Published ahead of print on October 27, 2005, doi:10.1165/rcmb.2005-0117OC
© 2006 American Thoracic Society DOI: 10.1165/rcmb.2005-0117OC Cigarette Smoking Induces Overexpression of Hepatocyte Growth Factor in Type II Pneumocytes and Lung Cancer CellsDepartment of Health, Feng-Yuan Hospital, Feng-Yuan; Department of Surgery, Chang-Hua Christian Hospital, Chang-Hua; Graduate Institute of Biomedical Sciences and Graduate Institute of Veterinary Microbiology, National Chung Hsing University, and Department of Surgery, Taichung Veterans General Hospital; Department of Health Care Administration, Central Taiwan University of Science and Technology, Taichung; and Department of Health, Hsin Chu General Hospital, Hsin Chu, Taiwan Correspondence and requests for reprints should be addressed to Kuan-Chih Chow, Graduate Institute of Biomedical Sciences, National Chung Hsing University, 250 Kuo-Kuang Road, Taichung, 40227 Taiwan. E-mail: kcchow{at}dragon.nchu.edu.tw
We examined gene expression of hepatocyte growth factor (HGF) and HGF receptor (HGFR), or product of proto-oncogene c-met (c-met), in smokers and nonsmokers with adenocarcinoma (ADC) by suppression subtractive hybridization and microarray techniques. Expression of HGF and c-met was confirmed by RT-PCR. HGF content in the respective tumor mass and nontumor lung tissue was measured by ELISA. HGF in pathologic samples was localized by immunohistochemistry and in situ hybridization. Our results indicate that overexpression of HGFR was frequently detected in ADC cells, whereas overexpression of HGF was detected in alveolar type II (ATII) cells. Overexpression of HGF was correlated with cigarette smoking and tumor stages. In vitro, HGF expression was evaluated in isolated murine ATII cells and in 12 ADC cell lines, and we found that nicotine activated HGF expression in ATII cells and lung cancer cells.
Key Words: immunohistochemistry • in situ hybridization • microarray • suppression subtractive hybridization • type II pneumocyte
Lung cancer is the second leading cause of cancer death in Taiwan. The annual mortality rate of lung cancer (  6,000 deaths) is 20% of total cancer-related deaths (1). In the United States, the annual death rate of lung cancer (> 150,000 deaths) is 30% of total cancer-related deaths (2). Nearly 90% of deaths from lung cancer are associated with tobacco smoking (3). According to the presence or the absence of neuroendocrine features, lung carcinoma can be categorized into small cell lung cancer or non–small cell lung cancer (NSCLC) (4). Based on the histopathologic characteristics, NSCLC can be subcategorized into adenocarcinoma (ADC), squamous-cell carcinoma, and large-cell carcinoma (4). Among these, ADC, which is most commonly found in women and smokers, is associated with a higher frequency of drug resistance and mortality of lung cancer in Taiwan.
The peripheral distribution and intrapulmonary spreading pattern have indicated that the pathologic origin of lung ADC could be bronchioalveolar. Detection of surfactant gene expression in lung ADC cell lines H441 and A549 has suggested that the cellular basis of ADC could be alveolar type II (ATII) cells (type II pneumocyte or type II alveolar epithelial cells) (5). The fact that, after local alveolar injury (e.g., by cigarette smoking), ATII cells can be induced to actively proliferate and differentiate into ATI pneumocytes to repair the damaged alveolar epithelium has indicated that ATII cells are capable of regeneration (6, 7). The regeneration could be mediated by growth factors, including epidermal growth factor, transforming growth factor-
Biochemically, HGF is synthesized and secreted as an inactive single-chain glycoprotein that is activated by serine protease cleavage to form heterodimeric peptides consisting of a 69-kD In this study, we combined suppression subtractive hybridization (SSH) and microarray (15) to examine the gene expression of HGF and c-MET in smokers and nonsmokers with ADC. Expression of HGF and c-MET was confirmed by RT-PCR. HGF content in the respective tumor mass and nontumor lung tissue (NTLT) was measured by ELISA. HGF signal in pathologic samples was localized by immunohistochemistry and in situ hybridization (ISH). Correlation between HGF expression and cigarette smoking was evaluated statistically. In vitro, HGF expression was measured in 12 lung ADC cell lines. The effect of nicotine on cancer cells and on ATII cells was also evaluated.
Tissue Specimens, Lung ADC Cell Lines, and Isolation of ATII Cells From January to December 2001, sera and tissue specimens were collected from 264 patients with newly diagnosed NSCLC. Samples from all patients, for whom at least one follow-up examination or death was documented, were pathologically confirmed NSCLC. Of the 264 patients, 107 were diagnosed as having lung ADC. The stage of the disease was classified according to the new international staging system for lung cancer (16). The Medical Ethical Committee approved the protocol, and written informed consent was obtained from every patient before surgery. All patients had undergone surgical resection and radical N2 lymph node dissection. Tumor size, lymph node number, differentiation, vascular invasion, and mitotic number were evaluated. Patients with lymph node involvement or locoregional recurrence received irradiation at the afflicted areas. Patients with distant metastasis were treated with chemotherapy. After treatment, patients were routinely followed every 3–6 mo in the Outpatients department. Tumor recurrence and metastasis were detected when blood examination, biochemical studies, chest radiography, abdominal sonography, whole-body bone scan, and computerized tomography chest scans showed evidence of the disease. The average age of the male patients (n = 62) was 60.6 ± 1.41 yr, and that the average age of the female patients (n = 15) was 53.3 ± 1.27 yr (P = 0.0154). ELISA, ISH, and immunohistochemical staining were performed using a single-blind procedure. Twelve ADC cell lines (H23, H226, H838, H1355, H1437, H2009, H2087, H2126, A549, CL1–0, CL1–1, and CL1–5) (17) were used for evaluation of HGF expression in vitro. Cells were grown at 37°C in a monolayer in RPMI 1640 supplemented with 10% FCS, 100 IU/ml penicillin, and 100 µg/ml streptomycin. Mouse and human ATII cells were isolated and characterized by a previously reported method (18). Purity of the isolated ATII cells was determined by morphology and immunocytochemistry to surfactant protein B (Chemicon, Temecula, CA), an authentic marker for ATII cells.
RNA Extraction, SSH, and Microarray The resultant SSH cDNA libraries were labeled with fluorescent nucleotides, and the reaction mixture was hybridized to microarray slides (Taiwan Genome Sciences, Taipei, Taiwan) to determine the correlated gene(s). Microarray was done in triplicate, and each correlated gene was mapped out. The ratio between normalized fluorescent signal intensities of lung cancer tissues and nontumor tissues was measured on the individual spot of microarray, and values of the three readings were averaged. The cut-off ratio for highly expressed genes was set at 2.0 (with 99.5% coefficient variance in a scatter plot), and the cut-off ratio for downregulated genes was set at 0.5. The expression ratio of a gene that fell between 0.5 and 2.0 was considered constitutive. The resultant genes that were identified by microarray were further confirmed with colony membrane arrays and RT-PCR. Specificity of the amplified fragment was determined by DNA sequencing (Protech Technology, Taipei, Taiwan). The sequence was matched with sequences in the GenBank database.
RT-PCR
ELISA The concentration of HGF was determined by ELISA kits (human HGF immunoassay, Quantikine; R&D Systems Inc., Minneapolis, MN). The frozen sample was quickly weighed, thawed on ice, and placed in 0.5 ml of PBS before being homogenized on ice for 10 s. The homogenate was centrifuged at 10,000 x g for 10 min at 4°C, and the supernatant (cytoplasmic extract) was collected. Before HGF assay, 150 µl of assay diluent was added to equilibrate a well that had been precoated with monoclonal antibody specific to HGF before the addition of 50 µl of sample. The reaction was incubated at 37°C for 2 h. The microtiter plate was washed in PBS four times, and 200 µl of alkaline phosphatase–conjugated polyclonal antibody specific to HGF was added. The reaction was further incubated at 37°C for 2 h. After washing in PBS, 200 µl of tetramethylbenzidine was added, and the reaction was incubated at 37°C for 30 min. Positive signal was developed in 50 µl of 2N sulfuric acid at 37°C for 2 min. The chromogenic reaction was identified by reading at A450 nm (Sunrise, Tecan, Austria). The individual sample was done in duplicate. Samples with overscaled readings were diluted before further determination. HGF concentration of each sample was calculated according to reaction standards. The final HGF level was adjusted by weight of the tissue and expressed as pg/mg of tissue.
Immunohistochemistry, Immunocytochemistry, and ISH
Statistical Analysis
Expression of HGF, c-MET, and GAPDH in ADC and NTLT To determine gene expression profiles of HGF and HGF receptor (c-met) in smokers and nonsmokers with ADC, we used SSH (Figure 1A) and microarray (Figure 1B) to examine 10 pairs of ADC and NTLT. The differential expression patterns of genes in different groups are listed in Table 2. HGF and c-MET, a pulmotrophic factor and its receptor, were located in two highly distinct groups: c-MET was expressed mainly in the ADC fraction (normalized expression ratio = 13.89), whereas HGF was expressed in the NTLT fraction (normalized expression ratio = 0.023). Results were confirmed by colony membrane arrays (Figure 1C), RT-PCR (Figure 2A, upper panel), and immunoblotting (Figure 2A, lower panel) in paired tumor and NTLT from 77 patients with lung ADC, in which glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was used as a monitoring standard of lung cancer fraction (20). Among 77 paired samples, HGF expression was identified in 42 (54.5%) tumors and in 61 (79.2%) NTLT. HGF was not expressed in five pairs of tumors or NTLT (Figure 2B). Expression of c-MET was detected in 14 (18.1%) NTLT and in 61 (79.2%) of 77 tumor samples. In an arbitrary grouping in which a smoker was defined as a person who smoked more than 20 packs per year, our data indicated that elevated HGF was associated with tumor staging and cigarette smoking (Figure 2B). Statistical analysis showed that HGF expression in tumor was correlated with gender, cigarette smoking, tumor staging, cell differentiation, and lymphovascular involvement. Elevated HGF in NTLT was correlated with gender, cigarette smoking, and tumor staging (Table 3) but not with cell differentiation and lymphovascular involvement.
Increased Tissue HGF Levels in Patients with Lung ADC HGF concentration in 77 paired tumor and NTTL was measured by an ELISA method. The average HGF level in the tumor fraction was 24.13 ± 1.84 pg/ml, and that in the nontumor fraction was 18.77 ± 1.91 pg/ml (P = 0.043). Because the previous results indicated that HGF concentration was correlated with cigarette smoking and because HGF levels in 33 (42.9%) of the NTLT fractions were higher than that in the ADC fractions, patients were divided into groups based on their smoking history, tumor stage, and tissue source. Increased HGF levels in NTLT fractions were associated with cigarette smoking and tumor staging (Table 4).
Identification of HGF Expression by ISH and Immunohistochemistry Protein signals of growth factor are rarely detected in the cells because growth factors can be released from cells immediately after completion of protein synthesis. In contrast, mRNA remains inside the cell. Therefore, ISH was used to determine cell types that expressed HGF mRNA in the pathologic sections. Among 61 NTLT samples that had high HGF content determined by ELISA, HGF mRNA was detected in ATII cells (Figure 3A) in 54 (88.5%) tissues. The ATII cells were further characterized by immunoreactivity to surfactant protein B (Figures 3B and 3C). In contrast, HGF mRNA was only detected in 37 (48.05%) tumor specimens (Figure 3D). In 18 (29.5%) pathologic samples, HGF mRNA was identified in endothelial cells. These results indicate that higher levels HGF mRNA were correlated well with elevated HGF contents in tissues. These data further suggest that besides lung cancer cells, ATII cells could be a source of locally increased HGF.
Expression of HGF and c-MET in Cultured Lung Cancer and ATII Cells Our data indicate that ATII and ADC cells were potential sources for the locally increased levels of HGF and that cigarette smoking could have induced HGF expression in these cells. To further characterize such correlations, we screened 12 cultured lung ADC cells for the in vitro expression of HGF and c-MET. Spontaneous HGF expression was detected in five lung cancer cells by RT-PCR, and c-MET expression was detected in all of the lung cancer cell lines (Figure 4A). HGF secreted from ADC cells, which was determined by ELISA, was significantly elevated in RT-PCR–positive cells (Figure 4B). Content of c-MET determined by immunoblotting did not match that determined by RT-PCR (Figure 4C). Although we suspect that this could be a result of post-transcriptional regulation, this has yet to be determined.
We used a crisscross ELISA method to determine the optimal concentration and timing for nicotine to have an effect on HGF expression in ADC cells. In a 96-well plate, 5 x 104 cells were seeded into each well for 18 h. To a set of quadruplicate wells, various concentrations of nicotine (0, 0.1, 0.25, 1.0, 4.0, and 10 µM) were added and incubated with cells for different lengths of time (0, 6, 12, 18, 24, and 48 h); treatment with 10 µM nicotine (Sigma) (Figure 4D) for 24 h (Figure 4E) was found to be the best condition for inducing HGF expression. A slight suppression of HGF was detected in H1437 cells. The results were further confirmed by RT-PCR (Figure 4F, upper panel). The respective changes in HGF expression determined by quantitative real-time RT-PCR were 5.6-fold in H226, 17.2-fold in H838, 0.87-fold in H1437, 19.03-fold in H2087, and 3.7-fold in A549 cells (Figure 4F, lower panel).
In isolated murine ATII cells, which were determined by immunoreactivity to surfactant protein B to have purity of 93–96%, treatment with 10 µM of nicotine induced HGF expression (Figure 5A). The induction effect of nicotine was suppressed by the addition of 100 µM mecamylamine. To study the in vivo effect of cigarette smoking on ATII cells and HGF expression, Balb/c mice (National Animal Center, National Science Council, Taipei, Taiwan) at 6 wk of age were treated with passive cigarette smoking daily (5 min/d in a 40 x 40 x 60 cm chamber; each cigarette contained 0.9 mg of nicotine) for 4 wk before histopathologic and biochemical examinations. Compared with mice that had not been exposed to cigarette smoke (Figures 5B-a and 5B-b), the thickness of bronchial epithelium and cellularity in alveoli increased in mice that had been exposed to cigarette smoke (Figure 5B-c). The increased cell type was mostly HGF positive (Figure 5B-d). Nicotine acetylcholine receptor
Our results show that an increased HGF level in tissue from patients with lung ADC correlated with cigarette smoking and tumor staging. Detection of HGF mRNA in type II pneumocytes and tumor cells further indicated that these two cell types could be the important local source of elevated HGF in patients with NSCLC. In particular, from ATII cells, HGF, which is an autocrine factor for alveolar regeneration and repair (5, 6, 11), can turn into a potent paracrine factor for enhancing matrix angiogenesis and increasing the metastatic potential of lung cancer cells (21). By demonstrating that in patients with more advanced lung cancer, especially in those who smoked more than 20 packs per year, HGF expression was highly expressed in NTLT and tumor cells, our data indicate that, in addition to pulmonary carcinogenesis, cigarette smoking could induce HGF expression in ATII and tumor cells. The mechanism of how cigarette smoking activates gene expression of HGF remains to be determined.
After embryo implantation, HGF has been shown to play a crucial role in the development of the placenta, and the major source of HGF was identified to be from stromal villus cells (22). The production of HGF then gradually increases, reaches its maximum concentration in the second trimester, and decreases at term (23). During embryonic formation, HGF also plays a vital role in the morphogenesis of the fetus, in particular in angiogenesis and organogenesis of lung, kidney, and liver (21). In vitro, supplementation of HGF enhances branching expansion of pulmonary tubules (21) and increases conversion of metanephric mesenchymal cells into epithelial progenitors to renovate organ morphogenesis (21). The mitogenic and morphogenic activity of HGF was thus thought to be involved in the regeneration of these organs. Shigemura and colleagues (24) showed that in the adult rat, ectopic HGF could improve the pulmonary function of elastase-induced emphysema by inhibiting the apoptosis of alveolar cells and inducing proliferation of endothelial and ATII cells. However, elevated HGF, which was repeatedly found in tracheal effluents of premature infants, might be harmful to fragile, underdeveloped lung (25). Ectopic HGF expression was also shown to enhance murine liver regeneration when the animal was partially hepatectomized. In the presence of tumor, HGF increased the invasiveness of cancer cells (21). By using in situ RT-PCR, Hata and colleagues (26) demonstrated that, in chemical-induced liver cirrhosis, the increased HGF mRNA was identified mainly in sinusoidal endothelial cells and Kupffer cells. These results further support the finding that, within limited damage of an organ, the source of HGF could be mostly local. Although systemically increased HGF was frequently observed under different pathologic conditions (27), the omnipresence of c-MET on endothelial cells (28) and the high affinity of HGF to c-MET (dissociation constant Kd These observations, considered together with our results, indicate that the source of HGF in NSCLC must be local if increased HGF level is a pathologic basis of lung regeneration to regain pulmonary function in response to tumor invasion. The presence of atypical adenomatous hyperplasia in precancer lesions and at the leading front of tumor nests would favor such an interpretation (31). A study of idiopathic pulmonary fibrosis by Sakai and colleagues has shown that hyperplastic ATII cells can express HGF mRNA (32). Our results confirm their findings and show a close relationship between cigarette smoking and increased HGF levels in patients with NSCLC. In those patients, the effect of cigarette smoking on ATII cells or lung cancer cells should be comparable to that of N-nitrosodimethylamine or dimethylnitrosamine on HCC and sinusoidal endothelial cells (27, 33). Moreover, identification of nicotine acetylcholine receptor (34) on ATII and NSCLC cells and the induction effect of nicotine on HGF expression, which could be suppressed by mecamylamine, suggest that cigarette smoking might have a direct effect on these cells per se. The impact of cigarette smoking on HGF expression in ATII and cancer cells remains to be clarified if this is one of the pathophysiologic factors that affect tumor growth and invasion (21, 27, 33). Although these issues are being evaluated in an ongoing study, other explanations are possible. In conclusion, our data show that HGF expression was frequently increased in NTLT and tumor cells in patients with NSCLC, particularly in patients who smoked or with previous smoking history. Elevated HGF concentration in NTLT correlated with increased HGF mRNA in ATII cells. Statistical analysis further showed that increased HGF levels in NTLT and tumor nest were closely associated with disease progression of lung cancer. In vitro, nicotine upregulated HGF expression in lung cancer tissues and in isolated ATII cells. Although the details have yet to be determined, our results suggest that, in addition to initiating carcinogenesis, cigarette smoking may play a key role in promoting tumor progression via activation of HGF expression in ATII cells and tumor cells in patients with NSCLC.
The authors thank Ms. Ya-Lin Chen, Mr. Alex Y. H. Shih (Taiwan Genome Sciences, Taipei, Taiwan), and Dr. Kai Wang (PhenoGenomics, Bothell, WA) for technical assistance.
This study was supported by National Science Council grant NSC93-2320-B-005-014 (K.C.C.) and NSC93-2314-B-371-004 and NSC94-2314-B-371-001 (T.S.L.), Taiwan.
* These authors contributed equally to this work. Originally Published in Press as DOI: 10.1165/rcmb.2005-0117OC on October 27, 2005 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form March 28, 2005 Accepted in final form September 27, 2005
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
(
-chain. The heterodimeric peptides stimulate epithelial cell proliferation and migration (
5). The chi-square test for trend was used when corresponding pathologic factors (e.g., cell differentiation or tumor stage) exceeded two categories. Statistical significance was set at P < 0.05. Statistical analysis was performed by Prism4 statistical software (GraphPad, San Diego, CA). 
