Metformin suppresses lung adenocarcinoma by downregulating long non-coding RNA (lncRNA) AFAP1-AS1 and secreted phosphoprotein 1 (SPP1) while upregulating miR-3163.

AFAP1-AS1 plays a pro-tumor role in lung cancer. However, no investigation has focused on whether it is involved in the anticancer activity of metformin (Met) in the treatment of lung adenocarcinoma (LUAD). Reverse transcription quantitative polymerase chain reaction (RT-qPCR) was performed to detect the expression of long non-coding (lnc)RNA AFAP1-AS1, the microRNA (miR)-3163, and secreted phosphoprotein 1 (SPP1) in LUAD tissues, or of A549 and H3122 cells. Cell Counting Kit-8, wound scratch, and cell invasion assays were performed to evaluate the effect of the overexpression of lncRNA AFAP1-AS1, miR-3163, and SPP1 on the malignant behaviors of A549 and H3122 cells. Phosphatidylinositol 3-kinase/protein kinase B/mammalian target of rapamycin (PI3K/Akt/mTOR) signaling pathway-related proteins were detected by Western blot analysis. Dual luciferase reporter or RIP assays were used to determine the interplay between AFAP1-AS1 and miR-3163, or of miR-3163 and SPP1. Met inhibits the malignant characteristics of A549 and H3122 cells in vitro. GEPIA database analysis showed that AFAP1-AS1 is a highly expressed lncRNA in LUAD tissues, which was validated by RT-qPCR. Overexpression of AFAP1-AS1 suppressed the met-mediated anti-tumor activity in A549 and H3122 cells, while AFAP1-AS1 silencing promoted it. Met inhibited AFAP1-AS1 expression, which resulted in reduced proliferation, migration, and invasion in A549 and H3122 cells. This led to AFAP1-AS1-mediated suppression of miR-3163 and, subsequently, the upregulation of SPP1. Met exerts its antitumor activities by regulating the AFAP1-AS1/miR-3163/SPP1/PI3K/Akt/mTOR axis. Our findings deepen our understanding of mechanisms underlying anti-tumor effect of Met in LUAD.

Highlights

Metformin inhibited the expression of the lncRNA AFAP1-AS1, and resulted in the reduced proliferation, migration, and invasion of LUAD cells.

Overexpression of AFAP1-AS1 suppressed the anti-tumor effect of Met on LUAD cells, while lncRNA AFAP1-AS1 silencing promoted it.

lncRNA AFAP1-AS1 acts as a sponge of miR-3163 targeted to SPP1 activated PI3K/Akt/mTOR signaling pathway.

Background

The highly aggressive and unfavorable prognosis of lung cancer allows this disorder to be listed as one of the most serious malignancy types, despite the great advances in curative treatment modalities [1]. Lung adenocarcinoma (LUAD) is the most typical histologic subtype of lung cancer, with over one million deaths globally each year [2]. Apart from environmental tobacco smoke, pulmonary infections, and western lifestyle, epidemiological investigations have shown that diabetes, especially type 2 diabetes (T2D), increases the incidence of LUAD [3,4]. Notably, maintaining constant blood glucose levels in individuals with T2D has been demonstrated to reduce the risk of lung cancer development. Therefore, the anti-tumorigenic abilities of glucose-lowering drugs have been noted because they can attenuate postprandial blood glucose levels or stabilize blood sugar levels.

Metformin (Met) is an oral anti-hyperlipidemic drug used as a first-line treatment for diabetes [5]. The mechanism of blood sugar control is to control the liver glucose output, improve the sensitivity of peripheral tissues to insulin, thereby increasing the uptake and utilization of glucose, and simultaneously controlling blood sugar at the same time from the source and destination. Recently, emerging evidence has confirmed the encouraging anti-tumor effect of Met [6-8]. Furthermore, its long-term application attenuates the risk of multiple types of malignancies [9]. In lung cancer, several retrospective cohort studies revealed that Met application could reduce the risk of developing lung cancer and improve the prognosis of lung cancer patients [10]. In addition, the anti-tumor effect of Met has been validated both in vitro and in vivo [11]. Mechanistically, Met is thought to exert an anti-tumor activity in lung cancer by inactivating the phosphatidylinositol 3-kinase/protein kinase B/mammalian target of the rapamycin (PI3K/Akt/mTOR) signaling pathway [12]. However, the detailed mechanism remains unclear.

Long non coding RNAs (lncRNAs) are transcripts over 200 nucleotides that do not encode proteins. The involvement of lncRNAs in cancer biology through the modulation of various pathophysiological processes is gradually being accepted. Aberrantly expressed lncRNAs function as tumor-promoting and tumor-suppressive factors during cancer progression [13]. The lncRNA, AFAP1-AS1, was found to be upregulated in diverse tissues and is involved in the regulation of tumor malignancies [14-16]. The in vitro and in vivo tumor-promoting effects of AFAP1-AS1 were investigated in lung cancer [17-19]. However, recent investigations have shown that Met can participate in the expression of lncRNA, thereby interfering with pathological processes contributing to multiple disorders [20-23]. Therefore, we hypothesized that Met exerts its anticancer activity by regulating the axis of lncRNA AFAP1-AS1. In the present study, we examined the role of AFAP1-AS1 during the effect of Met on LUAD cell proliferation, migration, and invasion. Using subsequent bioinformatics analyses and functional assays, we also elucidated the underlying mechanisms of this action.

We hypothesized that Met is involved in regulating lncRNA AFAP1-AS1 to inhibit the proliferation, migration, and invasion of LUAD cells, and explored the downstream miR-3163/SPP1/PI3K/Akt/mTOR signaling axis of lncRNA AFAP1-AS1. The purpose of this study was to explore the molecular mechanisms of Met in the treatment of LUAD.

Material and methods

Patient samples

Thirty-two LUAD tissues and 32 adjacent uncancerous lung tissues were obtained from our local hospital. Informed consent forms were provided by all patients. Ethical approval was granted by the Institutional Ethics Committee.

Cells and cell treatment

A549 human non-small cell lung carcinoma cells and H3122 human LUAD cells were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). A549 cells and H3122 cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum (FBS). Cells were grown at 37°C in a 5% CO2 incubator for 24 h and were divided into two groups: treated with Met (Sigma-Aldrich, USA) at the indicated concentrations (Met group) or with dimethylsulfoxide (CON group) [24].

Plasmids and antibodies

The recombinant pcDNA3.1 plasmid containing the full sequence of lncRNA AFAP1-AS1 (OE-lnc) or SPP1 (OE-SPP1) and the empty control vectors were purchased from Ubigene (Shanghai, China) along with small interfering RNA (siRNA) targeting AFAP1-AS1(si-lnc) and siRNA negative control (si-NC), miR-3163 mimic, and mimic NC. All transfections were performed according to the protocol of the Lipofectamine 3000 kit (Invitrogen, USA).

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

Total RNA was isolated from cells or tissue samples using MolPure Cell RNA Kits (Yeason, China). Purity was detected using an OD260/OD280 ratio > 1.8. The One Step Primer Script® miRNA cDNA Synthesis Kit (Haigene, China) and PrimeScript™ RT Reagent kit (TaKaRa) were used for reverse transcription of RNA to cDNA. SYBR Green qPCR Super Mix (Invitrogen) was used to quantify the indicated genes. Uracil6 (U6) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used for normalization. The 2−∆∆CT method was employed to analyze the relative expression of the indicated genes [25]. The primer sequences used for RT-qPCR analysis are shown in Table 1.

Table 1.

Primer sequences used for RT-qPCR analysis

Gene	Primer sequence (5ʹ-3ʹ)
lncRNA AFAP1-AS1	F:TCGCTCAATGGAGTGACGGCA
	R:CGGCTGAGACCGCTGAGAACTT
miR-3163	F:GCCGAGTATAAAATGAGGGCA
	R:CTCAACTGGTGTCGTGGA
SPP1	F:TAGCCAGGACTCCGTTGACT
	R:ACACTATCACCTCGGCCATC
GAPDH	F:GTCAACGGATTTGGTCTGTATT
	R:AGTCTTCTGGGTGGCAGTGAT
U6	F:CTCGCTTCGGCAGCACA
	R:AACGCTTCACGAATTTGCGT

Western blot

The assay was performed as described previously [26]. Cells were processed using radioimmunoprecipitation assay (RIPA) buffer, and the obtained cell lysate was centrifuged at 12,000 rpm for 20 min at −4°C. The cell lysate supernatant was collected for quantification using a bicinchoninic acid (BCA) kit (BCA Protein Assay Kit, P0010). Protein samples (25 µg) were fractionated using 10% sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) before polyvinylidene fluoride (PVDF) membrane transfer. Primary antibodies were diluted as instructed and then incubated with the membrane in a cool room for 24 h, followed by continuous incubation with horseradish peroxidase (HRP)-coupled secondary antibody at room temperature for another 1 h. Pierce electrogenerated chemiluminescence (ECL) Western blot substrate (Thermo Scientific, USA) was used to develop the membrane on X-ray films. The antibodies used were as follows: anti PI3K (ab278545), anti-Akt (ab38449), anti-phosphorylated (p)-Akt (ab38449), anti-p-mTOR (ab137133), and anti-mTOR (ab134903) antibodies obtained from Abcam (Cambridge, United States).

Cell counting kit-8 (CCK-8)

First, a CCK-8 kit (GlpBio, USA) was used to detect the median inhibitory concentration (IC50) of Met against the two LUAD cells. Briefly, untransfected A549 and H3122 cells (2.0 × 105 cells/well) were maintained in a 96-well plate for 24 h until they reached 70–80% confluence. Subsequently, the cells were continuously cultivated with 0, 5, 10, 20, 40, 80, and 160 μM/L Met. After 24 h, 10 μL CCK8 solution was added to each well and incubated with the cells for an additional 1 h. Optical density (OD) values were read at 450 nm using a plate reader (Thermo Fisher Scientific). Cell proliferation and inhibitory concentration (IC)50 were determined as previously reported [27] for the next cell functional assays. In another set of assays, transfected cells were exposed to Met at the concentrations of IC50 for 24 h and enumerated using a series of functional assays.

Scratch wound-healing migration assay

After 24 h of Met exposure, cells (3 × 104 cells/well) were maintained in six-well plates until they reached 95% confluence as a monolayer. Next, a new 1 mL pipette tip was used to create a cross in the center of the well perpendicular to the bottom of the well. After the detached cells were removed, the fresh medium was replenished. Following continuous 24 h cultivation, the cells were fixed with 3.7% paraformaldehyde for 30 min, before staining with 1% crystal violet in 2% ethanol for 30 min. A microscope was used to capture the photographs at 0 and 24 h [26].

Transwell invasion assay

After 24 h of Met exposure, 3 × 105 of the transfected cells were trypsinized and resuspended in fresh medium before being added to the upper compartment of the Millicell insert pre-coated extracellular matrix (ECM) gel. The prepared inserts were positioned onto the wells of the plate containing 1.5 ml of DMEM supplemented with 10% FBS. After 24 h of incubation, the upper compartments were removed. The cells on the upper side were removed. The invasive cells underneath were fixed with 5% glutaraldehyde for 10 min, stained with 1% crystal violet for 20 min, and then counted under a microscope [26].

Dual luciferase assay

The luciferase reporter constructs carrying wild-type lncRNA AFAP1-AS1 (WT AFAP1-AS1) or 3ʹUTR SPP1 (WT SPP1) and their corresponding mutant vectors (Mut AFAP1-AS1 and Mut SPP1) were purchased from Yuanjin Biology Company, Guangdong, China, to verify the interplay between lncRNA AFAP1-AS1, SPP1, and miR-3163. Briefly, the recombinant reporter vectors were transfected with miR-3163 mimic or mimic NC along with Renilla luciferase vectors. The fluorescence signals were detected using Dual-Lucy Assay Kits (Thermo Fisher, USA) [28].

RNA immunoprecipitation (RIP) assay

An EZ-Magna RIP Kit (Millipore, China) was used to further verify the interaction between lncRNA AFAP1-AS1 and miR-3163. Briefly, A549 and H3122 cells were lysed before incubation with magnetic beads precoated with anti-Ago2 antibody (Boster, Wuhan, China) for 24 h at 4°C. The pull-down Ago2-RNA complex was digested by proteinase K following RT-qPCR analysis [29].

Statistical analysis

Three biological replicates were implemented, and the values are presented as means ±  standard deviation. Unpaired Student’s t-test was used to analyze the unparied clinical sample groups. One-way analysis of variance (ANOVA) or two-way ANOVA was used to analyze the multi-group measurements. Prism 9.00 software was applied for statistical analyses. Statistical significance was set at P < 0.05.

Results

Here, we aimed to explore the molecular mechanisms of Met in the treatment of LUAD. We hypothesized that Met suppresses LUAD by downregulating lncRNA AFAP1-AS1, and secreted SPP1 activated the PI3K/Akt/mTOR signaling pathway, while upregulating miR-3163. The GEPIA database showed that AFAP1-AS1 was highly expressed in LUAD. A series of functional experiments revealed that lncRNA AFAP1-AS1 suppressed anti-tumor activity of Met in A549 and H3122 cells, while lncRNA AFAP1-AS1 silencing promoted it. Our findings deepen the theoretical knowledge about Met- AFAP1-AS1/miR-3163/SPP1/PI3K/Akt/mTOR application in LUAD.

The identification of SPP1 and miR-3163 as the potential participants in LUAD

PI3K/AKT signaling has been considered to play a pivotal role in lung cancer progression, including cases of drug application [30-33]. By analyzing differentially expressed genes in lung cancer from GSE85841 database, we identified 20 significantly upregulated genes with adjusted P < 0.05, and logFC ≥ 1.5. Among the 20 genes, we identified SPP1 as a critical gene in the PI3K/AKT signaling pathway according to the Kyoto Encyclopedia of Genes and Genomes database (Figure 1(a)). SPP1 has been reported to be a significant tumor facilitator in lung cancer [34-37]. By interrogating the GEPIA database, we found that SPP1 was significantly upregulated in both lung squamous cell carcinoma (LUSC) and LUAD (Figure 1(b)). The survival data of LUAD patients obtained from the KMplotter database (http://kmplot.com/) showed that SPP1 high expression was significantly associated with poorer overall survival outcomes (Figure 1(c)). According to GEPIA LUAD data, AFAP1-AS1 was the most significantly upregulated lncRNA with a logFC of 3.297 (Figure 1(d)). AFAP1-AS1 has been revealed to be a potent oncogene and a participant in the chemoresistance of lung cancer [14,19,38-41]. However, whether AFAP1-AS1 participates in the progression of LUAD remains unclear. By intersecting the target miRNAs of SPP1 and AFAP1-AS1, we identified three miRNAs that potentially link SPP1 and AFAP1-AS1: miR-520 g-3p, miR-520 h, and miR-3163 (Figure 1(e)). The expression of the three miRNAs was detected in our collected samples, and it was found that miR-3163 showed significantly downregulated expression levels, whereas the other two miRNAs did not show different expression levels when compared to the healthy control samples (Figure 1(f)). In addition, miR-3163 was found to be a potent tumor suppressor in other human cancers such as colorectal cancer [42,43] and cervical cancer [44].

Figure 1.

The identification of SPP1 and miR as the potential participants in LUAD. (a) The identification of genes that were both differentially expressed in lung cancer and critical members in PI3K’AKT signaling. (b) The expression of SPP1 in lung cancer according to GEPIA database. LUSC: lung squamous cell carcinoma. LUAD: lung adenocarcinoma. (c) The overall survival outcomes of LUAD patients according to KMplotter database using three probes against SPP1. (d) The significantly upregulated lncRNAs in lung adenocarcinoma (LUAD) according to GEPIA LUAD data. (e) A Venn diagram showing the common miRNAs that were targets of both SPP1 and AFAP1-AS1. (f) The miRNAs expression in LUAD tissues and normal tissues.

Met curbs the malignant characteristics of LUAD cells by inactivating PI3k/AKT/mTOR signaling pathway

As a T2M drug, Met also displays anti-cancer activity and can improve the poor prognosis of LUAD. Met treatment suppressed the proliferative potential of A549 and H3122 cells in a dose-dependent manner (Figure 2(a)). The IC50 of Met against A549 and H3122 cells was 13.5 mM and 21.8 mM (Figure 2(b)), respectively, which were then chosen as the optimal concentrations for subsequent assays. Furthermore, Met treatment led to low migratory and aggressive capacities of these LUAD cells compared to the untreated groups (Figure 2(c,d)). A previous investigation suggested that the PI3K/AKT/mTOR signaling pathway is involved in the anticancer activity of Met [45]. Consequently, we further examined the phosphorylation status of this signaling cascade by Western blotting. As expected, the increased levels of phosphorylated PI3K, AKT, and mTOR were observed in Met-treated A549 and H3122 cells (Figure 2(e)). Therefore, Met treatment inhibited the malignant phenotypes of LUAD cells in vitro.

Figure 2.

Metformin inhibits the LUAD cell proliferation, migration and invasion. (a) Molecular structure of Metformin. (b) Proliferation and IC50 analysis of Metformin. The IC50 of Metformin for A549 and H3122 cells was about 13.5 mM. The IC50 of Metformin for G401 cells was about 21.8 mM. (c) Cell scratch experiment analysis of the migration of A549 and H3122 cells treated with 13.5 mM and 21.8 mM Metformin. (d) Transwell invasion analysis of A549 and H3122 cells treated with 13.5 mM and 21.8 mM Metformin. (e) Western bolt analysis of the critical effectors in PI3K signaling pathway in A549 and H3122 cells treated with 13.5 mM Metformin**P < 0.01 compared with CON group. CON, control; Met, Metformin.

High expression of lncRNA AFP1-AS1 negatively correlated with LUAD cell sensitivity to Met

To further explore the function of lncRNA AFAP1-AS1 in Met anti-cancer activity, we first detected the expression status of AFAP1-AS1 in LUAD patients. As shown in Figure 3(a), high expression was observed in LUAD tissues compared with that in non-cancerous lung tissues. RT-qPCR analysis showed that Met treatment caused an almost four-fold reduction in the level of AFAP1-AS1 expression (Figure 3(b)), indicating that lncRNA AFAP1-AS1 expression might be associated with the anti-tumor activity of Met. To further confirm this idea, we exogenously expressed or knocked down lncRNA AFAP1-AS1 in A549 and H3122 cells via cell transfection of lncRNA AFAP1-AS1 overexpression or siRNA vectors. Transfection efficiency was confirmed by high or low expression of lncRNA AFAP1-AS1 indicated in the OE-lnc or si-lnc groups of both LUAD cells by RT-qPCR (Figure 3(c)). Next, a series of cell functional assays were employed to detect the malignant characteristics of LUAD cells after Met treatment. As shown in Figure 3(d), despite the dose-dependent decrease in cell proliferation detected in A549 and H3122 cells after treatment with Met, the exogenous expression of lncRNA AFAP1-AS1 attenuated the reduced cell proliferation, whereas downregulation of lncRNA AFAP1-AS1 inhibited cell proliferation. The results of the scratch and transwell assays demonstrated that overexpression of lncRNA AFAP1-AS1 weakened the strong anti-migratory and anti-invasive capacities of Met in A549 and H3122 cells, while low expression of lncRNA AFAP1-AS1 reinforced it (Figure 3(e,f)). These findings suggest that lncRNA AFAP1-AS1 might be involved in the anti-tumor activity of Met.

Figure 3.

Overexpression of lncRNA AFAP1-AS1 weakens the sensitivity of A549 and H3122 cells to Met in vitro. (a) RT-qPCR analysis of lncRNA AFAP1-AS1 expression level in LUAD tissues and noncancerous lung tissues. (b) RT-qPCR analysis of lncRNA AFAP1-AS1 expression level in A549 and H3122 cells after Met treatment for 24 h. (c) RT-qPCR analysis of lncRNA AFAP1-AS1 overexpressing or low-expressing efficiency after cell transfection with lncRNA AFAP1-AS1-overexpression vectors. (d) Comparison of IC50 of Met in A549 and H3122 cells after lncRNA AFAP1-AS1-overexpression or low-expression by CCK8 assays. (e) The migratory potential was assessed by scratch wound assay. (f) The invasive potential was measured by transwell invasion assay. *P < 0.05, **P < 0.01. CON, control; OE-lnc, overexpression of lncRNA AFAP1-AS1; empty vector, negative control of OE-lnc; si-lnc, knockdown of lncRNA AFAP1-AS1; si-NC, negative control of si-lnc.

MiR-3163 is sponged by lncRNA AFP1-AS1 and increases the sensitivity of LUAD cells to Met

It is widely accepted that lncRNAs may attenuate miRNA function through their ceRAN activity. Therefore, we employed a StarBase 3.0 search and found a novel miRNA, miR-3163, which might be involved in the Met treatment to control LUAD progression (Figure 4(a)). The dual luciferase system clearly confirmed the significant reduction of AFP1-AS1-mediated luciferase activity as a consequence of the overexpression of miR-3163 in A549 and H3122 cells (Figure 4(b)). Furthermore, the enrichment of miR-3163 and AFAP1-AS1 on Ago2 was monitored in A549 and H3122 cells compared to that in A549 and H3122 cells (Figure 4(c)). A negative association between miR-3163 and AFAP1-AS1 expression in LUAD tissues was found by Spearman analysis (Figure 4(d)), further supporting their interplay in LUAD progression. We also detected miR-3163 expression in A549 and H3122 cells after treatment with Met. Contrary to the upregulation of lncRNA AFAP1-AS1 expression, miR-3163 levels were considerably enhanced in A549 and H3122 cells after exposure to Met for 24 h (Figure 4(e)). Transfection with miR-3163 mimics could obviously recover the suppression of miR-3163 expression by lncRNA AFAP1-AS1 (Figure 4(f)). These data suggest that lncRNA AFAP1-AS1 might bind miR-3163 and be involved in LUAD progression.

Figure 4.

miR-3163 is a target of lncRNA AFAP1-AS1. (a) Specific binding regions between lncRNA AFAP1-AS1 sequence and miR-3163 sequence detected. (b) the binding of miR-3163 to AFAP1-AS1 verified by dual‐luciferase assays. (c) The enrichment of miR-3163 to lncRNA AFAP1-AS1 in Ago2 and IgG detected by RIP assay. (d) the correlation between miR-3163 to lncRNA AFAP1-AS1 levels. (e) RT-qPCR analysis of miR-3163 expression in A549 and H3122 cells after exposure to Met for 24 h. (f) RT-qPCR analysis of miR-3163 in A549 and H3122 cells after transfected with empt vetor, OE-lnc, OE-lnc+mimic-NC, OE-lnc+mimic.*P < 0.05, **P < 0.01. CON, control; Met, Metformin; OE-lnc, overexpression of lncRNA AFAP1-AS1; empty vector, negative control of OE-lnc; OE-lnc+mimic, overexpression of lncRNA AFAP1-AS1 and miR-3163; OE-lnc+mimic-NC, overexpression of lncRNA AFAP1-AS1 combined with negative control of mimic.

Next, we further verified the dependence of the function of lncRNA AFAP1-AS1 on miR-3163 during LUAD malignancy by assessing phenotypic alterations in LUAD cells in vitro. As shown in Figure 5(a), miR-3163 ectopic expression reduced the proliferation of A549 and H3122 cells overexpressing AFAP1-AS1 to levels comparable to those in cells treated with Met alone. Furthermore, miR-3163 mimic transfection inhibited the lncRNA AFAP1-AS1 overexpression-induced migration and invasion of LUAD cells (Figure 5(b,c)).

Figure 5.

Overexpressing of lncRNA AFAP1-AS1 and overexpressed miR-3163 regulates LUAD cell proliferation, migration and invasion. Cells were transfected with empty vector, lncRNA AFAP1-AS1 overexpression vectors (Lnc-OE), Lnc-OE+miR-3163 mimic NC or Lnc-OE+miR-3163 mimic before Met treament. (a) Cell proliferation determined by CCK8. (b) Cell migration determined by scratch wound assay. Cell invasion determined by transwell invasion assay. (c) Cell invasion determined by Transwell invasion assays. *P < 0.05, **P < 0.01. CON, control; OE-lnc, overexpression of lncRNA AFAP1-AS1; empty vector, negative control of OE-lnc; OE-lnc+mimic, overexpression of lncRNA AFAP1-AS1 and miR-3163; OE-lnc+mimic-NC, overexpression of lncRNA AFAP1-AS1 combined with negative control of mimic.

MiR-3163 targets SPP1

As previously predicted, SPP1 was the target of miR-3163 (Figure 6(a)). Subsequently, the 3ʹUTR-SPP1-WT luciferase reporter constructs or 3ʹUTR-SPP1-MUT luciferase reporter constructs were co-transfected with miR-3163 mimic or mimic NC for luciferase activity evaluation. The miR-3163 mimic significantly impaired the 3ʹUTR-SPP1-WT luciferase signaling in the two LUAD cell lines, but showed no influence on the 3ʹUTR-SPP1-MUT luciferase signaling (Figure 6(b)). Moreover, an increased expression of SPP1 in LUAD tissues was detected (Figure 6(c)), and Spearman’s correlation indicated a clear negative correlation between SPP1 and miR-3163 expression levels (Figure 6(d)). Met treatment resulted in a four-fold reduction in SPP1 expression in untransfected LUAD cells (Figure 6(e)). To further confirm the targeted relationship between miR-3163 and SPP1, we co-transfected SPP1-overexpressing vectors (SPP1-OE) and miR-3163 mimic or mimic NC in A549 and H3122 cells. At the mRNA and protein levels, ectopic expression of miR-3163 not only inhibited the expression of SPP1 in A549 and H3122 cells, but also resulted in the attenuation of increased expression of SPP1 in A549 and H3122 cells (Figure 6(f,g)), indicating that miR-3163 exerts its function by targeting the 3’UTR SPP1. Therefore, we performed several functional assays to verify the effect of the miR-3163/SPP1 axis on LUAD cell malignant phenotypes. As shown in Figure 7(a), a high proliferative rate of LUAD cells was detected when SPP1 was overexpressed, whereas this effect was reversed by the addition of the miR-3163 mimic. Furthermore, the elevated migration and invasion of LUAD cells triggered by overexpression of SPP1 were abrogated by co-transfection with SPP1-OE and the mimic (Figure 7(b,c)). These results suggest that lncRNA AFAP1-AS1 weakens the anticancer activity of Met by sequestering miR-3163 and upregulating SPP1 expression.

Figure 6.

MiR-3163 regulated the sensitivity of met on LUAD cells by targeting SPP1. (a) Bioinformatics analyses were used to identify the complementary binding sites of miR-3163 in the 3ʹ‑UTR sequence of SPP1. (b) Wild‑type (SSPP1-wt) or mutant SPP1 (SPP1‑mut) reporter plasmids were co‑transfected into A549 and H3122 cells with miR‑3136 mimic or NC mimic. Luciferase activity was determined at 48 h following transfection. (c) RT‑qPCR analysis detected SPP1 mRNA expression in LUAD tissues and matched normal tissues. (d) Pearson’s correlation coefficient was used to determine the correlation between SPP1 and miR‑3136 expression in LUAD tissues. (e) RT‑qPCR analysis detected SPP1 mRNA expression in A549 and H3122 cells after treated by Met. (f and g) RT‑qPCR and Western blot analysis were employed to verify the effect of miR-3136 overexpression on the SPP1 mRNA level and protein expression.SPP1 overexpressing constructs (SPP1-OE) was co‑transfected with miR‑3163 mimic or NC mimic into A549 and H3122 cells. Subsequent alternation in the expression of SPP1 mRNA and protein were determined by RT‑qPCR and Western blot analysis, respectively. **P < 0.01 vs. NC mimic or mimic. *P < 0.05, **P < 0.01. CON, control; Met, Metformin; mimic, overexpression of miR-3163; OE-SPP1, overexpression of SPP1; OE-SPP1+ mimic, overexpression of miR-3163 and SPP1; empty vector+mimic-NC, negative control of OE-lnc and mimic.

Figure 7.

Overexpression of SPP1 offsets the regulatory effect of miR-3136 overexpression in LUAD cells. Cells were transfected with empty vectors, SPP1 overexpressing vectors (SPP1-Ov), SPP1-Ov+miR-3136 mimic, SPP1-Ov+mimic NC before Met treatment. (a) Cell proliferation was determined by CCK8. (b and c) migration and invasion assays were used to assess the migration and invasion capabilities, respectively of A549 and H3122 cells. *P < 0.05 and **P < 0.01 vs. NC inhibitor or empty vector group. *P < 0.05, **P < 0.01. CON, control; Met, Metformin; mimic, overexpression of miR-3163; OE-SPP1, overexpression of SPP1; OE-SPP1+ mimic, overexpression of miR-3163 and SPP1; empty vector+mimic-NC, negative control of OE-lnc and mimic.

As mentioned above, we further examined the phosphorylation status of the key effectors of the KI3K/AKT/mTOR signaling pathway. As shown in Figure 8, the activation of miR-3163 mimic reduced the phosphorylation levels of PI3K, AKT, and mTOR compared to cells transfected with an empty vector after Met treatment, indicating that miR-3163 overexpression could promote the anti-tumor activity. However, this inactivation of PI3K/AKT/mTOR signaling was completely nullified by SPP1 overexpression, as evidenced by the significantly increased phosphorylation levels of PI3K, AKT, and mTOR in SPP1-overexpressing cells.

Figure 8.

miR-3163/SPP1 reduced the anticancer activity of Met by inactivating PI3K/AKT/mTOR signaling pathway. A549 and H3122 cells were transfected with empty vectors, SPP1 overexpressing vectors (SPP1-Ov), SPP1-Ov+miR-3136 mimic, SPP1-Ov+mimic NC before Met treatment. Western blots were performed to assess the PI3K, p-PI3K, AKT, p-AKT, mTOR, p-TOR (a) in A549, (b) in H3122. *P < 0.05, **P < 0.01. CON, control; mimic, overexpression of miR-3163; OE-SPP1, overexpression of SPP1; OE-SPP1+ mimic, overexpression of miR-3163 and SPP1; empty vector+mimic-NC, negative control of OE-lnc and mimic.

Discussion

Recent experimental and clinical investigations have demonstrated the antineoplastic potential of Met [46-48]. However, its clinical significance remains limited. Therefore, it is critical to understand the mechanism underlying its anticancer activity. Consistent with a previous report, Met also displayed antineoplastic properties in LUAD in vitro. We further verified that this action of Met in LUAD is partially dependent on the lncRNA AFAP1-AS1/miR-3163/SPP1 axis. Our findings suggest that targeting the lncRNA AFAP1-AS1/miR-3163/SPP1 axis might be a novel approach for treatment of LUAD.

Recent studies have shown that the therapeutic benefits of Met in multifactorial diseases in vitro or in vivo may be lncRNA-dependent [49,50]. For example, Zeng et al. revealed that Met exerts antioxidant effects and alleviates ischemic stroke-induced oxidative stress injuries in vivo and in vitro by suppressing the expression of lncRNA-H19 [20]. Met curbs the lncRNA TUG1-modulated activation of vascular cell autophagy and might contribute to delaying the progression of atherosclerosis. Several lncRNAs are defined as Met-responsive lncRNAs in mice with nonalcoholic fatty liver disease [21]. Therefore, we analyzed the influence of Met administration on lncRNA AFAP1-AS1, which was found to be the most significantly upregulated lncRNA in LUAD tissues by GEPIA database analysis. Furthermore, the tumor-promoting role of lncRNA AFAP1-AS1 has been found in multiple cancers, including LUAD. Interestingly, our data showed that the expression of lncRNA AFAP1-AS1 was obviously reduced in Met-treated A549 and H3122 cells. To further analyze whether the anticancer activity of Met is AFAP1-AS1-dependent, we rescued the expression of AFAP1-AS1 and examined the alteration in LUAD cell behaviors after Met treatment. The ectopic expression of lncRNA AFAP1-AS1 could increase the suppressed malignant behavior of LUAD cells caused by Met treatment. Therefore, Met might be exerting its anti-tumor effects by inhibiting the pro-tumorigenic properties of AFAP1-AS1.

SPP1 is recognized as a multifunctional matricellular protein that influences diverse cell-biological properties, such as cell proliferation, adhesion, migration, and ECM accumulation. As a matricellular protein, SPP1 robust expression has been detected in various types of cancer, including lung cancer, and is associated with cancer progression [51-53]. Previous in vitro studies have demonstrated that SPP1 displays oncogenic activities in breast cancer, gastric cancer, glioma, and renal cell carcinoma. In LUAD, SPP1 proteolytic inactivation plays a putative role in combating LUAD malignancies [54]. Furthermore, silencing of SPP1 can significantly suppress the expression of programmed death ligand 1 in A549 cells and thereby negatively modulate macrophage polarization, resulting in the reversal of the immune escape of tumors [34]. Interestingly, SPP1 knockdown attenuates acquired resistance to drugs such as cisplatin and afatinib [55]. On the other hand, the pathologic potential of SPP1 in type 2 diabetes has also been investigated. For example, SPP1 deficiency can inhibit the expression of inflammatory factors in adipose tissue, resulting in improved sensitivity to insulin and glucose metabolism in HFD-induced obesity models [56]. SPP1-defecient mice showed diminished adipocyte hypertrophy, which is a predominant risk factor for diabetes [57]. Additionally, SPP1 and Met are both involved in the PI3K/AKT signaling event in lung cancer [45,58,59]. Therefore, we chose SPP1 for further investigations. Further bioinformatics analysis showed that miR-3163 is a target of SPP1 and lncRNA AFAP1-AS1. Consistent with a previous investigation in lung cancer cells, we found that miR-3163 was significantly downregulated in LUAD tissues. Moreover, the high expression of miR-3163 repressed oncogenicity in several LUAD cells. Luciferase reporter assays and RIP assays confirmed the lncRNA AFAP1-AS1/miR-3163/SPP1 axis, which was supported by the negative trend in the expression of lncRNA AFAP1-AS1 and miR-3163 in LUAD tissues. More importantly, miR-3163 mimic transfection significantly reduced the proliferative, metastatic, and invasive properties of A549 and H3211 cells resulting from lncRNA AFAP1-AS1 or SPP1 overexpression under Met exposure. In addition, we confirmed that the suppressed of PI3K/AKT/mTOR signaling caused by miR-316 overexpression was re-activated by SPP1 overexpression. These findings suggest that lncRNA AFAP1-AS1 functions as a competing endogenous RNA (ceRNA) of miR-3163 to upregulate SPP1 expression, which is involved in Met-treated LUAD in vitro.

Conclusion

In conclusion, Met exerts anti-cancer activities, at least in part, by regulating the AFAP1-AS1/miR-3163/SPP1 axis. Therefore, our findings may offer a new theoretical basis for the application of Met against LUAD.