Immunohistochemical expression of epithelial mesenchymal transition proteins ZEB1 and E-cadherin in oral leukoplakia and oral squamous cell carcinoma

Abstract

BACKGROUND

Oral squamous cell carcinoma (OSCC) is a frequent cancer affecting the oral cavity. OSCC is usually preceded by occurrence of precursor lesions, which demonstrate a varying degree of malignant transformation potential. Epithelial-mesenchymal transition is a significant biological phenomenon that facilitates tumor growth and metastasis through the involvement of diverse epithelial and mesenchymal proteins. E-cadherin down regulation and over-expression of zinc-finger E-box-binding homeobox 1 (ZEB1) has been reported in several cancers.

AIM

To evaluate the role of E-cadherin and ZEB1 in oral leukoplakia and OSCC.

METHODS

A total of 60 cases i.e., oral leukoplakia/oral epithelial dysplasia (OED) (n = 30) and OSCC (n = 30) were included. Immunohistochemistry was performed utilizing two markers: E-cadherin and ZEB1. The Mann-Whitney U test was employed to compare individual markers across the study groups. Spearman’s correlation was done followed by discriminant function analysis.

RESULTS

Reduction in the expression E-cadherin and altered localization were noted from OED to OSCC. Overexpression of ZEB1 along with cytoplasmic accumulation was noted in OED, with marked expression in OSCC. ZEB1 epithelium intensity and ZEB1 connective tissue percentage contributed to the discriminant function with an accuracy of 85% for the classification of OED from OSCC.

CONCLUSION

Loss of E-cadherin and up-regulation of ZEB1 from OED to OSCC, predisposes to induction of epithelial-mesenchymal transition. Discriminant formulas are developed to classify cases of OED and OSCC with 85% accuracy.

Key Words: Oral squamous cell carcinoma; Oral leukoplakia; Oral epithelial dysplasia; Epithelial mesenchymal transition; E-cadherin; Zinc-finger E-box-binding homeobox 1

Core Tip: The study demonstrates the presence of the epithelial-mesenchymal transition (EMT) phenomenon in oral premalignant and malignant lesions suggesting it to be an early change in carcinogenesis. The loss of E-cadherin and up-regulation of zinc-finger E-box-binding homeobox 1 from oral epithelial dysplasia to oral squamous cell carcinoma corresponds to its role in the induction of EMT and correlated to regional lymph node metastasis, thus promoting tumor progression, invasion, and metastasis. Discriminant formulas have been established to ascertain the malignant potential and to categorize instances of oral epithelial dysplasia and oral squamous cell carcinoma based on the expression of EMT markers which is not previously reported.

INTRODUCTION

Oral squamous cell carcinoma (OSCC) accounts for the greatest frequency among all the oral cavity cancers, amounting to 2%-4%, and is the 2^nd most common cancer in India[1,2]. It is characterized by a substantial level of local aggressiveness, invasion, and regional lymph node involvement, as well as a high fatality rate[1,2]. In South Asian countries, the use of smokeless tobacco and areca nut chewing remain the chief risk factors associated with OSCC[3]. Regardless of significant advances in treatment, the 5-year overall survival is less than 50%, and management is associated with severe functional and cosmetic defects associated with substantial morbidity[2].

OSCC is usually preceded by the occurrence of potentially malignant lesions, which demonstrate a varying degree of malignant transformation potential[3]. Timely identification and treatment of potentially malignant lesions is essential to improve patient quality of life and prolong survival duration[4]. Oral leukoplakia is a potentially malignant disorder of the oral mucosa[5]. The histological characteristics of oral leukoplakia manifest as varying degrees of oral epithelial dysplasia (OED) and cancer[6]. OED reflects histological changes characterized by the loss of uniformity or of the architecture of the epithelial cells[6].

The multi-step process of cancer growth is marked by the accumulation of genetic flaws and mutations that are reflected in changes observed at the molecular level[7]. These molecular changes are evident even before any clinical or histological abnormalities occur. Recognizing these changes may enable early detection of high-risk lesions and predict disease aggressiveness and clinical course[7]. Epithelial-mesenchymal transition (EMT) is a multi-step morphogenetic process that enables cancer cells to separate from the main primary tumor and traverse to distant organs, resulting in metastasis. It is a crucial phase during tumor invasion and metastatic cascade[8]. The characteristic of EMT is the loss of cell polarity and cell-cell adhesion of epithelial cells, which results in the development of migratory and invasive capacities[8,9]. Therefore, the loss of epithelial cell junction proteins during EMT causes the non-motile epithelial cells to become motile mesenchymal ones, making them more susceptible to metastasis[8,9]. SNAIL, TWIST, SLUG, Zinc-finger E-box-binding homeobox 1 (ZEB1), ZEB2, and other transcription factors are among those that control EMT[9].

A key step in EMT is the loss of E-cadherin, an essential “cell adhesion molecule” that maintains epithelial integrity[10,11]. E-cadherin repression has been linked to enhanced tumor migration and invasion[10]. Many transcription factors influence E-cadherin expression, including ZEB1/2, SIP1, SNAIL, SNUG, and TWIST[10,11]. These molecules bind to the E-box domain, thus inhibiting E-cadherin[11].

ZEB1 is a “transcription factor” that belongs to the zinc finger E-box binding homeobox family. This molecule enhances tumor invasion and metastasis by activating EMT[12-14]. The regulation of target genes through its protein-binding domains, particularly that of E-cadherin, is one of the ways ZEB1 promotes EMT[13-15]. Thus, ZEB1 plays a crucial part in E-cadherin down-regulation[16]. Many human cancers, including uterine cancer, pancreatic cancer, osteosarcoma, lung cancer, breast cancer, etc., exhibit aberrant ZEB1 expression[16].

E-cadherin downregulation and overexpression of ZEB1 have been reported in several cancers, including colorectal carcinoma, lung carcinoma, esophageal carcinoma, breast carcinoma, etc.[16]. This has been associated with poorer prognosis in terms of decreased survival rates and regional as well as distant metastasis[16]. Currently, there are just a few studies examining the clinical importance or link between ZEB1 and E-cadherin in OSCC, and no research has been conducted on potentially malignant disorders of the oral cavity. Thus, the present study aimed to evaluate the expression and correlation of EMT markers E-cadherin and ZEB1 in OED and OSCC.

MATERIALS AND METHODS

Ethical clearance and sample

Following institutional ethics committee permission, 60 paraffin-embedded tissue blocks of clinically and histologically confirmed cases of OSCC and OED were taken from the department’s archives for the retrospective investigation. The study was reviewed and approved by the Karanataka Lingayat Education, Vishwanath Katti, Institute of Dental Sceinecs Institutional Review Board, approval No. 1321.

Laboratory materials

From each block, three 4 μm tissue slices were cut and placed onto slides coated with “aminopropyltriethoxysilane”. Hematoxylin and eosin were used to stain one slide, and anti-ZEB1 antibody (1:100 Thermo Fischer Scientific United States) and anti-E-cadherin antibody (pre-diluted PathnSitu) were used to stain the remaining slides immunohistochemically. The PolyExcel HRP/DAB Detection System Two-Step Universal Kit (PathnSitu Catalogue no. #PEH002/ United States) was used for visualization.

Demographic data and histological analysis

The clinical data for the cases was gathered and organized from the archival registers. The Hematoxylin and eosin-stained slides of OED slides were graded as per the World Health Organization 2017 classification, and OSCC cases were classified according to Bryne’s classification. “Intensity (mild, intense), localization (membrane, cytoplasm, nuclear), and percentage of staining positivity (0: Absent, 1: 1%-25%, 2: 25%-50%, 3: > 50%)” were the three primary criteria used to examine all of the slides. For E-cadherin, the evaluation was only in the epithelium. Since ZEB1 expression was in both epithelium and connective tissue, the evaluation was done in both the tissues. Controls for the immunohistochemical staining in each batch were used as per the manufacturer’s instructions, i.e., normal oral epithelium for E-cadherin and high-grade ductal carcinoma of the breast for ZEB1.

Statistical analyses

The Mann-Whitney U test was used to evaluate the variations in ZEB1 and E-cadherin expression among the study groups. The relationship between ZEB1 immuno-expression and E-cadherin was evaluated using Spearman correlation. Univariate analysis for difference in expression of E-cadherin and ZEB1 in OED and OSCC was assessed using the χ² test. Discriminant functional analysis for discriminations of two groups (OED and OSCC) was done, and classification accuracy was evaluated.

RESULTS

Expression of E-cadherin in OED and OSCC

Intensity: In OED, 15 cases had mild staining, while 15 cases demonstrated strong staining. In OSCC, 10 cases showed mild and 19 cases showed intense staining. E-cadherin intensity was almost similar in OSCC and OED (P = 0.4420).

Location: OED cases showed E-cadherin expression predominantly located in the membrane (n = 24). In OSCC, most of the cases showed membranous with cytoplasmic location (n = 25). A statistically significant change of location from membranous to cytoplasm in OSCC as compared to OED was evident (P = 0.0001).

Percentage: In the OED, the majority of cases (n = 24) exhibited over 50% positivity for E-cadherin, whereas in OSCC, 15 cases shown over 50%, 9 instances displayed 25%-50%, and 5 cases indicated 1%-25% positivity. There was a significant reduction in the percentage of positivity of E-cadherin expression in OSCC as compared to OED (P = 0.0315) (Table 1, Figures 1 and 2).

Figure 1 Immunoexpression of E cadherin and zinc-finger E-box-binding homeobox 1 in oral epithelial dysplasia. A: Immuno-expression of E-cadherin is intense and membranous in mild dysplasia [immunohistochemistry (IHC), 10 ×]; B: Immuno-expression of E-cadherin is intense and membranous with focal areas of cytoplasmic relocation in moderate dysplasia (IHC, 10 ×); C: Immuno-expression of E-cadherin is mild and membranous with focal areas of cytoplasmic relocation in severe dysplasia (IHC, 10 ×); D: Immuno-expression of zinc-finger E-box-binding homeobox 1 (ZEB1) is intense and nuclear in mild dysplasia (IHC, 10 ×); E: Immuno-expression of ZEB1 is intense and nuclear as well as cytoplasmic in moderate dysplasia (IHC, 10 ×); F: Immuno-expression of ZEB1 is intense and nuclear as well as cytoplasmic in severe dysplasia (IHC, 10 ×).

Figure 2 Immuno-expression of E cadherin in oral squamous cell carcinoma. A: Immuno-expression of E-cadherin is reduced with cytoplasmic relocation in oral squamous cell carcinoma (OSCC) [immunohistochemistry (IHC), 4 ×]; B: Immuno-expression of E-cadherin is reduced with cytoplasmic relocation in OSCC (IHC, 10 ×); C: Immuno-expression of E-cadherin is reduced with cytoplasmic relocation in OSCC (IHC 40 ×).

Table 1 Comparison of E-cadherin expression with respect to intensity, location and percentage scores among the major study groups (oral epithelial dysplasia and oral squamous cell carcinoma) by Mann Whitney U test.

E-cadherin expression	Intensity				Location						Percentage of positivity
	0	1	2	Value	0	1	2	3	4	Value	0	1	2	3	Value
OED (n = 30)	0	15	15	-	0	24	0	6	0	-	0	1	5	24	H = 17.898
OSCC (n = 30)	1	10	19	P = 0.4420	1	4	0	25	0	P = 0.0001b	1	5	9	15	P = 0.0315a

^aP < 0.05.

^bP < 0.01.

Intensity: 0: Negative, 1: Mild, 2: Moderate. Location: 0: Absent, 1: Membrane, 2: Membrane + cytoplasmic, 3: Cytoplasm. Percentage: 0: Absent, 1: 1%-25%, 2: 25%-50%, 3: > 50%. OED: Oral epithelial dysplasia; OSCC: Oral squamous cell carcinoma.

Expression of ZEB1 in epithelium of OED and OSCC

Intensity: In the OED, 21 cases exhibited mild expression, whereas 9 cases demonstrated severe intensity. Eleven cases of OSCC had intense ZEB1 staining in the epithelium, while 19 cases had mild staining. There was no statistically significant difference noted (P = 0.6627).

Location: OED to OSCC showed a change in site from nuclear to mostly cytoplasmic expression, however this change was not statistically significant (P value = 0.773).

Percentage: Eighteen OED cases had > 50% ZEB 1 positivity in the epithelium, ten had 25%-50% positivity, and two had 1%-25%. In OSCC, the majority of the cases (n = 23) showed > 50%, 6 cases 25%-50%, and 1 case 1%-25% positivity. Overall, the percentage of positivity from OED to OSCC increased significantly in both groups (Table 2, Figures 1 and 3).

Figure 3 Immunoexpression of zinc-finger E-box-binding homeobox 1 in oral squamous cell carcinoma. A: Immuno-expression of zinc-finger E-box-binding homeobox 1 (ZEB1) is intense and shows cytoplasmic expression in oral squamous cell carcinoma (OSCC) [immunohistochemistry (IHC) 4 ×]; B: Immuno-expression of ZEB1 is intense and shows cytoplasmic expression in OSCC (IHC 10 ×); C: Immuno-expression of ZEB1 is intense and shows cytoplasmic expression in OSCC (IHC 40 ×).

Table 2 Comparison of zinc-finger E-box-binding homeobox 1 expression in epithelium with respect to intensity, location and percentage scores among the major study groups (oral epithelial dysplasia and oral squamous cell carcinoma) by Kruskal Wallis analysis of variance.

ZEB1 expression	Intensity				Location						Percentage of positivity
	0	1	2	Value	0	1	2	3	4	Value	0	1	2	3	Value
OED (n = 30)	0	21	9	-	0	0	22	4	4	-	0	2	10	18	-
OSCC (n = 30)	0	19	11	P = 0.6627	0	0	30	0	0	P = 0.7730	0	1	6	23	P = 0.2643

Intensity: 0: Negative, 1: Mild, 2: Moderate. Location: 0: Absent, 1: Membrane, 2: Cytoplasm + nucleus, 3: Cytoplasm. Percentage: 0: Absent, 1: 1%-25%, 2: 25%-50%, 3: > 50%. OED: Oral epithelial dysplasia; OSCC: Oral squamous cell carcinoma; ZEB1: Zinc-finger E-box-binding homeobox 1.

ZEB1 expression in connective tissue OSCC and OED

Intensity: The majority of cases (n = 28) in OED showed intense staining. However, in OSCC, only 11 cases showed intense staining, while the rest (n = 18) showed mild expression. A progressive reduction in intensity of stromal expression from OED to OSCC was evident (P = 0.0038).

Percentage: Twenty-three OED instances had > 50% positivity, six had 25%-50% positivity, and one had 1%-25%. However, in OSCC, the majority (20 cases) had decreased expression (1%-25% positive), while only 5 cases had > 50% and 4 cases had 25%-50%. A significant difference in the percentage of positivity was seen, with maximum stromal expression observed in OED and a marked reduction seen in OSCC (P = 0.0001) (Table 3, Figure 3).

Table 3 Comparison of zinc-finger E-box-binding homeobox 1 expression in connective tissue with respect to intensity and percentage scores among two study groups (oral epithelial dysplasia and oral squamous cell carcinoma) by Kruskal Wallis analysis of variance.

ZEB1 expression in connective tissue	Intensity				Percentage of positivity
	0	1	2	Value	0	1	2	3	Value
OED (n = 30)	0	2	28	-	0	1	6	23	-
OSCC (n = 30)	1	18	11	P = 0.0038a	1	20	4	5	P = 0.0001b

^aP < 0.05.

^bP < 0.01.

Intensity: 0: Negative, 1: Mild, 2: Moderate. Location: 0: Absent, 1: Membrane, 2: Cytoplasm + nucleus, 3: Cytoplasm. Percentage: 0: Absent, 1: 1%-25%, 2: 25%-50%, 3: > 50%. OED: Oral epithelial dysplasia; OSCC: Oral squamous cell carcinoma; ZEB1: Zinc-finger E-box-binding homeobox 1.

Correlation between intensity, location, and percentages between E-cad and ZEB1 in OED and OSCC by Spearman’s rank correlation

Regarding the intensity of expression in both groups, a significant positive correlation (r = 0.2195) was found (P = 0.0376); that is, an increase or decrease in E-cadherin intensity was correlated with an associated increase or decrease in ZEB1 intensity. E-cadherin membrane expression and ZEB1 cytoplasmic localization were shown to be negatively correlated (r = -0.1861), indicating that the two markers have an inverse connection with regard to their cellular localization. There was an adverse relationship (r = -0.0751) between the percentage of positivity for ZEB1 and E-cadherin; that is, when the percentage of positivity for E-cadherin decreased, ZEB1 expression increased.

Univariate comparison of different immunohistochemical parameters (variables) in OED and OSCC groups

Five of the eight factors that were examined indicated a significant difference between the OSCC and OED groups. ZEB1 epithelial location (mostly cytoplasmic expression was noted in OSCC as compared to nuclear and cytoplasmic expression in OED), ZEB1 connective tissue intensity (intense expression in OED as compared to OSCC), ZEB1 CT percentage (increased expression in OED as compared to OSCC), and E-cadherin location (cytoplasmic location in OSCC as compared with membranous expression in OED) were all significantly different between the OED and OSCC groups (χ² test, Table 4).

Table 4 Univariate comparison of different immunohistochemical parameters (variables) in oral epithelial dysplasia and oral squamous cell carcinoma groups (χ² test).

Variables	OED	%	OSCC	%	Total	%	χ2	P value
E-Cad intensity	18	60.00	19	63.33	37	61.67	0.0710	0.7910
E-Cad location	8	26.67	25	83.33	33	55.00	19.4610	0.0001b
E-Cad percentage	25	83.33	15	50.00	40	66.67	7.5000	0.0060b
ZEB1 EPI intensity	9	30.00	11	36.67	20	33.33	0.3000	0.5840
ZEB1 EPI location	8	26.67	1	3.33	9	15.00	9.2310	0.0020b
ZEB1 EPI percentage	18	60.00	23	76.67	41	68.33	1.9260	0.1650
ZEB1 CT intensity	28	93.33	15	50.00	43	71.67	13.8710	0.0001b
ZEB1 CT percentage	23	76.67	6	20.00	29	48.33	19.2880	0.0001b

^bP < 0.01.

ZEB1: Zinc-finger E-box-binding homeobox 1; EPI: Epithelium; CT: Connective tissue; OED: Oral epithelial dysplasia; OSCC: Oral squamous cell carcinoma.

Discriminant function analysis for OED and OSCC

Considering that several variables showed differences in the two groups, a stepwise discriminant analysis was planned to study the efficacy of these variables for classification of the cases in OED and OSCC. Only 3 variables entered into the analysis, i.e., E-cadherin location, ZEB1 CT intensity, and ZEB1 CT percentage represented the best discriminators (Table 5). The classification matrix demonstrated 85% accuracy in correctly classifying a case as either OED or OSCC based on these three parameters (Table 6).

Table 5 Forward step wise discriminant function analysis for oral epithelial dysplasia and oral squamous cell carcinoma.

Variables	Wilks’ Lambda	Partial Lambda	F-remove (1, 56)	P level
E-cadherin location	0.5762	0.7532	18.3489	0.0001a
ZEB1 CT intensity	0.5607	0.7740	16.3487	0.0002a
ZEB1 CT intensity	0.4894	0.8869	7.1442	0.0098a
Step 3, N of vars in model: 3; grouping: Groups (2 groups)
Wilks’ Lambda: 0.43402 approx. F (3, 56) = 24.342 P < 0.0000
Variables	Un-standardized coefficients	Standardized coefficients	Sectioning points
E-cadherin location	-1.5935	-0.6628	-1.1228	1.1228
ZEB1 CT intensity	1.5171	0.6352	-
ZEB1 CT intensity	1.1142	0.4478	-
Constant	-0.6553	-	-
Eigen value	1.3040	1.3040	-

^aP < 0.05.

ZEB1: Zinc-finger E-box-binding homeobox 1; CT: Connective tissue.

Table 6 Demonstrates the classification accuracy using the discriminant function.

Predicted classifications	Observed classifications
	Percentage	OED group	OSCC group
OED group	86.7	26	4
OSCC group	83.3	5	25
Total	85.0	31	29

OED: Oral epithelial dysplasia; OSCC: Oral squamous cell carcinoma.

Association of lymph node metastases with clinicopathologic parameters as well as E-cadherin and ZEB1 protein expression (data not shown)

ZEB1 epithelial percentage and ZEB1 connective tissue intensity were the only two parameters that showed a significant association with lymph node metastasis in OSCC in this study.

DISCUSSION

EMT is a crucial molecular process that plays a vital role in OSCC progression (EMT)[16,17]. This is a complex process in which epithelial cells relinquish most of their morphological characters while obtaining mesenchymal characteristics[17]. EMT is known to be induced by numerous markers, viz., epithelial markers, mesenchymal markers, and transcription factors[18,19]. Epithelial markers are E-cadherin, occludin, claudins, plakoglobin, desmoplakin, desmoglein, and cytokeratins, and reduction of these markers is noted in EMT[17]. Mesenchymal markers include N-cadherin, fibronectin, alpha-smooth muscle actin, vimentin, and matrix metalloproteinases, and these are known to be overexpressed in EMT[18]. Various transcription factors are also involved, such as TWIST1/2, SNAIL1/2, ZEB1/2, and SLUG. Among all these markers, E-cadherin is the best known and most reliable marker in EMT[18,19]. Reduction in its expression is considered the earliest change suggesting loss of epithelial phenotype. It is influenced by various transcription factors[18,19]. ZEB1 is the master transcription regulator factor that has been associated with most of the signaling pathways, including the Wnt pathway, phosphatidylinositol 3-kinase/protein kiinase B pathway, transforming growth factor β, Rat sarcoma virus/extracellular signal-regulated kinase, and nuclear factor kappa B pathway in the process of EMT[18-28]. ZEB1 and E-cadherin are interdependent on each other in the process of EMT. Thus, using immunohistochemistry, the current study sought to assess the expression and association of the EMT markers ZEB1 and E-cadherin in OED and OSCC.

E-cadherin expression

The OED group observed a decrease in intensity and positive% from mild to moderate to severe dysplasia. A switch in location of E-cadherin from membrane to cytoplasmic expression was noted, which was consistent with previous research[29-31]. In advanced OED, the loss of cell adhesion mediated by E-cadherin is correlated to the loss of epithelial architecture[29-31]. Its participation in beta-catenin activation and nuclear translocation could explain its change from membrane to cytoplasmic expression. This will activate signaling pathways such as the Wnt pathway, which will enhance the malignant potential in OED[29-31].

In OSCC, there was a considerable decrease in intensity and proportion of positivity and a shift from membrane to cytoplasmic location consistent with findings from literature[32]. Loss of differentiation and higher grades presented a decrease in E-cadherin expression, which is in line with histological observations of epithelial phenotypic loss. Reduced E-cadherin expression is therefore crucial for the change of OED to OSCC[32]. In OSCC, changes in E-cadherin expression are related to the development of an invasive phenotype and a poor clinical outcome[33].

E-cadherin membrane loss and cytoplasmic translocation occur when the E-cadherin beta-catenin complex is disrupted[33]. Thus, there is increased cytoplasmic free beta-catenin that forms a multi-protein degradation complex with “serine/threonine kinases casein kinase 1 and glycogen synthase kinase 3 beta (GSK3), the adenomatous polyposis coli protein, and the scaffold protein axin”[33,34]. Beta-catenin is directed for proteasomal breakdown by GSK3 after casein kinase 1 phosphorylates N-terminal serine/threonine residues[33,34]. Following Wnt particle attachment to serpentine frizzled receptors, a signaling cascade is initiated, and low-density lipoprotein receptor-related protein 5 and 6 co-receptors send signals to the cytoplasm by blocking GSK3[33,34]. As a result, cytoplasmic catenin escapes degradation and collects within the cytoplasm before translocating to the nucleus, activating the Wnt signaling pathway[33,34].

ZEB1 expression in epithelium

The intensity and percentage of ZEB1 positivity increased in OED epithelium with the increase in grade of dysplasia. This finding is similar to the research done by Ahmed et al[35] in 2017. The expression of ZEB1 was in the basal layer in moderate and severe dysplasia, as well as some cytosolic expression in the spinous layer. The most plausible explanation for such an increase in expression and shift into the cytoplasm is its function in activating target genes involved in gaining the mesenchymal phenotype and promoting malignant transformation.

In OSCC, ZEB1 expression exhibited higher intensity and percentage of positive cells as well as predominant cytoplasmic expression. ZEB1 switches from a predominantly nuclear subcellular site in dysplasia to a heavily cytosolic subcellular location in cancer. Experiments have shown that phosphorylation of ZEB1 by particular kinase pathways causes cytosolic localization[35,36]. Changes in the cancer cells' surrounding environment may be causing this alteration in signaling. ZEB-1 activates mesenchymal differentiation target genes via interacting with the acetyltransferases p300/p300/CBP-associated factor and SMADs[36]. Increased ZEB-1 expression is associated with a worse overall survival rate and recurrence in colorectal, esophageal squamous cell, pancreatic, gastric, and hepatocellular cancers[35]. According to Ahmed et al[35], the presence of ZEB1 in the cytosol of certain cells indicates a molecular mechanism for subcellular distribution regulation. SNAIL and TWIST, two functionally related EMT-transcription factors, are regulated in part by phosphorylation, which affects nuclear localization directly[36].

The increase in the percentage of positivity of ZEB1 expression in OSCC was consistent with various studies reported in the literature[36]. The probable reason for enhanced expression of ZEB1 could be attributed to aberrant activation of EMT, thereby causing relocation and elevated accumulation of ZEB1 in the cytoplasm.

ZEB1 expression in connective tissue

ZEB1 expression in the stroma is analogous to that of immune cells and/or a specific subgroup of fibroblasts. In comparison to OSCC, ZEB1 stromal expression was greater in OED in terms of both intensity and positive percentage. It is unknown what the real cause or process is. The most plausible theory is that it controls the differentiation of stromal fibroblasts. ZEB1 in stromal carcinoma-associated fibroblasts (CAFs), leads to enhanced production and secretion of several paracrine signaling molecules into the surrounding stroma, such as fibroblast growth factor 2, fibroblast growth factor 7, vascular endothelial growth factor A, and interleukin 6.

It has been demonstrated that these paracrine signaling molecules, which are produced from stromal CAFs, bind to certain receptors on the tumor epithelial cells and activate p53, resulting in increased tumor growth and accelerated metastatic development in a variety of cancer types. Therefore, the ZEB1/p53 axis is a crucial stroma-specific signaling route that predisposes to carcinogenesis, and ZEB1 may act as an upstream regulator of p53 in stromal CAF.

The stromal cells with ZEB1 positivity have been reported in colon, bladder, breast, and lung carcinomas[37,38]. It has been hypothesized that ZEB1-dependent paracrine signaling from the stroma may influence E-cadherin regulation in tumor regions. Moreover, ZEB1 may predispose to EMT-associated carcinogenesis due to its role as a transcriptional repressor that negatively regulates E-cadherin expression, as evidenced in OSCC[38,39]. The existence of ZEB1-positive stromal cells implies that ZEB1 is significantly involved in carcinogenesis; nevertheless, the exact mechanism underlying ZEB1 expression in the stroma remains undisclosed.

Possible effects related to ZEB1 and E-cadherin protein expression

The upregulation of ZEB1 and the diminished expression of E-cadherin indicate that the loss of E-cadherin may lead to a reduction in cell-cell adhesion, which is the preliminary event in EMT. The expression of E-cadherin is controlled at several levels, as demonstrated in different types of cancer. “ZEB1, ZEB2, SNAI1 (SNAIL), SNAI2 (SLUG), E12/E47, and TWIST1/2” are examples of nuclear factors or “E-cadherin transcriptional repressors” that limit E-cadherin transcription[39].

Lymph node metastasis in OSCC with association of EMT markers

EMT has been linked to a higher incidence of both local and distant metastases in a number of malignancies[40]. ZEB1 epithelial percentage and ZEB1 connective tissue intensity were the only two parameters that showed a significant association with lymph node metastasis in OSCC in this study. ZEB1 thus inhibits E-cadherin expression, predisposing to the development of metastasis via direct as well as micro-environmental signals.

Discriminant functional analysis

In the current study, discriminant function analysis was performed to describe the clinical relevance of the EMT markers expression. The main discriminators for OED and OSCC, according to the stepwise analysis, were E-cadherin location, ZEB1 CT intensity, and ZEB1 CT percentage, which were used to generate a discriminant function. Based on this, the protein expression of the new case can be added into the equation, and the discriminant function established can confidently categorize it into either OSCC or OED with 85% accuracy. It can also be used to predict the malignant potential in OED. No discriminant functions have been previously reported using these two markers.

CONCLUSION

The present study has shown the presence of the EMT phenomenon in oral premalignant and malignant lesions, which suggests that EMT is an early change and can predispose to carcinogenesis. The study showed loss of E-cadherin and up-regulation of ZEB1 from OED to OSCC, corresponding to its role in induction of EMT and correlated to regional lymph node metastasis, thus promoting tumor progression, invasion, and metastasis. Discriminant formulas have been established to ascertain the malignant potential and classify instances of OED and OSCC based on EMT marker expression, specifically E-cadherin localization, ZEB1 connective tissue intensity, and ZEB1 connective tissue percentage, achieving 85% accuracy.