Targeting the miR-200c/LIN28B axis in acquired EGFR-TKI resistance non-small cell lung cancer cells harboring EMT features

MicroRNA (miR)-200 family members (miR-200s) are frequently silenced in advanced cancer and have been implicated in the process of epithelial-to-mesenchymal transition (EMT). We previously reported that miR-200s were silenced through promoter methylation in acquired EGFR-tyrosine kinase inhibitor (TKI) resistant non-small cell lung cancer (NSCLC) cells harboring EMT features. In this study, we examined the functional role of miR-200s in NSCLC cells and investigated a novel approach to overcoming acquired EGFR-TKI resistance. In the analysis of NSCLC cell lines, each of the miR-200s expression-silenced cell lines showed promoter methylation. Significant correlations between miR-200c silencing and several oncogenic pathway alterations, including EMT-changes and LIN28B overexpression, were observed in the database analysis. In addition, EGFR-wild type cell lines had lower miR-200s expression levels than EGFR-mutant cell lines. The introduction of miR-200c using pre-miR-200c caused LIN28B suppression in cells with acquired EGFR-TKI resistance that harbored EMT features. Interestingly, both the introduction of miR-200c and the knockdown of LIN28B produced an antitumor effect in acquired EGFR-TKI resistance cells, whereas these manipulations were not effective in parental cells. The miR-200c/LIN28B axis plays an important role in cells with acquired resistance to EGFR-TKI that harbor EMT features and might be a useful therapeutic target for overcoming resistance.

members in acquired EGFR-TKI resistance cells with EMT features. However, the detailed mechanisms underlying the EMT-related acquired resistance to EGFR-TKIs remains unclear.
MicroRNAs are small noncoding, endogenous, single-stranded RNAs that are commonly deregulated in human cancers. Several miRs show abnormal expression patterns in cancers with consequent alteration of the target oncogenes or tumor suppressor genes 16 . MicroRNA profiling for NSCLC has been previously conducted by several groups to predict patient survival [17][18][19] . These profiles have also been correlated with clinicopathological parameters of NSCLC patients 20 24 . They are frequently silenced in advanced cancer and have been implicated in EMT and tumor invasion through the targeting of transcriptional repressors of E-cadherin, ZEB1 and ZEB2 25 . ZEB1 is also known to repress miR-200s transcription in a negative feedback loop. MiR-200c silencing was also reported in cancer stem cells (CSCs), suggesting that miR-200c deregulation is a key event in multiple levels of tumor biology 26 .
Several of the miRs identified from these studies have been associated with key regulatory pathways including EGFR and KRAS in NSCLC. In this study, we examined the functional role of miR-200s in NSCLC cell lines, and investigated the mechanisms underlying the EMT and acquired EGFR-TKI resistance to develop a novel approach for overcoming acquired resistance to EGFR-TKIs in NSCLC.

Functional analysis of miR-200s in NSCLC cells.
To define the functional significance of genes regulated by the miR-200c status, we performed a gene set enrichment analysis (GSEA) using the Cancer Cell Line Encyclopedia (CCLE) database. Among 34 NSCLC cell lines, comprehensive gene expression data for 28 cell lines were available in the CCLE database, and these cell lines were categorized into an miR-200c-low group  Table S1). We found that several pathways, including genes defining an EMT (HALLMARK_ EPITHELIAL_MESENCHYMAL_TRANSITION), and a subgroup of genes regulated by Myc (HALLMARK_ MYC_TARGETS_V1), were positively enriched in the miR-200c-low group (Table 1). We then comprehensively analyzed the correlation between miR-200c expression and the gene expression status (Supplementary Table S2). The epithelial marker CDH1 was positively correlated with miR-200c expressions (Pearson r = 0.70, P < 0.0001) ( Fig. 2A), and the mesenchymal marker ZEB1 was negatively correlated (Pearson r = − 0.66, P < 0.0001) (Fig. 2B). Interestingly, LIN28B, an oncogenic stem-cell factor 27 , showed a negative correlation with miR-200c expression  to the same extent as the correlation between CDH1 or ZEB1 and miR-200c (Pearson r = − 0.73, p < 0.0001) (Fig. 2C). In western blots of 34 NSCLC cell lines and HBEC-5KT, miR-200c-silenced NSCLC cells exhibited a low E-cadherin expression level and high vimentin or ZEB1 expression levels ( Supplementary Fig. S2). We also investigated the correlation between miR-200s and EMT markers in breast cancer, colon cancer and gastric cancer cell lines. As shown in Supplementary Fig. S3, in all three carcinomas, the expression of miR-200s was low in the cell lines harboring EMT features. These findings are presumed to supplement the relevance of this miR-200c expression based classification and importance of LIN28B.

MiR-200c introduction showed an antitumor effect in cells with acquired EGFR-TKI resistance.
As we previously reported, EMT features are correlated with the acquisition of EGFR-TKI resistance, but the mechanisms responsible for EGFR-TKI resistance remain unclear. We utilized experimentally established cells with acquired EGFR-TKI resistance and EMT features. By continuously exposing the cells to gefitinib, we established the EGFR-TKI-resistant clones HCC4006-GR-step and HCC4006-GR-high from the EGFR-mut HCC4006 cell line 15 . The established cell line showed EMT features, while lacking both T790M secondary EGFR-mut and MET-amplification. The IC values against gefitinib determined using an MTS assay was 0.024 μ mol/L for HCC4006 and over 10 μ mol/L for HCC4006-GR cells. MiR-200s expression was suppressed in HCC4006-GR cells compared with the parental HCC4006 (Fig. 3A). In western blots, the HCC4006-GR cells showed EMT features with the downregulation of E-cadherin and the upregulation of N-cadherin, vimentin, and ZEB-1 compared with HCC4006 (Fig. 3B). We examined the effect of miR-200s introduction against cell viability in HCC4006 and HCC4006-GR-high. Although the cell viability of HCC4006 was not affected by miR-200c introduction, the cell viability of HCC4006-GR cells was highly suppressed by miR-200c introduction ( Fig. 3C and Supplementary Fig. S3).

MiR-200c introduction suppressed LIN28B expression and produced an antitumor effect in EGFR-TKI resistant cells.
From the database analysis, miR-200c silencing was found to be correlated with the upregulation of LIN28B. Indeed, LIN28B was upregulated in HCC4006-GR cells, compared with that in HCC4006, as determined using qRT-PCR (Fig. 4A). When we introduced miR-200c into HCC4006 and HCC4006-GR cells, the expressions of not only mesenchymal protein ZEB1, but also LIN28B were significantly suppressed. On the other hand, expression of E-cadherin, which is known as an epithelial marker, was induced in HCC4006-GR cells (Fig. 4B).

Knockdown of LIN28B suppressed cell viability in cells with acquired EGFR-TKI resistance.
To examine the role of LIN28B on cell survival in cells with acquired EGFR-TKI resistance and EMT features, we suppressed the expression of LIN28B using two kinds of si-RNA. Interestingly, LIN28B suppression using siRNA produced an antitumor effect in HCC4006-GR cells, whereas it was not effective in HCC4006 cells when evaluated using the tetrazolium salt, 3-4,5 dimethylthiazol-2,5 diphenyl tetrazolium bromide (MTT) (Fig. 5A). In a western blot analysis, the knockdown of LIN28B led to the expression of the apoptosis marker c-PARP in HCC4006-GR cells (Fig. 5B). Regarding EMT features, In HCC4006-GR cells, even though the difference is not so clear, the knockdown of LIN28B seems to induce the E-cadherin expression and to reduce the vimentin and ZEB1 expression. This result suggests the possibility that LIN28B is one of the mediator of epithelial to mesenchymal transition in TKI resistance cell line. In addition, we examined phosphorylation of STAT3 as an indicator of TGF-beta/IL6 pathway. As a result, the knockdown of LIN28B doesn't seem to affect the phosphorylation of STAT3, suggesting that LIN28B doesn't regulate TGF-beta/IL6 pathway.

MiR-200c downregulation and LIN28B upregulation after EGFR-TKI resistance acquisition on NSCLC tissues.
We randomly selected the 3 non-small cell lung cancer patients who were treated at our hospital. All 3 patients underwent pulmonary resection as an initial treatment. After relapse, they were treated with EGFR-TKI and acquired resistance finally. We investigated the alteration of EMT marker and the expression of LIN28B and miR-200c before and after EGFR-TKI treatment. The alteration of e-cadherin, vimentin and LIN28B was analyzed by immunohistochemistry and that of miR-200c was analyzed by qRT-PCR. As a result, we confirmed that the expression of miR-200c was significantly suppressed, compared to its expression before resistance acquisition in all three patients (Fig. 6A). On the other hand, LIN28B was highly expressed, compared to before resistance acquisition (Fig. 6B). And then, the expression of E-cadherin was reduced and that of vimentin was increased, indicating that tumors acquired the EMT feature in the process of EGFR-TKI resistance acquisition (Fig. 6B). These results were consistent with those of HCC4006 parental and GR cells.

Discussion
In this study, we found that the miR-200c/LIN28B axis plays a critical role in the cell viability of acquired EGFR-TKI resistance cells harboring EMT features. Acquired resistance cells showed miR-200s silencing and LIN28B overexpression, which was the same for clinical samples. MiR-200c introduction caused LIN28B suppression, and both miR-200c introduction and LIN28B suppression had an anti-proliferative effect in HCC4006-GR cells. These effects were not observed in parental HCC4006 cells lacking miR-200s silencing and LIN28B overexpression, suggesting that the survival of acquired resistance cells depends on the miR-200c/ LIN28B axis (Fig. 5C). It is still inconclusive that miR-200c directly or indirectly regulates LIN28B. Therefore, further investigation is mandatory. We found that miR-200c expression was not silenced in most of the EGFR-mut NSCLC cell lines. Several reports have shown that EGFR-mut is associated with epithelial characteristics in NSCLC 28,29 . The detailed mechanisms of how EGFR-mut leads to the maintenance of epithelial characteristics have been unclear. Takeyama et al. hypothesized that EGFR-mut cells are resistant to EMT-inducing signals, or that EGFR-mut cells specifically express genes that retain cells in the epithelial state 29 . On the other hand, several reports have shown that the chronic activation of EGFR can promote an EMT-like change 30,31 . Further investigation is required regarding EGFR-mut and epithelial characteristics, including E-cadherin expression, because E-cadherin expression has been proposed as a biomarker predicting the clinical activity of EGFR-TKIs in NSCLC 12 .
We and other groups have reported that EGFR-TKIs cause an EMT change with miR-200s silencing in a subset of NSCLC harboring EGFR-mut 14 . EMT is a phenomenon in which cells with epithelial phenotypes acquire mesenchymal characteristics, and EMT plays an important role in cancer metastasis and drug resistance. EMT has been shown to be correlated with a poor prognosis in multiple epithelial-derived solid tumors. In preclinical models and clinical samples, EMT features were observed after the acquisition of resistance to EGFR-TKIs. Although the detailed mechanisms explaining how these alterations cause EGFR-TKIs resistance have been unclear, we have identified LIN28B as a candidate oncogenic driver molecule in acquired EGFR-TKI resistance cells.
LIN28B and its paralog LIN28A are RNA-binding proteins that mediate diverse biological functions. The LIN28 family regulates mammalian stem cell self-renewal, and LIN28A in combination with NANOG, OCT4, and SOX2, can reprogram human somatic cells to become pluripotent stem cells 32 . LIN28A participates in these key biological processes by blocking the maturation of the tumor suppressor miR let-7 family 33,34 . Furthermore, LIN28A plays a functional role in the maintenance of the ALDH1-positive CSC population in tumors 35 . LIN28B also influences multiple oncogenic signaling networks in diverse cellular contexts, and is being recognized as oncogenic stem-cell factor 27,36 . Indeed, CD166-positive tumor-initiating cells obtained from primary NSCLC tumor expressed high levels of LIN28B 27 . In addition, LIN28B has been reported as a candidate biomarkers of resistance to platinum-based chemotherapy and of a poor clinical outcome in patients with epithelial ovarian carcinoma 37 , suggesting the function of LIN28B as a drug-resistant factor.
LIN28B expression is reportedly induced by the c-Myc oncogenic transcription factor in multiple human and mouse tumor models 38 . The upregulation of LIN28 family and Myc or the downregulation of let-7 may promote the conversion of epithelial cells to a more undifferentiated stage and maintain the tumor cells in this stem-like stage 35,38 . In our analysis, genes regulated by Myc were positively enriched in the miR-200c-low group when examined using GESA (Table 1). These findings are compatible with the result that miR-200c-silenced cells had a high level of LIN28B expression as well as EMT features.
MiR-200c-introduced cells showed the downregulation of LIN28B in NSCLC cells, and miR-200c introduction produced an anti-proliferative effect in HCC4006-GR cells. In addition, both miR-200c introduction and LIN28B downregulation inhibit the expression of EMT markers. In prostate cancer, Kong et al. also reported that miR-200b and miR-200c introduction downregulated the LIN28B protein 39 . Although the exact mechanism involving the miR-200s and LIN28B interaction remains unclear, it is possible that both the miR-200s and the let-7 miR families are composed of multiple members, with each member regulating largely overlapping sets of targets 40 .
In conclusion, the miR-200c/LIN28B axis plays an important role in acquired resistance to EGFR-TKI and might be a useful therapeutic target for overcoming EMT-related EGFR-TKI resistance. To date, there are no drugs directly targeting miR-200c or LIN28B expression. Further investigation of the modulation of the miR-200c/LIN28B axis may lead to further successes in EGFR-TKIs treatment.

Materials and Methods
Ethics statement.  from Thermo Fisher Scientific) 41 . In addition, we used an experimentally established EGFR-TKI resistant cell line with an EGFR activation mutations (19del), HCC4006-GR-step and HCC4006-GR-high, as previously reported 15 . The identities of all the cell lines were confirmed by analyzing the short tandem repeat profile using the Cell ID System (Promega, Madison, WI) and an ABI Prism 310 Genetic Analyzer (Thermo Fisher Scientific), according to the manufacturer's instructions. The EGFR-TKI gefitinib was purchased from InvivoGen (San Diego, CA). DNA, RNA and miRNA extraction. Genomic DNAs were extracted from cell lines using a DNeasy Blood and Tissue Kit (Qiagen, Venlo, Netherlands). Total RNAs were extracted from cell lines using an RNeasy Mini Kit (Qiagen). Complementary DNA (cDNA) was synthesized from the total RNA using High-Capacity cDNA Reverse Transcription Kits (Thermo Fisher Scientific), according to the manufacturer's instructions. The miRNA was isolate from cell lines with TaqMan MicroRNA Cells-to-CT Kit (Ambion) and total RNA was extracted from microdissected formalin-fixed, paraffin-embedded (FFPE) tissues by using a miRNeasy FFPE Kit (Qiagen). Reverse transcription was conducted with TaqMan Micro-RNA Reverse Transcriptional Kit systems (Applied Biosystems).
DNA direct sequencing and methylation analysis. We determined the mutational status of EGFR, HER2, KRAS, NRAS, and BRAF genes in NSCLC cell lines using direct sequencing, as previously reported 15 . In a methylation analysis, DNA was subjected to bisulfate treatment using the Epitect Bisulfite Kit (Qiagen), according to the manufacturer's protocol. The DNA methylation statuses of miR-200ba429 and miR-200c141 were examined using MSP, as previously reported (Supplementary Table S4A) 15 . To restore the gene expression that was reduced by methylation, cells were treated with 5-Aza (Sigma-Aldrich, St. Louis, MO) at a concentration of 5 μ mol/L for 6 days. mRNA and miRNA expression analysis using quantitative reverse transcription-PCR. Gene expression was analyzed using qRT-PCR and cDNAs, TaqMan Gene Expression Assays, and the ABI StepOnePlus Determination of cell proliferation. The IC50 value against gefitinib was determined using a modified MTS assay with CellTiter 96 AQueous One Solution Reagent (Promega) as previously reported 15 . For experiments testing the effect of miR-200c transfection or the knockdown of siRNA on cell proliferation, the MTT (Sigma-Aldrich, St. Louis, MO) dye reduction method was used. Cells were cultured at 37 °C with 5% CO 2 , in 6-well plates at a concentration of 1 × 10 5 cells/mL for 72-96 h. MTT was dissolved in RPMI-1640, and 100 μL of the MTT solution were added to each well; the plates were then incubated at 37 °C with 5% CO 2 for 2 h. Subsequently, 100 μ L of dimethyl sulfoxide were added to each well. The cell viability was assessed by measuring the optical density at 570 nm and 690 nm on a plate reader. Three independent experiments consisting of triplicate runs (at least) were performed. Immunohistochemical analysis of clinical samples. Lung cancer tissues were obtained from the patients who underwent surgery at Okayama University Hospital (Okayama, Japan). This experimental protocol was approved by the Institutional Review Board/Ethical Committee of Okayama University (No. 147) and informed consent were obtained from all the patients. Tissue samples were fixed in 10% formaldehyde and embedded in paraffin.