Key Points
TRPV4 mediates the host defense and lung injury response to bacterial pneumonia.
TRPV4 mediates these effects through MAPK switching from JNK to p38 through DUSP1.
Abstract
Mechanical cell–matrix interactions can drive the innate immune responses to infection; however, the molecular underpinnings of these responses remain elusive. This study was undertaken to understand the molecular mechanism by which the mechanosensitive cation channel, transient receptor potential vanilloid 4 (TRPV4), alters the in vivo response to lung infection. For the first time, to our knowledge, we show that TRPV4 protects the lung from injury upon intratracheal Pseudomonas aeruginosa in mice. TRPV4 functions to enhance macrophage bacterial clearance and downregulate proinflammatory cytokine secretion. TRPV4 mediates these effects through a novel mechanism of molecular switching of LPS signaling from predominant activation of the MAPK, JNK, to that of p38. This is accomplished through the activation of the master regulator of inflammation, dual-specificity phosphatase 1. Further, TRPV4’s modulation of the LPS signal is mechanosensitive in that both upstream activation of p38 and its downstream biological consequences depend on pathophysiological range extracellular matrix stiffness. We further show the importance of TRPV4 on LPS-induced activation of macrophages from healthy human controls. These data are the first, to our knowledge, to demonstrate new roles for macrophage TRPV4 in regulating innate immunity in a mechanosensitive manner through the modulation of dual-specificity phosphatase 1 expression to mediate MAPK activation switching.
Introduction
Overwhelming bacterial pneumonia affects ∼1.2 million people per year (1). Of these, approximately one in three develop acute lung injury (ALI)/acute respiratory distress syndrome (ARDS) with a total mortality of 40,000–70,000 per year (1–3). The pathogenesis of ARDS from an infectious stimulus is complex. Infection-associated ARDS is a consequence of endothelial and alveolar epithelial injury followed by recruitment and accumulation of inflammatory cells in the injured alveolus (4–7). To clear bacteria effectively, macrophages undergo cytoskeletal rearrangements that depend on their interaction with the lung tissue matrix to drive the phagocytic process (8–11). Studies show that proinflammatory cytokines and chemokines secreted by macrophages also play a key role in the pathogenesis of ALI/ARDS and determine mortality (12–14). As such, macrophages are important effector cells for bacterial clearance and the tissue inflammatory response (15). The key biological processes that underlie lung injury and their molecular drivers remain an active area of investigation, given the lack of effective molecular pathway–targeted therapeutics (16).
The innate immune system requires precise and finely tuned regulation of multiple signals to either maintain homeostasis or implement host defense (17, 18). LPS and Gram-negative bacteria initiate immune cell activation through TLR4 signaling using several intracellular signaling pathways, including NF κ L chain enhancer of activated B cells, IFN regulatory factor 3, and MAPKs pathways (19). MAPKs include p38, ERK, and JNK, all of which have both unique and redundant functions in inflammation (20). Persistent and exuberant MAPK activation can potentially skew the inflammatory response toward collateral tissue damage, as seen in lung injury/ARDS (20). As such, MAPK activity is limited by the regulated phosphatase action of dual-specificity phosphatase (DUSP) and dual-specificity serine threonine phosphatases (MKPs) (21, 22). However, the discrete molecular pathways by which 1) the mechanical properties of the extracellular matrix are sensed and transduced, 2) the inflammatory and mechanical signals are integrated, and the 3) MAPK specificity and 4) downstream pathophysiological consequences are poorly understood.
Studies from our laboratory and others show that particle binding to macrophages induces calcium transients that are required for phagocytosis (23–26). Intracellular calcium is tightly regulated in a spatiotemporal manner through a system of membrane pumps and ion channels, such as transient receptor potential vanilloid 4 (TRPV4) (27). TRPV4 is a ubiquitously expressed, cation-permeable channel that is sensitized and/or activated by both chemical (5,6-epoxyeicosatrienoic acid and 4 α-phorbol 12,13-didecanoate) and physical stimuli (temperature, stretch, and hypotonicity) (28–31). TRPV4 also possesses intracellular NH2 and COOH tails that can interact with intracellular signaling proteins to modulate important biological functions (32). We have previously shown that the mechanosensitive channel, TRPV4, mediates LPS-induced macrophage phagocytosis and alters the cytokine response in vitro (23). However, the role of TRPV4 in infection-associated lung injury and the molecular signaling mechanism by which TRPV4 exerts its effect remain to be elucidated. This study is the first, to our knowledge, to uncover that TRPV4 enhances bacterial clearance and changes cytokine production to protect the lung from infection-associated injury and the first, to our knowledge, to identify the molecular signaling mechanism that mediates these effects.
Materials and Methods
Abs and reagents
The following primary Abs were purchased: intracellular TRPV4 (Alomone Labs, Jerusalem, Israel), extracellular TRPV4 (Alomone Labs), anti-phospho p38 (Thr180/Tyr182; Cell Signaling Technology, Danvers, MA), anti-p38 (Santa Cruz Biotechnology, Dallas, TX), anti-phospho JNK (Cell Signaling Technology), anti-JNK (Cell Signaling Technology), anti-phospho ERK (Santa Cruz Biotechnology), anti-ERK (Cell Signaling Technology), anti-phospho MAPKAPK2 (MK2) (Cell Signaling Technology), anti-MK2 (Cell Signaling Technology), anti-phospho MKK3/MKK6 (Cell Signaling Technology), anti-MKK3 (Cell Signaling Technology), anti-MKK6 (Cell Signaling Technology), anti-GAPDH (Fitzgerald Industries International, Acton, MA), anti-DUSP1/MKP1 (Santa Cruz Biotechnology), anti-CD45 (BD Biosciences), and purified rabbit IgG from mouse serum (Sigma-Aldrich, St. Louis, MO). Secondary Ab to rabbit was obtained from The Jackson Laboratory, and rat Alexa Fluor 594 was obtained from Life Technologies (Carlsbad, CA). The following selective inhibitors of TRPV4 (HC067047 [HC]), p38 MAPK (SB203580), and JNK (SP600125) were purchased from Sigma-Aldrich. The p38 MAPK inhibitor (BIRB796), caspase inhibitor (ZVAD-FMK), TLR4 inhibitor (CLI-095), and DUSP1 (BCI) were obtained from MilliporeSigma (Burlington, MA). Escherichia coli LPS 0111:B4 (LPS) for the in vitro experiments was obtained from Sigma-Aldrich. The GFP Pseudomonas aeruginosa was purchased from American Type Culture Collection (Manassas, VA), and the clinical strain P. aeruginosa (PAM 57-15) was obtained from Dr. T. Bonfield at Case Western Reserve University.
Chronic P. aeruginosa pneumonia infection model
P. aeruginosa PAM57-15 (mucoid clinical isolate) and GFP P. aeruginosa (American Type Culture Collection) were used at a sublethal dose of 105−6; and sterile culture embedded into agarose beads was instilled intratracheally (IT) in TRPV4 null mice and age-matched female congenic wild-type (WT) C57/BL6 (The Jackson Laboratory) adapted from published protocols with assistance from Lung Infection and Inflammation Modeling Core at Case Western Reserve University directed by T. Bonfield (33, 34). The original strain of the TRPV4 null mice were given to us by D. Zhang (Medical College of Wisconsin) (35). TRPV4 null mice were created from a 129/Sv chromosome genomic library (Genome Systems, St. Louis, MO) identified by the TRPV4 cDNA probe. The vector containing PGK neocassette was transfected into RW4 embryonic stem cells (Genome Systems) and chimeras bred with C57BL/6 females. Heterozygous mice were crossed to achieve TRPV4 null homozygotes. Genotyping via Southern blot or PCR was used to identify TRPV4 homozygotes for experimental use (36). In our laboratory, we have further backcrossed to C57BL/6 animals to maintain the background. Briefly, P. aeruginosa culture grown in the log phase and added to agarose/mineral oil solution that was cooled down sequentially. The P. aeruginosa–laden or sterile bead controls were separated from mineral oil via a separatory funnel. Mice were anesthetized with pentobarbital (50 mg/kg), and P. aeruginosa or sterile beads (40 μl) were instilled IT. These methods are identical to those previously published by the Lung Infection and Inflammation Modeling Core at Case Western Reserve University (33, 34). Total bacterial concentration instilled was measured pre– and post–bead preparation (CFU/ml) as published (33, 34). Lung bacterial concentration (CFU/ml) was measured by culturing whole lung lavage fluid and whole lung homogenates at sacrifice during acute infection (day 3). Lung injury was measured by inflammatory cell infiltration, total protein, and histologic percentage consolidation as described previously (37).
Study approval
All animal protocols were performed as approved by the Cleveland Clinic Institutional Animal Care and Use Committee. Healthy controls (nonsmokers with normal pulmonary function tests, six individuals) were recruited, informed consent was obtained per the Cleveland Clinic Institutional Review Board no. 16–343, and samples were deidentified.
Cell culture, small interfering RNA transfection, nuclear and cytoplasmic fractionation, Western blot analysis, and cytokine measurement
Deidentified monocyte-derived and alveolar macrophages were obtained and maintained in RPMI 1640 5% FBS or human serum plated on tissue culture-treated plastic. Primary murine bone marrow–derived macrophages (BMDMs) and alveolar macrophages were harvested from 8- to 12-wk-old C57BL/6 WT or TRPV4 null mice or TLR4 null mice provided by Dr. R. Fairchild (Department of Inflammation and Immunity, Lerner Research Institute). BMDMs were differentiated in recombinant mouse M-CSF (50 ng/ml; PeproTech) as previously published (23). BMDMs and alveolar macrophages were plated on fibronectin-coated (10 mg/ml), tissue culture-treated plastic or polyacrylamide hydrogels with varying stiffness (1, 8, or 25 kPa) (Matrigen, Bream, CA). Cells were plated on tissue culture-treated plastic unless otherwise specified. Cells were treated with LPS (100 ng/ml) alone or with LPS and pretreated for 1 h with TRPV4 inhibitor (HC), p38 inhibitor (SB203580, BIRB796), MK2 inhibitor (PF3644022), or JNK inhibitor (SP600125) for a total of 6–24 h. Primary isolates of alveolar macrophages obtained from whole lung lavage were purified by adherence and cultured in DMEM/10% FBS as previously described (40). p38 MAPK expression was downregulated by transfecting BMDMs with p38-specific mouse small interfering RNA (siRNA) duplexes or scrambled siRNA controls for 96 h (GE Healthcare Dharmacon) using electroporation, as previously published (2323P. aeruginosa and WT BMDMs ± LPS ± SB203580 ± SP600125.
In vivo and in vitro macrophage phagocytosis
Cytospin preparations were performed for each lavage from WT and TRPV4 KO mice after P. aeruginosa infection per a previously published protocol (23). Neutrophils and macrophages were distinguished by nuclear shape (polymorphonuclear cell: multilobular nucleus; macrophage: single concentric nucleus). The attached cells were fixed with 4% paraformaldehyde, permeabilized by 0.1% Triton X-100, and blocked with 10% normal goat serum. To examine the primary TRPV4 expressing whole lung lavage cell, whole lung lavage cells were incubated with anti-TRPV4 followed by Alexa Fluor–conjugated secondary Ab. To determine the primary phagocytic cell of P. aeruginosa, the whole lung lavage cells were incubated with anti-CD45 (membrane), DAPI (nucleus), and anti-GFP (P. aeruginosa). Phagocytosis was analyzed by quantifying the number of P. aeruginosa per cell using ImageJ software. Digital fluorescent microscopic images were analyzed by ImageJ software. BMDMs were stimulated in vitro then incubated with heat-inactivated E. coli (K-12 strain, Vybrant Phagocytosis Assay Kit, V-6694; Molecular Probes) for 2 h per manufacturer’s instructions at a concentration and time course previously titrated as published (23). In selected conditions, p38 MAPK inhibitor (SB203580 or BIRB796) or MK2 inhibitor (PF3644022) was added 1 h before LPS stimulation. Nonopsonized phagocytosis was measured as fluorescence intensity in the FlexStation system (Molecular Devices) as published (23). A dose response of the p38 MAPK inhibitors and MK2 inhibitors determined its maximal effects, which were noted at 10, 10, and 20–50 μM, respectively. Opsonized phagocytosis was measured by the uptake of IgG-coated latex beads (F8853, 2-μm beads; Molecular Probes) and imaged via confocal (Leica TCS-SP8-AOBS inverted confocal microscope) as previously published.
Statistical analysis
All data are presented as means ± SEM, unless otherwise specified. Comparison of data from two groups was performed with the Student t test. Comparing change scores of more than two groups was performed via ANOVA followed by Dunnett test or Student–Newman–Keuls. Significance was accepted at the p ≤ 0.05 level.
Results
TRPV4 function protects against lung injury in a murine P. aeruginosa pneumonia model
To determine the significance of TRPV4 to bacterial clearance and lung injury, sterile beads or a clinical strain of P. aeruginosa (PAM 57-15) was embedded in agarose beads was instilled IT into WT and TRPV4 KO mice (33). Mice were examined in the lung injury phase, day 2–3 after inoculation, as published (33). TRPV4 KO mice have increased lung injury as measured by 2-fold increases in inflammatory cell infiltration (Fig. 1A, *p < 0.05), increases in total protein in whole lung lavage (Fig. 1B, *p < 0.05), and a 2.5-fold increase in lung parenchymal consolidation (Fig. 1G, *p < 0.05) as compared with WT mice after P. aeruginosa administration (n = 20 per group). In addition, TRPV4 KO mice exhibit impaired lung bacterial clearance as measured by a 1–2 log increase in CFU/ml of P. aeruginosa retained in the lung (Fig. 1C, *p = 0.012) as compared with WT mice. Sterile beads alone do not explain the differential inflammatory cell infiltration or cytokine secretory effect in WT versus TRPV4 KO mice we observed after P. aeruginosa–laden beads (Fig. 1).
TRPV4 function protects against lung injury in a murine P. aeruginosa pneumonia model. Sterile beads or P. aeruginosa (PA, PAM57-15) were instilled IT in TRPV4 KO and age-matched female congenic WT mice with bronchoalveolar lavage (whole lung lavage), and tissue harvest was performed at day 3 (injury phase). TRPV4-deleted mice (TRPV4 KO) have greater (A) inflammatory cell infiltration and (B) Whole lung lavage total protein compared with WT (*p < 0.05). (C) TRPV4 KO mice have decreased bacterial clearance as measured by retained bacterial CFU in the combined whole lung lavage/lung homogenate as compared with WT (*p = 0.012). TRPV4 KO mice have greater whole lung lavage content of (D) IL-6 (*p = 0.028), (E) CXCL2 (MIP-2) (*p = 0.049), and (F) CXCL1 (KC) (*p = 0.009) than WT control by ELISA. (G) TRPV4 KO H&E lung sections show greater parenchymal inflammatory cell infiltration (quantified as percentage of lung consolidation) as compared with WT. n ≥ 5 per sterile bead group and n = 20 per P. aeruginosa group on day 2–3. The box plots (B–F) indicate the 25th–75th percentile for each measure. The error bars denote maximum and minimum values (5th–95th percentile). The horizontal white line denotes the median value. * denotes WT versus TRPV4 KO.
The balance of pro- to anti-inflammatory cytokines has been shown to be a critically important driver of lung inflammation (13, 14, 41). Interestingly, administration of P. aeruginosa to TRPV4 KO mice results in increases in the levels of several proinflammatory cytokines in whole lung lavage fluid, including IL-6 by 100 ± 6%, (Fig. 1D, *p = 0.028), CXCL2 (MIP-2) by 46 ± 9% (Fig. 1E, *p = 0.049), and CXCL1 (KC) by 88 ± 12% (Fig. 1F, *p = 0.009) as compared with that of WT mice. With the anti-inflammatory cytokine, IL-10, secretion was below the level of detection in WT and TRPV4 KO mice (data not shown). Taken together, these data demonstrate that TRPV4 is an important driver of bacterial clearance and regulator of proinflammatory cytokine secretion and, as such, functions to enhance host defense and limit acute inflammation in experimental P. aeruginosa pneumonia.
TRPV4 mediates clearance of P. aeruginosa by macrophages
We have previously shown that TRPV4 is important for the LPS-induced phagocytosis of opsonized and nonopsonized particles in vitro in macrophages; however, the complexity of the process of live bacterial clearance in the lung precludes direct extrapolation from in vitro studies (23). In this study, we now test if TRPV4 in macrophages plays a role in bacterial clearance (phagocytosis) in a clinically relevant in vivo model of bacterial pneumonia. First, P. aeruginosa–laden bead instillation induces TRPV4 expression to a 3-fold greater extent in alveolar macrophages compared with neutrophils in WT mice (Fig. 2A, 2B). To then determine if alveolar macrophages were the primary phagocytic cell, we analyzed cytospins of whole lung lavage by immunofluorescence after P. aeruginosa in mice. The administered P. aeruginosa is found predominantly in morphologically identifiable macrophages from WT mice (Fig. 2C). The capacity of alveolar macrophages (derived from whole lung lavage) to phagocytose IT-administered P. aeruginosa was largely lost in alveolar macrophages from TRPV4 KO mice (WT mice, 30% versus TRPV4 KO mice, 4%; *p < 0.05) (Fig. 2C, 2D, *p < 0.05). To further confirm the cellular identity and quantify the amount of P. aeruginosa phagocytosis, cells derived from collagenase-digested lung from infected animals were sorted by immunophenotyping into macrophages (CD45+, F4/80+, and CD64+) and neutrophils (CD45+ and Ly6G+) by flow cytometry (Fig. 2E). As above, the capacity of whole lung macrophages from TRPV4 KO mice to phagocytose P. aeruginosa was largely lost (WT mice, 8% versus TRPV4 KO mice, <0.1%; *p = 0.035), whereas phagocytosis of P. aeruginosa by neutrophils was barely detectable in either group (Fig. 2E, 2F). Collectively, these results show, for the first time to our knowledge, that TRPV4 is predominantly expressed in alveolar macrophages and that TRPV4 mediates macrophage phagocytosis of P. aeruginosa in vivo.
TRPV4 mediates clearance of P. aeruginosa by macrophages. WT and TRPV4 KO mice were IT administered ± GFP P. aeruginosa for 3 d. Representative confocal images of whole lung lavage cytospins of macrophages (open arrowhead) and neutrophils (filled arrowhead) in WT mice given IT sterile beads or GFP P. aeruginosa after immunofluorescence with (A) TRPV4 extracellular Ab (green, TRPV4) and (C) anti-GFP (green, GFP P. aeruginosa; anti-CD45, red; and DAPI, blue). (B and D) Quantification of (A) and (C) *p < 0.05, #p < 0.05, percentage of WT versus KO. (E) Flow cytometry of macrophage populations (+F4/80, CD64) from collagenase digested lung ± GFP P. aeruginosa from WT and TRPV4 KO mice. Cell debris was excluded on an FSC-A/SSC-A plot, and cell aggregates were excluded on a FSC-A/FSC-H plot. Viable cells were selected on a DAPI/SSC-A plot. Pseudocolor plots for CD45, neutrophil, and macrophage gating thresholds are shown. Gate boundaries for CD45-positive leukocytes and F4/80-positive macrophages were set using fluorescence minus one controls. Immunofluorescence with anti-GFP (green) performed on cytospins. (F) Quantification of the percentage of cell phagocytosis (*p = 0.035). n = 20 per group. All images original magnification ×63. Scale bar, 10 μm. * and # denote WT versus TRPV4 KO. macs, macrophages; PMNs, neutrophils.
TRPV4 mediates the molecular switch activation of p38 MAPK and JNK through DUSP1
To define the molecular pathway by which TRPV4 mediates P. aeruginosa/LPS–stimulated macrophage phagocytosis, we investigated putative candidates based on prior work implicating the MAPK pathway in macrophage responses to both mechanical signals and LPS (42). Confirming prior work, LPS rapidly induces the activation of p38, JNK, and ERK (Fig. 3A). Interestingly, the activation/phosphorylation of p38 MAPK after LPS stimulation is reduced by 2-fold in TRPV4 KO as compared with that from WT BMDMs (Fig. 3A, 3B, *p < 0.001). In contrast, LPS-induced JNK phosphorylation is enhanced by 3.5-fold in TRPV4 KO as compared with that from WT BMDMs (*p = 0.027), although ERK phosphorylation is not different (Fig. 3A, 3C, 3D). Because loss of TRPV4 decreases p38 MAPK activation after LPS and p38 MAPK is known to phosphorylate and thereby activate MK2, we next determined if MK2 is similarly downstream of TRPV4. MK2 activation is similarly decreased by 2-fold in TRPV4 KO BMDMs upon LPS exposure compared with that of WT (Fig. 3E, 3F, *p < 0.05). Taken together, our data demonstrate that TRPV4 mediates molecular switching from JNK to p38 activation upon LPS exposure in BMDMs.
p38 MAPK and JNK are differentially regulated by TRPV4 after LPS or P. aeruginosa. BMDMs were incubated ± LPS as above for the indicated time cultured on tissue culture-treated plastic, and cells were lysed and analyzed by immunoblot for (A) phosphorylated and total p38, ERK, and JNK compared with WT BMDMs (whole cell lysate). Band density quantified from immunoblot (n = 3–6) for (B) p-p38/total p38 (*p < 0.001), (C) p-ERK/total ERK, and (D) p-JNK/total JNK (*p = 0.027). (E) Representative immunoblot for phosphorylated and total MK2 in WT versus TRPV4 KO BMDMs. (F) Band density quantified from immunoblot (n = 5, *p < 0.05). (G) Representative immunoblot for phosphorylated and total MKK3/MKK6. (H) Band density quantified from immunoblot (n = 3). (I) Representative immunoblot for phosphorylated and total p38 in homogenized mouse lung after sterile or P. aeruginosa beads (3 d). (J) Band density quantified from p38 immunoblot (n = 6) (*p < 0.001). * denotes WT versus TRPV4 KO.
To determine the molecular level of cross-talk between the LPS–TLR4 and the TRPV4 signals, we examined the kinases immediately upstream of p38 (MKK3 and MKK6). LPS-induced MKK3 and MKK6 activation to an equal extent in both WT and TRPV4 KO macrophages (Fig. 3G, 3H), indicating the LPS and TRPV4 signals converge downstream of MKK3/MKK6 at the level of p38 MAPK. As expected, deletion of TLR4 abrogated p38 MAPK activation in response to LPS (see Supplemental Fig. 1A). To determine if TRPV4 plays a role in p38 MAPK activation in vivo after bacterial infection, p38 MAPK phosphorylation was measured in whole lung homogenates after intratracheal administration of P. aeruginosa in mice. As with LPS in vitro, p38 MAPK phosphorylation is similarly decreased in TRPV4 KO murine lungs compared with WT (Fig. 3I, 3J, *p < 0.001). Taken together, our data demonstrate that the TRPV4-initiated signal is integral to and converges with the LPS–TLR4 signal through molecular switching from JNK activation to predominantly p38 activation.
The MAPK pathway is a key response node and therefore must be finely tuned for optimal homeostatic cell function as well as for responses to external cues (20, 43–45). One such negative regulatory mechanism occurs through the action of a family of DUSP/MKPs (18, 22). Hence, we investigated the role of DUSP1 on TRPV4-mediated JNK to p38 MAPK molecular switching after LPS. First, DUSP1/MKP1 protein expression is reduced by 3-fold in BMDMs from TRPV4 KO mice as compared with that from WT BMDMs (Fig. 4A, 4B, *p < 0.05). Furthermore, inhibition of DUSP1 (BCI 5 μM) in WT BMDMs after treatment with LPS selectively increased JNK activation while having no effect on p38 activation (Fig. 4C–E, *p = 0.004). Thus, TRPV4 functions to enhance the LPS induction of DUSP1, which then selectively dephosphorylates (deactivates) JNK.
TRPV4-mediated p38 MAPK and JNK molecular switch are regulated by DUSP1. WT and TRPV4 KO BMDMs were incubated ± LPS as above for the indicated time, and cells were lysed and analyzed by (A) immunoblot for DUSP1 and (B) band density quantified as DUSP1/GAPDH from immunoblot (n = 4) (*p < 0.05). (C) Representative immunoblot ± DUSP1 pharmacologic inhibitor, BCI 5 μM, for p-p38/total p38 and p-JNK/total JNK in WT BMDMs. Band density quantified for (D) p-p38/total p38 or (E) p-JNK/total JNK from immunoblot (n = 4) (*p = 0.004). * denotes WT versus TRPV4 KO, + denotes ± pharmacologic inhibitor.
TRPV4 exhibits a threshold effect for matrix stiffness activation of p38 and LPS-induced phagocytosis
Macrophage phenotype is influenced by the biophysical properties of the surrounding matrix; however, the signaling pathway by which this occurs has yet to be fully elucidated (9–11). In WT BMDMs, we validated that macrophage phagocytosis is dependent on p38 MAPK as evidenced by its abrogation with two structurally distinct small molecule inhibitors of p38 (SB203580 and BIRB796) (Fig. 5A, 5B) and upon downregulation of p38α using p38α-specific siRNA (Fig. 5C). In contrast, small molecule inhibition of JNK had no significant effect on LPS-induced phagocytosis (Fig. 5F). Macrophage phagocytosis was similarly reduced upon inhibition of the p38-interacting kinase, MK2 (PF3644022), and the MAPK inhibitory phosphatase, DUSP1 (BCI) (Fig. 5D, 5E, *p < 0.05). As expected, we confirmed that TLR4 is necessary for the LPS effect on macrophage phagocytosis of both E. coli particles and IgG-coated latex beads (see Supplemental Fig. 1B–E).
Macrophage phagocytosis in response to LPS is mediated by p38 MAPK and DUSP1. BMDMs were incubated ± LPS (100 ng/ml, 24 h) ± small molecule inhibitors of p38 (SB203580 and BIRB796) and MK2 (PF3644022). Inhibition of p38 by (A) SB203580 and (B) BIRB796 induces a dose-dependent decrease in LPS-induced phagocytosis (*p < 0.05). Immunoblots show the phosphorylation of p38 in the presence of BIRB796 and MK2 in the presence of SB203580 per manufacturer-stated mechanism. (C) Downregulation of p38α by siRNA (96 h) decreases LPS-induced macrophage phagocytosis (*p = 0.019, maximal 60–70% decrease p38α protein at 96 h). Small molecule inhibition of (D) MK2 by PF3644022 and (E) DUSP1 by BCI blocks LPS-induced phagocytosis. (F) In contrast, small molecule inhibition of JNK by SP600125 had no significant effect on LPS-induced phagocytosis (*p < 0.05). n = 3 for all experiments. * denotes difference in LPS response ± inhibition or downregulation of p38/MK2.
Whereas p38 phosphorylation has been shown to be responsive to some forms of cyclic cell stretch (43, 46–48), the upstream mechanosensor and the details of the mechanical signal have yet to be elucidated. We show that the LPS and TRPV4 signals converge on the molecular switch from JNK to p38 (Fig. 3A), albeit under standard culture conditions that exhibit supraphysiologic substrate stiffness (glass/polystyrene is 103 kPa) as compared with normal lung (1 kPa). We further show that the stiffness threshold for the TRPV4-dependent LPS effects on p38 and JNK is in the range of that measured in inflamed/fibrotic lung (25 kPa) (Fig. 6A, 6B). Furthermore, under ambient matrix stiffness conditions below this threshold, as present in normal/uninflamed lung (1 kPa), activation of p38 and JNK by LPS is reduced as compared with that on injured lung stiffness and independent of the presence of TRPV4 (Fig. 6A, 6B).
p38 activation and macrophage phagocytosis in response to LPS is dependent on matrix stiffness. Phosphorylated and total (A) p38 and (B) JNK on various matrix stiffnesses in the physiologic range (1, 8, and 25 kPa) from WT versus TRPV4 KO BMDMs quantified for LPS 15 min (*p = 0.031). (C) Macrophage phagocytosis of E. coli particles ± p38 inhibition (SB, BIRB) on various matrix stiffnesses (*p < 0.05). n = 3–5 for all experiments. * denotes difference in LPS response ± inhibition of p38.
Given the observed impact of the mechanosensitive ion channel TRPV4 on LPS-induced p38 activation and our prior findings (23), we posited that inhibition of p38 MAPK might affect phagocytosis in a stiffness-dependent manner. Indeed, macrophage phagocytosis after LPS was abrogated by p38 inhibition using two mechanistically distinct, small molecule inhibitors of p38 MAPK (SB203580 [p38α/β] and BIRB796 [p38α-δ]) but only under conditions of pathophysiologic range stiffness (>8 kPa; Fig. 6C, *p < 0.05). Taken together with the selective MAPK activation pattern shown in Fig. 3, these data show that TRPV4 senses pathophysiologic range matrix stiffness and thereby modifies the LPS signal through molecular MAPK switching as regulated by DUSP1 to enhance phagocytosis. Furthermore, these data would also predict that the macrophage phagocytic responses to LPS are restrained under conditions of normal tissue integrity.
TRPV4 modulates proinflammatory cytokine production through JNK
Many studies show that the balance of pro- and anti-inflammatory cytokines contribute to the pathogenesis of ALI (14, 41). This study demonstrates that TRPV4 suppresses JNK activation in response to LPS (Fig. 3), and JNK has been implicated in the pathogenesis of ALI (49, 50). Thus, we next sought to determine if the observed increase in IL-6, CXCL2 (MIP-2), and CXCL1 (KC) in our P. aeruginosa lung injury model in TRPV4 KO mice (Fig. 1) was due to activation of the JNK pathway. First, we confirmed that TRPV4 KO macrophages had increased secretion of IL-6 (6404.804 pg/ml versus 4855.1 pg/ml), CXCL2 (7533.2 pg/ml versus 4346.69 pg/ml), and CXCL1 (1298.11 pg/ml versus 891.90 pg/ml) compared with WT macrophages (Fig. 7A, *p < 0.05). We found that small molecule inhibition of JNK (SP600125) suppresses macrophage secretion of IL-6, CXCL2, and CXCL1 by 2-fold (Fig. 7B, #p < 0.05), whereas inhibition of p38 (SB203580) had no significant effect, indicating that TRPV4’s modulation of the LPS signal leads to skewing of cytokine secretion through the induction of JNK activation.
TRPV4 modulates proinflammatory cytokine production through JNK. BMDMs were incubated ± LPS (100 ng/ml, 24 h) ± JNK inhibitor, SP600125 (20 μM, 25 h), ± p38 inhibitor, SB203580 (10 μM, 25 h), cultured on cell culture-treated plastic, and cytokines were measured via ELISA. IL-6, CXCL2, and CXCL1 secretion ± LPS in (A) WT and TRPV4 KO BMDMs and (B) WT BMDMs ± SP600125 ± SB203580 (*p < 0.05, #p < 0.05). n = 3–5, one-way ANOVA followed by Dunnett test or Student–Newman–Keuls used for statistical analysis. * denotes WT versus TRPV4 KO, and # denotes difference in LPS response ± inhibitor.
TRPV4 mediates phagocytosis after LPS in healthy human macrophages through p38 MAPK
Cellular and animal models can only provide clues to human disease mechanisms; thus, we examined the role of TRPV4 in LPS responses of human monocyte-derived macrophages and human alveolar macrophages from healthy controls. First, TRPV4 inhibition with a small molecule (HC) abrogates the LPS effect on phagocytosis by 100 ± 8% and 50 ± 4% in monocyte-derived macrophages and alveolar macrophages from healthy controls, respectively (Fig. 8A, 8B, *p < 0.05, #p < 0.05). The TRPV4 inhibitor (HC) alone had no effect on basal phagocytosis (data not shown). Similar to the mouse, p38 MAPK activation was induced by LPS, and this effect was almost completely abrogated upon inhibition of TRPV4 (HC) in healthy control macrophages (Fig. 8C, 8D, *p < 0.05, #p < 0.05). In addition, TRPV4 expression remained unchanged with LPS (Supplemental Fig. 2). Thus, we show for the first time, to our knowledge, that TRPV4 mediates LPS-stimulated human macrophage phagocytosis and is likely associated with p38 MAPK activation in healthy human macrophages (Fig. 9).
TRPV4 mediates phagocytosis after LPS in healthy humans through p38 MAPK. Monocyte-derived and alveolar macrophages from healthy (n = 6) control subjects were incubated, and ± LPS ± TRPV4 inhibitor, HC, and phagocytosis of E. coli particles was measured in (A) monocyte-derived and (B) alveolar macrophages in healthy controls. HC alone had no effect. Representative immunoblot for phosphorylated and total p38 in (C) healthy monocyte-derived macrophages ± LPS 15 min and (D) band density quantified (*p < 0.05). One-way ANOVA followed by Dunnett test or Student–Newman–Keuls used for statistical analysis. * denotes ± LPS, and # denotes difference in LPS response ± inhibitor.
Matrix mechanical signal transduction through TRPV4 modulates the LPS signal through MAPK switching. (A) In the presence of a subthreshold mechanical signal, as seen in normal lung, TRPV4 does not influence the LPS–TLR4 signal, which results in limiting the phagocytic response to LPS, thereby maintaining lung homeostasis. (B) In the presence of an above-threshold mechanical signal, as seen with lung stiffening during injury, TRPV4 influences the LPS–TLR4 signal. We have previously published that TRPV4 regulates the stiffness-dependent responses of increased macrophage phagocytosis and cytokine secretion in response to LPS (23). We now show a molecular switch from JNK activation to predominantly p38 activation, which results in abrogation of enhanced DUSP1 expression. DUSP1 regulates the MAPK molecular switch by deactivating JNK, resulting in enhanced bacterial clearance, inhibiting proinflammatory cytokine secretion, and thereby alleviating lung injury/ARDS. This defines a novel molecular mechanism linking inflammation-induced changes in the mechanical properties of the extracellular matrix with innate immunity.
Discussion
The current work extends our prior in vitro findings by elucidating the biological consequences of TRPV4 function in clearing infection and its associated lung injury in vivo and by identifying intracellular signaling pathway by which TRPV4 mediates these effects (23). Using complementary loss of function approaches, our data reveal for the first time, to our knowledge, that the TRPV4 and the LPS–TLR4 signals converge on p38, whereby DUSP1 regulates MAPK switching from JNK to p38 activation. TRPV4 senses matrix stiffness mechanical signals over the pathophysiologic range to drive MAPK switching, which mediates macrophage phagocytosis in a matrix stiffness-dependent manner. Furthermore, JNK suppression, via the integrated TRPV4 and LPS–TLR4 converging signals, mediates control of proinflammatory cytokines (IL-6, CXCL2, and CXCL1). Importantly, TRPV4 enhances host defense (innate immune defenses) and protects the lung from injury after P. aeruginosa pneumonia by enhancing macrophage bacterial clearance and downregulating proinflammatory cytokine secretion. Findings in normal human macrophages essentially mirror the results seen in the murine cells. Taken together, these data demonstrate that TRPV4 in macrophages is required to protect the lung from bacterial pneumonia and infection-associated lung tissue injury/ARDS through its ability to modulate the LPS–TLR4 signal pathway via MAPK activation switching by DUSP1 (Fig. 9). TRPV4 provides the molecular link between matrix biophysical properties and innate immunity.
TRPV4 has been implicated in other mouse models of lung inflammation because of noninfectious causes including hydrochloric acid–induced lung injury and in lung diseases associated with changes in lung parenchymal stretch, such as mechanical ventilator–induced pulmonary parenchymal overdistension, pulmonary edema due to pulmonary venous hypertension, and most recently from our group, pulmonary fibrosis (37, 51–53). However, in direct contrast to the findings in this study, TRPV4 was found to exacerbate lung inflammation in sterile or noninfectious models of lung injury (37, 51–53). Furthermore, a recent study using a single pharmacologic inhibitor of TRPV4 showed decreased lung injury after intratracheal instillation of LPS for 24 h (54). The data presented in this study of a clinically relevant infectious model of lung injury show that TRPV4 is required to protect the lung from injury. In general support of our findings, epithelial cell TRPV4 similarly protects the lung but in a somewhat distinct, rapid direct LPS-induced lung injury model (3 h) (55). Our data reveal, for the first time to our knowledge, that TRPV4 protects the lung from injury through dual beneficial functions of enhancing alveolar macrophage–dependent bacterial clearance as well as suppressing the cytokine response to infection/LPS in macrophages (56–58).
TRPV4 can initiate cell type– and context-specific intracellular signals that depend on intracellular calcium, activation of kinases, and/or direct interaction with cytoskeletal proteins via its intracellular amino- and C-terminal tails (59–62). In this study, we uncover a novel mechanism whereby TRPV4 modulates the LPS response in macrophages via switching of MAPK activation through DUSP1. Activation of p38 MAPK has been documented to occur through both canonical (MAP2K–MKK3/6) phosphorylation of Thr180 and Tyr182 by upstream MAPKKs as well as noncanonical mechanisms in a stimulus and cell type–specific manner. Thus, our finding that TRPV4 has no effect on MKK3/MKK6 activation after LPS suggests that TRPV4 alters LPS-induced p38 activation by a noncanonical mechanism. Many extracellular cues and environmental stresses have been shown to activate MAPKs (42, 45, 46, 63–65). Specifically, the p38 MAPK homolog, Hog1, has long been identified as a key regulator of the osmotic stress response in Saccharomyces cerevisiae, and cyclic stretch activates p38 MAPK in mammalian smooth muscle cells (66). However, the upstream sensors and transducers of the p38 MAPK activation in response to the mechanical signals and the precise mechanical signals largely remain to be determined (43, 44, 47). Our data, for the first time to our knowledge, convincingly show that the stretch-activated cation channel, TRPV4, senses lung matrix mechanical signal stiffness over the pathophysiological range and thereby modulates the LPS-induced p38 MAPK activation.
p38 MAPK activation is also a key regulatory hub for macrophage innate immune functions, such as cytokine production, bacterial clearance/phagocytosis, production of inflammation-modulating molecules (e.g., inducible NO synthase, COX2), and apoptosis (42, 45, 46, 63–65). The MAPK phosphatase, DUSP1/MKP1, is ubiquitously expressed and has been identified as an important negative regulator of MAPK activity through its ability to dephosphorylate MAPKs (17, 21, 67). Our data demonstrate that the DUSP1 induction response to LPS is reduced in the absence of TRPV4 and that the downregulated DUSP1 exhibits less MAPK phosphatase activity. An individual MAPK selectively of DUSP1’s action has been shown by others after LPS (21). This downregulated phosphatase action acts selectively to enhance JNK activation in the absence of TRPV4 (Fig. 4). Thus, we conclude that TRPV4 action mediates the MAPK molecular switch that was observed (Fig. 9). Given the attendant changes of additional proinflammatory cytokines, these results may have broad applicability to other p38/JNK MAPK-mediated inflammatory processes, in which changes in tissue stiffness signals through TRPV4 and TLR4 signals (DAMPs and PAMPs) seen in tissue inflammation coexist (e.g., atherosclerosis, tissue fibrosis, malignancy) (68–70).
The current paradigm in human ARDS, as based on cellular, animal, and human studies, implicates increased levels of proinflammatory cytokines as both causal and predictive of poor outcomes (3, 12–14). The mechanism of release of key proinflammatory cytokines such as IL-6, CXCL2, and CXCL1 remains unclear. As we found no evidence of changes in plasma membrane permeability ± LPS ± TRPV4 (lactate dehydrogenase release, data not shown) as such seen with cell death; thus, other potential regulatory mechanisms such as transcriptional/translational regulation might explain our findings (71). As alveolar vessel wall stiffness increases >10-fold (from 3 to 45 kPa) early after intratracheal LPS-induced lung injury in mice, we hypothesize that the stiffness–TRPV4–p38–DUSP1 axis controls the acute macrophage activation responses seen in our model (72). We have previously published that TRPV4 regulates increased macrophage phagocytic response and proinflammatory cytokine secretion response to LPS in a stiffness-dependent manner (23). Our new data also show that the macrophage phagocytic, MAPK activation, and cytokine secretory response to LPS is downregulated when macrophages sense conditions that mimic the biophysical properties of normal lung (i.e., 1–3 kPa). This was evident in that the matrix stiffness threshold for TRPV4’s impact on LPS-induced p38 activation and p38-dependent macrophage phagocytosis was above that of normal lung (Fig. 6). Collectively, the data demonstrate the importance of the TRPV4–p38–DUSP1–JNK pathway on innate immunity under conditions of inflammation-induced changes in tissue biophysical properties.
Although we have convincingly shown through multiple complementary methods that TRPV4 protects the lung from bacterial pneumonia, there are some limitations to our study. Although we have initially focused on the MAPK pathway because of its known responsiveness to extracellular cues, the possibility that TRPV4 also modulates other LPS-initiated signals, such as NF κ L chain enhancer of activated B cells, is an area of active investigation. We demonstrate that TRPV4 in macrophages are the key overall phagocytic cell and p-p38–expressing cell after P. aeruginosa in vivo, it remains possible that TRPV4 actions through p38 MAPK in other cell types may contribute to our in vivo lung injury findings. Based on our DUSP1 inhibitor data, there is selectivity of the effect of DUSP1 inhibition on JNK. In addition, small molecule inhibitors of p38 MAPK used in this study can potentially target multiple p38 MAPK subunits as well as have off-target effects. To mitigate this limitation, we used two structurally distinct inhibitors with independent p38 MAPK binding sites and distinct mechanisms of inhibition, as well as p38α-specific siRNA to downregulate p38α, all of which yielded similar results.
In summary, our study, for the first time to our knowledge, reveals that inflammation-associated matrix stiffness sensing, through TRPV4, leads to a molecular switch from JNK to p38 MAPK activation via DUSP1 in macrophages. This molecular switch mediates macrophage phagocytosis and downregulation of proinflammatory cytokine secretion in response to LPS (Fig. 9). In vivo, we further show that TRPV4 functions to fine-tune macrophage activation in response to an infectious stimulus and thereby limit lung injury by integrating the mechanical and infectious signals. As chronic administration of a TRPV4 agonist has been shown to alleviate atherosclerosis in mice, manipulating TRPV4 function may be an adjunct to antibiotics to alleviate infection-associated lung injury (73). Our findings may have broad impact as TRPV4 and macrophage activation functions are implicated in other inflammatory lung diseases, including bronchiectasis, chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis, and granulomatous lung diseases as well as vascular and malignant diseases (74).
Disclosures
The authors have no financial conflicts of interest.
Acknowledgments
We acknowledge the assistance of the Cleveland Clinic Lerner Research Institute Imaging and Flow Cytometry Core in providing microscopy and flow cytometry services and Robert Fairchild for critically reading the manuscript.
Footnotes
This work was supported by National Institutes of Health (NIH) Grants HL133380 (to R.G.S.), HL132079 (to B.D.S.), and HL-133721 and HL-119792 (to M.A.O.). This publication was made possible by the Clinical and Translational Science Collaborative of Cleveland (Award UL1TR000439), the National Center for Advancing Translational Sciences component of the NIH, and the NIH Roadmap for Medical Research (to R.G.S.). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH. This work utilized the Leica SP8 confocal microscope that was purchased with funding from NIH Shared Instrumentation Grant 1S10OD019972-01.
The online version of this article contains supplemental material.
Abbreviations used in this article:
- ALI
- acute lung injury
- ARDS
- acute respiratory distress syndrome
- BMDM
- bone marrow–derived macrophage
- DUSP
- dual-specificity phosphatase
- FSC
- forward light scatter
- IT
- intratracheally
- KO
- knockout
- MK2
- MAPKAPK2
- siRNA
- small interfering RNA
- SSC
- side light scatter
- TRPV4
- transient receptor potential vanilloid 4
- WT
- wild-type.
- Received August 23, 2019.
- Accepted December 22, 2019.
- Copyright © 2020 by The American Association of Immunologists, Inc.
References
In this issue
