Cofilin-1 and profilin-1 expression in lung microvascular endothelial cells exposed to titanium dioxide nanoparticles

Abstract

Background. Air pollutants exacerbate chronic airway diseases, such as asthma and chronic obstructive pulmonary disease (COPD). However, the underlying mechanisms are yet to be determined. While a number of studies have reported adverse effects of nanoparticles on humans, little is known about their effects on the respiratory system.

Objectives. To examine the protein expression in human lung microvascular endothelial cells (HMVEC-L) exposed to titanium dioxide (TiO₂) nanoparticles, a common air pollutant.

Materials and methods. A proteomics approach using two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and matrix-assisted laser desorption/ionization time-of-flight/time-of-flight mass spectrometry (MALDI-TOF/TOF MS) was used to determine the differences in protein expression at 8 h and 24 h, following the treatment of HMVEC-L with 20-μM or 40-μM TiO₂ nanoparticles.

Results. Human lung microvascular endothelial cells treated with 20-μM TiO₂ nanoparticles showed alterations of 7 protein spots, including molecules related to calcium regulation, transport, cytoskeleton, and muscle contraction. The treatment of HMVEC-L with 40-μM TiO₂ nanoparticles resulted in alterations of 4 protein spots, with molecular functions related to the cytoskeleton, myosin regulation, actin modulation, as well as guanosine diphosphate (GDP) and guanosine triphosphate (GTP) regulation. To validate these results, immunohistochemical staining and western blotting analyses were performed on lung tissues collected from mice exposed to TiO₂ nanoparticles. Cofilin-1 and profilin-1 were expressed in the endothelium, epithelium and inflammatory cells, and decreased in lung tissues of TiO₂ nanoparticle-exposed mice compared to sham-treated controls.

Conclusions. These results suggest that some of the differentially expressed proteins may play important roles in airway diseases caused by TiO₂ nanoparticle exposure.

Key words: proteomics, air pollutants, airway disease, nanoparticles

Background

Air pollutants encompass also toxic particles and gases emitted in large quantities from many different sources, including vehicles and factories.¹ Major pollutants include particulate matter (PM), ozone, nitrogen dioxide (NO₂), and sulfur dioxide (SO₂). Air pollution negatively impacts human health and increases the burden of disease and demand for healthcare services.^{2, 3, 4}

The principal component of indoor and outdoor air pollution is PM, which can vary widely in particle size, and includes coarse, fine and ultrafine particles. Furthermore, PM is a complex mixture of materials with a carbonaceous core and associated materials, such as organic compounds, acids and fine metal particles.^{5, 6}

Inflammation exacerbates pulmonary disease and causes lung dysfunction. The most important issue regarding environmental air pollution is its toxic effect, which can result in inflammatory disease,^{7, 8} with the lung function of patients with airway disease showing greater deterioration among those living in communities exposed to high levels of air pollution.⁹ Toxic titanium dioxide (TiO₂) nanoparticles are produced by human activities, such as driving automobiles and charcoal burning, as well as by natural processes.^{10, 11, 12} Titanium dioxide nanoparticles have toxic effects that manifest in inflammation and lead to disease.^{13, 14, 15} Furthermore, the exposure to TiO₂ nanoparticles induces both acute and chronic inflammation, increases inflammatory cell expression and mucosal gland hyperplasia in the airway and alveolar spaces in animal models,^{13, 16, 17} and can cause itching, bronchoconstriction and mucus hypersecretion.¹⁸ The TiO₂ nanoparticles achieve this by penetrating the cell membrane and exerting toxic effects in organs.^{19, 20} Our previous studies demonstrated that the exposure to TiO₂ nanoparticles induced pulmonary inflammation and activated the inflammasome in mice, causing differential expression of NLRP3, caspase-1, and cytokines IL-1β and IL-18 in the lungs.⁵ Moreover, TiO₂ nanoparticles increase the production of reactive oxygen species (ROS) and oxidative products, as well as the depletion of cellular antioxidants.^{21, 22}

Tight junctions (TJs) are the apical-most constituents of the junctional complex in epithelial cell sheets, also present in vascular endothelial cells and mesothelial cells.²² Tight junctions act as a semipermeable barrier to the paracellular transport of ions, solutes and water, and function as a barrier between the apical and basolateral domains of plasma membranes.²² Tight junctions coordinate a variety of signaling and trafficking molecules involved in the regulation of cell differentiation, proliferation and polarity, and thereby serve as multifunctional complexes.^{21, 23} The disruption of the function of TJs can cause or contribute to a variety of pathological conditions, such as infections, cancers and blood-borne metastasis.²⁴ Endothelial cells are among the main cellular constituents of blood vessels, and one of their most important functions is the formation of a barrier between blood and underlying tissues. These endothelial junctional proteins play important roles in tissue integrity, vascular permeability, leukocyte extravasation, and angiogenesis.²¹

The actin-severing activity of cofilin is an important factor in the stretch-induced cytoskeletal fluidization and may account for an appreciable part of the bronchodilatory effects of a deep inspiration.²⁸ Both cigarette smoke-exposed and chronic obstructive pulmonary disease (COPD) patient-derived epithelia display quantitative evidence of cellular plasticity, along with a loss of specialized apical features with distinct cell motion indicative of cellular jamming. These injured/diseased cells have an increased fraction of polymerized actin, due to the loss of the actin-severing protein, cofilin-1.²⁹ Counteracting cofilin-1 is profilin-1, an actin-binding protein that promotes actin polymerization in vitro,^{25, 26} endothelial cell migration and proliferation.²⁷ A requirement for dynamic actin association and dissociation is rendered by reversible Ser⁷¹ phosphorylation and dephosphorylation. The phosphorylation of Ser⁷¹ inhibits actin-binding of profilin-1.³⁰

The mechanisms by which toxic nanoparticles induce pulmonary disease remain to be elucidated. In this study, we examined whether nanoparticles play a role in lung disease employing two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) and matrix-assisted laser desorption/ionization time-of-flight/time-of-flight mass spectrometry (MALDI-TOF/TOF MS) of endothelial cells, and examined protein expression in the lungs of a mouse model of asthma.

Objectives

This study investigated whether TiO₂ nanoparticle exposure altered protein expression in vitro and in vivo.

Materials and methods

Cell culture and stimulation
with TiO₂ nanoparticles

Primary human lung microvascular endothelial cells (HMVEC-L; 5000 cells/cm², cat No. CC-2527; Lonza, Basel, Switzerland) were grown in EGM™-2MV Microvascular Endothelial Cell Growth Medium BulletKit™ (Lonza). The medium was replaced every 48 h until cells reached 90% confluence at 37°C in 5% CO₂. The cells were seeded in 6-well plates. Twenty-four hours before the experiment, the medium was changed to EBM-2 supplemented with 0.1% fetal bovine serum (FBS). After 30 min, cells were stimulated with TiO₂ nanoparticles (SRM 1898, 20 or 40 µM; National Institute of Standards & Technology (NIST), Gaithersburg, USA) for 8 h or 24 h. The endotoxin concentration in nanoparticles was <0.064 ng/mL (0.32 EU/mL), based on the limulus amebocyte lysate assay (QCL-1000; BioWhittaker, Walkersville, USA).

Two-dimensional electrophoresis
and image analysis

Human lung microvascular endothelial cells were harvested by centrifugation and then disrupted with lysis buffer containing 5 mM Tris-HCl (pH 7.4), 100 mM NaCl, 1% Triton X-100 and 2 mM phenylmethylsulfonyl fluoride (PMSF). Cell lysates were centrifuged at 12,000 rpm for 30 min and the supernatant was collected. The protein concentration was determined using a commercial BCA assay kit (Thermo Fisher Scientific, Rockford, USA) and the samples were stored at −70°C until use. Immobiline DryStrips (Amersham Biosciences, Piscataway, USA) were used for isoelectric focusing (IEF), which was carried out with 1 mg of the protein on an IPGphor system (Amersham Biosciences). The proteins were then separated in the second dimension by sodium dodecyl sulfate (SDS)-PAGE. Staining was performed with Coomassie brilliant blue G250 for the visualization of proteins. The 2-D gels were scanned with an ImageScanner (Amersham Biosciences) in a transmission mode. Spot detection and matching were performed using ImageMaster 2D v. 5.0 (Amersham Biosciences). The 2D spot intensity was calculated by integrating the optical density over the spot area. The data normalization was performed using ImageMaster 2D v. 5.0 (GE Healthcare, Seoul, South Korea) and then the values were exported to SPSS software v. 18.0 (SPSS Inc., Chicago, USA) for statistical analysis.

Protein identification
with MALDI-TOF/TOF MS

Mass spectrometry was performed at Yonsei Proteome Research Center (YPRC; Seoul, South Korea). Synovial fluid and serum peptide profiles were analyzed with a MALDI-TOF/TOF mass spectrometer (Micromass, Wilmslow, UK). The following reagents and solutions were used for sample preparation: 70% acetonitrile (ACN), 2% formaldehyde and 2% fluorescent antibody (FA) buffer. Finally, the peptides attached to the resin were eluted in a drop-by-drop manner using an elution buffer.

Database search and protein identification

Peptide matching and protein searches against the Swiss-Prot and NCBI databases were performed using the Mascot software (Matrix Science; http://www.matrixscience.com).

Animal experiment

All experimental protocols were approved by the Animal Care and Use Committee of Soonchunhyang University, Asan, South Korea (approval No. SCHBCA-2015-06).³¹ Six-week-old female BALB/c mice (n = 8 per group) were sensitized by intraperitoneal (ip.) injection on days 0 and 14, with grade V chicken egg ovalbumin (OVA, 50 μg; Sigma-Aldrich, St. Louis, USA) emulsified in 10 mg of Al(OH)₃ plus Dulbecco’s phosphate-buffered saline (D-PBS, 100 μL). On days 21–23, all mice received an intranasal challenge with grade III OVA (150 μg; Sigma-Aldrich) in D-PBS (50 μL). Control mice were sensitized and challenged with saline. In the TiO₂ nanoparticle groups, mice were exposed to TiO₂ nanoparticles by inhalation (200 µg/m³ for 2 h) before OVA challenge, every day for 3 days. The airway hyperresponsiveness was measured on day 23. Bronchoalveolar lavage fluid (BALF) and lung tissues of mice were then collected for further analysis, including measurement of cofilin-1 and profilin-1 in lung protein, as well as BALF.

Determination of airway responsiveness and morphological analysis

Mice were challenged with 0 mg/mL (baseline), 5 mg/mL, 20 mg/mL, or 100 mg/mL methacholine. After challenge, enhanced pause (Penh) was measured using methacholine in a barometric plethysmographic chamber (All Medicus Co., Anyang, South Korea). Mice were anesthetized the next day and BALF was obtained and stored at −20°C until use. Differential cell counts within BALF samples were determined with Diff-Quick staining of cytospin slides (500 cells per animal). A portion of the lung was fixed in 4% buffered paraformaldehyde and embedded in paraffin. The tissue was cut into 4-μm-thick sections for histological analysis. The images were captured using an ECLIPSE Ci-L microscope (Nikon Corp., Tokyo, Japan) equipped with a DS-Ri2 digital camera (Nikon Corp.).

Western blotting analysis

The extracted lung tissue was homogenized in protein lysis solution containing 50 mM Tris-HCl (pH 7.4), 50 mM NaCl, 0.1% SDS, 1% TritonX-100, 0.5 mM ethylenediaminetetraacetic acid (EDTA), and 100 mM PMSF in distilled water, and centrifuged at 14,000 rpm for 30 min at 4°C. Then, the soluble fraction was collected. Mouse lung proteins were separated with SDS-PAGE and transferred onto polyvinylidene difluoride (PVDF) membranes, which were then blocked with 5% bovine serum albumin (BSA) plus 0.1% Tween 20 in Tris-buffered saline (TBS) for 2 h at room temperature. The membranes were incubated with rabbit anti-cofilin-1 (1:5000; Abcam, Cambridge, UK), mouse anti-profilin-1 (1:100; Santa Cruz Biotechnology, Santa Cruz, USA) or rabbit anti-profilin-1 (1:1000; ECM Biosciences, Versailles, USA) antibody overnight at 4°C. The next day, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies. The detection was performed using EzWestLumi plus western blot detection reagent (ATTO Corporation, Tokyo, Japan). The relative abundances of the proteins were determined using quantitative densitometry with normalization relative to β-actin (Sigma-Aldrich).

Immunohistochemistry

Mouse lung sections were deparaffinized and rehydrated by passage through a series of increasing concentrations of ethanol. The sections were treated with 1.4% H₂O₂ in methanol for 30 min to block endogenous peroxidase activity, and nonspecific binding was then blocked with 1.5% normal horse serum. Next, the samples were incubated with rabbit anti-cofilin-1 (1:2000; Abcam) or mouse anti-profilin-1 (1:50; Santa Cruz Biotechnology) antibody. The next day, the sections were incubated using an ABC kit (Vector Laboratories, Burlingame, USA), and the color reaction was developed by staining with a DAB substrate kit (Golden Bridge International Inc., Mukilteo, USA). After immunohistochemical staining, the slides were counterstained with Harris hematoxylin for 1 min. Staining was quantified using ImageJ software (National Institutes of Health (NIH), Bethesda, USA).

Statistical analyses

The researchers hypothesized that the TiO₂ affects asthma pathogenesis through regulating cofilin-1 and profilin-1. Alternatively, a null hypothesis was proposed by the researchers, which advocated that there was no statistically significant difference between the groups. The data were expressed as median and interquartile range (IQR). All data were analyzed using IBM SPSS v. 22.0 software (IBM Corp., Armonk, USA). The differences between groups were determined using the nonparametric Kruskal–Wallis test, followed by Dunn’s post hoc test. A value of p < 0.05 was considered statistically significant.

Results

2D-PAGE of HMVEC-L exposed
to TiO₂ nanoparticles

A proteomics approach was used to determine the differential expression of proteins 8 h and 24 h after treating HMVEC-L with 20- or 40-μM TiO₂ nanoparticles (Figure 1). A total of 11 spots were detected on each gel, all identified spots were localized in the 3–10 pH range, and the molecular mass ranged from 10 kDa to 150 kDa.

Protein spots selected on 2D-PAGE

Following the treatment for 8 h or 24 h with TiO₂ nanoparticles, 7 and 4 spots were found to have a differential, greater than two-fold expression on 2D-PAGE gel extracts (concentrations of 20 μM and 40 μM, respectively). These spots were excised from the gel and incubated with trypsin to digest the proteins in the gel, and then analyzed using MALDI-TOF/TOF MS (Table 1).

Identification of protein in each spot

Moesin, calreticulin precursor, eukaryotic initiation factor 4A-I isoform 1, heat shock protein β-1, profilin-1, myosin light polypeptide 6, and chain A were identified in the spots showing a differential expression. The differential expression (up or down) was shown in the Table 1 (average at 8 h and 24 h) on exposure to TiO₂ nanoparticles at a concentration of 20 μM. Furthermore, changes in the expression of cofilin-1, rho GDP-dissociation inhibitor 1 isoform A, cytokeratin 7, and glycyl-tRNA synthetase were detected in cells exposed to 40-μM TiO₂ nanoparticles. The identified spots had biological functions related to the cytoskeleton, calcium regulation, initiation factors, transport, muscle contraction, actin modulation, guanosine diphosphate/guanosine triphosphate (GDP/GTP) regulation, and protein coding.

Airway hyperresponsiveness
and inflammatory changes on exposure
to OVA and OVA plus TiO₂ nanoparticles

Airway hyperresponsiveness and inflammatory cells were increased in mice exposed to OVA and TiO₂ nanoparticles, compared to sham-treated control mice. Moreover, the increase was more pronounced with OVA+TiO₂ nanoparticles compared to OVA alone (Figure 2, Table 2, Table 3).

Protein validation with western blotting
and immunohistochemical analyses

Western blotting analysis revealed decreases in cofilin-1 and profilin-1 expressions in HMVEC-L exposed to TiO₂ nanoparticles for 24 h, but not in cells exposed for 8 h (Figure 3A, Table 4, Table 5). Furthermore, there were no changes in protein expression after 8 h or 24 h of starving the cells. The expression of profilin-1 was increased by exposure to TiO₂ nanoparticles in a time-dependent manner (Figure 3A, Table 4, Table 5). The expression levels of cofilin-1 and profilin-1 were increased in lung tissue of asthmatic mice compared to control mice and were decreased in the OVA+TiO₂ nanoparticles group compared to the OVA group (Figure 3B, Table 2, Table 3). The level of profilin-1 was significantly decreased in the sham-treated group, but it was increased in both the OVA group and the TiO₂ groups (Figure 3B, Table 2, Table 3). The immunohistochemical analysis indicated that cofilin-1 and profilin-1 levels in lung tissue were increased in the OVA group compared to the control group, but the increase was less pronounced in the OVA+TiO₂ nanoparticles group (Figure 4, Table 2, Table 3).

Discussion

This study revealed changes in protein expression in endothelial cells exposed to TiO₂ nanoparticles using a proteomics approach. Cofilin-1 and profilin-1 were shown to contribute to airway inflammation in the lungs of mice exposed to these potent air pollutants.

Ambient pollution has both a particulate component and a gaseous component, with the major fractions relevant to health designated as PM₁₀ (10 μm), PM_2.5 (2.5 μm) and ultrafine PM.³² Moreover, air pollution causes and exacerbates asthma, as demonstrated in a meta-analysis that revealed the associations between the exposure to traffic-related pollutants (black carbon, NO), PM_2.5 and PM₁₀, and the development of asthma in children.³² Meta-analyses of the associations between air pollution and asthma exacerbations requiring emergency healthcare utilization in children and/or adults also showed significant associations for NO₂, PM_2.5, CO, and O₃.^{33, 34}

Particulate matter with a diameter smaller than 100 nm is defined as ultrafine particles or nanoparticles. Nanoparticles are mainly found in urban air as single and aggregated particles, and can be divided into 2 major categories based on their source. Typically, nanoparticles are generated incidentally, often as byproducts of fossil fuel combustion and condensation of semi-volatile substances or industrial emissions, although they can also be produced through engineering processes.^{19, 35} The exposure to PM induces oxidative stress and inflammation, stimulating both the innate and acquired immune responses in laboratory animals and humans. Oxidant-mediated cellular damage, including the production of ROS and oxidative stress, and innate and adaptive immunity, can lead to PM-mediated adverse health effects.¹⁹

Our previous studies showed that air pollutants promote lung disease through various mechanisms, including innate immunity (e.g., via macrophage migration inhibitory factor and inflammasomes) and neuroinflammation.^{8, 31, 36} Nanoparticles also alter lung structure, reflected in goblet cell hyperplasia and airway remodeling.^{13, 37}

In this study, proteomic analyses showed that the treatment of HMVEC-L with 20-μM TiO₂ nanoparticles altered the expression of 7 proteins, including moesin, calreticulin precursor, eukaryotic factor 4A-isoform 1, heat shock protein β-1, profilin-1, myosin light polypeptide 6, and chain A, while the treatment with 40-μM TiO₂ nanoparticles altered the expression of 4 proteins, including cofilin-1, rho GDP-dissociation inhibitor 1 isoform A, cytokeratin 7, and glycyl-tRNA synthetase, that have roles in calcium regulation, transport, cytoskeleton, muscle contraction, myosin regulation, actin modulation, and GDP and GTP regulation.^{38, 39} Further studies are needed to clarify the role of this variety of expressed proteins in air pollutant-induced lung diseases.

Cofilin is a widely distributed intracellular actin-modulating protein that binds and depolymerizes filamentous (F)-actin and inhibits the polymerization of monomeric globular (G)-actin in a pH-dependent manner. It is involved in the translocation of actin-cofilin complex from the cytoplasm to the nucleus.³⁹ Cofilin modulates actin transport and function in the nucleus as well as actin organization associated with mitochondrial fission and mitophagy. Cofilin-saturated F-actin fragments under stress conditions can undergo oxidative cross-linking and bundle together to form cofilin-actin rods.⁴¹ Cofilin-1 is a part of a novel mechanism linking mechano-transduction and transcription.⁴² Profilin-1 is a 140-amino acid protein that regulates the growth of F-actin by binding to monomeric G-actin.³⁸ In the present study, cofilin-1 and profilin-1 protein expression levels were changed in endothelial cells and lungs exposed to TiO₂ nanoparticles, suggesting that these proteins may be useful markers of nanoparticle exposure in the lung. These results can be partly explained by recent experiments.^{40, 41, 42} Injured epithelial cells exposed to cigarette smoke in COPD have an increased fraction of polymerized actin, due to loss of the actin-severing protein cofilin-1.⁴⁰ Further studies are needed to clarify the underlying signaling pathways and implications of these findings for airway diseases.

The respiratory epithelia function as a selective barrier between the outside environment and underlying tissue. Tight junctions are considered to function as a barrier between the apical and basolateral domains of plasma membranes.^{21, 43} In this study, TiO₂ nanoparticles altered the lung structure and disrupted the function of TJs in epithelial cells and endothelial cells, suggesting that nanoparticles can penetrate cell barriers and promote airway inflammation and hyperresponsiveness, thus contributing to airway remodeling.

Limitations

Future studies are needed for the multi-omics approaches which would lead to a more comprehensive overview, as compared to single-omics layer with slight advantages for combinations that complement each other directly, e.g., proteome and metabolome.^{44, 45, 46, 47}

Conclusions

In conclusion, nanoparticles are involved in airway inflammation and responsiveness (mediated by cofilin-1 and profilin-1), and in the change of cell membranes. Further studies are necessary to determine the mechanisms underlying these effects.