Blocking MMP-12-modulated epithelial-mesenchymal transition by repurposing penfluridol restrains lung adenocarcinoma metastasis via uPA/uPAR/TGF-β/Akt pathway

Abstract

Purpose

Metastasis of lung adenocarcinoma (LADC) is a crucial factor determining patient survival. Repurposing of the antipsychotic agent penfluridol has been found to be effective in the inhibition of growth of various cancers. As yet, however, the anti-metastatic effect of penfluridol on LADC has rarely been investigated. Herein, we addressed the therapeutic potential of penfluridol on the invasion/metastasis of LADC cells harboring different epidermal growth factor receptor (EGFR) mutation statuses.

Methods

MTS viability, transwell migration and invasion, and tumor endothelium adhesion assays were employed to determine cytotoxic and anti-metastatic effects of penfluridol on LADC cells. Protease array, Western blot, immunohistochemistry (IHC), immunofluorescence (IF) staining, and expression knockdown by shRNA or exogenous overexpression by DNA plasmid transfection were performed to explore the underlying mechanisms, both in vitro and in vivo.

Results

We found that nontoxic concentrations of penfluridol reduced the migration, invasion and adhesion of LADC cells. Protease array screening identified matrix metalloproteinase-12 (MMP-12) as a potential target of penfluridol to modulate the motility and adhesion of LADC cells. In addition, we found that MMP-12 exhibited the most significantly adverse prognostic effect in LADC among 39 cancer types. Mechanistic investigations revealed that penfluridol inhibited the urokinase plasminogen activator (uPA)/uPA receptor/transforming growth factor-β/Akt axis to downregulate MMP-12 expression and, subsequently, reverse MMP-12-induced epithelial–mesenchymal transition (EMT). Subsequent analysis of clinical LADC samples revealed a positive correlation between MMP12 and mesenchymal-related gene expression levels. A lower survival rate was found in LADC patients with a SNAl1^high/MMP12^high profile compared to those with a SNAl1^low/MMP12^low profile.

Conclusions

Our results indicate that MMP-12 may serve as a useful biomarker for predicting LADC progression and as a promising penfluridol target for treating metastatic LADC.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13402-021-00620-1.

Introduction

Lung cancer is the most commonly diagnosed cancer and the leading cause of cancer-related death worldwide [1]. Non-small-cell lung cancer (NSCLC) comprises 85 % of all lung cancers, and lung adenocarcinomas (LADCs) and squamous cell carcinomas (SCCs) are its two major histotypes. Most patients diagnosed with lung cancer have advanced stage disease (stage IV) or develop distant metastasis, and the median survival time of this population is only 8–12 months [2]. Although tremendous efforts have recently been devoted to improving treatment strategies by combining novel chemotherapies with targeted agents, most patients ultimately do not respond effectively to such treatments due to the development of drug resistance [3]. Therefore, developing new targets and/or alternative approaches to current treatment modalities of advanced NSCLC is urgently needed.

A decisive hallmark in cancer progression is crossing of the extracellular matrix (ECM) and basement membrane (BM) by cancer cells. To penetrate the ECM, cancer cells secrete various proteolytic enzymes of the matrix metalloproteinase (MMP) family [4]. This family has more than 22 members. Currently, the clinical relevance of MMPs in NSCLC is still a matter of investigation. MMP-1, MMP-2, MMP-7, MMP-9 and MMP-10 are promising biomarkers for lung cancer, with MMP-2 and MMP-9 being the most widely studied MMPs in lung cancer. These two MMPs have been reported to promote locally invasive tumors, and their expression has been correlated with lung cancer invasiveness [5]. MMP-12 is a 22-kDa metal-dependent proteinase and has been detected in alveolar macrophages [6]. The role of MMP-12 is well documented in several lung diseases, including chronic obstructive pulmonary disease, emphysema and asthma, and recent studies have indicated that MMP-12 is critical to promoting the emphysema-to-LADC transition that is facilitated by inflammation and immune suppression [7, 8]. In addition to macrophages, MMP-12 has been found to be highly expressed in lung cancer tissues and to be correlated with local recurrence, advanced stage and lymph node metastasis in patients with NSCLC [9–11]. Therefore, MMP-12 may serve as an attractive therapeutic target for lung cancer.

Epithelial–mesenchymal transition (EMT) is a phenotypical alteration in which epithelial cells detach from their neighbors and the underlying BM, enabling cancer cells to acquire aggressive features such as invasiveness and the ability to metastasize [12]. MMPs have been reported to induce the EMT in different cancer types. MMP-3, MMP-7 and MMP-28 have, for example, been found to induce EMT in A549 LADC cells [13], but the role of MMP-12 in the EMT of LADC cells has remained unclear.

Drug repurposing is a promising strategy for overcoming “financial toxicity” for patients with anticancer therapy, because it may provide new therapeutic options for diseases at lower cost, higher safety, and within shorter periods of time before marketing the drug. A successful example of drug repurposing is thalidomide, a sleep-inducing agent that is used for multiple myeloma chemotherapy [14]. Anti-diabetic and nonsteroidal anti-inflammatory drugs are other successful examples of anticancer drug repurposing [15, 16]. Recently, several reports indicated that patients with schizophrenia who are treated with antipsychotic drugs have a lower risk of cancer, including lung cancer [17], which implies that antipsychotics may serve as repurposing drugs for cancer treatment. Penfluridol is a long-acting oral antipsychotic drug used for treating schizophrenia [18] and has recently been reported to exhibit anticancer activities in several cancer types in in vitro cell and in vivo animal models. Most studies indicated that penfluridol may exert antitumor effects by inhibiting tumor cell growth through induction of cell cycle arrest, apoptosis and autophagy, as well as by reducing the numbers of regulatory T cells and myeloid-derived suppressor cells [19, 20]. We have recently shown that penfluridol at cytotoxic concentrations (up to 5 µM) can suppress the growth of LADC cells through inducing autophagosome accumulation-mediated ATP energy deprivation [21]. Besides breast cancer and glioblastoma [22, 23], few studies have investigated the impact of penfluridol on the invasive and metastatic abilities of cancer cells. Here, we investigated the effects of penfluridol on the migration and invasion of several highly invasive LADC cell types and examined the possible underlying mechanisms.

Materials and methods

Reagents, chemical inhibitors, antibodies and plasmids

Penfluridol (P3371) and its solvent dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Penfluridol was dissolved in DMSO to prepare a stock solution of 40 mM. The Akt-specific inhibitor MK-2206 was obtained from MedChemExpress (Monmouth Junction, NJ, USA). Primary antibodies including those directed against Snail, Slug, and the phosphorylated (p-) or unphosphorylated forms of ERK1/2, JNK and p38 were obtained from Cell Signaling Technology (Danvers, MA, USA). Anti-vimentin and anti-integrin α6 antibodies were purchased from BD Biosciences (San Jose, CA, USA) and Abcam (Cambridge, MA, USA), respectively. Anti-E-cadherin, anti-N-cadherin, anti-MMP-12, anti-TGF-β, anti-uPA, anti-integrin α5 and anti-GAPDH antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Another anti-MMP-12 antibody was obtained from Millipore (Billerica, MA, USA). Protein marker (TD-PM10315) was obtained from BIOTOOLS Co. Ltd. (New Taipei City, Taiwan). Plasmids encoding myr-Akt, pLEX-Snail and pCIneo-Slug were kindly provided by Dr. C.C. Chen and Dr. T.C. Kuo (National Taiwan University, Taipei, Taiwan). Plasmid pLX304-MMP-12 was purchased from the DNASU plasmid repository (HsCD00436679; DNASU, Tempe, AZ, USA).

Cell lines and culture conditions

LADC cell lines (A549, H23, HCC827, PC9 and H1975) with wild-type EGFR or different EGFR mutation statuses were obtained from the American Type Culture Collection (ATCC; Manassas, VA, USA) or the National Cancer Center Hospital (Tokyo, Japan) and cultured in RPMI-1640 medium supplemented with 10 % fetal bovine serum (FBS) and a 1 % penicillin, streptomycin and glutamine mixture (all from Life Technologies). Human microvascular endothelial cells (HMEC-1) and lung epithelial cells (BEAS-2B) were obtained from the ATCC and cultured in Dulbecco’s Modified Eagle Medium (DMEM; Invitrogen, Carlsbad, CA, USA) supplemented with 10 % FBS and a 1 % penicillin, streptomycin and glutamine mixture. All cultured cells were maintained in a CO₂ incubator (37 °C, 5 % CO₂ and 95 % air atmosphere).

MTS cell viability assay

The effect of penfluridol on cell viability was measured using a MTS assay (Promega, Madison, WI, USA). LADC cells (5000 cells/well) were seeded in 96-well plates for 24 h and next treated with the indicated concentrations of penfluridol for another 24 h. At the end of the incubation period, 20 µl CellTiter 96 AQueous One Solution reagent that contained MTS was added to each well for 2 h at 37 °C. MTS is reduced by living cells to a colored formazan product and was detected at 490 nm absorbance using a microplate reader (MQX200; Bio-Tek Instruments, Winooski, VT, USA).

In vitro migration and invasion assays

Cell motility (migration or invasion) assays were performed using 24-well transwells (8-µm pore size; Corning Life Sciences, Tewksbury, MA, USA) as described previously [24]. Briefly, 3 × 10⁴ LADC cells were seeded in a noncoated top chamber for the migration assay or a Matrigel (BD Biosciences, Bedford, MA, USA)-coated top chamber for the invasion assay. Before cell seeding, cells were pretreated for 24 h with penfluridol (1.25 and 2.5 µM) and then plated in a top chamber that contained medium without serum. In the lower chamber, medium was added supplemented with 10 % serum as a chemoattractant for cells in the top chamber. Cells were allowed to migrate through or invade the membrane for 24 h, after which the migrated and invaded cells were fixed with methanol and stained with crystal violet (Sigma). The number of migrated and invaded cells was counted using a light microscope (100× or 200×; three random fields per well).

Cell adhesion assay

LADC cells were treated with 2.5 or 5 µM penfluridol for 1 h and then washed twice with phosphate-buffered saline (PBS). Next, 2.5 × 10⁴ cells were seeded in 24-well dishes and incubated for 30 min at 37 °C. The nonadherent cells were washed off using PBS, whereas the adherent cells were fixed with paraformaldehyde for 10 min at room temperature, after which they were rinsed with PBS and stained with 0.4 % crystal violet for 10 min. The number of adherent cells was counted under a light microscope.

Tumor endothelium adhesion assay

LADC A549 cells and HMEC-1 cells were respectively labeled with 1 µmol/L Cell Tracker Green CMFDA and Red CMTPX dye (Molecular Probes; Invitrogen, CA, USA) in serum-free RPMI-1640 and DMEM. After washing, the A549^Green cells (1 × 10⁵) were resuspended and incubated with vehicle or 5 µM penfluridol for 1 h. Next, 2.5 × 10⁴ cells were added to the HMEC-1^Red monolayers, grown in 24-well plates, and incubated in a CO₂ incubator at 37 °C for 30 min. Nonadherent cells were removed from the plate through gentle washing twice with PBS, after which the adherent cells were visualized and counted from three random fields using fluorescence microscopy at 100× magnification (Carl Zeiss Microimaging, Gottingen, Germany).

Human protease array assay

To identify ECM proteases through which penfluridol induces motility inhibition, we used a Proteome Profiler Human Protease Array (R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s instructions. LADC cells were treated with vehicle or 5 µM penfluridol for 24 h after which 200 µg cell lysate was extracted per array set comprising two nitrocellulose membranes with spotted capture antibodies. After adding the biotinylated detection antibody cocktail following horseradish-peroxidase-conjugated streptavidin, signal development was achieved through chemiluminescence detection using a multi-functions MultiGel-21 imaging system (TOP BIO, New Taipei City, Taiwan). Pixel densities of spots were analyzed using Image-Pro Plus. Spot densities were normalized against respective reference array spots and then against vehicle controls.

Extraction of RNA and RT-PCR

Total RNA was isolated from LADC cells using TRIzol reagent (Invitrogen, CA, USA) and amplified as described previously [25]. The primer sequences used for RT-PCR were as follows: Cathepsin B forward: ATA CCC CCA AGT GTA GCA AGA T and reverse: TGG TAC ACT CCT GAC TTG TAG A; Cathepsin L forward: AGG AAC CTC TGT TTT ATG AGG C and reverse: CTA CCA GAT TCT GCT CAC TCA G; Cathepsin C forward: CGA TGT CAA CTG CTC GGT TAT G and reverse: TTG TCT CGT TGC AGT AAG TGG T.

Protein extraction and Western blot analysis

The procedures for total protein lysate extraction and Western blot analysis have been described previously [26]. The primary antibodies used for Western blot analysis include those directed against p-Akt, Akt, p-p38, p38, p-ERK1/2, ERK1/2, p-JNK, JNK, TGF-β, uPA, MMP-12, E-cadherin, N-cadherin, vimentin, Snail, Slug integrin α5, integrin α6, GAPDH and β-actin.

Plasmid transfection

To identify possible regulatory targets of activated Akt, Snail and Slug, we exogenously overexpressed these proteins in LADC cells. Briefly, LADC cells were cultured in 6-cm petridishes till semiconfluency. Next, an empty or expression vector (3 µg) was transfected into the cells using Lipofectamine 3000 Transfection Reagent (Invitrogen, Carlsbad, CA, USA) for 6 h according to the manufacturer’s instructions. Cell lysates were harvested 24 h after transfection to check for Snail, Slug and p-Akt expression and their possible regulatory targets.

Lentiviral production and infection

293T packaging cells were transfected with 10 µg pLX304-MMP-12 or shMMP-12-expression plasmids together with 10 µg pCMVDR8.91 (the packaging vector) and 1 µg pMD.G (the envelope vector). The transfection medium was removed and replaced with fresh culture medium after 5 h of incubation. Another 48 h later, the lentivirus-containing medium was collected through centrifugation at 1500 rpm for 5 min. Next, LADC cells were infected with fresh lentivirus-containing medium (supplemented with 8 µg/ml polybrene) for 24 h and subjected to the indicated assays. MMP12 shRNA was obtained from the National RNAi Core Facility at Academic Sinica (Taipei, Taiwan).

Dot blot assay

Briefly, 10⁶ LADC cells were seeded in a 6-cm petridish overnight. Next, 3 µg of the indicated plasmid was transfected into the cells using Lipofectamine 3000 Transfection Reagent according to the manufacturer’s instructions. After 48 h of transfection, the cell culture medium was replaced by serum-free medium for another 24 h and harvested for dot blotting. Each dot presented the proteins in 300 µl soup. Hybridization and detection were performed according to the Western blot protocol.

Bioinformatics analysis

The survival Z-scores for MMP12 in various types of cancer including LADC were obtained from PRECOG (http://precog.stanford.edu). MMP7, MMP8 or MMP12 levels and survival data of LADC and SCC subjects were obtained from KM plotter (http://kmplot.com/analysis/index.php?p=background). MMP12 mRNA levels in normal lung and LADC tissues were obtained from TCGA and GEO (GSE10072) datasets. Clinical analysis of molecular expression using RNA sequencing in a cohort of patients with LADC was obtained from the TCGA University of California Santa Cruz Xena website (https://xenabrowser.net/). The prognostic significance of MMP12 and PLAUR, PLAU, PIK3CA, SNAl1 or CDH2 levels in patients with LADC was determined using Kaplan–Meier analysis. Correlations between MMP12 and TGFB1, PLAU, PLAUR, PDK1, PIK3CA, CDH2, FN1, SNAI1 or SNAI2 in LADC were evaluated using the cBioportal platform (https://www.cbiop ortal.org/).

Immunofluorescence (IF) microscopy

LADC cells were incubated with vehicle or penfluridol (1.25 µM) for 72 h, re-seeded on coverslips and fixed in 4 % paraformaldehyde, permeabilized, and stained with Alexa Fluor 594 Phalloidin (Thermo Fisher Scientific) to observe actin rearrangements in the cytoplasm. Cell nuclei were counterstained with 4′,6-diamino-2-phenylindole (DAPI) and visualized using a Zeiss Axiophot fluorescence microscope (Carl Zeiss Microimaging, Gottingen, Germany).

IHC staining

All primary tumor tissue samples were processed as described previously [21]. PBS was used for washing the slides before antibody incubation. Next, anti-E-cadherin, anti-N-cadherin and anti-MMP-12 antibodies were applied to the slides at a dilution of 1:100, incubated at 4℃ overnight, and washed twice with PBS. The slides were subsequently developed using a VECTASTAIN ABC (avidin-biotin complex) peroxidase kit (Vector Laboratories, Burlingame, CA) and a 3,3’-diaminobenzidine (DAB) peroxidase substrate kit (Vector Laboratories) according to the manufacturer’s instructions. Nuclei were counterstained with hematoxylin.

In vivo metastasis model

Age-matched male nonobese diabetic (NOD)-SCID mice were used for tumor metastasis assays in an experimental model. 10⁶ luciferase-tagged A549 cells (A549-Luc) were pretreated with vehicle or 2.5 µM penfluridol for 24 h and introduced into the circulation by injection in the tail vein using a 30-gauge needle. After 5 weeks, lung metastases were quantified using a Xenogen IVIS Spectrum system (Caliper; Xenogen). These experiments were performed in accordance with guidelines of the Institutional Animal Care and Use Committee (IACUC) of Wan Fang Hospital, Taipei Medical University (WAN-LAC-109-002).

Statistical analysis

Values are presented as mean ± standard deviation. Statistical analyses were performed using SPSS v20, and quantified data were analyzed using GraphPad Prism 7. Differences between two groups were analyzed using Student’s t-test and were considered significant at p < 0.05.

Results

Penfluridol attenuates the migratory, invasive and adhesive abilities of LADC cells with wild-type (WT) EGFR or different EGFR mutation statuses

To investigate the pharmacological potential of penfluridol against LADC, we first evaluated the effect of low concentrations of penfluridol with little or no cytotoxicity on the motility of LADC cells. A MTS assay was used to first determine the cytotoxic effects of penfluridol treatment for 24 h at various concentrations (0, 1.25, 2.5, 5, 10 and 20 µM) on four metastatic LADC cell lines, A549, H23, H1975 and HCC827, which harbor WT or mutant epidermal growth factor receptors (EGFRs). We found that 24-h treatment with penfluridol (1.25 and 2.5 µM) elicited very little or no cytotoxicity in both types of LADC cells tested (Fig. 1a). In addition, we found that penfluridol did not affect the viability of normal BEAS-2b lung epithelial cells at concentrations up to 2.5 µM (Fig. 1a). Thus, in all subsequent experiments we used this penfluridol concentration range. Next, the antimotility effect of penfluridol in A549 cells was tested using transwell migration and Matrigel invasion assays. We found that after treating these cells with 1.25 and 2.5 µM penfluridol for 24 h, their migratory and invasive abilities were suppressed in a concentration-dependent manner (Fig. 1b). The antimotility activity of penfluridol was also observed in LADC cells other than A549, i.e., HCC827 (EGFR exon 19 deletion) and H1975 (EGFR L858R/T790M mutations; Fig. S1). These results suggest that even nontoxic concentrations of penfluridol abolished the motility of these LADC cells. In addition to cell motility increases during the process of metastasis, circulating tumor cells must adhere to the endothelium in order to subsequently invade tissue [27]. Incubation of A549 cells with 2.5 or 5 µM penfluridol for 1 h significantly inhibited their ability to adhere to culture dishes (Fig. 1c) and to monolayers of vascular endothelial cells (Fig. 1d). This anti-adhesive ability of penfluridol was also observed in H1975 cells (Fig. S2). Penfluridol at a non- or low-cytotoxic concentration thus acts as an effective inhibitor of the motility and adhesion in LADC cells with different EGFR mutation statuses.

Fig. 1.

Effect of low concentrations of penfluridol on the migratory, invasive and adhesive abilities of lung adenocarcinoma (LADC) cells. a Four LADC cell lines, A549, H23, HCC827 and H1975, and one normal lung epithelial cell line, BEAS-2b, were treated with the indicated concentrations of penfluridol (1.25-20 µM) or dimethyl sulfoxide (vehicle control) for 24 h, after which a MTS assay was performed to determine cell viability. Data are presented as the mean ± standard deviation (SD) from three independent experiments. b Transwell migration and Matrigel invasion assays were performed with A549 cells which were treated with penfluridol at the indicated concentrations. Upper panel: representative photomicrographs (100×). Lower panel: data presented as the mean ± SD (n = 3). ** p < 0.01 and *** p < 0.001 compared with that of the vehicle group. c A549 cells were pretreated with the indicated concentrations of penfluridol for 30 min and subjected to an adhesion assay. Left panel: representative photographs of cell adhesion assay. Right panel: average adhered number of A549 cells pretreated with penfluridol or vehicle. ** p < 0.01 and *** p < 0.001 compared with those of the vehicle group. d A549 cells were prestained with Cell Tracker Green and pretreated with 5 µM penfluridol for 1 h. The labeled A549 cells were seeded onto a human microvascular endothelial cell monolayers prestained with Cell Tracker Red and cocultured for 30 min. After the nonadherent cells were removed, the remaining adherent cells were visualized as green bright dots and counted (left panel). Right panel: data shown as the mean ± SD (n = 3). ** p < 0.05 vs. vehicle group

Penfluridol induces protease alterations in LADC cells

Previous studies have shown that different MMPs are required for cancer cell migration, adhesion and EMT [28]. Therefore, we next evaluated the effect of penfluridol on alterations in 34 proteases using a human protease array and found that several of them, i.e., urokinase plasminogen activator (uPA), MMP-7, MMP-8, MMP-12, cathepsin B, cathepsin C, cathepsin V and cathepsin L, were downregulated in A549 cells after penfluridol treatment (Fig. 2a). Next, RT-PCR analysis was performed to confirm whether penfluridol can modulate cathepsins in A549 cells. We found, however, that the mRNA expressions of cathepsins B, L and C were only slightly inhibited or not affected at all by penfluridol (Fig. 2b). According to the protease array data, MMP-7, MMP-8 and MMP-12 were markedly downregulated after penfluridol treatment. Through Western blot analysis, we actually observed MMP-12 downregulation by penfluridol treatment in four LADC cell lines (A549, PC9, HCC827 and H1975) (Fig. 2c). This suggests that MMP-12 suppression by penfluridol generally occurs in LADC cells harboring WT or mutant EGFR. In clinical practice only high levels of MMP12, but not MMP7 or MMP8, are correlated with shorter overall survival times in patients with lung cancer. We also found that lung cancer patients with high MMP12 expression exhibited a shorter first progression-free survival than those with a low MMP12 expression (Fig. 2d). Next, we used a pancancer prognostic database, PREdiction of Clinical Outcomes from Genomic profiles (PRECOG; https://precog.stanford.edu/index.php) [29], to analyze the prognostic value of MMP-12. We found that MMP-12 has the potential to predict adverse outcomes in various human cancers, with the most significant prognostic effect in LADC (Z score = 4.29; Fig. 2e). Furthermore, in silico analysis of data in the Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) revealed significantly higher MMP12 transcript levels in LADC tumors compared with adjacent nontumor tissues (p < 0.0001, TCGA set) and normal lung tissues (p < 0.0001, GSE10072; Fig. 2f). Furthermore, using the Kaplan–Meier plotter (KM plotter) database, we found that high levels of MMP12 were correlated with a poor outcome in patients with LADC, even when they had negative surgical margins (Fig. 2g). The unfavorable prognostic role of MMP12 was not observed in SCC (Fig. 2g), which suggests that MMP-12 may be a LADC-specific prognostic factor.

Fig. 2.

Matrix metalloproteinase (MMP)-12 is downregulated by penfluridol in lung adenocarcinoma (LADC) cells and serves as a poor prognostic factor in patients with LADC. a Changes in the expression of proteases in A549 cell lysates following 24-h treatment with 2.5 µM penfluridol. Protein expression was determined using an antibody array (R&D Systems, Minneapolis, MN, USA) that contained 34 antibodies directed against proteases; n = 2 tests. Quantitative analysis of the protease array with a densitometer is shown as multiple of change of protease expression levels between the penfluridol- and vehicle-treated groups. b A549 cells were treated with 2.5 µM penfluridol for 8 h and then subjected to RT-PCR to analyze cathepsin mRNA expression. GAPDH was used as an internal control. c Four LADC cell lines, A549 (wild-type EGFR), H1975 (L858R/T790 M) and HCC827 and PC9 (del E746-A750) were subjected to 24 h penfluridol (2.5 µM) treatment, after which Western blot analysis was performed. d Correlation between MMP7, MMP8 and MMP12 expression and overall survival (OS) or first progression-free survival (FPS) in patients with lung cancer determined using a Kaplan–Meier plotter (KM plotter) database. Gene expression was dichotomized into high and low values using the median as a cut-off. HR, hazard ratio. e Pancancer expression of MMP12 from meta-Z analysis from the PREdiction of Clinical Outcomes from Genomic profiles (PRECOG) website. f Gene expression of MMP12 in the paired adjacent (TCGA dataset; right panel) and unpaired (GSE10072; left panel) normal and tumor tissues from patients with LADC. Statistical significance was analyzed using a paired t-test in the right panel and a t-test in the left panel. g Correlation between MMP12 expression and OS in LADC and squamous cell lung carcinomas using the KM plotter database

Penfluridol suppresses LADC cell motility by targeting MMP-12

To next determine whether MMP-12 modulates tumor cell motility and adhesion, we knocked down MMP12 expression in A549 cells. The knockdown efficiency of two specific shRNAs, detected by Western blot analysis, is shown in Fig. 3a. We found that after knocking down MMP12 in A549 cells, their migratory (Fig. 3b), invasive (Fig. 3c) and adhesive (Fig. 3d) abilities were significantly attenuated compared with those of control cells. Concordantly, we found that MMP12 overexpression significantly increased the invasive ability of A549 cells and reversed the inhibitory effect of penfluridol on the invasive ability of A549 cells (Fig. 3e). In clinical data obtained from the TCGA, we found that the MMP12 expression levels tend to correlate with lymph node metastasis in patients with LADC (p = 0.087) (Fig. 3f). These results suggest that MMP-12 plays a crucial role in the penfluridol-modulated invasive ability of LADC cells and may be correlated with advanced LADC progression.

Fig. 3.

Matrix metalloproteinase (MMP)-12 is a critical regulator in penfluridol-modulated motility of lung adenocarcinoma (LADC) cells. a-d A549 cells were infected with a lentivirus that carried specific MMP12 small hairpin (sh)RNAs or shGFP (shCtrl) and subsequently subjected to Western blot (a), transwell migration (b), Matrigel invasion (c) and adhesion (d) assays. Quantitative results were obtained by counting migrated, invaded and adhered cells in a 100× field. Multiples of differences are shown as mean ± standard deviation (SD) (n = 3). * p < 0.05, ** p < 0.01 and *** p < 0.001 vs. vehicle group. e Invasive ability of A549 cells infected with lentiviruses that carry a vector control (A549/Neo) or pLX304-MMP12 (A549/MMP12), followed by an additional 24 h treatment with penfluridol or vehicle. Quantitative results were obtained by counting invaded cells in a 200× field. Representative micrographs of the invasion assays are displayed on the left of the quantitative plot. Values are presented as mean ± SD of three independent experiments. ** p < 0.01 compared with that of the vehicle group. ^##p < 0.01 compared with that of the penfluridol-treated group. f MMP12 expression levels in LADC tissues from TCGA compared according to lymph node status (N stages)

Penfluridol inhibits MMP-12 expression by downregulating the uPA/TGF-β/Akt axis

To investigate the possible underlying mechanisms involved in the penfluridol-mediated inhibition of MMP-12 expression, we first dissected the interacting neighbors of MMP-12 using the STRING database (http://string-db.org/). We found that the protein–protein interaction networks for MMP12 and plasminogen activator, urokinase receptor (PLAUR), had the highest confidence (score = 0.959; Fig. 4a). Previously, it has been fund that PLAUR encodes the uPA receptor (uPAR), which binds and activates serine proteinase, uPA, and induces plasmin-mediated upregulation of activated forms of transforming growth factor-β (TGF-β) and MMPs, including MMP-12 [30, 31]. In various types of cancer cells, the PI3K/Akt pathway has been reported to be involved in the TGF-β-mediated expression of MMPs such as MMP-8 [32] and MMP-9 [33]. We found that treating A549 cells with low concentrations of penfluridol suppressed the expression of TGF-β and the active form of uPA, as well as phosphorylation of Akt in a concentration-dependent manner (Fig. 4b). The inhibitory effect of penfluridol on TGF-β and Akt was also observed in HCC827 or H1975 cells (Fig. S3). Next, a constitutively activated form of Akt, myr-Akt, was exogenously expressed in A549 cells to investigate whether Akt is important for regulating MMP-12 expression. Our data revealed that overexpressing activated Akt increased MMP-12 expression and secretion compared to control cells (Fig. 4c and d). Moreover, we found that myr-Akt-promoted MMP-12 secretion can also be observed in HCC827 cells (Fig. S4). Conversely, we tested whether inhibition of Akt activity affects MMP-12 expression. We found that treating A549 cells with the Akt-specific inhibitor MK-2206 concentration-dependently suppressed MMP-12 expression (Fig. 4e). Together, these data suggest that the ability of penfluridol to inhibit MMP-12 expression may be brought about through the uPA/TGF-β/Akt axis. In addition, we analyzed 566 primary human LADC samples retrieved from TCGA using the cBioportal platform and found that MMP-12 (MMP12) expression significantly correlated with TGF-β (TGFB1), uPAR (PLAUR), uPA (PLAU), PDK1(PDK1) and PI3K (PIK3CA) expression (Fig. 4f). Using TCGA data, we found that shorter survival times were observed in patients with LADC with high levels of MMP12 and high levels of PLAU, PLAUR or PIK3CA in tumors compared with patients with low levels of MMP12 and low levels of PLAU, PLAUR or PIK3CA in tumors (Fig. 4g). Together, these data indicate a considerable positive correlation between the uPA/uPAR/TGF-β/Akt axis and MMP-12 expression in patients with LADC, and that upregulation of MMP-12 and the uPA/uPAR/TGF-β/Akt axis may be critical events in promoting LADC progression.

Fig. 4.

Targeting the uPA/uPA receptor/TGF-β/Akt axis with a low concentration of penfluridol induces inhibition of matrix metalloproteinase (MMP)-12 in lung adenocarcinoma (LADC) cells. a MMP-12 protein–protein interaction network of 10 differentially expressed genes from the STRING database. b Levels of TGF-β, active uPA and phosphorylated (p)-Akt assessed using Western blot analysis after treatment with penfluridol (1.25 and 2.5 µM) for 24 h. Quantitative results of TGF-β, uPA and p-Akt protein levels were adjusted to β-actin and total Akt protein levels, respectively. c & e Western blot analysis of p-Akt and MMP-12 expression in A549 cells exogenously overexpressing myr-Akt (c) or treated with Akt inhibitor, MK-2206, for 24 h (e). Quantitative MMP-12 protein results were adjusted to that of β-actin. d Dot blot analysis of secreted MMP-12 using conditioned media of myr-Akt-transfected A549 cells as described in Materials and methods. f Correlation analysis of the TCGA LADC database (TCGA, PanCancer Atlas) using the cBioPortal revealing a positive correlation between MMP12 and the mRNA levels of the indicated genes. g Combination of high MMP12 expression with high PLAUR, PLAU or PIK3CA expression is correlated with poorer overall survival in patients with LADC compared to those with low MMP12 expression and low PLAUR, PLAU or PIK3CA expression. The LADC dataset was retrieved from TCGA

Penfluridol blocks EMT of LADC cells by targeting the Akt-MMP-12 axis

In addition to degradation of the ECM, EMT represents another key mechanism in tumor metastasis initiation and promotion [34]. We, therefore, evaluated the effect of penfluridol on EMT in LADC cells. After treating A549 cells with a low concentration of penfluridol (1.25 and 2.5 µM) for 24 h, the expression of the epithelial marker E-cadherin was found to be upregulated and the mesenchymal markers N-cadherin and vimentin to be downregulated, suggesting that penfluridol may also act as an EMT inhibitor (Fig. 5a). The molecular mechanism involved in penfluridol-mediated suppression of EMT was elucidated by investigating known EMT-promoting transcription factors, Snail and Slug. We found that the expression of Snail and Slug was also concentration-dependently suppressed by penfluridol in A549 cells, suggesting that their downregulation may be critical for penfluridol-mediated EMT suppression (Fig. 5a). After having undergone EMT, cancer cells become more fibroblast-like, which allows polarized assembly of the actin cytoskeleton into protrusive and invasive structures [35]. To determine whether penfluridol-suppressed EMT is accompanied by the acquisition of cell morphological changes, Alexa Fluor 594 phalloidin and DAPI were used to, respectively, stain F-actin filaments and nuclei. We found that A549 control cells displayed actin-spike protrusions in the form of filopodia, whereas the number of these protrusions decreased after penfluridol (1.25 µM) treatment for 72 h. (Fig. 5b). These morphological changes are characteristic of cells undergoing the reverse process of EMT. Next, we assessed E-cadherin, N-cadherin and MMP-12 expression by immunohistochemistry (IHC) in A549 orthotopic xenograft tumor tissues harvested from vehicle- or penfluridol-treated mice from our previous study [21]. Consistent with our in vitro findings, we found that the E-cadherin expression levels were increased and the N-cadherin and MMP-12 levels decreased in tumor tissues from penfluridol-treated mice (Fig. 5c and S5). Hematoxylin and eosin (H&E) staining was used for routine pathological evaluation of the tumor tissues (Fig. 5c). To rule out the possibility that the anti-metastastic effect of penfluridol in the orthotopic xenograft mouse model was due to penfluridol-mediated growth inhibition of the primary tumor, we further investigated the anti-metastatic effect of penfluridol in an experimental metastasis model using a bioluminescence system. To this end, A549-Luc cells were pretreated with 2.5 µM penfluridol or vehicle for 24 h and then injected into the lateral tail vein of NOD-SCID mice. Five weeks postinjection, penfluridol-treated cells exhibited a relatively lower lung colony forming capacity, as revealed by bioluminescence imaging analysis (Fig. S6). Previous studies indicated that the PI3K/Akt signaling pathway is involved in regulating Snail-mediated EMT in liver cancers [36]. Here, we found that overexpression of activated Akt significantly increased the expression of Snail and Slug in A549 and H1975 cells compared with that in control cells (Fig. 5d and S6), which suggests that PI3K/Akt signaling can trigger EMT in LADC cells. Our results indicate that PI3K/Akt signaling activation is crucial for MMP-12 expression and EMT progression in LADC cells. Previous studies have reported that overexpression of several MMPs (MMP-2, MMP-3, MMP-9, MMP-13 and MMP-14) is associated with EMT [28]. Hence, we investigated whether MMP-12 modulates EMT progression in LADC cells. After overexpressing MMP-12 in A549 cells, we noted that the expression of N-cadherin, Snail, and Slug was significantly enhanced compared with that in control cells (Fig. 5e). To further confirm that MMP-12 is critical for activated Akt-modulated EMT, we found that transfection and expression of MMP-12 shRNA dramatically reversed activated Akt-induced upregulation of Snail and Slug in A549 cells, suggesting that PI3K/Akt signaling triggers EMT at least partially through MMP-12 (Fig. 5f). From the same aforementioned TCGA dataset, we found that MMP12 (MMP-12) expression in LADC significantly correlated with the expression of several mesenchymal markers including CDH2 (N-cadherin), FN1 (fibronectin), SNAl1 (Snail) and SNAl2 (Slug; Fig. 5g). Moreover, we found that LADC patients with MMP12^high/SNAl1^high or MMP12^high/CDH2^high exhibited shorter survival times than those with MMP12^low/SNAl1^low or MMP12^low/CDH2^low (Fig. 5h). Taken together, these results indicate that the Akt-MMP-12 axis may be vital in mediating EMT progression in LADC cells, and that penfluridol harbors the potential to block this pathway. Clinical data additionally revealed that MMP-12-induced EMT is associated with a poor prognosis in patients with LADC.

Fig. 5.

Penfluridol suppresses MMP-12-modulated EMT progression in lung adenocarcinoma (LADC) cells. a A549 cells were treated with vehicle or penfluridol (1.25 and 2.5 µM) for 24 h, after which the expression of epithelial and mesenchymal markers was detected using Western blot analysis. b After treating A549 cells with vehicle or penfluridol (1.25 µM) for 72 h, cells were fixed and stained for F-actin using Alexa Fluor 594 Phalloidin (red). Nuclei were counterstained with DAPI (blue). White arrow: actin-spike protrusions (filopodia). Original magnification, 400x. Scare bar, 10 μm. c H&E staining and IHC analysis of E-cadherin, N-cadherin and MMP-12 expression in A549 orthotopic tumors treated with vehicle or penfluridol (5 mg/kg). Original magnification, 400x. Scare bar, 30 μm. d & e A549 cells were transiently transfected with a vector control (A549/Neo) or myr-Akt (A549/myr-Akt) (d) or infected with a lentivirus that carried a control vector or MMP-12-V5 (e). f A myr-Akt-expressing plasmid was transfected into A549 cells expressing shCtrl or shMMP-12 as indicated. Expression of the indicated EMT-related proteins from d-f was determined through Western blot analysis. The extent of EMT-related protein and β-actin expression from a, d, e and f were determined using a densitometer and Image-Pro Plus. Quantitative results of EMT-related protein levels were adjusted to that of β-actin. g Correlation analysis of the TCGA LADC databases (TCGA, PanCancer Atlas) using the cBioPortal revealing a positive correlation between MMP12 expression and the EMT-related genes CDH2, FN1, SNAl1 and SNAl2. h Combined high MMP12 expression with high SNAl1 or CDH2 expression correlated with a poorer overall survival in patients with LADC compared with patients with a low MMP12 expression combined with a low SNAl1 or CDH2 expression. The LADC dataset was retrieved from TCGA

Discussion

Accumulating evidence indicates that MMPs play critical roles in the tumor microenvironment to induce changes during EMT and to facilitate invasive and metastatic behavior [28]. Several MMPs have been reported to induce EMT and to increase the invasive ability of liver cancer cells, including MMP-1, MMP-2, MMP-3, MMP-7, MMP-8 and MMP-9 [37]. In lung cancer cells, MMP-3, MMP-7 and MMP-28 have been reported to induce EMT through multiple signaling pathways. MMP-28 can, for example, induce EMT by upregulating active TGF-β, MMP-9 and MT1-MMP [38]. Overexpression of MMP-3 has been found to cause EMT by inducing E-cadherin degradation and Rac1b upregulation in lung cancers [13]. Until now, limited information is available regarding the potential role of MMP-12 in the EMT of cancer cells. Our current study indicates that overexpression of MMP-12 induces a Snail-family-mediated EMT in A549 LADC cells and enhances the invasive ability of these cells. These results indicate a novel role of MMP-12 in regulating the phenotypic change of LADC cells. MMP-12 has been reported to activate other MMPs, including MMP-2 and MMP-3 [37]. Therefore, we suggest that upregulation of active TGF-β, MMP-2, MMP-3 or Rac1b may participate in the process of MMP-12-induced EMT, and that this possibility should be investigated in the future. Clinically, we found that in patients with LADC, but not in those with SCC, MMP-12 serves as a poor prognostic factor and correlates with lymph node metastasis. Similarly, previous studies have indicated that MMP-12 expression may be correlated with local recurrence and tumor metastasis in patients with NSCLC [9]. Moreover, we found that MMP-12 expression significantly correlated with expression of the mesenchymal-related markers N-cadherin (CDH2), fibronectin (FN1), Snail (SNAl1) and Slug (SNAl2) in 566 LADC samples retrieved from TCGA. MMP-12 may thus be a specific biomarker for forecasting EMT-regulated invasion of LADC cells and be a valuable therapeutic target for treating LADC patients.

Missing or low MMP levels were observed in healthy tissues, whereas upregulation of MMPs was observed in lung cancer tissues. This phenomenon supports the concept of adjuvant therapy with MMP inhibitors in lung cancer. At present, several MMP inhibitors have been developed and are being applied in clinical trials. However, most of these trials were stopped during phase III because of poor solubility, low oral bioavailability and/or dose-limiting toxicity [39]. Drug repurposing is a promising strategy for discovering new MMP inhibitors since preclinical safety profiles of these drugs have already been established. The antibiotic doxycycline (Periostat) was, for example, recently approved by the US Food and Drug Administration (FDA) as a MMP inhibitor for treating periodontal disease using a mechanism unrelated to its antimicrobial activity [40]. In the present study, we noted that a nontoxic concentration of the antipsychotic drug penfluridol elicited an anti-metastatic effect in LADC cells by targeting MMP-12-induced EMT. Our data revealed that penfluridol significantly suppressed the invasive and migratory abilities of various LADC cells harboring wild-type or mutant EGFR. Also, adhesion of cancer cells to the vascular endothelium and their interaction are crucial steps during metastasis and invasion. We found that penfluridol remarkably inhibited LADC cell adhesion to endothelial cells.

The PI3K/Akt signaling pathway is known to regulate the expression of Snail family members by promoting the GSK-3β-mediated stabilization of Snail and Slug [41] or nuclear factor-κB-mediated transcription of Snail [42]. Our current results indicate that overexpression of activated Akt can upregulate MMP-12, Snail and Slug expressions in A549 cells. Conversely, we found that treating A549 cells with Akt-specific inhibitors suppressed MMP-12 expression, and that overexpression of MMP-12 induced upregulation of Snail and Slug. Moreover, Akt activation was found to be suppressed using penfluridol treatment of A549 and H1975 cells, suggesting that penfluridol suppresses EMT of LADC cells by targeting the Akt-MMP-12-Snail/Slug pathway. In addition to the MMP-12-modulated expression of Snail family members, Snail has been reported to induce upregulation of several MMPs in different cancer types. For example, Snail overexpression has been shown to induce upregulation of MMP-1, MMP-2, MMP-7 and MT1-MMP in liver cancers [43] and MMP-2 in skin cancers [44]. Our previous work indicated that Snail family members exhibited positive feedback mechanisms to activate Akt through transcriptional inhibition of the Akt deactivators maspin and phosphatase and tensin homolog (PTEN) in A549 cells [45, 46]. Snail/Slug may feedback-regulate MMP-12 expression by activating Akt in LADC cells, and we actually observed that overexpression of both Snail or Slug can promote secretion of MMP-12 in HCC827 LADC cells (Fig. S4). However, the functional roles of maspin and PTEN in penfluridol-regulated MMP-12 expression require further study.

In addition to targeting the interplay of p-Akt and Snail/Slug, which attenuated MMP-12-mediated EMT by penfluridol, MMP-12 expression has been found to be suppressed by the natural product oridonin in LADC cells by downregulating the EGFR/extracellular-signal-regulated kinase (ERK)/AP-1 pathway and upregulating protein phosphatase 2 A (PP2A) activity [47]. AP-1 is a critical transcription factor that activates MMP-12 transcription [48] and is regulated by mitogen-activated protein kinases (MAPKs) such as ERK and c-Jun N-terminal kinase (JNK). Herein, we observed that penfluridol treatment attenuated the activity of JNK, but not ERK, in A549 and HCC827 cells (Fig. S8), which suggests that targeting of JNK/AP-1 by penfluridol may be another way to regulating MMP-12. TGF-β is known to be a crucial upstream inducer stimulating Akt-mediated EMT in cancer. Moreover, it is known that TGF-β can regulate uPA expression in cancer cells, and that uPA can activate latent TGF-β, thereby producing a pernicious cycle that contributes to tumor progression, including EMT [30]. Here, we discovered that penfluridol treatment can reduce the expression levels of TGF-β and active uPA in A549 or HCC 827 cells, suggesting that the uPA system-mediated TGF-β production that triggers the Akt-MMP-12-Snail/Slug pathway is a major target of penfluridol to prevent EMT-induced invasive abilities of LADC cells. It has been reported that TGF-β can promote CD44/EGFR expression and colocalization and, subsequently, induce Akt-mediated EMT in A549 cells [49]. We suggest that the role of CD44/EGFR in penfluridol-mediated suppression of Akt-MMP-12 and EMT progression should be investigated in more detail in the future.

PP2A is a serine/threonine phosphatase that deactivates multiple components of growth- and metastasis-related signaling pathways, such as Akt and MAPKs [50]. Previously, we found that penfluridol may serve as a PP2A activator to suppress Akt and JNK activation and induce apoptosis of leukemic cells [51]. Whether penfluridol can also trigger PP2A activity to negatively regulate Akt and JNK and suppress MMP-12-mediated EMT in LADC should be further verified.

Conclusions

Our previous study has shown that penfluridol exerts anti-metastatic effects in an A549 LADC orthotopic xenograft model [21]. Our current study revealed that MMP-12 is highly expressed in LADC tumors but not in normal lung tissues and that MMP-12 serves as a poor prognostic factor in LADC, but not in other NSCLC subtypes. Functionally, MMP-12 can enhance cell migratory and invasive abilities by inducing Snail/Slug-mediated EMT progression of LADC cells, and penfluridol at nontoxic concentrations was identified as a MMP-12 inhibitor that reduces the motility of LADC cells. Mechanistic investigations revealed that Akt activation can induce MMP-12 expression and, subsequently, trigger Snail/Slug-mediated EMT. Penfluridol treatment can suppress the activation of Akt and expression of its upstream regulators, uPA/uPAR/TGF-β, to further downregulate MMP-12. Moreover, our previous study indicated that upregulation of Snail/Slug triggered by active Akt showed positive feedback mechanisms to strengthen Akt activity in A549 LADC cells [45]. Therefore, we suggest that MMP-12-induced upregulation of Snail family members may induce feedback activation of Akt and further enhance MMP-12 expression in LADC cells. The mechanisms deduced from our current and previous studies are schematically illustrated in Fig. 6. Similar to our study, Kim et al. indicated that penfluridol at 2 µM can inhibit the migration and invasion of glioma sphere-forming cells by reducing the expressions of integrin α6 and uPAR and EMT progression [23]. Integrins α6 and α5 have also been reported to serve as targets of penfluridol in breast cancer cells [52]. Here, we found that penfluridol can suppress integrin α5 and α6 expression in A549 and H1975 cells (Fig. S9), but the role of these integrins in penfluridol-regulated LADC cell motility requires further investigation. The results from our present and previous studies [21] support the concept of repurposing penfluridol as a drug for treating LADC metastasis, particularly in cases with a high expression of MMP-12.

Fig. 6.

Schematic overview of the putative molecular mechanisms underlying penfluridol-mediated inhibition of the metastasis of lung adenocarcinoma (LADC) cells. Penfluridol-mediated suppression of the metastasis of LADC cells is caused by attenuation of MMP-12-mediated Snail/Slug expression, which results in an ultimate restrain of EMT progression and subsequent suppression of the metastasis of LADC cells. The uPA/uPA receptor/TGF-β/Akt axis is critical for penfluridol-modulated MMP-12 expression in LADC cells. The dashed gray line indicates a hypothetical pathway that may be regulated by penfluridol

Abbreviations

BM

basement membrane

ECM

extracellular matrix

EMT

epithelial–mesenchymal transition

GEO

Gene Expression Omnibus

KM plotter

Kaplan–Meier plotter

LADC

lung adenocarcinoma

MMP

matrix metalloproteinase

NSCLC

non-small cell lung cancer

SCC

squamous cell carcinoma

TCGA

The Cancer Genome Atlas

TGF-β

transforming growth factor-β

uPA

urokinase plasminogen activator

uPAR

uPA receptor

WT

wild-type

Author contributions

Conceptualization, M-HC, J-HC and W-YH; methodology, W-JL, G-ZC, C-HT, Y-CY, T-CL and J-QC; writing, original draft preparation, M-HC, J-HC and C-LC; writing, review and editing, M-HC and W-YH. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the Taipei Medical University Research Center of Cancer Translational Medicine from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education in Taiwan (to M.-H. Chien). This study was also supported by a grant (110-eva-14) from Wan Fang Hospital, Taipei Medical University (to J.-H. Chang).

Data availability

All data generated or analyzed during this study are included in this article and its supplemental files.

Declarations

Conflict of interest

The authors declare that no competing interests exist.

Ethics approval and consent to participate

All animal experiments were carried out in accordance with guidelines of a protocol approved by the Taipei Medical University Animal Ethics Research Board.

Consent for publication

Not applicable.