“Loss of polarity of basal cells” – The term revisited in the light of genomics

Abstract

The concept of polarity in development, homeostasis and pathological alteration of tissues has emerged as an interesting aspect of the concerned biology. The epithelial cells exhibit apicobasal polarity which is maintained by Crumbs complex located at apical region of tight junction, ‘PAR’ complex at sub-apical region of tight junction and Scribble complex at adherens junction. Any functional perturbation of these proteins cause alteration of normal epithelial physiology en-route to epithelial pathology. In this maiden scientific exercise, we have tried to explore the association of expression of cell polarity proteins in OPMD and OSCC. Here, we have chosen DLG1 as a representative of polarity protein. RNA was isolated from the tissue samples. Then cDNA was prepared by RTPCR technique. qPCR was performed on cDNA samples. Expression data was analysed on the basis of Ct values. Paired t-test was performed with normalized Ct values of disease and normal tissue to determine whether there was any significant difference in expression between them. The statistical tests were done using SPSS software. Results of this study reflected increase in DLG1 expression in high grade dysplasia. There was no significant alteration in expression of DLG1 in WDSCC where there is formation of cluster of neoplastic cells ultimately producing epithelial islands and keratin pearls. But, in case of MDSCC, when the same neoplastic cells keep on invading with minimal keratin pearl formation, they again gain the mesenchymal character in full potential. And this phenomenon supports the upregulation of DLG1 in MDSCC in our study.

1. Introduction

The function of any tissue depends on the structural organization of the cells that build up that tissue. This organized arrangement of cells is possible when individual cell exhibits a unique feature called cell polarity. Cell polarity may be defined as the asymmetric organization of different aspects of the cell such as the cell surface, intracellular organelles, the cytoskeleton,¹ intracellular protein distributions and cell functions.² Hence, cell polarity is the reflection of complex mechanisms that maintain specialized domains in the plasma membrane and cytoplasm of a cell which facilitate the organization and functioning of any cell as well as tissue.³

Two types of polarity have been conceptualized in epithelial cells and tissues - apicobasal polarity and planar polarity. Apicobasal polarity refers to a specialized apical membrane domain facing external environment or lumen of internal cavities and a basolateral membrane domain localized at basal and lateral surface of individual epithelial cell. Planar cell polarity is the polarized organization of individual cell within an epithelial tissue.

Apicobasal polarity is maintained by different polarity protein complexes at two different domains of an individual epithelial cell - apical domain and basolateral domain. Apical domain is maintained by PAR complex and basolateral domain is maintained by SCRIBBLE and CRUMBS complex. These polarity protein complexes are closely associated with cell junctions also.

Par3, Par6 and aPKC protein together constitute PAR complex, Crumbs complex include Crb3, PALS1 and PATJ whereas Scribble complex has LGL, DLG1 and Scribble proteins.^{4, 5, 6} Any alteration in expression of polarity protein genes or defect in post-translational modification or mis localization of polarity proteins in the cytosol results in perturbation in cell polarity. This altered polarity will lead to dysplastic epithelium. This epithelial dysplasia may lead to EMT (Epithelial Mesenchymal Transition) and carcinogenesis.

Till date, universally “loss of polarity of basal cells”⁷ is considered as an important histopathological criterion to describe dysplasia, but it can be hypothesized that true loss of polarity of basal cells is not possible, rather “alteration of polarity” should be the criterion to describe dysplasia. However, it is sometimes difficult for the pathologists to assess the altered polarity or so-called “loss of polarity of basal cells” because there are no distinct criteria to describe the polarity alteration/loss except visual interpretation which again varies from one pathologist to other.

So, to make a clear distinction between polarity alteration and loss, it may be ideal to compare the expression of these genes in normal and dysplastic epithelium of oral precancer and cancer. This will not only help for objective evaluation of dysplasia but also guide us in grading of dysplasia.

Review of literature reveals that researchers had explored the expression of different polarity genes in different cancers like Breast, Prostrate, Lung and Ovaries⁸ but no attempt have been made in oral precancer and cancer.

DLG1 gene encoding the DLG1 protein maintains the adherens junction assembly and basolateral domain of the cell by mutual antagonism. Besides these, DLG1 also helps in cell migration. Various studies have revealed that DLG was downregulated in cancers of cervical and oesophageal epithelium. As DLG1 expression alteration is noted in various epithelial cancers, so we have chosen DLG1 in our study. Hence, in oral precancer and oral squamous cell carcinoma (OSCC) it is a maiden attempt to evaluate the expression of DLG1 gene.

So, our objectives were to determine the expression of DLG1 (Discs large 1) gene in epithelial cells of normal oral mucosa, dysplastic oral epithelium in oral precancer (leukoplakia) and OSCC. For this purpose, the expression data was assessed and compared among the normal oral mucosa, dysplastic oral epithelium in leukoplakia and squamous cell carcinoma.

2. Materials and methods

This was a hospital-based study with a sample size of 80 (40 leukoplakia and 40 OSCC). Purposive and incidental type of probability sampling technique was used in this study for selection of sample. Considering the footfall of leukoplakia and OSCC patients as well as infrastructure of the genetic laboratory, such a sample size was chosen for conducting the study. The study was conducted after getting proper approval from the ethics committee. Besides that, informed consent of the patients for using their samples in the study as well as for publication was taken and their identity were kept confidential.

Patients suffering from either leukoplakia or OSCC were included in the study based on the inclusion and exclusion criteria.

Those patients having sufficient exposure to tobacco, areca nut, betel leaf, alcohol and suffering from histopathologically confirmed leukoplakia and OSCC involving buccal mucosa, labial mucosa, floor of the mouth, alveolar ridge and palate were included in the study.

The patients suffering from any infectious or contagious disease and/or any intractable medical or radiological abnormality were not included in the study. Leukoplakia and OSCC involving tongue were excluded as because tongue mucosa represented a different variation in expression of polarity proteins in comparison to other parts of oral mucosa. Verrucous carcinoma patients were discarded because histopathologically the epithelium in verrucous carcinoma shows bland cytology with minimal atypia and lack of invasion. OSCC and leukoplakia patients with history and clinical findings suggestive of oral submucosal fibrosis (OSMF) and/or lichen planus were not included in this study. Any other white lesion like hypertrophic lichen planus, candidiasis etc. were not selected for the study. If the clinically normal tissues of the patients were found to be histopathologically normal with non-descript epithelium, then only those normal tissues were considered for comparative evaluation of polarity genes in normal tissue, leukoplakia and OSCC.

After general medical check-up and getting informed consent, incisional biopsy was performed. During biopsy, both diseased as well as normal tissue samples were collected in both Neutral Buffered Formalin (NBF) and RNA later solution for histopathological study and genetic study respectively.

For histopathological study, NBF fixed biopsied sample were processed, stained with haematoxylin and eosin and finally examined under microscope to confirm the diagnosis.

For genetic study, RNA was isolated from the tissue samples (normal and diseased) by using QIAGEN RNeasy mini kit. The steps of RNA isolation from the tissue samples include - disruption and homogenization of tissues with lysis buffer RLT; centrifugation; RNA binding; RNA washing and DNAase treatment; RNA eluting. RNA concentration and quality was checked by using Nanodrop 2000. Only those RNA samples which had good concentration and quality (minimum concentration 60 ng/μl; A₂₆₀: A₂₈₀ = 2) were included in the study. From those RNA samples, cDNA was prepared by RTPCR technique. Then, qPCR was performed on cDNA samples by using SYBR Green assay.

Expression data was analysed on the basis of Ct values and thereafter fold change was calculated. Statistical analysis with paired t-test was performed with normalized Ct values of disease and normal tissue to determine whether there was any significant difference in expression between them. Normal distribution of the normalized Ct values was confirmed by Kolmogorov-Smirnov test before performing paired t-test. The statistical tests were done using SPSS software.

3. Results and analysis

3.1. Results and analysis in case of leukoplakia samples

In samples of leukoplakia, expression of DLG1 was found to be upregulated in 20 patients (average fold change of 2.35) while it was downregulated in rest 20 (average fold change of 1.73) (Table 1). On the basis of grading of dysplasia (as per WHO classification of head and neck tumours 2017), all the leukoplakia samples were divided into two subtypes – mild/low grade dysplasia and severe/high grade dysplasia. It was observed that 29 leukoplakia samples had mild/low grade dysplasia amongst which expression of DLG1 were upregulated (↑) in 12 samples with average fold change of 2.6 and down-regulated (↓) in 17 samples with average fold change of 1.8. severe/high grade dysplasia was observed in 11 leukoplakia samples amongst which expression of DLG1 were upregulated (↑) in 8 samples with average fold change of 1.96 while downregulated (↓) in 3 samples with average fold change of 1.25.

Table 1.

Demographic details of leukoplakia patients with fold change.

Sample No.	Age in years and Sex		Histological Grading of Dysplasia	Fold Change 2^ΔΔCT	Direction of Fold change
01	63	M	LOW	1.19	↓
02	55	F	LOW	1.69	↓
03	30	M	LOW	1.44	↑
04	45	M	LOW	1.83	↑
05	82	M	LOW	1.53	↓
06	51	F	LOW	1.20	↑
07	34	M	LOW	1.84	↑
08	51	M	LOW	2.94	↑
09	37	F	LOW	4.05	↓
10	42	M	LOW	1.26	↓
11	55	M	LOW	1.37	↓
12	56	M	LOW	2.08	↑
13	36	M	LOW	1.62	↓
14	51	M	LOW	1.69	↑
15	44	M	LOW	1.02	↓
16	37	M	LOW	2.86	↑
17	47	M	LOW	2.56	↑
18	64	M	LOW	1.80	↓
19	51	M	LOW	2.32	↑
20	49	M	LOW	10.25	↑
21	47	M	LOW	3.82	↓
22	45	M	LOW	1.39	↑
23	48	M	LOW	1.59	↓
24	55	M	LOW	3.77	↑
25	52	M	LOW	1.21	↓
26	26	M	LOW	1.14	↑
27	58	M	LOW	3.31	↓
28	47	M	LOW	1.08	↓
29	60	M	LOW	1.13	↓
30	68	M	HIGH	1.64	↑
31	54	M	HIGH	1.42	↑
32	51	M	HIGH	2.08	↑
33	66	M	HIGH	1.07	↑
34	70	F	HIGH	1.30	↓
35	53	M	HIGH	1.49	↑
36	67	M	HIGH	1.18	↓
37	35	M	HIGH	1.93	↑
38	47	M	HIGH	1.72	↓
39	48	M	HIGH	1.36	↓
40	46	M	HIGH	1.52	↓

[ M − Male; F – Female; Y – Yes; ].

According to the gradation of dysplasia, two groups were created for statistical analysis. One of the groups included mild/low grade dysplasia and normal samples while the other group contained severe/high grade dysplasia (Fig. 1) and normal samples.

Fig. 1.

Hematoxylin and Eosin-stained section of moderate (high grade) dysplasia (10 X view).

After carrying out the Kolmogorov Smirnov test it was found that none of the groups showed deviation from normal distribution. This implied that the mean ΔCt of diseased tissue and normal tissue for each of them could now be compared.

As the ΔCt of normalized expression of DLG1 with respect to dysplastic leukoplakia samples are not independent (i.e., collected from same patients), “paired t-test” was carried out to compare their means ΔCt. The output of paired t-test gives us the difference of mean of the two groups followed by their standard deviation and 95% confidence interval along with a P-value for the t -test. The P-value <0.05 implies significant difference in mean ΔCt of two groups of leukoplakia samples.

Expression of DLG1 was not found to be significantly different when paired t-test was performed between leukoplakia and its adjacent normal tissues (P value > 0.05) in all samples. However, when the samples were classified on the basis of histopathological feature (mild/low-grade and severe/high grade), significant difference in expression of DLG1 was observed (paired t-test value 0.017) between disease tissue and its adjacent normal for severe/high grade dysplasia patients but not for mild/low grade dysplasia.

3.2. Results and analysis in case of OSCC samples

In samples of OSCC, expression of DLG1 was upregulated in 18 samples with average fold change of 3.11 while in remaining 22 samples, the expression was downregulated with 1.79 average fold change (Table 2).

Table 2.

Demographic details of OSCC patients with fold change.

Sample No.	Age in years and Sex		Tumour Size	Histological subtype	Fold Change 2^ΔΔCT	Direction of Fold change
01	32	M	T3	WDSCC	1.36	↑
02	55	M	T2	WDSCC	1.78	↑
03	55	F	T2	WDSCC	1.19	↓
04	40	F	T3	WDSCC	1.39	↑
05	45	F	T2	WDSCC	2.28	↓
06	51	F	T2	WDSCC	1.04	↑
07	75	M	T2	WDSCC	1.30	↓
08	50	F	T2	WDSCC	1.03	↑
09	45	M	T2	WDSCC	1.31	↓
10	64	M	T2	WDSCC	2.09	↓
11	30	M	T2	WDSCC	2.03	↓
12	46	M	T2	WDSCC	2.12	↓
13	65	M	T2	WDSCC	3.35	↑
14	50	M	T2	WDSCC	1.25	↓
15	53	M	T2	WDSCC	1.24	↓
16	70	F	T3	WDSCC	3.55	↑
17	54	M	T1	WDSCC	1.56	↑
18	60	F	T2	WDSCC	1.24	↓
19	70	F	T2	WDSCC	1.82	↑
20	68	M	T1	WDSCC	1.23	↑
21	42	M	T2	WDSCC	1.70	↓
22	37	M	T2	WDSCC	1.00	↓
23	45	M	T2	WDSCC	1.32	↓
24	45	F	T2	WDSCC	1.33	↓
25	45	M	T3	WDSCC	1.47	↓
26	60	F	T3	WDSCC	2.35	↑
27	45	M	T2	WDSCC	3.32	↓
28	48	F	T2	WDSCC	2.66	↓
29	65	M	T2	WDSCC	4.40	↓
30	63	M	T2	WDSCC	1.08	↑
31	50	F	T2	WDSCC	1.40	↑
32	66	F	T1	WDSCC	18.94	↑
33	35	M	T1	WDSCC	1.73	↓
34	58	F	T2	WDSCC	2.34	↑
35	60	F	T2	MDSCC	8.40	↑
36	55	M	T3	MDSCC	2.24	↑
37	45	F	T2	MDSCC	1.22	↓
38	52	M	T2	MDSCC	1.25	↓
39	43	M	T1	MDSCC	1.17	↑
40	38	M	T3	PDSCC	2.00	↓

[ M − Male; F- Female; Y –Yes; T1 – Tumour size is less than equal to 2 cm in greatest dimension; T2 – Tumour size is greater than 2 cm but less than equal to 4 cm in greatest dimension; T3 – Tumour size greater than 4 cm in greatest dimension].

Histopathologically, on the basis of degree of differentiation, all the OSCC samples were divided into three subtypes – Well Differentiated SCC (WDSCC), Moderately Differentiated SCC (MDSCC) and Poorly Differentiated SCC (PDSCC).

WDSCC was observed in 34 OSCC samples amongst which expression of 14 samples were upregulated (↑) with average fold change of 3.6 and expression in remaining 20 samples were downregulated (↓) with average fold change of 1.73.

MDSCC was observed in 5 OSCC samples in which expression of 4 samples were upregulated (↑) with average fold change of 1.4 while expression in remaining 1 sample was downregulated (↓) with 2.12-fold change. Only one PDSCC sample was observed in our study which showed down regulated DLG1 expression (2.66-fold change).

Based on the histopathological degree of differentiation, all the samples were divided into two groups for statistical analysis. One of the groups were WDSCC and the other was MDSCC (Fig. 2).

Fig. 2.

Hematoxylin and Eosin-stained section of MDSCC (10 X view).

After Kolmogorov Smirnov test, it was found that none of the groups showed deviation from normal distribution, followed by which paired t-test was done. However, DLG1 expression was not found to be significantly different between tumour and its adjacent normal for all OSCC samples (P value > 0.05). Similar results were obtained when the samples were classified on the basis of histopathological features, lesion types and tobacco consumption habit.

But the significant finding was that, out of 5 MDSCC histological subtype of OSCC samples, 4 samples were upregulated (↑) and remaining 1 sample was downregulated (↓).

4. Discussion

The concept of polarity in development, epithelial homeostasis and pathological alteration of tissues has emerged as an interesting aspect of the concerned biology.

The epithelial cells always exhibit apicobasal polarity in order to carry out different physiological functions. This polarity, a unique cellular phenomenon, is maintained by mainly three complexes viz. Crumbs Complex, ‘PAR’ Complex and Scribble Complex. Any functional perturbation of the proteins forming these complexes cause alteration of normal epithelial physiology en-route to epithelial pathology. Many epithelial pathologies involving colon, breast, oesophagus^{9, 10, 11} has been found to be associated with cell polarity alteration. But no scientific work has been found relating to oral cancer and/or pre-cancer with cell polarity after a thorough search of known scientific databases.

Even after extensive research, till date, the pathogenesis of different precancers as well as progression of precancers to malignancies is not clear, rather blurred with desperate scientific findings. Likewise in the field of oral oncology, many attempts also have been made to explain the pathological progression of oral precancer to cancer. Recently molecular aspect of cell polarity is gaining importance for unravelling the pathogenesis of epithelial dysplasia, forerunner of true cancer. And from the review of literature, association between carcinogenesis and polarity alteration is exhibited in various cancers. Many scientific reports have been published supporting this association in various cancers like loss of Scrib in breast cancer,¹² downregulation of DLG and PAR3 in cervical¹³ cancer and oesophageal cancer.¹⁴

In this maiden scientific exercise, we have tried to explore the association of expression of cell polarity proteins in OPMD and OSCC. The chief histopathological feature concerned with OPMD and OSCC is dysplasia with/without invasion. In this context, a new objective dimension of histopathological assessment of dysplasia in OPMD and OSCC can be introduced on the basis of deregulation of polarity. Here, we have chosen DLG1 as a representative of polarity protein. Mammals have 5 different DLG proteins and DLG1 is most closely related to drosophila DLG.¹⁵

Besides maintaining cell polarity, DLG1 also plays an important role in cell migration. The binding and colocalization of DLG and APC (Anaphase Promoting Complex) at the leading edge of a migrating cell contribute to the stabilization of microtubules and cell migration.¹⁶ Depletion of DLG, resulting in disintegration of APC, ultimately lead to microtubule destabilization towards the leading edge and subsequently, cells lose their migratory capacities.¹⁷ So, a cell, which is going to migrate should have a higher level of DLG1 expression.

In neoplasia, EMT is an indispensable phenomenon. An epithelial-mesenchymal transition (EMT) is a biologic process that allows a polarized epithelial cell, which normally interacts with basement membrane via its basal surface, to undergo multiple biological changes that enable it to assume a mesenchymal cell phenotype, which includes enhanced migratory capacity, invasiveness, elevated resistance to apoptosis, and greatly increased production of ECM components.¹⁸

If the flow of events of EMT can be traced back, altered cell polarity should be considered as starting point. In EMT, the apicobasal polarity must be transformed into front rear polarity. And in this process of acquiring front rear polarity, upregulation of DLG1 expression plays a vital role.^19,20 Thus, a severely dysplastic epithelial cell must have to undergo EMT on the way to become a cancer cell. According to the results of this study, it is evident that there is marked increase in DLG1 expression in high grade dysplastic epithelium of leukoplakia in comparison to low grade dysplasia. This finding supports the increased expression of DLG1 in our study as severely dysplastic epithelium are considered to be on the verge of EMT.

In case of carcinogenesis, after the invasion, the neoplastic invading cell start gaining some of the epithelial characteristics, attain cohesiveness and ultimately form neoplastic epithelial islands. This is called partial EMT.²¹ During partial EMT, there is downregulation of DLG1 because these neoplastic cells again start forming clusters instead of migration. The neoplastic epithelial cells first loose apicobasal polarity, then gain front-rear polarity. Finally, during formation of neoplastic islands, the said cells acquire apicobasal polarity. Again, they gain front-rear polarity by acquiring mesenchymal characters and migrate by the process of secondary EMT. During secondary EMT, the neoplastic cells loose cohesiveness and further migrate. So, during migration, there is upregulation of DLG1.

In this study, the results depict that there is no significant alteration in expression of DLG1 in WDSCC in comparison to normal tissue where there is formation of cluster of neoplastic cells ultimately producing epithelial islands and keratin pearls.

But, in case of MDSCC, when the same neoplastic cells keep on invading with minimal keratin pearl formation, they again start gaining the mesenchymal character. In our study, the upregulation of DLG1 in MDSCC can be seen as contributing factor for the shift of epithelial to mesenchymal characters.

It can be concluded that this alteration of expression of DLG1 in leukoplakia and OSCC samples can be considered as a significant factor in the progression of oral precancer to cancer and in pathogenesis of oral cancer itself.

As far as the clinical significance of the study is concerned, it can be concluded that increased expression of DLG1 leads to a poorer prognosis for both in leukoplakia and OSCC. This is because, the dysplastic epithelial cells with high DLG1 expression in leukoplakia has a tendency for invasion after undergoing EMT. Similarly, in OSCC, the dysplastic epithelial cells have a more tendency towards migration into deeper tissues thereby favouring metastasis.

If we consider the future prospective of this study, immunohistochemistry will be an indispensable part in order to identify mis localization of polarity proteins in OPMD and OSCC. It would also have been interesting if we can identify any association between cell polarity alteration with other hallmarks of cancer.

Declaration of competing interest

None.