TEK Suppresses Lung Adenocarcinoma Cell Phenotypes by Interacting with miR-19a-3p


 Background: We identified TEK as a key gene that that participates in lung adenocarcinoma cell migration and adhesion, and miR-19a-3p a potential upstream regulator of TEK. Both TEK and miR-19a-3p have been reported during lung cancer development. However, how TEK/miR-19a-3p interactome regulates lung adenocarcinoma remains unraveled. We herein aim to report a novel TEK/miR-19a-3p interactome in lung adenocarcinoma.Methods: The mRNA and protein expression of TEK in tissues and cell lines were determined using qPCR and Western blot, respectively. CCK-8 assay, EDU assay, flow-cytometry cell apoptosis assay, scratch assay, and cell-to-extracellular matrix adhesion assay were performed to detect the proliferation, apoptosis, migration, and adhesion of A549 and H1975 cell lines. Results：Both mRNA and protein levels of TEK were down-regulated in tumor tissues and cell lines. Compared with the control, the transfection of TEK overexpression plasmids into H1975 and A549 cells led to the significant inhibition of cancerous phenotypes. On the other hand, miR-19a-3p promoted lung adenocarcinoma cancerous cell phenotypes by downregulating TEK.Conclusions: TEK can be a potential LUAD tumor suppressor by interacting with miR-19a-3p. This novel interactome can be used as a novel therapy target for LUAD.

. Interestingly, some claimed that TEK could not be used as a chemotherapy responsive predictor (Naumnik, Chyczewska and Ossolinska 2009), whereas some claimed that TEK level might re ect the patients' response to chemotherapy (Mroz et al. 2013). It was not until 2018 when TEK was rstly identi ed as a core gene participating in angiogenesis, cell growth, cytokine secretion, and in ammatory response in LUAD (Tian et al. 2018). Then, in 2019, a subset of circulating monocytic myeloid-derived suppressor cells overexpressing ANGPT2 and TEK was identi ed as a participant in tumor immune evasion that could enlighten future immunotherapy approaches for non-small cell lung in non-small cell lung cancer was also reported (Fan et al. 2018). For the rst time, we explored the involvement of the interaction between miR-19a-3p and its downstream effector TEK in regulating the cancerous phenotypes of lung adenocarcinoma cells.
In the study, we aimed to investigate the effects of TEK and the interacting miRNA, miR-19a-3p, in LUAD. Our study revealed a novel interactome of miR-19a-3p and TEK that modi ed the malignancy phenotypes of LUAD cell.

Tissue samples
Lung adenocarcinoma tissues and normal lung tissues in our research were collected from 20 patients in the Edong Healthcare Group, Huangshi Central Hospital (A liated Hospital of Hubei Polytechnic University). The hospital Ethics Committee approved our study. The baseline characteristics of these patients were shown in Table 1. ampli cation using SYBR Green PCR kit on an ABI 7300 system (Applied Biosystems, USA). The 2 −△△Ct method was employed to calculate the relative expression. The primer sequences for TEK, miR-19a-3p and the corresponding internal reference genes were provided in Table 2.

Cell Migration And Adhesion Assays
For wound healing assay to detect cell migration, scratches on cell culture surfaces were made using plastic pipette tips. The cells were cultured without FBS for 48 hours, and the migrated lengths symbolized the cell migration ability. Cell adhesion mechanism is complex. It is involved in a variety of aggressive processes including cell migration and invasion and potential cell-cell communication.
To determine the cell adhesion to extracellular matrix, a colorimetric method was employed. Brie y, 96-well plates were coated with collagen type I (50 lg/ml; Sigma Chemical Co., Steinheim, Germany) at 4 °C overnight. A549 and H1975 cells were allowed to adhere at 37 °C. Cell culture was washed with DMEM for three times. At 30 min and 60 min, unbound A549 and H1975 cells were washed away, and adherent cells were xed with 4% paraformaldehyde, and stained with crystal violet (0.5%) for 10 minutes. After stain extraction, the relative cell attachment was determined using absorbance readings at 620 nm.

Statistical Analysis
The GraphPad PRISM Version 8.0.1 statistical program (San Diego, CA, USA) was employed for our data (expressed as mean ± standard deviation) analysis, and diagram construction. Statistical signi cance was determined by one-way ANOVA followed by Dunnett's post hoc tests for multiple-group data, and Student's t test for two independent-group data. Probability smaller than 0.05 was regarded statistically signi cant. Each experiment was performed at least three times independently.

Results
Microarray analysis of potentially core participants of LUAD GSE118370 (Normal: n=6; Tumor: n=6) and GSE89039 (Normal: n=8; Tumor: n=8) went through GEO2R analyses. A Venn diagram was drawn to identify the 540 overlapping differentially expressed genes (DEGs) of GSE118370 and GSE89039 (Fig. 1A). These 540 DEGs then underwent functional enrichments using STRING (Fig. 1B) and Metascape (Fig. 1C) algorithms. STRING algorithm showed that these DEGs were signi cantly related to regulation of migration (Fig. 1B), whereas Metascape algorithm showed that the DEGs were signi cantly related to regulation of adhesion (Fig. 1C). The built-in MCODE algorithm of Metascape identi ed the core genes (Fig. 1D). Cell migration and adhesion phenotypes were signi cant cancerous cell phenotypes, thus were chosen for further key genes identi cation. 31 genes were identi ed to participate in both cell migration and adhesion of lung adenocarcinoma (Fig. 1E). WEBGESTALT software was nally used to enrich KEGG terms of the 31 DEGs. And they were shown to be signi cantly related to PI3K-Akt signaling pathway. To be speci c, ANGPT1, LAMA2, LAMA3, LAMA4, ITGA6, PIK3R1 and TEK were enriched to be participants of PI3K-Akt signaling (Fig. 1F). Among them, ANGPT1, LAMA3, ITGA6 and TEK were also identi ed by the MCODE algorithm in the last step. Then ENCORI algorithm ) was employed to analyse the relationship between the expression of the 7 genes and the overall survival outcomes of LUAD patients. The results showed that LUAD patients with lower level of IPK3R1 and TEK displayed signi cantly poorer survival outcomes compared to patients with high expression of PIK3R1 and TEK. And the other ve DEGs did not relate to the survival of LUAD patients.
Interestingly, a high level of ITGA6 seemed to predict a signi cantly better survival outcome ( Fig. 2A-G).
The low expression of TEK in lung adenocarcinoma tissues and cell lines GEPIA was used to preliminarily analyse the expression of ITGA6, PIK3R1 and TEK. ITGA6 was shown to be signi cantly upregulated in LUSC but not in LUAD, in which type of cancer it was not found signi cantly differentially expressed in tumor tissues and healthy tissues (Fig. 3A). PIK3R1 and TEK, on the other hand, were both shown to be signi cantly downregulated in both LUSC and LUAD. In particular, TEK ( Fig. 3B-C). Fig. 3D gives an overall insight into the expression levels of the three genes in LUAD and LUSC. We further veri ed the mRNA expression of TEK and PIK3R1 in LUAD tissues and healthy tissues that we collected. Both TEK and PIK3R1 exhibited an approximately half level in tumor tissues of normal tissues (Fig 3E-F). Similarly, the expression of TEK and PIK3R1 mRNA (Fig. 3G) and protein (Fig. 3I)  TEK suppressed the progression of LUAD in vitro Before the cellular experiments, the transfection e ciency of TEK overexpression plasmids was con rmed in both A549 and H1975 cell lines (Fig 4A-B). CCK-8 and EDU assay were performed to study the suppressive effect of TEK on LUAD cell proliferation. Compared with the blank control group, TEK overexpression signi cantly suppressed A549 cell line proliferation at 48 h and 72 h (Fig 4C), and H1975 cell line at 72 h ( Fig 4D). Then, EDU assay results con rmed the CCK-8 results (Fig 4E-F). On the other hand, the overexpression of TEK distinctly increased the cell apoptosis rate in A549 ( Fig 4G) and H1975 (Fig 4H) cells. Scratch assay results showed that exogenous TEK overexpression led to signi cant impairment of migration of LUAD cells. The migration rate of A549 cells was reduced by over 50%, and that of H1975 cells was reduced by more than 40% (Fig 5A-B). In addition, the results of cell adhesion assay showed that TEK overexpression had no effect on the adhesion ability of LUAD cells at 30 min, but at 60 min, the adhesion ability of LUAD cells was obviously weaker (approximately 1/3 weaker in A549 cells and 1/2 in H1975 cells) (Fig 5C-D).
MiR-19a-3p directly targeted TEK in LUAD miRDB and TargetScan Human 7.2 algorithms were used to predict potential upstream miRNAs for TEK. 71 common candidates were identi ed (Fig. 6A). dbDEMC database was used to do a meta pro ling of the 71 miRNAs in lung cancer. 12 candidates were found signi cantly (Fig. 6B), whilst the others were found signi cantly downregulated upregulated in lung cancer (data not shown). miRDB predicted that miR-19a-3p might target sequences 827-833 and 892-898 on the 3'UTR of TEK (Fig 6C). The 3'UTR reporter assay results con rmed that. The relative luciferase activity of the cells with wild-type 3'UTR sequences (827-833 and 892-898), particularly 827-833 region, and miRNA mimic was signi cantly lower ( Fig 6D). The high level of miR-19a-3p was determined in both LUAD tissues ( Fig 6E) and LUAD cell lines ( Fig 6F). As expected, correlation analysis indicated an inverse correlation relationship between miR-19a-3p and TEK mRNA expression (Fig 6G). In addition, the transfection e ciency of miR-19a-3p inhibitor and miR-19a-3p mimic in A549 and H1975 cells were analyzed by qRT-PCR. The e ciency of inhibitor reached approximately 75% (Fig 6H).
MiR-19a-3p negatively regulated TEK in LUAD Finally, rescue experiments were carried out to illuminate whether TEK was the potential downstream effector of miR-19a-3p in LUAD. TEK overexpression led to signi cantly impaired cell proliferation, which was reconstructed by miR-19a-3p mimic in A549 (at 48 h and 72 h) and H1975 cells (at 72 h) (Fig 8A). TEK overexpression obviously suppressed the ability of cell migration, which was abrogated by miR-19a-3p mimic in A549 and H1975 cells ( Fig 8B). As expected, miR-19a-3p mimic signi cantly restored the inhibition on cell adhesion ability induced by TEK upregulation in A549 and H1975 cells (Fig 8C).

Discussion
We have used A549 and H1975, two lung adenocarcinoma cell lines, to investigated into the potential therapy target, the miR-19a-3p/TEK axis. We found that TEK was signi cantly downregulated in LUAD tissues and cell lines, and hypothesized that TEK might be a LUAD suppressor. On the other hand, we identi ed a potential regulator of TEK, miR-19a-3p. How miR-19a-3p promoted LUAD phenotypes by targeting TEK was studied in our work.
Previous studies indicated that TEK might be promising prognosis markers in cancers such as clear cell renal carcinoma (Nakashima et al. 2019). TEK was reported to be associated with the complete response to bevacizumab in breast cancer patients (Makhoul et al. 2017). TEK also could modulate the interaction of glioma and brain tumor stem cells, and promote an invasive phenotype (Liu et al. 2010). Furthermore, TEK is a novel prognostic marker for clear cell renal cell carcinoma (Pan et al. 2019). However, how TEK affects the invasive cancerous phenotypes of LUAD remains largely unknown. We herein reported that TEK was signi cantly downregulated in LUAD tissues and cell lines compared with normal tissues and cell lines. The low level of TEK not only predicted a poorer survival outcome of LUAD patients, but also was related to tumor staging. The overexpression of TEK results in signi cant impaired proliferation and aggression abilities of LUAD cell lines, suggesting its possible LUAD suppressor role.
High serum miR-19a level could be an independent poor prognostic indicator for non-small cell lung cancer because it is signi cantly related to the malignancy characteristics of non-small cell lung cancer patients (Lin et al. 2013). miR-19a-3p has also been found to ful l its regulation effects by targeting its downstream mRNAs in lung cancer. For instance, miR-19a-3p was reported to downregulate suppressor of cytokine signaling 1 (SCS1) to promote non-small cell lung cancer cell canceration (Wang and Chen 2015). miR-19a-3p has also constantly been reported to be related to chemotherapy resistance and epithelial mesenchymal transition (EMT). . TEK was identi ed to participate in PI3K/Akt signaling in our bioinformatics analyses process; however, we did not study how TEK regulated PI3K/Akt signaling at this stage of our research. Then, again, to provide a solid basis for further studies on how miR-19a-3p regulates chemo-resistance, we thoroughly studied its effect on LUAD cell phenotypes. We herein reported that miR-19a-3p promoted LUAD cancerous phenotypes by targeting a novel downstream effector, TEK.
Certain limitations of the current research existed. On one hand, in vivo experiments should be conducted to con rm the in vitro results in the coming future. Then, epithelial-mesenchymal transition (EMT) is a hall marker of lung cancer-genesis. miR-19a-3p has been reported to induce EMT in A549 cell line and HCC827 cell line by targeting PTEN (Li et al. 2015). Whether miR-19a-3p could compel EMT by targeting TEK remains to be answered. In addition, the reported relationship of miR-19a-3p with chemotherapy resistance in lung cancer also calls for further study of how miR-19a-3p mediates drug resistance by targeting TEK. Finally, a thorough study into the effects of TEK and miR-19a-3p on PI3K-Akt signaling is warranted.

Conclusions
We

Declarations
The datasets used during the current study are available from the corresponding author on reasonable request.

Competing interests
There is no con ict of interest existed among the authors.  The overall survival analyses of the seven genes that participate in PI3K-Akt signaling. A-G represents the overall survival outcomes of LUAD patients with high and low levels of ANGPT1, LAMA2, LAMA3, LAMA4, ITGA6, PIK3R1 and TEK, respectively. Blot of TEK expression in Figure 3I and Uncropped Blot of GAPDH expression in Figure 3I. *P<0.05, **P<0.01, compared with BEAS-2B cell line using one-way ANOVA method.    H1975 cells (B) during the following 72 h after transfecting miR-19a-3p mimic and inhibitor was measured using CCK-8 assays. C. Wound healing assay of A549 and H1975 cell lines after transfected with miR-19a-3p mimic and inhibitor for 48 h. The migration rate was calculated as (the distance at 0 h-the distance at 48 h)/the distance at 0h. D. Cell adhesion assay revealed that cell adhesion ability increased in mimic group but decreased in inhibitor group 60 min after the cells were plated to the matrix. *P<0.05, **P<0.01, compared with blank control group using one-way ANOVA method.