Mast Cell Chymase and Human Lung Fibroblast Interaction: Mechanisms and Implications for Asthma

TL;DR

This study explores how mast cell chymase affects human lung fibroblasts, revealing new mechanisms that may contribute to asthma development.

Contribution

The study identifies novel intracellular effects of chymase on fibroblast signaling and metabolism, independent of PAR activation.

Findings

Chymase degrades fibronectin and alters signaling factors like integrin αVb3, FAK, Src, and Akt in lung fibroblasts.

Chymase reduces fibroblast motility and suppresses CREB and c-Jun levels or phosphorylation.

Chymase enters fibroblasts, degrades HSP27, and suppresses metabolic activity without inducing cell death.

Abstract

Mast cells play a crucial role in the pathogenesis of asthma, by releasing inflammatory mediators including mast cell-specific proteases such as tryptase and chymase. However, the exact role of these proteases in asthma is not fully understood. We showed previously that chymase imposes multiple effects on primary human lung fibroblasts (HLFs), and in the present report we addressed the underlying mechanisms. The effects of chymase on HLFs were found to be independent of protease-activated receptors (PARs), as judged by employing PAR agonists and antagonists, despite PAR1 and PAR3 being highly expressed. Further, Western blot analysis revealed that chymase degraded fibronectin and affected the levels and phosphorylation status of multiple signalling factors related to fibronectin, including integrin αVb3, focal adhesion kinase, Src and Akt. This was associated with a decreased motility of chymase-treated HLFs, as assessed by a cell migration assay. Chymase treatment also suppressed the levels and/or the phosphorylation status of CREB and c-Jun. Intriguingly, chymase was shown to induce the degradation of HSP27, an intracellular protein, suggesting that chymase may exert proteolytic activity in the cell interior. In agreement with this, confocal microscopy analysis revealed that chymase was taken up by the HLFs. By using Seahorse technology, we also demonstrate that chymase has the capacity to suppress the metabolic activity of HLFs, without causing cell death. Altogether, these findings reveal a novel role for mast cell chymase in affecting signalling pathways in HLFs, thereby providing insight into how mast cells potentially can affect the lung microenvironment in asthma.

The online version contains supplementary material available at 10.1007/s10753-026-02460-3.

Figures

Chymase affects primary HLFs via a PAR-independent mechanism. (A) Representative phase-contrast images of primary HLFs treated with chymase (5 nM, 24 h) at the indicated time points. (B) Quantification of cell area following treatments in A. Data represent mean values ± SEM (*n* = 10–12 fields). *P* < 0.01, *P* < 0.0001. (C) AmpliSeq transcriptomic analysis of PAR gene expression in HLFs. Chymase significantly reduced F2RL1 (PAR2) and F2RL2 (PAR3) expression; F2R (PAR1) and F2RL3 (PAR4) shown for reference. P* ≤ 0.001, *P* ≤ 0.0001. (D) Representative phase-contrast images of HLFs treated for

Chymase modulates integrin and syntenin-1-associated signalling pathways in HLFs. (A-B) AmpliSeq transcriptome analysis illustrating the gene expression levels of various heparan sulphate proteoglycans (HSPGs), including glypicans (GPC) and syndecans (SDC), as well as syntenin-1 (SDCBP), in primary HLFs. P* < 0.001. (C-D) Effect of chymase treatment on fibronectin (FN), integrin αV (ITGAV) and integrin β3 (ITGB3), phosphorylation of FAK (Tyr397), phosphorylation of Src family kinases (Tyr416), and Akt (Ser473). GAPDH served as a loading control. Representative blots from at least three indepen

Chymase reduces the migratory capacity of primary HLFs. (A) Representative phase-contrast micrographs illustrating the cell-free area at indicated time points (0, 6, 24, 48 h) in untreated and chymase-treated HLFs. (B) Quantification of relative cell-free area over time. Data are presented as mean values ± SEM, pooled from three independent experiments. *P* ≤ 0.01; P* ≤ 0.001

Chymase modulates intracellular kinase phosphorylation and heat shock protein levels in primary HLFs. (A) Representative Proteome Profiler Phospho-Kinase Array membranes showing phosphorylation profiles in whole cell lysates from untreated primary HLFs and HLFs treated with 5 nM chymase for 24 h. A scheme is provided for target identification. Blue dotted boxes highlight targets with decreased phosphorylation after chymase treatment, while red dotted boxes indicate targets with increased phosphorylation. (B-C) Western blot analysis of total and phosphorylated forms of CREB (p-CREB Ser133), c-J

Chymase reduces the metabolic activity and mitochondrial respiration of primary HLFs without affecting viability or mitochondrial protein abundance. (A) Oxygen consumption rate (OCR) profiles of primary HLFs, untreated or treated with 5 nM chymase for 24 h. Sequential injections include Oligomycin A (ATP synthase inhibitor), FCCP (protonophore), and a Rotenone/Antimycin A mixture (Complex I/III inhibitors). Data represent means ± SD. (B) Quantification of basal respiration in HLFs following 24-hour treatment with 5 nM chymase. Data are presented as means + SD, pooled from three independent exp

Full text

Introduction

Asthma is a heterogeneous, chronic inflammatory disease characterized by airflow obstruction, airway hyperresponsiveness, and airway remodelling [1, 2]. The underlying immune response is Th2-mediated, involving mast cell activation and infiltration of effector T lymphocytes and eosinophils [3, 4]. This inflammatory cascade ultimately leads to excessive deposition of extracellular matrix (ECM), resulting in thickening of the airway wall, subepithelial fibrosis and airway narrowing [5, 6]. Fibroblasts represent a major source of ECM components such as collagens, fibronectin and proteoglycans and, based on this, fibroblasts are implicated as major players in the progression of airway remodelling in the context of asthma [7–9]. In asthma, fibroblasts typically undergo a transition into myofibroblasts, associated with increased expression of smooth muscle actin and maintained production of ECM components [8].

Previous studies have revealed that fibroblasts can interact with various immune cells, which can affect the fibroblast functionality. For example, interaction of fibroblasts with eosinophils or with factors released by eosinophils led to the upregulation of proinflammatory cytokine expression by the fibroblasts [10, 11]. Further, co-culture of bronchial fibroblasts from asthmatics with CD4^+^ T cells resulted in the induction proinflammatory cytokine production [12]. There are also studies indicating that airway fibroblasts can interact with mast cells. This is supported by a study showing that coculture of a mast cell line with human lung fibroblasts led to increased production of IL-6 by the fibroblasts [13]. Further, an increased number of mast cells in close proximity to fibroblasts in lung tissue samples has been observed in patients with various lung diseases [14, 15], supporting direct interactions between human lung fibroblasts and degranulating mast cells. Notably, mast cell granules contain large quantities of preformed compounds, which are released upon mast cell activation [16]. A plausible scenario would thus be that such mast cell-released compounds could have an impact on the airway fibroblasts, such that their functional properties are modulated during the progression of asthmatic disease.

The preformed mast cell granule constituents include histamine and other biogenic amines, certain preformed cytokines, serglycin proteoglycans, various lysosomal enzymes and also remarkably large amounts of mast cell-restricted proteases, the latter including tryptase, chymase and carboxypeptidase A3 [16, 17]. In previous studies we have reasoned that the mast cell-restricted proteases might have the capacity to influence the function of fibroblasts residing in the vicinity of activated mast cells in the airways. Indeed, we recently demonstrated that mast cell chymase has a profound impact on primary human lung fibroblasts, in particular by inducing a proinflammatory phenotype and by influencing their production of ECM components [18]. However, the molecular mechanisms underlying this interaction have remained unclear. In the present study, we have addressed this issue. Our data show that chymase acts on fibroblasts through a pathway involving fibronectin and cell associated fibronectin-binding partners. Moreover, our findings reveal that chymase is taken up by the airway fibroblasts, raising the notion that mast cell chymase, in addition to having effects on extracellular targets, also can execute proteolytic events in the cell interior.

Materials and Methods

Reagents and Antibodies

Recombinant human chymase (C8118-50UG) was purchased from Sigma-Aldrich (St. Louis, MO, USA), unless otherwise stated. Chymostatin (11004638001), a chymase inhibitor, was purchased from Roche (Basel, Switzerland) and used at a concentration of 10 µM. The PAR1 agonist TFLLRN-NH₂ (AS-62937) and PAR3 agonist SFNGGP-NH₂ (AS-62938) were purchased from AnaSpec (Fremont, CA, USA) and used at a concentration of 50 µM. The PAR1 antagonist Vorapaxar (SML3834) was purchased from Sigma-Aldrich and used at a concentration of 50 nM. For the Seahorse analysis, Oligomycin A (75351), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (FCCP; C2920), Rotenone (R8875), and Antimycin A (A8674) were all purchased from Sigma-Aldrich. Phosphate-buffered saline (PBS; 70011-044) was purchased from Gibco (Thermo Fisher Scientific, Waltham, MA, USA). Tris-buffered saline (TBS) tablets (09–7500-100) were purchased from Medicago (Uppsala, Sweden). Tween-20 (8.22184.0500) was purchased from Sigma-Aldrich. For chymase detection by immunofluorescence, Anti-CMA1 Antibody (1:1000, HPA052634) from Atlas Antibodies (Bromma, Sweden) were used.

The following antibodies were used for Western blot analysis:

• From Cell Signaling Technology (Danvers, MA, USA): Fibronectin/FN1 (1:1000, #26836), FAK (1:1000, #3285), Phospho-FAK (Tyr397) (1:1000, #8556), Syntenin-1 (1:1000, #27964), Src (1:1000, #2109), Phospho-Src Family (Tyr416) (1:1000, #2101), Akt (1:1000, #9272), Phospho-Akt (Ser473) (1:2000, #4060), CREB (1:1000, #9197), Phospho-CREB (Ser133) (1:1000, #9198), c-Jun (1:1000, #9165), Phospho-c-Jun (Ser63) (1:1000, #2361), HSP27 (1:1000, #2402), Phospho-HSP27 (Ser15) (1:1000, #2404), Phospho-HSP27 (Ser78) (1:1000, #2405), Phospho-HSP27 (Ser82) (1:1000, #9709), HSP60 (1:1000, #12165), and Anti-rabbit IgG, HRP-linked Antibody (1:1000, #7074). From Santa Cruz Biotechnology (Dallas, TX, USA): Integrin αV (1:200, #sc-10719) and GAPDH (1:1000, #sc-32233). From BD Transduction Laboratories (San Jose, CA, USA): Integrin β3 (1:250, #I19620). From Thermo Fisher Scientific (Waltham, MA, USA): OxPhos Human WB Antibody Cocktail (1:500, #45–8199). From Proteintech (Rosemont, IL, USA): COXIV Polyclonal antibody (1:5000, #11242-1-AP). From ABclonal (Woburn, MA, USA): MTCO1 Rabbit pAb (1:1000, #A17889) and MTCO2 Rabbit pAb (1:1000, #A11522). From Bio-Rad Laboratories (Hercules, CA, USA): StarBright Blue 700 Goat Anti-Rabbit IgG (1:5000, #12004161) and StarBright Blue 700 Goat Anti-Mouse IgG (1:5000, #12004159). All primary antibodies for Western blots were diluted in Odyssey^®^ blocking buffer (LI-COR, 927–40000). Secondary antibodies were diluted in TBS-T.

Cell Culture

Two sources of primary human lung fibroblast (HLF) were used in this study. Normal Human Lung Fibroblasts (NHLF) (Lonza, CC-2512, donor 1 LOT: 22TL094289) were maintained at 37 °C with 5% CO_2_, in Fibroblast Growth Basal Medium (FBM) (Lonza, CC-3131) supplemented with FGM-2 Fibroblast Growth Medium-2 SingleQuots Supplements and Growth Factors (Lonza, CC-4126), 100 U/mL penicillin, and 100 µg/mL streptomycin (Sigma-Aldrich, P0781). Primary Human Lung Fibroblasts (ATCC^®^ PCS-201-013™, donor 2 LOT: 64128489, donor 3 LOT: 70041445) were purchased from the American Type Culture Collection (ATCC). These cells were cultured at 37 °C with 5% CO_2_ in Fibroblast Basal Medium (ATCC^®^ PCS-201-030™) supplemented with Fibroblast Growth Kit-Low Serum (ATCC^®^ PCS-201-041™), 100 U/mL penicillin, and 100 µg/mL streptomycin (Sigma-Aldrich, P0781).

For subculturing, cells were detached at approximately 80% confluency using trypsin-EDTA solution (Sigma-Aldrich, T3924). Detached cells were pelleted by centrifugation (5 min, 4 rcf, 20 °C), resuspended in fresh medium. Cell counts were determined by trypan blue (Thermo Fisher Scientific, 15250061) using an automated cell counter (Countess II FL, Thermo Fisher Scientific). All experiments were performed using cells between passages 2 and 9 to ensure consistency.

Fluorescence Microscopy for Chymase-Fibroblast Interaction

To visualize the interaction of chymase with primary human lung fibroblasts (HLFs), recombinant human chymase/CMA1 protein (CMA1; R&D Systems, 4099-SE-010) was fluorescently labelled using the Alexa Fluor™ 488 Protein Labelling Kit (Thermo Fisher Scientific, A10235). Labelling was performed according to the manufacturer’s protocol, with a modification due to limited protein availability: 50 µL of 1 M sodium bicarbonate solution was added to 71 µL of the protein solution (0.14 mg/mL). The subsequent protein purification and dye separation steps were completed as instructed.

HLFs were seeded into µ-Slide 4 Well chamber slides (Ibidi, 80426) and allowed to adhere overnight at 37 °C with 5% CO_2_. Cells were then either left untreated or treated with labelled chymase (5 nM). Cells were fixed with 4% formaldehyde (Histolab, 02176) for 15 min at room temperature, rinsed 3 x with PBS, and permeabilized with 0.5% Triton X-100 (Sigma-Aldrich, T9284-100ML) in PBS for 10 min at room temperature. To detect unlabelled chymase, cells were incubated with an Anti-CMA1 Antibody (1:1000; Atlas Antibodies, Bromma, Sweden; catalogue no. HPA052634) overnight at 4 °C. Nuclei were stained with NucBlue ReadyProbes Reagent (Hoechst 33342; Thermo Fisher Scientific, R37605; 2 drops/mL in PBS, 10 min), and F-actin was stained with ReadyProbes Phalloidin Conjugate (Thermo Fisher Scientific, R37112; 2 drops/mL in PBS, 20 min) in the dark. Confocal images were acquired on a Leica Stellaris 5 confocal microscope (Leica Microsystems, Wetzlar, Germany). Cells from donor 1 and donor 3 were used for this experiment.

Western Blot Analysis

Cells were lysed on ice in RIPA buffer (1 × 10^6^ cells per 100 µL) supplemented with phosphatase inhibitor cocktail (1×; Roche, 04906845001) and protease inhibitor cocktail (1×; Roche, 04693132001). Cells were scraped using a 2-position blade cell scraper (Sarstedt, 83.3950) and incubated on ice for 45 min. Lysates were clarified by centrifugation at 16,000 × g (approx. 13,000 rpm) for 15 min at 4 °C, and supernatants were collected. Samples were mixed with 4x Laemmli Sample Buffer (Biorad, 1610747) with 10% NuPAGE Sample Reducing Agent (10X) (Thermo Fisher, NP0009), heated at 95 °C for 5 min, and resolved by Mini-PROTEAN^®^ TGX™ Precast Gels (Bio-Rad). Proteins were transferred by semi-dry transfer using Trans-Blot Turbo Mini 0.2 μm Nitrocellulose Transfer Packs (Bio-Rad, 1704158). Membranes were blocked in Odyssey^®^ blocking buffer for 1 h at room temperature.

Membranes were incubated with primary antibodies overnight at 4 °C, followed by 3 × 10-minute washes in TBS-T. Targets detected by chemiluminescence were probed with HRP-conjugated secondary antibodies and visualized using ECL Prime detection reagent (Cytiva, GERPN2236). All other targets were detected using StarBright Blue 700 Fluorescent Secondary Antibodies (Bio-Rad) and imaged under fluorescence settings. Images were acquired using a ChemiDoc MP Imaging System (Bio-Rad, 17001402). All experiments were performed with at least three replicates. The specific HLF donors utilized for each target quantification were: Src, Phospho-Src Family (Tyr416), OxPhos Human WB Antibody Cocktail (subunits), COXIV, MTCO1, MTCO2: Donor 1; Integrin αV, Integrin β3: Donor 2; FAK, Phospho-FAK (Tyr397), Syntenin-1, HSP27, Phospho-HSP27 (Ser15), Phospho-HSP27 (Ser78), Phospho-HSP27 (Ser82), HSP60: Donors 1, 2; Akt, Phospho-Akt (Ser473), CREB, Phospho-CREB (Ser133), c-Jun, Phospho-c-Jun (Ser63): Donors 1, 3; Fibronectin/FN1: Donors 1, 2, 3.

Cell Migration Assay

Cell migration was assessed using the Ibidi Culture-Insert 2 Well system (Ibidi, P81176). Primary HLFs were prepared as a cell suspension at a concentration of 8.5 × 10^5^ cells/mL. For each well of the Ibidi Culture-Insert, 70 µL of this cell suspension was seeded. Following overnight adhesion at 37 °C with 5% CO_2_, the Ibidi Culture-Insert was carefully removed. The wells were gently washed with PBS to remove any non-adherent cells or debris. Cell migration into the wound area was monitored and evaluated over a span of 48 h. Images of the wound area were captured at regular intervals. Cells from donor 1 and donor 2 were used for this experiment.

Phospho-Array

To assess intracellular signalling responses to chymase, protein phosphorylation profiles were measured using the Proteome Profiler Human Phospho-Kinase Array Kit (R&D Systems, ARY003C). Primary human lung fibroblasts (HLFs) were treated for 24 h with vehicle (untreated control) or recombinant human chymase (5 nM). Following treatment, cells were placed on ice, washed twice with ice-cold PBS, and lysed in the kit-provided lysis buffer 6 supplemented with 1× phosphatase inhibitor cocktail and 1× protease inhibitor cocktail. The following steps were performed according to the manufacturer’s protocol. Cells from donor 3 were used for this experiment.

Seahorse Analysis

Oxygen consumption rates (OCR) and extracellular acidification rates (ECAR) in primary HLFs were determined using an Agilent Seahorse XFe96 Extracellular Flux Analyzer (Seahorse Bioscience, 101991-100). HLFs were seeded at a density of 2.5 × 10^4^ cells/well in XF96 Cell Culture Microplates (Seahorse Bioscience, 103794-100) and incubated overnight at 37 °C with 5% CO_2_. Prior to the assay, the culture medium was replaced with XF DMEM Medium (Seahorse Bioscience, 103575-100) supplemented with 4.5 g/L D-(+)-Glucose (Sigma-Aldrich, G7021) and 6 mM L-Glutamine (Sigma-Aldrich, G7513). Cells were either left untreated or treated with 5 nM recombinant human chymase for 24 h. A mitochondrial stress test was performed following the manufacturer’s protocol, involving sequential injections of specific modulators through the sensor cartridge. The final concentrations of the inhibitors in the well were: 1.2 µM Oligomycin A, 1.8 µM FCCP, and a mixture of 1.8 µM Rotenone and 1.8 µM Antimycin A. Following the Seahorse analysis, cell quantity per well was determined using a bicinchoninic acid (BCA) assay. OCR and ECAR values were normalized to the total protein content (µg/well) for each corresponding well. Data were analysed using Agilent Seahorse Analytics online software (Version: 1.0.0). Cells from donor 1 and donor 2 were used for this experiment.

BCA Assay

Cells were washed with ice-cold PBS and lysed on ice for 45 min with gentle shaking using Pierce™ Luciferase Cell Lysis Buffer (Thermo Fisher Scientific, 89900). Total protein concentration was then determined using a BCA assay. A working BCA reagent was prepared by mixing 50 parts of Bicinchoninic Acid solution (Sigma-Aldrich, B9643) with 1 part of Copper(II) sulphate solution (Sigma-Aldrich, C2284). For quantification, 20 µL of each cell lysate was added to individual wells of a 96-well microplate (Sarstedt, 82.1581.200), followed by the addition of 100 µL of the working BCA reagent. A standard curve was prepared using serially diluted bovine serum albumin (BSA) protein assay standards (Thermo Scientific, 23210). The plate was incubated at 37 °C for 20 min, followed by an additional 10 min at room temperature. Absorbance was then measured at a wavelength of 562 nm using a Tecan Infinite M200 microplate reader (Tecan Group Ltd., Männedorf, Switzerland). Protein concentrations were calculated from the standard curve. Cells from donor 1 and donor 2 were used for this experiment.

Cell Viability Assessment

Cell viability and apoptosis were quantified by flow cytometry using a BD Accuri C6 Plus (BD Biosciences). Following 24 h of treatment, cells were harvested with trypsin/EDTA, and pelleted by centrifugation at 400 × g for 5 min at 4 °C. Supernatants were carefully aspirated, and pellets were resuspended in 1× Annexin V binding buffer (diluted from 10× concentrate; BD Pharmingen, 51-66121E) at 300 µL per tube. Staining was performed by adding 3 µL Annexin V- FITC (BD Pharmingen, 556419) and 3 µL DRAQ7 (BioStatus, DR71000) to each tube, followed by gentle mixing. Samples were acquired within 1 h of staining. Unstained, single-stained (Annexin V only; DRAQ7 only) samples were included for gating. Cells from donor 1 and donor 3 were used for this experiment.

Cell Proliferation Assessment

Cell proliferation was quantified by incorporation of 5-ethynyl-2′-deoxyuridine (EdU) and detection with Alexa Fluor 647. Primary HLFs were suspended at 1 × 10^5^ cells/mL and incubated for 24 h with EdU at 0.5 µL/mL (Component A, Click-iT EdU Alexa Fluor 647 Imaging Kit; Thermo Fisher Scientific, C10340) in the presence or absence of chymase (5 nM). Following incubation, cells were harvested by trypsinization (Trypsin/EDTA), transferred to FACS tubes, and pelleted by centrifugation at 400 × g for 5 min at 4 °C. Supernatants were aspirated, and cells were washed once with 1% BSA (Roche, 10735094001) in PBS. EdU detection was then performed according to the manufacturer’s instructions (Click-iT EdU Alexa Fluor 647 Imaging Kit). Cells were finally resuspended in 300 µL PBS per tube. Unstained cells (resuspended directly in PBS without EdU labelling) were included for gating.

PrestoBlue Assay

The metabolic activity of primary HLF was assessed using the PrestoBlue Cell Viability Reagent. HLFs were seeded at a density of 3 × 10^4^ cells/well into flat-bottom 96-well culture plates (Sarstedt, 83.3924) and left to attach overnight. Following a 24-hour incubation period with the respective treatments, 10 µL of PrestoBlue Cell Viability Reagent (Thermo Fisher Scientific, A13261) was added directly to each well. The plate was then incubated at 37 °C for 30 min. Fluorescence intensity was measured using an Infinite M200 microplate reader (Tecan Group Ltd., Männedorf, Switzerland) with wavelengths set at 570 nm and normalized to the 600 nm value. Cells from donor 1 and donor 3 were used for this experiment.

Quantification and Statistical Analysis

Individual cells were initially segmented using cellpose-SAM (cellpose.org). The resulting outlined cell areas were then measured using ImageJ software (version 1.53c). For each experimental condition and time point, measurements were performed on 12 representative cells (4 cells per independent experiment, pooled from 3 independent experiments). Western blot analyses were quantified by densitometric analysis of band intensities using ImageJ software. For each target protein, band intensities were normalized to the corresponding GAPDH loading control and expressed as fold change relative to the untreated control group, unless otherwise specified. All quantifications were performed on blots obtained from three independent experiments.

Two-way ANOVA followed by Dunnett’s multiple comparison test was employed. Adjusted P values of 0.05 or lower were deemed statistically significant. Figures were generated using GraphPad Prism 8.0 software (GraphPad Software Inc, San Diego, California). The presented results are from individual experiments and are representative of a minimum of three experiments, unless stated otherwise.

Results

Chymase Can Induce Morphological Changes in HLFs Via PAR-Independent Pathways

We have previously reported that mast cell chymase induces dramatic morphological changes in primary HLFs, accompanied by profound effects on their gene expression profile [18]. We consistently observed these effects across multiple batches of HLFs derived from different donors. Consistent with our previous findings, phase-contrast microscopy analysis showed that chymase-treated HLFs exhibited a contracted phenotype, suggesting increased cell thickness (Fig. 1A). In agreement with this, quantification of cell area demonstrated a significant reduction in cell size following chymase treatment (Fig. 1B).Fig. 1. Chymase affects primary HLFs via a PAR-independent mechanism. (A) Representative phase-contrast images of primary HLFs treated with chymase (5 nM, 24 h) at the indicated time points. (B) Quantification of cell area following treatments in A. Data represent mean values ± SEM (n = 10–12 fields). **P < 0.01, ****P < 0.0001. (C) AmpliSeq transcriptomic analysis of PAR gene expression in HLFs. Chymase significantly reduced F2RL1 (PAR2) and F2RL2 (PAR3) expression; F2R (PAR1) and F2RL3 (PAR4) shown for reference. ***P ≤ 0.001, ****P ≤ 0.0001. (D) Representative phase-contrast images of HLFs treated for 24 h with chymase (5 nM), PAR3 agonist SFNGGP-NH2 (50 µM), PAR1 agonist TFLLRN-NH2 (50 µM), PAR1 antagonist vorapaxar (50 nM), vehicle control (DMSO), or chymase + vorapaxar (5 nM + 50 nM)

To elucidate the mechanism underlying the effects of chymase on HLFs, we initially focused on the possibility that chymase may modulate fibroblast function by proteolytically activating any of the protease-activated receptor (PAR) family members (PAR-1–4). To address this possibility, we first assessed the expression levels of the various PARs in the airway fibroblasts, by reanalysing transcriptome data from our previous study [18]. This showed that primary HLFs express relatively high levels of PAR1 (F2R) and PAR3 (F2RL2), but barely detectable levels of PAR2 (F2RL1) and no detectable PAR4 (Fig. 1C). Interestingly, chymase treatment led to a significant downregulation of the F2R and F2RL2 transcripts (Fig. 1C). Given the known interaction between chymase and PAR1 [19], we hypothesized that direct activation of PAR1 using a PAR1 agonist might mimic the morphological alterations induced by chymase. Further, since PAR3 was highly expressed by the fibroblasts and chymase affected the expression of the PAR3 gene (F2RL2), we reasoned that PAR3 could be a potential target for chymase. However, treatment of the fibroblasts with agonists to either PAR1 or PAR3 did not phenocopy the chymase-induced morphological changes (Fig. 1D), indicating that the effects of chymase on the HLFs are not dependent to a major extent on PAR1/3 activation. In agreement with this notion, treatment with a PAR1 antagonist failed to prevent the chymase-induced morphological alterations (Fig. 1D).

Chymase May Exert Intracellular Effects Via a Heparan Sulphate Proteoglycan-Mediated Pathway

One possible mechanism through which chymase may influence fibroblast function could be via interacting with cell surface proteoglycans, given the high affinity of chymase for such compounds [20], as mediated by electrostatic interactions between the highly positively charged chymase and highly negatively (sulphated) proteoglycans [21, 22]. To explore this possibility, we revisited transcriptome sequencing data from our previous study to evaluate the proteoglycan expression profile of the airway fibroblasts. These analyses showed that the HLFs express several members of the heparan sulphate proteoglycan family, including syndecans (SDC2, SDC3, SDC4) and glypicans (GPC1, GPC6) (Fig. 2A-B). Moreover, we noted that a related protein, SDCBP (syndecan binding protein; also known as syntenin-1), was expressed at high levels (Fig. 2A). It was also notable that chymase treatment significantly upregulated SDCBP expression (Fig. 2A). Given the role of SDCBP in coordinating intracellular signalling downstream of extracellular matrix (ECM) and integrin interactions [23], we hypothesized that chymase may affect the functional properties of the fibroblasts through pathways associated with SDCBP. To examine this possibility, we assessed the impact of chymase on ECM components and integrin-associated signalling. Consistent with previous findings, chymase demonstrated strong proteolytic activity against fibronectin derived from HLFs, an effect that was largely inhibited by the chymase inhibitor chymostatin (Fig. 2C). Since fibronectin is a known ligand for several integrins, we next investigated the expression of integrin αV (ITGAV) and integrin β3 (ITGB3) proteins, i.e., both components of the integrin  \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha{V}\beta{3}$$\end{document}  heterodimer which is known to participate in fibronectin binding [24]. As seen in Fig. 2C (quantification shown in Suppl. Figure 1 A-B), chymase treatment caused a marked reduction in the levels of both  \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha{V}$$\end{document} (ITGAV) and  \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha\beta{3}$$\end{document} (ITGB3) in the HLFs. These effects were abrogated by chymostatin (Fig. 2C), indicating that the effect of chymase on these integrins was dependent on its enzymatic activity. We also observed a modest increase in SDCBP levels in HLF culture supernatants following chymase treatment, whereas the levels of cell-associated SDCBP were not affected by chymase (Fig. 2D; quantification shown in Suppl. Figure 1E). We next focused on signalling molecules acting downstream of SDCBP, by first assessing whether chymase affects focal adhesion kinase (FAK). Indeed, chymase treatment caused a marked reduction in the phosphorylation of FAK at Tyr396, without affecting the levels of total FAK (Fig. 2C; quantification shown in Suppl. Figure 1 C-D). Chymase treatment significantly reduced both total proto-oncogene tyrosine-protein kinase (Src) levels and its phosphorylation at Tyr416 (Fig. 2D; quantification shown in Suppl. Figure 1G-H). Similarly, chymase treatment of the HLFs caused a significant decrease in total levels of Protein Kinase B (Akt) and on its phosphorylation at Ser473 (Fig. 2D; quantification shown in Suppl. Fig. I-J).Fig. 2. Chymase modulates integrin and syntenin-1-associated signalling pathways in HLFs. (A-B) AmpliSeq transcriptome analysis illustrating the gene expression levels of various heparan sulphate proteoglycans (HSPGs), including glypicans (GPC) and syndecans (SDC), as well as syntenin-1 (SDCBP), in primary HLFs. ***P < 0.001. (C-D) Effect of chymase treatment on fibronectin (FN), integrin αV (ITGAV) and integrin β3 (ITGB3), phosphorylation of FAK (Tyr397), phosphorylation of Src family kinases (Tyr416), and Akt (Ser473). GAPDH served as a loading control. Representative blots from at least three independent experiments are shown

Chymase Treatment Reduces Cell Motility in Primary HLFs

The data above suggest that chymase acts on a pathway involving fibronectin and a syntenin-1 signalling axis involving FAK and Akt. Since such signalling pathways are known to have a role in cellular migration/motility [25], we next reasoned that chymase could have an impact on the capacity of the lung fibroblasts to migrate. To assess this possibility, we conducted an assay to evaluate the migratory capacity of HLFs following chymase treatment. Consistent with our hypothesis, chymase significantly delayed the migration of the HLFs (Fig. 3A). Quantification of the remaining cell free area confirmed a marked reduction in migration efficiency in the chymase-treated group (Fig. 3B).Fig. 3. Chymase reduces the migratory capacity of primary HLFs. (A) Representative phase-contrast micrographs illustrating the cell-free area at indicated time points (0, 6, 24, 48 h) in untreated and chymase-treated HLFs. (B) Quantification of relative cell-free area over time. Data are presented as mean values ± SEM, pooled from three independent experiments. **P ≤ 0.01; ***P ≤ 0.001

Chymase Treatment Influences Multiple Intracellular Cell Signalling Pathways

The results above introduce the notion that mast cell chymase can affect intracellular signalling in human lung fibroblasts. To provide more comprehensive insight into the effects of chymase on intracellular signalling in the fibroblasts, we next adopted an unbiased phospho-array approach (Proteome Profiler Human Phospho-Kinase Array) to examine phosphorylation changes in response to chymase treatment. The results from this approach revealed that chymase affected the phosphorylation status of several signalling molecules. Hence, we observed that chymase induced an increase in the phosphorylation of signal transducer and activator of transcription 5 (STAT5a/b) and glycogen synthase kinase-3 beta (GSK3β) (Fig. 4A). Conversely, chymase treatment led to decreased phosphorylation of multiple signalling proteins, including cAMP-response element-binding protein (CREB, Ser133), transcription factor c-Jun (Ser63), 90 kDa ribosomal S6 kinases (RSK1/2/3, Ser380/Ser386/Ser377), and proline-rich Akt substrate of 40 kDa (PRAS40, Thr246) (Fig. 4A). The reduction in CREB phosphorylation was independently validated via Western blot analysis (Fig. 4B; quantification shown in Suppl. Figure 1 K-L). For c-Jun, we observed a decrease in total protein levels following chymase treatment. This was reversed by co-treatment with the chymase inhibitor chymostatin, suggesting that this effect was dependent on the enzymatic activity of chymase (Fig. 4B; quantification shown in Suppl. Figure 1 M-N).Fig. 4. Chymase modulates intracellular kinase phosphorylation and heat shock protein levels in primary HLFs.** (A)** Representative Proteome Profiler Phospho-Kinase Array membranes showing phosphorylation profiles in whole cell lysates from untreated primary HLFs and HLFs treated with 5 nM chymase for 24 h. A scheme is provided for target identification. Blue dotted boxes highlight targets with decreased phosphorylation after chymase treatment, while red dotted boxes indicate targets with increased phosphorylation. (B-C) Western blot analysis of total and phosphorylated forms of CREB (p-CREB Ser133), c-Jun (p-c-Jun Ser63), and HSP27 (p-HSP27 Ser15, Ser78, Ser82), as well as total HSP60, in primary HLFs treated with 5 nM chymase for 24 h. (D) AmpliSeq transcriptomic analysis of heat shock protein (HSP) gene expression in HLFs after 24 h of 5 nM chymase treatment. HSPB1 (HSP27), HSPD1 (HSP60), and HSPH1 (HSP105) gene expression levels were not significantly affected by chymase treatment. (E) Confocal images showing intracellular Alexa Fluor 488-labelled chymase (white arrows) in HLFs. (F) Immunofluorescence detection of unlabelled chymase (white arrows) within HLFs using anti-CMA1 and Alexa Fluor 488-conjugated secondary antibody. Nuclei, blue (Hoechst 33342); F-actin, red (phalloidin); chymase, green

The phospho-array approach indicated that chymase induced a reduction in phosphorylation of heat shock protein 27 (HSP27; Ser78/Ser82) and HSP60 (Fig. 4A). However, independent Western blot analyses did not confirm a profound effect of chymase on the phosphorylation status of HSP27 or HSP60 (Fig. 4C). We are at present not able to explain this apparent discrepancy, but likely reasons could be related to the exact source and nature of the antibodies used in the respective experimental approaches. Interestingly, we noted the appearance of a truncated form of HSP27 after chymase treatment of the cells, suggesting that chymase has the capacity to act proteolytically on HSP27 (Fig. 4C; quantification shown in Suppl. Figure 1O-S). This effect was abolished in the presence of chymostatin, confirming that chymase’s enzymatic activity is required for such HSP27 degradation.

To determine whether these effects could be translated to the level of gene transcription, we re-examined transcriptomic data from our previous study. However, despite the marked changes in HSP27 protein stability and phosphorylation imposed by chymase, chymase treatment did not alter HSPB1 (HSP27) mRNA levels (Fig. 4D), indicating that chymase modulates HSP27 predominantly at the post-transcriptional level. Notably though, HSP27 was highly expressed by the HLFs (Fig. 4D).

The finding that chymase promotes the degradation of HSP27, an intracellular protein, suggests that chymase can exert proteolytic activity in the cell interior. In turn, this would require that chymase is taken up by the cells. To gain more insight into this issue, we next investigated whether chymase can interact physically with the cells. To examine this, we labelled chymase with Alexa Fluor 488, added the labelled chymase to the fibroblast cultures and then used confocal microscopy to examine its possible association with the cells. These analyses revealed that chymase was indeed associated with the fibroblasts (Fig. 4E). Further, an intriguing finding was that the labelled chymase in fact could be observed in the cell interior, indicating that chymase is taken up by the cells (Fig. 4E). To validate this finding with an independent method, we added non-labelled chymase to the cells, and then used an anti-chymase antibody to localize the protease by immunocytochemical analysis. As seen in Fig. 4F, these analyses confirmed the presence of chymase in the cell interior of the fibroblasts. Collectively, these findings indicate that chymase is taken up by primary airway fibroblasts, introducing the possibility that chymase can have a direct impact on intracellular targets.

Chymase Treatment Impairs Mitochondrial Respiration in Primary HLFs

Given the established roles of heat shock proteins, particularly HSP27 and HSP60, in maintaining mitochondrial protein homeostasis and supporting mitochondrial respiration [26], we next investigated whether chymase-induced reductions in these proteins impact mitochondrial function in primary HLFs. To assess this, we performed a Seahorse XF Real-Time Cell Metabolic Analysis to directly evaluate mitochondrial respiration following chymase treatment. Chymase-treated HLFs exhibited an overall reduction in mitochondrial respiratory activity compared to untreated controls (Fig. 5A). Specifically, the basal oxygen consumption rate (OCR) was significantly decreased in chymase-treated cells (Fig. 5B). While not reaching statistical significance, a consistent decrease was also observed in ATP-linked respiration (Fig. 5C) and maximal respiratory capacity (Fig. 5D). In addition, treatment of the HLFs with chymase led to a reduced rate of glycolysis, as measured by the extracellular acidification rate (ECAR) (Fig. 5E). Together, these findings suggest that chymase has an overall suppressive impact on the basal metabolism of HLFs. All Seahorse assay results were normalized to total protein content to control for potential differences in cell numbers.Fig. 5. Chymase reduces the metabolic activity and mitochondrial respiration of primary HLFs without affecting viability or mitochondrial protein abundance. (A) Oxygen consumption rate (OCR) profiles of primary HLFs, untreated or treated with 5 nM chymase for 24 h. Sequential injections include Oligomycin A (ATP synthase inhibitor), FCCP (protonophore), and a Rotenone/Antimycin A mixture (Complex I/III inhibitors). Data represent means ± SD. (B) Quantification of basal respiration in HLFs following 24-hour treatment with 5 nM chymase. Data are presented as means + SD, pooled from three independent experiments. *P < 0.1 (paired Student’s t-test). (C) Quantification of ATP-linked respiration in HLFs following 24-hour treatment with 5 nM chymase. Data are presented as means + SD, pooled from three independent experiments. (D) Quantification of maximal respiration in HLFs following 24-hour treatment with 5 nM chymase. Data are presented as means + SD, pooled from three independent experiments. (E) Extracellular acidification rate (ECAR) profiles of primary HLFs, untreated or treated with 5 nM chymase for 24 h. Sequential injections include Oligomycin A, FCCP, and a Rotenone/Antimycin A mixture (Complex I/III inhibitors). Data represent means ± SD. (F) Flow cytometric assessment of HLF viability and apoptosis after 24-hour treatment with 5 nM chymase, stained with Annexin V and DRAQ7. Populations include viable (Annexin V−/DRAQ7−), early apoptotic (Annexin V+/DRAQ7−), and late apoptotic/necrotic (Annexin V+/DRAQ7+) cells. Data are shown as means ± SEM. (G) Quantification of EdU-positive proliferating HLFs after 24-hour treatment with 5 nM chymase, relative to untreated cells. Data are presented as means ± SEM. (H) Quantification of HLF metabolic activity assessed by an independent PrestoBlue assay following 24-hour treatment with 5 nM chymase. Data are presented as means ± SEM. ****P ≤ 0.0001. (I) Representative Western blots showing the protein levels of subunits for oxidative phosphorylation (OXPHOS) complexes (Complex I, II, III, IV, V) in primary HLFs after 24-hour treatment with 5 nM chymase. (J) Total protein levels of COXIV, MTCO1, and MTCO2 are shown. GAPDH serves as a loading control

To ensure that the observed reduction in mitochondrial function was not secondary to effects on cell viability or proliferation, we conducted Annexin V/DRAQ7 staining (Fig. 5F) and EdU proliferation assays (Fig. 5G) using flow cytometry. Consistent with our previous findings [18], chymase treatment did not affect HLF viability or proliferation. To further validate the observed decline in mitochondrial activity using an independent method, we employed the PrestoBlue metabolic assay. This assay measures the reduction of resazurin to resorufin, a process dependent on mitochondrial metabolic activity. Consistent with the Seahorse results, chymase-treated HLFs showed significantly reduced PrestoBlue fluorescence, in agreement with impaired mitochondrial function (Fig. 5H).

We then examined whether chymase directly affects mitochondrial protein complexes. Western blot analysis of representative components from mitochondrial complexes I–V revealed no significant degradation or loss of expression following chymase treatment (Fig. 5I, J; quantification shown in Suppl. Figure 1T-V), suggesting that chymase impairs mitochondrial function through signalling and regulatory mechanisms rather than by direct proteolytic damage to the electron transport chain.

Discussion

Airway remodelling is both a consequence and driver of asthma exacerbations. Understanding the regulatory mechanisms underlying this process is therefore crucial. In our previous investigations we showed that mast cell chymase has a profound impact on primary HLF function [18], and the present study was undertaken to provide increased insight into the underlying mechanism.

PARs are known to be major targets for a wide range of proteases including mast cell proteases [27, 28], the latter exemplified by the activation of PAR2 by tryptase [29–31] and the known capacity of chymase to cleave and thereby activate PAR1 [18]. Hence, we hypothesized that the effects of chymase on the HLFs could be mediated via effects downstream of PAR activation, a scenario which was supported by the demonstrated high expression of PAR1 and PAR3 by the HLFs. However, exposure of the HLFs to agonists of either PAR1 or PAR3 did not replicate the effects induced by chymase, arguing against that the effects of chymase on HLFs are mediated by PAR ligation.

To search for alternative scenarios of chymase action we then evaluated if chymase could act on the HLFs through mechanisms related to modification of the ECM, based on the known ability of chymase to regulate various ECM-related events [32]. These analyses revealed that proteoglycans of syndecan and glypican type were expressed by the HLFs. Moreover, syntenin-1 (SDCBP), a syndecan-binding adaptor protein [33], was highly expressed by the cells, and was shown to be induced by exposure of the cells to chymase. The latter led us to investigate whether syntenin-1 and components known to participate in signalling via syntenin-1 were affected by chymase treatment. Syntenin-1 is known to interact with  \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha{V}\beta{3}$$\end{document} integrins [34], and such cell surface integrins are known to interact with ECM components including fibronectin [34–36]. Hence, we assessed whether such a fibronectin/integrin/syntenin-1 axis could be affected by chymase. Indeed, chymase was shown to cause extensive degradation of HLF-produced fibronectin. Chymase was also shown to cause a reduction in the levels of both ITGAV and ITGB3, i.e., both components of the integrin  \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha{V}\beta{3}$$\end{document} heterodimer, as well as release of syntenin-1 into the extracellular space. Importantly, all of these effects were abrogated in the presence of a chymase inhibitor, suggesting that these effects were dependent on the enzymatic activity of chymase.

To assess whether such effects can translate into intracellular signalling events, we investigated whether signalling molecules downstream of syntenin-1 were affected by chymase treatment. Syntenin-1 is known to be linked to signalling via a FAK/Src/Akt axis [23, 37] and we thus assessed whether chymase could have an impact on these signalling molecules. Indeed, our findings reveal that the phosphorylation of FAK, Src and Akt was markedly suppressed after chymase treatment. Again, all of these effects were abolished by applying a chymase inhibitor, indicating that the effects on these signalling molecules were dependent on the proteolytic activity of chymase. Altogether, these findings are compatible with a scenario in which chymase, by proteolytic action on fibronectin and cell surface integrins, affects syntenin-1, which in turn affects major signalling events downstream of syntenin-1 (Fig. 2C). Most likely, this can lead to major effects on HLF functionality, such as those observed here and in our previous study [18].

One notable aspect of syntenin-1/Akt signalling axis is the regulation of cell migration [36], and we thus assessed whether chymase could affect the migratory capacity of the HLFs. Indeed, chymase treatment was shown to markedly suppress the migration of HLFs in a wound closure assay, in agreement with effects on intracellular signalling events involved in the regulation of cell migration.

To provide enhanced insight into the effects of chymase on intracellular signalling events, we used an unbiased phospho-array approach. By using this approach, we identified additional intracellular events mediated by chymase, including suppressed phosphorylation of CREB and reduced levels of c-Jun protein. It is thus conceivable that chymase-mediated effects on these signalling molecules could contribute to our noted effects of chymase on the morphology, transcriptome and secretome of HLFs. In agreement with this notion, c-Jun has previously been linked to syntenin-1 signalling [23]. More strikingly, we noted that HSP27 was partially degraded in cells treated with chymase. Importantly, HSP27 degradation was completely abolished in the presence of a chymase inhibitor, indicating that the HSP27 degradation was dependent on the enzymatic activity of chymase. The mechanism by which HSP27 is degraded in response to chymase is intriguing. Considering that chymase was added to the extracellular milieu, a direct proteolytic effect of chymase on HSP27 would require that chymase is taken up by the cells and that chymase exerts proteolytic activity within the intracellular milieu. In line with such a scenario, we show that chymase is in fact taken up by the HLFs, introducing the possibility that chymase can execute direct proteolytic events within the interior of the HLFs. However, we cannot exclude the possibility that the HSP27 degradation is executed by indirect effects initiated by chymase.

We also asked whether the noted effects of chymase on the functional properties of HLFs could be reflected by effects on the cellular metabolism. These analyses revealed that chymase markedly suppressed the basal metabolism of the HLFs, as measured both by the PrestoBlue assay, and by measurements of mitochondrial respiration by Seahorse technology. Notably though, no effects on the levels of mitochondrial markers were observed, suggesting that the effect of chymase on mitochondrial metabolism is not mediated by compromising the integrity of mitochondrial proteins. It was also notable that the chymase-mediated effects on basal metabolism were not associated with induction of cell death or with effects on HLF proliferation. We cannot with certainty explain the mechanism by which chymase affects the metabolic parameters in HLF. However, considering that heat shock proteins including HSP27 have a role in stabilizing mitochondrial function [26], we may propose that the chymase-catalysed degradation of heat shock proteins such as HSP27 might impose effects on the mitochondria such that their metabolic capacity is suppressed. However, further investigations are warranted to outline the precise molecular mechanisms underlying the effect of chymase on mitochondrial function.

Altogether, this study reveals novel insight into how chymase can impact on primary HLFs, by identification of signalling events that are triggered by treatment of the cells with chymase. However, a weakness of the study is that only fibroblasts from healthy donors were used. Hence, to provide further support for the findings reported here, it will be important to study to what extent the present observations can be translated into effects on lung fibroblasts derived from asthmatics. It will also be important to establish whether interaction between chymase and lung fibroblasts can be verified in vivo in an asthma context, through the analysis of tissue samples taken from healthy donors vs. asthmatics. Another key extension of the present findings will be to investigate whether signs of chymase action can be detected in vivo, in samples taken from asthmatic vs. healthy subjects. For example, it would of interest to assess whether aberrant fibroblast morphology can be seen in the vicinity of degranulated, chymase-positive mast cells, and whether increased fibronectin and/or HSP27 degradation can be seen at sites in which chymase has been released into the tissue microenvironment.

Noteworthy, several previous studies, both based on animal experiments and on clinical investigations, have indicated that mast cell chymase has a beneficial impact on asthma.

[38–42]. The exact mechanism behind these effects is not clear, but based on the present investigation we may propose that chymase might serve such an anti-asthma impact, at least partly, by executing effects on airway fibroblasts.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (PDF 3.43 MB)

Supplementary Material 2 (PDF 331 KB)