Gene Ther Mol Biol Vol 10, 123-132, 2006

Epstein-Barr Virus downregulates expression of DNA-Double strand break repair proteins in nasopharyngeal cancer

Research Article

Prabha Balaram^1,*, Smriti M Krishna¹, Susan James², Vino. T. Cheriyan¹, Sreelekha Therakathinal Thankappan¹, Aleyamma Mathew³

¹Division of Cancer Research, Regional Cancer Centre, Medical College. P.O, Trivandrum 695 011, Kerala, India.

²Department of ENT, Medical College, Trivandrum 695-011, Kerala, India

³Department of Statistics, Regional Cancer Centre, Trivandrum, Kerala, India __________________________________________________________________________________

*Correspondence: Dr. Prabha Balaram, Professor and Head, Division of Cancer Research, Regional Cancer Centre, Trivandrum 695 011, Kerala, India; Phone: 91-0471-2522203; Fax: 91-0471-2447454; E-mail: prabhabalaram@yahoo.co.in

Key words: DNA-PKcs, ATM, Nasopharyngeal carcinoma, EBV, HPV, Radiotherapy, DNA-double-strand break (DSB) repair proteins.

Abbreviations: 5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium, (BCIP/NBT); bovine serum albumin, (BSA); Epstein Barr Virus, (EBV); Epstein-Barr nuclear antigen-1, (EBNA-1); head and neck squamous cell carcinoma, (HNSCC); ionizing radiation, (IR); Nasopharyngeal carcinoma, (NPC); Radiotherapy, (RT); severe combined immunodeficiency, (SCID); sodium dodecyl sulfate-polyacrylamide gels, (SDS-PAGE)

Received: 15 December 2005; Revised: 13 March 2006

Accepted: 27 March 2006; electronically published: April 2006

Summary

Nasopharyngeal carcinoma (NPC) is a unique cancer due to its etiology, and incidence and viral infection pattern. Radiotherapy (RT) is the main treatment modality and these lesions vary considerably in treatment response. In the present study, the effect of EBV and HPV infection on response to treatment based on the expression of DNA-repair proteins, ATM and DNA-PKcs was assessed in NPC based on the rationale that the expression of these proteins play important roles in the response to RT in NPC. 103 NPC biopsies and 26 benign adenoid lesions of the Nasopharynx were collected before treatment and graded according to WHO classification. Response to treatment(radiation) was evaluated clinically. Western blotting and immuno histochemical analysis were used for evaluating the expression of DNA-PKcs and ATM and PCR using specific primers was used for detection of EBV and HPV. EBV presence was also confirmed by EBER-ISH. 63% of NPC were EBV+ve and 30% HPV+ve in the samples studied. Expression of ATM and DNA-PKcs was significantly increased in NPC when compared to benign samples (p<0.001). However, NPC showing EBV infection showed downregulation of ATM and DNA-PKcs expression (p=0.009 and p=0.011), in comparison to HPV infected NPC’s (p=0.001 and p=0.003). NPC patients with HPV positivity also showed a significantly poor response to treatment (p=0.002)with a higher rate of recurrence and lower disease free survival. The study indicates that EBV infection down regulates the expression of DNA-repair proteins and renders NPC sensitive to RT while HPV infection up regulates their expression making the tumors resistant to therapy. The results of the study also indicate that assessment of expression of DNA-PKcs and ATM in biopsy specimens can be used as criteria to identify radio-resistant NPC’s and selection of appropriate therapy regimens.

I. Introduction

Recent research provides ample evidence that viruses affect response of tumour cells toward anti-cancer drugs and irradiation. Viruses interfere with specific cellular genes, and their interaction with the tumor suppressor genes abrogate cell cycle arrest and disturb repair of radiation and drug-induced DNA lesions (Efferth and Grassman, 2000) thus leading to enhanced susceptibility to carcinogenesis in tissue cells persistently infected with viruses.

DNA-PKcs and ATM are two crucial proteins in the DNA-damage repair pathway. Cells lacking DNA-PK activity as a result of a mutation in any of the subunits are radiosensitive and deficient in the rejoining of radiation induced DNA-DSBs and reduced activity in mice leads to ‘severe combined immunodeficiency (SCID), which is coupled with extreme hypersensitivity to ionizing radiation (IR). Complete loss increases the risk of developing lymphomas and impaired V(D)J recombination (Smith and Jackson 1999). Up regulation, on the other hand, was recently reported to correlate with radiation resistance and suggests its potential as a molecular target for novel radiation sensitization therapy for oral squamous cell carcinomas (Shintani et al, 2003).

At the cellular level, ATM deficiency is manifested by increased sensitivity to IR or other agents that yield DSBs, chromosomal instability, cellular and humoral immunodeficiency, developmental defects in various organ systems and predisposition to cancer (Khanna, 2000; Becker-catania and Gatti, 2001). A-T carriers also represent a large proportion of patients with enhanced radiation sensitivity (Oppitz et al, 1999). AT deficient cells are impaired in IR induced G1, intra-S, and G2-M cell cycle check points and DNA damage repair capacity. Thus it has a surveillance role in maintaining genomic integrity.

NPC is a unique subset of head and neck squamous cell carcinoma (HNSCC) and is a rare malignancy in most parts of the world (Fandi et al, 1994). Geographical variation in incidence rates, rare occurrence of this tumor and the persistent association with Epstein Barr Virus (EBV), suggest that environmental factors and genetic susceptibility play crucial roles in the etiology of the disease (Vokes et al, 1997). Sporadically occurring NPC in the West usually belong to the WHO Type I histology and is associated with alcohol and smoking habits (Vokes et al, 1993). In many parts of Asia, including Southern China and South East Asia, NPC occurs as an endemic disease, and histologically, majority of cases belong to WHO Types II and III (Nonkeratinising and Undifferentiated types) with persistent association with EBV (Hui et al, 2002). Radiotherapy (RT) is the mainstay of treatment for NPC. However, biological aggressiveness and the radiation sensitivity of tumors within the same stage varies considerably and cannot be predicted by conventional histopathological evaluation. These facts indicate that there is a strong need for additional predictive and prognostic factors in order to improve therapy results of these patients. This study addresses the influence of the viruses on overall treatment response of NPC and the potential use of crucial DNA-DSB repair proteins, namely DNA-PKcs and ATM as biological tumor markers in predicting clinical outcome, mainly response to RT.

II. Materials and methods

A. Samples

Nasopharyngeal carcinoma samples (103) were collected from the ENT department, Medical College Hospital, and Regional Cancer Centre, Trivandrum, Kerala, India, with the approval of local ethical committees before start of any treatment. The histological diagnosis was confirmed by a pathologist and grading was performed according to WHO classification. One part of the biopsy sample was snap frozen for protein extraction and the other part processed for routine paraffin embedding. Radiation therapy was the sole treatment given to these patients. Clinical follow up period was from 1 to 71 months, the median being 18.5 months. We considered in particular, the amount of tumor mass reduction at the end of primary treatment with no evidence of residual disease as a good response to treatment. Loco-regional failures, recurrence, distant metastasis etc. were considered as signs of radio-resistance (poor response). Majority of the lesions with poor response had recurrence of the disease and hence recurrence was taken as the criteria of a poor treatment response for analysis. The benign lesions included cases of adenoids of the nasopharynx.

B. Western blotting

Extraction of total protein was done from biopsy samples snap frozen and stored in liquid nitrogen. Protein extracts were prepared in 1ml of lysis buffer [50mM Tris (pH 7.5), 0.1mM EDTA, 150mM NaCl, 1% Triton-X-100, 0.5mM PMSF, 200U/ml aprotonin, 10mg/ml leupeptin, 0.5% NP-40] for 30 minutes on ice. Protein lysates were quantified using Bradford assay with bovine serum albumin (BSA) as a reference standard and resolved by 12% SDS-PAGE (sodium dodecyl sulfate-polyacrylamide gels). The proteins were transferred to nitrocellulose membrane (Millipore Co. USA) and the membrane blocked with TBST (20mM Tris base, 135 mM NaCl, 0.1%Tween-20, pH 7.6) plus 3% powdered non-fat milk. After incubation with primary antibodies against DNA-PKcs (N-20), ATM (H-248) and b-actin (Santacruz Biotech, USA), in optimal concentrations, the membranes were incubated successively with alkaline phosphatase conjugated secondary antibody and visualized with alkaline phosphatase specific chromogens BCIP/NBT (5-bromo-4-chloro-3-indolyl phosphate/nitroblue tetrazolium).

C. Immuno histochemistry

Indirect immuno histochemical staining for assessing the subcellular localization of DNA-PKcs and ATM proteins was performed using monoclonal antibodies N-20 and H-248 respectively (Santacruz Biotech). Paraffin sections (5mm) from tumor tissue were deparaffinized and rehydrated through a series of graded ethanol. For antigen unmasking, sections were incubated in 10mM citrate buffer; pH6 and heat treatment given for 5 min. in a microwave oven. Inactivation of endogenous peroxidase activity was done with H₂O₂ and blocking of nonspecific binding by normal mouse serum and BSA. This was followed by incubation with the primary antibody, linking antibody and development of the reaction with H₂O₂ as substrate and DAB (diaminobenzidine) as chromogen. As negative controls, sections treated similarly with PBS substitution instead of the primary antibodies were included. The sections were evaluated using a visible microscope and the number of positive nuclei was expressed as percentage and graded as follows; 0=score1, 1-25%=score 2, 26-50%=score 3, 51-75%=score 4 and >75%=score 5.

D. DNA extraction

Paraffin embedded tissues were used for DNA extraction. The first and last sections (5 mm) of paraffin embedded tissues were stained with haematoxylin and eosin and histologically examined for confirmation of diagnosis and estimation of percentage of normal and neoplastic cells and inflammatory infiltrate. 10 mm sections from paraffin blocks were collected and dewaxed by xylene immersion, followed by graded alcohol and sterile distilled water washes. DNA was extracted using the standard phenol-chloroform method. 100 ng of template DNA was used for each set of PCR amplification.

E. Detection of viruses

All samples were amplified with b-globin primers to assess DNA integrity and only samples that tested positive for b-globin were further analyzed. Primer pairs were constructed that amplified the non-polymorphic Epstein-Barr nuclear antigen-1 (EBNA-1) gene and the presence of EBV was evident by a 262 bp product. EBV positivity was further confirmed by EBER ISH assay (Novocastra,UK) carried out as per the kit instructions. Presence of HPV DNA was identified with highly conserved Late-1 gene of HPV, which encoded a viral capsid protein (HPV L1) using consensus primers MY09 and MY11. The cases scored as HPV positive using consensus primers were subjected to two multiplex PCR reactions (TS1-PCR and TS2-PCR) for the determination of types of HPV infection. After electrophoresis in 3 % agarose gels along with marker, the HPV types were scored according to the length of the amplified fragment (450 bp for consensus HPV, 280bp for HPV 6, 152 bp for HPV16, 216 bp for HPV 18 and 360 bp for HPV 11). Negative controls consisted of a mixture of all the reagents used in the PCR mixture preparation, adding 1ml of sterile distilled water instead of the DNA and also a known EBV and HPV negative sample. The positive control was a previously known EBV positive Hodgkins lymphoma sample and HPV positive Condyloma tissue. These samples were processed in parallel to patient samples to exclude the detection of contamination.

F. Statistical analysis

The statistical analysis of the protein status in relation to clinical and pathological data was performed by Chi-square test for trend analysis. Students t test was used to analyze the difference in mean staining scores of the proteins among the different variables. All p values were two-sided and considered significant at p<0.05. Kaplan Meier survival analysis was carried out to evaluate the influence of various DNA-DSB repair proteins and viruses on survival.

III. Results

A. Patient details

Briefly, the study samples included patients with a mean age of 40.4 (range 9-79). Majority of the patients were males (68%) and stage III and IV patients (64%), followed by early disease stages, I and II (36%). 67% Patients showed presence of cervical lymph node involvement. Local or regional recurrence of the disease was observed in 35.6% of the NPC cases (Table 1). Follow up was available in 91 patients (6-71 months) and of these, 66% patients were alive and 34% died of the disease during the study period.

1. Histology

Formalin fixed, paraffin embedded biopsy tissues were used for histopathological exmination. Based on the WHO classification, the distribution of cases were as follows; KSCC (WHO Type I)-8, differentiated NKSCC (WHO Type II)-56, and UDC (WHO Type III)-39 (Table 1). KSCC’s were moderately to well differentiated squamous cell carcinoma.

Table 1. Expression of DNA damage proteins in relation to malignancy and Stage, histology and viral status of the lesions

	No. of cases	DNA-PKcs	ATM
Benign Cancer	26 103	6.34 ± 1.6 15.20 ± 2.9 *	1.00 ± 0.01 9.32 ± 2.1**
Histological Classification WHO I WHO II WHO III	8 56 39	14.12 ± 9.32 15.32 ± 4.10 15.25 ± 4.66	6.25 ± 6.25 10.26 ± 3.24** 8.58 ± 3.06 *
Stage Classification Stage I Stage II Stage III Stage IV	5 33 25 40	55.75 ± 16.52** 9.33 ± 4.10 14.72 ± 6.18 21.38 ± 5.90*	30.00 ±19.14* 5.75 ± 2.9 8.05 ± 3.94* 12.63 ± 4.40*
Recurrence Nil Present	46 57	5.74±4.08 31.30±5.61^êê	1.85±0.85 20.47±4.51^êê
Node status Negative Positive	30 73	19.00±6.54 15.00±3.44	11.40±4.36 9.31±2.68
EBV Negative Positive	38 65	24.84 ± 5.80 9.56 ± 2.90^ê	16.57 ± 4.5 5.07 ± 1.90^ê
PV Negative Positive	72 31	8.12 ± 2.50 31.64 ± 6.8 ^êê	3.23 ± 1.40 23.45 ± 5.5 ^êê

*p value <0.05 – comparison with benign lesions

** p value <0.001 – comparison with benign lesions

^ê p value <0.05 - comparison between positive and negative lesions

^êê p value <0.001 – comparison between positive and negative groups

2. Detection of viruses

i. EBV detection

Among the 103 NPC tissues and 26 benign nasopharyngeal tissues analyzed, 65 NPC samples (63%) and 7 (27%) benign samples showed EBV positivity (Table 2) in both PCR and EBER ISH assays. The EBV positivity of the benign lesions was restricted to the lymphoid cells except in one case which showed mild positivity of epithelial cells along with lymphoid cells in EBER-ISH. 2 cases (25%), 30 cases (53%) and 33 cases (85%) were positive for EBV in WHO Type I, WHO Type II and WHO Type III respectively (Table 2). EBV positivity was found to be highest in the WHO III tumours and Chi-square analysis showed this association to be highly significant (p<0.001). No significant difference was noticed in EBV positivity in the different stages of the disease. EBV positivity was associated with a statistically higher rate of recurrence (45% cases in EBV negative vs 60% in EBV positive cases p=0.045) and did not show any statistical significance with nodal status. In the EBV positive group, the recurrence was independent of the WHO type with similar distribution in WHOII and WHO III. The number of EBV+ patients in WHOI group was very low.

ii. HPV detection

With the consensus primers, 31 of 103 (30%) NPC samples and 4% (1/26) of the benign lesions were positive for HPV (Table 2). Upon further typing by TS1-PCR for HPV 6 and HPV 16 and TS2-PCR for HPV 11 and HPV 18, it was found that the separate infection rate of the nasopharyngeal lesions with HPV 6, HPV 16, HPV 11 and HPV 18 was 2%(2/103), 19.4%(20/103), nil, and 9% (9/103) respectively. Highest frequency of HPV infection was seen in WHOI (62.5%, 5/8) followed by 34% (19/56) in WHO II and 18% (7/39) in WHO III (Table 2) ( Chi-square significance p=0.031). HPV 16 was the most common type with 62.5% in WHO I, 16% (9/56) in WHO II and 15.4% (6/39) in WHO III. HPV 18 infection was found in 14.3% (8/56) in WHO II and 2.6% (1/39) in WHO III and absent in WHO I. 3.5% (2/56) of WHO II cases were found to be positive for HPV 6. None of the nasopharyngeal lesions showed presence of HPV 11 and no double or multiple infections was observed in any of the lesions. HPV positivity was associated with a higher rate of recurrence (78.3% in HPV+ vs 57.1% in HPV-ve. p=0.04) while no relation was seen to nodal involvement or stage of the disease.

iii. EBV and HPV co-infection

A total of 11 NPC cases (13%) showed co-infection of EBV and HPV. Highest number of co-infected samples was seen in WHO Type II with 19.5% (8/56) followed by that in WHO III (8%, 3/39) and no co-positive sample in WHO I Type lesions (Table 2). Percentage of samples showing co-infection was higher in node positive cases (82%), advanced stage of disease (73%), and among group showing recurrence (71.4%), which indicates an aggressive clinical course. Majority of cases showed single infection by either EBV or HPV and the mutually exclusive pattern of EBV and HPV infection in NPC is further evident from the significant negative bivariate correlation (r=-0.376, p=0.009) between the positivity for the two viruses.

C. Expression of DNA-damage repair proteins

1. Immunohistochemistry

i. Expression of DNA-PKcs

a. Benign vs NPC

Staining pattern to anti-DNA-PKcs antibody was mostly nuclear with occassional lesions showing cytoplasmic staining. Benign nasopharyngeal epithelium expressed predominantly moderate to intense nuclear expression of DNA-PKcs with few samples showing cytoplasmic staining (Plate 1A-F). The malignant lesions also showed moderate to intense nuclear positivity with diffuse staining throughout the nucleus. Lesions showing higher expression than the mean value in benign lesions +1 SD (14% in this case, the arbitrary cut off for normal expression) were taken as cases of over expression. Based on this cut off, Chi-square analysis did not yield any significant difference in the number of lesions over expressing the protein in cancer lesions taken as a whole (23% in benign lesions vs 24% in Cancer). The percentage of positive cells were, however, higher in the cancer lesions than that in the benign lesions (range: 20-100%, p=0.01). Detectable levels of DNA-PKcs were also detected in 28% cancer cases by western blot (Figure 1).

b. Effect of viral involvement

Opposite effects on expression of this protein were noticed in the lesions infected by HPV and EBV. The percentage over expression was seen in 45.2% cases in HPV positive lesions in contrast to 16.9% cases in the EBV positive lesions. The percentage of positive cells was higher (p<0.001) in the HPV+ve lesions when compared to HPV–ve lesions while a lower percentage cellular positivity was observed in the EBV+ve lesions when compared to EBV-ve lesions (p=0.047) (Table 1).

Table 2. Summary of Virus infection in NPC patients

Groups	EBV + (%)	HPV + (%)	HPV+/EBV+ (%)	Co-infection (%)
NPC (103)	65 (63)	31 (30)	85 (82.5)	11 (13)
WHO Type I (8)	2 (25)	5 (62.5)	7 (87.5)	0
WHO Type II (56)	30 (53.5)	19 (34)	41 (73.2)	8 (14.3)
WHOType III (39)	33 (85)	7 (18)	37 (95)	3 (8)
Controls (26)	7 (27)	1 (4)	8 (31)	0

Plate 1. Immunohistochemical staining of NPC lesions showing staining with anti- DNA-PKcs antibody (Santa Cruz) X 400.

A. Benign Nasopharyngeal epithelium showing low nuclear positivity

B. WHO type I lesion showing nuclear positivity (400X)

C. WHO type II lesion showing nuclear positivity (400X)

D. WHO type III lesions showing decreased nuclear positivity (400X)

E. EBV+ WHO type II lesion with low percentage of nuclear positivity (400X)

F. HPV+ WHO type II lesion showng high nuclear positivity (400X

Figure 1. Western Blots of study samples showing expressions of (a) DNA-PKcs, (b) ATM and (c) b-Actin proteins. Samples 1, 2, 3& 7 – HPV+ NPC lesions, 4 & 8 – EBV+ NPC lesions, 5 & 6 – Adenoid (benign) lesions

.

c. Correlation with clinico-pathological variables

The expression of DNA-PKcs were lower in the advanced stages of the disease when compared to stage I disease but no difference in expression was noticed between the different histological grades (Table 1). This observation is to be confirmed further as the number of samples in WHO I type is very small. However, this goes with the finding that DNA-PKcs is over expressed in HPV +ve lesions and 4 of the five lesions in WHO I are HPV +ve.

d. Correlation with recurrence and response to treatment

Expression of DNA-PKcs was lower in lesions (p=0.006) showing no recurrence and good overall response (no residual disease, metastasis or recurrence) to radiation treatment in comparison to that in the lesions showing recurrence and poor treatment response (Table 3). High expression of DNA-PKcs, on the other hand, was positively associated with recurrence (r=0.402, p=0.001) and poor treatment response (Table 3). Most HPV+ve lesions showed high expression while majority of the EBV+ve lesions showed low expression of this protein correlating very well with recurrence pattern (80% of HPV+ve tumours in contrast to 40% of EBV+ve tumours showing recurrence) Kaplan Meyer survival analysis showed that the expression of DNA-PKcs was closely associated with disease free survival with majority of the EBV positive lesions showing a lower expression (Table 4, Figure 2) having a better overall survival. These observations point towards the potential use of DNA-PKcs expression levels along with the viral status as a marker for radiation response and disease free survival in NPCs.

ii. Expression of ATM

a. Benign vs NPC

The benign lesions showed very low levels of ATM expression in contrast to higher levels in the malignant lesions (p<0.001, Table 1 Plate 2 A-F). Most of the benign lesions expressed cytoplasmic positivity. In NPC lesions, the expression of ATM protein was mainly nuclear with some cases showing both intense nuclear and mild to moderate cytoplasmic positivity. Increased expression of this protein was observed in all the less differentiated histological grade lesions and different stages of the disease (Table 1). No significant difference was observed in ATM protein expression in node positive and negative lesions suggesting that the expression of this protein is not directly related to invasion process.

b. Effect of viral involvement

Lesions showing higher expression than the arbitrarily fixed cut off of mean value in benign lesions +1 SD (1.84% in this case) were taken as cases of over expression. Accordingly, Chi-square analysis show EBV infection to be associated with a lower expression of this protein (84.5% of the lesions showing <1.84% positive cells being EBV+ve. p<0.001).

Figure 2. Survival graph showing the difference in survival in DNA-PKcs over expressing and those with low or normal expression group among NPC patients.

Table 3. Expression of proteins in relation to treatment response

Variables	Treatment Response
	-Poor (Mean ± S.E).	-Good (Mean ± S.E).
ATM(total) Virus – ve EBV alone + ve HPV alone +ve	16.60 ± 5.60 2.00 ± 2.00 1.50 ± 1.50 43.75 ± 12.38*§	1.42 ± 1.40* 0.00^# 0.00^# 25.00 ± 25.00*§
DNA-Pkcs (total) Virus – ve EBV alone + ve HPV alone +ve	23.96 ± 6.70 6.00 ± 6.00 12.60 ± 7.03 46.25 ± 14.99*§	2.85 ± 2.03** 12.50 ± 8.39 0.00^# 0.00^#
	Recurrence +ve (Mean ± S.E).	Recurrence –ve (Mean ± S.E).
DNA-PKcs(total) EBV-ve EBV+ve	36.47 ± 9.07 27.04 ± 7.06	25.83 ± 16.9 0.0^#
HPV-ve HPV+ve	17.87 ± 6.38 49.22 ± 8.43**	3.00 ± 3.00 19.00 ± 19.00
ATM(total) EBV-ve EBV+ve	27.89 ± 7.59 14.34 ± 5.16	8.33 ± 8.00 0.0^#
HPV-ve HPV+ve	0.0 34.83 ± 7.8**	9.70 ± 4.24 10.00 ± 10.00

Table 4. Details of survival analysis of DNA-PKcs showing relation to disease-free survival and viral status.

Variables	Protein Status	Mean survival time ± S.E.	95% CI	%Censored	p value (Log-rank test)
Total DNA-PKcs	-	48.57 ± 3.71	41.30, 55.84	76.74	0.0180
	+	26.60 ± 4.66	17.46, 35.73	52.94
Stage 1 DNAPKcs	-	45.84± 4.04	37.92, 53.77	75.00
	+	19.00± 1.04	17.04, 20.96	78.00
Stage II DNAPKcs	-	41.25± 6.71	28.10, 54.40	80.00
	+	24.84± 5.64	22.40, 28.75	75.00
Stage III DNAPKcs	-	34.57± 2.64	19.40, 29.75	57.14
	+	13.50± 1.25	11.05, 15.95	50.00
Stage IV DNAPKcs	-