Abstract

Background

Elevated levels of antibodies against antigens in the Epstein–Barr virus (EBV) lytic phase are important predictive markers for nasopharyngeal carcinoma (NPC) risk. Several lifestyle factors, including smoking, have also been associated with NPC risk. We hypothesized that some specific lifestyle factors induce transformation of EBV from the latent to the lytic stage and contribute to NPC occurrence.

Methods

We conducted a case–control study using data from male case patients (n = 1316) and control subjects (n = 1571) living in Guangdong Province, an area in China at high risk for NPC, to study potential NPC risk factors and EBV inducers. Two independent healthy male populations from a second high-risk area (n = 1657) and a low-risk area (n = 1961) were also included in the analysis of potential EBV inducers using logistic regression models. In vitro assays were performed to investigate the effect of cigarette smoke extract on EBV activation in two EBV-positive cell lines. All statistical tests were two-sided.

Results

Smoking was associated with an increased risk of NPC among the Guangdong participants with 20–40 and 40 or more pack-years vs never smokers (OR = 1.52, 95% CI = 1.22 to 1.88 and OR = 1.76, 95% CI = 1.34 to 2.32, respectively; Ptrend < .001). Smoking was the only factor linked to EBV seropositivity among the expanded control group and the independent low-risk population. In vitro experiments showed that cigarette smoke extract promoted EBV replication, induced the expression of the immediate-early transcriptional activators Zta and Rta , and increased transcriptional expression levels of BFRF3 and gp350 in the lytic phase.

Conclusion

Smoking is not only associated with NPC risk in individuals from China but is also associated with EBV seropositivity in healthy males and is involved in EBV activation.

Nasopharyngeal carcinoma (NPC) is rare in most populations around the world, but it is common in South China and Southeast Asia where the incidence can reach 20–50 cancers per 100 000 person-years ( 1–5 ). Epstein–Barr virus (EBV) is associated with the occurrence and development of NPC, and several specific serologic biomarkers have been identified and widely used for screening and early diagnosis of NPC in high-risk populations in some endemic areas such as South China ( 6–8 ), which include antibodies against antigens in the EBV lytic phase, ie, IgA antibodies against EBV capsid antigen (VCA-IgA) and Zta-IgA ( 9–14 ). A few long-term prospective studies have reported that elevated levels of antibodies against EBV antigens in serum precede the development of NPC and might have a causal role in the development of NPC, although there are other causal etiological factors that might play a role ( 9 ). Also, the molecular mechanism behind EBV seroreactitivity is largely unknown.

EBV establishes a life-long persistent infection in more than 90% of the population worldwide ( 15 ). The virus remains in a latent phase in healthy individuals carrying the virus with a low copy number of episomal virus maintained in resting memory B cells ( 16 , 17) . The virus can be periodically reactivated throughout an individual’s lifetime during periods of environmental stress. No specific natural inducer of EBV reactivation has been well defined although studies have shown that several different chemicals, such as phorbol-12-myristate-13-acetate (TPA) ( 18–20 ), sodium butyrate ( 19 , 21) , and trichostatin A ( 22 ), have been found to induce the EBV lytic cycle. In addition, a few studies have searched for inducers of EBV reactivation among environmental factors in human life. Two previous in vitro studies have observed that Cantonese-style salted fish ( 23 ) and Chinese herbs ( 24 ) may contain EBV-inducing substances, but these observations have not been made using epidemiological data.

After EBV is activated, it switches from a latent phase to a lytic phase. This switch is marked by the ability of serological tests to identify a series of antibodies against viral antigens expressed exclusively in the lytic phase. Previous studies have shown that elevated antibody titers against various EBV lytic gene products ( 25 ) and increasing viral DNA loads ( 26 ) have been detected in the serum of NPC patients, indicating that the lytic cycle and virus replication have been reactivated ( 19 ). Antibodies in the lytic stage of EBV are markedly elevated in NPC patients and high-risk individuals. However, the cause of this phenomenon has not yet been fully understood, and the etiological mechanism remains unclear. The switch from the latent phase to the lytic phase may be one of the crucial steps in the course of NPC development ( 27 ).

Previous epidemiological studies have documented that consumption of salted fish, consumption of other preserved foods, and smoking tobacco may be associated with elevated NPC risk ( 28 ). We hypothesized that specific factors among the confirmed or potential risk factors of NPC may act as environmental inducers to initiate the switch from the EBV latent stage to a lytic stage, therefore resulting in a series of elevated antibodies against viral antigens in the lytic phase.

To investigate the epidemiological relationship behind transformation of EBV from the latent to the lytic stage and NPC, we focused our study on eight suspected risk factors for NPC, including family history of NPC, smoking, and consumption of alcohol, tea, Chinese herbal tea, Cantonese slow-cooked soup, salted fish, and preserved vegetables. In vitro experiments in human EBV-positive cell lines were done to determine the molecular mechanism behind this transformation.

Materials and Methods

Study Populations

The study was reviewed and approved by the Human Ethics Approval Committee at Sun Yat-sen University Cancer Center. Written informed consent was obtained from each study subject. Characteristics of the study populations are summarized in Supplementary Table 1 (available online).

Guangdong Case–Control Study and 21-RCCP Population.

A large-scale case–control study population has been previously described ( 29–31 ). Briefly, NPC case patients were identified from Sun Yat-sen University Cancer Center in Guangzhou of Guangdong province, China between October 1, 2005 and October 1, 2007. During the same period, healthy control subjects were recruited from physical examination centers in 21 regions in Guangdong Province (defined as 21 regions collected from the Cantonese population and abbreviated hereafter as [21-RCCP]) and matched to NPC case patients by geographical area, age (±5 years), and sex. All subjects were asked to complete an in-person interview. Demographic data including age, education, and family history of NPC in first-degree relatives were collected. Also, data on eight putative risk factors including family history of NPC, smoking, consumption of alcohol, tea, Chinese herbal tea, Cantonese slow-cooked soup, salted fish, and preserved vegetables were collected. The lifestyle questionnaire ( Supplementary Materials , available online) was based on questionnaires issued by The University of Arizona Cancer Center ( 32 ) and Lifetime Drinking History issued by the European Monitoring Centre for Drugs and Drug Addiction ( 33 ). In detail, the dietary questionnaire was based on the Arizona Food Frequency Questionnaire ( 32 ); the tea questionnaire was based on the Arizona Tea Frequency Questionnaire ( 32 ); the smoking questionnaire was based on the Arizona Smoking Assessment Questionnaire ( 32 ). Alcohol consumption was measured using the Lifetime Drinking History ( 33 ) and Arizona Food Frequency Questionnaire ( 32 ). Regarding the intake of certain foods (salted fish, preserved vegetables, and slow-cooked soup) and drinks (tea and herbal tea), subjects were asked to choose from three intake frequency categories of less than monthly, monthly, and weekly or more.

For cigarette smoking, individuals who had smoked at least 100 cigarettes in their lifetime were defined as ever smokers, including current smokers and ex-smokers. Ex-smokers were defined as those subjects who had quit smoking more than 1 year before the interview. The cumulative amount of cigarette smoking was defined in pack-years by multiplying the number of packs of cigarettes smoked per day by the number of years of smoking. We investigated various cigarette smoking categories as follows: age at which the individual started smoking, the cumulative amount, and type of smoking (defined as filtered or nonfiltered cigarettes).

The consumption of alcohol was recorded as the average amount of all alcoholic beverages (including beer, wine, and liquor) that participants consumed during each time period in which regular drinking was consistent. A medium serving size (one drink) was defined as 12 ounces of beer, 5 ounces of wine, or 1.5 ounces of liquor ( 34 ).

A total of 1871 NPC patients and 2275 healthy control subjects were enrolled, and the response rates were 96% for the case patients and 66% for the control subjects. For the 21-RCCP study, 1316 male case patients ranging from age 14 to 80 years (mean age = 46.3 years) and 1571 male control subjects ranging from age 14 to 80 years (mean age = 46.9 years) were analyzed. Fifty of these case patients and 45 of the healthy control subjects were excluded from the analysis because of missing data.

Shihui and Yangquan Populations.

To confirm the observations made in the 21-RCCP, we extended our study to two additional independent healthy populations. First, in Sihui City, Guangdong Province in a NPC endemic area (defined as Sihui population), participants were enrolled from September 1 to December 31, 2008. The subjects included in the Sihui population were healthy male residents from several administrative villages in Sihui City, and the educational levels of all study subjects were at the high school level or below. Initially, 1674 healthy males were enrolled, but 17 subjects were excluded for missing data. The response rate of the Sihui population was 72%, and 1657 healthy subjects ranging from age 18 to 67 years with a mean age of 45.5 years were included in the analysis.

Data from a second independent healthy population from Yangquan City, Shanxi Province in a NPC nonendemic area (defined as Yangquan population), were collected between March 1 and August 1, 2010. For the Yangquan population, the subjects were healthy male residents who attended the First General Hospital of Yangquan City for health examinations. Initially, 2016 healthy males were recruited, but 55 subjects were excluded from the study because of missing data. The response rate was 67%. Overall, data from 1961 healthy subjects ranging from age 16 to 87 years with a mean age of 47.0 years were analyzed.

All of the participants answered a structured questionnaire including demographic information and potential risk factors during a face-to-face interview. Three factors that are specific for Cantonese (21-RCCP and 1403 subjects of the Sihui population) were not included in the Yangquan population, including consumption of Cantonese herbal tea, slow-cooked soup, and salted fish. Only healthy males were included in the analysis because of the exceptionally low percentage of females who have reported their smoking history (less than 2% in the female population).

VCA-IgA Antibody Titers of EBV in Blood Specimens

For all three populations (the Guangdong, Shihui, and Yangquan populations), 5–10mL of each subject’s blood was collected. The VCA-IgA antibody titers of EBV were measured using a commercial kit (Zhongshan Bio-tech Co Ltd, Zhongshan City, China) based on standard immunoenzymatic techniques, which were described previously ( 35 , 36) . In brief, the protocol steps are as follows: 1) B95-8 cell smears were prepared and aliquoted in the wells of slides to serve as antigen; 2) diluted sera were added to the wells; 3) the sera were incubated for 30 minutes at 37°C; 4) wells were washed with phosphate buffered saline three times; 5) peroxidase-conjugated antihuman IgA antibody (Zhongshan Bio-tech Co Ltd) was added, the wells were incubated at 37°C for 1 hour, and then washed three times with phosphate buffered saline; 6) the slides were flooded with 0.3mg/mL aminoethylcarbazole solution (Sigma-Aldrich Co, St. Louis, MO) and 30% H 2 O 2 (Guangzhou Chemical Reagent Factory, Guangzhou City, China) for 15 minutes. The slides were then examined under a light microscope (10×). Brown staining was considered positive. To minimize bias, samples were blinded to laboratory personnel. Serial dilutions of quality control sera (1:10, 1:20, 1:40, 1:80) were applied to each assay for the evaluation of intraset variability. To minimize experimental error, all tests were conducted in the same laboratory by the same technicians.

Serum Concentration of Total IgA

The serum concentration of total IgA was measured by immunoturbidimetry method using automatic analyzer 7600-020 (Hitachi Ltd, Tokyo, Japan) in the clinical laboratory of SYSUCC. The kits used in the assay were supplied by Sichuan Maker Biotechnology Co Ltd (Chengdu City, China).

Cell Culture

EBV-positive Akata, EBV-negative Akata ( 37 ), CNE2, and B95-8 cell lines were cultured as described in the Supplementary Methods (available online). The previous research reported that in vitro latently infected EBV-harboring cells could be reactivated to the lytic replication phase by treatment with TPA ( 20 ). The green fluorescent protein (GFP)-EBV-positive Akata cells start to produce EBV efficiently after cross-linking of cell surface immunoglobulin G (IgG) using anti-IgG antibodies ( 38 , 39) . Treatment with 20ng/mL TPA (stock dilution was prepared using ethanol) (Sigma-Aldrich) and 100 μg/mL rabbit anti-human IgG (Dako-Cytomation, Carpinteria, CA) served as positive controls for induction of EBV reactivation.

Preparation and Characterization of CSE

The CSE preparation process and determination of nicotine level in the CSE by high-performance liquid chromatography are summarized in the Supplementary Methods and Results and Supplementary Figure 1 (available online).

Figure 1.

The relationship between cigarette smoking and EBV VCA-IgA antibody serostatus among current smokers and ex-smokers. Odds ratios (ORs) for the VCA-IgA seropositive vs seronegative current smokers and ex-smokers were calculated. Because of the small number of ex-smokers, we stratified smoking categories into two instead of three levels for age: started smoking (<20, ≥20 years) and cumulative amount (<20, ≥20 pack-years). Both groups were adjusted for age (years, continuous variable) and education (high school or less, university or more) using logistic regression analysis with never smokers as a reference ( ref. ), and the tests were two-sided. Squares = study-specific odds ratios; horizontal lines = study-specific confidence intervals (CIs); and solid vertical line = odds ratio of 1.0. EBV VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

Figure 1.

The relationship between cigarette smoking and EBV VCA-IgA antibody serostatus among current smokers and ex-smokers. Odds ratios (ORs) for the VCA-IgA seropositive vs seronegative current smokers and ex-smokers were calculated. Because of the small number of ex-smokers, we stratified smoking categories into two instead of three levels for age: started smoking (<20, ≥20 years) and cumulative amount (<20, ≥20 pack-years). Both groups were adjusted for age (years, continuous variable) and education (high school or less, university or more) using logistic regression analysis with never smokers as a reference ( ref. ), and the tests were two-sided. Squares = study-specific odds ratios; horizontal lines = study-specific confidence intervals (CIs); and solid vertical line = odds ratio of 1.0. EBV VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

Flow Cytometry

Exponentially growing EBV-negative and EBV-positive Akata cell lines were treated with TPA (20ng/mL), anti-IgG (100 µg/mL), CSE (0.5%–2%), or were untreated. The cells were washed twice with phosphate buffered saline at 24 or 48 hours posttreatment and were then evaluated by flow cytometry (FACS Diva Option, Becton Dickinson, Mountain View, CA) using signal wave length analysis.

Immunofluorescence Analysis

Immunofluorescence analysis was performed as a modification of the method described previously ( 40 ). EBV capsid antigen expression in the cell lines was evaluated by indirect immunofluorescence following sequential incubation with a serum mixture obtained from ten NPC patients with high VCA-IgA titers, a polyclonal rabbit anti-human IgA (1:100) (Dako, Glostrup, Denmark), and a cyc3-conjugated anti-rabbit IgG (1:1000) (Jackson Laboratory, West-Grove, PA). Cellular expression of Zta or Rta was assessed using a monoclonal antibody specific for the Zta protein (1:40) (Dako) or anti-Rta (1:40) (Argene, Varilhes, France) and rhodamine-conjugated goat anti-mouse IgG (1:1000) (Jackson Laboratory). After three washes, the slides were stained with 0.5 µg/mL 4-,6-diamidino-2-phenylindole (1mg/mL stock prepared in deionized water; Sigma-Aldrich) to visualize the nuclei and were examined using an Olympus confocal imaging system (Olympus FV1000) at 40× objective (oil immersion). Images were then analyzed via the Olympus FV 1000 software (Olympus, Fluoview, version 1.7, Tokyo, Japan). Experiments were performed in triplicate.

Quantitative Reverse-Transcription Polymerase Chain Reaction (qRT-PCR) Analysis

Cell lines treated with anti-IgG, TPA, or CSE or untreated cell lines were harvested at the indicated times. qRT-PCR was performed as described previously ( 41 ). Sequence data for the EBV genome were acquired from the GenBank sequence database (accession number V01555). The primers used for real-time RT-PCR to detect transcriptional levels of BZLF1, BRLF1, BFRF3, and gp350/220 are listed in Supplementary Table 2 (available online). Also, the primers used for qRT-PCR to detect the relative amount of EBV DNA are also shown in Supplementary Table 2 (available online). qRT-PCR and data collection were performed on the ABI Prism 7900HT Sequence Detection System (Applied Biosystems Inc, Foster City, CA). The data were analyzed using the RQ Manager Software (version 1.2, Applied Biosystems Inc) system.

Plasmid Construction, Transfection, and Luciferase Assay

The entire BRLF1 promoter sequence of the EBV genome from B95-8 was amplified with primers (5′-primer: 5′-GGCTCGAGA AAAAGGCCGGCTGACAT GGATTACTGG-3′; 3′-primer: 5′-GGGAGCTCTAAAGGCCCTGTCGTCGGGA GATAAAAGTG-3′) to generate a fragment from nucleotide 107144 to 106177 upstream of the starting codon ATG with XhoI and SacI sites at each end, respectively ( 42 ). The amplified PCR fragment was cloned into the promoterless pGL3-basic luciferase reporter plasmid (Promega, Madison, WI) to generate the pGL3-BRLF1-promoter. Genomic DNA from B95-8 cells was used as a template. Plasmids were prepared from Escherichia coli by CsCl–ethidium bromide centrifugation ( 43 ). Transfection of plasmids into CNE2 was performed by using the FuGENE HD transfection reagent (Roche, Mannheim, Germany) by following the manufacturer’s protocol. Cells were treated with 20ng/mL TPA or 2% CSE added to the culture medium 5 hours after transfection. The BRLF1 promoter luciferase assay was performed with the Dual-Luciferase Reporter Assay System (Promega) as per the manufacturer’s instructions.

Statistical Analyses

Throughout data processing, double entry of EpiData Entry software (EpiData Association, http://www.epidata.dk ) was used. Logistic regression analysis was used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) after adjusting for age (years, continuous variable) and education (high school or less, university or more). Linear trend tests for associations between exposure and EBV VCA-IgA antibodies were performed for continuous variables including smoking, alcohol, tea, herbal tea, salted fish, and preserved vegetables. The possible synergistic effect of smoking status and having lived in high/low-risk NPC areas was evaluated using the method proposed by Rothman ( 44 ). The synergism index ( S ) was calculated by the following equation: S = (OR 11 − 1)/(OR 01 + OR 10 − 2). STATA 10.0 (Stata Corp, College Station, TX) software and SAS 9.1 (SAS Institute, Cary, NC) were used.

The statistical power was calculated using software PASS 2008 (NCSS, LLC, Kaysville, UT) based on three parameters including α (0.05), relevant sample size, and effect size. For the analysis of the relationship between eight risk factors and NPC risk, all power values were greater than 0.999. Whereas for the analysis of the relationship between the factors and seropositivity of EBV VCA-IgA antibodies, power of greater than 0.999 was achieved for the factor of smoking. However, power ranged from 0.080 to 0.344 for other factors, suggesting that under the test level and sample sizes (α = 0.05, n = 3228), we could not determine the statistical significance. All P values were calculated using two-sided tests, and P values less than .05 were considered to be statistically significant.

Results

Study of Risk Factors Related to NPC Risk in the Guangdong Case–Control Population

Individuals who reported a high cumulative amount of smoking (20–40 or ≥40 pack-years) had an increased risk of NPC (OR = 1.52, 95% CI = 1.22 to 1.88 and OR = 1.76, 95% CI = 1.34 to 2.32, respectively) compared with never smokers after adjustment for age and education ( Ptrend < .001). Only 4 of 1316 patients were diagnosed with differentiated NPC (World Health Organization type I); therefore, the histopathological type was not considered in the analysis.

The relationship between VCA-IgA antibody status and eight putative or potential risk factors with NPC risk was investigated. We observed that VCA-IgA positivity was statistically significantly associated with an increased risk of NPC (OR = 144, 95% CI = 103 to 202, P < .001) ( Table 1 ). Consumption of salted fish and preserved vegetables (monthly, weekly or more) and a family history of NPC were also associated with an increased risk for NPC ( P < .001), whereas consumption of tea or slow-cooked soup (monthly, weekly or more) were both associated with a statistically significantly reduced risk of NPC ( P < .001). However, although consumption of herbal tea monthly or one or fewer alcoholic beverages per day was associated with a decreased risk of NPC ( P < .001), increased levels of consumption of herbal tea or alcohol were not statistically significantly associated with NPC risk ( Ptrend = .58 and .31, respectively). These findings are consistent with previous reports of the relationship between smoking, consumption of alcohol, tea, Chinese herbal tea, Cantonese slow-cooked soup, salted fish, and preserved vegetables and NPC ( 29–31 ).

Table 1.

Multivariable analyses of Epstein–Barr virus (EBV), family history of nasopharyngeal cancer, and environmental factors in Guangdong men

Variable No. of case patients No. of control subjects OR (95% CI)* P* 
EBV VCA-IgA     
    Negative 44 1296 1.00 (referent)  
    Positive 1272 275  144 (103 to 202) <.001 
Family history of NPC     
    No 1157 1517 1.00 (referent)  
    Yes 146 54 3.50 (2.52 to 4.85) <.001 
Cigarette smoking, pack-years     
    Never 357 519 1.00 (referent)  
    <20 382 576 0.86 (0.71 to 1.05)   .13 
    <40 358 311 1.52 (1.22 to 1.88) <.001 
    ≥40 192 159 1.76 (1.34 to 2.32) <.001 
    Ptrend    <.001 
Alcohol     
    Nondrinker 615 627 1.00 (referent)  
    ≤1 drink per day 487 745 0.70 (0.59 to 0.83) <.001 
    >1 drink per day 210 199 1.09 (0.87 to 1.37)   .47 
    Ptrend      .31 
Tea     
    Less than monthly 351 257 1.00 (referent)  
    Monthly 246 329 0.58 (0.46 to 0.73) <.001 
    Weekly or more 697 982 0.55 (0.45 to 0.67) <.001 
    Ptrend    <.001 
Herbal tea     
    Less than monthly 312 293 1.00 (referent)  
    Monthly 591 867 0.64 (0.53 to 0.78) <.001 
    Weekly or more 398 394 0.89 (0.72 to 1.11)   .30 
    Ptrend      .58 
Slow-cooked soup     
    Less than monthly 199 62 1.00 (referent)  
    Monthly 162 154 0.33 (0.23 to 0.48) <.001 
    Weekly or more 933 1338 0.24 (0.18 to 0.32) <.001 
    Ptrend    <.001 
Salted fish     
    Less than monthly 1046 1398 1.00 (referent)  
    Monthly 150 80 2.50 (1.87 to 3.33) <.001 
    Weekly or more 115 85 1.74 (1.29 to 2.35) <.001 
    Ptrend    <.001 
Preserved vegetable     
    Less than monthly 957 1364 1.00 (referent)  
    Monthly 194 96 2.82 (2.17 to 3.67) <.001 
    Weekly or more 163 102 2.22 (1.70 to 2.90) <.001 
    Ptrend    <.001 
Variable No. of case patients No. of control subjects OR (95% CI)* P* 
EBV VCA-IgA     
    Negative 44 1296 1.00 (referent)  
    Positive 1272 275  144 (103 to 202) <.001 
Family history of NPC     
    No 1157 1517 1.00 (referent)  
    Yes 146 54 3.50 (2.52 to 4.85) <.001 
Cigarette smoking, pack-years     
    Never 357 519 1.00 (referent)  
    <20 382 576 0.86 (0.71 to 1.05)   .13 
    <40 358 311 1.52 (1.22 to 1.88) <.001 
    ≥40 192 159 1.76 (1.34 to 2.32) <.001 
    Ptrend    <.001 
Alcohol     
    Nondrinker 615 627 1.00 (referent)  
    ≤1 drink per day 487 745 0.70 (0.59 to 0.83) <.001 
    >1 drink per day 210 199 1.09 (0.87 to 1.37)   .47 
    Ptrend      .31 
Tea     
    Less than monthly 351 257 1.00 (referent)  
    Monthly 246 329 0.58 (0.46 to 0.73) <.001 
    Weekly or more 697 982 0.55 (0.45 to 0.67) <.001 
    Ptrend    <.001 
Herbal tea     
    Less than monthly 312 293 1.00 (referent)  
    Monthly 591 867 0.64 (0.53 to 0.78) <.001 
    Weekly or more 398 394 0.89 (0.72 to 1.11)   .30 
    Ptrend      .58 
Slow-cooked soup     
    Less than monthly 199 62 1.00 (referent)  
    Monthly 162 154 0.33 (0.23 to 0.48) <.001 
    Weekly or more 933 1338 0.24 (0.18 to 0.32) <.001 
    Ptrend    <.001 
Salted fish     
    Less than monthly 1046 1398 1.00 (referent)  
    Monthly 150 80 2.50 (1.87 to 3.33) <.001 
    Weekly or more 115 85 1.74 (1.29 to 2.35) <.001 
    Ptrend    <.001 
Preserved vegetable     
    Less than monthly 957 1364 1.00 (referent)  
    Monthly 194 96 2.82 (2.17 to 3.67) <.001 
    Weekly or more 163 102 2.22 (1.70 to 2.90) <.001 
    Ptrend    <.001 

* Odds ratios (ORs) and P values were calculated using logistic regression analysis and adjusted for age (years, continuous variable) and education (high school or less, university or more). All tests were two-sided. CI = confidence interval; EBV VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

Relationship Between Risk Factors and EBV VCA-IgA Antibody Serostatus in High-Risk Areas for NPC

To identify potential risk factors for NPC that are associated with EBV, smoking, family history, consumption of alcohol, tea, herbal tea, slow-cooked soup, salted fish, and preserved vegetables were compared between EBV VCA-IgA antibody seropositive and seronegative groups. A cutoff value of 1:10 was used to differentiate between seropositive and seronegative subjects, and it has been widely used in the screening of high-risk individuals in the endemic regions of Guangdong, China in other studies ( 35 , 36 , 45) . Because we did not observe any heterogeneity between the 21-RCCP and the Sihui populations when we examined the effects of the factors on EBV serostatus ( Pheterogeneity = .25 to .97) (data not shown), we combined the subjects in the 21-RCCP and the Sihui population to maximize the statistical power. A total of 275 and 328 participants had a positive antibody titer (≥1:10), and a total of 1296 and 1329 participants had a negative antibody titer among the 21-RCCP and Shihui healthy populations, respectively ( Supplementary Table 1 , available online). We found that smoking was the only factor associated with seropositivity for the VCA-IgA antibody. Table 2 shows that individuals who started smoking at a younger age were associated with a higher risk of being seropositive for the VCA-IgA antibody when compared with never smokers (started smoking when age ≥ 30 years: OR = 1.83, 95% CI = 1.22 to 2.73; age 20–29 years: OR = 1.98, 95% CI = 1.53 to 2.58; age ≤ 20 years: OR = 2.14, 95% CI = 1.64 to 2.79; Ptrend < .001). Moreover, the likelihood of being seropositive for VCA-IgA antibody increased with the cumulative amount of smoking (OR = 1.89, 95% CI = 1.45 to 2.45; OR = 2.04, 95% CI = 1.54 to 2.70; OR = 2.74, 95% CI = 1.96 to 3.82 for <20, 20–40, and ≥40 pack-years, respectively; Ptrend < .001). With regard to the type of smoking, smoking filtered or nonfiltered cigarettes was associated with VCA-IgA antibody seropositivity (OR = 2.13, 95% CI = 1.57 to 2.90; OR = 2.01, 95% CI = 1.57 to 2.58, respectively).

Interestingly, the relationship between NPC risk and serum VCA-IgA antibody status among ex-smokers was weaker than that among current smokers (OR = 1.51, 95% CI = 1.09 to 2.11; OR = 2.17, 95% CI = 1.70 to 2.78, respectively; Ptrend < .001), compared with the never smokers ( Table 3 ). Strong associations between cigarette smoking and the VCA-IgA antibody serotype in current smokers were observed in all subgroups with odds ratios ranging from 2.02 to 2.43, and weaker associations were observed for ex-smokers with odds ratios ranging from 1.30 to 1.79 ( Figure 1 ). The risk of being seropositive for the VCA-IgA antibody for ex-smokers was much lower than that for current smokers (adjusted OR = 0.69, 95% CI = 0.52 to 0.91, P = .008).

Table 2.

The dose–response relationship between smoking and Epstein–Barr virus VCA-IgA antibodies among two populations in China

 Guangdong population *  Shanxi population Pooled 
Variable No of VCA-IgA- positive men No. of VCA-IgA- negative men OR (95% CI) †  P †  No of VCA-IgA- positive men No. of VCA-IgA- negative men OR (95% CI) †  P †  No of VCA-IgA positive men No. of VCA-IgA negative men OR (95% CI) †  P †  
Smoking             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    Ex-smoker 76 364 1.51 (1.09 to 2.11) .02 13 174 1.35 (0.68 to 2.66) .39 89 538 1.60 (1.19 to 2.15) .002 
    Current smoker 432 1512 2.17 (1.70 to 2.78) <.001 92 1062 1.64 (1.08 to 2.51) .02 524 2574 2.01 (1.63 to 2.47) <.001 
    Ptrend    <.001    .02    <.001 
Age started
    smoking, y             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    ≥30 44 173 1.83 (1.22 to 2.73) .003 94 0.58 (0.17 to 1.94) .38 47 267 1.69 (1.17 to 2.43) .005 
    20–29 232 887 1.98 (1.53 to 2.58) <.001 45 563 1.47 (0.92 to 2.36) .11 277 1450 1.90 (1.52 to 2.38) <.001 
    <20 232 816 2.14 (1.64 to 2.79) <.001 57 579 1.92 (1.22 to 3.04) .005 289 1395 2.03 (1.62 to 2.55) <.001 
    Ptrend    <.001    .003    <.001 
Cumulative
    amount,
    pack-years             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    <20 214 894 1.89 (1.45 to 2.45) <.001 47 640 1.41 (0.88 to 2.27) .15 261 1534 1.76 (1.41 to 2.21) <.001 
    <40 185 688 2.04 (1.54 to 2.70) <.001 33 404 1.50 (0.90 to 2.51) .12 218 1092 1.94 (1.53 to 2.46) <.001 
    ≥40 108 286 2.74 (1.96 to 3.82) <.001 24 180 2.52 (1.39 to 4.56) .002 132 466 2.67 (2.01 to 3.55) <.001 
    Ptrend    <.001    .004    <.001 
Type of
    smoking             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    Filtered
    cigarettes 365 1386 2.01 (1.57 to 2.58) <.001 103 1209 1.61 (1.06 to 2.43) .03 468 2595 1.82 (1.48 to 2.25) <.001 
    Nonfiltered
    cigarettes 126 422 2.13 (1.57 to 2.90) <.001 20 1.80 (0.40 to 8.17) .45 128 442 2.65 (2.00 to 3.49) <.001 
    Ptrend    <.001    .03    <.001 
 Guangdong population *  Shanxi population Pooled 
Variable No of VCA-IgA- positive men No. of VCA-IgA- negative men OR (95% CI) †  P †  No of VCA-IgA- positive men No. of VCA-IgA- negative men OR (95% CI) †  P †  No of VCA-IgA positive men No. of VCA-IgA negative men OR (95% CI) †  P †  
Smoking             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    Ex-smoker 76 364 1.51 (1.09 to 2.11) .02 13 174 1.35 (0.68 to 2.66) .39 89 538 1.60 (1.19 to 2.15) .002 
    Current smoker 432 1512 2.17 (1.70 to 2.78) <.001 92 1062 1.64 (1.08 to 2.51) .02 524 2574 2.01 (1.63 to 2.47) <.001 
    Ptrend    <.001    .02    <.001 
Age started
    smoking, y             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    ≥30 44 173 1.83 (1.22 to 2.73) .003 94 0.58 (0.17 to 1.94) .38 47 267 1.69 (1.17 to 2.43) .005 
    20–29 232 887 1.98 (1.53 to 2.58) <.001 45 563 1.47 (0.92 to 2.36) .11 277 1450 1.90 (1.52 to 2.38) <.001 
    <20 232 816 2.14 (1.64 to 2.79) <.001 57 579 1.92 (1.22 to 3.04) .005 289 1395 2.03 (1.62 to 2.55) <.001 
    Ptrend    <.001    .003    <.001 
Cumulative
    amount,
    pack-years             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    <20 214 894 1.89 (1.45 to 2.45) <.001 47 640 1.41 (0.88 to 2.27) .15 261 1534 1.76 (1.41 to 2.21) <.001 
    <40 185 688 2.04 (1.54 to 2.70) <.001 33 404 1.50 (0.90 to 2.51) .12 218 1092 1.94 (1.53 to 2.46) <.001 
    ≥40 108 286 2.74 (1.96 to 3.82) <.001 24 180 2.52 (1.39 to 4.56) .002 132 466 2.67 (2.01 to 3.55) <.001 
    Ptrend    <.001    .004    <.001 
Type of
    smoking             
    Never
    smoker 95 749 1.00 (referent)  33 587 1.00 (referent)  128 1336 1.00 (referent)  
    Filtered
    cigarettes 365 1386 2.01 (1.57 to 2.58) <.001 103 1209 1.61 (1.06 to 2.43) .03 468 2595 1.82 (1.48 to 2.25) <.001 
    Nonfiltered
    cigarettes 126 422 2.13 (1.57 to 2.90) <.001 20 1.80 (0.40 to 8.17) .45 128 442 2.65 (2.00 to 3.49) <.001 
    Ptrend    <.001    .03    <.001 

* The two populations of men from 21 regions collected from the Cantonese population and Sihui population combined.

† Odds ratios (ORs) with 95% confidence intervals (CIs) and P values were calculated using logistic regression analysis and adjusted for age (years, continuous variable) and education (high school or less, university or more). All statistical tests were two-sided. VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

Table 3.

The association between suspected risk factors for nasopharyngeal carcinoma and Epstein–Barr virus VCA-IgA antibody serostatus among healthy men in Guangdong*

 VCA-IgA antibody serostatus*   
Variable No. of positive men No. of negative men OR (95% CI) P 
Age, y     
    ≤30 21 126 1.00 (referent)  
    31–40 137 710 1.14 (0.69 to 1.87)   .61 
    41–50 198 882 1.30 (0.79 to 2.12)   .30 
    51–60 194 751 1.48 (0.91 to 2.42)   .12 
    ≥61 53 156 2.03 (1.16 to 3.55)   .01 
    Ptrend    <.001 
Family history of NPC     
    No 531 2331 1.00 (referent)  
    Yes 22 84 1.15 (0.71 to 1.86)   .56 
Smoking     
    Never smoker 95 749 1.00 (referent)  
    Ex-smoker 76 364 1.51 (1.09 to 2.11)   .02 
    Current smoker 432 1512 2.17 (1.70 to 2.78) <.001 
    Ptrend    <.001 
Alcohol     
    Nondrinker 248 1054 1.00 (referent)  
    ≤1 drink per day 212 1016 0.94 (0.76 to 1.15)   .53 
    >1 drink per day 93 351 1.12 (0.86 to 1.47)   .40 
    Ptrend      .64 
Tea     
    Less than monthly 127 593 1.00 (referent)  
    Monthly 108 472 1.13 (0.85 to 1.51)   .40 
    Weekly or more 315 1345 1.10 (0.87 to 1.38)   .44 
    Ptrend      .50 
Herbal tea     
    Less than monthly 88 411 1.00 (referent)  
    Monthly 311 1406 1.08 (0.83 to 1.41)   .56 
    Weekly or more 151 587 1.27 (0.94 to 1.70)   .12 
    Ptrend      .10 
Slow-cooked soup     
    Less than monthly 166 700 1.00 (referent)  
    Weekly or more 383 1702 0.96 (0.78 to 1.17)   .67 
Salted fish     
    Less than monthly 514 2204 1.00 (referent)  
    Monthly 21 129 0.69 (0.43 to 1.11)   .12 
    Weekly or more 17 79 0.84 (0.49 to 1.43)   .51 
    Ptrend      .19 
Preserved vegetable     
    Less than monthly 431 1930 1.00 (referent)  
    Monthly 93 355 1.17 (0.91 to 1.50)   .23 
    Weekly or more 28 126 0.96 (0.63 to 1.47)   .86 
    Ptrend      .58 
 VCA-IgA antibody serostatus*   
Variable No. of positive men No. of negative men OR (95% CI) P 
Age, y     
    ≤30 21 126 1.00 (referent)  
    31–40 137 710 1.14 (0.69 to 1.87)   .61 
    41–50 198 882 1.30 (0.79 to 2.12)   .30 
    51–60 194 751 1.48 (0.91 to 2.42)   .12 
    ≥61 53 156 2.03 (1.16 to 3.55)   .01 
    Ptrend    <.001 
Family history of NPC     
    No 531 2331 1.00 (referent)  
    Yes 22 84 1.15 (0.71 to 1.86)   .56 
Smoking     
    Never smoker 95 749 1.00 (referent)  
    Ex-smoker 76 364 1.51 (1.09 to 2.11)   .02 
    Current smoker 432 1512 2.17 (1.70 to 2.78) <.001 
    Ptrend    <.001 
Alcohol     
    Nondrinker 248 1054 1.00 (referent)  
    ≤1 drink per day 212 1016 0.94 (0.76 to 1.15)   .53 
    >1 drink per day 93 351 1.12 (0.86 to 1.47)   .40 
    Ptrend      .64 
Tea     
    Less than monthly 127 593 1.00 (referent)  
    Monthly 108 472 1.13 (0.85 to 1.51)   .40 
    Weekly or more 315 1345 1.10 (0.87 to 1.38)   .44 
    Ptrend      .50 
Herbal tea     
    Less than monthly 88 411 1.00 (referent)  
    Monthly 311 1406 1.08 (0.83 to 1.41)   .56 
    Weekly or more 151 587 1.27 (0.94 to 1.70)   .12 
    Ptrend      .10 
Slow-cooked soup     
    Less than monthly 166 700 1.00 (referent)  
    Weekly or more 383 1702 0.96 (0.78 to 1.17)   .67 
Salted fish     
    Less than monthly 514 2204 1.00 (referent)  
    Monthly 21 129 0.69 (0.43 to 1.11)   .12 
    Weekly or more 17 79 0.84 (0.49 to 1.43)   .51 
    Ptrend      .19 
Preserved vegetable     
    Less than monthly 431 1930 1.00 (referent)  
    Monthly 93 355 1.17 (0.91 to 1.50)   .23 
    Weekly or more 28 126 0.96 (0.63 to 1.47)   .86 
    Ptrend      .58 

* Odds ratios (ORs) and P values were calculated using logistic regression analysis and adjusted for age (years, continuous variable) and education (high school or less, university or more). All statistical tests were two-sided. CI = confidence interval; VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

We also evaluated the associations of the remaining seven risk factors and EBV serotype but did not detect any statistically significant findings ( Table 3 ). So far, among the eight factors that we focused on in our study, which were previously shown to be statistically significantly associated with NPC risk, smoking is the only factor associated with elevated titers of VCA-IgA antibody. In addition, associations between these eight factors and EBV serostatus were further adjusted for age, education, family history of NPC, and consumption of salted fish and preserved vegetables, and all the results were similar to the above results that included adjustment for age and education only (data not shown).

Cigarette Smoking and Seropositivity of VCA-IgA Antibodies in the Yangquan Population in a Low-Risk Area

All of the above findings prompted an investigation of whether or not the relationship was consistent between smoking habit and serostatus of the VCA-IgA antibody not only in populations in high-risk areas but also in low-risk areas for NPC. Thus, a third population in a low-risk area of Yangquan City in North China was investigated (Yangquan population). A total of 1961 subjects were recruited, including 138 subjects who were seropositive for VCA-IgA antibody, and the seropositive rate (7.0%) was much lower than that of the 21-RCCP (17.5%) and the Sihui population (19.8%). Nevertheless, a similar result that cigarette smoking was also related to the seropositivity of the VCA-IgA antibody in this area was observed, and it was also uniquely statistically significantly related to elevated titers of anti-VCA-IgA ( Table 2 ).

However, the association between cigarette smoking and an increased risk of VCA-IgA antibody seropositivity in the Guangdong high-risk population (21-RCCP and Sihui population) was higher than that in the low-risk population (Yangquan population) among ever smokers in Guangdong and Yangquan (OR = 5.09, 95% CI = 3.47 to 7.48; OR = 1.60, 95% CI = 1.06 to 2.42, respectively), compared with never smokers among the Yangquan population ( Table 4 ). The interaction effect (synergism “ S ” index) between the smoking status and regions of high-risk and low-risk areas for NPC on the VCA-IgA antibody seropositivity was 1.96 (95% CI = 1.30 to 2.97) when an additive model was used, but it was not statistically significant when a multiplicative model was used ( P = .15).

Table 4.

An interaction analysis of smoking status and Epstein–Barr virus VCA-IgA antibody serostatus among high- and low-risk male populations*

 Low-risk area (Yangquan population) High-risk area (21-RCCP and Sihui population) 
Smoking status No of VCA-IgA-
 positive men No. of VCA-IgA-
 negative men OR (95%CI) No of VCA-IgA-
 positive men No. of VCA-IgA-
 negative men OR (95%CI) 
Never smoker 33 587 1.00 (referent) 95 749 2.49 (1.63 to 3.80) 
Ever smoker 105 1236 1.60 (1.06 to 2.42) 508 1876 5.09 (3.47 to 7.48) 
Interaction effect (synergism S index)†  1.96 (95% CI = 1.30 to 2.97) 
 Low-risk area (Yangquan population) High-risk area (21-RCCP and Sihui population) 
Smoking status No of VCA-IgA-
 positive men No. of VCA-IgA-
 negative men OR (95%CI) No of VCA-IgA-
 positive men No. of VCA-IgA-
 negative men OR (95%CI) 
Never smoker 33 587 1.00 (referent) 95 749 2.49 (1.63 to 3.80) 
Ever smoker 105 1236 1.60 (1.06 to 2.42) 508 1876 5.09 (3.47 to 7.48) 
Interaction effect (synergism S index)†  1.96 (95% CI = 1.30 to 2.97) 

* Odds ratios (ORs) with 95% confidence intervals (CIs) and P values were calculated using logistic regression analysis and adjusted for age (years, continuous variable) and education (high school or less, university or more). All tests were two-sided, and P values less than .05 were considered to be statistically significant. 21- RCCP = 21-regions collected from the Cantonese population; VCA-IgA = IgA antibodies against Epstein–Barr virus capsid antigens.

† The synergism S index was calculated by equation S = (OR 11 − 1)/(OR 01 + OR 10 − 2).

Relationship of Cigarette Smoking and VCA-IgA Antibody Seropositivity Among the 21-RCCP, Sihui, and Yangquan Populations

To increase the power of detecting the association ( 46 ) between EBV serostatus and smoking, we jointly analyzed the data from the three study populations after finding no heterogeneity between the Guangdong (21-RCCP and Sihui population) high-risk populations and the low-risk Yangquan population ( Pheterogeneity = .15), except for one specific age group of those older than 55 years ( Pheterogeneity = .03). We observed a consistent trend indicating that as the amount of smoking increases, so does the likelihood of VCA-IgA antibody seropositivity among both the high-risk and low-risk populations ( Table 2 ). The level of statistical significance of the results for each subgroup from the pooled analysis was increased (the corresponding P values decreased dramatically); for example, for the subgroup consisting of individuals whose cumulative amount of smoking was 40 pack-years or more, the P value decreased from 1 × 10 −9 (in single analysis of Guangdong populations) to 1 × 10 −12 , suggesting that the power was increased.

Relationship Between Smoking and Serum Concentration of Total IgA

We assessed the relationship between smoking and general IgA levels. Serum samples positive for the VCA-IgA antibody (n = 100, including 51 ever smokers and 49 never smokers) as well as serum samples negative for the antibody (n = 100, including 50 ever smokers and 50 never smokers) were identified. We did not observe any difference in the serum concentration of total IgA between seropositive (mean = 1.87g/L, SD = 0.92g/L) and seronegative (mean = 1.67g/L, SD = 0.79g/L) groups ( P = .10). However, when the samples were divided into subgroups of ever smokers and never smokers, the serum concentration of total IgA in ever smokers (mean = 1.54g/L, SD = 0.82g/L) was lower than that in never smokers (mean = 2.01g/L, SD = 0.84g/L) ( P < .001), suggesting that smoking is not associated with increasing IgA levels in general.

CSE and EBV Replication in EBV-Positive B-Cell Lines

The results obtained from the epidemiological study encouraged us to investigate whether the association of smoking exposure with elevated VCA-IgA antibodies was attributed to a direct reactivation of a switch from latent to lytic EBV. Thus, we performed in vitro experiments in cell lines to test this hypothesis.

To elucidate the effect of CSE on EBV replication, we tested whether or not CSE promotes EBV replication in Akata and B95-8 cells. Akata cells were first treated with CSE. As shown in Figure 2 , A , the proportion of GFP-EBV-positive cells increased from 5.7% to 9.9%. Similarly, the intensity of GFP was enhanced in a dose-dependent manner following CSE treatment at both 24 and 48 hours ( P < .05) ( Figure 2 , B–D). Moreover, an additive effect in the enhancement of fluorescent intensity was observed in cells treated with both CSE and TPA compared with cells treated with TPA alone. The increased expression of GFP was consistent with the enhanced EBV replication. The amount of EBV DNA in Akata cells was increased after treatment with anti-IgG, TPA, or CSE for 24 hours compared with untreated cells ( P < .05) ( Figure 2 , E and F ). Similar results were obtained in B95-8 cells ( Figure 2 , E ). The relative number of EBV DNA copies increased as early as 4 hours after treatment with CSE and peaked after 12 hours; a 4-fold to 21-fold increase in the replication of EBV DNA was detected in Akata cells at different time points after CSE treatment ( Figure 2 , F ). Taken together, these results suggest that CSE promotes EBV replication in Akata and B95-8 cells.

Figure 2.

Effect of cigarette smoke extract (CSE) on Epstein–Barr virus (EBV) replication in EBV-positive B-cell lines. A ) The proportion of GFP-EBV-positive cells was determined by immunofluorescence after EBV-positive and -negative AKATA cells were treated with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or 2% CSE for 24 hours. The percentage of GFP-EBV-positive cells ( green ) was calculated relative to total cell populations ( blue ) in five random fields of view after treating the cells for 24 hours. Representative fields are shown. Scale bar = 100 µm. Data represent the mean and SD (n = 5). B ) A flow histogram overlay of flow cytometric detection of green fluorescent protein (GFP)-EBV-positive cells is shown. The fluorescence intensities of EBV-positive and -negative AKATA cells treated with 0%, 0.5%, 1%, 1.5%, or 2% CSE, 20ng/mL TPA, a combination of 0.5% CSE and 20ng/mL TPA, or 100 µg/mL anti-IgG were compared with those of untreated cells for 24 or 48 hours. Black lines represent the untreated cells and colored lines represent treated cells. C and D ) The relative fold increase of GFP expression among groups treated with 0%, 0.5%, 1%, 1.5%, or 2% CSE, 20ng/mL TPA, a combination of 0.5% CSE and 20ng/mL TPA, or 100 µg/mL anti-IgG for 24 or 48 hours is given. The relative fold increase of GFP expression was normalized to the amount for untreated GFP-EBV-positive Akata cells. E and F ) The relative fold increase of EBV DNA in the BamHI-W region of GFP-EBV-positive Akata and B95-8 cell lines that were untreated or treated with 100 µg/mL anti-IgG, 20ng/mL TPA, or 2% CSE after harvesting at the indicated times (0, 4, 8, 12, 24, 36, 48, and 72 hours) by real-time quantitative PCR are shown. β-globin served as the internal control. The data represent the mean and SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 compared with untreated cells. † P < .01 compared with untreated cells as the control.

Figure 2.

Effect of cigarette smoke extract (CSE) on Epstein–Barr virus (EBV) replication in EBV-positive B-cell lines. A ) The proportion of GFP-EBV-positive cells was determined by immunofluorescence after EBV-positive and -negative AKATA cells were treated with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or 2% CSE for 24 hours. The percentage of GFP-EBV-positive cells ( green ) was calculated relative to total cell populations ( blue ) in five random fields of view after treating the cells for 24 hours. Representative fields are shown. Scale bar = 100 µm. Data represent the mean and SD (n = 5). B ) A flow histogram overlay of flow cytometric detection of green fluorescent protein (GFP)-EBV-positive cells is shown. The fluorescence intensities of EBV-positive and -negative AKATA cells treated with 0%, 0.5%, 1%, 1.5%, or 2% CSE, 20ng/mL TPA, a combination of 0.5% CSE and 20ng/mL TPA, or 100 µg/mL anti-IgG were compared with those of untreated cells for 24 or 48 hours. Black lines represent the untreated cells and colored lines represent treated cells. C and D ) The relative fold increase of GFP expression among groups treated with 0%, 0.5%, 1%, 1.5%, or 2% CSE, 20ng/mL TPA, a combination of 0.5% CSE and 20ng/mL TPA, or 100 µg/mL anti-IgG for 24 or 48 hours is given. The relative fold increase of GFP expression was normalized to the amount for untreated GFP-EBV-positive Akata cells. E and F ) The relative fold increase of EBV DNA in the BamHI-W region of GFP-EBV-positive Akata and B95-8 cell lines that were untreated or treated with 100 µg/mL anti-IgG, 20ng/mL TPA, or 2% CSE after harvesting at the indicated times (0, 4, 8, 12, 24, 36, 48, and 72 hours) by real-time quantitative PCR are shown. β-globin served as the internal control. The data represent the mean and SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 compared with untreated cells. † P < .01 compared with untreated cells as the control.

Expression of Lytic-Phase Late Gene Products of EBV After CSE Treatment

As the rate of replication of EBV DNA in Akata and B95-8 cell lines was increased after treatment with CSE, we tested if the expression of lytic-phase late genes, including the VCA p18 encoded BFRF3 and gp350, was induced in CSE-treated cells. As shown in Figure 3 , A and B , the transcriptional levels of these genes in cells treated with anti-IgG, TPA, or CSE were elevated compared with untreated cells ( P < .05). CSE induced a time-dependent increase in BFRF3 and gp350 expression in Akata cells. Furthermore, the relative transcriptional levels of these two genes increased 12 hours after induction by CSE and reached a peak at 36 hours ( P < .01) ( Figure 3 , C and D ). In addition, immunofluorescence staining showed an increased population of VCA-positive cells after 48 hours of CSE treatment in both Akata (1.70-fold induction; P < .05) and B95-8 (1.76-fold induction; P < .01) cells ( Figure 3 , E–G). Together, these data suggest that CSE induces the expression of late lytic gene products of EBV after CSE treatment.

Figure 3.

Expression of lytic-phase late gene products of Epstein–Barr virus (EBV) after treatment with cigarette smoke extract (CSE). ( AD ) Expression of the viral lytic-phase late genes BFRF3 and gp350 was measured in green fluorescent protein (GFP)-EBV-positive Akata and B95-8 cell lines after treatment with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or 2% CSE and harvested at 0, 4, 8, 12, 24, and 36 hours. RNA was extracted from the cells, and quantitative real-time polymerase chain reaction was done. The relative fold increase of transcripts was normalized to the amount of RNA harvested from untreated GFP-EBV-positive Akata cells. β-actin served as the internal control. Data represent the mean and SD ( whisker bars ) from three independent experiments performed in triplicate. E ) Expression of virus capsid antigen protein (VCA) was detected by immunofluorescence after 2% CSE treatment for 48 hours. The percentage of cyc3-VCA-positive Akata EBV-positive and B95-8 cells shown in F and G ( red ) was calculated relative to the total cell number measured by 4-,6-diamidino-2-phenylindole (DAPI) staining ( blue ) in five random fields of view after treating the cells for 48 hours. Representative images are shown. Scale bar = 100 µm. Data represent the mean and SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 and † P < .01 compared with untreated cells as the control.

Figure 3.

Expression of lytic-phase late gene products of Epstein–Barr virus (EBV) after treatment with cigarette smoke extract (CSE). ( AD ) Expression of the viral lytic-phase late genes BFRF3 and gp350 was measured in green fluorescent protein (GFP)-EBV-positive Akata and B95-8 cell lines after treatment with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or 2% CSE and harvested at 0, 4, 8, 12, 24, and 36 hours. RNA was extracted from the cells, and quantitative real-time polymerase chain reaction was done. The relative fold increase of transcripts was normalized to the amount of RNA harvested from untreated GFP-EBV-positive Akata cells. β-actin served as the internal control. Data represent the mean and SD ( whisker bars ) from three independent experiments performed in triplicate. E ) Expression of virus capsid antigen protein (VCA) was detected by immunofluorescence after 2% CSE treatment for 48 hours. The percentage of cyc3-VCA-positive Akata EBV-positive and B95-8 cells shown in F and G ( red ) was calculated relative to the total cell number measured by 4-,6-diamidino-2-phenylindole (DAPI) staining ( blue ) in five random fields of view after treating the cells for 48 hours. Representative images are shown. Scale bar = 100 µm. Data represent the mean and SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 and † P < .01 compared with untreated cells as the control.

Expression of EBV Transcription Factors After Treatment With CSE

Elevated mRNA levels of VCA and gp350, in addition to elevated protein levels of VCA, presumably reflect EBV reactivation in GFP-EBV-positive Akata and B95-8 cells after CSE treatment. Thus, we investigated if CSE induces the expression of BZLF1, which encodes Zta (BZLF1 transcription activator), and BRLF1, which encodes Rta (BRLF1 transcription activator). The Zta and Rta transcription factors play an important role in the switch from the latent cycle to the lytic cycle during EBV infection. The mRNA expression levels of BZLF1 and BRLF1 were statistically significantly enhanced in both cell lines after treatment for 24 hours with anti-IgG, TPA, or CSE. The induction of BZLF1 and BRLF1 expression by CSE treatment in GFP-EBV-positive Akata cells was 3.8-fold and 3.5-fold, respectively ( P < .01), and the induction of both BZLF1 and BRLF1 in B95-8 cells was 4.5-fold ( P < .01) ( Figure 4 , A and B ). The expression of BZLF1 had the same trend, although there was no statistically significant increase at 4 hours ( P = .12) ( Figure 4 , C ). The increased BRLF1 mRNAs were observed as early as 4 hours after treatment ( P < .01) and peaked at 12 hours ( P < .01) compared with untreated cells ( Figure 4 , D ). Also, immunofluorescence staining showed an increased population of Rta-positive cells and Zta-positive cells in both cell lines after 24 hours of CSE treatment ( Figure 4 , E–J).

Figure 4.

Expression of Epstein–Barr virus (EBV) transcriptional factors (Zta and Rta) in vitro after cell lines were treated with cigarette smoke extract (CSE). A–D ) Expression of Zta and Rta was detected by quantitative real-time polymerase chain reactions after treatment with 2% CSE. EBV-positive Akata and B95-8 cell lines were treated or untreated with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or CSE and harvested at 0, 4, 8, 12, 24, 36, and 48 hours, and RNA was extracted for quantitative real-time polymerase chain reactions. The relative fold increase of transcript expression was normalized to that of untreated Akata or B95-8 cells, respectively. β-actin was used as an internal control. Data represent the mean with SD ( whisker bars ) from three independent experiments. E–J ) Expression of Rta and Zta protein was detected by immunofluorescence after treatment with 2% CSE for 24 hours. The percentage of Rta-positive cells or Zta-positive cells ( red ) was calculated relative to the total cell population ( blue ) in five random fields of view. Representative images are shown. Scale bar = 100 µm. Data represent the mean with SD ( whisker bars ) from three independent experiments. K ) The EBV immediate-early BRLF1 gene promoter was activated by TPA or CSE. pGL3-BRLF1-promoter construct (0.5 µg) or pGl3-Basic (0.5 µg) was cotransfected with 0.01 µg of Renilla luciferase plasmid. Luciferase activity was monitored 24 hours after transfection. The data are represented as the ratio of firefly to Renilla luciferase activity. CNE2 cells, rather than EBV-negative Akata cells, were used in transfection experiments as these lymphoid cells exhibited low transfection efficiency. Data represent the mean with SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 and † P < .01 compared with the untreated cells as the control.

Figure 4.

Expression of Epstein–Barr virus (EBV) transcriptional factors (Zta and Rta) in vitro after cell lines were treated with cigarette smoke extract (CSE). A–D ) Expression of Zta and Rta was detected by quantitative real-time polymerase chain reactions after treatment with 2% CSE. EBV-positive Akata and B95-8 cell lines were treated or untreated with 100 µg/mL anti-IgG, 20ng/mL phorbol-12-myristate-13-acetate (TPA), or CSE and harvested at 0, 4, 8, 12, 24, 36, and 48 hours, and RNA was extracted for quantitative real-time polymerase chain reactions. The relative fold increase of transcript expression was normalized to that of untreated Akata or B95-8 cells, respectively. β-actin was used as an internal control. Data represent the mean with SD ( whisker bars ) from three independent experiments. E–J ) Expression of Rta and Zta protein was detected by immunofluorescence after treatment with 2% CSE for 24 hours. The percentage of Rta-positive cells or Zta-positive cells ( red ) was calculated relative to the total cell population ( blue ) in five random fields of view. Representative images are shown. Scale bar = 100 µm. Data represent the mean with SD ( whisker bars ) from three independent experiments. K ) The EBV immediate-early BRLF1 gene promoter was activated by TPA or CSE. pGL3-BRLF1-promoter construct (0.5 µg) or pGl3-Basic (0.5 µg) was cotransfected with 0.01 µg of Renilla luciferase plasmid. Luciferase activity was monitored 24 hours after transfection. The data are represented as the ratio of firefly to Renilla luciferase activity. CNE2 cells, rather than EBV-negative Akata cells, were used in transfection experiments as these lymphoid cells exhibited low transfection efficiency. Data represent the mean with SD ( whisker bars ) from three independent experiments. Two-sided P values were calculated by Student t test. * P < .05 and † P < .01 compared with the untreated cells as the control.

Furthermore, we investigated the ability of CSE to activate EBV immediate-early promoters in CNE2 nasopharyngeal carcinoma cells. CSE treatment increased BRLF1 promoter activity by 2.87-fold, and TPA treatment also increased BRLF1 promoter activity by 2.5-fold ( Figure 4 , K ). As expected, the control plasmid, pGl3-Basic, remained at background levels regardless of CSE or TPA treatment, and the background levels were statistically significantly lower than those exhibited by the pGL3-BRLF1-promoter ( P < .01). Thus, these results suggested that CSE treatment enhances the expression of the EBV transcriptional factors, Zta and Rta, which may be responsible for the EBV latent-to-lytic switch.

Discussion

Cigarette smoking, a major risk factor for mortality in China ( 47 ), contributes to a variety of cancers, including lung cancer ( 48–50 ), hepatocellular carcinoma ( 51 ), colorectal cancer ( 52 ), head and neck cancer ( 53 ), and throat cancer ( 54 ). Cigarette smoke is a complex mixture of more than 4000 compounds, and many of these compounds, especially nitrosamine, are mutagenic or carcinogenic ( 57 ). Many studies have linked the exposure of tobacco smoking to NPC risk ( 8 , 55–66) . In our study, we conducted a case–control study in males in high-risk areas of South China, and we observed a statistically significant association between cigarette smoking and the risk of NPC, which was consistent with the literature.

Cigarette smoking produces several reactive forms of agents, such as polycyclic aromatic hydrocarbons, aromatic amines, and N-nitrosamines, resulting in the formation of DNA adducts, which cause DNA damage ( 67 , 68) . This mechanism may be important in NPC pathogenesis.

However, in NPC carcinogenesis, cigarette smoking may play an alternative role via induction of EBV reactivation. In this study, we verified that cigarette smoking and elevated VCA-IgA antibodies were both associated with NPC risk. Moreover, we observed consistently that cigarette smoking was related to seropositivity of the VCA-IgA antibody in three independent populations. In vitro assays provided direct evidence that CSE effectively induced EBV reactivation in both EBV-positive Akata and B95-8 cell lines. In our study, we further assessed the relationship between cigarette smoking and general IgA levels. The results suggest that cigarette smoking may be associated with a decreased serum concentration of total IgA, which was consistent with previous reports ( 69 , 70) . These findings indicate that smoking may increase seropositivity for the EBV VCA-IgA antibody, but not general IgA levels.

Few studies have indicated that cigarette smoking is associated with seropositivity of EBV. A 20-year follow-up study conducted in Taiwan found that long-term heavy cigarette smokers have a higher anti-EBV seropositivity rate than never smokers and short-term light smokers. However, the number of NPC case patients was small (n = 32) ( 8 ). In an analysis of 42 NPC patients in 2008, Abdulamir et al. ( 71 ) suggested that there may be an association between the serum levels of EBV IgA antibodies and smoking. In addition, in 2004, Chang et al. ( 72 ) reported that cigarette smoking is more common among EBV-positive Hodgkin lymphoma patients than among EBV-negative Hodgkin lymphoma patients. Although there has been a consistent trend showing a link between cigarette smoking and EBV-positivity in NPC or Hodgkin lymphoma patients, the sample size of these studies was usually small and could not demonstrate a statistically significant association between cigarette smoking and EBV because of inadequate statistical power. Moreover, these studies were conducted in NPC or Hodgkin lymphoma patients, and tumor burden is known to contribute to the reactivation of EBV ( 73 ). Therefore, it may be inappropriate to search for EBV-inducing agents in patients who are affected with EBV-associated tumors. Thus, healthy males were used in our study to avoid potential confounders.

Also, in vitro assays were used to investigate if EBV reactivation can be induced by CSE. Two B-cell lines, Akata and B95-8, with EBV latent infection were utilized. We demonstrated that EBV replication was promoted by CSE treatment and that the expression levels of lytic-phase gene products were enhanced, including VCA p18, gp350, and two key immediate early genes, BZLF1 and BRLF1, which stimulate a cascade of viral lytic gene expression ( 16 , 74 , 75) .

It has been hypothesized that activated EBV might have a different impact on NPC tumorigenesis during different stages. In the early stage of tumorigenesis, the reactivation of EBV in B cells might facilitate epithelial cell infection with EBV. In vitro experiments suggested that producing massive viral particles of B cells and cell-to-cell contact might play a pivotal role for efficient infection of EBV ( 76 , 77) . The frequent reactivation of EBV may not only increase the viral load in the individual but also change cytokine levels to effectively enhance EBV infection in nasopharyngeal epithelial cells ( 78 ). EBV reactivation in B cells could be triggered by cellular products of tumor cells ( 79 , 80) . Conversely, EBV might act as a promoter for tumorigenesis in a feedback loop. EBV reactivation is associated with the elevations of levels of cytokines and growth factors, ie, interleukin-6 ( 81 ), interleukin-10 ( 81 ), transforming growth factor-β1 ( 82 ), and vascular endothelial growth factor ( 83 ), which could contribute to cell proliferation, immune system perturbation, and angiogenesis. Therefore, EBV reactivation in serum might establish cross-talk with tumor cells via modulating interactions between the infiltrating cells and the tumor cells in the microenvironment.

A previous study suggested that genetic variation may affect an individual’s propensity for seropositive status of the VCA-IgA antibody ( 36 ). In our study, the seropositivity rate for the VCA-IgA antibody in ever smokers in a high-risk area was statistically significantly higher than that in a low-risk area when we examined the interaction effect by using an additive model. These results may be attributed to the specific genetic background of the subjects in the different areas. However, the effect did not reach statistical significance when a multiplicative model was used. Until now, no large-scale study has been conducted to evaluate genetic variants influencing VCA-IgA antibody titers.

To our knowledge, our study is the first attempt to systematically conduct a large-scale investigation aimed at characterizing the relationship between cigarette smoking and EBV reactivation as a potential novel mechanism of NPC etiology. Although cigarette smoking and the anti-EBV antibody have been reported to be related to NPC risk, the novelty of the current study is the establishment of a robust link between cigarette smoking and EBV activation. On the basis of our findings, cigarette cessation programs should be advocated for the primary prevention of NPC in China, especially in NPC endemic areas.

Our study is not without limitations. The individual contribution of smoking and EBV-positivity to the etiology of NPC was not investigated in this study. Future investigations are planned. Also, viruses other than EBV could also be activated by smoking, although none have been investigated. Here, we have provided evidence only for EBV. Studies of other viruses are being planned.

To summarize, we believe that searching for more environmental inducers specific to NPC high-risk areas has great value in NPC prevention. Future studies of the mechanisms by which cigarette smoking modulates the reactivation of EBV and further induces carcinogenesis of the nasopharynx are needed.

Funding

Ministry of Science and Technology of China (2011CB504300; 2011ZX09307-001-04), National Natural Science Funds for Distinguished Young Scholar (81025014), the National Natural Science Foundation of China (81000925), and Guangdong Department of Science and Technology (2008A060201009).

References

1.
Yu
MC
Yuan
JM
Epidemiology of nasopharyngeal carcinoma.
Semin Cancer Biol
 
2002
12
(
6
):
421
429
2.
Ou
SH
Zell
JA
Ziogas
A
Anton-Culver
H
Epidemiology of nasopharyngeal carcinoma in the United States: improved survival of Chinese patients within the keratinizing squamous cell carcinoma histology.
Ann Oncol
 
2007
18
(
1
):
29
35
3.
Busson
P
Ooka
T
Corbex
M
Nasopharyngeal carcinomas and Epstein-Barr virus: from epidemiology and detection to therapy.
Med Sci (Paris)
 
2004
20
(
4
):
453
457
4.
Ng
WT
Choi
CW
Lee
MC
Chan
SH
Yau
TK
Lee
AW
Familial nasopharyngeal carcinoma in Hong Kong: epidemiology and implication in screening.
Fam Cancer
 
2009
8
(
2
):
103
108
5.
Takeshita
H
Furukawa
M
Fujieda
S
et al
Epidemiological research into nasopharyngeal carcinoma in the Chubu region of Japan.
Auris Nasus Larynx
 
1999
26
(
3
):
277
286
6.
Pattle
SB
Farrell
PJ
The role of Epstein-Barr virus in cancer.
Expert Opin Biol Ther
 
2006
6
(
11
):
1193
1205
7.
Raab-Traub
N
Epstein-Barr virus in the pathogenesis of NPC.
Semin Cancer Biol
 
2002
12
(
6
):
431
441
8.
Hsu
WL
Chen
JY
Chien
YC
et al
Independent effect of EBV and cigarette smoking on nasopharyngeal carcinoma: a 20-year follow-up study on 9,622 males without family history in Taiwan.
Cancer Epidemiol Biomarkers Prev
 
2009
18
(
4
):
1218
1226
9.
Chien
YC
Chen
JY
Liu
MY
et al
Serologic markers of Epstein-Barr virus infection and nasopharyngeal carcinoma in Taiwanese men.
N Engl J Med
 
2001
345
(
26
):
1877
1882
10.
Zeng
Y
Zhong
JM
Li
LY
et al
Follow-up studies on Epstein-Barr virus IgA/VCA antibody-positive persons in Zangwu County, China.
Intervirology
 
1983
20
(
4
):
190
194
11.
Zeng
Y
Zhang
LG
Wu
YC
et al
Prospective studies on nasopharyngeal carcinoma in Epstein-Barr virus IgA/VCA antibody-positive persons in Wuzhou City, China.
Int J Cancer
 
1985
36
(
5
):
545
547
12.
Zong
YS
Sham
JS
Ng
MH
et al
Immunoglobulin A against viral capsid antigen of Epstein-Barr virus and indirect mirror examination of the nasopharynx in the detection of asymptomatic nasopharyngeal carcinoma.
Cancer
 
1992
69
(
1
):
3
7
13.
Joab
I
Triki
H
de Saint Martin
J
Perricaudet
M
Nicolas
JC
Detection of anti-Epstein-Barr virus trans-activator (ZEBRA) antibodies in sera from patients with human immunodeficiency virus.
J Infect Dis
 
1991
163
(
1
):
53
56
14.
Tedeschi
R
Pin
E
Martorelli
D
et al
Serum antibody response to lytic and latent Epstein-Barr virus antigens in undifferentiated nasopharyngeal carcinoma patients from an area of nonendemicity.
Clin Vaccine Immunol
 
2007
14
(
4
):
435
441
15.
Rickinson
AB,
Kieff
E
Epstein-Barr virus
. In:
Knipe
DM
Howley
PM
, eds.
Field’s Virology
 
4th
ed.
Philadelphia, PA
Lippincott
Williams and Wilkins;
2001
2575–
2–
627
16.
Feng
P
Ren
EC
Liu
D
Chan
SH
Hu
H
Expression of Epstein-Barr virus lytic gene BRLF1 in nasopharyngeal carcinoma: potential use in diagnosis.
J Gen Virol
 
2000
81
(p
t 10
):
2417
2423
17.
Pepperl
S
Benninger-Döring
G
Modrow
S
Wolf
H
Jilg
W
Immediate-early transactivator Rta of Epstein-Barr virus (EBV) shows multiple epitopes recognized by EBV-specific cytotoxic T lymphocytes.
J Virol
 
1998
72
(
11
):
8644
8649
18.
zur Hausen
H
O’Neill
FJ
Freese
UK
Hecker
E
Persisting oncogenic herpesvirus induced by the tumour promotor TPA.
Nature
 
1978
272
(
5651
):
373
375
19.
Fang
CY
Lee
CH
Wu
CC
et al
Recurrent chemical reactivations of EBV promotes genome instability and enhances tumor progression of nasopharyngeal carcinoma cells.
Int J Cancer
 
2009
124
(
9
):
2016
2025
20.
Jiang
JH
Wang
N
Li
A
et al
Hypoxia can contribute to the induction of the Epstein-Barr virus (EBV) lytic cycle.
J Clin Virol
 
2006
37
(
2
):
98
103
21.
Luka
J
Kallin
B
Klein
G
Induction of the Epstein-Barr virus (EBV) cycle in latently infected cells by n-butyrate.
Virology
 
1979
94
(
1
):
228
231
22.
Yoshida
M
Kijima
M
Akita
M
Beppu
T
Potent and specific inhibition of mammalian histone deacetylase both in vivo and in vitro by trichostatin A.
J Biol Chem
 
1990
265
(
28
):
17174
17179
23.
Shao
YM
Poirier
S
Ohshima
H
et al
Epstein-Barr virus activation in Raji cells by extracts of preserved food from high risk areas for nasopharyngeal carcinoma.
Carcinogenesis
 
1988
9
(
8
):
1455
1457
24.
Zeng
Y
Zhong
JM
Ye
SQ
et al
Screening of Epstein-Barr virus early antigen expression inducers from Chinese medicinal herbs and plants.
Biomed Environ Sci
 
1994
7
(
1
):
50
55
25.
Liu
MY
Chang
YL
Ma
J
et al
Evaluation of multiple antibodies to Epstein-Barr virus as markers for detecting patients with nasopharyngeal carcinoma.
J Med Virol
 
1997
52
(
3
):
262
269
26.
Lo
YM
Chan
LY
Lo
KW
et al
Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma.
Cancer Res
 
1999
59
(
6
):
1188
1191
27.
Tong
JH
Lo
KW
Au
FW
Huang
DP
To
KF
Re: discrete alterations in the BZLF1 promoter in tumor and non-tumor-associated Epstein-Barr virus.
J Natl Cancer Inst
 
2003
95
(
13
):
1008
–100–
9; author reply 1009
28.
Chang
ET
Adami
HO
The enigmatic epidemiology of nasopharyngeal carcinoma.
Cancer Epidemiol Biomarkers Prev
 
2006
15
(
10
):
1765
1777
29.
Ren
ZF
Liu
WS
Qin
HD
et al
Effect of family history of cancers and environmental factors on risk of nasopharyngeal carcinoma in Guangdong, China.
Cancer Epidemiol
 
2010
34
(
4
):
419
424
30.
Jia
WH
Luo
XY
Feng
BJ
et al
Traditional Cantonese diet and nasopharyngeal carcinoma risk: a large-scale case-control study in Guangdong, China
BMC Cancer
 
2010
10
446
http://www.biomedcentral.com/1471-2407/10/446 . Accessed May 26, 2012
31.
Ruan
HL
Xu
FH
Liu
WS
et al
Alcohol and tea consumption in relation to the risk of nasopharyngeal carcinoma in Guangdong, China.
Front Med China
 
2010
4
(
4
):
448
456
32.
Arizona Cancer Center.
Questionnaires
http://www.azcc.arizona.edu/research/shared-services/bmss/questionnaires . Accessed May26
2012
33.
European Monitoring Centre for Drugs and Drug Addiction
Lifetime drinking history
http://www.emcdda.europa.eu/html.cfm/index4163EN.html . AccessedMay26
2012
34.
Centers for Disease Control and Prevention.
Alcohol
http://www.cdc.gov/alcohol/faqs.htm . AccessedMay26
2012
35.
Qin
HD
Jia
WH
Zhang
LL
et al
Elevated Epstein-Barr virus seroreactivity among unaffected members of families with nasopharyngeal carcinoma.
J Med Virol
 
2011
83
(
10
):
1792
1798
36.
Shen
GP
Pan
QH
Hong
MH
et al
Human genetic variants of homologous recombination repair genes first found to be associated with Epstein-Barr virus antibody titers in healthy Cantonese.
Int J Cancer
 
2011
129
(
6
):
1459
1466
37.
Shimizu
N
Tanabe-Tochikura
A
Kuroiwa
Y
Takada
K
Isolation of Epstein-Barr virus (EBV)-negative cell clones from the EBV-positive Burkitt’s lymphoma (BL) line Akata: malignant phenotypes of BL cells are dependent on EBV.
J Virol
 
1994
68
(
9
):
6069
6073
38.
Shimizu
N
Yoshiyama
H
Takada
K
Clonal propagation of Epstein-Barr virus (EBV) recombinants in EBV-negative Akata cells.
J Virol
 
1996
70
(
10
):
7260
7263
39.
Lan
K
Verma
SC
Murakami
M
Bajaj
B
Robertson
ES
Epstein-Barr virus (EBV): infection, propagation, quantitation, and storage.
Curr Protoc Microbiol
 
2007
Chapter 14
Unit 14E.2
40.
Song
LB
Zeng
MS
Liao
WT
et al
Bmi-1 is a novel molecular marker of nasopharyngeal carcinoma progression and immortalizes primary human nasopharyngeal epithelial cells.
Cancer Res
 
2006
66
(
12
):
6225
6232
41.
Song
LB
Li
J
Liao
WT
et al
The polycomb group protein Bmi-1 represses the tumor suppressor PTEN and induces epithelial-mesenchymal transition in human nasopharyngeal epithelial cells.
J Clin Invest
 
2009
119
(
12
):
3626
3636
42.
Chang
LK
Liu
ST
Activation of the BRLF1 promoter and lytic cycle of Epstein-Barr virus by histone acetylation.
Nucleic Acids Res
 
2000
28
(
20
):
3918
3925
43.
Sambrook
J
Fritsch
EF
Maniatis
T
Molecular Cloning
A Laboratory Manual
 
2nd
ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press;
1989
142–
152
44.
Rothman
KJ
The estimation of synergy or antagonism.
Am J Epidemiol
 
1976
103
(
5
):
506
511
45.
Liu
Y
Huang
Q
Liu
W
et al
Establishment of VCA and EBNA1 IgA-based combination by enzyme-linked immunosorbent assay as preferred screening method for nasopharyngeal carcinoma: a two-stage design with a preliminary performance study and a mass screening in southern China.
Int J Cancer
 
2012
131
(
2
):
406
416
46.
Skol
AD
Scott
LJ
Abecasis
GR
Boehnke
M
Joint analysis is more efficient than replication-based analysis for two-stage genome-wide association studies.
Nat Genet
 
2006
38
(
2
):
209
213
47.
Gu
D
Kelly
TN
Wu
X
et al
Mortality attributable to smoking in China.
N Engl J Med
 
2009
360
(
2
):
150
159
48.
Tyczynski
JE
Bray
F
Parkin
DM
Lung cancer in Europe in 2000: epidemiology, prevention, and early detection.
Lancet Oncol
 
2003
4
(
1
):
45
55
49.
Yuan
J
Ma
J
Zheng
H
et al
Overexpression of OLC1, cigarette smoke, and human lung tumorigenesis.
J Natl Cancer Inst
 
2008
100
(
22
):
1592
1605
50.
Haiman
CA
Stram
DO
Wilkens
LR
et al
Ethnic and racial differences in the smoking-related risk of lung cancer.
N Engl J Med
 
2006
354
(
4
):
333
342
51.
Jee
SH
Ohrr
H
Sull
JW
Samet
JM
Cigarette smoking, alcohol drinking, hepatitis B, and risk for hepatocellular carcinoma in Korea.
J Natl Cancer Inst
 
2004
96
(
24
):
1851
1856
52.
Giovannucci
E
Rimm
EB
Stampfer
MJ
et al
A prospective study of cigarette smoking and risk of colorectal adenoma and colorectal cancer in U.S. men.
J Natl Cancer Inst
 
1994
86
(
3
):
183
191
53.
Brennan
JA
Boyle
JO
Koch
WM
et al
Association between cigarette smoking and mutation of the p53 gene in squamous-cell carcinoma of the head and neck.
N Engl J Med
 
1995
332
(
11
):
712
717
54.
Liu
X
Togo
S
Al-Mugotir
M
et al
NF-kappaB mediates the survival of human bronchial epithelial cells exposed to cigarette smoke extract
Respir Res
 
2008
9
66
http://respiratory-research.com/content/9/1/66 .Accessed May26
2012
55.
Vaughan
TL
Shapiro
JA
Burt
RD
et al
Nasopharyngeal cancer in a low-risk population: defining risk factors by histological type.
Cancer Epidemiol Biomarkers Prev
 
1996
5
(
8
):
587
593
56.
Friborg
JT
Yuan
JM
Wang
R
Koh
WP
Lee
HP
Yu
MC
A prospective study of tobacco and alcohol use as risk factors for pharyngeal carcinomas in Singapore Chinese.
Cancer
 
2007
109
(
6
):
1183
1191
57.
West
S
Hildesheim
A
Dosemeci
M
Non-viral risk factors for nasopharyngeal carcinoma in the Philippines: results from a case-control study.
Int J Cancer
 
1993
55
(
5
):
722
727
58.
Cheng
YJ
Hildesheim
A
Hsu
MM
et al
Cigarette smoking, alcohol consumption and risk of nasopharyngeal carcinoma in Taiwan.
Cancer Causes Control
 
1999
10
(
3
):
201
207
59.
Chow
WH
McLaughlin
JK
Hrubec
Z
Nam
JM
Blot
WJ
Tobacco use and nasopharyngeal carcinoma in a cohort of US veterans.
Int J Cancer
 
1993
55
(
4
):
538
540
60.
Yuan
JM
Wang
XL
Xiang
YB
Gao
YT
Ross
RK
Yu
MC
Non-dietary risk factors for nasopharyngeal carcinoma in Shanghai, China.
Int J Cancer
 
2000
85
(
3
):
364
369
61.
Zhu
K
Levine
RS
Brann
EA
Hall
HI
Caplan
LS
Gnepp
DR
Case-control study evaluating the homogeneity and heterogeneity of risk factors between sinonasal and nasopharyngeal cancers.
Int J Cancer
 
2002
99
(
1
):
119
123
62.
Shao
JY
Li
YH
Gao
HY
et al
Comparison of plasma Epstein-Barr virus (EBV) DNA levels and serum EBV immunoglobulin A/virus capsid antigen antibody titers in patients with nasopharyngeal carcinoma.
Cancer
 
2004
100
(
6
):
1162
1170
63.
Chen
CJ
Liang
KY
Chang
YS
et al
Multiple risk factors of nasopharyngeal carcinoma: Epstein-Barr virus, malarial infection, cigarette smoking and familial tendency.
Anticancer Res
 
1990
10
(
2B
):
547
553
64.
Nam
JM
McLaughlin
JK
Blot
WJ
Cigarette smoking, alcohol, and nasopharyngeal carcinoma: a case-control study among U.S. whites.
J Natl Cancer Inst
 
1992
84
(
8
):
619
622
65.
Yu
MC
Garabrant
DH
Huang
TB
Henderson
BE
Occupational and other non-dietary risk factors for nasopharyngeal carcinoma in Guangzhou, China.
Int J Cancer
 
1990
45
(
6
):
1033
1039
66.
Lin
TM
Chen
KP
Lin
CC
et al
Retrospective study on nasopharyngeal carcinoma.
J Natl Cancer Inst
 
1973
51
(
5
):
1403
1408
67.
Hecht
SS
Cigarette smoking and lung cancer: chemical mechanisms and approaches to prevention.
Lancet Oncol
 
2002
3
(
8
):
461
469
68.
Takahashi
H
Ogata
H
Nishigaki
R
Broide
DH
Karin
M
Tobacco smoke promotes lung tumorigenesis by triggering IKKbeta- and JNK1-dependent inflammation.
Cancer Cell
 
2010
17
(
1
):
89
97
69.
Gulsvik
A
Fagerhoi
MK
Smoking and immunoglobulin levels.
Lancet
 
1979
1
(
8113
):
449
70.
Arnson
Y
Shoenfeld
Y
Amital
H
Effects of tobacco smoke on immunity, inflammation and autoimmunity.
J Autoimmun
 
2010
34
(
3
):
J258
J265
71.
Abdulamir
AS
Hafidh
RR
Abdulmuhaimen
N
et al
The distinctive profile of risk factors of nasopharyngeal carcinoma in comparison with other head and neck cancer types
BMC Public Health
 
2008
8
400
http://www.biomedcentral.com/1471-2458/8/400 .Accessed May26
2012
72.
Chang
ET
Zheng
T
Lennette
ET
et al
Heterogeneity of risk factors and antibody profiles in Epstein-Barr virus genome-positive and -negative Hodgkin lymphoma.
J Infect Dis
 
2004
189
(
12
):
2271
2281
73.
Halprin
J
Scott
AL
Jacobson
L
et al
Enzyme-linked immunosorbent assay of antibodies to Epstein-Barr virus nuclear and early antigens in patients with infectious mononucleosis and nasopharyngeal carcinoma.
Ann Intern Med
 
1986
104
(
3
):
331
337
74.
Kudoh
A
Fujita
M
Kiyono
T
et al
Reactivation of lytic replication from B cells latently infected with Epstein-Barr virus occurs with high S-phase cyclin-dependent kinase activity while inhibiting cellular DNA replication.
J Virol
 
2003
77
(
2
):
851
861
75.
Lu
CC
Jeng
YY
Tsai
CH
et al
Genome-wide transcription program and expression of the Rta responsive gene of Epstein-Barr virus.
Virology
 
2006
345
(
2
):
358
372
76.
Chang
Y
Tung
CH
Huang
YT
Lu
J
Chen
JY
Tsai
CH
Requirement for cell-to-cell contact in Epstein-Barr virus infection of nasopharyngeal carcinoma cells and keratinocytes.
J Virol
 
1999
73
(
10
):
8857
8866
77.
Imai
S
Nishikawa
J
Takada
K
Cell-to-cell contact as an efficient mode of Epstein-Barr virus infection of diverse human epithelial cells.
J Virol
 
1998
72
(
5
):
4371
4378
78.
Tsang
CM
Zhang
G
Seto
E
et al
Epstein-Barr virus infection in immortalized nasopharyngeal epithelial cells: regulation of infection and phenotypic characterization.
Int J Cancer
 
2010
127
(
7
):
1570
1583
79.
Mochanko
K
Fejes
M
Breazavscek
DM
Suarez
A
Bachmann
AE
The relation between Epstein-Barr virus antibodies and clinical symptomatology and immunodeficiency in patients with Hodgkin’s disease.
Cancer
 
1979
44
(
6
):
2065
2070
80.
Fan
H
Nicholls
J
Chua
D
et al
Laboratory markers of tumor burden in nasopharyngeal carcinoma: a comparison of viral load and serologic tests for Epstein-Barr virus.
Int J Cancer
 
2004
112
(
6
):
1036
1041
81.
Tan
EL
Selvaratnam
G
Kananathan
R
et al
Quantification of Epstein-Barr virus DNA load, interleukin-6, interleukin-10, transforming growth factor-beta1 and stem cell factor in plasma of patients with nasopharyngeal carcinoma
BMC Cancer
 
2006
6
227
http://www.biomedcentral.com/1471-2407/6/227 .Accessed May26
2012
82.
Xu
J
Menezes
J
Prasad
U
Ahmad
A
Elevated serum levels of transforming growth factor beta1 in Epstein-Barr virus-associated nasopharyngeal carcinoma patients.
Int J Cancer
 
1999
84
(
4
):
396
399
83.
Hong
GK
Gulley
ML
Feng
WH
Delecluse
HJ
Holley-Guthrie
E
Kenney
SC
Epstein-Barr virus lytic infection contributes to lymphoproliferative disease in a SCID mouse model.
J Virol
 
2005
79
(
22
):
13993
14003

Notes

The authors have no conflict of interest to declare. The sponsors of the study had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

We thank Dr Maria Masucci for kindly providing Epstein–Barr positive and negative Akata cell lines. We also thank Hui-Jun Li at the First General Hospital of Yangquan City, Shanxi Province, for his help with the Yangquan population enrollment and Wan-Li Liu at Sun Yat-sen University Cancer Center for his technical assistance in detection of Epstein–Barr virus serum antibodies.