CircACAP2 promotes cell proliferation and migration in lung adenocarcinoma via LASP1-mediated TGF ‐ β/Smad3 pathway

Lung adenocarcinoma (LUAD), a common malignant tumor, has led to a great number of deaths around the world. Circular RNAs (circRNAs) have been certied as essential players in the progression of diverse cancers. CircRNA ACAP2 (hsa_circ_0068568) is an oncogene in several cancers. However, the role of circACAP2 in LUAD remains unknown. This study revealed that the expression of circACAP2 was signicantly elevated in LUAD tissues and cell lines, especially in the tissues of LUAD patients at advanced stage. Additionally, circACAP2 enhanced cell proliferation, migration, invasion abilities and epithelial-mesenchymal transition (EMT) process in LUAD. Moreover, miR-342-3p interacted with circACAP2 in LUAD cells. Importantly, we found that miR-342-3p targeted LIM and SH3 protein 1 (LASP1), and circACAP2 positively regulated LASP1 expression by competing for miR-342-3p in LUAD. Further, it was conrmed that circACAP2 promoted the malignant behaviors and stimulated the activation of TGF-β/Smad3 pathway in LUAD by modulating the miR-342-3p/LASP1 axis. To conclude, the molecular regulatory mechanism of circACAP2 in LUAD was under discussion in the current study. The ndings revealed that circACAP2 facilitated malignant phenotypes in LUAD via the activation of the TGF ‐ β/Smad3 pathway.


Introduction
Lung carcinoma, the most frequent serious malignancies characterized by high occurrence, poor prognosis and high mortality, is a great threat to global human health (Dong, Liu, Sun, & Ping, 2020;Ho & Leung, 2018). There are two types of lung cancer, non-small cell lung cancer accounting for about 80% and small cell lung cancer accounting for approximately 20% (Lemjabbar-Alaoui, Hassan, Yang, & Buchanan, 2015;Yuan, Liu, Qu, Liu, & Li, 2019). Lung adenocarcinoma (LUAD) is the most common type of non-small cell lung cancer (Inamura, 2018). Despite progress made in LUAD therapies, the ve-year survival rate of LUAD patients is still low because of inapparent early symptoms as well as high recurrence and metastasis rates after advanced LUAD diagnosis and treatment (Behera et al., 2016;Denisenko, Budkevich, & Zhivotovsky, 2018). Accordingly, it is urgent to explore the molecular regulatory mechanism of LUAD tumorigenesis and growth. Circular RNAs (circRNAs), non-coding RNA molecules with closed loop structures, play important roles in diverse human diseases, including cancers (Liu, Li, Luo, & Zhu, 2019;Wei & Liu, 2019). For instance, circDDX17 acts as a tumor suppressor by suppressing proliferation and accelerating apoptosis in colorectal cancer cells (Xiao et al., 2019). Besides, circRNA hsa_circ_0000263 is upregulated and played carcinogenic role through promoting cell cycle and tumor growth in cervical cancer (Cai et al., 2019).
We aimed to study the molecular mechanism and function of circACAP2 in LUAD. The in uence of circACAP2 on proliferation, migration, invasion, and EMT of LUAD cells was under investigation, which may offer an innovative theoretical basis for LUAD.

Materials And Methods
Tissue specimens Fifty paired tissues of LUAD and adjacent noncancerous tissues were obtained from LUAD patients at Harbin Medical University Cancer Hospital. The collected tissues were immediately frozen in liquid nitrogen and conserved at -80℃ for experimental needs. No patients had received corresponding radiotherapy or chemotherapy before surgery, and all tissue specimens were con rmed pathologically. Written informed consents were got from every patient. The study was permitted by the Ethics Committee of Harbin Medical University Cancer Hospital.

Cell lines and reagent
Three LUAD cell lines (A549, H1975 and PC9) and one human bronchial epithelial cell line BEAS-2B were provided by the Cell Bank of Type Culture Collection of the Chinese Academy of Sciences, Shanghai Institute of Cell Biology (Shanghai, China). All cell lines were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Thermo Fisher, Shanghai, China) consisting of 10% fetal bovine serum (FBS; Thermo Fisher) in humidi ed conditions with 5% CO 2 at 37℃. For blocking transcription, Actinomycin D (2 mg/ml; Sigma-Aldrich) was used with dimethylsulphoxide (Sigma-Aldrich) as a control. Besides, 3 U/ μg of RNase R (Epicentre Technologies) was used for RNase R treatment.

Cell transfection
For knocking down circACAP2/LASP1, sh-circACAP2#1/2 and sh-LASP1 were constructed with sh-NC as control. Besides, pcDNA3.1/LASP1 was transfected into LUAD cells for the overexpression of LASP1, and empty pcDNA3.1 was seen as an internal control. In addition, miR-342-3p mimics and NC mimics (control) were also constructed to increasing miR-342-3p expression. All above-mentioned plasmids were provided by GenePharma (Shanghai, China). Then the cell transfection was conducted with Lipofectamine 3000 (Invitrogen, USA) for 48 hours.
RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR) Total RNA was collected from LUAD tissues and cells by TRIzol reagent following the manufacturer's instructions. The extracted RNA was reverse transcribed into complementary DNA (cDNA) using a Reverse Transcription Kit (Takara, Dalian, China). Then, qRT-PCR was performed with Applied Biosystems™ TaqMan and Taqman microRNA RTKit (ABI, USA). Relative expression of genes was calculated with the 2 −ΔΔCt method with GAPDH or U6 as the internal control.

Western blot analysis
Total protein in LUAD cells was separated by RIPA lysis buffer (Beyotime, Shanghai, China) and the protein concentration was detected with a Bio-Rad Protein Assay Kit (Bio-Rad, USA). Then proteins were separated with 10% SDS-PAGE and transferred to PVDF membranes (Millipore, USA). Blocked in 5% nonfat milk for 2 hours at room temperature, the membranes were incubated with primary antibodies (Ecadherin, N-cadherin, Vimentin, LASP1, TGF-β, Smad3 and p-Smad3, GAPDH) at 4℃ overnight. Subsequently, after washing the membranes with TBST solution, secondary antibodies were added to coculture with the membranes at 37℃ for another 2 hours. All the antibodies were obtained from Abcam (Shanghai, China). Finally, the blots were detected with an ECL Detection System (Thermo Fisher Scienti c, USA).

MTT assay
The transfected LUAD cells were cultured in 96-well plates at 37℃. After 0, 24, 48 and 72 hours, each well was added with 10 μl of MTT solution. After 4 hours, every well was supplemented with 150 μl of DMSO to dissolve formazan crystal. Then, the optical density (OD) value at 490 nm was measured by a microplate reader (Bio-Rad).

Colony formation assay
The transfected LUAD cells (1000 per well) were placed into 6-well plates with RPMI 1640 medium comprising 10% FBS at 37℃. Two weeks later, the cells were washed twice with phosphate buffered saline (PBS, Thermo Fisher Scienti c), xed with 5% paraformaldehyde for 30 minutes and stained with 0.1% crystal violet (Beyotime) for another 30 minutes. Then, PBS was used to wash the cells until the solution was clear and cell colonies were counted.

Transwell assays
Transwell assays were used to examine the migration and invasion abilities of LUAD cells. For invasion assay, the transfected cells in serum-free RPMI 1640 medium were added to the upper chambers (Millipore) with Matrigel (BD Biosciences, USA). The lower chambers contained RPMI 1640 medium with 10% FBS. After 2 days, crystal violet was used to stain the invaded cells in the lower chamber and cell number was counted manually in ve random elds with a uorescence microscope (Olympus, Beijing, China). To detect cell migration ability, the procedures were the same as cell invasion assay except for the upper chambers precoated without Matrigel.

Wound healing assay
Page 5/19 The transfected LUAD cells were seeded into 6-well plates, and then incubated in RPMI 1640 medium overnight. After a linear scratch wound was created with a sterile pipette tip, serum-free RPMI 1640 medium was used to culture the cells. At 0 h and 24 h, wound areas were imaged and assessed by ImageJ software (National Institutes of Health, USA).
RNA pull down assay A549 and H1975 cells were transfected with circACAP2 and NC-circRNA labeled with biotin. Then, streptavidin magnetic beads were used to incubate with A549 and H1975 cell lysates for 4 h at 4 °C. Subsequently, the beads were rinsed in precooled lysis buffer and salt buffer. Later, levels of miRNAs interacting with circACAP2 were detected using qRT-PCR after the puri cation of pull-down RNAs.

RNA immunoprecipitation assay (RIP assay)
With the application of Magna RIP Kit (Millipore), RIP assay was carried out. A549 or H1975 cells were treated with RIPA lysis buffer, and the mixture was added with magnetic beads conjugated with Ago2 or IgG antibodies (Abcam). Afterwards, RNAs were puri ed and assessed by qRT-PCR analysis.
Fluorescence in situ hybridization assay (FISH assay) FISH assay was conducted to determine the subcellular location of circACAP2 in LUAD cells. CircACAP2 probes were designed and synthesized by RiboBio (Guangzhou, China). The probe signals were detected with a FISH Kit (RiboBio). Brie y, LUAD cells were fastened in 4% paraformaldehyde for 15 minutes. After prehybridization in PBS, the cells were hybridized in hybridization solution (RiboBio) for 30 minutes and counterstained by DAPI (Beyotime), followed by visualization under an Olympus uorescence microscope.

Statistical analysis
All statistical analyses were performed using GraphPad Prism 7 Software (GraphPad Software, USA). Differences between groups were assessed by Student's t test (comparison evaluation between two groups) or one-way ANOVA (comparison evaluation over two groups). Spearman's correlation analysis displayed the expression correlation between genes. Data were shown as the mean ± standard deviation (SD) based on at least three experiments. p < 0.05 was regarded as statistical signi cance.

CircACAP2 expression is signi cantly upregulated in LUAD and correlated with LUAD progression
To identify the association between circACAP2 and LUAD, qRT-PCR analysis showed that circACAP2 expression in LUAD tissues was higher than that in adjacent normal tissues ( Figure 1A). Additionally, the circACAP2 was highly expressed in LUAD cell lines (A549, H1975 and PC9) relative to that in BEAS-2B cell line ( Figure 1B). Further, it was suggested that circACAP2 expression in LUAD patients at advanced stage was upregulated in comparison with those at early stages ( Figure 1C). Later, the stability and distribution of circACAP2 in A549 and H1975 cells were investigated. After treating with Actinomycin D, we observed that the half-life of circACAP2 was over 24 h, while that of ACAP2 mRNA was about 4 h in both A549 and H1975 cells ( Figure 1D). In addition, we discovered the resistance of circACAP2 to RNase R digestion. Both data suggested the circular feature of circACAP2 ( Figure 1E). Later, FISH assay con rmed that circACAP2 was mainly distributed in the cytoplasmic section of A549 and H1975 cells ( Figure 1F), revealing the post-transcriptional regulation of circACAP2 in LUAD. To conclude, data above suggested that circACAP2 was a cytoplasmic RNA which harbored a loop structure and upregulated in LUAD tissues and cell lines.
CircACAP2 promotes the malignancy of LUAD Further, the biological function of circACAP2 in LUAD was explored. In A549 and H1975 cells transfected with sh-circACAP2#1/2 vectors, the expression of circACAP2 was more decreased by sh-circACAP2#1 than by sh-circACAP2#2 (Figure 2A), so sh-circACAP2#1 vectors were used to conduct follow-up assays. MTT assay exhibited that cell viability was reduced by circACAP2 knockdown ( Figure 2B). Consistently, the number of colonies was decreased after sh-circACAP2#1 transfection, indicating that circACAP2 downregulation weakened the proliferation capacity of LUAD cells ( Figure 2C-D). Then, as presented by wound healing assay, cell migration suffered an evident reduction because of circACAP2 silencing ( Figure  2E). The inhibition of circACAP2 knockdown on migration and invasion abilities in LUAD cells was further con rmed using transwell assay ( Figure 2F). In Figure 2G, the expression of E-cadherin was raised while levels of N-cadherin and Vimentin were lessened in circACAP2-repressed A549 and H1975 cells, which indicated that EMT in LUAD was hindered by circACAP2 suppression. Taken together, circACAP2 promotes cell proliferation, migration, invasion and EMT in LUAD.
CircACAP2 activates the TGF-β/Smad3 pathway via modulating LASP1 in LUAD It has been reported that the TGF-β/Smad3 pathway promotes the progression of LUAD (Jiang et al., 2019), and LASP1 can activate the TGF-β/Smad3 pathway in lung cancer (Shen, Yang, & Li, 2020). Here, we wondered whether LASP1 could affect the TGF-β/Smad3 pathway in LUAD. First, the expression of LASP1 was decreased by sh-LASP1 in A549 and H1975 cells ( Figure 6A). Then, the effect of LASP1 on TGF-β/Smad3 pathway was investigated using western blot. The protein levels of phosphorylated Smad3 were reduced after LASP1 knockdown in A549 and H1975 cells, whereas the total protein expression of TGF-β and Smad3 were unchanged ( Figure 6B). Further, we found that circACAP2 knockdown reduced p-Smad3 protein levels, but this effect was counteracted by overexpressing LASP1 (Figure 6C), indicating that circACAP2 could activate the TGF-β/Smad3 pathway by mediating LASP1. To further con rm this, TGF-β was used to carry out rescue assays in A549 and H1975 cells transfected with sh-circACAP2#1. The results revealed that the treatment of TGF-β reversed the suppressive effect caused by silenced circACAP2 on cell proliferation, migration and invasion ( Figure 6D-F). In summary, circACAP2 promotes LUAD cell growth and migration via activating the TGF-β/Smad3 pathway.

Discussion
A mounting body of evidence has suggested that circACAP2 is a key regulator in malignant tumors, consisting of breast cancer and colon cancer (J. H. He et al., 2018;B. Zhao et al., 2020). However, the biological function and regulatory mechanism of circACAP2 in LUAD remain uncertain. Our study rstly unveiled that circACAP2 was upregulated in LUAD tissues and cell lines. Further, the circular feature and cytoplasmic localization of circACAP2 in LUAD were con rmed. Moreover, circACAP2 was validated to propel cell proliferation, migration, invasion and EMT in LUAD. These suggested the oncogenic role and post-transcription of circACAP2 in LUAD.
In conclusion, circACAP2 activated the LASP1-mediated TGF-β/Smad3 pathway to accelerate proliferation, migration, invasion, and EMT of LUAD cells via sponging miR-342-3p. This discovery may provide a new clue to improve therapeutic methods of LUAD.

Declarations Acknowledgement
We thank all participators for their help.

Competing interests
There are no competing interests.

Funding
The work was supported by National Cancer Center fund project NCC201808B015 Ethics approval and consent to participate Written informed consents were got from every patient. The study was permitted by the Ethics Committee of Harbin Medical University Cancer Hospital.

Not applicable
Availability of data and material The data underlying this article will be shared on reasonable request to the corresponding author. expression in A549 and H1975 cells treated with actinomycin D or RNase R. (F) The distribution of circACAP2 in A549 and H1975 cells was determined via FISH assay. *p < 0.05, **p < 0.01, ***p < 0.001.  CircACAP2 interacts with miR-342-3p in LUAD. (A) Thirteen potential circACAP2-binding miRNAs were found on the starBase database. (B) qRT-PCR was used to examine levels of predicted miRNAs in LUAD cell lines. (C) RNA pull down assay was conducted for selecting the potential miRNAs binding to circACAP2. (D) The expression correlation between circACAP2 and miR-342-3p was accessed using Spearman's correlation analysis. (E-F) The binding site and interaction between miR-342-3p and circACAP2 was separately examined using starBase and luciferase reporter assay. (G) RIP assay was used to explore the interaction between miR-342-3p and circACAP2. **p < 0.01, ***p < 0.001.
(B-C) The qRT-PCR analysis was utilized to test expressions of seven putative mRNAs in miR-342-3p mimics-transfected A549 and H1975 cells. (D) The LASP1 expression in LUAD tissues and cell lines was evaluated via qRT-PCR analysis. (E) Spearman's correlation analysis was conducted to evaluate the expression correlation between LASP1 and circACAP2 (miR-342-3p). (F-G) StarBase website and luciferase reporter assay were respectively employed to verify the binding sequence and interaction between miR-342-3p and LASP1. (H) The binding of miR-342-3p to LASP1 was subjected to RIP assay. (I) The qRT-PCR and western blot analyses were performed to measure the mRNA and protein expressions of LASP1 after upregulating miR-342-3p or downregulating circACAP2. **p < 0.01, ***p < 0.001.

Figure 5
CircACAP2 accelerates LUAD malignancy through targeting miR-342-3p/LASP1 axis. (A) The qRT-PCR analysis was performed to examine the e ciency of LASP1 overexpression in A549 cells. (B-C) MTT and colony formation assays were performed to assess A549 cell viability and proliferation in each group. (D-G) Wound healing and transwell assays were used to detect migration and invasion of A549 cells transfected with sh-NC, sh-circACAP2#1 or sh-circACAP2#1+pcDNA3.1/LASP1. (H) Western blot analysis was carried out to evaluate the expressions of E-cadherin, N-cadherin and Vimentin in A549 cells with indicated transfections. **p < 0.01, ***p < 0.001.