GRP78 is a novel receptor initiating a vascular barrier protective response to oxidized phospholipids

Abstract

Vascular integrity and the maintenance of blood vessel continuity are fundamental features of the circulatory system maintained through endothelial cell-cell junctions. Defects in the endothelial barrier become an initiating factor in several pathologies, including ischemia/reperfusion, tumor angiogenesis, pulmonary edema, sepsis, and acute lung injury. Better understanding of mechanisms stimulating endothelial barrier enhancement may provide novel therapeutic strategies. We previously reported that oxidized phospholipids (oxidized 1-palmitoyl-2-arachidonoyl-sn-glycero-3-phosphocholine [OxPAPC]) promote endothelial cell (EC) barrier enhancement both in vitro and in vivo. This study examines the initiating mechanistic events triggered by OxPAPC to increase vascular integrity. Our data demonstrate that OxPAPC directly binds the cell membrane-localized chaperone protein, GRP78, associated with its cofactor, HTJ-1. OxPAPC binding to plasma membrane-localized GRP78 leads to GRP78 trafficking to caveolin-enriched microdomains (CEMs) on the cell surface and consequent activation of sphingosine 1-phosphate receptor 1, Src and Fyn tyrosine kinases, and Rac1 GTPase, processes essential for cytoskeletal reorganization and EC barrier enhancement. Using animal models of acute lung injury with vascular hyperpermeability, we observed that HTJ-1 knockdown blocked OxPAPC protection from interleukin-6 and ventilator-induced lung injury. Our data indicate for the first time an essential role of GRP78 and HTJ-1 in OxPAPC-mediated CEM dynamics and enhancement of vascular integrity.

FIGURE 1:

GRP78 and OxPAPC-induced EC barrier enhancement. (A) HPAECs plated on microelectrodes were treated with vehicle (IgG) or GRP78-blocking antibody (50 μg/ml), followed by OxPAPC (10 μg/ml) stimulation and TER measurements. (B) HPAECs were transfected with HTJ-1–specific siRNA or nonspecific RNA, followed by stimulation with OxPAPC and TER measurements. (C) HTJ-1 knockdown did not affect the TER increase in HPAEC caused by iloprost (200 ng/ml). (D, E) HPAECs grown in 96-well plates (D) or on glass coverslips (E) with immobilized biotinylated gelatin (0.25 mg/ml) were transfected with HTJ1-specific siRNA or nonspecific RNA. After 72 h, the cells were stimulated with OxPAPC for 30 min, followed by addition of FITC-avidin (25 μg/ml, 3 min). Unbound FITC-avidin was removed, and FITC fluorescence was measured; *p < 0.05; n = 5.

FIGURE 2:

Analysis of GRP78–HTJ1 and OxPL–GRP78 interactions. (A) GRP78 and HTJ1 expression in HPAEC and human lung microvascular endothelial cells was detected by Western Blot. (B, C) GRP78 interactions were analyzed in coimmunoprecipitation assays using lysates from control or DMPC- (10 μg/ml, 15 min) or OxPAPC-stimulated (10 μg/ml) cells with antibody to GRP78 (B, top), HTJ1 (B, bottom), or EO6 antibody recognizing OxPL (C). (D) Human recombinant GRP78 was incubated with OxPAPC, OxPAPS, or their oxidation-resistant analogues DMPC or DMPS. Left, native gel electrophoresis, followed by Western blot with anti-GRP78 antibody. Shift in electrophoretic mobility of GRP78 incubated with OxPAPS, but not DMPS, indicates formation of GRP78–OxPAPS complex. Right, SDS–PAGE, followed by Western blot with EO6 antibody and reprobing with anti-GRP78 antibody. Positive EO6 immunoreactivity of GRP78 preincubated with OxPAPC indicates formation of GRP78–OxPAPC complex. (E) ELISA plates coated with OxPAPS or DMPS or control uncoated plates incubated with PBS incubated with human recombinant GRP78 (left) or HPAEC lysates (middle and right). The bound GRP78 was detected using anti-GRP78 antibody. *p < 0.05 vs. DMPS or PBS; n = 4.

FIGURE 3:

Analysis of OxPAPC-induced GRP78 membrane translocation. (A) Endothelial cells grown on glass coverslips were stimulated with vehicle of OxPAPC (10 μg/ml, 15 min), followed by immunofluorescence labeling of GRP78 (green) and plasma membrane with wheat germ agglutinin (red). Merged images show GRP78 colocalization with plasma membrane labeled by wheat germ agglutinin (yellow). (B, C) Cells were stimulated with OxPAPC for indicated periods of time, and cell surface proteins were labeled with Sulfo-NHS-SS-Biotin. Biotinylated proteins were collected using streptavidin-agarose and evaluated by Western blot (B). CEM fractions (20% OptiPrep layer) were isolated, followed by GRP78 immunoprecipitation and detection of biotinylated GRP78 in precipitates using avidin-HRP–labeled antibody (C).

FIGURE 4:

Effect of HTJ1 knockdown and MβCD treatment on OxPAPC-induced activation of signaling complex in the CEMs. (A) CEM fractions (20% OptiPrep layer) were isolated from HPAECs treated with 10 μg/ml OxPAPC. S1P-R1, GRP78, HTJ1, Src, Fyn, phospho-Src family kinases, Rac1, and caveolin as a CEM marker and detected by Western blot. (B) HPAECs treated with nonspecific or HTJ1-specific siRNA were stimulated with OxPAPC (10 μg/ml, 15 min), and CEM fractions were isolated. Accumulation of GRP78, HTJ1, Src, Fyn, Rac1, and increased phosphorylation of phospho-Src family kinases in CEM fractions was detected by Western blot. Bottom, verification of HTJ1 depletion by specific siRNA. (C) HPAECs treated with nonspecific siRNA, caveolin-1–specific siRNA, or MβCD (2 mM) were stimulated with OxPAPC, followed by isolation of CEM fractions. Accumulation of GRP78, HTJ1, calnexin, and calreticulin was detected by Western blot analysis. Bottom, verification of caveolin-1 depletion. (D, E) HPAECs treated with vehicle or MβCD (2 mM) were stimulated with OxPAPC (10 μg/ml, 15 min), followed by (D) isolation of plasma membrane fractions. Transactivation of S1P-R1 was evaluated by increased Thr-236 phosphorylation levels. Total S1P-R1, HTJ1, and GRP78 in plasma membrane fractions were detected by Western blot or (E) immunoprecipitation of GRP78 from membrane fractions. GRP-78–S1P-R1 association and S1P-R1 transactivation were monitored by Western blot.

FIGURE 5:

GRP78 and OxPAPC-induced activation of Rac signaling. HPAECs were treated with nonspecific or HTJ-11–specific siRNA before OxPAPC (10 μg/ml, 15 min) stimulation. (A) Effect of HTJ-1 knockdown on OxPAPC-induced membrane translocation of GRP78 and cortactin. HTJ-1 protein depletion was verified by Western blot. (B) Rac-GTP pull-down assay of control and HTJ-1–depleted HPAEC. (C) Effect of HTJ-1 knockdown on OxPAPC-induced site-specific phosphorylation of Src, PAK1, and cortactin. Autophosphorylation of Src at Tyr-416, indicating its activation, phosphorylation of Rac target PAK1 at Ser-423 and Ser-199, as well as cortactin phosphorylation at Tyr-421 and Tyr-486, reflecting activation of Rac signaling, were detected by immunoblotting. (D) Effect of HPAEC preincubation with GRP78-blocking antibody on OxPAPC-induced site-specific phosphorylation of PAK1 and cortactin. ECs were pretreated with GRP78-blocking antibody (50 μg/ml), vehicle, or control IgG before OxPAPC stimulation (10 μg/ml).

FIGURE 6:

HTJ-1 knockdown blocks OxPAPC-induced activation of cortical actin dynamics and cortactin accumulation. (A) Live-cell imaging of HPAEC labeled with GFP-cortactin and stimulated with OxPAPC (10 μg/ml). Consecutive images were taken after 1, 2, 5, 15, and 30 min of OxPAPC stimulation. (B) OxPAPC-induced, time-dependent lamellipodia dynamics and peripheral cortactin accumulation were blocked by HTJ-1 knockdown. (C) HPAECs grown on glass coverslips were transfected with nonspecific or HTJ-1–specific siRNA (top) or incubated with control IgG or GRP78 blocking antibody (bottom), followed by OxPAPC stimulation (10 μg/ml, 30 min). Immunofluorescence staining with Texas red–phalloidin detects actin filaments. Arrows indicate peripheral F-actin enhancement caused by OxPAPC.

FIGURE 7:

Knockdown of mouse homologue of HTJ-1 attenuates protective effects of OxPAPC in animal models of acute lung injury. Mice were transfected with nonspecific or MTJ-1–specific siRNA. (A) Tissue-specific MTJ-1 depletion was verified by RT-PCR analysis of lung samples and Western blot analysis of lung, heart, liver, and kidney extracts. (B) Protein concentration and cell count in BAL samples of control and IL-6–exposed mice (5 μg/kg, intrathecal) with or without OxPAPC treatment (1.5 mg/kg, intravenous; *p < 0.05 vs. IL-6; n = 6. (C, D) Mice were transfected with nonspecific or MTJ1-specific siRNA, followed by IL-6 administration with or without OxPAPC (1.5 mg/kg, intravenous). Protein concentration and cell count in BAL samples were analyzed. *p < 0.05 vs. IL-6; n = 4 (C). Vascular leak was assessed by measurements of Evans blue accumulation in the lung parenchyma. *p < 0.05 vs. nonspecific RNA; n = 4 (D). (E) Mice were transfected with nonspecific or MTJ1-specific siRNA, followed by HTV with or without OxPAPC intravenous injection (1.5 mg/kg). Protein concentration and cell count in BAL samples of control and HTV-exposed mice were analyzed. *p < 0.05 vs. HTV; n = 6.

FIGURE 8:

Proposed model of OxPAPC-induced human EC barrier enhancement. OxPAPC binds GRP78 and induces membrane localization of GRP78/HTJ-1 complex. OxPAPC-bound GRP78/HTJ-1 interacts with S1P-R1 and induces its activation via translocation to CEMs and consequent Src and Fyn tyrosine kinase–dependent phosphorylation of Akt, resulting in Akt-mediated S1P-R1 transactivation (threonine phosphorylation). Activated S1P-R1 receptor induces full activation of Akt via mTOR and PI3K-dependent serine and threonine phosphorylation required for Rac1 activation, cortical actin cytoskeletal rearrangement, and consequent OxPAPC-mediated EC barrier enhancement.