Human Papillomavirus Type 16/18 Up-Regulates the Expression of Interleukin-6 and Antiapoptotic Mcl-1 in Non–Small Cell Lung Cancer

The prevalence of lung cancer has been increasing in Taiwan and Western countries during the past two decades. Lung cancer is the leading and second cause of mortality among female and male cancer patients in Taiwan, respectively (Department of Health, ROC). Cigarette smoking is a predominant risk factor of lung cancer (1, 2). However, about one-half of Taiwanese lung cancer is not correlated with cigarette smoking, particularly in women, owing to the comparatively low prevalence of female smokers in Taiwan (<10%). The gender discrepancy in Taiwanese lung cancer mortality (male to female ratio, 2.0) is different from that in other geographic areas, in which the ratio ranges from 2.3 to 8.6 (1, 3). The above observations implicate that certain environmental factors other than smoking may contribute to the development of lung cancer in Taiwan. Nevertheless, the etiology of nonsmoking lung cancer remains unknown.

Human papillomavirus (HPV) is closely associated with human neoplasms such as uterine cervix, vulva, skin, etc. (4). Among the abovementioned neoplasms, HPV is closely correlated with cervical carcinomas. In particular, HPV 16/18 are frequently detected in high-grade squamous intraepithelial lesions and invasive carcinomas. Abundant evidence regarding interactions between HPV infection and inflammatory cytokines, especially interleukin-6 (IL-6), has been established in cervical cancer. Wei et al. (5) discovered that most of the HPV-infected basal cells, squamous epithelial cells surrounding the basal cells, and invasive tumorigenic cervical cells could secrete mass amount of IL-6. The expression of IL-6 is correlated with patients' age (older than 45 years old), tumor size (larger than 2 cm), HPV 16 or HPV 18 infection, and squamous epithelial carcinoma. Cultured human keratinocyte cell line (SK-v cells) harboring and expressing integrated HPV16 DNA also has the ability to constitutively secrete IL-6 (6). IL-6 released by human cervical cancer cells is indicated to have an antiapoptotic effect through the up-regulation of its downstream regulator Mcl-1 (7). Taken the above studies together, the HPV-related cervical tumorigenesis is suggested to be associated with the up-regulation of antiapoptotic Mcl-1 expression by the autocrine and/or paracrine IL-6.

A number of reports have revealed that HPV DNA can be detected in lung carcinomas (reviewed in ref. 8). Although the HPV infection rates in lung cancer patients varied greatly among different countries (0-100%), 21.7% of the total 2,458 carcinomas analyzed contain HPV DNA, with HPV 16 as the predominant type, followed by HPV 18. Our previous report showed that HPV 16/18 infection is associated with nonsmoking Taiwanese female lung cancer (9). More recently, our study revealed that HPV16/18 E6 may be involved in p53 inactivation to down-regulate p21^WAF1/CIP1 and mdm2 transcription (10). IL-6 levels in sera of lung cancer patients are significantly increased, and the IL-6 levels higher than 130 pg/mL is correlated with a poorer prognosis of lung cancer (11–13). It is therefore believed that IL-6 takes part in the regulation of lung cancer cell growth. Accordingly, we hypothesized that the up-regulation of IL-6 and Mcl-1 expression associated with HPV infection might promote lung cancer tumorigenesis through the similar scenario revealed in cervical cancer. Therefore, the present study aimed to examine the expression of IL-6 and its downstream antiapoptotic protein Mcl-1 in non–small cell lung carcinoma from clinical patients with or without HPV infection. Putative correlation between clinical manifestations of the study subjects with the IL-6 and Mcl-1 expressions was statistically analyzed. In addition, the expressions of IL-6 and Mcl-1 were investigated in HPV 16/18 E6- and E7-transfected A549 lung cancer cells and HPV16-harboring TL-1 lung cancer cells, established from the pleural effusions of lung cancer patient, for further illustrating the role of HPV in lung tumorigenesis.

Study subjects. Paraffin-embedded lung cancer tissues from 79 patients with primary lung cancer (27 females and 52 males; average age, 64.3 ± 11.4 y) who had undergone thoracic surgery were collected. None of the subjects had received radiation therapy or chemotherapy before surgery. Information regarding patients' smoking history was obtained by questionnaire. Smokers and nonsmokers were defined as current smokers who had smoked up to the day of pulmonary surgery and life-time nonsmokers, respectively. The characteristics of study subjects including age, gender, smoking status, tumor type, and tumor stage of lung cancer patients were listed in Table 1. The procedures followed were in accordance with the current revision of the Helsinki Declaration, and the informed consent was obtained from each study subject.

Detection of HPV 16/18 infection in lung tumor tissues by nested PCR. Genomic DNA was prepared from tissue sections by phenol-chloroform extraction, ethanol precipitation, and finally dissolved in 20 μL of sterile distilled water. HPV viral DNA was first amplified with type consensus primers MY09 and MY11, followed by a second round of amplification with type-specific primers flanking the L1 region to identify the subtype (14). Ten microliters of the final PCR products 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 of the samples was amplified to exclude false-negative results, whereas DNA preparations from the SiHa cells (containing HPV 16) and the HeLa cells (containing HPV 18) were used as positive controls.

Immunohistochemical staining of IL-6 and Mcl-1. The IL-6 and Mcl-1 proteins were immunohistochemically assessed on air-dried 5-mm formalin-fixed, paraffin-embedded sections using commercially available anti-human IL-6 and Mcl-1 rabbit polyclonal antibodies (Santa Cruze Biotechnology), respectively. Briefly, the sections were placed in a microwaveable container, submerged in 10 mmol/L citrate buffer (pH 6.0), wrapped in vented cling film, and incubated for two 5-min periods at a maximum power in a domestic microwave. After microwaving, the sections were allowed to equilibrate to room temperature in the buffer, and then rinsed in distilled water. The IL-6 and Mcl-1 antibodies were respectively applied to the sections at a 1:50 and 1:500 dilutions for overnight at 4°C. Immunoreactivities of both IL-6 and Mcl-1 were shown using the universal-labeled streptavidin-biotin horseradish peroxidase kit (Dako Denmark A/S, Produktionsvej) according to the manufacturer's instructions. The sections were counterstained in hematoxylin and scored semiquantitatively by microscopic examination of the entire tumor field. Cytosolic immunoreactivities of IL-6 and Mcl-1 were scored as negative if <10% of tumor cells showed cytosolic positivity, and positive if >10% of tumor cells showed cytosolic positivity.

Cell culture. A human lung adenocarcinoma cell line A549 and a cervical carcinomas cell line CasKi were obtained from American Type Culture Collection. Cells were maintained in DMEM supplemented with 10% fetal bovine serum, penicillin (100 units/mL), streptomycin (0.1 mg/mL), and 2 mmol/L glutamine at 37°C in a 5% CO₂ atmosphere.

Transfection of HPV 16/18 E6 and E7 into A549 lung cancer cells. Full-length E6 and E7 of HPV 16 or 18 was amplified by PCR from CasKi and HeLa cells, containing the viral genome of HPV 16 and HPV18, respectively. The resulted PCR products were purified by GENECLEAN III kit (Qbiogene). Purified fragments were cloned into a eukaryotic expression vector, pcDNA3.1/V5-His TOPO TA Expression kit (Invitorogen), and DNA preparations of resultant recombinants were transfected into A549. On the day before transfection, A549 cells were seeded at 1 × 10⁵ cells per well. Cells at 30% to 50% confluence were washed twice with phenol red–free DMEM medium without fetal bovine serum after an overnight incubation. After the incubation in 1 mL phenol red–free DMEM medium with 10% FBS for 3 h, cells were further incubated with calcium chloride-HEPES-buffer saline containing recombinant DNA solution. The medium was aspirated after 4 h of incubation, the cells were shocked with glycerol solution for 30 s, and washed twice with PBS. Stable transfectants were selected by culturing cells in medium containing antibiotic G418.

Establishment of HPV16-infected TL-1 lung cancer cell lines. HPV-infected TL-1 lung tumor cells were established from pleural effusions of lung cancer patients by the Ficoll-Paque method. The clinical characteristics of these patients and the identification of isolated cell lines were as previously described (10).

Silencing of endogenous HPV 16 E6 or E7 expression by RNA interference. The RNA interference (RNAi) target sequences for HPV16 E6 or E7 have been previously verified (10, 15, 16). To respectively suppress transcription of the endogenous HPV16 E6 and E7 gene, SiHa cells and TL-1 cells were transiently transfected with synthetic small interfering RNA against HPV16 E6 and E7, respectively, using Oligofectamine reagent (Invitrogen) according to the manufacturer's instructions. The detailed procedures were done as described previously (10).

Western blot. Cell lysates were harvested using SDS lysis buffer [20 mmol/L Tris (pH 8.0) and 1% SDS]. Twenty micrograms of cell lysates were applied and separated by 12% SDS-PAGE, transferred to polyvinylidene difluoride membrane, and immunoblotted overnight with 0.2 μg/mL antibodies against Mcl-1, HPV16 E6, HPV 18 E6, HPV 16 E7, HPV 18 E7, or β-actin, respectively (Santa Cruz Biotechnology). The membrane was then washed thrice and incubated with secondary antibody conjugated with horseradish peroxidase for 90 min at room temperature. After washing five times with TBS-Tween buffer [10 mmol/L Tris-HCl (pH 7.5), 150 mmol/L NaCl, and 0.05% Tween 20], the membrane was developed with enhanced chemiluminescence kit according to the manufacturer's protocol (ECL kit; Amersham Biosciences). β-actin in all of the samples was probed to exclude false-negative results and cell lysates from CasKi cells (containing HPV 16) were used as positive controls.

Statistical analysis. Differences of IL-6 and Mcl-1 expressions among gender, smoking status, tumor type, tumor stage, grade, and tumor-node-metastasis value were calculated by χ² test. Logistic regression analysis was used to identify important variables for IL-6 and Mcl-1 expressions, as well as the risk factors of lung cancer. The difference between groups was evaluated by χ² test or Fisher's exact test for categorical data (sex, histologic subtype, differentiation, HPV infection, T factor, N factor, and M factor). Level of statistical significance was defined as having a P value of <0.05. Survival analyses were done using Kaplan-Meier method (with 95% confidence intervals).

IL-6 and Mcl-1 expressions in non–small cell lung carcinoma were significantly associated with HPV infection. The representative results of IL-6 and Mcl-1 immunostainings were shown in Fig. 1. Both IL-6 and Mcl-1 were uniformly detected in cytoplasma of lung tumor cells, whereas no IL-6 and Mcl-1 expressions were observed in the adjacent normal cells (Fig. 1G and H). The positive immunostaining rate of IL-6 and Mcl-1 expressions in lung tumors was 70.9% (56 of 79) and 57.0% (45 of 79), respectively (Table 2). Both IL-6 and Mcl-1 expressions in non–small cell lung carcinoma were significantly correlated with HPV16 or 18 infection (P = 0.027 for IL-6 and P = 0.006 for Mcl-1), and IL-6 expression was significantly correlated with tumor stage (P = 0.01; Table 2). No significant association between Mcl-1 expression and other demographic information or clinical manifestations was found. It suggested that HPV-infected lung tumors would express IL-6 and Mcl-1 that were possibly involved in lung tumorigenesis. Notably, IL-6 was detected in tumor tissues of all stage II patients (Table 2).

IL-6 expression was significantly correlated with Mcl-1 expression. In human multiple myeloma cells, IL-6 plays important roles in the tumorigenesis by up-regulating Mcl-1 expression through Janus-activated kinase/signal transducers and activators of transcription 3 pathways (17, 18). Twenty-three (29.1%), 11 (13.9%), 0 (0%), and 45 (57.0%) of the IL-6/Mcl-1 expression pattern in our lung tumor tissues were −/−, +/−, −/+, and +/+, respectively (Table 3). The expression of IL-6 in lung tumors was significantly correlated with Mcl-1 expression (P < 0.0001; Table 3). Our results suggested that a positive IL-6 expression was the prerequisite of Mcl-1 expression in lung tumors because none of the IL-6–negative lung tumors showed Mcl-1–positive immunostaining. This observation might reflect the possibility of IL-6 up-regulating Mcl-1 expression in lung tumor tissues, as reported in cervical cancer, hepatoma, and prostate cancer (7, 19–23).

Concurrent IL-6 and Mcl-1 expression in non–small cell lung carcinoma was significantly associated with HPV infection. For elucidating the correlation between IL-6/Mcl-1 expression and HPV 16/18 infection, the relationship between the above variables were further investigated. As shown in Table 4, whereas 16 of 39 (41.0%) HPV noninfected patients had concurrent negative IL-6 and Mcl-1 expressions, only 7 of 40 (17.5%) of HPV-infected patients had concurrent IL-6/Mcl-1–negative expression. The presence of negative and positive HPV 16/18 infection in patients with IL-6+/Mcl-1− staining was only 18.0% and 10.0%, respectively. Intriguingly, no HPV 16/18 infection in patient with IL-6−/Mcl-1+ expression was identified. Another noteworthy observation was that 16 (41.0%) and 29 (72.5%) patients, who were respectively identified to be negative and positive for HPV, showed concurrent IL-6/Mcl-1 positivity. In addition to the respectively significant association between IL-6 and Mcl-1 expression with HPV16/18 infection in lung tumors (Table 2), a significant correlation between concurrent IL-6/Mcl-1 expression with HPV infection was also identified (P = 0.018; Table 4). The results were consistent with a previous study showing that both Mcl-1 and IL-6 expressions could be detected in human cervical cancer tissues and cell lines (7). Our observation revealed that IL-6 and Mcl-1 expressions were colocalized with HPV DNA in lung tumors.

Expressions of IL-6 and Mcl-1 were up-regulated in HPV16/18 E6- and E7-transfected A549 cells. According to the above observations, we hypothesized that the elevated IL-6 expression in HPV-infected lung tumor tissues may promote cell growth by up-regulating the antiapoptotic Mcl-1 expression. Therefore, it was intriguing to investigate IL-6 and Mcl-1 levels in HPV 16/18 E6- and E7-expressing lung cancer cells for elucidating the role of HPV infection in lung tumorigenesis. Stable clones of HPV 16/18 E6- and E7-transfected A549 cells were established, respectively, and a putative modulation of Mcl-1 and IL-6 expressions by HPV 16/18 E6 and E7 was subsequently investigated. IL-6 secretions were significantly elevated in HPV 16/18 E6–transfected (1,084 ± 102 pg/mL for HPV 16, column 3; 1,616 ± 48 pg/mL for HPV 18, column 4; Fig. 2A) and HPV16/18 E7–transfected (698 ± 54 pg/mL for HPV 16, column 3; 720 ± 36 pg/mL for HPV 18, column 4; Fig. 2B) A549 cells, compared with that of the parental (330 ± 6 pg/mL) and control vector–transfected (358 ± 4 pg/mL) counterparts (Fig. 2A and B). The Mcl-1 expression was also dramatically increased in HPV 16/18 E6- and E7-transfected cells (Fig. 2C and D). Intriguingly, the IL-6 levels seemed significantly higher in HPV 18 E6-transfected cells, compared with that in HPV 16 E6 counterparts (Fig. 2A). Whereas, the elevated IL-6 levels in HPV 16 E7 and HPV 18 E7 transfectants as well as the up-regulated Mcl-1 expressions in HPV 16/18 E6- and E7-transfected cells are comparable (Fig. 2). These results indicated that IL-6 and Mcl-1 were concomitantly up-regulated by both HPV 16/18 E6 and E7 in lung cancer cells.

IL-6 and Mcl-1 expressions were concomitantly up-regulated by E6 and E7 in HPV16-infected TL-1 lung cancer cells through phosphatidylinositol-3-OH kinase pathway. To verify whether IL-6 and Mcl-1 were up-regulated by HPV infection, we further examined IL-6 and Mcl-1 expressions in our previously established HPV16-infected TL-1 lung cancer cells (10) by using RNAi techniques to specifically knockdown E6 or E7 expression. Our data indicated that not only IL-6 (Fig. 3A) but also Mcl-1 expressions (Fig. 3B) were significantly reduced in E6 and E7 knockdown TL-1 cells, compared with that in the nonspecific RNAi transfectant or parental cells. Interestingly, the up-regulating effects of E6 to both IL-6 and Mcl-1 were more potent than that of E7. Specific inhibitors for phosphatidylinositol-3-OH kinase (PI3K; LY294002), Janus-activated kinase/signal transducers and activators of transcription 3 (AG490), and mitogen-activated protein kinase (PD98059) signal pathways were subsequently applied to verify the pathway responsible for IL-6 signaling to up-regulate Mcl-1 in HPV16-infected TL-1 cells. Our results showed that Mcl-1 expression was significantly reduced by LY294002, instead of AG490 and PD98059 (Fig. 4). These results suggest that Mcl-1 up-regulation by IL-6 in HPV16-infected TL-1 lung cancer cells is predominately mediated by PI3K signaling pathway.

Cytokines in tumor microenvironment may originate from autocrine supply by the tumor cells themselves or from the environment induced by infection, inflammation, or tissue damage. Our observation regarding IL-6 expression in lung tumors was comparable with previous reports (24, 25), in which the authors considered tumor cells were the major IL-6–secreting cells. Cervical epithelial cells have been shown to be potential sources of IL-6, which is detected in the cytoplasm of cervical carcinoma cells and high-grade dysplasias (5). In vitro, HPV-positive cervical carcinoma cells can constitutively produce IL-6. Our results supported the conclusion from previous study (5) that IL-6 is produced mainly by tumor cells because IL-6 and Mcl-1 can be uniformly detected in the cytoplasma of tumor cells but not detected in the adjacent normal cells (Fig. 1G and H). The expressions of IL-6 and Mcl-1 are correlated with HPV 16/18 infection in our patients. Taking the study from Wei et al. (5) and our observation together, it reveals that HPV-infected tumor tissues could secret higher levels of IL-6. Nevertheless, we were not able to find the significant correlation between IL-6 expression and squamous cell carcinoma as described by Wei et al. (5), although a trend of higher IL-6 expression frequency in squamous cell carcinoma (34 of 44; 77.3%) than that in adenocacinoma (22 of 35; 62.9%; P = 0.21; Table 2) was observed (5). Our results agreed with previous reports regarding cervical cancer that indicate that during host defense responses to infections, IL-6 is induced in the infected keratinocytes and in immune cells by microbial products such as bacterial lipopolysaccharides, viral RNAs, and viral proteins.

It is intriguing to understand the implications of HPV infection to induce highly local expression of IL-6 in lung tumor cells. Study of cervical cancer can serve as a good model to address and further infer the correlation between IL-6 secretion and HPV infection in lung cancer because growing amount of evidence has verified that higher IL-6 levels may promote tumor development of cervical cancer, in particular, among the patients with HPV 16/18 infections. In addition to cervical cancer, the up-regulated Mcl-1 mRNA and proteins by IL-6 stimulation are also observed in basal cell carcinoma and human hepatoma cells (23, 24). In regard to the study of IL-6 in lung cancer, clinical investigations show that increased serum levels of IL-6 can be found in patients with lung cancer, whereas IL-6 is not detected in the serum of patients with benign lung diseases (13). Cancer cell lines of non–small cell origin have been found to express IL-6 mRNA and proteins, and antisense oligonucleotides targeting IL-6 resulted in reduced IL-6 synthesis and cellular proliferation (26). These results suggest a possible autocrine role of IL-6 in the growth regulation of lung cancer cells. Nevertheless, the clinical implications of microenvironmental IL-6 expression in lung cancer patients are seldom addressed. Systemic IL-6 levels can be affected by various conditions such as infection, wound healing, or malignancy at distal lesions in lung cancer patients. Although analysis of local IL-6 levels in tumor lesions is quite invasive and not suitable for clinical applications, we believe the investigation of microenvironmental cytokine levels should be able to faithfully reflect local pulmonary situations. Our study provides the first evidence that IL-6 and its downstream Mcl-1 proteins can be concurrent detected in lung tumor tissues and may participate in the tumorigenesis of HPV-infected lung cancer.

Accumulative studies have shown that Mcl-1 can be up-regulated by IL-6. Antiapoptotic protein Mcl-1 is concomitantly expressed with IL-6 in human cervical cancer tissues and cell lines but not in normal cervix tissues (7). Mcl-1, but not other Bcl-2 family members, is rapidly up-regulated in human cervical cancer C33A cells upon IL-6 treatment (7). Study from Brocke-Heidrich et al. (27) further indicated that among members of the Bcl-2 family, Mcl-1 gene is the only one that can be regulated by IL-6 in human myeloma cells. The study also revealed that Mcl-1 gene expression is decreased in association with myeloma cell apoptosis upon IL-6 starvation, although this decreased expression and the cell survival can be reversed by adding IL-6. These abovementioned data further emphasize the major role of antiapoptotic Mcl-1 protein in the IL-6–induced survival of human cancer cells.

The positivity of Mcl-1 in our study (57%) was comparable with a previous report in which Mcl-1 is present in 58% of 49 specimens from patients with radically resected non–small cell lung carcinoma (28). Higher IL-6 and Mcl-1 production is observed in cancer tissues than in adjacent normal tissues, particularly, in patients with HPV 16/18 infections. In this study, we further prove that Mcl-1 expression levels are up-regulated in HPV16/18 E6- and E7-expressing A549 lung cancer cells (Fig. 2). Notably, levels of IL-6 and Mcl-1 expression in E6- and E7-transfected A549 cells (Fig. 2) and full HPV 16 genome–containing TL-1 cells (Fig. 3) were quite comparable. It implied that only E6 or E7 alone might be required to up-regulate the expressions of IL-6 and Mcl-1. The above observations of a significant correlation between concurrent IL-6/Mcl-1 expression and HPV 16/18 infections as well as the up-regulated Mcl-1 expression by HPV16/18 in lung cancer cells support the findings in cervical cancer and suggest that HPV 16/18 infections in lung cancers might play a similar role, as in cervical cancer, to promote tumor cell proliferation through the induction of IL-6 and downstream antiapoptotic Mcl-1 protein expression. In this context, IL-6 may have a synergistic effect to oncogenic HPV infection in the pathogenesis and progression of lung cancer by activating the downstream antiapoptotic Mcl-1 expression, as evidenced by the significant correlation between concurrent IL-6 and Mcl-1 expression with HPV infection among our patients (P = 0.018; Table 4).

Another interesting issue is the IL-6 signaling to up-regulate Mcl-1 expression in lung cancer cells. In human cervical cancer, IL-6 regulates Mcl-1 expression via a PI3K/Akt–dependent pathway that facilitates tumorigenesis (7). However, IL-6 mediated Mcl-1 up-regulation in human multiple myeloma cells is through Janus-activated kinase/signal transducers and activators of transcription 3 pathway (18). In this study, the up-regulated Mcl-1 expression by IL-6 in HPV16-infected TL-1 lung cancer cells is predominately by PI3K pathway, instead of Janus-activated kinase/signal transducers and activators of transcription 3 and mitogen-activated protein kinase pathways. Therefore, the signaling pathway(s) leading to the Mcl-1 induction may vary according to cell type or stimuli used (29). However, the underlying mechanisms of E6 and E7 to up-regulate IL-6 and Mcl-1 in lung cancer cells needs further investigation because Mcl-1 expression is not entirely repressed by the application of PI3K inhibitor (Fig. 4).

In conclusion, our study provides both in vivo and in vitro evidence of an association between inflammation and HPV infection in lung cancer. In addition to demonstrating that IL-6 and Mcl-1 are coexpressed in human lung cancer tissues, this study is the first to indicate that both HPV 16/18 E6 and E7 can up-regulate IL-6 and antiapoptotic Mcl-1 expressions, which, subsequently, may promote the tumor progression of HPV-infected lung cancer. The microenvironmental high IL-6 levels expressed by lung cancer cells in response to HPV stimulus, therefore, is likely to play a role in lung tumorigenesis through autocrine and/or paracrine mechanisms. Therefore, the HPV-mediated IL-6 and Mcl-1–dependent antiapoptotic effects may play an important role in HPV-associated lung tumorigenesis. To the best of our knowledge, this is the first work to show that HPV can up-regulate the proinflammatory cytokine IL-6 and antiapoptotic Mcl-1 expressions, which may potentially influence lung tumorigenesis. Hopefully, the present study can provide novel clues and shed new insights to the mysterious role of HPV infection in lung cancer.

No potential conflicts of interest were disclosed.