Nitric-oxide Synthase-2 Linkage to Focal Adhesion Kinase in Neutrophils Influences Enzyme Activity and β2 Integrin Function*

Background: Reactive nitrogen species increase by an unclear mechanism with exposure to hyperbaric oxygen, which inhibits neutrophil β2 integrin adherence. Results: Nitric-oxide synthase-2 activity increases when held in proximity to short filamentous actin by focal adhesion kinase. Conclusion: Protein associations are transient because of actin S-nitrosylation. Significance: Nitric-oxide synthase regulation in neutrophils depends on cytoskeletal protein associations. This investigation was to elucidate the basis for augmentation of nitric-oxide synthesis in neutrophils exposed to hyperbaric oxygen. Hyperoxia increases synthesis of reactive species leading to S-nitrosylation of β-actin, which causes temporary inhibition of β2 integrin adherence. Impaired β2 integrin function and actin S-nitrosylation do not occur in neutrophils from mice lacking type-2 nitric-oxide synthase (iNOS) or when incubated with 1400W, an iNOS inhibitor. Similarly, effects of hyperoxia were abrogated in cells depleted of focal adhesion kinase (FAK) by treatment with small inhibitory RNA and those exposed to a specific FAK inhibitor concurrent with hyperoxia. Nitric oxide production doubles within 10 min exposure to hyperoxia but declines to approximately half-maximum production over an additional 10 min. Elevated nitric oxide production did not occur after FAK depletion or inhibition, or when filamentous actin formation was inhibited by cytochalasin D. Intracellular content of iNOS triples over the course of a 45-min exposure to hyperoxia and iNOS dimers increase in a commensurate fashion. Confocal microscopy and immunoprecipitation demonstrated that co-localization/linkage of FAK, iNOS, and filamentous actin increased within 15 min exposure to hyperoxia but then decreased below the control level. Using isolated enzymes in ex vivo preparations an association between iNOS and filamentous actin mediated by FAK could be demonstrated and complex formation was impeded when actin was S-nitrosylated. We conclude that iNOS activity is increased by an FAK-mediated association with actin filaments but peak nitric oxide production is transient due to actin S-nitrosylation during exposure to hyperoxia.

This investigation was to elucidate the basis for augmentation of nitric-oxide synthesis in neutrophils exposed to hyperbaric oxygen. Hyperoxia increases synthesis of reactive species leading to S-nitrosylation of ␤-actin, which causes temporary inhibition of ␤ 2 integrin adherence. Impaired ␤ 2 integrin function and actin S-nitrosylation do not occur in neutrophils from mice lacking type-2 nitric-oxide synthase (iNOS) or when incubated with 1400W, an iNOS inhibitor. Similarly, effects of hyperoxia were abrogated in cells depleted of focal adhesion kinase (FAK) by treatment with small inhibitory RNA and those exposed to a specific FAK inhibitor concurrent with hyperoxia. Nitric oxide production doubles within 10 min exposure to hyperoxia but declines to approximately half-maximum production over an additional 10 min. Elevated nitric oxide production did not occur after FAK depletion or inhibition, or when filamentous actin formation was inhibited by cytochalasin D. Intracellular content of iNOS triples over the course of a 45-min exposure to hyperoxia and iNOS dimers increase in a commensurate fashion. Confocal microscopy and immunoprecipitation demonstrated that co-localization/linkage of FAK, iNOS, and filamentous actin increased within 15 min exposure to hyperoxia but then decreased below the control level. Using isolated enzymes in ex vivo preparations an association between iNOS and filamentous actin mediated by FAK could be demonstrated and complex formation was impeded when actin was S-nitrosylated. We conclude that iNOS activity is increased by an FAK-mediated association with actin filaments but peak nitric oxide production is transient due to actin S-nitrosylation during exposure to hyperoxia.
The goal of this study was to determine the mechanism for augmented nitric oxide ( ⅐ NO) production by neutrophils exposed to hyperbaric oxygen (HBO 2 ). The motivation for this research effort is to understand how HBO 2 temporarily inhibits neutrophil ␤ 2 integrin adhesion molecules, which participate in regulating neutrophil activation and endothelial adhesion (1). When animals or humans are exposed to HBO 2 at 2.8 to 3.0 atmospheres absolute (ATA), 2 ␤ 2 integrins on circulating neutrophils are temporarily inhibited (2)(3)(4)(5)(6). This effect offers a unique therapeutic opportunity because inhibition of neutrophil ␤ 2 integrin adhesion by hyperoxia ameliorates reperfusion injuries of brain, skeletal muscle, and intestine, as well as smoke-induced lung injury, decompression sickness, and encephalopathy due to carbon monoxide poisoning in animal models (2, 6 -13). Inhibited ␤ 2 integrin adhesion may be the basis for benefits of HBO 2 shown in clinical trials involving coronary artery thrombolytic therapy, balloon angioplasty/ stenting (14 -17), and the reduction of encephalopathy seen after cardiopulmonary bypass and carbon monoxide poisoning (18,19).
Previous studies have shown that although the function of ␤ 2 integrins is temporarily inhibited (in humans the effect resolves within 24 h (3)) neutrophil antibacterial responses remain intact (20,21). HBO 2 does not reduce neutrophil viability or cause immunocompromise, as functions such as degranulation and oxidative burst in response to chemoattractants are intact (2,3). In fact, HBO 2 causes a degree of cell activation based on elevation of ␤ 2 integrin surface expression and oxidative burst without altering ␤ 2 integrin affinity, as measured by binding beads coated with intercellular adhesion molecule-1 (22).
When neutrophils are exposed to hyperoxia, there is an increased production of reactive species derived from nitricoxide synthase (NOS) and myeloperoxidase, which cause S-nitrosylation (SNO Ϫ ) of ␤ actin (22). This increases actin polymerization because vasodilator-stimulated phosphoprotein (VASP) has higher affinity for the S-nitrosylated form of short filamentous actin (sF-actin) (23). VASP bundles Rac1, Rac2, cyclic AMP-dependent and cyclic GMP-dependent protein kinases in close proximity to short actin filaments and subsequent Rac activation increases actin-free barbed end formation. HBO 2 -exposed cells exhibit greater actin filament formation and turnover, which inhibits ␤ 2 integrin clustering (also called avidity) and thus ␤ 2 integrin adhesion (22,23).
Increased actin polymerization with exposure to hyperoxia causes an increase in linkage of focal adhesion kinase (FAK) to sF-actin (24). The scaffold protein FAK coordinates many integrin-actin activities (25)(26)(27)(28). Under normal conditions the antioxidant protein NADPH-dependent thioredoxin reductase-1 (TrxR) is linked to actin filament-bound FAK (24). We found, however, that with exposure to hyperoxia this linkage is transient and FAK dissociates after 10 to 15 min. Therefore, denitrosylation of SNO-actin fails to occur spontaneously in HBO 2 -exposed neutrophils because when FAK dissociates, the TrxR secondarily dissociates from FAK and thus becomes physically separated from the SNO-actin. If actin turnover is slowed by incubating cells with fMLP or 8-Br-cGMP (which trigger phosphorylation of VASP causing its dissociation from sF-actin) then the FAK-TrxR complex with sF-actin is restored and actin S-nitrosylation is reversed (23,24). Thus, SNO-actin impedes its own eradication because of the hastened actin filament turnover.
HBO 2 fails to inhibit ␤ 2 integrin function when cells are incubated in the presence of the nonspecific NOS inhibitor N Gnitro-L-arginine methyl ester (L-NAME) or with 1400W (N-3-(aminomethyl)benzyl acetamine), which specifically inhibits NOS-2, inflammatory or iNOS (22). These findings suggest iNOS is the source for reactive nitrogen species that mediate SNO-actin formation. There are alternative sources, however, because neutrophils also contain small amounts of both NOS-1, neuronal or nNOS, and NOS-3, endothelial or eNOS (29 -31). Moreover, if nitrite were available from any source, myeloperoxidase can generate reactive nitrogen species (32)(33)(34).
There are a number of proteins that interact with NOS isoforms and either increase or decrease catalytic activity. All three NOS isoforms exhibit increased activity when associated with actin filaments; some interactions are direct and some involve intermediate protein linkages (35,36). Specifically with regard to iNOS, activity in various cell types is heightened by F-actin linkage mediated by ezrin-radixin-moesin-binding phosphoprotein (EBP50) or ␣-actinin 4, and associations with either Rac1 or Rac2 enhance enzyme activity and redistribution to cytoskeletal structures (36).
FAK consists of an N-terminal FREM (band 4.1 radixin, ezrin, moesin) homology domain, followed by a tyrosine kinase domain and a C-terminal focal adhesion targeting domain. The FERM homology domain coordinates FAK activation by several growth factors (37,38). Links to the actin cytoskeleton and regulation of cell motility, proliferation, and oncogenic transformation involve adaptor proteins such as paxillin or talin that bind integrin cytoplasmic tails and the C-terminal focal adhesion targeting region of FAK (39 -41). There are no reports of a linkage between iNOS and FAK.
The purpose of this investigation was to verify iNOS as the source for reactive nitrogen species and elucidate the mechanism for enhanced enzyme activity in neutrophils exposed to hyperoxia. In the course of these studies, it became clear that iNOS-FAK linkage played a central role in modulating iNOS intracellular localization and activity.
Animals-Mice (Mus musculus) were purchased (Jackson Laboratories, Bar Harbor, ME), fed a standard rodent diet and water ad libitum, and housed in the animal facility of the University of Pennsylvania. A colony of iNOS knock-out mice was maintained from breading pairs purchased from Jackson Laboratories. After anesthesia (intraperitoneal administration of ketamine (100 mg/kg) and xylazine (10 mg/kg)) skin was prepared by swabbing with Betadine and blood was obtained into heparinized syringes by aortic puncture.
Isolation of Neutrophils and Exposure to Various Agents-Mice were anesthetized and neutrophils were isolated from heparinized blood as previously described (22). A concentration of 5 ϫ 10 5 neutrophils/ml of PBS ϩ 5.5 mM glucose was exposed to either air or 2.0 ATA O 2 for 45 min (we have shown that ex vivo exposures to 1-2.8 ATA O 2 are equivalent to in vivo exposures to 2.8 ATA O 2 (22). Where indicated, prior to air/ HBO 2 exposures some cell suspensions were exposed for 20 h at room temperature to 0.08 nM siRNA following the manufacturer's instructions using control siRNA or siRNA specific for mouse FAK (23,24). Expressed as the ratio of FAK to actin in cell lysate Western blots, FAK siRNA cause a 52.5 Ϯ 5.3% (S.E., n ϭ 10) reduction in FAK content. Note that this is a different paradigm for cell manipulation versus that used in prior studies (24). In the former studies FAK depletion was performed after neutrophils had been exposed to hyperoxia.
Fibrinogen-coated Plate Adherence-Preparation and use of fibrinogen-coated plates to measure ␤ 2 integrin-specific neutrophil adherence in calcein AM-loaded cells was as previously described (22). Suspensions of 25,000 cells in 100 l of PBS were added to plate wells and at the end of the 10-min incubation, wells were washed and adherence was calculated as described in Ref. 22.
Microelectrodes-Microelectrodes selective for ⅐ NO were fabricated and mounted in a hyperbaric chamber as described in previous studies (42,43). Two-point calibrations for each electrode were made at 37°C in physiological buffer equilibrated with either 100% N 2 or 1,800 ppm ⅐ NO (balance N 2 ). The electrodes were polarized at an oxidation potential of ϩ850 mV relative to an Ag/AgCl reference electrode. Electrochemical oxidation currents were amplified with a sensitive electrometer (Keithley, model 610). The electrometer voltage output was low pass-filtered (analog circuit with 5-Hz cutoff) and digitized (1 sample/s) by computer. Current sensitivities ranged between 0.5 and 5 pA/M. Suspensions of 5 ϫ 10 5 neutrophils/0.5 ml of PBS ϩ 5.5 mM glucose in microwell plates containing a small stir bar were exposed to 2.0 ATA O 2 for 45 min while ⅐ NO production was monitored.
NOS Activity Assay in Permeabilized Neutrophils-Isolated neutrophils were subjected to permeabilization using 0.2% n-octyl ␤-glucopyranoside exactly as described in a previous publication (23). Cells were suspended in PBS (9 ϫ 10 5 /ml) with 40 M N-hydroxy-L-arginine to inhibit arginase. After 10 min, 20 mM L-[ 3 H]arginine was added and 200-l samples without or with 0.1 M 1400W were parceled for 10-min exposures to air (control) of 2.0 ATA O 2 . Immediately after the incubations 0.5 ml of 1 M trichloroacetic acid was added to quench the reaction. Cells were pelleted by centrifugation at 5000 ϫ g for 5 min, washed three times with ethyl ether, and then passed through 2 ml of Dowex 50WX8 resin. L-[ 3 H]Citrulline was eluted with two 0.5-ml washings with water and then analyzed using a scintillation counter.
Cell Extract Preparation and Biotin-switch Assay-Isolated neutrophils previously exposed to air (control) or HBO 2 were suspended in 2 ml of HEN buffer (250 mM Hepes, pH 7.7, 1 mM EDTA, 0.1 mM neocuproine), sonicated on ice for 30 s and then passed through a 28-guage needle five times. Lysates were centrifuged at 2000 ϫ g for 10 min, supernatant was recovered and made 0.4% CHAPS using the 10% stock solution. The biotinswitch assay was carried out following published methods (22).
Cytoskeletal Protein Analysis Based on Triton Solubility-Neutrophil were processed following our published protocol (22). In brief, cells were suspended in cytoskeleton stabilization buffer (300 l) (25 mM Hepes, pH 6.9, 0.2% Triton X-100, 1 M glycerol, 1 mM EGTA, 1 mM PMSF, 1 mM MgCl 2 ), incubated for 10 min at room temperature, then centrifuged at 15,000 ϫ g for 5 min to obtain the Triton-insoluble pellets. Supernatant was centrifuged at 366,000 ϫ g for 5 min and the supernatant Triton-soluble G-actin was set aside. The Triton-soluble F-actin pellet was resuspended in cytoskeleton stabilization buffer and centrifuged at 300 ϫ g for 10 min to remove debris. Where indicated both the Triton-soluble and Triton-insoluble proteins were electrophoresed in 4 -15% gradient SDS-PAGE gels and Western blotting (22) or subjected to immunoprecipitation. Triton-insoluble proteins were dissolved with SDS buffer heated to 95°C and then electrophoresis followed by Western blotting.
Immunoprecipitation of Protein Complexes-Suspensions of G-actin or short F-actin containing 250 g of protein were precleared and then incubated with 5 g of antibodies on a shaker overnight at 4°C, then 30 l of 20% (w/v) protein G-Sepharose (pre-blocked with 2% BSA) was added and incubated for 1.5 h at 4°C. Samples were washed twice in cytoskeleton stabilization buffer, pelleted, suspended in 20 l of heated SDS buffer (62.5 mM Tris-HCl, 2% SDS, 10% glycerol, 20% ␤-mercaptoethanol), incubated at 95°C for 15-20 min, electrophoresed, and then analyzed by Western blotting (22).
Confocal Microscopy-Isolated neutrophils exposed to air or HBO 2 were placed on slides coated with fibrinogen following published methods (22). Cells were permeabilized by incubation for 1 h at room temperature with PBS containing 0.1% (v/v) Triton X-100 and 5% (v/v) fetal bovine serum. Cells were then incubated overnight with 1:200 dilutions of Alexa 488-conjugated phalloidan plus primary antibodies to either FAK or iNOS. The next morning slides were rinsed three times with PBS and counterstained with a 1:500 dilution of activated protein C and R-phycoerythrin-conjugated secondary antibodies. Images of neutrophils were acquired using a Zeiss Meta510 confocal microscope equipped with a Plan-Apochromat ϫ63/ 1.4NA oil objective. Fluorophore excitation was provided by 488-and 543-nm laser lines and the resulting fluorescence was separated using 500 -530-and 560 -615-nm band pass filters.
Ex Vivo FAK/iNOS/F-actin Interactions-Solutions of glutathione S-transferase (GST)-tagged active human FAK (0.28 M) were prepared from wash solution provided as a component of a GST protein interaction pulldown kit purchased from Pierce Biotechnology. His-tagged human kinase domain of FAK (3.6 M) was prepared in 50 mM sodium phosphate, pH 7.5.
Rabbit skeletal muscle ␣-actin was polymerized to filamentous (F-actin) in modified C-buffer (25 mM Hepes, pH 7.0, 15 mM KCl, 25 mM NaCl, 2 mM MgCl 2 , 0.2 mM CaCl 2 ) as described previously (23). S-Nitrosylated actin (SNO-actin) was prepared either by first forming F-actin or allowing G-actin to incubate with 100 M S-nitroso-N-acetyl-D,L-penicillamine for 1 h as previously described (23). Purified mouse iNOS was prepared in modified C-buffer (40 g/100 l). For studies involving GST-FAK pulldown, columns were first loaded with 50 l of glutathione-agarose slurry plus 400 l of wash solution (Pierce) following the manufacturer's instructions, then with GST-FAK (50 l). Either 10 l of 10 M F-actin or G-actin was added to these preparations without or with 10 l (4 g) of iNOS and incubated for 1 h at room temperature. Columns were then washed three times with wash solution and proteins were eluted with 250 l of manufacturer-supplied elution buffer. Samples were combined with 2 ϫ SDS buffer for electrophoresis in 4 -15% gradient SDS-PAGE gels followed by Western blotting. Blots were probed for iNOS, FAK, and actin.
For studies with His-FAK, a portion was first acetylated by incubation for 1 h with 2 mM N-acetylimidazole followed by incubation with Sephadex G-25, centrifugation, and the supernatant was used in studies. Protein interactions were assessed by incubating 10 l of His-FAK solution with 10 l of iNOS and/or F-actin for 1 h at room temperature with constant shaking. The mixture was then combined with 100 l of washed cobalt resin following our published method (24). Samples were centrifuged at 700 ϫ g for 2 min, protein was eluted from the resin by incubation with elution buffer (50 mM sodium phosphate, 300 mM sodium chloride, 150 mM imidazole, pH 7.4) for 1 h at room temperature and then centrifuged for 2 min at 700 ϫ g. Samples were combined with 2 ϫ SDS buffer for electrophoresis in 4 -15% gradient SDS-PAGE gels followed by Western blotting. Blots were probed for iNOS, FAK, and actin.
Statistical Analysis-Results are expressed as the mean Ϯ S.E. for three or more independent experiments. To compare data, we used analysis of variance using SigmaStat (Jandel Scientific, San Jose, CA) and Newman-Keuls post hoc test. The level of statistical significance was defined as p Ͻ 0.05.

RESULTS
Neutrophil ␤ 2 Integrin-dependent Adherence-Neutrophil adherence and inhibition of ␤ 2 integrin function after cells were exposed to HBO 2 is shown in Table 1. As with previous studies (24), the effect of HBO 2 was abrogated if cells were incubated with NOS inhibitors L-NAME or 1400W concurrent with exposure to hyperoxia. In this study we also found that cells obtained from iNOS knock-out mice (iNOS-KO) did not exhibit impaired function of ␤ 2 integrins after exposure to HBO 2 . We conclude, therefore, that iNOS activation is required for the HBO 2 -mediated inhibition.
Interest in the role FAK may play in HBO 2 -induced ␤ 2 integrin inhibition was prompted by immunoprecipitation and imaging studies as described below. The effect of HBO 2 was abrogated if neutrophils were depleted of FAK by siRNA treatment prior to hyperoxia, or if cells were incubated with PF-573228, a small chemical inhibitor of FAK kinase activity (46), during exposure to HBO 2 .
NO-sensitive Electrode-In a recent study we showed that exposure to hyperoxia for just 10 min will reduce ␤ 2 integrin adherence by 50%, with complete inhibition occurring after a 45-min exposure (24). Therefore, we were interested in exam-ining production of ⅐ NO by cells over a 45-min incubation interval. The concentration of ⅐ NO in suspensions of neutrophils exposed to HBO 2 was monitored with a ⅐ NO-specific electrode as described under "Experimental Procedures." The baseline ⅐ NO concentration while cells were exposed to ambient air was 47.5 Ϯ 21.9 nM and a peak concentration of 97.0 Ϯ 32 nM (p Ͻ 0.05) occurred after 6.4 Ϯ 0.6 min of HBO 2 exposure.
Activation of NOS was corroborated by monitoring 1400Winhibitable conversion of L-[ 3 H]arginine to L-[ 3 H]citrulline. Permeabilized neutrophils exposed to HBO 2 for 10 min generated 4-fold greater L-[ 3 H]citrulline than air-exposed cells. In these trials, there was a lag to cell fixation so the reaction actually went on for 15 min after L-[ 3 H]arginine was added and a similar relationship in time was present in subsequent studies.
The time course for ⅐ NO production by neutrophils exposed to hyperoxia is shown in Fig. 1. Comparison across multiple experiments was aided by setting the ordinate scale as the relative elevation in ⅐ NO concentration. Each experiment was normalized to the maximum concentration reached during exposure to HBO 2 and the figure shows mean Ϯ S.E. for 7 experiments. As shown, the peak ⅐ NO production was transient and dropped to 50% of the peak level at 20.5 Ϯ 5.3 min of exposure.
Studies have shown that protein associations involving FAK are modified when cells are exposed to HBO 2 (24). The effect of hyperoxia on ⅐ NO production was abrogated if FAK was depleted using siRNA, whereas incubation with control siRNA had no significant effect. After incubation with FAK siRNA the mean change in ⅐ NO concentration was just 0.1 Ϯ 0.2% (n ϭ 3) over the baseline ambient air value. Similarly, when cells were incubated with the FAK inhibitor PT-573228, concurrent with exposure to hyperoxia, the value at 6.4 min was 5.6 Ϯ 6.9% over the baseline (not significant, NS) and mean change in the ⅐ NO concentration across the 45-min experiments was 7.4 Ϯ 1.9% (NS). Incubation with cytochalasin D to inhibit actin filament formation prevents a number of HBO 2 -mediated effects (23,24). If cells were incubated with 2 M cytochalasin D concurrent with exposure to HBO 2 the ⅐ NO production value at 6.4 min was only 12.0 Ϯ 4.3% over the baseline and mean change in

Integrin-specific neutrophil adherence (%), inhibitor effects
Adherence to fibrinogen-coated plates was measured using neutrophils obtained from wild type or iNOS knock-out air-breathing mice. Cell suspensions were exposed for 45 min to air or HBO 2 while incubated with chemical agents. Where indicated, cells were incubated with control siRNA or siRNA to FAK for 24 h prior to adherence studies. Data are mean Ϯ S.E., n ϭ 5-15 separate studies using neutrophils from different animals.

alone).
Cell iNOS Content and Dimer Formation-We next compared iNOS content between control and HBO 2 -exposed cells. Western analysis assessing the amount of iNOS in cells normalized to ␤-actin after a 15-min exposure to HBO 2 was 1.33 Ϯ 0.11-fold (NS, n ϭ 6) greater than control but after 45 min the ratio versus control was 2.46 Ϯ 0.39-fold greater (n ϭ 8, p Ͻ 0.01).
Activation of iNOS involves homodimer formation so we were interested in evaluating the iNOS status in cells after exposure to HBO 2 . Compared with air-exposed control cells, the fraction of iNOS dimers to total iNOS present in cells exposed to HBO 2 did not differ. Following a 15-min exposure to hyperoxia the mean value was 1.00 Ϯ 0.18-fold (NS, n ϭ 6) and after 45 min of HBO 2 the value was 1.08 Ϯ 0.20-fold (NS, n ϭ 9) greater than control. However, due to greater iNOS content the amount of iNOS dimer assessed as the 260-kDa band density versus actin was 3.98 Ϯ 0.6-fold (p Ͻ 0.05, n ϭ 12) greater than control in cells exposed to HBO 2 for 45 min. The ratio after just the 15-min exposure to HBO 2 was 1.50 Ϯ 0.29-fold (NS, n ϭ 9) greater than control.
Augmented iNOS protein synthesis by hyperoxia was inhibited when cells were incubated with cytochalasin D or 1400W concurrent with 45 min of HBO 2 . A representative experiment is shown in Fig. 2. The iNOS/actin ratio versus control when cells were incubated with cytochalasin D was 1.15 Ϯ 0.03 (n ϭ 3, NS) and with 1400W, 1.02 Ϯ 0.05 (n ϭ 3, NS).
Protein Co-localization by Confocal Microscopy-Because iNOS activity is augmented by association with F-actin, co-localization in neutrophils exposed to air (control) or 2.0 ATA O 2 for 15 or 45 min was assessed by confocal microscopy. Our prime interest was associations among iNOS, FAK, and F-actin. Fig. 3 shows images as well as co-localization quantified by the magnitude of fluorescence in merged images. Fluorescence was significantly elevated in cells exposed to hyperoxia for 15 min but lower than control (air-exposed) cells when incubated for 45 min. This finding led us to question whether exposure to fMLP or 8-Br-cGMP may reverse alterations in co-localization  . Protein co-localizations with F-actin in neutrophil images. Neutrophils were exposed to air or HBO 2 for 15 or 45 min then placed on fibrinogencoated slides with PBS or with PBS plus 100 nM fMLP or 100 M 8-Br-cGMP. After a 15-min incubation, cells were fixed, permeabilized, and stained as described under "Experimental Procedures." Images are shown for cells exposed to only PBS on slides. There were three groupings of studies: FAK and iNOS, phalloidan and iNOS, phalloidan and FAK. The bar graphs show merged fluorescence intensity, which reflects protein co-localizations. These data were obtained with cells from mice in 4 to 8 independent experiments by analyzing from 30 to 65 neutrophils in each trial. Values in bar graphs are mean Ϯ S.E., *, p Ͻ 0.001 versus air-exposed cells.
due to HBO 2 because these agents inhibit actin turnover and other aspects of HBO 2 -mediated inhibition of ␤ 2 integrin function (22,23). As shown, the elevation at 15 min and reduction in co-localization at 45 min were abrogated by these agents.
Immunoprecipitation Studies-Quantitative evaluations of immunoprecipitated proteins were conducted on G-and short F-actin fractions of air and HBO 2 -exposed samples. Using antibodies to iNOS, the amount of actin or FAK relative to the iNOS band density on Western blots of short F-actin fractions is shown in Fig. 4, A and B. Lysates of cells incubated with antibodies to iNOS after exposure to HBO 2 for 15 min showed an elevated co-precipitation of actin and FAK and significantly less co-precipitation following exposure to HBO 2 for 45 min compared with control (air-exposed) cells. Reciprocal immunoprecipitation studies examining iNOS precipitation with antibodies to actin or FAK resulted in similar patterns (Fig. 4, A  and B). These patterns were not found when cells were co-incubated with fMLP or 8-Br-cGMP (supplemental Table S1). That is, the significant elevation in protein associations seen at 15 min and reduction at 45 min were abrogated by cell activation with fMLP or 8-Br-cGMP. We did not find significant differences in co-precipitated proteins when G-actin fractions were used versus the sF-actin fractions (data not shown).
Actin S-Nitrosylation in HBO 2 -exposed Cells-Prior work has shown that S-nitrosylated actin (SNO-actin) formation is the proximal event leading to HBO 2 -mediated inhibition of neutrophil ␤ 2 integrin adherence (22,23). S-Nitrosylation of neutrophil proteins was surveyed by the biotin-switch assay, which covalently adds a disulfide-linked biotin to the labile S-nitrosylation sites on proteins. Western blots were probed with anti-biotin antibodies, and in keeping with previous work (24), an actin band at 42 kDa was reliably visualized (Fig. 5). For serial studies, the magnitude of biotin was normalized to actin loaded onto the gels. As described under "Experimental Procedures," this required Western blotting using a separate sample from the cell lysate because the biotinylation procedure obscures protein recognition by anti-actin antibodies. As with previous studies, if the biotin-switch analysis was performed on cell lysates treated with N- [6-(biotinamido)hexyl]-3Ј-(2Ј-pyridyldithio)propionamide or with ascorbate (but not both), with 1 mM HgCl 2 , or if cells were exposed to UV light prior to cell lysis and biotin switch, the bands were not visualized (data not shown).
No elevation in SNO-actin occurred in HBO 2 -exposed cells from iNOS-KO mice, contrary to effects observed in wild type neutrophils. If cells from wild type mice were incubated with 1400W or PT-573228 was included with the neutrophil suspension during exposure to hyperoxia, SNO-actin was not elevated. We also examined SNO-actin formation when cells had been incubated with siRNA. When control siRNA-incubated cells were exposed to hyperoxia the biotin-actin ratio was 2.2 Ϯ 0.16fold (n ϭ 4, p Ͻ 0.05) greater than found in cells exposed to just air, whereas cells incubated with siRNA to FAK then exposed to hyperoxia exhibited a ratio just 0.96 Ϯ 0.05-fold (NS) that of air-exposed cells incubated with control siRNA, and cells exposed to only air (no siRNA treatment) had a 1.07 Ϯ 0.06-fold biotin/actin ratio (NS).
FAK Linkage to iNOS and F-actin ex Vivo-The data suggested that FAK might serve as a bridge linking iNOS and actin. To further examine this possibility, full-length human FAK was incubated with F-actin and iNOS followed by immunoprecipitation using anti-actin antibodies. Fig. 6 (first lane) shows all three proteins were precipitated, when FAK and F-actin are FIGURE 4. Immunoprecipitation of short F-actin fraction neutrophil lysates. Neutrophils exposed to air (control) or 2.0 ATA O 2 for 15 or 45 min were lysed, the short F-actin fraction was isolated and subjected to immunoprecipitation using antibodies to iNOS, FAK, or actin. Western blots at the top show typical results. In A, the left image shows immunoprecipitation using anti-iNOS and blots were also probed for actin, and the right image shows lysates immunoprecipitated with anti-actin and blots were also probed for iNOS. In B, the left image shows immunoprecipitation using anti-FAK and blots were also probed for iNOS, and the right image shows lysates immunoprecipitated with anti-iNOS and blots also probed for FAK. The ratios of band densities are shown and values are mean Ϯ S.E. for 4 to 8 independent trials, *, p Ͻ 0.05 versus air, analysis of variance. present both precipitate (second lane), but if iNOS was incubated with F-actin no iNOS was precipitated and only the actin band appears on the Western blot (third lane).
As an alternative approach glutathione transferase-tagged (GST) FAK was incubated with purified iNOS and F-actin. As shown in Fig. 7, resin-linked GST-FAK could bind F-actin (fourth lane) but if the F-actin had been previously incubated with S-nitroso-N-acetyl-D,L-penicillamine to cause S-nitrosylation actin was not linked (fifth lane). The sixth lane shows that GST-FAK can bind with both iNOS and F-actin. These studies were performed in triplicate with consistent results.
An alternative set of experiments was performed using actin that was S-nitrosylated before filament formation, rather than after F-actin formation as shown in Fig. 7. This allowed us to vary the amount of SNO-actin in filaments. Fig. 8 shows linkage between FAK and F-actin (fifth lane) and a marked reduction when GST-tagged FAK was incubated with F-actin made using mixtures of 25 or 50% S-nitrosylated actin. In replicate experiments using 50% S-nitrosylated actin, linkage to GST-FAK was reduced to 35 Ϯ 9% (n ϭ 4, p Ͻ 0.05) of that resulting with F-actin containing no SNO-actin. Finally, we found that if GST-FAK was first incubated with 5 M PT573228 no F-actin linkage occurred (eighth lane).
To further explore linkages involving FAK and assess whether inhibition could be achieved with an alternative protein manipulation versus S-nitrosylation, we carried out a series of studies with a histidine-tagged (His) fragment of human FAK that includes the kinase domain but not N-or C-terminal segments (amino acids 393-698). The efficiency for pulling down iNOS and F-actin with this fragment was comparable with fulllength FAK. This comparison was achieved by taking the ratio of band densities for iNOS or actin in pulldown studies and comparing them with band densities when just iNOS or actin was placed on the SDS gel at the same concentration as used in the pulldown samples. For example, in experiments using GST-FAK the actin band was 89.1 Ϯ 5.9% (n ϭ 4) as dense as with using only the actin solution and with His-FAK studies the actin band density was 95.0 Ϯ 4.5% (NS versus GST-FAK studies).
There are 16 lysine residues rather evenly distributed in the His-FAK segment and studies were conducted with un-manipulated protein and also with His-FAK that was first acetylated (see "Experimental Procedures"). Cobalt resin pulldown of His-FAK co-precipitated only 0.49 Ϯ 0.2% (n ϭ 3, p Ͻ 0.001)) as much iNOS and 26.1 Ϯ 7.5% (n ϭ 3, p Ͻ 0.05) co-precipitated as much F-actin as un-manipulated His-FAK. We conclude that associations involving iNOS and F-actin occur within the kinase domain of FAK and they can be abrogated by acetylating FAK.

DISCUSSION
Results from this study provide an explanation for why ⅐ NO production is increased in HBO 2 -exposed neutrophils. There is strong precedence for cytoskeletal associations enhancing NOS activity (35,36). The novel aspect to neutrophil dynamics in response to hyperoxia is that the cytoskeletal linkage is mediated by FAK. Moreover, inhibition by cytochalasin D indicates that both actin polymerization and FAK are required for iNOS dimer formation and NOS activation. This adds further insight into the mechanism for neutrophil ␤ 2 integrin inhibition by HBO 2 . The steps in the process based on current and previously published findings are as follows (22-24): 1) S-nitrosylated  The pattern shown in the figure was identical in four replicate trials. The first three lanes show blots using only the single protein solutions (iNOS, F-actin or FAK). Solutions of GST-tagged FAK combined with either F-actin, F-actin first S-nitrosylated by incubation with S-nitroso-N-acetyl-D,L-penicillamine (SNAP), or F-actin plus iNOS were incubated for 1 h and then eluted from glutathioneagarose slurry as described under "Experimental Procedures" to assess protein-protein associations. actin is formed when reactive species are generated by iNOS and myeloperoxidase; 2) VASP exhibits enhanced affinity for S-nitrosylated sF-actin, which 3) increases actin-free barbed end formation because of subsequent cyclic AMP-dependent and cyclic GMP-dependent protein kinase-mediated Rac1 and -2 activation; 4) increased actin polymerization and filament turnover increases linkage of FAK to the short F-actin fraction; 5) iNOS becomes associated with F-actin via FAK and as dimers form, iNOS exhibits increasing activation; 6) enhanced actin turnover inhibits ␤ 2 integrin clustering and thus ␤ 2 integrin adhesion; but 7) S-nitrosylation of actin causes FAK to dissociate from sF-actin, decreasing iNOS activity; 8) TrxR that is also linked to FAK and thus in proximity to short F-actin can affect de-nitrosylation, but when FAK dissociates the TrxR no longer remains linked to FAK and thus does not reduce SNO-actin; 9) VASP phosphorylation by protein kinases in cells treated with either fMLP or 8-Br-cGMP abrogates these events by reducing VASP binding to SNO-actin, diminishing Rac activity, which restores more normal actin polymerization; thus allowing FAK to re-associate to sF-actin, increasing TrxR linkage, which drives SNO removal and restores ␤ 2 integrin function.
There is a discrepancy in magnitude of iNOS activation during the first ϳ10 min exposure to hyperoxia such that the rate of conversion of arginine to citrulline is 2-fold greater than the increase in rate assessed by observations made with the ⅐ NO electrode. A portion of this difference may arise because the electrode response is rapid, whereas the imaging and biochemical investigations require cells to be decompressed from the hyperbaric chamber. Hence, chemical interactions will progress before the cell samples are fixed or processed. It is also likely that once ⅐ NO forms, it undergoes a variety of reactions that will limit its detection using the electrode.
An additional response to HBO 2 is an elevation in iNOS content. As it is inhibited when cells are incubated with 1400W or cytochalasin D we conclude that iNOS synthesis is triggered as a consequence of the HBO 2 -induced increased enzyme activity. The mechanism for this response will require additional work. Given the rapidity of the elevation, within 45 min while cells are exposed to hyperoxia, we suspect this response is due to enhanced translation of pre-formed mRNA or post-translational events such as alterations in protease activity.
The data in this and our prior studies (22)(23)(24) indicate that iNOS activation is among the very first responses to hyperoxia and clearly actin S-nitrosylation drives ongoing events. Mention needs to be made of the time course for events, given that ⅐ NO production peaks within 10 min of HBO 2 . In this study we settled on examining cytoskeletal associations at 15 and 45 min. Preliminary studies were done at times as early as 6 min of HBO 2 exposure, but we found an unacceptable degree of variability in the results. This was especially true with immunoprecipitation experiments that are relatively complex. Possible reasons may include a maturation of protein-protein associations within cells once reactive species are generated and it is also likely there is variation in the time course for responses across the cell populations.
Data demonstrate that iNOS does not remain linked to F-actin in cells exposed to hyperoxia for 45 min. Peak enzyme activity at ϳ6.5 min generates 97.0 nM ⅐ NO and by ϳ30 min of HBO 2 the level drop to about 50% of the peak concentration. This activity at 30 min is almost exactly the same production as seen with the air-exposed, control cells. Therefore, although iNOS dimer content remains nearly 4-fold elevated over control, the enzyme no longer is active once dissociated from the actin cytoskeleton.
Findings in this study add to recognized roles for FAK. FAK functions predominantly as a scaffolding protein. In a previous study we found that the FAK kinase domain inhibitor (PT-573228) blocks TrxR linkage to FAK leading us to suggest the kinase domain played a role (24). We have now found that fulllength FAK and also a protein fragment that includes the kinase domain but not the amino-or carboxyl-terminal ends will bind iNOS and F-actin. Inhibition of the protein associations by PT-573228 is consistent with the kinase domain playing a role. Moreover, protein modification via either S-nitrosylation or acetylation abrogates protein-protein associations.
In summary, this study adds to our knowledge of how ␤ 2 integrin function is regulated in neutrophils. The results demonstrate further complexity to the role cytoskeleton plays in this important cell function. Further work is necessary to better define the mechanisms for protein-protein associations. The data indicate that the FAK kinase domain is necessary. We can also say that the C-terminal portion of actin is involved, as we have previously shown that S-nitrosylation occurs at the four cysteine moieties closest to the carboxyl-terminal end of actin (22).
Clinical utility for HBO 2 in a variety of situations where neutrophil interactions mediate tissue injury is suggested by multiple animal studies and several clinical trials. Obviously, further work is necessary along these lines and better elucidation of mechanisms should assist in this effort. Finally, given that ⅐ NO plays such a complex role in cytoskeletal regulation and ␤ 2 integrin function, it is feasible that some events uncovered using HBO 2 are aspects of normal physiology and perhaps, one or more ⅐ NO-donor drug may function in a similar manner as HBO 2 .