The role of Thyroid Transcription Factor-1 and Tumor differentiation in Resected Lung Adenocarcinoma

To investigate the role of thyroid transcription factor-1 (TTF-1) and tumor differentiation in resected lung adenocarcinoma. A total of 520 patients with clinical early stage lung adenocarcinoma who underwent surgical resection were reviewed retrospectively. Clinical data and outcomes were evaluated with an average follow-up of 117 months. The results were validated via lung cancer cell line studies. The clinical parameters did not differ between relapse and nonrelapse patients. Exceptions were tumor differentiation, lymphovascular space invasion, F18-fluorodeoxyglucose maximum standard uptake value, tumor size, and pathological stage (p < 0.001). Poor tumor differentiation was the independent prognostic factor (odds ratio: 2.937, p = 0.026). The expression of TTF-1 was correlated with tumor differentiation in resected lung adenocarcinoma patients (p < 0.001). Five-year survival was 60.0% for score 1 TTF-1 expression patients, 80.1% for score 2 TTF-1 expression patients, and 86.1% for score 3 TTF-1 expression group patients. The lung cancer cell line study of knockdown and overexpression of TTF-1 revealed TTF-1 mediated High Mobility Group AT-Hook 2 (HMGA2) protein involved with epithelium-mesenchymal transformation. The chromatin immunoprecipitation revealed TTF-1 regulated HMGA2 via direct binding. TTF-1/HMGA2 axis was associated with tumor differentiation and mediated the aggressiveness of the tumor and prognosis.

Lung cancer remains the leading cause of cancer-related death. Despite advances in therapeutic drugs and newly developed target therapies, long-term survival is still unsatisfactory. Mortality following surgical resection is thought to be associated with tumor relapse 1,2 . Many studies have investigated the clinical and basic parameters associated with relapse after surgical resection of lung cancer. Tumors graded as poorly differentiated are found to be an independent prognostic factor in patients with stage I non-small-cell lung cancer (NSCLC) 3,4 , but the mechanism is not clear.
Thyroid transcription factor-1 (TTF-1), a member of the homeodomain-containing transcription factor family, activates the expression of selected genes in the thyroid, lung, and brain 5 . TTF-1 is known to play key roles in the control of embryonic development and differentiation 6,7 . The expression of TTF-1 is restricted to alveolar type II cells 8 . It has been found in all types of lung cancer but is reported frequently in adenocarcinoma 9,10 . TTF-1 has been shown to be a favor prognostic factor for survival in NSCLC 11 . Some studies have shown that differentiation grade is not a prognostic factor for lung cancer survival 12,13 . A dual role of TTF-1 expression in lung adenocarcinoma has been reported 14 . The mechanisms which control tissue-specific expression of TTF-1, tumor differentiation, and their prognostic significance are not fully understood. One study showed the possible mechanism of suppression of lung adenocarcinoma progression by TTF-1 via regulation of High Mobility Group AT-Hook 2 (HMGA2) protein 15 . We, therefore, aimed to clarify the role of TTF-1 in tumor differentiation and the aggressiveness of lung adenocarcinoma.

RNA Extraction and Reverse Transcription-Polymerase Chain Reaction.
Total RNA was isolated from cultured cells using TRIzol reagent (Thermo Scientific Life Sciences, Carlsbad, California, USA) following the manufacturer's protocol. cDNA synthesis was performed using a cDNA reverse transcription kit (ABI, Foster City, CA, USA). In the reverse transcription-polymerase chain reaction (RT-PCR) step, 50 pmol of each forward and reverse primer was used. Standard RT-PCR amplification conditions were applied as follows: 95 °C for 5 min, followed by 30 cycles of 95 °C for 30 s, 58 °C for 30 s, and 72 °C for 30 s, and a final amplification at 72 °C for 5 min. Each measurement was performed in duplicate, and the threshold cycle value was determined for each amplification curve. The geometric mean of endogenously expressed glyceraldehyde-3-phosphate dehydrogenase was used to normalize the expression. RT-PCR products were separated on 2% agarose gels and visualized with Health View nucleic acid stain under ultraviolet light.
Ectopic expression of TTF-1. The human TTF-1 cDNA ORF clone was designed and obtained from Origene Technologies, Rockville, MD, USA. Briefly, cells were seeded at 2 × 10 5 in 6-well culture plates and incubated overnight. The human TTF-1 cDNA ORF clone was transiently transfected into cells for 24 hr using Polyjet transfection reagents (SignaGen Laboratories, Gaithersburg, MD, USA) and followed by changing with the complete medium. Following 48 hr incubation, the cells were utilized for following experimental analysis.
SCIeNTIfIC RepoRTs | 7: 14222 | DOI:10.1038/s41598-017-14651-y Transwell invasion assay. Invasion ability was investigated by using Transwell invasion chambers. The culture inserts (8-μm pore size; BD Biosciences, Franklin lakes, NJ, USA) was precoated with 30 μg/insert Matrigel (BD Biosciences, Franklin lakes, NJ, USA) on top of the membrane. A total of 2.5 × 10 4 cells were seeded onto the upper chamber in medium containing 10% NuSerum. The lower chamber contained complete medium (RPMI supplemented with 10% FBS). The cells were allowed to invade toward the lower chamber overnight. On the second day, the invasive cells were fixed with methanol, stained with 50 μg/mL propidium iodide (Sigma, St. Louis, MO, USA) in ddH 2 O and analyzed using immunofluorescent microscopy. The amounts of invasive cells were quantified with ImageJ software.
Quantification of miRNA expression. The MicroRNAs was extracted, purified, and converted to cDNA using TaqMan TM Advanced miRNA cDNA Synthesis Kit (Thermo Fisher Scientific, Waltham, MA, USA). The expression of miR-33a was quantified by quantitative real time polymerase chain reaction (qPCR). The miR-33a primer was designed and synthesized by Origene Technologies, Rockville, MD, USA. qPCR was performed in a 20 μl reaction solution containing of 4 μl cDNA, 4 μl ddH 2 O, 1 μl of each primer, and 10 μl of SYBR Green I master mix (Roche, Basel, Switzerland) using the Roche LightCycler 480 Real-Time PCR system (Roche Applied Science, Penzberg, Bavaria, Germany).

Chromatin Immunoprecipitation. The ChIP assay was performed according to the instructions of Pierce
Magnetic ChIP kit (Pierce, Rockford, IL, USA). Briefly, cells were cross-linked by the addition of 16% formaldehyde to obtain a final concentration of 1% formaldehyde for 10 minutes at room temperature, and the reaction was terminated by the incubation with 1X glycine for 5 minutes at room temperature. Subsequently, the crosslinked cells were lysed with membrane extraction buffer containing proteinase/phosphatase inhibitor for 10 minutes on ice. Genomic DNA was fragmented into 200 bp segments using the combination of Micrococcal Nuclease and Bioruptor USD-200TM sonicator (Diagenode, Liege, Belgium). 10% of digested chromatin was used for input control and the remaining sample was further incubated with 5 μg TTF-1 antibody or normal rabbit IgG with rotation at 4 °C overnight, respectively. After incubation, 20 μL of the magnetic beads were added into each sample and incubated for 2 hours at 4 °C with mixing. The magnetic beads were washed sequentially with provided wash buffer and pelleted by centrifuging at 3000 × g for 5 minutes. The chromatin was eluted from the magnetic beads by incubation with elution buffer at 65 °C for 30 minutes with vigorous shaking. DNA-protein cross-links reversal was performed by using a high salt solution containing 6 μL of 5 M NaCl and 2 μL of 20 mg/ mL Proteinase K. The precipitated DNA was recovered using the DNA Clean-Up Column and collection tube and eluted with 30 μL of DNA Column Elution Solution. PCR was performed using the primers designed to recognize HMGA2 promoter containing TTF-1 recognition site, which were5′-CCCTTGTTATAAATTCAAAAGGGG-3′ and 5′-GGTAACGTGTACAGAGGTCAGCAA-3′.
Statistical Analysis. Descriptive data are expressed as mean ± standard deviation. Student's t test was used to investigate continuous variables, and the χ 2 test was used to compare categorical variables between groups. Overall survival (OS) and disease-free survival (DFS) were calculated using Kaplan-Meier survival analysis. Multiple logistic regression analysis was used to identify independent risk factors for patients with relapse. For statistical analysis, each cell experiment was performed in triplicate and repeated at least three times. SPSS 14.0 software (SPSS, Inc., Chicago, IL, USA) was used for all analyses, and statistical significance was defined as p < 0.05.

Results
Of 520 lung adenocarcinoma patients, 94 (18.77%) developed relapse after surgery. The common relapses included locoregional recurrence (lung, mediastinum, and pleura) and distant metastasis (brain, bone, liver, and adrenal gland). Sex distribution, age, smoking status, visceral pleural invasion, and serum CEA levels did not differ significantly between the patients with relapse and nonrelapse. In the relapse group, patients had a more advanced pathological stage, more central tumor location. The patients with relapse had more poorly differentiated tumors (p < 0.001), a higher percentage of lymphovascular space invasion (22.08%; 7.30% for nonrelapse groups, respectively, p < 0.001), higher maximum standard uptake values (SUVmax: 6.09 ± 4.23; 3.74 ± 3.70 for nonrelapse groups, respectively, p < 0.001), larger tumor size (2.89 ± 1.35 cm; 2.09 ± 1.29 cm for nonrelapse groups, respectively, p < 0.001), and high prevalence of lymph node involvement. The relapse group patients had high prevalence of positive lymph node both for N1 and N2. The results were showed in Tables 1 and 2. Patients developed with relapse had a poor prognosis with a short OS. 5-year OS was 38.4% in the relapse groups, and 97.3% in the nonrelapse groups respectively (p < 0.001) (Fig. 1A).
Multiple logistic regression revealed that poor tumor differentiation was an independent prognostic factor (odds ratio [OR]: 2.937, 95% confidence interval [CI]: 1.134-7.607; p = 0.026). Tumor size was also an independent prognostic factor (OR: 3.371, 95% CI: 1.427-7.959; p = 0.006) ( Table 3). The grade of tumor differentiation correlated with F 18 -fluorodeoxyglucose uptake, expressed as maximal standard uptake value (SUVmax). The SUVmax of well-differentiated tumors was 2.66 ± 2.64, that of moderately differentiated tumors was 4.75 ± 3.64, and that of poorly differentiated tumors was 7.59 ± 5.25 (p < 0.001) (Fig. 1B). OS and DFS significantly correlated with the grade of tumor differentiation. 5-year OS was 86.8% in the well-differentiated group, 82.5% in the moderately differentiated group, and 58.6% in the poorly differentiated group, p < 0.001 ( Fig. 2A). 5-year DFS was 95.1% in the well-differentiated group, 66.5% in the moderately differentiated group, and 50.4% in the poorly differentiated group, p < 0.001 (Fig. 2B). The expression of TTF-1 IHC staining was scored after excluded advanced after excluding advanced disease (pathological stage 3 & 4). 5-year OS was 60.0% in score 1 group, 80.1% in score 2 group, and 86.1% in score 3 TTF-1 group (Fig. 2C). Well-differentiated tumors exhibited stronger TTF-1 expression in IHC stain (Fig. 3A). Moderately differentiated tumors showed focally positive TTF-1 expression  3B). Poorly differentiated tumors showed negative TTF-1 expression (Fig. 3C). The expression of TTF-1 was associated with tumor differentiation in resected lung adenocarcinoma patients (p < 0.001, Fig. 3D). The result was validated via lung cancer cell line study. CL1-0 cells exhibited stronger expression of TTF-1 at both the RNA and protein levels compared with CL1-5 cells (Fig. 4A). CL1-5 cells exhibited more aggressive behavior. Levels of E-cadherin, vimentin as shown in Western blot analysis confirmed the tumor aggressiveness of CL1-5 cells (Fig. 4B). The less aggressive cell lines CL1-0 had higher TTF-1 expression and lower high-mobility group AT-hook 2 expression. It revealed reciprocal change in epidermal growth factor receptor (EGFR) expression (Fig. 4C). The expression of HMGA2, mesenchymal markers and EGFR increased in CL1-0 and HCC827 cells but expression of epithelial markers decreased after knocked down TTF-1 by shRNA (Fig. 5A). The expression of HMGA2, mesenchymal markers and EGFR decreased but expression of epithelial markers increased in CL1-5 and HCC827-GR cells following overexpression of TTF-1 by human TTF-1 cDNA ORF clone (Fig. 5B). The invasion ability of CL1-0 cells increased when knockdown TTF-1 expression with shTTF-1. In contrast, the invasion ability of CL1-5 cells decreased when overexpression of TTF-1( Fig. 6A and B). The invasion abilities were significantly inversely correlated with the expression of TTF-1. The precipitated chromatin was analyzed by PCR to investigate the interaction between TTF-1 and HMGA2 promoter. TTF-1 co-immunoprecipitated with HMGA2 promoter containing TTF1-recognition site and the effect was blocked after knockdown TTF-1 by TTF-1 shRNA (Fig. 7A). The expression of miR-33a was positively correlated with TTF-1 expression and this was validated by knockdown and overexpression of TTF-1 (Fig. 7B).

Discussion
Lung cancer is the leading cause of cancer deaths worldwide. Recurrent rate after complete resection for patients with stage I disease was ranged from 25% to 50% 18 . Some studies have reported that poor tumor differentiation is an independent prognostic factor in patients with stage I NSCLC 3,4 . However, other studies have shown that differentiation grade is not a prognostic factor for lung cancer survival 12,13 . The role of tumor histological grade as a prognostic factor is controversial.  The molecular mechanisms underlying tumor cell differentiation and the biological effects of prognosis are not clear. In previous study, poor differentiation of tumors was an independent prognostic factor for patients with pathological stage I NSCLC after surgical resection 19 . The grade of tumor differentiation as a prognostic factor in pulmonary adenocarcinoma has not been well clarified. In this study, we found that tumor differentiation was a prognostic factor for adenocarcinoma patients who underwent surgical resection. Poor tumor differentiation was an independent prognostic factor with (OR = 2.937). The result of metabolic PET scan supported this finding. Poorly differentiated tumors had a higher SUVmax of glucose uptake than moderately and well-differentiated tumors. This result is compatible with the metabolic concept of tumor biology.
TTF-1 plays key roles in the control of tissue-specific gene expression and in lung morphogenesis and differentiation 5 . Haploinsufficiency of TTF-1 is associated with tumor formation in transgenic mice 20 . TTF-1 expression is higher in lung adenocarcinoma than in normal lung tissue but decrease in metastatic lesions 15 . TTF-1 is enigmatic with dual roles in cancer cell survival and progression 14 . The mechanisms by which TTF-1 and its transcriptional activity are related to tumor aggressiveness are not fully understood. In clinical observations, the TTF-1 expression presented reciprocal correlation with the grade of tumor differentiation (Fig. 3). After scoring the TTF-1 expression for 520 resected lung adenocarcinoma patients, the expression of TTF-1 was correlated with tumor differentiation (p < 0.001). In this study, tumor differentiation was associated with TTF-1 expression and resulted in different surgical outcomes. The conclusion based on the clinical data was confirmed through in vitro studies of lung cancer cell lines. The expression of TTF-1 at both the RNA and protein level were high in CL1-0 cells and low in CL1-5 cells. CL1-5 cells displayed more aggressive behavior than CL1-0 cells. Western blot analysis of E-cadherin, vimentin, and HMGA2 expression confirmed the tumor aggressiveness of CL1-5 cells. The less aggressive cell lines (CL1-0) had higher TTF-1 expression and lower high-mobility group AT-hook 2 expression.
The highly invasive tumor cell line (CL1-5) had strong HMGA2 expression. HMGA2 is a member of the high-mobility group protein family. It plays a critical role in the invasion and metastasis of human cancers 21 and is involved in the epithelial-mesenchymal transition (EMT) process 22 . The functional studies via knockdown and overexpression of TTF-1 resulted in the respondent change of EMT marker, HMGA2 expression and the invasion ability. How does TTF-1 expression involve in the regulation of HMGA2 expression? Suppression of lung adenocarcinoma progression by TTF-1 via regulation of HMGA2 has been reported 15 . However, the mechanism by which TTF-1 regulates HGMA2 is not well known. In this study, TTF-1 co-immunoprecipitated with HMGA2 promoter containing TTF1-recognition site and the effect was blocked after knockdown TTF-1 by TTF-1 shRNA. The possible mechanism of TTF-1 involved in the regulation of HMGA2 expression via direct binding to promotor region. In addition, the expression of miR-33a was positively correlated with TTF-1 expression and this was validated by knockdown/overexpression of TTF-1. Micro RNAs are small non-protein-coding RNA that regulate gene expression by binding to the target mRNA. The miR 365 regulates lung cancer and development gene TTF-1 was reported 23 . The miR33a expression decreased after the suppressive effect of TTF-1. Our data suggest that TTF-1 is an important regulator of miR 33a. The TTF-1, miR33a and HMGA2 may have cross talk in regulation of the lung adenocarcinoma. The further study to investigate the miR33a will be worked.
In this study, TTF-1 expression was also associated with EGFR expression in lung adenocarcinoma cell lines. TTF-1-induced receptor tyrosine kinase-like orphan receptor 1 is required to sustain EGFR survival signaling in lung adenocarcinoma 24 . Transforming growth factor β1 (TGF-β1) induces stemness and plasticity through the EGFR/Ras signaling pathway and specifically activates HMGA2 25 . The TGF-β1 induction of HMGA2 is abrogated by EGFR inhibitors 26 . A blockade of TGF signaling induces a subset of genes that regulate hormone-induced type II cell differentiation, including TTF-1 27 . Tumor-associated macrophages also promote the EMT process in tumor cells by producing TGF-β. An analysis of 491 patients with NSCLC revealed positive correlations between intratumoral macrophage density, EMT markers, intraepithelial TGF-β levels, and tumor grade 28 . We speculated that tumor-associated macrophages may regulate TTF-1 expression by producing TGF-β and that the TTF-1/ HMGA2/EFGR pathway may play a role in the carcinogenesis of lung adenocarcinoma. Further work is needed to clarify this hypothesis.
The limitations of this study are that it was a single-institution retrospective study. The study focused on the lung adenocarcinoma. The subtype of lung adenocarcinoma did not dissect clearly. The mucinous type adenocarcinoma had different TTF-1 expression. Further study should be done to clarify the different mechanism of different subtype adenocarcinoma in this issue. In addition, the protein and RNA expression of clinical patient's specimen did not demonstrated. Further study should be conducted in clarify this issue.

Conclusion
The grade of tumor differentiation is a crucial prognostic factor in lung adenocarcinoma. TTF-1/HMGA2 asix is associated with tumor differentiation and aggressiveness. TTF-1-mediated tumor differentiation is involved in the tumorigenesis of pulmonary adenocarcinoma.