Differential Expression of Desmocollin2, Desmoglein2, and Plakophilin2 in the Progression of Esophagitis to Esophageal Adenocarcinoma

Background Desmosomes play a key role in intercellular adhesive, but also contribute to tumorigenesis. This study aimed to examine the differential expression of desmocollin2 (DSC2), desmoglein2 (DSG2), and plakophilin2 (PKP2) in the progression of reux esophagitis to esophageal adenocarcinoma (EAC) in the rat model of reux disease established by esophagogastroduodenal anastomosis (EGDA). GAPDH-Reversed primer:

represents the strongest risk factor [8] that has been demonstrated in several population-based studies [6,9] . GERD causes re ux esophagitis, leading to Barrett's esophagus and the dysplasia of the esophagus, which predisposes to esophageal adenocarcinoma [10,11] . The molecular mechanisms underlying such progression of a chronic benign condition to esophageal carcinogenesis remain to be elucidated.
Previously, we presented a functional module-based approach to explore link between in ammation and esophagus cancer based on GEO database [12] . We observed a functional module containing genes Desmocollin2(DSC2), Desmoglein2(DSG2), and Plakophilin2(PKP2), was served as both in ammation and cancer module. DSG2, PKP2 and DSC2 supposed to play a vital role in progression from esophagitis to esophageal cancer. However, the mechanism underlying the effect of the three genes on tumor progression remains unknown, with only a limited number of studies reporting on the role of them in esophageal cancer.
As previously reported, DSG2, PKP2 and DSC2 are major components of cell desmosomes [13,14] in mediating desmosomal cell-cell adhesion in epithelial tissues and cardiac muscle and maintaining epithelial homeostasis. In addition, they play a key role in regulating epithelial cell proliferation and tumorigenesis [15,16] . Desmosomal cadherins have been shown to be downregulated in several cancers to promote tumor progression [17] but upregulated in other cancer types. For instance, in patients with non-small cell lung cancer (NSCLC), the DSG2 expression has been associated with a poorer prognosis, suggesting that overexpression of DSG2 promotes the development and progression of NSCLC [18] . While in another study, downregulation of DSG2 and DSG3 expression has been found in lung cancer cell lines, which might be contributed by DNA metheylation [19] . The composition of desmosomes may be dependent on the tissue-speci c or differentiation-speci c expression of particular isoforms of the constituent proteins [20] . Therefore, the role of desmosomes-adhesion complexes in carcinogenesis remains elusive [14] . However, to date, the changes in the expression of DSC2, DSG2, and PKP2 with their roles in the progression of esophagitis to esophageal adenocarcinoma have not been fully elucidated. Therefore, this study aimed to examine the expression levels of these proteins in esophageal tissues (of esophagitis to esophageal adenocarcinoma) harvested from re ux rat models simulating human gastroesophageal re ux disease using the esophagogastroduodenal anastomosis (EGDA) method.

Animal experiment
This study was approved by the Medical Ethics Committee of the Fourth A liated Hospital of Harbin Medical University. Eight-week-old male Sprague-Dawley rats (Beijing Vital River Laboratory Animal Technology Co., Beijing, China) were housed ve per cage. They were given water and food ad libitum, maintained on a twelve-hour light/dark cycle, and allowed to acclimate for two weeks. Eight-hour prior to surgery, the rats were kept nil per os (NPO). Anesthetics were administered in an acrylic anesthetizing chamber (2% iso urane in 1 ml/L oxygen) [8] . Based on previously described, EGDA was performed in an end-to-side esophagoduodenostomy manner with gastric preservation through a left upper abdominal incision. Brie y, two 1.0 cm incisions were made each on the gastroesophageal junction and the duodenum on the antimesenteric border and then were anastomosed together with precise mucosal-to-mucosal opposition. All surgically treated rats (n = 55) were randomly allocated into two groups: group A rats treated with EGDA only (n = 27); group B treated with EGDA and iron dextran (4 mg Fe/Kg/week, i.p.), starting from 2 weeks after surgery and continuing for the duration of the experiment (n = 28). On the other hand, 40 pseudo-surgery rats were randomly assigned into two groups: Group C received pseudo-surgery only (n = 20); group D received pseudo-surgery and iron dextran (4 mg Fe/Kg/week, i.p.) starting 2 weeks after surgery. The animals were weighed weekly and euthanized at 8 weeks (n = 10 in every group), and the rest were kept for 32 weeks following surgery. The animals were euthanized by using iso urane overdose.

Histopathology
To examine the morphological changes, a longitudinal incision was made in the middle and lower esophagus of rats. The esophagus was xed with 40 g/L paraformaldehyde for 12-24 hours and then embedded in para n. Para n tissue sections were stained with hematoxylin and eosin (H&E) for histopathological analysis.
Based on the established classi cation criteria, the histopathological changes of the rat esophagus were de ned as the following: Type 1. Normal esophagus: esophageal mucosa was a laminated squamous epithelium, where the basal cells were con rmed in size, low in the nucleus/cytoplasm ratio, cells were aligned and good polarity; Type 2. Esophagitis: the presence of in ammatory cell in ltration in the esophageal mucosa and submucosa, which were categorized into mild, moderate, and severe according to the number of in ammatory cells; Type 3. Esophageal dysplasia: basal cells were enlarged, the nucleoplasmic ratio was increased, the cell arrangement was disordered, and the polar orientation was poor; Type 4.  The IHC slides were analyzed by the pathologist. The cellular localization and the staining degree of DSC2, DSG2, PKP2 were assessed, with the expression of these proteins graded according to the intensity of the staining: no staining (-), mild staining (yellow) (+), medium staining (light brown) (++), and strong staining (dark brown) (+++).

Page 5/14
The total liver RNA was isolated from the esophageal tissue by using the E.Z.N.A.®Total RNA Kit I (OMEGA, United States; R6834-1) according to the manufacturer's protocols. The isolated RNA samples were then converted to cDNA using the ReverTra Ace qPCR RT Kit (TOYOBO CO., Osaka, Japan; Code No.FSQ-101). All PCR reactions were performed on the Applied Biosystems QuantStudio™ 3 Real-Time PCR System The differential expression of DSC2, DSG2, and PKP2 among the groups were compared using the Kruskal-Wallis test or Shapiro-Wilk normality test, and the results of indicators between the four groups were analyzed using One-Way ANOVA multiple comparisons. A P-value of < 0.05 was considered statistically signi cant.

Pathological analyses
Most mice (79.17%, 95/120) survived the surgery. The rest died of anesthesia, bleeding or unknown reasons during the surgery. All histopathological features were categorized into four: normal esophagus, esophagitis, dysplasia, and esophageal adenocarcinoma (EAC) (Fig. 1 and Table 1). As outlined in Table 1, there was no dysplasia or EAC observed in all groups at 8 weeks. In group A and group B, 80% and 100% of rats, respectively, had esophagitis type. However, at 32 weeks, 17.65% (3/17)

Immunohistochemical analyses
The positive stainings of DSC2, DSG2, and PKP2 were localized in the esophageal epithelium. In normal esophagus and esophagitis, the staining background of DSC2, DSG2, and PKP2 appeared mild in the cell membrane and/or cytoplasm of the esophageal epithelium ( Fig. 2A, 2B). Compared to that of normal esophagus and esophagitis, the staining was increased slightly in esophageal dysplasia (Fig. 2C). However, strong staining (dark brown) in the cell cytoplasm and/or nucleus was observed in esophageal adenocarcinoma (Fig. 2D), indicating the alterations in the levels of protein expression and locations of DSC2, DSG2, and PKP2 in esophageal carcinogenesis.

Gene expression analyses
The relative expression of the DSC2 gene was the highest in EAC and the lowest in esophagitis. Compared with the normal esophagus, the expression level of DSC2 was signi cantly higher in EAC (P < 0.0001), while no signi cant difference was observed in esophagitis and dysplasia. Additionally, there was a signi cant difference in the DSC2 gene expression between esophagitis and EAC (P < 0.0001) (Fig. 3A). For the DSG2 gene, the relative expression was the highest in EAC with signi cant differences when compared with other groups (P < 0.0001) (Fig. 3B), while no signi cant difference was observed between normal esophagus and esophagitis. Although the DSG2 expression was slightly higher in dysplasia compared with the normal esophagus, this difference was not statistically signi cant. Also, the expression of the PKP2 gene appeared progressively increased from the normal esophagus to EAC, with its expression signi cantly higher in EAC than normal esophagus and esophagitis (P < 0.0001) (Fig. 3C).

Western blot analyses
To validate the above ndings, western blot was performed to detect the differential expression of DSC2, DSG2, and PKP2 protein expression in the normal esophagus, esophagitis, esophageal dysplasia, and EAC.
After normalizing to the expression of GAPDH, the expression of DSC2, DSG2, and PKP2 were signi cantly upregulated in EAC compared with normal esophagus or esophagitis (p < 0.05), which were consistent with the results of qRT-PCR and IHC.

Discussion
Following the description and introduction of the surgically induced re ux model of esophagitis in rats, it has been extensively used to study esophageal carcinogenesis in order to develop novel preventive and treatment strategies [8,[21][22][23] . Among various surgical approaches available, EGDA is currently the most effective surgical method to induce animal re ux models [8,24] . Different from other surgical approaches that encourage acid or alkaline re ux only, EGDA diverts both the gastric contents and some duodenal secretions to the esophagus, which is superior in promoting and simulating the process of human gastroesophageal re ux disease (GERD) from esophagitis to Barret' esophagus (BE), leading to esophageal dysplasia and consequently adenocarcinoma. In our study, the combination of EGDA with frequent iron injection (not as a carcinogenic agent) promoted tumorigenesis while the application of iron alone had a small in uence, indicating that iron dextrin could potentiate further the oxidative stress in the in amed esophageal epithelium [21] . Higher oxidative stress could in turn aggravate the in ammatory reaction, resulting in abnormal cell proliferation and transformation, with eventual esophageal carcinogenesis [3,8,25] .
We reported previously DSG2, DSG2 and PKP2 were differently expressed in both esophagitis and esophageal cancer based on bioinformatics. DSC2 and DSG2 are members of the desmosomal cadherin family. Desmosomes not only contribute to an intercellular adhesive which is essential in the normal organization and stabilization of epithelial tissues, but also participate in cellular differentiation [26] , transformation, and tumorigenesis [16] . Moreover, during malignant transformation of epithelial cells, alterations in the expression and function of intercellular junctions might result in tumor invasion and metastasis. There are four types of intercellular junction presenting in vertebrates: desmosomes (DSMs), adherens junctions (AJs), tight junctions (TJs), and gap junctions (GJs),which are crucial for maintaining epithelial homeostasis [27] . All these complexes work as a single unit, interdependently rather than individually. The formation of DSMs is dependent on classic cadherin-mediated adhesion [28] . E-cadherin is the classical cadherins of adherens junctions. It has been reported that E-cadherin can associate with some Dsgs (DSG2), and this interaction may help initiate early stages of DSM assembly. Gap junctions metabolically and electrically connect the cytoplasm of adjacent cardiomyocytes. Meanwhile, some articles summarized the involvement of gap junctions in carcinogenesis [29] . Connexin 43 (Cx43), an important gap junction protein, has been reported associated with esophageal squamous cell carcinoma [30] . DSMs could contribute to Cx assembly and GJ function. The relationship between the DSM and Cxs was rst studied in the heart, showing that PKP2 is critical for Cx43 expression and function. The above reports suggested that DSMs might impact gap junction. The connection between them will be explored further in the following study. As outlined above, there were tight connection among these intercellular junctions. The aim of our study is to elucidate the mechanism underlying the effect of DSG2, DSC2 and PKP2 on tumor progression in esophageal adenocarcinoma.
Our ndings con rmed the differential expression of DSC2, DSG2, and PKP2 from benign esophagitis to esophageal adenocarcinoma. The upregulation of all these at mRNA expression levels in esophageal adenocarcinoma and the differences when compared to the normal esophagus or benign esophagitis were signi cant. Besides, the analyses of immunohistochemistry and western blot further validated the above observation, in which the protein expression levels of DSC2, DSG2, and PKP2 were all signi cantly elevated in EAC. Furthermore, in addition to quantitative changes of these protein expressions in the development of esophageal adenocarcinoma, immunohistochemical staining also demonstrated changes in their cellular localization. Several studies have described alterations of protein expression leading to structural changes of desmosomes, which can induce cell-transformation phenotype and promote carcinogenesis [19,31,32] .
However, the mechanism involved in the changes of various desmosomal components remains unclear.
Both DSC2 and DSG2 are members of the desmosomal cadherin family. They are transmembrane proteins that maintain intercellular connectivity. The ectodomains of DSC are connected with the DSG at the extracellular region of apposing cells, and the intracellular regions of DSC and DSG directly or indirectly interact with the armadillo proteins including Plakophilins (PKPs), Junction plakoglobin (JUP), Desmoplakin (DSP), Intermediate laments (IFs), and other desmosomal complexes, which mediate intercellular adhesion in epithelial tissues [12,19] and may have a direct or indirect role in regulating cell differentiation [16] and tumorigenesis [14] . JUP, which is an homologous protein withβ-catenin, can replace β-catenin in adherens junctions and stimulate the transcription of WNT target genes, including oncogenic targets such as CCND1 (encoding cyclin D1). So, we inferred that desmosomes dysfunction can promote cancer by WNT-β-catenin signalling pathway indirectly. Additionally, PKP2 has been reported functioned as an intracellular inhibitor of the Wnt/β-catenin pathway in colon cance [52] . We will explore the connection between desmosomal cadherin family and Wnt/β-catenin pathway furtherly.
Through modi ed cell adhesive strength or changes in intracellular and intercellular signaling, the expression patterns of desmosomal cadherin may have been altered, which can affect cell behavior and drive proliferation under some circumstances [33] . In our study, the levels of mRNA and protein expression of DSC2 and DSG2 progressively increased from the epithelial of esophagitis to dysplasia and carcinoma, suggesting an important role of DSC2 and DSG2 overexpression in tumorigenesis of esophageal epithelia. These ndings were consistent with previous studies, in which both the DSC2 and DSG2 were overexpressed in squamous cell cancers of the skin and head and neck (HNSCCs) [34] . This could enhance internalization, modulate extracellular vesicles (Evs) secretion and paracrine signaling, and increase local and distant mitogenic effects that encourage tumor progression [23,35] . Besides, the loss of desmoglein-2 (DSG2) has been associated with decreased epithelial cell proliferation and suppressed xenograft tumor growth in mice [14] . In addition, DSC2-positive urothelial carcinomas (UC) have a more advanced stage disease than those tumors without DSC2 [36] . Meanwhile, other studies have reported that DSG2 overexpression may promote tumorigenesis in basal cell carcinoma (BCC) [37] by activating Stat3 in the basal layer of the skin of mice [38] , which potentiates the proliferation of CRC through up-regulation of PNN and activating EGFR/ERK signaling pathway [39] . Furthermore, DSG2 has been shown to induce and activate urokinase-type plasminogen activator receptor-related signaling cascade and accelerate cutaneous wound healing [40] , which plays a critical role in the vasculature that is independent of its canonical role as a component of desmosomes in a distinct subpopulation of progenitor cells [41] .
PKP2 is a member of the armadillo family generally localized in the nucleus and cytoplasm of epithelial tissues and cardiomyocytes, which binds the desmosomal cadherins and desmoplakin [42,43] . Lately, PKP2 has been implicated in tumorigenesis and/or invasion and metastasis of cancer. For instance, in some soft tumors and highly proliferative colonies of cultured mesenchymal stem cells, the upregulation of PKP2 and its integration into adherent junctions (Ajs) have been observed [44] . PKP2 may excite Top ash activity, leading to the translocation of desmosomes [33] . Another study has revealed that PKP2 can promote actin recombination and regulate the assembly of the desmosomal complex by connecting Ras homolog A and protein kinase C-dependent pathways [45] . In our study, the mRNA and protein levels of PKP2 were signi cantly upregulated in esophageal adenocarcinoma tissues compared with tissues of normal esophagus and esophagitis in rats. Particularly, the expression of PKP2 progressively increased from esophagitis to esophageal adenocarcinoma, and signi cantly elevated in EAC. These ndings were consistent with several previous studies and suggested that PKP2 might be essential for carcinogenesis.
Indeed, reduced expression of PKP2 has been associated with inhibition of glioma cell proliferation and migration [46] . Moreover, decreased PKP2 expression reduces Epidermal growth factor (EGF)-dependent and EGF-independent epidermal growth factor receptor (EGFR) dimerization and phosphorylation, resulting in a signi cant decrease in proliferation and migration of cancer cells and tumor development [47,48] . Besides, PKP2 has been implicated in the invasiveness of bladder cancer as a result of loose adhesion via the epithelial-mesenchymal transition (EMT) and β-catenin-mediated signaling pathways [49] . The redistribution of PKPs could promote adhesion, differentiation, and the localization of PKPs from desmosomes to the nucleus that contribute to tumorigenesis [20] . In a previous study, strong PKP2 immunoreactivity has been observed in 85.7% of metastatic tumors of oropharyngeal carcinoma [50] . Thus, an increase of plakophilin could promote carcinogenesis via the stimulation of translation and proliferation, while the loss of plakophilin may contribute to carcinogenesis and/or metastasis via loss of contact inhibition and increased motility [51] .

Conclusion
In summary, this study has demonstrated that a combination of the EGDA method with frequent iron injection increases the incidence of EAC. Upregulations of DSC2, DSG2, and PKP2 expression have been observed in EAC, suggesting that these genes could be important in the development of EAC. Therefore, DSC2, DSG2, and PKP2 can be used as novel molecular biomarkers for early diagnosis and targeted therapy for esophageal adenocarcinoma. To further shed light on the complete complexities of these proteins in the progression of esophagitis to EAC, additional studies are certainly necessary. Human tissue is going to be Consent for publication: Not applicable.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests:
The authors declare that they have no competing interests. Authors' contributions: NL analyzed and interpreted the data regarding the operated animals and controls. YW performed the histological examination and was a major contributor in writing the manuscript. LZ and WB designed the study. All authors read and approved the nal manuscript.