The MZF1/c-MYC axis mediates lung adenocarcinoma progression caused by wild-type lkb1 loss

L-H Tsai1,6, J-Y Wu2,6, Y-W Cheng3, C-Y Chen4, G-T Sheu1, T-C Wu5 and H Lee3

Abstract

Liver kinase B1 (LKB1) loss in lung adenocarcinoma is commonly caused by genetic mutations, but these mutations rarely occur in Asian patients. We recently reported wild-type LKB1 loss via the alteration of NKX2-1/p53-axis-promoted tumor aggressiveness and predicted poor outcomes in cases of lung adenocarcinoma. The mechanistic action of wild-type LKB1 loss within tumor progression remains unknown. The suppression of MYC by LKB1 controls epithelial organization; therefore, we hypothesize that MYC expression can be increased via wild-type LKB1 loss and promotes tumor progression. Here, MYC transcription is upregulated by LKB1-loss-mediated MZF1 expression. The wild-type LKB1-loss-mediated MZF1/MYC axis is responsible for soft-agar growth, migration and invasion in lung adenocarcinoma cells. Moreover, wild-type LKB1 loss-induced cell invasiveness was markedly suppressed by MYC inhibitors (10058-F4 and JQ1). Patients with low-LKB1/high-MZF1 or low-LKB1/high-MYC tumors have shorter overall survival and relapse-free-survival periods than patients with high-LKB1/low-MZF1 or high-LKB1/low-MYC tumors. In summary, MZF1-mediated MYC expression may promote tumor progression, resulting in poor outcomes in cases of lung adenocarcinoma with low-wild-type-LKB1 tumors.

INTRODUCTION

Liver kinase B1 (LKB1) inactivation via loss of heterozygosity (LOH) and genetic mutation frequently occurs in lung adenocarcinoma.1 However, East Asian populations have a low percentage of LKB1 mutations, including Taiwanese (3–7%).2–6 We recently reported that wild-type LKB1 can be inactivated through the aberration of the NKX2-1-mediated p53 pathway, leading to poor outcomes in lung adenocarcinoma patients.6 Mutant LKB1 has been shown to promote tumor growth via enhanced cyclin D1 expression.7
Tumor progression induced by LKB1 loss because of LOH and genetic mutations is mediated through increased NEDD9 and lysyl oxidase expression.8,9 However, the underlying mechanism of wild-type LKB1 loss in tumor progression in cases of lung adenocarcinoma remains unclear.
The loss of LKB1 alone did not promote tumor formation in transgenic mice.10 Tumor formation efficacy, however, was markedly elevated in LKB1-knockout mice with the addition of a KRAS mutation or p53 deletion.10 Klefstrom’s group reported that LKB1 suppressed the oncogenic properties of c-MYC (MYC) to control epithelial organization.11 The epithelial integrity defects caused by LKB1 loss may promote MYC-mediated tumor progression.12 Nonetheless, the link between LKB1 and MYC during lung tumorigenesis is not fully understood. We, therefore, hypothesized that MYC could be increased via wild-type LKB1 loss to promote tumor malignancy. This hypothesis was generated based on our preliminary data and showed that MYC expression was markedly elevated by wild-type LKB1 knockdown in lung cancer cells. Myeloid zinc-finger 1 (MZF1) was concomitantly increased in these wild-type LKB1-knockdown cells. MZF1 is a transcription factor of the Krupple family of zinc-finger proteins, which were originally cloned from the cDNA library of a patient with chronic myeloid leukemia.13,14 We expected that MYC could be elevated by wild-type LKB1 loss via increased MZF1 expression. Here, cell modeling was performed to explore whether MZF1 and MYC could be concomitantly increased by wild-type LKB1 loss, consequently promoting soft-agar growth, migration and invasion. To understand whether MZF1-mediated MYC expression due to wild-type LKB1 loss could be associated with a given outcome among lung adenocarcinoma patients, the prognostic value of MZF1 and MYC for overall survival (OS) and relapse-free survival (RFS) in low-LKB1/high-MZF1 tumors and low-LKB1/ high-MYC tumors was assessed using Cox regression analysis.

RESULTS

Wild-type LKB1 loss may contribute to the upregulation of MYC expression

We explored the possibility that MYC expression could be elevated by wild-type LKB1 loss. To test this hypothesis, LKB1 and MYC expressions in nine LKB1-wild-type cells and three LKB1-mutated lung adenocarcinoma cells were evaluated by western blotting. Low-LKB1 expression exhibited high-MYC expression in LKB1mutated A549 and H23 cells, but not in H1355. Interestingly, the reverse relationship of LKB1 with MYC expression was also observed in LKB1-wild-type cells, but not in CL3 and H1650 cells (Figure 1a, upper panel). Moreover, MYC mRNA expression, which was evaluated by real-time RT–PCR, was correlated with its protein expression in LKB1-wild-type cells, but not in LKB1-mutated A549 cells (Figure 1a, lower panel). These results suggested that MYC expressions in LKB1-wild-type cells might be elevated by LKB1 loss at a transcriptional level. We further used a cDNA plasmid and small hairpin (sh)RNA to overexpress and knockdown wild-type LKB1 in the CL1-5 and CL1-0 cells to examine whether MYC expression could be attenuated by LKB1 overexpression and knockdown. As expected, LKB1 expression was elevated in the LKB1-overexpressing CL1-5 cells and reduced in the LKB1knockdown CL1-0 cells (Figure 1b). Interestingly, MYC expression decreased markedly in the LKB1-overexpressing CL1-5 cells as compared with CL1-5 cells with an empty vector transfection (VC). The MYC protein expression modulated by the LKB1 overexpression and knockdown were consistent with MYC mRNA expression levels (Figure 1b). These results suggest that the increase in MYC expressions via wild-type LKB1 loss might be regulated at a transcriptional level.

MYC expression is transcriptionally activated by LKB1-loss-mediated MZF1

ALGGEN Research Software analysis (http://alggen.lsi.upc.es/ cgi-bin/promo_v3/promo/promoinit.cgi?dirDB=TF_8.3) showed the eleven putative MZF1 binding sites on the MYC promoter (−1697/+1) (Figure 2a, upper panel).15,16 We therefore expected that MYC expression could be transcriptionally activated by wildtype LKB1 loss. Two MYC promoter fragments (−1697/+1 and − 200/+1) were constructed and introduced into the CL1-0 and CL1-5 cells for the luciferase reporter assay. The reporter activity of the − 200/+1 promoter had ~ 80% of the − 1697/+1 promoter activity (Figure 2a, lower panel), suggesting that the four MZF1 binding sites located at the − 200/+1 promoter could have contributed more greatly than other binding sites to the wild-type LKB1 loss-mediated MYC transcription. Quantitative chromatin immunoprecipitation assay (ChIP-qPCR) analysis further showed that the binding activity of MZF1 on the − 200/+1 promoter increased in the MZF1-overexpressing CL1-0 cells and decreased in the MZF1-knockdown CL1-5 cells (Figure 2b, middle panel). Western blotting showed that the expressions of MZF1 and MYC were elevated by MZF1 overexpression in the CL1-0 cells and reduced by MZF1 knockdown in the CL1-5 cells. Consistently, MYC mRNA and MYC promoter activities (−200/+1) were dosedependently increased in the MZF1-overexpressing CL1-0 cells and decreased in the MZF1-knockdown CL1-5 cells (Figure 2b, lower panel). We further mutated the MZF1 binding sites on the − 200/+1 promoter via site-directed mutagenesis. A decrease of − 200/+1 promoter activity in the CL1-0 and CL1-5 cells was dependent on the number of MZF1 binding site mutations (Figure 2c). We further used two paired cell lines (CL1-0/CL1-5, TL10/TL13) to evidence that MZF1-mediated MYC expression was caused by wild-type LKB1 loss at the transcriptional level (Figure 2d). These results clearly indicate that the upregulation of MYC expression via wild-type LKB1 loss occurs through increased MZF1 binding to the MYC promoter.

MZF1-mediated MYC expression is responsible for the migration, invasion and soft-agar growth of lung adenocarcinoma cells with wild-type LKB1 loss

We then investigated whether MZF1-mediated MYC expression could be responsible for the migration, invasion and soft-agar growth of wild-type LKB1-loss lung adenocarcinoma cells. To test this hypothesis, MZF1 was ectopically expressed in the CL1-0 cells, and MYC expression was further silenced in the MZF1overexpressing CL1-0 cells. MZF1 expression in the CL1-5 cells was directly silenced via shRNA. As expected, MZF1 and MYC expressions increased in the MZF1-overexpressing CL1-0 cells and decreased in the MZF1-knockdown CL1-5 cells. MYC-downstream genes CDK4 and p21 were increased and decreased, respectively, in the MZF1-overexpressing CL1-0 cells. However, the change in both molecules caused by MZF1 overexpression can be rescued via MYC knockdown in the CL1-0 cells (Figure 3a, left panel), and the opposite was seen in the MZF1-knockdown CL1-5 cells (Figure 3a, right panel). Wound healing, Boyden chamber and softagar growth assays showed that migration, invasion and soft-agar growth increased markedly in the MZF1-overexpressing CL1-0 cells, but that these capabilities were almost restored by MYC knockdown in the MZF1-overexpressing CL1-0 cells (Figure 3d, left panel). Conversely, migration, invasion and soft-agar growth were dose-dependently decreased by MZF1 knockdown in the CL1-5 cells (Figure 3b, right panel). The representative migration, invasion and soft-agar growth in the MZF1-overexpressing CL1-0 and MZF1-knockdown CL1-5 cells are shown in Figures 3c and d. Small molecular inhibitors of MYC have been shown to inhibit the binding of MYC/MAX dimmers to its transcriptional targets or directly suppress MYC transcription, 10058-F4 and JQ1.17–19 Both inhibitors were used to treat with LKB1-knockdown CL1-0 and CL1-5 parental cells. Western blotting revealed that MYC expression levels in both cell types were dose-dependently decreased by both MYC inhibitors (10058-F4 and JQ1) and MYC-downstream genes (CDK4 and p21) were concomitantly reduced and elevated by MYC inhibitors in both cell types (Figure 4a, upper panel). The invasion capability induced by wild-type LKB1-loss-mediated MYC expression was nearly rescued by both MYC inhibitors in LKB1knockdown CL1-0 cells. Consistently, the invasiveness of CL1-5 cells was almost suppressed by both MYC inhibitors. Expectedly, MYC mRNA expression levels were dose-dependently decreased by JQ1, but not by 10058-F4 (Figure 4a, middle panel). However, the decrease of wild-type LKB1 loss-mediated MYC protein expression by 10058-F4 were nearly restored by MG132 treatment in both cell types (Figure 4b). These results clearly indicate that wild-type LKB1-loss-mediated MYC expression is responsible for lung adenocarcinoma cells with aggressive phenotypes.

Wild-type LKB1 expression was negatively correlated with MZF1 and MYC expression in lung tumors from lung adenocarcinoma patients

One hundred twenty-eight lung adenocarcinoma patients were enrolled to examine whether MZF1 and MYC expressions could be associated with wild-type LKB1 expression in lung tumors. Immunohistochemical data showed that low-LKB1 tumors were more likely to have high levels of MZF1 mRNA and protein expression as compared with high-LKB1 tumors (65% vs 35%, P = 0.004 for MZF1 mRNA, 60% vs 40%, P = 0.036 for MZF1 protein; Table 1). Low-LKB1 expression in lung tumors more frequently occurred in cases of high-MYC mRNA and protein expressions. Conversely, high levels of LKB1 expression were more common in cases of low-MYC mRNA and protein expressions (63% vs 37%, P = 0.012 for MYC mRNA; 67% vs 33%, P = 0.001 for MYC protein; Table 1). A positive correlation between MZF1 protein and MYC mRNA and protein expressions was observed in the study population (Table 1). MZF1 and MYC mRNA expressions were also associated with the related protein expressions (Po0.001 for MZF1 and MYC; Table 1). The positive correlation between low-LKB1 and low-MZF1 and low-MYC expressions in lung tumors was consistent with the findings from the cell model.

High MZF1/high MYC may predict poor OS and RFS in lung adenocarcinoma patients with low-LKB1 tumors

As mentioned above, MZF1-mediated MYC expression is responsible for migration, invasion and soft-agar growth in wild-type LKB1-loss lung adenocarcinoma cells. We next examined whether low-LKB1 tumors with high-MZF1 and -MYC expressions could have shorter OS and RFS periods than high-LKB1 tumors and lowLKB1 tumors with high levels of MZF1 or MYC expression. A Cox regression analysis showed that LKB1, MZF1 and MYC expressions may independently predict OS and RFS in this study population (Table 2). Low-LKB1/high-MZF1 or low-LKB1/high-MYC tumors had high hazard ratios (HR) regarding OS and RFS as compared with low-LKB1, high-MZF1 and low-MYC tumors (Table 2). These results suggest that high-MZF1 expression combined with high-MYC expression may predict a high risk of tumor recurrence and poor survival in lung adenocarcinoma patients with low-LKB1 tumors.

DISCUSSION

The CRTC1–NEDD9 signaling axis has been shown to promote tumor progression and metastasis in cases of lung cancer with LKB1 mutation.8,20 Aberrant lysyl oxidase (LOX) expression has also been linked to lung tumor progression,21 and a higher level of LOX expression was seen in LKB1-mutated cells than in LKB1-wildtype cells.9 We expected that NEDD9 and LOX would contribute to tumor progression in the LKB1-mutated cells, but not in the wildtype-LKB1 lung adenocarcinoma cells. NEDD9, LOX, MZF1 and MYC were silenced in the LKB1-mutated A549 and H23 cells and in wild-type CL1-5 and TL10 cells to evaluate the change in invasiveness that would be caused in these cells. As shown in Supplementary Figure 2, the invasion capability decreased markedly in the LOX- and NEDD9-knockdown A549 and H23 cells, but this invasiveness change caused by the LOX and NEDD9 knockdown was not observed in the CL1-5 and TL10 cells. Conversely, invasiveness was significantly reduced by MZF1 and MYC knockdown in the CL1-5 and TL10 cells, but not in the A549 and H23 cells. These results clearly indicate that MZF1-mediated MYC expression is responsible for tumor progression in LKB1-wildtype cells, but not in LKB1-mutated lung adenocarcinoma cells.
AXL expression has been found to be upregulated by MZF1 and to promote tumor progression in lung adenocarcinoma cells.22 We examined whether MZF1-mediated AXL could have a greater effect on the tumor progression than MZF1-mediated MYC in lung adenocarcinoma cells. As expected, AXL and MYC expressions were concomitantly increased by MZF1 overexpression in CL1-0 and TL13 cells (Supplementary Figure 3). However, the invasiveness decrease caused by the MYC knockdown was more pronounced than that caused by the AXL knockdown in both MZF1-overexpressing cell types (Supplementary Figure 3). These results clearly indicate that MZF1-mediated MYC expression is responsible for the invasiveness of LKB1-loss lung adenocarcinoma cells. We also examined the possibility that the elevation of MYC levels via wild-type LKB1 loss could occur via the Wnt/ β-catenin pathway because the nuclear translocation of β-catenin could be activated by LKB1 loss.23,24 As shown in Supplementary Figure 4, MYC expression was not changed by mutant β-catenin transfection in LKB1-overexpressing CL1-5 cells or by β-catenin knockdown in the LKB1-silencing CL1-0 cells, as compared with the VC and nonspecific shRNA (NC) cells, respectively. In the present study, we provide evidence demonstrating that MZF1-mediated MYC expression may be responsible for wild-type LKB1-loss-induced migration, invasion and soft-agar growth capability in lung adenocarcinoma cells. Therefore, we suggest that the MZF1/MYC axis has a critical role in wild-type-LKB1-lossmediated tumor progression and metastasis in cases of lung adenocarcinoma via p53 alteration.
The role of MYC in tumor metastasis is controversial. For example, MYC suppresses tumor metastasis by direct transcriptional silencing of the αv- and β3-integrin subunits.25 Conversely, MYC-mediated microRNA-9 suppresses E-cadherin to promote tumor metastasis.26 Wolfer et al.27 reported that MYC oncoprotein coordinately regulates the expression of 13 ‘poor outcome’ cancer signatures. The inactivation of MYC in human breast cancer cells specifically inhibits distant metastasis in vivo and its invasive capability.28 More interestingly, MYC was recently identified as controlling the metastatic conversion of KRASG12D-expressing cells for the generation of self-renewing metastatic cancer cells.29 LKB1 mutations do not concomitantly occur in cases of KRAS-mutated lung adenocarcinoma.2 Recently, the alteration of the NKX2-1/p53 axis has been shown to be involved in wild-type LKB1 loss in cases of lung adenocarcinoma.6 Therefore, we expected that tumor metastasis due to the alteration of NKX2-1/p53 axis-mediated wild-type LKB1 loss might occur via the MZF1/MYC axis in lung adenocarcinoma cells. This novel tumorigenic pathway is distinct from KRAS-mediated lung adenocarcinoma progression due to LKB1 mutation and/or LOH.
Although the LKB1 mutation was common in lung adenocarcinoma cases, the mutation frequency of LKB1 was still lower than that of wild-type LKB1.2,4,5,10,30,31 Moreover, rare mutations of KRAS (0%, 0 of 128) and LKB1 (1%, 1 of 128) were observed in Taiwanese lung adenocarcinomas (n = 128). The patient with LKB1 mutation is a 53-year-old lifetime nonsmoking female with stage IIIa (T2N2M0). To the best our knowledge, this mutation pattern of LKB1 gene at exon 1 (Arg42Leu) in this patient was not yet reported. We recently reported that wild-type LKB1 loss is caused at a transcriptional level via the alteration of the NKX2-1/p53 pathway, thus promoting tumor malignancy.6 As shown in Figure 4, MYC expression and invasiveness were markedly reduced by the following MYC inhibitors: 10058-F4 and JQ1. Interestingly, Wang’s group reported that MYC expression’s suppression by MYC inhibitor JQ1 was only observed in LKB1wild-type H441 and H1734 cells, but not seen in LKB1-mutated A549 and H460 cells.32 In animal models, the tumor burden in KRAS-mutated transgenic mice was more greatly suppressed by JQ1 than in KRAS/LKB1-double-mutated transgenic mice.32 We examined the cytotoxic effects of two MYC inhibitors, 10058-F4 and JQ1 to A549, LKB1-knockdown CL1-0 and parental CL1-5 cells using an MTT (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay. As shown in Supplementary Figure 5A, the cytotoxic effects of 10058-F4 to LKB1-knockdown CL1-0 and parental CL1-5 cells with wild-type LKB1 loss was greater than A549 cells with mutant LKB1 (half-maximal inhibitory concentration: 176.6 μM, LKB1-knockdown CL1-0 and 176.8 μM, CL1-5 vs >400 μM, A549 cells). Consistent finding was observed in the cytotoxic effects of JQ1 to these cells (half-maximal inhibitory concentration: 0.6 μM, LKB1-knockdown CL1-0, 0.8 μM, parental CL1-5 and 1.5 μM, A549 cells; Supplementary Figure 5B). Therefore, MYC inhibitors have greater cytotoxic effects to lung adenocarcinoma cells with wild-type LKB1 loss than those with mutant LKB1. Moreover, the cytotoxic effect of JQ1 to lung adenocarcinoma cells with wild-type-LKB1 loss was greater than 10058-F4. These results suggest that JQ1 might have more potentially therapeutic significance in lung adenocarcinoma patients who had low wildtype LKB1 tumors. The findings from Wang’s group and our present study strongly suggest that MYC inhibitor is potentially useful in suppressing wild-type-LKB1-loss-mediated, MYC-induced tumor aggressiveness and consequently improve outcomes in lung adenocarcinoma patients with low-wild-type-LKB1 tumors.
In summary, we provide evidence to demonstrate that MZF1-mediated MYC expression due to wild-type-LKB1 loss is responsible for tumor progression in lung adenocarcinoma cells. A consistent correlation between wild-type-LKB1 loss with MZF1 and MYC expressions in lung tumors and poorer outcomes were observed in lung adenocarcinoma patients whose tumors had low-LKB1/high-MYC or low-LKB1/low-MZF1 expressions. Therefore, the MZF1/MYC axis has a novel role in LKB1 inactivation for lung adenocarcinoma-harbored wild-type LKB1, but not for mutated LKB1. We suggest that MYC might be targeted to improve outcomes via suppressing tumor progression in lung adenocarcinoma patients with wild-type-LKB1 loss.

MATERIALS AND METHODS

Tissue specimens

The lung tumor specimens used in this study were collected from 128 lung cancer patients at the Department of Thoracic Surgery in Taichung Veteran’s General Hospital (TVGH, Taichung, Taiwan) between 1993 and 2004. This study was approved by the hospital’s institutional review board (Institutional Review Board, Buddhist Tzu Chi General Hospital; No: IRB102-95). The inclusion criteria for patients included the following: primarily diagnosed with lung adenocarcinoma, no metastatic disease at diagnosis, no previous diagnosis of carcinoma, no neoadjuvant treatment before primary surgery and no evidence of disease within 1 month of primary surgery. Written informed consent was obtained from the patients before sample collection. The tumor stages were determined according to the World Health Organization (WHO) classifications. The lung tumor specimens were collected by surgical resection, and surgically resected tissues were stored at − 80 °C immediately after resection. The tumor node metastasis stage, tumor type and stage of each specimen collected were histologically determined according to the WHO classification system. The age of all patients was between 26 and 84 years (mean ± s.d. = 64.3 ± 10.4). The clinical parameters and OS data were collected from the chart review and the Taiwan Cancer Registry, Department of Health, Executive Yuan, Taiwan, ROC. The survival time was defined as the period of time from the date of primary surgery to the date of death. The median follow-up time after surgery was 21.5 months, and the median OS of all patients was 33.1 months. During this survey, 87 patients died. Based on the follow-up data, 49 patients relapsed (17 had local recurrences, 44 had distant metastases and 12 had local and distant metastases). Among these patients, tumors frequently relapsed in the bilateral lung (17 patients) and metastasized in the bones other than the spine (11 patients), brain (12 patients), liver (3 patients), pleura (3 patients), chest wall (1 patient), mediastinum (2 patients), adrenal (1 patient) and spine (1 patient). In total, six patients had tumors that metastasized to more than one organ. The association between the clinical parameters and LKB1, MZF1 and MYC expressions is shown in Supplementary Table 1.

Cell lines and reagents

A549 cells were maintained in Dulbecco’s modified Eagle’s medium. CL1-0, CL1-5, CL3, H1355, H1650, H1975, H23, H358, H441, TL10 and TL13 cells were maintained in RPMI-1640. The CL1-0, CL1-5 and CL3 cells were kindly supplied by Dr Pan-Chyr Yang (Department of Internal Medicine, National Taiwan University College of Medicine, Taipei, Taiwan). The A549, H1355, H1650, H1975, H23, H358 and H441 cells were obtained from American Type Culture Collection (ATCC, Manassas, VA, USA). The TL10 and TL13 cells were primarily established from the pleural effusions of Taiwanese lung cancer patients from the Chung Shan Medical University Hospital (Taichung, Taiwan). These cells were further identified to be adenocarcinoma cells using the procedures and methods described previously.33,34 TL10 and TL13 cells were confirmed to be lung adenocarcinoma cells via specific antibodies (TTF-1, CEA, DAPAS and mucin) that have been described in our previous reports.33,34 The medium contained 10% fetal bovine serum supplemented with penicillin (100 U/ml) and streptomycin (100 mg/ml). Cells were grown at 37 °C in a humidified incubator at 5% CO2. 10058-F4 was purchased from Sigma (St Louis, MO, USA). (+)-JQ1 was purchased from Cayman Chemical (Ann Arbor, MI, USA).

Immunohistochemistry and scoring

Immunohistochemistry was used to detect LKB1, MZF1 and MYC protein expressions. The LKB1 and MYC antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). The MZF1 antibodies were purchased from GeneTex (Hsinchu City, Taiwan). The immunohistochemical procedures were conducted as described previously.34 Negative controls were obtained by performing all of the immunohistochemistry steps, but leaving out the primary antibody. The immunohistochemical staining scores were defined as described previously,34 and the intensities of the signals were evaluated independently by three observers. Immunostaining scores were defined as the cell staining intensity (0 = nil, 1 = weak, 2 = moderate and 3 = strong) multiplied by the percentage of labeled cells (0–100%), leading to scores from 0 to 300. For LKB1, a score higher than 100 was defined as ‘high’ immunostaining, whereas a score equal to or lower than 100 was categorized as ‘low’. For MZF1 and MYC, a score higher than 150 was defined as ‘high’ immunostaining, whereas a score equal to or lower than 150 was categorized as ‘low’. The representation of immunostaining results of LKB1, MZF1 and MYC protein expressions were shown in Supplementary Figure 1.

RNA isolation and real-time quantitative RT–PCR analysis

The methods and procedures for total RNA extraction from tissues and cells were described previously.35 The primers used for real-time RT–PCR analysis were as follows: MZF1 (forward, 5′-CCCGGAGATGGGTCACAGT-3′; reverse: 5′-TGCATAGTCCTAGGAGGTGTCTTG-3′); MYC (forward: 5′-AGC GACTCTGAGGAGGAACAAG-3′; reverse: 5′-CCTGCCTCTTTTCCACAGAAA-3′). The MZF1 and MYC mRNA levels in lung tumors that were higher than the median value were defined as ‘high’, whereas levels lower than the median value were defined as ‘low’.

Western blotting assay and antibodies

The methods and procedures were described previously.35 LKB1 (Ley37G/ D6), MYC (9E10), p21 (F-5), CDK4, cyclin D1 (H-295), β-catenin (H-102) and NEDD9 antibodies were obtained from Santa Cruz Biotechnology. β-Actin (AC-15) antibody was obtained from Sigma. MZF1 (N2C1), AXL (C2C3) and LOX antibodies were obtained from GeneTex.

Plasmid construction and transfection reaction

The wild-type MZF1 plasmids were kindly provided by Dr Yi-Hsien Hsieh, from the Department of Biochemistry in the School of Medicine at Chung Shan Medical University. The LKB1 overexpression plasmid was purchased from Addgene Technologies (Cambridge, MA, USA). The shLKB1 (TRCN0000000411), shMZF1 (TRCN0000017137), shMYC (TRCN0000039642), shLOX (TRCN0000045991), shAXL (TRCN0000195353), shNEDD9 (TRCN0000004967), shβ-catenin (TRCN0000314991) and pLKO.1 vector plasmids were purchased from the National RNAi Core Facility at Academic Sinica, Taipei, Taiwan. The various concentrations of expression plasmids, as indicated, were transiently transfected into the lung cancer cells (1 × 106) using the Turbofect reagent (Fermentas, Hanover, MD, USA). After 48 h, the cells were harvested, and the whole-cell extracts were assayed in subsequent experiments.

Site-directed mutagenesis

The methods and procedures for the site-directed mutagenesis assay were described previously.36 Site-directed mutagenesis was performed to generate mutant MZF1 binding sites on the MYC promoter constructs using the complementary oligos as follows: − 93 to − 58 (forward: 5′-CAAAGCAGAGGGCGTGGtGGAAAAGAAAAAAGATCC-3′; reverse: 5′-GGA TCTTTTTTCTTTTCCaCCACGCCCTCTGCTTTG-3′); − 136 to − 105 (forward: 5′-CCCACCCTCCaCACCCTCCaCATAAGCGCC-3′; reverse: 5′-GGCGCTT ATGtGGAGGGTGtGGAGGGTGGG-3′). All plasmid clones were verified through DNA sequencing. (underline: gene binding sites, lower case letters: mutated bases).

Luciferase reporter assay

For the luciferase reporter assay, the appropriate number of cells were transfected with sufficient reporter plasmid, MZF1-Luc or its derivatives, MYC-Luc or its derivatives and either the control vector or the shLKB1 and LKB1 expression plasmid. To normalize the transfection efficiency, β-galactosidase was also co-transfected. Transfected cells were harvested 48 h after transfection, and a luciferase assay was performed according to the manufacturer’s instructions. The luciferase activity was measured with an AutoLumat LB953 luminometer (Berthold Technologies, Bad Wildbad, Germany) and normalized via co-transfected β-galactosidase activity.

ChIP-qPCR

ChIP analysis was performed as described in a previous report.35 ChIP-qPCR analysis was performed using an ABI-7500 with SYBR Green real-time PCR MasterMix (Fermentas), following the detailed procedures described previously.6,37,38 The primer sequences for the MZF1 binding sites on the MYC promoter were the forward primer (5′-GGCTGCCCGGCTGAGTCTCC-3′) and the reverse primer (5′-GGGGCGCTTATGGGGAGGGTGG-3′). All signals from the immunoprecipitated DNA samples were normalized to the input of the total cell DNA.

Statistical analysis

The statistical analysis was performed using the SPSS statistical software program (Version 15.0 SPSS Inc., Chicago, IL, USA). The associations between LKB1, MZF1, MYC and the clinical parameters were analyzed using a χ2-test. A multivariate Cox regression analysis was performed for both OS and RFS. The analysis was stratified for all known prognostic variables (age, gender, smoking status and stage), as well as LKB1, MZF1 and MYC protein expressions.

REFERENCES

1 Makowski L, Hayes DN. Role of LKB1 in lung cancer development. Br J Cancer 2008; 99: 683–688.
2 Koivunen JP, Kim J, Lee J, Rogers AM, Park JO, Zhao X et al. Mutations in the LKB1 tumour suppressor are frequently detected in tumours from Caucasian but not Asian lung cancer patients. Br J Cancer 2008; 99: 245–252.
3 Sun Y, Ren Y, Fang Z, Li C, Fang R, Gao B et al. Lung adenocarcinoma from East Asian never-smokers is a disease largely defined by targetable oncogenic mutant kinases. J Clin Oncol 2010; 28: 4616–4620.
4 Onozato R, Kosaka T, Achiwa H, Kuwano H, Takahashi T, Yatabe Yet al. LKB1 gene mutations in Japanese lung cancer patients. Cancer Sci 2007; 98: 1747–1751.
5 Kim MJ, Jin G, Jheon HS, Lee SY, Cha SI, Kim CH et al. LKB1 mutations are extremely rare in Korean non-small cell lung cancers. Cancer Genet Cytogenet 2010; 196: 204–206.
6 Tsai LH, Chen PM, Cheng YW, Chen CY, Sheu GT, Wu TC et al. LKB1 loss by alteration of the NKX2-1/p53 pathway promotes tumor malignancy and predicts poor survival and relapse in lung adenocarcinomas. Oncogene (e-pub ahead of print 2 September 2013; doi:10.1038/onc.2013.353).
7 Scott KD, Nath-Sain S, Agnew MD, Marignani PA. LKB1 catalytically deficient mutants enhance cyclin D1 expression. Cancer Res 2007; 67: 5622–5627.
8 Feng Y, Wang Y, Wang Z, Fang Z, Li F, Gao Y et al. The CRTC1-NEDD9 signaling axis mediates lung cancer progression caused by LKB1 loss. Cancer Res 2012; 72: 6502–6511.
9 Gao Y, Xiao Q, Ma H, Li L, Liu J, Feng Y et al. LKB1 inhibits lung cancer progression through lysyl oxidase and extracellular matrix remodeling. Proc Natl Acad Sci USA 2010; 107: 18892–18897.
10 Ji H, Ramsey MR, Hayes DN, Fan C, McNamara K, Kozlowski P et al. LKB1 modulates lung cancer differentiation and metastasis. Nature 2007; 448: 807–810.
11 Partanen JI, Nieminen AI, Makela TP, Klefstrom J. Suppression of oncogenic properties of c-Myc by LKB1-controlled epithelial organization. Proc Natl Acad Sci USA 2007; 104: 14694–14699.
12 Partanen JI, Tervonen TA, Myllynen M, Lind E, Imai M, Katajisto P et al. Tumor suppressor function of Liver kinase B1 (Lkb1) is linked to regulation of epithelial integrity. Proc Natl Acad Sci USA 2012; 109: E388–E397.
13 Hromas R, Morris J, Cornetta K, Berebitsky D, Davidson A, Sha M et al. Aberrant expression of the myeloid zinc finger gene, MZF-1, is oncogenic. Cancer Res 1995; 55: 3610–3614.
14 Hromas R, Collins SJ, Hickstein D, Raskind W, Deaven LL, O’Hara P et al. A retinoic acid-responsive human zinc finger gene, MZF-1, preferentially expressed in myeloid cells. J Biol Chem 1991; 266: 14183–14187.
15 Messeguer X, Escudero R, Farre D, Nunez O, Martinez J, Alba MM. PROMO: detection of known transcription regulatory elements using species-tailored searches. Bioinformatics 2002; 18: 333–334.
16 Farre D, Roset R, Huerta M, Adsuara JE, Rosello L, Alba MM et al. Identification of patterns in biological sequences at the ALGGEN server: PROMO and MALGEN. Nucleic Acids Res 2003; 31: 3651–3653.
17 Delmore JE, Issa GC, Lemieux ME, Rahl PB, Shi J, Jacobs HM et al. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 2011; 146:904–917.
18 Huang MJ, Cheng YC, Liu CR, Lin S, Liu HE. A small-molecule c-Myc inhibitor, 10058-F4, induces cell-cycle arrest, apoptosis, and myeloid differentiation of human acute myeloid leukemia. Exp Hematol 2006; 34: 1480–1489.
19 Wang H, Hammoudeh DI, Follis AV, Reese BE, Lazo JS, Metallo SJ et al. Improved low molecular weight Myc-Max inhibitors. Mol Cancer Ther 2007; 6: 2399–2408.
20 Gu Y, Lin S, Li JL, Nakagawa H, Chen Z, Jin B et al. Altered LKB1/CREB-regulated transcription co-activator (CRTC) signaling axis promotes esophageal cancer cell migration and invasion. Oncogene 2012; 31: 469–479.
21 Payne SL, Hendrix MJ, Kirschmann DA. Paradoxical roles for lysyl oxidases in cancer—a prospect. J Cell Biochem 2007; 101: 1338–1354.
22 Mudduluru G, Vajkoczy P, Allgayer H. Myeloid zinc finger 1 induces migration, invasion, and in vivo metastasis through Axl gene expression in solid cancer. Mol Cancer Res 2010; 8: 159–169.
23 Miyaki M, Iijima T, Hosono K, Ishii R, Yasuno M, Mori T et al. Somatic mutations of LKB1 and beta-catenin genes in gastrointestinal polyps from patients with Peutz–Jeghers syndrome. Cancer Res 2000; 60: 6311–6313.
24 Ossipova O, Bardeesy N, DePinho RA, Green JB. LKB1 (XEEK1) regulates Wnt signalling in vertebrate development. Nat Cell Biol 2003; 5: 889–894.
25 Liu H, Radisky DC, Yang D, Xu R, Radisky ES, Bissell MJ et al. MYC suppresses cancer metastasis by direct transcriptional silencing of alphav and beta3 integrin subunits. Nat Cell Biol 2012; 14: 567–574.
26 Ma L, Young J, Prabhala H, Pan E, Mestdagh P, Muth D et al. miR-9, a MYC/MYCNactivated microRNA, regulates E-cadherin and cancer metastasis. Nat Cell Biol 2010; 12: 247–256.
27 Wolfer A, Wittner BS, Irimia D, Flavin RJ, Lupien M, Gunawardane RN et al. MYC regulation of a ‘poor-prognosis’ metastatic cancer cell state. Proc Natl Acad Sci USA 2010; 107: 3698–3703.
28 Cho KB, Cho MK, Lee WY, Kang KW. Overexpression of c-myc induces epithelial mesenchymal transition in mammary epithelial cells. Cancer Lett 2010; 293: 230–239.
29 Ischenko I, Zhi J, Moll UM, Nemajerova A, Petrenko O. Direct reprogramming by oncogenic Ras and Myc. Proc Natl Acad Sci USA 2013; 110: 3937–3942.
30 Fernandez P, Carretero J, Medina PP, Jimenez AI, Rodriguez-Perales S, Paz MF et al. Distinctive gene expression of human lung adenocarcinomas carrying LKB1 mutations. Oncogene 2004; 23: 5084–5091.
31 Sanchez-Cespedes M, Parrella P, Esteller M, Nomoto S, Trink B, Engles JM et al. Inactivation of LKB1/STK11 is a common event in adenocarcinomas of the lung. Cancer Res 2002; 62: 3659–3662.
32 Shimamura T, Chen Z, Soucheray M, Carretero J, Kikuchi E, Tchaicha JH et al. Efficacy of BET bromodomain inhibition in Kras-mutant non-small cell lung cancer. Clin Cancer Res 2013; 19: 6183–6192.
33 Cheng YW, Wu MF, Wang J, Yeh KT, Goan YG, Chiou HL et al. Human papillomavirus 16/18 E6 oncoprotein is expressed in lung cancer and related with p53 inactivation. Cancer Res 2007; 67: 10686–10693.
34 Sung WW, Wang YC, Cheng YW, Lee MC, Yeh KT, Wang L et al. A polymorphic -844T/C in FasL promoter predicts survival and relapse in non-small cell lung cancer. Clin Cancer Res 2011; 17: 5991–5999.
35 Wu DW, Liu WS, Wang J, Chen CY, Cheng YW, Lee H. Reduced p21(WAF1/CIP1) via alteration of p53-DDX3 pathway is associated with poor relapse-free survival in early-stage human papillomavirus-associated lung cancer. Clin Cancer Res 2011; 17: 1895–1905.
36 Wu HH, Cheng YW, Chang JT, Wu TC, Liu WS, Chen CY et al. Subcellular localization of apurinic endonuclease 1 promotes lung tumor aggressiveness via NF-kappaB activation. Oncogene 2010; 29: 4330–4340.
37 Steger DJ, Lefterova MI, Ying L, Stonestrom AJ, Schupp M, Zhuo D et al. DOT1L/KMT4 recruitment and H3K79 methylation are ubiquitously coupled with gene transcription in mammalian cells. Mol Cell Biol 2008; 28: 2825–2839.
38 Gong F, Sun L, Wang Z, Shi J, Li W, Wang S et al. The BCL2 gene is regulated by a special AT-rich sequence binding protein 1-mediated long range chromosomal interaction between the promoter and the distal element located within the 3′-UTR. Nucleic Acids Res 2011; 39: 4640–4652.
39 Wu DW, Cheng YW, Wang J, Chen CY, Lee H. Paxillin predicts survival and relapse in non-small cell lung cancer by microRNA-218 targeting. Cancer Res 2010; 70:10392–10401.