Altered expression of junctional proteins as a potential biomarker in oral precancerous and cancerous patients

ABSTRACT

Due to a lower survival rate in patients with advanced clinical stages of oral cancer, discovering a biomarker that could diagnose and predict disease progression is vital. Cell–cell junctional proteins play a crucial role in the maintenance of tissue architecture but are often deregulated in different cancer. The present study investigates the expression of cell–cell junctional proteins viz: e-cadherin (E-cad) and zonula occludens-1 (ZO-1) in oral precancerous (OED) and cancerous (OSCC) patients to monitor if they can serve as practicable molecular markers. The ultrastructural junctional complex was studied by transmission electron microscopy, and the expression of proteins was performed by immunohistochemistry. The relationship between the expression of protein and clinicopathological features of the patients was checked by Pearson’s correlation test. Furthermore, the survival curve of the follow-up data was estimated by the Kaplan–Meier method. We observed a disrupted junctional complex and a significantly decreased immunoexpression of E-cad and ZO-1 in OED and OSCC when compared to the adjacent non-cancerous tissues. The expression of ZO-1 was associated with TNM stages, whereas E-cad was associated with histological grades as well as TNM stages. A positive correlation was observed between the expression of ZO-1 and E-cad proteins in OED and OSCC. Further, follow-up studies revealed that high ZO-1 and E-cad expressing patients survived longer than their low expressed counterparts. The present study shows disruption of junctional complex and alteration of junctional proteins expression that could draw the attention of health professionals to explore junctional proteins as a possible therapeutic target in oral cancer.

KEYWORDS:

• Immunohistochemistry
• oral epithelial dysplasia
• oral squamous cell carcinoma
• e-cadherin
• zonula occludens-1

1. Introduction

Despite extensive worldwide research and improvements in the field of surgery, radiation, and chemotherapy, the treatment of oral cancer still remains a major challenge. The incidence of oral squamous cell carcinoma (OSCC) is increasing worldwide and is most frequent in South Asian countries, with the highest incidence rate globally in both sexes.¹ Prolonged efforts have been made to wrestle the disease; however, due to lower survival index and delayed clinical detection, this hard work has yielded minimum improvements. Therefore, discovering a biomarker capable in predicting disease progression can play a crucial role to help cure the disease. Even though OSCC has been extensively studied,² the molecular characteristic associated with this malignancy remains to be identified. Studies have explored several genes/proteins that are transformed, amplified, or dysregulated in expression or deleted during epithelial–mesenchymal transition (EMT) which finally gets involved in the initiation of metastasis.^3–5 One of the characteristic features associated with tumor development is the deregulation of cell–cell adhesion.⁶ Progression of carcinoma and expression of mesenchymal genes are often associated with increased cellular motility and decreased cell–cell adhesion. These features are observed not only in cancerous (OSCC) but precancerous conditions i.e. oral epithelial dysplasias (OED) as well.^7,8 Loss of cellular contact promotes detachment of cancer cells from their site of origin and invades the adjacent tissues and finally metastasizes.⁹

Epithelial cell sheets form cell–cell adhesion that mainly involves tight junctions (TJs) and adherens junctions (AJs). Deregulation of these TJ and AJ proteins results in loss of cellular adhesion and leads to uncontrolled growth of cells.^4,10 TJ complex consists of various families of protein such as zonula occludens (ZO), junctional adhesion molecule-A (JAM-A), occludin, and claudins. Among them, the ZO family is the one involved in maintaining epithelial cell structure and function. Its family consists of ZO-1, ZO-2, and ZO-3.¹¹ Out of all the tight junction proteins, ZO-1 is the first to be reported in epithelial cells. It belongs to the membrane-associated guanylic acid protein and is the major structural and functional protein of tight junction to be associated with regulating the cell polarity as well as the permeability of epithelial cells.¹²

Cadherins are major components of AJs and serve a key role in the maintenance of epithelial tissue integrity.¹³ E-cadherin (E-cad) is a 120 KD transmembrane glycoprotein belonging to AJs and a major calcium-dependent cell surface adhesion molecule.¹⁴ It is encoded by the CDH1 gene located on chromosome 16q-22⁴ and is restricted to the surface of the epithelial cells.¹⁵ It is a widely distributed, intercellular adhesion molecule¹⁶ that through its cytoplasmic tail, associates with various intracellular proteins.¹⁷ It functions in cellular crosstalk and mediates cell-to-cell communication by protein interactions on the cytoplasmic membrane surfaces. Reports suggest the role of E-cad as a biomarker in detecting gene expression and predicting disease progression.¹⁸ Therefore, investigation of E-cad might be a promising area of interest and could be a potential biological marker.

Both ZO-1 and E-cad are considered as the major components of the junctional assembly. The role of E-cad has been studied in oral precancerous and cancerous conditions but reports on the role of ZO-1 are limited. However, their combined role in OED and OSCC remains to be validated. Therefore, the present study monitors the ultrastructural changes within the junctional complex during the progression of oral cancer followed by an immunohistochemical approach to examine changes in the expression level of E-cad and ZO-1 in OED and OSCC. Additionally, Pearson’s correlation χ2 test was performed to examine any possible association between the protein–protein expression and protein-clinicopathological features.

2. Material and methods

2.1. Study population

The characteristics of the variables are presented in Table 1. Experiments were undertaken with the understanding and written consent of each patient and in accordance with the ethical standards of the ethics committee of Assam University, Silchar and Silchar medical college and hospital, Silchar, Assam, India. Tissue samples have been collected from individuals suffering from OED and OSCC comprising 20 participants in each group. Samples were selected after the clinical and histopathological examination (Figure 1a-1c) of the patient. Clinical staging of lesions and classification of OED and OSCC was performed based on the guidelines proposed by the American Joint Committee on Cancer and International Agency for Research on Cancer, WHO respectively.^19,20 Control samples were collected from adjacent oral gingival tissue of OED and OSCC patients (10 cases each) who were not subjected to field change cancerization. The control samples from OED and OSCC patients have been merged as “control” given that no variation in any study parameter was observed.

Table 1. Socio-demographic and clinicopathological features of OED and OSCC patients

Variables	No. of cases (%)
	OED	OSCC
Age, years
<50	10 (50)	8 (40)
≥50	10 (50)	12 (60)
Sex
Male	12 (60)	11 (55)
Female	8 (40)	9 (45)
Lifestyle
Tobacco	16 (80)	10 (50)
Tobacco+alcohol	2 (10)	9 (45)
None	2 (10)	1 (5)
Histological grades
Mild dysplasia	5 (25)	-
Moderate dysplasia	11 (55)	-
Severe dysplasia	4 (20)	-
Poorly differentiated	-	9 (45)
Moderately differentiated	-	8 (40)
Well differentiated	-	3 (15)
TNM stage
T1	-	8 (40)
T2	-	6 (30)
T3	-	6 (30)

OED: oral epithelial dysplasia; OSCC: oral squamous cell carcinoma; TNM: tumor node metastasis

Figure 1. Histological and Transmission electron microscopic (TEM) images of control, oral epithelial dysplasia (OED) and Oral squamous cell carcinoma (OSCC). Histological sections of control tissues (1a) shows gingival epithelium characterized by an intact basement membrane. OED (1b) represents severe dysplasia demonstrated by abnormal cells occupying more than two-thirds of the epithelium, and pearl formation (arrow). OSCC (1 c) shows poorly differentiated epithelium characterized by tumor cells (arrow) infiltrating the sub epithelium (×20). TEM images of control tissue (1d), OED (1e) and OSCC (1 f). Arrows represents tight junctions (TJ), arrowheads represents adherens junctions (AJ) and arrowhead on both ends represents desmosomes, NJC: normal junctional complex, DJC: disrupted junctional complex (scale bar 0.5 µm).

2.2. Transmission electron microscopy

Primary fixation of tissue samples (0.5–1.0 mm) was performed in karnovsky’s fixative for 4–12 hours at 4°C followed by a secondary fixation in osmium tetraoxide for 1 hr. After dehydration and embedding in araldite, the polymerized specimen was sectioned at approximately 90 nm thickness in an ultramicrotome and finally stained using uranyl acetate and lead citrate. Sections were viewed on JOEL 2100 transmission electron microscope at an accelerating voltage of 120 kV. The junctional structure amid adjoining cells was observed from multiple sites that were clearly identified.

2.2.1. Immunohistochemical staining

Paraffin-embedded tissues sections were taken on poly-L-lysine coated slides and incubated for approximately 2 hours at 56°C. An immunohistochemical process was executed using an IHC Detection kit (Cell Marque, Rocklin, CA, USA), which is a three-step indirect immunoperoxidase technique. Antigen retrieval was carried out by using the tris-EDTA buffer in a microwave oven at pH 9. This was followed by blocking in an endogenous peroxidase solution (PeroxFreeTM block). Sections were incubated at room temperature in humidifying chamber for 60 minutes with primary antibodies obtained from Santa Cruz Biotechnology, USA, viz: ZO-1 (Rabbit ZO-1) and E-cad (Rabbit E-cadherin) prepared in 3% BSA (1:100). This was followed by washing the slides three times with IHC wash buffer and incubating with HiDef Detection TM Amplifier (known to universally detect mouse and rabbit primary antibodies) for 30 minutes. This was followed by incubation with polymer HiDef Detection TM HRP Label for 30 minutes. Subsequently, a substrate chromogen solution known as 3,3ʹ-diaminobenzidine tetrahydrochloride (DAB Substrate Kit) was added to the slides and counterstained with hematoxylin for approximately 10 minutes. Thereafter, sections were washed, dehydrated, and mounted with DPX, and observed under Leica DMLS light microscope.

2.2.2. Immunohistochemical scoring

Ten random fields were selected and a hundred cells were counted in each field making 1000 numbers of cells counted per slide. Two independent examiners scored the slides in a blinded fashion. Staining intensity (I) was ranged as weak (1), moderate (2), and high (3). The frequency of positive cells (F) was categorized as follows: 0, <10% positive cells; +1, ≥ 10-<25%; +2, ≥ 25-<50%; +3, ≥ 50-<75% and +4, ≥75%. Final scores were calculated by multiplying the frequency of positive cells (F) and the staining intensity (I) i.e. (FxI). Cases with a final expression score less than or equals to 6 were considered to be low expressed and greater than 6 were considered to be highly expressed for both the proteins.

2.3. Statistical analysis

Expression scores were expressed as mean ± SEM. Analyses were executed using IBM SPSS statistics version 22 (IBM Corporation) for windows. A one-way analysis of variance (ANOVA) was used to determine the significance and post hoc analysis was performed for multiple comparisons among groups using Tukey’s multiple-comparison tests. Correlation between potential markers and categorical analysis of patient’s clinicopathological parameters was analyzed by Pearson’s correlation χ2 test. Survival analysis was carried out by the Kaplan–Meier method and the log-rank test was employed to establish significant differences set at p < .05.

3. Results

3.1. Patient characteristics

Histopathological examination of tissues was performed where 25% of OED cases were categorized as mild dysplasia, 55% moderate dysplasia, and 20% severe dysplasia. In the case of OSCC, 45% of cases were graded as poorly differentiated, 40% moderately differentiated, and, 15% well differentiated (Table 1).

3.2. Ultrastructural changes in the junctional complexes

Ultrastructural microphotographs of control, OED, and OSCC tissues are shown by transmission electron microscope (Figure 1d-1f). The presence of electron-dense material (arrow) on the apical region of the adjoining cells represents TJ’s. AJ’s are located between TJ and desmosomes. The control population displayed an intact structure with distinct TJ, AJ, and normal junctional complex (Figure 1d) whereas, in the case of OED (Figure 1e) and OSCC (Figure 1f), disrupted junctional complex is evident as the cells proceed from normal to malignancy.

3.3. Immunohistochemical assessment

3.3.1. ZO-1 protein expression in OED and OSCC

Differences in expression of ZO-1 were observed between all groups. Localization of ZO-1 was membranous in 50% (10/20) of the control population and the expression was homologous all over the epithelium (Figure 2a-b) while in the case of OED only 35% (7/20) cases exhibited ZO-1 predominantly on the cell–cell border (Figure 2c-d). It is interesting to note that none of the OSCC samples exhibited membranous staining that indicates that its localization is disturbed as the cells proceed toward malignancy (Figure 2e-f). Low ZO-1 expression was observed in 25% (5/20) cases of control and 35% (7/20) cases of OED. However, OSCC exhibited 60% (12/20) cases low expressed ZO-1. We observed that the ZO-1 protein level was significantly lower in OSCC (p < .001) than in the adjacent non-cancerous gingival tissues. Pair-wise comparison between groups showed a significant decrease in ZO-1 expression in OSCC (p < .05) when compared to OED (Figure 3a).

Figure 2. Expression of ZO-1 and E-cad in control, OED and OSCC tissues. ZO-1 immunoexpression (a-f): a-b: ZO-1 expression in control sample, c-d: ZO-1 expression in OED, e-f: representing ZO-1 expression in OSCC. E-cad immunoexpression (g-l): g-h: E-cad expression in control sample, i-j: E-cad expression in OED and k-l: E-cad expression in OSCC.

Figure 3. Expression score of ZO-1(3a) and E-cad (3b) in control, OED and OSCC. Values are significantly different from control: ***p < .001. Values having similar superscripts are significantly different from each other: ^a p < .05.

Bivariate analysis (Table 2) showed that the intensity of ZO-1 immunoexpression was associated with tumor node metastasis (TNM) stage (p = .047) and male gender (p = .002) in OSCC. However, no statistically significant relationship was observed with other studied and clinical variables like age, lifestyle, pathological grade.

Table 2. Relationship between expression of proteins and clinicopathological parameters of OED and OSCC patients

		ZO-1						E-cad
		OED			OSCC			OED			OSCC
	Variables	Low (%)	High (%)	p- val.	Low (%)	High (%)	p- val.	Low (%)	High (%)	p- val.	Low (%)	High (%)	p- val.
Age, years	<50	3 (15)	7(35)	0.639	5 (25)	3 (15)	0.852	5 (25)	5 (25)	0.361	5 (25)	3 (15)	0.292
	≥50	4 (20)	6 (30)		7 (35)	5 (25)		7 (35)	3 (15)		10 (50)	2 (10)
Gender	Male	6 (30)	6 (30)	0.085	10 (50)	1 (5)	0.002	8 (40)	4 (20)	0.456	8 (40)	3 (15)	0.795
	Female	1 (5)	7 (35)		2 (10)	7 (35)		4 (20)	4 (20)		7 (35)	2 (10)
Life Style	Tobacco	5 (25)	11 (55)	0.781	7 (35)	3 (15)	0.370	10 (50)	6 (30)	0.112	6 (30)	4 (20)	0.292
	Tob. + Alc.	1 (5)	1 (5)		5 (25)	4 (20)		2 (10)	0 (0)		8 (40)	1 (5)
	None	1 (5)	1 (5)		0 (0)	1 (5)		0 (0)	2 (10)		1 (5)	0 (0)
Path.Grade	Mild Dys.	2 (10)	3 (15)	0.691	-	-	0.231	1 (5)	4 (20)	0.108	-	-	0.005
	Mod. Dys.	3 (15)	8 (40)		-	-		8 (40)	3 (15)		-	-
	Severe Dys.	2 (10)	2 (10)		-	-		3 (15)	1 (5)		-	-
	Poorly Diff.	-	-		7 (35)	2 (10)		-	-		8 (40)	1 (5)
	Mod. Diff.	-	-		3 (15)	5 (25)		-	-		7 (35)	1 (5)
	Well Diff.	-	-		2 (10)	1 (5)		-	-		0 (0)	3 (15)
TNM stage	T1	-	-		4 (20)	4 (20)	0.047	-	-		3 (15)	5 (25)	0.007
	T2	-	-		2 (10)	4 (20)		-	-		6 (30)	0 (0)
	T3	-	-		6 (30)	0 (0)		-	-		6 (30)	0 (0)

Note: Expression score ≤ 6: low expressed and > 6: high expressed; p- val.: p- value, ZO-1: Zonula occludens-1; E-cad: E-cadherin; OED: oral epithelial dysplasia; OSCC: oral squamous cell carcinoma; TNM: tumor node metastasis; Tob.: Tobacco; Alc.: Alcohol; Path. Grade: Pathological Grade; Dys: Dysplasia; Diff: Differentiated.

3.3.2. E-cad protein expression in OED and OSCC

In the control gingival mucosa, the expression of E-cad was observed only on the basal and parabasal layers, while the marking was absent on the upper third of the epithelium (Figure 2g). Expression of E-cad was continuous and homologous, featuring well-ordered junctional integrity (Figure 2h) in 80% (16/20) of control tissues. In the OED group, a significant reduction in E-cad expression was observed in all the epithelial layers (Figure 2i). Moreover, only 35% (7/20) cases show membrane staining patterns (Figure 2j). A complete disruption of the junctional complex was observed in the OSCC group (Figure 2k). A higher magnified image shows a discontinuous or absence of E-cad expression (Figure 2l) which demonstrates that as the cells proceed toward malignancy, the expression of E-cad junctional protein as well as junctional complex gets disrupted. Low E-cad expression was observed in 30% (6/20) of control cases, whereas OED exhibited 55% (11/20) cases of low expressed E-cad. In the case of OSCC, 75% (15/20) cases showed low expressed E-cad. Expression score shows a significant decrease in expression of both OED and OSCC (p < .001) when compared to control. Pairwise comparison between groups resulted in a significant decrease in OSCC (p < .05) when compared to OED tissues (Figure 3b).

When compared E-cad immunoexpression with clinicopathological parameters, we found that expression of E-cad was significantly associated with TNM stages (p = .007) and histological severity (p = .005) (Table 2).

3.3.3. The relationship between E-cad and ZO-1 immunoexpression

A positive correlation was observed between the expression level of E-cad and ZO-1 in OED (r = 0.385, p = .085) and OSCC (r = 0.471, p = .036) (Table 3) which demonstrates that both the proteins have similar expression patterns among variables.

Table 3. Correlation analysis between expression of ZO-1 and E-cad in OED and OSCC

	E.cad Expression
	OED					OSCC
ZO-1 Expression	Low (%)	High (%)	Total (%)	Pearson correlation (r)		Low (%)	High (%)	Total (%)	Pearson correlation (r)
Low	6 (30)	1 (5)	7 (35)			8 (40)	4 (20)	12 (60)
High	6 (30)	7 (35)	13 (65)	r = 0.385		7 (35)	1 (5)	8 (40)	r = 0.471
Total	12 (60)	8 (40)	20 (100)	(p = .085)		15 (75)	5 (25)	20 (100)	(p = .036)

Note: Expression score ≤ 6: low expressed and > 6: high expressed; E-cad: E-cadherin; ZO-1: Zonula occludens-1; OED: oral epithelial dysplasia; OSCC: oral squamous cell carcinoma

3.4. Survival analysis

The participants were followed up for more than 60 months. Plots for high and low expression of ZO-1 and E-cad are shown in Figure 4. The log-rank test was carried out on the followed up participants. 2 (10%) patients were lost to follow up and the remaining 18 (90%) were alive in OED cases. Eight (40%) patients died of OSCC during the end of the follow-up period. Three (15%) patients were lost for follow-up and the remaining 9 (45%) patients were alive. Kaplan–Meier survival curves in OSCC patients demonstrated that the overall survival rate in patients with low ZO-1 (Figure 4a) and E-cad expression was lower than their highly expressed counterparts studied longitudinally (Figure 4b).

Figure 4. Kaplan-Meier survival plots for high vs. low expression of ZO-1(4a) and E-cad (4b) in OSCC patients.

4. Discussion

Oral cancer severely threatens and affects the quality of life by rigorously disturbing major functions of the body and leading to a lower survival rate that gets worsens with the advancement of clinical stages.²¹ Therefore, this demands the development of novel biomarkers for the diagnosis and prediction of disease progression. Theoretically, any distinctive change in the biological structure and function can serve as a biomarker or a therapeutic target. We have earlier reported that certain TJ proteins like claudin-1 and Junctional adhesion molecule-A are deregulated during precancerous and cancerous conditions.²² ZO-1 is reported to have barrier and adhesive functions.²³ Apart from that, ZO-1 participates in the building of cellular coupling, forms major closely connected structural and functional protein through interconnecting its PDZ domain and actin.^24,25 In this study, reduced expression of ZO-1 protein in OED and OSCC tissues was observed when compared to control, although statistically significant difference was observed only in OSCC. Our results are in concordance with Qiu et al. where expression of ZO-1 on laryngeal squamous cell carcinoma (LSCC) tissues was reported to be significantly lower than their corresponding noncancerous tissues.²⁶ A high degree of LSCC differentiation in their study was associated with high expression of ZO-1 protein but we did not observe any correlation between the degree of differentiation and expression of ZO-1 in both OED and OSCC. However, we could show the association of ZO-1 immunoexpression with TNM stages in OSCC. Studies report that the rate of oral cancer is significantly higher in males than females²⁷ and interestingly, we observed a correlation between downregulated ZO-1 immunoexpression and male gender in OSCC. These findings depict that the gender-specific expression of E-cad could also serve as an independent factor affecting the prognosis of OSCC. However, more extensive work is required to confirm our observations.

E-cad has been the major focus of cancer studies as its alteration has been observed in OSCC, ovarian cancer, early-stage cervical carcinoma, and colorectal cancer.^28–31 Our study supports the previous study in OED and OSCC tissues that reported the reduction in E-cad expression as the severity increases.²⁸ Reduction in E-cad expression is in line with the observation of earlier studies that reported downregulation of E-cad in the precancerous stage³² and significantly decreased expression in OSCC.³³ Several epigenetic mechanisms proposed for E-cad loss are chromatin-mediated effects and alteration in expression and function of a trans-acting factor that regulates CDH1 gene transcription.³⁴ Downregulation of E-cad with increasing grade of a tumor can be explained by the fact that E-cad acts as a tumor suppressor gene and its downregulation enhances invasion of cancer in adjacent normal tissues.³⁵ We observed that reduced expression of E-cad was associated with the histological grades and TNM stages in OSCC which demonstrates that E-cad could be a potential marker of aggressive histological characteristics and might be involved in the progression of cancer. Further, its low expression in poorly and moderately differentiated OSCC (Table 3) features its character as an epithelial marker, therefore, suggesting its role as an indicator for the degree of epithelial differentiation. Kaplan–Meier analysis shows that low ZO-1 and E-cad expression in our study relates to lower survival time for the patients, which indicates that both the proteins may be a favorable prognostic marker and a monitoring indicator in patients with OSCC. In addition to ZO-1 and E-cad, other tight junctional proteins, viz. claudin-4 and claudin-1 were also reported to correlate with poor prognosis and disease recurrence, respectively, in OSCC patients.³⁶ Association of Claudin-4 with high histological grade was observed by de Vicente et al., 2015 which is consistent with our E-cad studies.³⁶ These suggest that not only E-cad and ZO-1, but other tight junctional proteins could also serve as biomarkers for OSCC. Besides, our previous study shows upregulation of claudin-1 in OSCC,²² also reported by other research groups,^37,38 while our current study shows downregulation of ZO-1 and E-cad in the precancerous and OSCC patients, thereby implying an inverse relationship between these proteins. In fact, there are studies that show overexpression of claudin-1 could induce epithelial–mesenchymal transition and eventually down-regulate the expression of E-cad in metastatic cancer colon cells.³⁹ Further investigation correlating ZO-1 and E-cad with different classes of claudins in OSCC patients at the molecular level is an utmost topic of interest that will be explored in the future.

The electron microscopy images support the protein immunoexpression data by demonstrating poorly visualized or disrupted junctional complexes (TJ and AJ) in OED and OSCC compared to the control tissues (Figure 1). Immunohistochemical results show an imbalance in the expression of TJ and AJ protein expression that ultimately resulted in the loss of the barrier integrity. Loss of junctional integrity of TJ and AJ can consequently encourage the EMT rate of cancer cells that can lead to loss of the cell polarity, therefore reducing intercellular adhesion, alter the epithelial morphology and as a result strengthen the invasiveness of cancer cells. Malfunction in the permeability barrier helps the tumor cells to absorb nutrients and encourage their abnormal proliferation thereby increasing the invasion and metastatic ability of cancer cells.⁴⁰ Downregulated expression of ZO-1 and E-cad and disrupted junctional complex in oral cancer can be considered as an early phenomenon and that loss of epithelial cohesion may be an indicator of epithelial to mesenchymal transition in cells. Our results confirm the role of ZO-1 and E-cad as a biomarker for the detection of OSCC. A positive correlation between ZO-1 and E-cad expression in OED and OSCC suggests that both the proteins together might act as a biomarker in the detection of OSCC.

Disclosure of interest

The authors report no conflicts of interest.

Acknowledgments

The authors thank the patients for their contribution to the study. The authors are also thankful to the Sophisticated Analytical Instrument Facility, North-Eastern Hill University, for the TEM (JOEL 2100) facility.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Department of Biotechnology, Ministry of Science and Technology [BT/358/NE/TBP/2012].