Human Papillomavirus 16/18 E6 Oncoprotein Is Expressed in Lung Cancer and Related with p53 Inactivation Free

Lung cancer is the leading cause of cancer death for Taiwanese women since 1982. Although cigarette smoking is the major cause of lung cancer worldwide, more than 90% of lung cancer in Taiwanese females is not related to cigarette smoking (1). Therefore, most Taiwanese women may have a unique etiology for lung cancer development. Our previous study indicated that human papillomavirus (HPV) oncogenic subtypes 16 and 18, which are involved in cervical cancer, may also be involved in the pathogenesis of lung cancer among Taiwanese because 55% of lung cancer patients had HPV16/18 DNA compared with 11% of noncancer control subjects. In addition, the odds ratio for lung cancer in nonsmoking females with HPV16/18 infection (∼10) was much higher than that for nonsmoking males (∼2; ref. 2). Additionally, HPV16/18 DNA was uniformly detected in lung tumor cells but not in the adjacent noninvolved lung tissue. These results strongly suggest that HPV infection with virus subtypes known to be oncogenic for cervical cancer is associated with lung cancer development in nonsmoking Taiwanese women. In addition, our recent case-control study also clearly revealed that an individual with HPV16 and HPV18 DNA in their blood was at a 76-fold risk for lung cancer compared with subjects without HPV16/18 DNA (3), further implicating HPV in lung tumorigenesis.

Although studies of viral-related lung cancer have been reported (4–6), the molecular pathogenesis of this disease type remains unclear. For example, the effect of the oncogenic DNA virus SV40 on the development of malignant mesotheliomas and the high risk of HPV16/18 in lung cancer were controversial until recently. The integration of high-risk HPV16/18 DNA into host chromosome to express E6 protein plays a crucial role in HPV-induced cervical carcinogenesis (7–9). E6 has many functions that may contribute to its oncogenic potential. The classic function of E6, which is relevant to cellular immortalization, is binding to the tumor suppressor p53, thereby inducing p53 degradation (10). The role of p53 is to safeguard the integrity of the genome by inducing cell cycle arrest or apoptosis on DNA damage (11). As a transcription factor, p53 up-regulates target genes involved in coordinating these responses [e.g., p21^WAF1/CIP1, a cyclin-dependent kinase (cdk) inhibitor that acts on cyclin E/cdk2 complexes and mdm2; refs. 12, 13]. Therefore, p53 inactivation by E6 leads to chromosomal instability and increases the probability of a HPV-infected cell evolving toward malignancy (10). Animal model experiment further showed that HPV16 E6 gene alone is sufficient to induce carcinomas in transgenic mice (14).

In this study, to understand whether p53 could be inactivated by E6 in HPV-infected lung cancer, the following experiments would be done (a) to examine whether E6 could express in lung tumors; (b) to understand whether E6 protein expression in lung tumor was associated with the inactivation of p53 pathway; and (c) to elucidate the role of E6 in p53 inactivation in HPV-infected lung cancer cell lines that have been successfully established from patients' pleural effusions.

Study subjects. Lung tumor specimens from 122 patients with primary lung cancer were collected. All of these patients, including 57 females and 65 males who were admitted into the Department of Thoracic Surgery, Taichung Veteran's General Hospital, Taichung, Taiwan between 1998 and 2002, were asked to submit a written informed consent approved by the Institutional Review Board. None of these subjects had received radiation therapy or chemotherapy before surgery. Collected lung tumors have previously been analyzed for the presence of HPV16 and/or HPV18 DNA (2, 3), and tumor types and stages were histologically determined according to the WHO (1981) classification. Pathologic material was processed for conventional histologic procedures.

Immunohistochemistry. Formalin-fixed and paraffin-embedded specimens were sectioned at a thickness of 3 μm. All sections were then deparaffinized in xylene, rehydrated through serial dilutions of alcohol, and washed in PBS (pH 7.2), the buffer which was used for all subsequent washes. For HPV16 E6, HPV18E6, and p53 detection, sections were heated in a microwave oven twice for 5 min in citrate buffer (pH 6.0), and then incubated with a monoclonal antihuman p53 antibody (DAKO, DO7; at a dilution of 1:250) for 60 min at 25°C or with polyclonal anti-HPV16 or HPV18 E6 antibody (Santa Cruz Biotechnology and Chemicon International, Inc.) for 90 min at 25°C. The conventional streptavidin peroxidase method (DAKO, LSAB Kit K675) was done to develop signals and the cells were counterstained with hematoxylin. Negative controls were obtained by leaving out the primary antibody. The intensities of signals were evaluated independently by three observers. Negative immunostaining was defined to be with 0% to 10% positive nuclei, and cases with >10% positive nuclei were decided to be positive for immunostaining. Positive control slides for p53 protein detection were purchased from DAKO and the cervical cancer tumor tissues with HPV16/18 were used as positive control for HPV16/18 E6. The antibody dilution buffer was used to replace antibodies to serve as a negative control.

Direct sequencing. Mutations in exons 5 to 8 of the p53 gene were determined by direct sequencing of PCR products amplified from the DNA of tumor cells isolated by microdissection of the lung tumor tissues. DNA lysis buffer was used to lyse cells and then the solution was subjected to proteinase K digestion and phenol-chloroform extraction. Finally, the DNA was precipitated by ethanol with the addition of linear polyacrylamide to increase the efficiency of DNA precipitation (15). Target sequences were amplified in a 50-μL reaction mixture containing 20 pmol of each primer, 2.5 units of Taq polymerase (TAKARA Shuzo), 0.5 mmol/L deoxynucleotide triphosphates, 5 μL of PCR reaction buffer, and 1 μL of genomic DNA as the template. Genomic DNAs extracted from the frozen sections were not adequate for an amplification of long fragment DNAs, and therefore PCR products ranging from 200 to 400 bp were amplified for p53 mutation analysis. Primers for β-actin, acting as an internal control, were included in each amplification reaction. The primers used in the reactions were E5S (5′-tgccctgactttcaactctg-3′) and E5AS (5′-gctgctcaccatcgctatc-3′) for exon 5, E6S (5′-ctgattcctcactgattgct-3′) and E6AS (5′-agttgcaaaccagacctcagg-3′) for exon 6, E7S (5′-cctgtgttatctcctaggttg-3′) and E7AS (5′-gcacagcaggccagtgtgca-3′) for exon 7, and E8S (5′-gacctgatttccttactgcc-3′) and E8AS (5′-tctcctccaccgcttcttgt-3′) for exon 8. An initial cycle was done for 5 min at 94°C, followed by 35 cycles of 40 s at 94°C, 40 s at 54°C, and 1 min at 72°C. The PCR products were sequenced using an Applied Biosystems 3100 Avant Genetic Analyzer (Applied Biosystems), and the same primers used for the PCR were used for the DNA sequencing. All p53 mutations were confirmed by direct sequencing of both DNA strands.

Protein extraction and Western blot. Total proteins were extracted from fresh lung tumor tissues with a lysis buffer [100 mmol/L Tris (pH 8.0), 1% SDS] and recovered protein concentrations were determined using the Bio-Rad protein assay kit followed by a separation with SDS-PAGE (12.5% gel, 1.5 mm thick). To detect HPV16 or HPV18 E6 in lung tumors, 50 μg of total protein were loaded in each lane of the gel. After an electrophoretic transfer onto a polyvinylidene difluoride membrane, nonspecific binding sites were blocked with 5% nonfat milk in TBS-Tween 20. The detection of HPV16 or HPV18 E6 and β-actin was conducted by incubating the membrane with anti-HPV16 E6, anti-HPV18 E6 (Santa Cruz Biotechnology and Chemicon International), and anti–β-actin antibodies (DAKO) for 60 min at room temperature, followed by subsequent incubation with a peroxidase-conjugated secondary antibody (1:5,000 dilution). Extensive washings with TBS-Tween 20 were done between incubations to remove nonspecific binding. The protein bands were visualized by enhanced chemiluminescence (NEN Life Science Products, Inc.).

Nested PCR. Genomic DNA was prepared from a tissue section and isolated by conventional phenol-chloroform extraction, ethanol precipitated, and finally dissolved in 20 μL of sterile distilled water. HPV viral DNA was first amplified with type consensus primers MY09 and MY11 (2), followed by second round of amplification with type-specific primers flanking the L1 region to identify the subtype. Ten microliters of the final PCR product were loaded onto a 2% agarose gel, stained with ethidium bromide, and visualized under UV illumination. Appropriate negative and positive controls were included in each PCR reaction. A part of the β-actin gene in all samples was amplified to exclude false-negative results whereas DNA preparations from SiHa cells (containing HPV16) and HeLa cells (containing HPV18) were used as positive controls.

Establishment of HPV16-infected and noninfected lung cancer cell lines from patients' pleural effusions. Lung tumor cells were culturally established from pleural effusions of three patients by the Ficoll-Paque method. The pathologic diagnosis of the patients was adenocarcinoma of lung including a 72-year-old female nonsmoker with T₄N₃M₁, a 54-year-old male smoker with T₂N₂M₁, and a 53-year-old male ex-smoker with T₄N₃M₁. Briefly, 10-mL Ficoll-Paque was added to the centrifuge tube and then 10 mL of pleural effusion were carefully layered on top. After centrifugation at 400 × g for 30 to 40 min at 18°C to 20°C, the cell layer was transferred to a culture dish. Cells were cultured in RPMI 1640 with 5% CO₂. Single clones were selected for culture for ∼8 months. Karyotyping was done to verify the cells from the clonal expansion to be lung tumor cells. After an antibody against TTF1 had a positive immunoreactive, the cells were identified to be lung adenocarcinoma cells. To verify whether the cells were infected by HPV16 or HPV18, nested PCR, reverse transcription-PCR (RT-PCR), and Western blotting were used to detect HPV16/18 DNA, E6 mRNA, and E6 protein expressions, respectively.

Silencing of endogenous HPV16 E6 expression by RNA interference. The target sequences for RNA interference (RNAi) for HPV16 E6 were previously verified (16, 17). The sequence of the siE6-1 sense strand–directed small interfering RNA (siRNA) was 5,-GAGGUAUAUGACUUUGCUUdTdT-3′ (16) and that of the siE6-2 sense strand–directed siRNA was 5′-GAAUGUGUGUACUGCAAGCdTdT-3′ (17). To suppress transcription of the endogenous HPV16 E6 gene, SiHa and TL-1 cells were transiently transfected with the synthetic siRNAs against HPV16 E6 using Oligofectamine reagent (Invitrogen) according to the manufacturer's instructions. Briefly, 24 h before transfection, 1.4 × 10⁵ cells were seeded in each well of a six-well plate. Oligofectamine reagent (3 μL) was added to 12 μL of OPTI-MEM (Invitrogen). After 5 min, 60 pmol of each siRNA in 175 μL of OPTI-MEM were combined with the Oligofectamine mixture. After a 20-min incubation at 25°C, siRNA/Oligofectamine mixtures were added to the cells. After 48 h of incubation at 37°C, the cells were harvested and subjected to real-time quantitative PCR and Western blot analysis.

Fluorescence in situ hybridization. Fluorescence in situ hybridization (FISH) was done by the metaphase preparation of established lung cancer cells. A FITC-dUTP HPV type 16 probe was used for HPV detection (Roche Diagnostics). Briefly, the hybridizing probe was prepared by PCR amplification with a HPV type 16–specific primer (16UP-5′-TACTAACTTTAAGGAGTACC-3′, 16DN-5′-GTGTATGTTTTTGACAAGCAATT-3′; ref. 2). Slides were treated with RNase for 30 min at 37°C, and FISH was done as previously described (18). The hybridization was done in a humidified chamber at 48°C for 16 h followed by washing with sodium chloride sodium citrate, and then counterstaining with 4′,6-diamidino-2-phenylindole was observed for signals under a microscope.

Preparation of RNA and real-time quantitative RT-PCR. Total RNA of lung tumors was extracted by homogenization in 1-mL TRIzol reagent, followed by chloroform reextraction and isopropanol precipitation. Three micrograms of total RNA from lung tumor tissues were reverse transcribed using SuperScript II Reverse Transcriptase (Invitrogen) and oligo d(T)₁₅ primer. Quantitative RT-PCR was done in a final volume of 25 μL containing 1 μL of each cDNA template, 10 pmol of each primer, and 12.5 μL of a SYBR Green master mix. The primers were designed using the ABI Prism 7000 SDS Software. Quantification was carried out using the comparative threshold cycle (C_T) method and water was used as the negative control. An arbitrary threshold was chosen on the basis of the variability of the baseline. C_T values were calculated by determining the point at which fluorescence exceeded the threshold limit. C_T was reported as the cycle number at this point. The average of target gene was normalized to 18S rRNA as endogenous housekeeping gene.

Flow cytometry analysis. Distribution of the cells in the cell cycle was determined by propidium iodide staining. In brief, cells (2 × 10⁵/mL) were fixed with PBS containing 80% ethanol, then incubated at 4°C in 0.2-mL PBS solution containing 0.05 mg/mL propidium iodide, 1 mmol/L EDTA, 0.01% Triton X-100, and 1 mg/mL RNase A for 1 h. Analysis was done with a FACSCalibur cytometer (Becton Dickinson). Cells with subdipolid DNA content were considered to be apoptotic cells. Cell cycle distributions were analyzed by CellQuest software.

Effects of E6 RNAi on the doubling time, plating efficiency, and cloning efficiency of TL-1 lung cancer cells. TL-1 cells (10³/mL) transfected with or without siE6-1 and siE6-2 were seeded in a 35-mm dish and cultured for 24, 48, 72, 96, 120, and 144 h, and then the cell number at each culture time point was counted for calculation of the doubling time. The plating and cloning efficiency assay was done with 10³ cells in a 35-mm dish cultured for 5 to 7 and 14 days, respectively.

Statistical analysis. Statistical analysis was done using the SPSS statistical software program (version 11.0, SPSS, Inc.). The χ² test, Fisher exact test (two tailed), and Mann-Whitney test were applied for statistical analysis.

E6 protein was indeed expressed in lung tumors and adjacent normal tissues: relationships between E6 and clinical parameters. Our preliminary restriction-specific PCR data showed that HPV16/18 DNA integration occurred in HPV DNA–positive lung tumors. We thus attempted to determine whether HPV16/18 E6 is expressed in lung tumors by Western blotting and immunohistochemistry to verify the association between p53 expression and HPV16/18 E6 expression. Western blotting was first used to detect the presence or absence of HPV16/18 E6 in 10 randomly selected HPV DNA–positive lung tumors and corresponding adjacent normal lung tissues. The data clearly showed that E6 was predominately expressed in lung tumors, although some of paired adjacent normal tissues had low-level E6 expression (Fig. 1A). Consequently, 122 lung tumors containing or lacking HPV16/18 DNA were tested for E6 expression by immunohistochemistry. Our present and previous data indicated that HPV16/18 E6 is only expressed in lung tumors that were previously shown to be positive for HPV16/18 DNA by nested PCR (Table 1). HPV16 or HPV18 E6 was indeed expressed in tumor cells as well as in adjacent normal cells in tumor tissues such as type II pneumocytes, bronchiole epithelia, blood vessel endothelia, lymphocytes, and alveolar macrophages (Fig. 1B).

The relationships between E6 expression and clinical parameters of lung tumors are shown in Table 1. The expression of HPV16 or HPV18 E6 in lung tumors of females, adenocarcinoma patients, and nonsmokers was significantly higher than in lung tumors of males, squamous cell carcinoma patients, and smokers, respectively (P = 0.001 for gender and tumor type; P = 0.002 for smoking status). E6 expression was not associated with other clinical parameters including age, tumor stage, and T and N factor, although HPV18 E6 expression was more common in advanced tumors and associated with T factor (Table 1).

E6 protein was negatively associated with p53 protein expression in lung tumors. To elucidate whether E6 affects p53 expression, p53 expression in lung tumor tissues was also determined by immunohistochemistry. p53 expression correlated inversely with HPV16 E6 (P = 0.011) and HPV16/18 E6 expression (P = 0.004), but was marginally associated with HPV18 E6 alone (P = 0.085; Table 2). To confirm the reciprocal relationship between HPV16 or HPV18 E6 and p53, serial paraffin sections of lung tumors were used to assess protein expression in vivo. p53 protein was not detected in tumors positive for HPV16 or HPV18 E6; conversely, HPV16 or HPV18 E6–negative tumors had positive p53 protein expression (Fig. 2). The reverse correlation between HPV16/18 E6 and p53 expression in vivo clearly revealed the possibility that HPV16/18 E6 may, at least in part, promote the degradation of p53 in HPV-positive lung tumors.

The levels of p21 and mdm2 mRNA in E6-positive tumors were lower than in E6-negative tumors. To elucidate whether p53 was inactivated by E6, mRNA levels of p21^WAF1/CIP1 and mdm2, which function downstream of p53, in lung tumors were measured by real-time RT-PCR. The mRNA levels of p21^WAF1/CIP1 and mdm2 in HPV16 E6–, HPV18 E6–, and HPV16/18 E6–positive tumors were significantly lower than those of negative tumors (Table 3). However, the expression of these genes did not correlate with p53 mutations and p53 expression, although a negative trend was apparent (Table 3). These results suggest that p53 inactivation caused by HPV16/18 E6 may play a more important role than p53 mutations or other mechanism(s) in causing p53 accumulation in HPV-positive lung tumors.

The involvement of E6 in p53 inactivation in HPV16-infected lung cancer cells. To elucidate the role of E6 in p53 inactivation in lung cancer, we established three HPV16-infected and one non–HPV16-infected lung adenocarcinoma cell lines from patients' pleural effusions. The HPV16 DNA copy numbers of these three cell lines were evaluated by FISH showing that two to three, one, and four to five of HPV16 DNA copies were revealed in TL-1, TL-2, and TL-3 cells, respectively. Thus, these results clearly showed that HPV16 DNA integrated into the chromosomes of these cells (Supplementary Fig. S1). The tumorigenicity of these established lung cancer cell was shown by soft agar assay. HPV16-infected cells had the ability to form larger colonies in soft agar assay compared with the noninfected cells after 14 days of culture (data not shown). To our knowledge, this is the first time to establish HPV-infected lung cancer cell lines from pleural effusions of lung cancer patients.

It is well known that E6 was derepressed by E2 splicing when HPV16 DNA integrated into host chromosomes (19). E6 protein was then evaluated by Western blotting showing that different levels of HPV16 E6 proteins were expressed in HPV16-infected TL-1, TL-2, and TL-3 cells. As expected, E6 was not detected in non–HPV16-infected TL-4 cells. Our data also revealed that p53 protein levels in E6-positive TL-1, TL-2, and TL-3 cells were significantly lower than in E6-negative TL-4 cells (Fig. 3A). Immunoprecipitation assay clearly showed that E6 protein interacted with p53 protein in E6-positive cells but not in E6-negative cells (Fig. 3B). To further verify whether the interaction between E6 and p53 could be responsible for p53 inactivation, TL-1 E6 was knocked down by two RNAis. Western blot showed that E6 protein in siE6-1 (the first RNAi) and siE6-2 (the second RNAi) transient cells was reduced compared with negative control cells; however, E6 was more efficiently reduced in siE6-2 cells than in siE6-1 cells (Fig. 3C). Conversely, p53 protein levels were markedly increased in both siE6 cells (Fig. 3C). The levels of p21^WAF1/CIP1 and mdm2 mRNA evaluated by real-time RT-PCR were significantly restored in siE6-2 cells but relatively renovated in siE6-1 cells compared with negative control cells (Fig. 3D). To explore the growth effects of E6 knockdown by RNAi, TL-1 cells with and without siE6-1 or siE6-2 transfection were evaluated by the doubling time, plating efficiency, and cloning efficiency assay (Supplementary Fig. S2A). As shown in Fig. S2, the doubling time of TL-1 cells with siE6-1 and siE6-2 was extended to 28 to 32 h and 36 to 38 h compared with 24 to 26 h of doubling time for parental TL-1 cells, respectively. The plating efficiency of TL-1 cells was decreased from 42 ± 4% to 22 ± 3% (siE6-1) and 15 ± 1% (siE6-2). The cloning efficiency was also reduced from 95 ± 7% to 46 ± 4% (siE6-1) and 25 ± 3% (siE6-2). In addition, flow cytometry showed that S-phase cell proportion was significantly decreased in E6-knockdown cells (22.98% for siE6-1, 21.17% for siE6-2) as compared with TL-1 parental cells (38.94%; Supplementary Fig. S2B). These results clearly indicated that p53 inactivation by E6 may increase cell proliferation and colony formation.

One of the key events of HPV-induced carcinogenesis is the integration of the viral genome into a host chromosome (20). HPV genome integration often occurs near common fragile sites of the human genome, such as FRA 3B, a site of frequent integration by HPV16, which causes loss of heterozygosity of fragile histidine triad in cervical cancer (21, 22). Interestingly, frequent fragile histidine triad loss of heterozygosity in HPV DNA–positive lung tumors has also been reported (23). In the present study, 28% (34 cases), 25% (31 cases), and 45% (55 cases) of 122 lung tumors were positive for HPV16 E6, HPV18 E6, and HPV16/18 E6, respectively (Table 1). Being stratified tumors in the presence of HPV DNA, the detection frequency of HPV16 E6, HPV18 E6, and HPV16/18 E6 was increased to 61%, 77%, and 67%, respectively. HPV16/18 E6 expression in cervical cancers is necessary for the maintenance of the transformed phenotype (24). Therefore, the high HPV16/18 E6 expression in HPV DNA–positive lung tumors strongly suggests that HPV16/18 may play a crucial role in lung tumorigenesis in Taiwanese women.

Our data show that p53 expression was not associated with mutant p53 in lung tumors (Table 2), which is inconsistent with previous study (25) showing that positive p53 expression was due to the increased protein stability by p53 missense mutations. Among eight tumors negative for HPV16/18 E6 and p53 protein expression, four had p53 deletion mutations. The other four tumors showed HPV16 E6 mRNA expression by in situ RT-PCR, suggesting that E6 expression levels in these four tumors may have been too low to be detected by immunohistochemistry. Among six tumors positive for HPV16/18 E6 and p53 expression, four were detected with HPV16 or HPV18 E6 variants as shown by direct sequencing (data not shown). Nevertheless, these results might partly support the observation that E6-positive tumors had positive p53 expression.

Malanchi et al. (12, 26) reported that HPV16 E6 could induce cellular proliferation, pRb phosphorylation, and accumulation of gene products that are negatively regulated by pRb, such as p16, cdc2, E2F-1, and cyclin A. Consistent with the hyperphosphorylated state of pRb, cyclin A/cdk2 activity is highly elevated in cells expressing E6 from either HPV16 or HPV18. Recently, microarray analysis indicated that a distinct and large subset of cell cycle and cell proliferation genes were up-regulated in HPV-positive head and neck cancer as well as cervical cancer compared with that observed in HPV-negative head and neck cancer, such as cyclin E2, cyclin B1, PCNA, E2Fs, and cdc2 (27). Our studies showed higher cell proliferation and S-phase cell proportion in E6-positive lung cancer cells than in E6-knockdown cells. These results support the findings of microarray data that E6 could up-regulate cell cycle– and cell proliferation–regulated gene expressions (Supplementary Fig. S2B). Malanchi et al. (12, 26) also showed that E6 may strongly down-regulate p21^WAF1/CIP1. Overexpression of p21^WAF1/CIP1 decreases E6-induced proliferation, indicating that the observed down-regulation of endogenous p21^WAF1/CIP1 in E6-expressing cells is a key mechanism for cell cycle dysregulation. Interestingly, all these events seem to be independent of p53 inactivation. This finding may support the present study showing that the decrease in p21^WAF1/CIP1 mRNA levels by HPV16/18 E6 through the p53-independent pathway was more pronounced than the decrease through the p53-dependent pathway on p53 mutation. The inactivation of p53 by a high-risk HPV E6 oncoprotein is a crucial event during cervical carcinogenesis (10, 24). In our present study, tissue in situ immunohistochemistry data clearly showed that E6-positive lung tumors were most often negative for p53 expression. In addition, real-time RT-PCR data revealed that p21^WAF1/CIP1 and mdm2 mRNA expression was significantly reduced in HPV16/18 E6–positive lung tumors as compared with E6-negative tumors. Collectively, our data show that most lung tumors that expressed HPV16/18 E6 were negative for p53 immunostainings. Moreover, E6 seems to down-regulate p21^WAF1/CIP1 and mdm2 mRNA expression, which strongly suggests that HPV16/18 E6 expression in lung tumors could be involved in p53 inactivation. It was well established that the prominent function of E6 stems from its interaction with p53 (followed by p53 degradation) and the proapoptotic protein Bak, which results in resistance to apoptosis and increased chromosomal instability (20, 28). Apart from resistance to apoptosis, many other functions for HPV16/18 E6 in human carcinogenesis have been reported (11). For example, the activation of telomerase and the postulated inhibition of degradation of Src family kinases seem to fulfill important functions in stimulating tumor growth (28, 29). Nevertheless, these results provide crucial evidence in support of our previous reports showing that HPV16/18 infection may be associated with lung tumorigenesis, especially for Taiwanese female nonsmokers.

Grant support: National Health Research Institute (NHRI93A1-NSCLC07-5; NHRI-EX93-9125BI), National Science Council (NSC91-3112-P-040-002; NSC92-2314-B-040-023; NSC93-2320-B-040-056) and Department of Health (DOH 91-7D-1083) of Taiwan, Republic of China.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Dr. J.H. Tsai for her valuable suggestions and critical reading of the manuscript.