Pseudomonas aeruginosa induces spatio-temporal secretion of IL-1β, TNFα, proMMP-9, and reduction of epithelial E-cadherin in human alveolar epithelial type II (A549) cells

1Department of Inmunobioquímica, Instituto Nacional de Perinatología “Isidro Espinoza de los Reyes” (INPerIER), Ciudad de México, México; 2Department of Inmunología e Infectología, INPerIER, Ciudad de México, México; 3Department of Salud Sexual y Reproductiva, INPerIER, Ciudad de México, México; 4Department of Fisiología y Desarrollo Celular, INPerIER, Ciudad de México, México; 5Department of Bioquimica, Escuela de Medicina, UNAM, Ciudad de México, México; 6Department of Patología Experimental, Hospital Infantil de México “Federico Gómez”, Ciudad de México, México; 7Laboratory of Patogenicidad Bacteriana, Unidad de Hemato-Oncología e Investigación, Hospital Infantil de México “Federico Gómez”, Ciudad de México, México; 8Department of Pharmacology, Emory University School of Medicine, Atlanta, GA 30322, USA


INTRODUCTION
Pseudomonas aeruginosa is an opportunistic Gram-negative bacterium that has been associated with chronic infections in airways (Beaudoin et al., 2012), cystic fibrosis (Holm et al., 2013) and pulmonary inflammation (Park et al., 2013). The pathogenicity of P. aeruginosa is mediated by several factors, including the production of diffusible molecules controlled by a mechanism known as quorum sensing (Chugani et al., 2012;Kownatzki et al., 1987;Perez et al., 2013;Rada and Leto, 2013). It was shown that lipopolysaccharides of P. aeruginosa induce in the alveolar and bronchial epithelium the secretion of nitric oxide (Pitt & St Croix, 2002), inflammatory cytokines (Wong & Johnson, 2013) and production of matrix metalloproteinases (MMPs) (Frisdal et al., 2001;Okamoto et al., 2002;Yao et al., 1996). MMPs are a family of zinc neutral endopeptidases produced in several pathological conditions (Churg et al., 2007;Holm et al., 2013) by a wide variety of cell types, including neutrophils (Bradley et al., 2012;Louhelainen et al., 2010), alveolar macrophages (Churg et al., 2007), and bronchial epithelial cells (Yao et al., 1996). MMPs induce degradations of various structural components of the extracellular matrix including collagen type I, IV, V, VII, X, fibronectin, elastin, proteoglycan (Woessner, 1991), basement membrane (Kargozaran et al., 2007) as well as cell-binding adhesion proteins (Allport et al., 2002;Nawrocki-Raby et al., 2003). Although the secretion of MMPs is well known in various lung diseases: bronchopulmonary dysplasia (Mizikova & Morty, 2015), adenocarcinomas (Canete-Soler et al., 1994), and chronic obstructive pulmonary disease (Louhelainen et al., 2010), the secretion profile of proMMP-2 and -9 produced by human pneumocytes secretory type II cells during infection with Pseudomonas aeruginosa is unknown. We chose the A549 cell line as it is a model of human lung alveolar epithelium which plays an important role in the immune response. We hypothesized that an increase in IL-1β and TNFα concentrations would be accompanied by a parallel increase Vol. 68, No 2/2021 207-215 https://doi.org/10. 18388/abp.2020_5509 in collagenolytic activity of MMP-2 and -9 in the culture medium, and thereby would induced changes in epithelial cadherin (E-cadherin) in A549 cells during transient P. aeruginosa stimulation.

Cell lines and culture
A549 cell line (American Type Culture Collections, Rockville, MD, USA) was obtained and its genetic profile corroborated by the amplification of 21 specific markers. The result showed a complete match with the A549 line (ATTC, CCL-185). A549 cells were cultivated on 12 well plates (Corning, Darmstadt, Germany) in RPMI 1640 medium (Roswell Park Memorial Institute; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS), an antibiotic-antimycotic solution (penicillin 100 U/mL, streptomycin 100 µg/mL; Gibco) and incubated at 37°C in 5% CO 2 . After reaching 95% of confluence, A549 cells were washed twice with sterile saline solution to remove RPMI-FBS, and 1 mL of RPMI with 0.2% lactoalbumin hydrolyzated (RP-MI-LHA; Gibco) was added with subsequent incubation at 37°C in 5% CO 2 .

Cell stimulation
After reaching 95% confluence, A549 cells were washed twice with sterile saline solution to remove RP-MI-FBS and 1 mL of RPMI with 0.2% lactoalbumin hydrolysate (RPMI-LHA; Gibco) was added before incubation at 37°C in 5% CO 2 . Next A549 cells were infected with live P. aeruginosa in serial dilutions (10 2 , 10 4 , 10 5 , and 10 6 colony-forming units (CFU/mL)). The CFU numbers were based on a turbidity equivalent to 0.5 Mc-Farland standard. After the infection, A549 cells were cultured for 3, 6 or 24 hours. At the end of the incuba-tion time, the medium was collected and samples were centrifugated at 1400 rpm at 4°C for 5 min, the supernatants were collected and stored at -70°C until further processing.

Cell viability assay
To evaluate A549 cells viability after incubation with P. aeruginosa we used the colorimetric assay of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MMT) as previously described by Zeng et al. (2017). Cells were washed twice with sterile saline solution to remove RPMI-LHA and P. aeruginosa, and then cultured for 3 hours in presence of 20 µl (5 mg/mL) of MMT in 5% CO 2 at 37°C. Subsequently, 150 µl of Dimethyl sulfoxide (DMSO; Merck KGaA, Darmstadt, Germany) was added into each well (Zeng et al., 2017). For negative control, a mitochondrial uncoupler carbonyl cyanide m-chlorophenylhydrazone (CCCP) was dissolved in dimethylsulfoxide at a concentration of 80 µM (Chaudhari et al., 2008) and added to the cells before the incubation at 37°C with 5% CO 2 , 95 % air. Blue formazan product in the culture medium from A549 cells was analyzed by spectrophotometric absorbance reading at 570 nm in Benchmark microplate (model 550; BioRad. Hercules, CA, USA). Five independent experiments were performed, each in duplicate.

Measurement of proinflammatory cytokines
To quantify IL-1β and TNFα secreted to the culture medium of A549 cells after each period of incubation with P. aeruginosa we used a specific DouSet enzymelinked immunosorbent assay (ELISA) (R&D Systems, Minneapolis, MN, USA) following the manufacturer´s instructions. This procedure was previously reported by our research group (Flores-Herrera et al., 2012;Osorio-Caballero et al., 2015). For IL-1β (DY201; R&D Systems) and TNF (DY210; R&D Systems), a standard curve was created from 4 to 260 pg/mL and 15 to 960 pg/mL, with a sensitivity of 2.0 and 5.0 pg/mL, respectively. The concentration of IL-1β and TNFα were expressed as pg/mL. The ELISA assay was performed in eight independent experiments.

Zymography gel activity
To evaluate the secretion of proMMP-2 and proMMP-9 into the culture media of A549 cells, SDS-polyacrylamide gels with porcine gelatin (1 mg/mL) were used as described previously (Flores-Herrera et al., 2012). Each well was loaded with 0.75 µg of protein and the activity band was determined by optical density using NIH ImageJ. We used a culture medium from, promyelocyte cells as a control of electrophoretic mobility (U937, ATCC, CRL-1593.2; Manassas, VA, USA). The gel activity assay was performed in eight independent experiments.

Inmunodetection of MMP-9 and E-cadherin in the A549 cells
To localize MMP-9 in A549 cells after infection with P. aeruginosa we used immunefluorescence as described previously (Flores-Herrera et al., 2012). After fixing the cells with 4% paraformaldehyde for 10 minutes, a primary mouse anti-MMP9 antibody (clone 56-2A4; Calbiochem Darmstadt, Germany) was added at 1:50 dilution. An appropriate fluorescent-labeled secondary antibody (Molecular Probes, USA) was used. The nucleus was stained with 1 ng/ml of Hoechst 33258 (Sigma-Aldrich). In another set of experiments, E-cadherin was immu-nodetected using mouse anti-human E-cadherin (clone NCH-38) antibody at a 1:100 dilution. The nucleus was stained using 4',6-diamidino-2-phenylindole (DAPI) for 7 minutes. Negative controls consisted of cells without primary antibody, and, as expected, they did not exhibit any staining (not shown). The immunostaining was analyzed using an epi-fluorescence microscope (Olympus, IX-81, Tokyo, Japan) and photographed with a CCD camera (Hamamatsu, ORCA-Flash 2.8, Tokio, Japan).

Statistical analysis
Data were analyzed by one-way ANOVA with multiple comparisons followed by Tukey´s test using Sig-maPlot version 11.0 (San Jose, CA, USA). Results are expressed as mean ± S.E.M. p<0.05 was considered significant. Immunostainings of proMMP-9 and E-cadherine were performed five times. Figure 1 shows the viability of A549 cells with and without P. aeruginosa stimulation after 3 (1A), 6 (1B) and 24 hours (1C). The viability was not affected by the different doses of P. aeruginosa when compared to the control group (p=0.65). In the same experiments, we included the mitochondrial inhibitor (CCCP), which significantly reduced the viability of A549 cells in comparison to the control group (p<0.05). Finally, MMT was not metabolized by P. aeruginosa (Fig. 1). These experiments demonstrated that infection with P. aeruginosa did not affect the viability of A549 cells. We then assessed the effect of P. aeruginosa on the secretion of IL-1β and TNFα.
TNFα Figure 3 shows that stimulation of A549 cells with P. aeruginosa increased the secretion profile of TNFα in a dose-dependent manner. After 3 hours of stimulation with P. aeruginosa at 10 4 , 10 5 , and 10 6 CFU/mL, A549 cells significantly increased the secretion of TNFα by 1.4-, 1.5-, and 1.6-fold, respectively when compared with the control (6.4±0.4; p≤0.05, Fig. 3A). A similar secre- Effect of different number of colony-forming units (CFU/mL) of Pseudomonas aeruginosa at 3 (A), 6 (B), and 24-hours (C) of stimulated of A549 cell (ashurated bars), the viability was determined with MMT assay. We included two negative controls: carbonyl cyanide m-chlorophenylhydrazone as mitochondrial inhibitor incubated with A549 cells (CCCP, 80 µM; black bar) and Pseudomonas aeruginosa (Pa). The assay was performed in five independent experiments with duplicates. Data represent the mean ± standar deviation. Statistically significant difference *p<0.05 vs. control group.

Figure 2. Secretion of IL-1β by A549 cells stimulated with P. aeruginosa.
After 3 (A), 6 (B), and 24 hours (C) of stimulation with or without P. aeruginosa (differentnumber of colony-forming units; CFU/mL), the culture medium of A549 cells was recovered and analyzed using ELISA. The concentration of IL-1β was expressed as pg/mL. The assay was performed in 8 independent experiments with duplicates. Data represent the mean ± standard deviation. Statistically significant difference *p≤0.05 vs. control group.
Interestingly, it was shown that IL-1β and TNFα induce the secretion of MMPs (Roomi et al., 2013). Therefore, our next step was to determine the effect of the inflammatory responses induced by P. aeruginosa on the secretion of extracellular matrix metalloproteases into the culture medium from A549 cells. Lysis bands for proMMP-2 and -9 were identified by taking the mobility of U937 standard as a reference point, as previously reported and validated by our research group (Flores-Herrera et al., 2012). Figure 4 shows the lysis bands of proMMP-2 and -9 secreted by A549 cells after stimulation with P. aeruginosa for 3 (4A), 6 (4E), and 24 hours (4I). The relative densitometric analysis indicated that after 3 (Fig. 4B) and 6 hours (Fig. 4F) of stimulation, significantly higher levels of proMMP-2 were detected when compared to the control. Maximal secretion of proMMP-2 was observed after 24 hours of stimulation with P. aeruginosa at 10 5 , and 10 6 CFU/mL, with a 1.3-fold increase compared to the control (48.7±2.8; p≤0.05 Fig. 4J). After the same period of stimulation, we observed a band of 66-KDa that corresponded to the MMP-2 active form (Fig. 4J).
Consistent with these findings, we observed morphological changes in A549 cells characterized by an increase in the number of spherical cells (Fig. 4L), when compared to the control group (Fig. 4K). This finding, together with the absence of the proMMP-9 band in activity gels, suggested that this enzyme can be located in the extracellular matrix of A549 cells, as previously reported under other pathological conditions (Flores-Herrera et al., 2012;Nawrocki-Raby et al., 2003). To explore this hypothesis, we performed immunolocalization with specific antibodies.

proMMP-9 detection in A549 cells by immunofluorescence
As shown in Fig. 5, proMMP-9 was immunodetected in the extracellular matrix of A549 cells after stimulation with 10 6 CFU/mL of P. aeruginosa. We observed a significant increase in immunoreactivity after 3, 6 and 24 hours compared to the respective controls (Fig. 5). As it was previously demonstrated in another cellular system, the active isoform of MMP-9 is able to degrade different support components, including collagen type I, IV, V, XI, elastin, and proteoglycan of the extracellular matrix (Morrison et al., 2009;Woessner, 1991), as well as cellbinding proteins such as vascular endothelial-cadherin (Allport et al., 2002) and E-cadherin (Nawrocki-Raby et al., 2003). After observing a change in the morphology of A549 cells, a reduction in the number of adhered cells (data not shown), and a decrease of proMMP-9 immunoreactivity, we complemented our approach by analyzing E-cadherin using immunodetection.

A549 cells incubated for 24 hours with P. aeruginosa
showed very low immunostaining intensity for E-cadherin compared to the respective controls (Fig. 6). Immunoreactivity was located around the cells and the nuclei.

DISCUSSION
Several in vivo and in vitro models of infection are able to release a diverse set of molecules that are associated with cellular stress (Osorio-Caballero et al., 2015), and the reduction of chemotactic (Henriquez et al., 2015) and proinflammatory cytokines (Keyel, 2014;van de Veerdonk et al., 2011), which are involved in the next phase of the inflammatory response through the secretion of degradatives enzymes, such as proMMPs (Flores-Herrera et al., 2012). In in vitro models, the induction of the degradative response affects cell integrity by decreasing the expression of cell-cell adhesion proteins, like E-cadherin and vascular endothelial-cadherin (Allport et al., 2002;Nawrocki-Raby et al., 2003). However, little evidence is available on the effect of Pseudomonas aeruginosa on the inflammatory-degradative response in human lung alveolar epithelial type II (A549 line) cells.
Our results showed that P. aeruginosa was able to increase the secretion of 1) the proinflammatory cytokines IL-1β and TNFα; and 2) the prodegradative enzyme

. Secretion of TNFα by A549 cells stimulated with P. aeruginosa.
After 3 (A), 6 (B), and 24 hours (C) of stimulation with or without P. aeruginosa (different number of colony-forming units; CFU/mL), the culture medium of A549 cells was recovered and analyzed using ELISA. The concentration of TNFα was expressed as pg/mL. The assay was performed in 8 independent experiments with duplicates. Data represent the mean ± standard deviation. Statistically significant difference *p≤0.05 vs. control group.
MMP-9 in a time-and concentration-dependent manner. This proinflammatory/prodegradative environment compromised cell viability through changes in cell morphology and decrease of E-cadherin expression in the A549 cells.
IL-1β is a pivotal cytokine in several second messenger signaling pathways. It is involved in the activation of the inflammatory response (Chen et al., 2017;Ledesma et al., 2004), acts as a modulator of the specialized cells of the immune system (Gabay et al., 2010;Rubartelli et al., 1990), and induces the expression of MMPs (Eberhardt et al., 2000;Nam & Kwon, 2014). The production of IL-1β by alveolar macrophages and epithelial cells is induced by different bacterial components that interact with Tolllike receptors 4 (TLR4). Interestingly, this receptor has high homology with the IL-1R receptor which amplifies the inflammatory response and promotes the activation of transcription factors, such as nuclear factor kappabeta (NFĸβ) and activator protein (AP-1), inducing the expression of genes related to the inflammatory response (Armstrong et al., 2004;Parker et al., 2016). Wong and others (Wong et al., 2012) showed that alveolar type I cells obtained from rats that were stimulated with LPS from E. coli for 18 hours, show high expression of TNFα and IL-1β, but a low expression of IL-6 (Wong & Johnson, 2013). Similarly, in our experiments A549 cells stimulated during with P. aeruginosa for 24 hours showed a 10-fold increase in secretion of TNFα (Fig. 3C) in comparison to IL-1β (Fig. 2C).
Saperstein and others (Saperstein et al., 2009) and Thorley and others (Thorley et al., 2007), demonstrated that the IL-1β signaling pathways modulate TNFα secretion. They used mouse lung epithelial type II and prima-ry human alveolar type II cells to show that increase of TNFα can be reversed by using small interfering RNA and by neutralizing IL-1β with a specific antibody, respectively.
Recently, Jayaraman and others (Jayaraman et al., 2013) proposed a hypothetical mechanism by which IL-1β increases the secretion of TNFα via interacting with the type-1 form of the TNF receptor (TNFR1) and increasing the secretion of the soluble form of TNFα (Jayaraman et al., 2013;MacEwan, 2002). However, a alternative mechanisms mediated by nuclear factor kappa-beta (NFĸβ) could also explain the link between IL-1β and TNFα (Fig. 7). NFĸβ plays an important role in the immunological pathway (Tak & Firestein, 2001), and mutations of cellular NFκβ induced changes in this immunological response (Picard et al., 2011;Sung et al., 2014). NFĸβ and mitogen-activated protein kinases (MAPKs) knockout mice displayed an altered inflammatory response of chemokines and cytokines after LPS stimulation (Picard et al., 2011;Sung et al., 2014).
The next phase of the inflammatory response promoted by IL-1β/TNFα is the expression and secretion of MMPs (Fang et al., 2006;Flores-Herrera et al., 2012). Our results suggest that an infectious and inflammatory process modulates the secretion of proMMP-2 and -9 in a dose-dependent manner and in relation to the stimulation time (Fig. 4).
There is evidence of the mechanism through which IL-1β (Eberhardt et al., 2000;Mon et al., 2017;Ruhul Amin et al., 2003) and TNFα (Fang et al., 2006;Jayaraman et al., 2013;Mon et al., 2006;Tsai et al., 2014) increase the activity of MMP-9 (Fig. 7). Recently, Mon et al. (2017) demonstrated that IL-1β activates MMP-9 Representative gelatin-gel zymography (A, E, and I) showing enzymatic activity of proMMP-2 and proMMP-9 secreted into the culture medium by A549 cells after stimulation with or without P. aeruginosa (different number of colony-forming units; CFU/mL). After 24 hours of stimulation with P. aeruginosa, we detected actMMP-2. The proMMP-9 form was not clearly visualized (I). Each lysis band was quantified by densitometric analysis after bacterial stimulation (B, F, and J). The baseline activity of media was evaluated using a promyelocyte cell line (U937, ATCC Manassas, VA, USA). The assay was performed of 8 independent experiments. Data represent the mean ± standard deviation. Statistically significant difference *p≤0.05 vs. control group. Phase-contrast images showing the change in the morphology of A549 cells after stimulation with 10 6 CFU/mL of P. aeruginosa (D, H, L) vs. control group (Ctrl, C, G, and K). The magnification of the main image is 10x and of the box is 40x. Scale bar=100 µm. through a series of intracellular signals initiated by the activation of the proto-oncogene tyrosine-protein kinase Src (Src) which phosphorylates two tyrosines (Y397 and Y925), activating the system mediated by the growth factor receptor-bound protein 2 (Grb2) and Ras-dependent MAPK protein. This complex activates the MMP-9 (Mon et al., 2017). In addition, it was also shown by Mon et al. (2006) that TNFα interacts with the focal adhesion kinase (FAK) directly involved in the MMP-9 expresion. FAK activation is mediated by the TNFR2 receptor in two tyrosine (Y398, and Y925). These findings were confirmed using an antibody against TNFR2, which inhibited FAK phosphorylation and by using FAK -/-cells, which prevented the degradative activity of MMP-9 (Mon et al., 2006).
Finally, after 24 hours of stimulation with P. aeruginosa we observed a 72 KDa band corresponding to proMMP-2 and a 62-KDa band corresponding to its active form (Fig. 4E). Unfortunately, the activity of MMP-2 could not be determined. Furthermore, proMMP-9 (92 KDa) could not be clearly identified in the activity gels ( Fig. 4E and F), but it was clearly detected in the extracellular matrix of A549 cells using a specific antibody (Fig. 5). Alterations in the morphology of A549 cells were also evident (Fig. 6). Frisdal et al. (Frisdal et al., 2001) and Jackson and others (Jackson et al., 2010) have shown higher ex- A bright signal from E-cadherin immunostaining was detected after 3, 6, and 24 hours in the control group. In contrast, weaker staining was observed after 24 hours of incubation with P. aeruginosa. In these assays, the nucleus was stained using 4´,6-diamidino-2-phenylindole (DAPI, blue color) and colocalization with E-cadherin immunostaining (green color) was shown. The assay was performed in five independent experiments. The magnification 20x.

Figure 5. Immunoreactivity of proMMP-9 in A549 cells.
Increased immunoreactivity of actMMP-9 was observed after 3, 6, and 24 hours of stimulation with P. aeruginosa when compared to the control group. In these assays, the nucleus was stained using Hoechst (blue color) and colocalization with actMMP-9 immunostaining (red color) was shown. The assay was performed in five independent experiments. The magnification is 20x.
pression of MMP-2 and -9 during pulmonary pathological. During physiological development, MMPs are secreted into the extracellular space in the form of proMMPs and are bound to specific tissue inhibitors (TIMPs), as well as to the membrane-type metalloproteases (MT-MMP) (Somerville et al., 2003). Their activation is triggered by the removal of the peptides associated with the active site of the proMMP-2 (72 KDa) and proMMP-9 (92 KDa), inducing conformational change (Defawe et al., 2005;Koo et al., 2012;Somerville et al., 2003) (Fig. 7). Moreover, evidence from different sources suggests that in pathological processes, actMMP-9 degrades the E-cadherin involved in cell-cell adhesion (Allport et al., 2002;Nawrocki-Raby et al., 2003). Using immunohistochemistry, Shaco-Levy et al. (2008) showed that an increase in the secretion of actMMP-9 reduces the level of E-cadherin and intracellular β-catenin protein. Our results showed a reduction of the E-cadherin with relation to the concentration of P. aeruginosa used for stimulation and the time of stimulation (Fig. 7). Carayol et al. (2002) and Kim et al. (2018) used human nasal epithelial cell to demonstrated the association between an increase of MMP-9 expression and a decrease in E-cadherin levels. Interestingly, the activity of MMP-9 was inhibited by preincubation with dexamethasone which was accompanied by increased levels of E-cadherin (Carayol et al., 2002;Kim et al., 2018).
Although in this study we did not examine the expression of NFĸβ, we are planning to do it as part of our research project in order to explore the potential links between inflammasome (IL-1β/TNFα) and NFĸβ.
The studies reported here demonstrated that P. aeruginosa induces mainly the secretion of TNFα, increasing the actMMP-9, and significantly reduces the level of Ecadherin in the A549 cells. The interaction of the structural components of P. aeruginosa lipopolisaccharide (LPS) with Toll-like receptor 4 activates a series of intracellular signals (myeloid differentiation primary response protein; MyD88 and TIR domain-containing adaptor inducing interferon-beta; TRIF) leading to the phosphorylation of the inhibitory protein (Iĸβα) that induces the activation of the nuclear transcription activator NFĸβ and as a consequence increases IL-1β expression (Liu et al. 2017). IL-β interacts with the receptor for IL-1β and with the receptor type-1 of TNF (TNFR1) (Jayaraman et al., 2013;Jackson et al., 2010), modulating the secretion of extracellular matrix metalloproteases (MMPs). At the extracellular level, proMMP-2 is modified by the membrane type-1 matrix metalloproteinase (MT1-MMP) increasing its degradative capacity (Fig. 4I). actMMP-2 is able to cut specific regions of proMMP-9 transforming it into its active form with the capability to degrade different substrates; among them, the cell-binding protein of the E-cadherin type (Fig. 6).