TNF-α-activated eNOS signaling increases leukocyte adhesion through the S-nitrosylation pathway

Reagents

TNF-α was purchased from Roche. 1H-[1,2, 4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ) and N^G-methyl-l-arginine (l-NMA) were from Sigma Chemicals (St. Louis, MO). PKCζ inhibitor peptide was from Tocris. The working concentrations and application times were TNF-α (1 nM) for the indicated times in each figure, ODQ (10 µM) for 10 min before TNF-α, l-NMA (150 µM) for 45 min before TNF-α, and PKCζ inhibitor peptide (1 µM) for 20 min before TNF-α.

Antibodies

Mouse anti-ICAM-1 and rabbit anti-phospho-ICAM-Y512 were purchased from Santa Cruz. Rabbit anti-PKCζ, rabbit anti-pPKCζ (Thr410), rabbit anti-p-threonine, and rabbit anti-cortactin were from Cell Signaling. Mouse anti-peNOS-Ser1177 was from BD Transduction Laboratories and mouse anti-β-actin was from Sigma.

Intravital Microscopy and Image Analysis

Protocols for experiments on animals were approved by the Institutional Bioethics and Biosecurity Committee of the Universidad Austral de Chile and conducted according to National Institutes of Health (NIH) Guidelines for the use of animals in research. We used five animals in each experimental group. C57BL/6J (Jackson Laboratory, Bar Harbor, MA) male mice were anesthetized with ketamine (100 mg/kg ip)-xylazine (10 mg/kg ip). The cremaster was gently exteriorized through a midline scrotum incision and visualized with an inverted microscope (objective, ×32, 0.4 numerical aperture) equipped with a TV camera (Hamamatsu, Middlesex, NJ) connected to the microscope. Leukocytes adhering to the venular endothelium were recorded for 10 min, whereas a phosphate-buffered saline (PBS) solution was applied topically. One venule was recorded per animal. l-NMA was administered at 50 mg/kg for 30 min via the caudal tail vein in experiments designed to assess the role of NO.

Cell Culture

Immortalized human venous endothelial cells EAhy926 (ATCC CRL2922) were grown in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (vol/vol) fetal bovine serum (FBS), 2 mM l-glutamine, 100 U/mL penicillin, 100 µg/mL streptomycin, 2.5 µg/mL fungizone and 100 µM sodium hypoxanthine, 0.4 µM aminopterin, and 16 µM thymidine (HAT). EAhy926 cells have been authenticated as endothelial cells by ATCC using morphology, karyotyping, and PCR-based approaches to confirm their identity and to rule out intra- and interspecies contamination. The assays include the detection of species-specific variants of the cytochrome-c oxidase I gene (COI analysis) and short tandem repeat (STR) profiling. We used EAhy926 cells between passages 6–10 and additionally tested them for von Willebrand factor.

Polymorphonuclear Leukocyte Isolation

Whole blood was drawn from healthy human donors (the informed consent protocol was applied previously) on sodium citrate/dextran (6% p/vol; Sigma-Aldrich, St. Louis, MO). Isolation of polymorphonuclear leukocyte (PMN) was performed under sterile conditions using a dextran-PBS density gradient centrifugation technique. The leukocyte-enriched plasma was centrifuged, and the cellular pellet was layered on 50% Percoll solution gradient (Sigma-Aldrich) to separate the neutrophils (33).

Adhesion Assay

EAhy926 cells were grown to confluence on 24-well plates and then stimulated with 1 nM TNF-α in the presence or absence of l-NMA, ODQ, or PKCζ inhibitor peptide. Isolated PMNs were suspended in DMEM media at 1.5 × 106 cells/mL, activated with 1 nM TNF-α for 30 min at 37°C, and incubated with endothelial cells for 30 min at 37°C. Subsequently, unbound PMN were washed three times to remove them. Bound cells were lysed and assayed for myeloperoxidase (MPO) activity adding Hank’s balanced salt solution, 3,3′,5,5′-tetramethylbenzidine, and H₂O₂. The product was measured spectrophotometrically (Titertek, Multiscan) at 460 nm (33).

Western Blot Analysis

Confluent cells grown in 60-mm or 100-mm plates were serum starved overnight in DMEM 2% (vol/vol) FBS). TNF-α (1 nM) was applied to the cells for different times. Cells were washed two times with ice cold PBS and scraped in 300 µL lysis buffer 1, consisting of 50 mM Tris·HCl (pH 7.4), 150 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 1% Triton X-100, and protease and phosphatases inhibitor mixture, and incubated on ice with shaking for 30 min. Lysates were obtained by centrifugation at 10,000 g for 15 min at 4°C. Separated proteins were blotted to PVDF membranes and detected with specific antibodies. Proteins of interest were detected by ECL (Pierce). We analyzed Western blots densitometrically using the NIH ImageJ program.

Real-Time Detection of NO

We used a protocol previously described (34). Confluent EAHy926 cells were loaded with DAF-FM Diacetate (Invitrogen, Cat. No. D-23844) and examined for increases in NO following addition of TNF-α (1 nM). Spinning disk confocal imaging parameters: a 488-nm laser line (20w) atenued at 2% was used for excitation. To minimize exposure time, resolution was reduced at 2 × 2 binning on a CCD-based (EMCCD iXon Life 888, ANDOR, Oxford Instruments) wide field system. Exposure times were 50 ms. Multiple exposures over time were collected at one image per second for 360 s. The imaging sequence was initiated, and 10 s later the stimulus was added. For the analysis, 10 individual regions of interest (ROI) were used for each plate. Data are expressed as the mean fluorescence intensity (F) relative to the mean fluorescence intensity before stimulus application (F0). This value is expressed as F/F0.

ICAM-1 Cell Surface Immunofluorescence Microscopy

EAhy926 cells were cultured on glass coverslips, treated with TNF-α for the indicated times, and then washed with cold Dulbecco’s phosphate-buffered saline (DPBS) Ca²⁺, Mg²⁺ pH 7.2 on ice. Coverslips were blocked with 3% BSA in DPBS for 30 min on ice. Thereafter, cells were incubated with anti-ICAM antibody for 4 h at 4°C. After three washes in DPBS, cells were fixed in 100% methanol for 45 min at −20°C. Coverslips were air-dried and incubated with the secondary Alexa Fluor-conjugated antibody for 2 h at room temperature. Cells were extensively washed in DPBS, mounted on glass slides, and images were obtained using an epifluorescence microscope (Axioscop; Carl Zeiss) and Axio Vision Rel. software (Zeiss, Germany). Eight-bit images were prepared for publication in Adobe Photoshop.

Immunoprecipitation

EAhy926 cells stimulated with the different agents were lysed in immunoprecipitation buffer, consisting of 50 mM Tris·HCl (pH 7.4), 1% NP-40, 20 mM EDTA, and protease and phosphatase inhibitors mixture, and passed five times through a 28G1/2 needle. Samples were centrifuged at 14,000 rpm for 3 min at 4°C. Equal amounts of protein in the supernatant from control and TNF-α-treated cells were incubated with specific antibodies 3 h or overnight at 4°C. Protein A/G beads were added to samples for 2 h at 4°C; they were then pelleted by centrifugation and washed five times with immunoprecipitation buffer. Proteins of interest were detected using Western blot analysis and chemiluminescence. Quantification of changes from control was evaluated by densitometric analysis of Western blots using the NIH ImageJ Program.

Subcellular Fractionation

We used protocols previously described (27). Briefly, monolayers were washed in cold Tris-buffered saline (TBS) and lysed in 10 mM Tris·HCl (pH 7.4), 1 mM MgCl₂, 5 mM EDTA, 10 mM EGTA, 1 mM NaVO₄, and antiprotease mixture. The cell suspension was sonicated at 4°C for 10 s and then centrifuged at 100,000 g for 1 h at 4°C. The supernatant was saved as the cytosolic fraction. The pellet was suspended in lysis buffer with Triton X-100 1% and sonicated and incubated for 30 min at 4°C. Thereafter, suspended pellets were centrifuged at 13,000 rpm for 3 min at 4°C, and this second supernatant was considered as the membrane fraction. Membrane and cytosolic fractions were analyzed by Western blot analysis using anti-ICAM-1 antibody.

ICAM-GFP Dynamics Measurements on the Cell Surface by FRAP

EAhy926 cells were seeded on MatTek live cell imaging glass bottom dishes (MatTek Corporation, Ashland, MA) and transfected transiently with lipofectin (Invitrogen) to express transiently human ICAM-GFP (kindly donated by Dr. Francisco Sánchez-Madrid; Spanish National Center for Cardiovascular Research, Spain). After 48 h, the cells were washed twice with DMEM 10 mM HEPES, pH 7.4, 2% FBS and placed on temperature-controlled chamber UNO-Controller (Okolab SRL, Pozzuoli, Italy) mounted on a TCS SP8 Leica confocal microscope running Leica LAS X controller software (Leica Microsystems, Wetzlar, Germany). Different treatments protocols were applied: TNF-α for 5 min, l-NMA for 45 min, and l-NMA 45 min plus TNF-α 5 min. Fluorescence recovery after photobleaching (FRAP) for GFP was performed as follows: a multiline Argon laser was adjusted to 80% of potency (50 mW) for a single-shot irradiation at 488 nm by using FRAP-booster and zoom-in bleaching on a selected circular 20-pixel diameter region of interest (ROI) placed on ICAM-GFP cell surface located fluorescent signal. Pre- and postbleaching 144.72 × 144.72 µm (1,024 × 1,024 pixels) images were acquired every 3 s using 63× oil-immersion objective and a photomultiplier tube (PMT) gain adjusted to 650, 2 average, and 700 Hz frequency scanning at 1 mW 488 laser potency. Fluorescence recovery was recorded after photobleaching, and the resulting recovery curve was fitted to a single exponential curve using Leica LAS X software quantify tool. Recovery time and mobile fraction were obtained directly from curve fitting.

Biotin-Switch Assay

Total protein (100 µg) obtained from cellular lysates from control and TNF-α treated cells were denatured with sodium dodecyl sulfate (SDS) in the presence of methyl methanothiosulfonate (MMTS) (35). After acetone precipitation to remove excess MMTS, 1 mM ascorbate and 4 mM N-[6-(biotinamido)-hexyl]-3′-(2′-pyridyldithio) propionamide (biotin-HPDP) were added to reduce the S-NO bond and label the reduced thiol with biotin, respectively. Biotinylated proteins were precipitated with acetone, suspended, captured with streptavidin-agarose beads, and then separated by sodium dodecyl sulfate-PAGE. Proteins of interest were detected using Western blot analysis and chemiluminescence. We analyzed Western blots densitometrically using the NIH ImageJ program.

eNOS siRNA Transfection

We used specific siRNA (Ambion, Austin, TX) to deplete eNOS as we described previously (36). We transfected EAHy926 cells with small interfering (si)RNA using lipofectin and following the protocol recommended by the manufacturer (Invitrogen, Carlsbad, CA). We perform the experiment 96-h posttransfection. Western blot analysis was done to confirm eNOS depletion.

Identification of S-Nitrosylated Cysteines in PKCζ

The site of S-nitrosylation in PKCζ was identified by MS/MS mass spectrometry. Five micrograms of recombinant PKCζ (Cat. No. P2273, ThermoFisher) were treated with S-nitrosoglutathione (GSNO) at 37°C for 30 min. After the S-nitrosylation reaction, GSNO was removed, and all the free SH- were blocked by MMTS. Biotin-switch on S-nitrosylated PKCζ was performed with Biotin-HPDP. The result protein was split into two parts. One part was loaded on SDS-PAGE for trypsin digestion; the other part was run in Western blot analysis and detected using antibiotin antibody. Since both MMTS and Biotin-HPDP blocking could be reduced by dithiothreitol (DTT), trypsin digestion was performed without reduction. The peptides were C18 desalted and analyzed by LC-MS/MS on Orbitrap Fusion Lumos instrument. The MS/MS spectra were searched against SwissProt mouse database using Sequest search engines on Proteome Discoverer (V2.4) platform. The protein false discovery rate was less than 1%.

Statistical Analysis

Experiments were conducted in groups with a minimum of five independent experiments (n = 5). Data are expressed as means ± SE. Apparent differences were assessed for statistical significance using GraphPad Prism and Sigma Plot software. Details of the specific method are indicated in each figure. Significance was accepted at P < 0.05.

NO Signaling Regulates Positively Leukocyte Adhesion Induced by TNF-α In Vivo at Short Times of Stimulation

We first determined in vivo the effect of NO inhibition on leukocyte adhesion induced by TNF-α using intravital microscopy in the mouse cremaster muscle (30). We applied 1 nM TNF-α topically to the muscle and recorded leukocyte adherence. Leukocyte adhesion increased significantly at 5 and 10 min after TNF-α, and pretreatment of the mice with the nitric oxide synthases inhibitor l-NMA (50 mg/kg, injected via caudal vein 30 min before muscle exposure) inhibited TNF-α triggered leukocyte adhesion at 5 and 10 min (Fig. 1, A and B). These results demonstrated in vivo that leukocyte adhesion induced by TNF-α requires NO production. Based on the short time of TNF-α administration, we infer that eNOS is the isoform involved in the process.

S-Nitrosylation Regulates TNF-α-Induced PMN Adhesion to Endothelial Cells at Short Times of Stimulation

Next, we tested the effects of NO on leukocyte adhesion in vitro. TNF-α, applied for 5 and 15 min, induced significant PMN adhesion measured by the MPO activity (Fig. 2A). To test whether this effect depends on eNOS activation, we inhibited eNOS by incubating EAhy926 cells with l-NMA (150 µM for 45 min), a global nitric oxide synthase inhibitor. Since NO can independently activate sGC-PKG pathway and the S-nitrosylation pathway, we applied ODQ (10 µM for 10 min), a recognized sGC-PKG inhibitor, to differentiate between these two pathways. Inhibition of eNOS, but not inhibition of sGC-PKG, inhibited TNF-α-induced PMN adhesion at 5 min (Fig. 2B) and 15 min (Fig. 2C) indicating that this process depends on S-nitrosylation. The treatment with l-NMA and ODQ alone did not have direct effects on leukocyte adhesion (Fig. 2D). We measured eNOS phosphorylation at Ser1177 to verify activation of the enzyme and to confirm the involvement of eNOS-derived NO. TNF-α induced eNOS phosphorylation at times corresponding with its effects on PMN adhesion (Fig. 2E). To corroborate that eNOS phosphorylation is leading to NO production, we measured NO levels in response to TNF-α using DAF-FM. Figure 3F shows a significant increase in the accumulated fluorescence signal for NO after application of the agonist. Figure 3G illustrates a representative result of one experiment showing the increase in the fluorescence signal for NO in the time. These results demonstrate that the eNOS-NO-S-nitrosylation pathway regulates TNF-α-induced PMN adhesion at short times of stimulation.

NO Regulates Positively ICAM-1 Clustering Induced by TNF-α

It has been previously reported that the fast leukocyte adhesion induced by TNF-α is mediated by ICAM-1 clustering, which involves changes in constitutively expressed ICAM-1 improving the affinity of this protein by its ligands in leukocytes (27–29). By immunofluorescence experiments using an anti-ICAM-1 antibody that labels an extracellular epitope of ICAM-1, we observed an increase in ICAM-1 surface label after TNF-α treatment (Fig. 3A) with a punctate label in the cell surface characteristic of clustering. To discard any possible effects of vesicular traffic of ICAM-1 cytosolic pool to the plasma membrane, we performed subcellular fractionation experiments. EAhy926 cells were treated with 1 nM TNF-α for 5 and 15 min and then cytosolic and membrane fractions were analyzed by Western blot analysis. Figure 3B shows that TNF-α treatment did not cause any significant changes in the plasma membrane/cytosol distribution of ICAM-1. These results indicate that the enhanced surface label for ICAM-1 in immunofluorescence experiments are not due to ICAM-1 vesicular traffic to the plasma membrane, but rather to changes in ICAM-1 that improve its affinity for the antibody, which is strongly suggestive of clustering (27). During clustering, ICAM-1 also associates with cytoskeletal proteins such as cortactin (21); therefore, we evaluated the association between ICAM-1 and cortactin after stimulation of EAhy926 cells with TNF-α for 5 and 15 min. We observed that TNF-α increased the association between ICAM-1 and cortactin, which was inhibited by blocking NO production with l-NMA (Fig. 3C).

ICAM-1 clustering also increases the immobile fraction of the protein (37); therefore, to further verify ICAM-1 clustering, we performed FRAP experiments. We transfected EAhy926 cells transiently with human ICAM-1-GFP; then, we bleached a small area of the plasma membrane and monitored the fluorescence recovery of ICAM-1-GFP in real time. The results in Fig. 4A show a representative image of ICAM-1-GFP before bleaching, immediately after bleaching, and after recovery in nonstimulated conditions (C, top), under TNF-α treatment (TNF, middle) and with pretreatment with NOS inhibitor, l-NMA, before TNF-α treatment (LNMA/TNF, bottom). After TNF-α treatment (square), fluorescence recovery was significantly less than control (triangle) indicating that TNF-α increased the immobile fraction of ICAM-1 (Fig. 4B). The l-NMA pretreatment (circles) showed a similar recovery curve as control indicating that l-NMA inhibited the effect of TNF-α on the lateral mobility (Fig. 4B). Figure 4C shows the quantification of the immobile fraction indicating that TNF-α treatment increased the immobile fraction, and this effect is inhibited in the presence of l-NMA. These data indicate that ICAM-1 clustering induced by TNF-α depends on NO signaling.

TNF-α Leads to S-Nitrosylation of PKCζ and Regulation of Its Activity

Next, we searched for possible S-nitrosylated proteins that might be responsible for ICAM-1 clustering. We first tested S-nitrosylation of ICAM-1. Figure 5A shows that application of TNF-α for 5 and 15 min failed to elicit S-nitrosylation of ICAM-1. We next tested PKCζ as a possible target because it phosphorylates ICAM-1 and promotes ICAM-1 clustering after stimulation with TNF-α (27, 28). Application of TNF-α for 5 min induced significant PKCζ S-nitrosylation in EAhy926 cells, which was inhibited by l-NMA (Fig. 5B). To confirm that PKCζ S-nitrosylation is being induced by eNOS, we deplete eNOS using siRNA. Figure 5C shows that in the presence of eNOS siRNA, there is an inhibition in PKCζ S-nitrosylation induced by TNF-α (top) coincident with lower expression of eNOS shown in the Western blot (bottom). Since NO-induced S-nitrosylation regulates protein phosphorylation (17, 38) and PKCζ is activated by phosphorylation, we investigated whether or not NO regulates PKCζ phosphorylation. Figure 5D shows that TNF-α significantly increases PKCζ phosphorylation in EAhy926 cells. The increase in PKCζ phosphorylation was blocked by l-NMA indicating that NO, and presumably S-nitrosylation, is required to activate PKCζ. Because S-nitrosylation regulates protein interactions (16, 17, 39) and PKCζ associates with ICAM-1 during ICAM-1 clustering (27, 40, 41), we investigated whether or not S-nitrosylation regulates PKCζ association with ICAM-1. Figure 5E shows that TNF-α (applied for 5 min) causes significant association between PKCζ and ICAM-1 in EAhy926 cells, as demonstrated by enhanced coimmunoprecipitation. Inhibition of NO blocks the association of PKCζ and ICAM-1 induced by TNF-α, indicating that NO production and signaling are required for PKCζ-ICAM-1 association. Finally, we tested ICAM-1 phosphorylation. Figure 5F shows that TNF-α significantly increases ICAM-1 phosphorylation at Tyr 512 and the inhibition of eNOS blocks TNF-α-induced ICAM-1 phosphorylation at this residue. In the same way, Fig. 5G shows that TNF-α treatment for 5 min induced ICAM-1 threonine phosphorylation, which was inhibited in the presence of l-NMA. Taken together, these data indicate that PKCζ is a target of S-nitrosylation by TNF-α and that NO signaling regulates PKCζ activation, the interaction of PKCζ with ICAM-1, and ICAM-1 phosphorylation at tyrosine and threonine.

PKCζ Regulates Leukocyte Adhesion

To demonstrate that PKCζ regulates leukocyte adhesion, we use a myristoylated membrane-permeable peptide that specifically inhibits PKCζ in adhesion assays (27). Figure 6 shows that the pretreatment of EAhy926 cells with the inhibitor peptide of PKCζ significantly reduces the early endothelial adhesion induced by TNF-α after 5 and 15 min of stimulation. These results demonstrate that PKCζ activation is required for leukocyte adhesion induced by TNF-α.

NO Leads to S-Nitrosylation of PKCζ in Cys 503

To determine the cysteines nitrosylated by NO in PKCζ, we exposed 5 µg of purified PKCζ to S-nitrosoglutathione as a source of bioavailable NO. Analysis by mass spectroscopy identified Cys 503 as the only S-nitrosylated cysteine in PKCζ (Fig. 7). Cys 503 locates at the protein kinase domain of the protein (unit prot), and therefore, the modification of this cysteine may affect the kinase activity of the protein.