed in Tris–HCl buffer solution and then incubated with the biotinylated secondary antibody and streptavidin conjugated to horseradish peroxidase for 30 min each (LSAB, Dako, Carpinteria, CA, USA). The antibody reaction was detected using the diaminobenzidine chromogen and the tissue samples then counterstained with Harris's hematoxylin. A human breast carcinoma sample was used as a positive control for Wnt-1 and Wnt-3A, and the stomach and testis for Shh and Gli-3, respectively. Omission of primary antibody was used as the negative control. 2.3. Immunohistochemical evaluation A semi-quantitative analysis was used to evaluate the immunoreactivity of each antibody employed in the present study. To this end, both the intensity and percentage of positive cells were considered in the analysis (Sinicrope et al., 1995). Immunostaining intensity was scored in three categories: weak (1), moderate (2), and intense (3). The percentage of positive cells was classified in four categories: 0 (b5%), 1 (5–25%), 2 (26–50%), 3 (51–75%), and 4 (N75%). To determine the mean score/lesion value, the percentage and immunostaining intensity of positive cells were established for each lesion's field as mentioned above, and then the values multiplied to obtain a preliminary score/ field. Lastly, the sum of each field's score was divided by the number of fields took from the lesion to give the mean score/lesion value. All lesions were independently evaluated by two observers (DOS and PRF), and those ones were differentially scored were reanalyzed together until achieving a consensus. Moreover, due to the small size
Table 1 Antibodies used in immunohistochemical analysis. Antibody Wnt-1 Wnt-3A Gli-3 Shh a b
Source a
Santa Cruz Miliporeb Santa Cruza Santa Cruza
Catalog number
Dilution
6280 09-162 20688 9024
1:200 1:200 1:50 1:100
Santa Cruz Biotechnology, Santa Cruz, CA, USA. Millipore Corporation, Billerica, MA, USA.
334
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
of the lesions, especially dysplastic ones, all fields obtained/lesion were analyzed. 2.4. Statistical analysis Statistical analyses were conducted using GraphPad Prism 6.00 software (GraphPad Software, San Diego, USA). The D'Agostino–Pearson normality test was applied for each test. The Kruskal Wallis test followed by the Dunn post-test was used to compare the expression of each protein in mild, moderate, severe dysplasia and carcinoma from Gal-3+/+ and Gal3−/− mice. To investigate the role of these proteins during 4NQO-induced tongue carcinogenesis, differences in their expression between dysplasia (as a group) and carcinoma were evaluated using the Mann–Whitney U test. Statistical significance was determined at p b 0.05. 3. Results The findings on 4NQO-induced tongue carcinogenesis in Gal3+/+ and Gal3 −/− mice have previously been published elsewhere (de Faria et al., 2011). The immunohistochemical staining of Wnt1 was negative in all dysplasias and carcinomas diagnosed in the tongue of Gal3 +/+ and Gal3−/− mice. The frequency of dysplasias and carcinomas showing positivity for Wnt-3A was 80% and 73.3% respectively in Gal3 +/+ mice and, 69.5% and 66.7% in Gal3 −/− mice respectively. Independently from lesion subtype and mouse genotype, the intensity of Wnt3A expression ranged from mild to strong and was observed in the nucleus, cytoplasm and membrane (Fig. 1A). Regardless of lesion subtype, the mean positivity was always higher in Gal3+/+ than Gal3 −/− mice, although the difference was not significant (Fig. 1B). Likewise, no
significant difference was found in the mean Wnt3A expression during 4NQO-driven tongue transformation in each group of mice (Fig. 1C). In terms of Shh expression, 96.4% and 54.3% of dysplastic lesions and 100% and 40.7% of carcinomas diagnosed in 4NQO-treated Gal3 +/+ and Gal3−/− mice, respectively, were positive for this protein. Overall, the intensity of Shh expression in dysplasias and carcinomas was similar in both genotypes, ranging from mild to strong. Likewise, a cytoplasmic pattern of immunoreactivity was observed in both lesions independently of mouse genotype, with some lesions showing nuclear and rarely membrane expression (Fig. 2A). The mean Shh expression showed a steady increment during 4NQO-induced tongue carcinogenesis in Gal3 +/+ mice, with a significant difference between mild dysplasia and carcinoma (p b 0.001), and between moderate dysplasia and carcinoma (p b 0.01). The mean Shh expression increased to a much lesser extent from mild to severe dysplasia in Gal3 −/− mice, although the difference was not significant. Next, comparison of all 4NQO-induced, Shh-expressing lesions revealed a significant difference between Gal3 +/+ and Gal3 −/− mice (p b 0.0001) (Fig. 2B). Finally, the enhancement of Shh expression during 4NQO-driven tongue tumorigenesis was statistically significant in Gal3+/+ mice (p = 0.0037) but not in Gal3−/− mice (p = 0.74) (Fig. 2C). In terms of Gli-3, the frequency of dysplastic lesions and carcinomas exhibiting positivity was 93.6% and 64.7% in Gal3+/+ mice, and 84.3% and 73.9% in Gal3−/− mice, respectively. Despite more frequently observing Gli3-positive lesions in the latter group, this difference was not statistically significant. As observed for Shh protein, the intensity of Gl3 immunostaining varied from mild to strong in all lesions studied independently of mouse genotype, and with respect to its immunolocalization, it was predominantly seen in the cytoplasm, less frequently in the nucleus and rarely in the plasma membrane (Fig. 3A). The mean Gli3 immunoreactivity showed a gradual reduction from mild dysplasia to
Fig. 1. Wnt-3A expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Wnt-3A intensity in each lesion analyzed (Kruskal–Wallis test, p N 0.05). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in each group of mice (Mann–Whitney test, p N 0.05).
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
335
Fig. 2. Shh expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Shh intensity in each lesion analyzed (Kruskal–Wallis test, p b 0.001). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in Gal3+/+ mice (Mann–Whitney test, p b 0.01) and Gal3−/− mice (Mann–Whitney test, p N 0.05).
carcinoma in Gal3+/+ mice and a slight increment from mild dysplasia to carcinoma in Gal3−/− mice, although the difference was not significant (Fig. 3B). Unlike Shh expression, Gli3 expression during 4NQOinduced tongue tumorigenesis was not significant in either mouse genotype (Fig. 3C). 4. Discussion In this study, we analyzed the potential role of some key regulatory proteins of two important signaling pathways during tongue carcinogenesis in mice that have been implicated in tumorigenesis in many tissues. To sum up, our findings revealed similar mean Wnt3A expression in the setting of tongue malignant transformation in both groups of mice, confirming a previously published report showing aberrant Wnt signaling pathway activity from dysplasia to carcinoma in Gal3 +/+ and Gal3−/− mice. Together, these results indicate that Wnt signaling pathway activation is not dependent on Gal3, particularly during tongue tumorigenesis, at least in mice. At the same time, it was found that Shh protein, but not Gli3 protein, was strongly associated with tongue carcinoma in Gal3+/+ mice, suggesting that Shh protein, the hallmark of the activated Hh pathway, is essential for tongue tumor formation in these mice and might be regulated at a certain level by Gal3. The relationship between the Wnt signaling pathway and Gal3 protein was first described by Shimura et al. (2004), who showed that this lectin acts as a chaperone by helping beta-catenin translocate into the nucleus to enhance Wnt target gene transcription, including that of CCND1 and MYC (Shimura et al., 2004, 2005). Its critical role is currently understood as involving the promotion of tumor development. Giving more insight into this matter, Abdel-Aziz et al. (2008) found a significantly higher incidence of lung tumors in Gal3 +/+ than Gal3 −/− mice and concluded that the absence of Gal3 restrained
lung carcinogenesis by inhibiting the Wnt signaling pathway. In contrast, Eude-Le Parco et al. (2009) showed no influence of Gal3 in the development of breast and colon tumors or in the occurrence of metastasis in LGALS3-silenced mice. Similarly, our previous report found no difference in the incidence of tongue tumor between Gal3 +/+ and Gal3−/− mice, which in part might be associated with equal and persistent Wnt signaling pathway activation in both genotypes (de Faria et al., 2011; Sant'ana et al., 2011). This observation was confirmed herein, because Wnt3 expression, an important activator of this pathway, was similarly expressed in all dysplasia and carcinoma samples from both genotypes. No study using 4NQO-induced tongue tumorigenesis model has evaluated the Wnt3A protein in Gal3−/− mice. In human oral leukoplakia samples, for instance, Wnt3 expression was detected in dysplasia-developing leukoplakias, consistent with the Wnt signalinginduced malignancy (Ishida et al., 2007). Unexpectedly, we found no Wnt1 expression in any lesions investigated despite all control samples being positive for this protein. This finding is at odds with the study by Fracalossi et al. (2010), who identified Wnt1 expression in the majority of dysplasias and carcinomas from rats aged 12–20 weeks, with no difference between the two lesions. Taken as a whole, these findings allow us to conclude that tongue tumorigenesis might be driven by the Wnt signaling pathway through Wnt3 mediation, even in the absence of Gal3. Regarding the Hh signaling pathway players, it was found that Shh immunoreactivity was significantly higher in Gal3+/+ than Gal3−/− mice in both dysplastic and carcinomatous lesions. Furthermore, Gal3+/+developing carcinomas exhibited much higher Shh expression than dysplasias and this finding was strikingly associated with tongue tumorigenesis. Similar results have been observed in gastric and cervical carcinomas when compared with their respective premalignant counterparts (Wang et al., 2006; Xuan et al., 2006). Additionally, recent data have demonstrated that Shh overexpression might be involved in
336
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
Fig. 3. Gli3 expression in tongue lesions from Gal3+/+ and Gal3−/− mice. (A): The pattern of immunostaining in dysplastic lesions and carcinomas of both mice (top to bottom: 100×, 100×, and 400×). (B): Box-plot showing the Gli3 intensity in each lesion analyzed (Kruskal–Wallis test, p N 0.05). (C): The relationship between Wnt-3A expression in dysplasia (as a group) and carcinoma in Gal3+/+ and Gal3−/− mice (Mann–Whitney test, p N 0.05).
small cell carcinoma of lung, adenocarcinoma of the digestive tract, and more recently OSCC (Berman et al., 2003; Wang et al., 2012; Watkins et al., 2003). In a study using OSCC cell lines, it was shown that the activated Hh signaling pathway induced augmented 50-bromo-20deoxyuridine (BrdU) incorporation by tumoral cells and then an increment in the cellular proliferation index (Nishimaki et al., 2004). Also, an established OSCC xenograft mouse model study revealed that nude mice transfected with Shh-inhibited cell lines presented reduced tumor growth when compared with control siRNA-transfected and parental cells (Honami et al., 2012). As a whole, these findings suggest that dysregulated Hh signaling pathway activation might be triggered during the transition from dysplasia to carcinoma in the tongue epithelial lining of Gal3+/+ mice, probably augmenting cellular proliferation. Another protein investigated in the present article was Gli3. This is a member of the Gli family of transcription factors, which together with Gli1 and Gli2, is responsible for triggering the transcription of genes under Hh signaling-dependent activity (Schneider et al., 2010). On the other hand, in the absence of Hh signaling, these proteins are phosphorylated by PKA, GSK-3b and CKI in a sequential manner, which, in turn, targets them for splitting by the ubiquitin–proteosome complex, therefore interrupting their transcriptional activities (Schneider et al., 2010; Tempe et al., 2006). Moreover, the occurrence of such inhibition is also seen in the presence of Hh signaling activity but in a GSK-3bindependent manner (Kise et al., 2009; Tempe et al., 2006). In the present study the level of Gli3 expression showed a slight reduction from dysplasia to carcinoma in Gal3+/+ mice, while an inverse pattern of expression was observed in Gal3 −/− mice. However, the difference was not significant. Furthermore, no association with tongue tumorigenesis was noted in either genotype. At first glance, this finding seems to be contradictory because significant overexpression of Shh
was detected in Gal3 +/+ mouse lesions, which might indicate increased Hh signaling pathway upregulation in this group when compared with Gal3 −/− mice. However, Gli3 expression is not always accompanied by uncontrolled Hh signaling activity and one recent study showed higher Gli3 expression in healthy mucosa than in head and neck carcinoma samples (Dimitrova et al., 2013). It is well known that Gli3 contains C-terminal activation and N-terminal repressor domains, which allows it to have a dual role in the Hh pathway under influence of the Hh ligand and its proteolytic processing after Sufu– GSK-3b-mediated phosphorylation (Amakye et al., 2013; Briscoe and Therond, 2013; Kise et al., 2009). Given a recent report revealing high inactive GSK-3b expression in Gal3+/+ mouse lesions, one might speculate that it could have influenced Gli3 expression (Mendonca et al., 2012). However, as a predominantly cytoplasmic rather than nuclear pattern of expression was observed in all samples studied, this suggests an additional control mechanism regulating the level of Gli3 in this setting. A recent study showed that Sufu, a well-known negative regulator of Hh signaling, is needed to efficiently promote the Gli3–GSK-3b interaction and then its phosphorylation and partial processing (Kise et al., 2009). Another study observed that Sufu, which also tethers Gli proteins in the cytoplasm, interacts physically with Gal3, preventing it from accumulating in the nucleus (Paces-Fessy et al., 2004). Whether the Sufu–Gal3 interaction is essential in the regulation of Gli protein activities, including Gli3, and whether it also occurs in the context of tongue carcinogenesis in these mice remains an open question of great importance for future research. In conclusion, our results show that Shh protein is frequently expressed in Gal3+/+ mouse lesions and is strongly associated with tongue tumorigenesis, establishing that the Hh signaling pathway plays an essential role in this context. Additionally, we confirmed our
D. de Oliveira Santos et al. / Experimental and Molecular Pathology 97 (2014) 332–337
previous report that Wnt signaling activation is not influenced by the Gal3 protein and seems to occur in both genotypes during malignant transformation through the Wnt3A signal. However, further study will be required to determine how Gal3 itself might lead to Hh signaling activation in the setting of tongue malignant transformation in Gal3+/+ mice. Conflict of interest statement All authors disclose any financial and personal relationships with other people or organizations that could inappropriately influence (bias) their work. We denied any potential conflicts of interest include employment, consultancies, stock ownership, honoraria, paid expert testimony, patent applications/registrations, and grants or other funding. Acknowledgments The authors would like to acknowledge the Research Supporting Foundation of Minas Gerais (FAPEMIG-CDS-APQ-00397-09) for the financial support. References Abdel-Aziz, H.O., et al., 2008. Targeted disruption of the galectin-3 gene results in decreased susceptibility to NNK-induced lung tumorigenesis: an oligonucleotide microarray study. J. Cancer Res. Clin. Oncol. 134, 777–788. Amakye, D., et al., 2013. Unraveling the therapeutic potential of the Hedgehog pathway in cancer. Nat. Med. 19, 1410–1422. Beachy, P.A., et al., 2004. Tissue repair and stem cell renewal in carcinogenesis. Nature 432, 324–331. Berman, D.M., et al., 2003. Widespread requirement for Hedgehog ligand stimulation in growth of digestive tract tumours. Nature 425, 846–851. Briscoe, J., Therond, P.P., 2013. The mechanisms of Hedgehog signalling and its roles in development and disease. Nat. Rev. Mol. Cell Biol. 14, 416–429. Cardesa, A., et al., 2005. Tumors of the hypopharynx, larynx and trachea. In: Barnes, L., et al. (Eds.), World Health Organization Classification of Tumors. Pathology and Genetics of Head and Neck Tumors. IARC, Lyon, pp. 118–121. da Silva, S.D., et al., 2011. Advances and applications of oral cancer basic research. Oral Oncol. 47, 783–791. de Faria, P.R., et al., 2011. Absence of galectin-3 does not affect the development of experimental tongue carcinomas in mice. Exp. Mol. Pathol. 90, 189–193. Dimitrova, K., et al., 2013. Overexpression of the Hedgehog signalling pathway in head and neck squamous cell carcinoma. Onkologie 36, 279–286. Eude-Le Parco, I., et al., 2009. Genetic assessment of the importance of galectin-3 in cancer initiation, progression, and dissemination in mice. Glycobiology 19, 68–75. Fracalossi, A.C., et al., 2010. Wnt/beta-catenin signalling pathway following rat tongue carcinogenesis induced by 4-nitroquinoline 1-oxide. Exp. Mol. Pathol. 88, 176–183. Giles, R.H., et al., 2003. Caught up in a Wnt storm: Wnt signaling in cancer. Biochim. Biophys. Acta 1653, 1–24. Hagen, T., Vidal-Puig, A., 2002. Characterisation of the phosphorylation of beta-catenin at the GSK-3 priming site Ser45. Biochem. Biophys. Res. Commun. 294, 324–328. Honami, T., et al., 2012. Sonic hedgehog signaling promotes growth of oral squamous cell carcinoma cells associated with bone destruction. Oral Oncol. 48, 49–55. Hsu, D.K., et al., 2000. Targeted disruption of the galectin-3 gene results in attenuated peritoneal inflammatory responses. Am. J. Pathol. 156, 1073–1083. Ishida, K., et al., 2007. Nuclear localization of beta-catenin involved in precancerous change in oral leukoplakia. Mol. Cancer 6, 62.
337
Kanojia, D., Vaidya, M.M., 2006. 4-Nitroquinoline-1-oxide induced experimental oral carcinogenesis. Oral Oncol. 42, 655–667. Kise, Y., et al., 2009. Sufu recruits GSK3beta for efficient processing of Gli3. Biochem. Biophys. Res. Commun. 387, 569–574. Liu, F.T., Rabinovich, G.A., 2005. Galectins as modulators of tumour progression. Nat. Rev. Cancer 5, 29–41. Lumerman, H., et al., 1995. Oral epithelial dysplasia and the development of invasive squamous cell carcinoma. Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 79, 321–329. Markowska, A.I., et al., 2010. Galectin-3 is an important mediator of VEGF- and bFGFmediated angiogenic response. J. Exp. Med. 207, 1981–1993. Mendonca, D.F., et al., 2012. The inactive form of glycogen synthase kinase-3beta is associated with the development of carcinomas in galectin-3 wild-type mice, but not in galectin-3-deficient mice. Int. J. Clin. Exp. Pathol. 5, 547–554. Mill, P., et al., 2005. Shh controls epithelial proliferation via independent pathways that converge on N-Myc. Dev. Cell 9, 293–303. Mimeault, M., Batra, S.K., 2010. Frequent deregulations in the hedgehog signaling network and cross-talks with the epidermal growth factor receptor pathway involved in cancer progression and targeted therapies. Pharmacol. Rev. 62, 497–524. Nishimaki, H., et al., 2004. A role of activated Sonic hedgehog signaling for the cellular proliferation of oral squamous cell carcinoma cell line. Biochem. Biophys. Res. Commun. 314, 313–320. Paces-Fessy, M., et al., 2004. The negative regulator of Gli, suppressor of fused (Sufu), interacts with SAP18, galectin3 and other nuclear proteins. Biochem. J. 378, 353–362. Perez-Sayans, M., et al., 2012. The role of the adenomatous polyposis coli (APC) in oral squamous cell carcinoma. Oral Oncol. 48, 56–60. Reibel, J., 2003. Prognosis of oral pre-malignant lesions: significance of clinical, histopathological, and molecular biological characteristics. Crit. Rev. Oral Biol. Med. 14, 47–62. Ruiz i Altaba, A., 1999. Gli proteins and Hedgehog signaling: development and cancer. Trends Genet. 15, 418–425. Sant'ana, J.M., et al., 2011. Activation of the Wnt/beta-catenin signaling pathway during oral carcinogenesis process is not influenced by the absence of galectin-3 in mice. Anticancer Res. 31, 2805–2811. Schneider, F.T., et al., 2010. Sonic hedgehog acts as a negative regulator of {beta}-catenin signaling in the adult tongue epithelium. Am. J. Pathol. 177, 404–414. Shimura, T., et al., 2004. Galectin-3, a novel binding partner of beta-catenin. Cancer Res. 64, 6363–6367. Shimura, T., et al., 2005. Implication of galectin-3 in Wnt signaling. Cancer Res. 65, 3535–3537. Sinicrope, F.A., et al., 1995. bcl-2 and p53 oncoprotein expression during colorectal tumorigenesis. Cancer Res. 55, 237–241. Song, S., et al., 2009. Galectin-3 mediates nuclear beta-catenin accumulation and Wnt signaling in human colon cancer cells by regulation of glycogen synthase kinase3beta activity. Cancer Res. 69, 1343–1349. Takenaka, Y., et al., 2004. Galectin-3 and metastasis. Glycoconj. J. 19, 543–549. Tang, X.H., et al., 2004. Oral cavity and esophageal carcinogenesis modeled in carcinogentreated mice. Clin. Cancer Res. 10, 301–313. Tang, J.Y., et al., 2007. Novel Hedgehog pathway targets against basal cell carcinoma. Toxicol. Appl. Pharmacol. 224, 257–264. Tempe, D., et al., 2006. Multisite protein kinase A and glycogen synthase kinase 3beta phosphorylation leads to Gli3 ubiquitination by SCFbetaTrCP. Mol. Cell. Biol. 26, 4316–4326. Wang, L.H., et al., 2006. Increased expression of sonic hedgehog and altered methylation of its promoter region in gastric cancer and its related lesions. Mod. Pathol. 19, 675–683. Wang, Y.F., et al., 2012. Expression of hedgehog signaling molecules as a prognostic indicator of oral squamous cell carcinoma. Head Neck 34, 1556–1561. Watkins, D.N., et al., 2003. Hedgehog signaling: progenitor phenotype in small-cell lung cancer. Cell Cycle 2, 196–198. Xuan, Y.H., et al., 2006. Enhanced expression of hedgehog signaling molecules in squamous cell carcinoma of uterine cervix and its precursor lesions. Mod. Pathol. 19, 1139–1147.
