TLR4 activation induces inflammatory vascular permeability via Dock1 targeting and NOX4 upregulation

Abstract

The loss of vascular integrity is a cardinal feature of acute inflammatory responses evoked by activation of the TLR4 inflammatory cascade. Utilizing in vitro and in vivo models of inflammatory lung injury, we explored TLR4-mediated dysregulated signaling that results in the loss of endothelial cell (EC) barrier integrity and vascular permeability, focusing on Dock1 and Elmo1 complexes that are intimately involved in regulation of Rac1 GTPase activity, a well recognized modulator of vascular integrity. Marked reductions in Dock1 and Elmo1 expression was observed in lung tissues (porcine, rat, mouse) exposed to TLR4 ligand-mediated acute inflammatory lung injury (LPS, eNAMPT) in combination with injurious mechanical ventilation. Lung tissue levels of Dock1 and Elmo1 were preserved in animals receiving an eNAMPT-neutralizing mAb in conjunction with highly significant decreases in alveolar edema and lung injury severity, consistent with Dock1/Elmo1 as pathologic TLR4 targets directly involved in inflammation-mediated loss of vascular barrier integrity. In vitro studies determined that pharmacologic inhibition of Dock1-mediated activation of Rac1 (TBOPP) significantly exacerbated TLR4 agonist-induced EC barrier dysfunction (LPS, eNAMPT) and attenuated increases in EC barrier integrity elicited by barrier-enhancing ligands of the S1P1 receptor (sphingosine-1-phosphate, Tysiponate). The EC barrier-disrupting influence of Dock1 inhibition on S1PR1 barrier regulation occurred in concert with: 1) suppressed formation of EC barrier-enhancing lamellipodia, 2) altered nmMLCK-mediated MLC2 phosphorylation, and 3) upregulation of NOX4 expression and increased ROS. These studies indicate that Dock1 is essential for maintaining EC junctional integrity and is a critical target in TLR4-mediated inflammatory lung injury.

Keywords: Dock1; EC barrier integrity; Elmo1; Oxidative stress; Vascular permeability; eNAMPT.

Fig. 1.

Expression of Dock1 and Elmo1 is decreased in preclinical models of ARDS/VILI-mediated inflammatory lung injury. A. Septic shock/VILI-challenged Yucatan minipigs received placebo (PBS; n = 3) or eNAMPT-neutralizing mAb (n = 3) 2 h after onset of injury as outlined in the Methods. Lung tissues isolated from these pigs and untreated control pigs (Ctrl; n = 3) were homogenized for immunoblotting analysis of Dock1 and Elmo1 expression. B. Dock1 and Elmo1 expression was examined by immunoblotting analysis in lung homogenates of rats receiving either placebo (lanes 4–6, PBS; n = 3) or eNAMPT-neutralizing mAb (lanes 7–9, mAb; n = 3) followed by LPS/VILI and compared to control rats (lanes 1–3, Ctrl; n = 3). C. Representative lung immunoblots of Dock1 and Elmo1 expression in rat LPS-challenged rats (intratracheal LPS, 1 mg/kg) in a time course (duplicate). D/E. In pig (D) and rat (E) lung tissues, active Rac1 was isolated with PAK-GST beads. Isolated complexes were separated by Immunoblotting and probed with anti-Rac antibodies (n = 3). Tissue homogenates were probed in parallel to monitor total cellular Rac1 levels. Changes in Rac1 activity assessed by a G-GLISA are expressed in percentages of control (Ctrl) values, and each bar graph represents the mean ± SD of triplicate. *p < 0.05, LPS/VILI+PBS vs. Ctrl; #p < 0.05, LPS/VILI+mAb vs. LPS/VILI+PBS.

Fig. 2.

Reduced Elmo1 expression promotes Dock1 protein degradation. A. Human pulmonary artery endothelial cells (ECs) were treated with LPS or Thrombin (Thr) for 1 h (hr). Total cellular lysates were subjected to IgG or Dock1 immunoprecipitation (IP) and analyzed by immunoblotting with the indicated antibodies. B. ECs were transfected with siCtrl, siDock1, or siElmo1. After 72 h transfection, silencing effects of each siRNAs were analyzed by immunoblotting with the indicated antibodies (two sets of independently prepared samples). C. qRT-PCR was performed for Dock1 and Elmo1 mRNA expression after transfection of siCtrl, siDock1, or siElmo1. Data shown were generated from two sets of independently prepared samples assessed by qRT-PCR twice. Target gene expression was normalized to 18S levels. *p < 0.05, siDock1 (or siElmo1) vs siCtrl. D. Immunoblot analysis of Dock1 accumulation after MG132 treatment. ECs were transfected with siCtrl or siElmo1. After 24 h, these transfectants were exposed to 10 μmol/L MG132 for the indicated time. Total cell lysates were subjected to immunoblot analysis. E. Dock1 and Elmo1 protein expression was determined by densitometry and the levels of Dock1 and Elmo1 relative to those of actin was plotted as fold changes from untreated control (0 h).

Fig. 3.

Inhibition of Dock1 activation of Rac GTPase increases EC barrier dysfunction. EC barrier function was assessed by measurements of trans-endothelial electrical resistance (TER). ECs were pretreated with TBOPP (50 μmol/L for A, 25 μmol/L for B–D) for 2 h followed by challenge with eNAMPT (1.5 μg/mL, A), LPS (1 μg/mL, B), thrombin (0.5-unit, C), and Tysiponate (tyS1P; 1 μg/mL, D). TER was recorded continuously for 8 h to 24 h. Shown graphs are representative tracings from four experiments (mean ± SEM). *p < 0.05, treatment vs. ctrl; **p < 0.05, single treatment vs. combo.

Fig. 4.

eNAMPT-mediated ROS generation is augmented by Dock1 inhibition in lung endothelial cells. A. Immunoblot analysis of Nox4 expression. ECs were treated with 1.5 μg/mL rhNAMPT (3 h) and 10 umol/LTBOPP (16 h) Total cell lysates were subjected to immunoblot analysis as the indicated antibodies. B. ECs were transfected with siCtrl, siDock1 or siElmo1. After 72 h, Nox4 expression in these transfectants were determined by immunoblotting analysis. C. Effects of siRNA Dock1 (siDock1) on Nrf2, Hmox1 and Nqo1. Human lung ECs were transfected with siCtrl or siDock1. After 72 h transfection, total RNA extracts were used for qRT-PCR analysis. Normalized fold expression of Nrf2, Hmox1 and Nqo1 mRNA (duplicate) is shown and 18S was used for normalization. D. Immunoblots of Nox4, Nrf2 and Hmox1 expression. ECs were pretreated for 16 h with DMSO or 10 μM TBOPP and further treated with rhNAMPT for the indicated times. E. ROS measurement. Relative fluorescent intensity (RFI) was counted after dichlorofluorescein (DCF) staining of ECs after rhNAMPT, TBOPP and the combination of rhNAMPT and TBOPP and then stained for 0.5 h with 5 umol/L CM-H2DCFDA. Data are shown as mean ± SEM of triplicates. Scale-500 μm. *p < 0.05, treatment vs. Ctrl; ^#p < 0.05, eNAMPT vs. combo.

Fig. 5.

Impaired redox regulation and inflammatory signaling is linked to loss of Dock1 expression in preclinical mouse ARDS lung tissues. A. qRT-PCR analysis of IL-6 and IL-1β mRNA expression. ECs were transfected with siCtrl or siDock1 for 72 h. *p < 0.05, siCtrl vs. siDock1 (mean +/− SD, n = 4). B. hTLR4-HEK-293T cells were cotransfected with IL-6-luc, renilla-luc and Dock1 (or Elmo1) plasmids with immunoblot detection of Dock1 and Elmo1 overexpression in these transfectants (right panel). IL-6 luciferase activity was increased by LPS challenge (16 h, mean+/−SD, n = 3) but reduced in Dock1 and Elmo1 silenced cells. C. Plasma levels of IL-6 and IL-1β measured by MesoScale Discovery platform were markedly increased in LPS/VILI-challenged mice. *p < 0.05, Ctrl vs. LPS/VIL (mean +/− SD, n = 5). D. qRT-PCR analysis of mouse lung IL-6 and IL-1β, Dock1 and Elmo1 mRNA expression. *p < 0.05, Ctrl vs. LPS/VILI (mean ± SD, n = 5). E. Lung tissue homogenates were obtained from mice exposed to LPS/VILI and from untreated control mice. Representative immunoblotting images of phosphorylated proteins of NF-kB (p-NF-κB), JNK (p-JNK), p38 (p-p38), ERK (p-ERK). F. Similarly, IL-6, NAMPT, Dock1 and Elmo1 protein expression in mouse lung tissues was increased in ARDS/VILI-challenged animals (representative immunoblot shown). G. Expression of 4-Hydroxynonenal (4-HNE) adduct formation, Nox4 and Nrf2 in the lung tissues of mice challenged with LPS/VILI and controls was assessed by immunoblot analysis.

Fig. 6.

Dock1 inhibition impairs nmMLCK/MLC2-regulated lamellipodia formation. A. Human pulmonary artery endothelial cells (ECs) were pretreated with TBOPP (10 μmol/L) and then exposed to S1P (1 μmol/L) for the indicated time. Total cell lysates were used for immunoblotting analysis of phsopho-MLC2 (p-MLC2), total MLC2 and b-actin protein expression. B. ECs were transfected with non-specific scrambled sequence (siCtrl), siDock1 or siElmo1. At 72 h post transfection, transfectants were exposed to S1P (1 μmol/L) for 5 min. Total cell lysates were used for immunoblotting analysis with the indicated antibodies. C. Immunoblot of phospho-MLC kinase (p-MLCK) is shown. β-Actin was used as a loading control. Preparation of samples are described for Panel A. D. Changes in p-MLCK expression are shown with or without S1P exposure (1 μmol/L, 5 min) in ECs after 72 h post-transfection with siCtrl, siDock1 or siElmo1. β-actin was used as a loading control. E. Immunoblot analysis of phospho-MLC kinase and MLC2 expression after S1P exposure of Nrf2-depleted ECs. ECs were infected with lentiviruses of shControl (shCtrl) or shNrf2. After 7 days, shCtrl and shNrf2 transduced ECs were exposed to S1P (1 μmol/L) for 5 min. Immunoblotting studies confirmed that shNrf2 transduction depleted Nrf2 and its transcriptional target Hmox1 expression. F. Lamellipodia formation in ECs. HPAECs in chamber slides were pre-treated for 1 h with 10 μmol/L TBOPP, a specific DOCK1 inhibitor, before exposing them to S1P (5 μmol/L), TySIPonate (2 μM) or HFG (20 ng/mL) for 1 h. Cells were formalin-fixed, permeabilized and stained with mouse anti-cortactin(green) and phalloidin (red). Arrow indicates lamellipodiae; Scale-40 μm. *p < 0.05, treatment vs. Ctrl; ^#p < 0.05, combo vs single treatment.