FASEB J.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by GARCIA, J. G. N.
Right arrow Articles by VERIN, A. D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by GARCIA, J. G. N.
Right arrow Articles by VERIN, A. D.
(The FASEB Journal. 2002;16:1064-1076.)
© 2002 FASEB

Critical involvement of p38 MAP kinase in pertussis toxin-induced cytoskeletal reorganization and lung permeability

JOE G. N. GARCIA1, PEIYI WANG, KANE L. SCHAPHORST, PATRICE M. BECKER, TALAIBEK BORBIEV, FENG LIU, ANNA BIRUKOVA, KERI JACOBS, NATALIA BOGATCHEVA and ALEXANDER D. VERIN

Division of Pulmonary and Critical Care Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA

1Correspondence: Johns Hopkins Asthma and Allergy Center, 5501 Hopkins Bayview Circle, 4B.77, Baltimore, MD 21224-6801, USA. E-mail: drgarcia{at}jhmi.edu


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Bordetella pertussis is an important cause of infection in humans worldwide, with full expression of the syndrome associated with characteristic increases in lung permeability and airway edema. The exact cellular mechanisms by which pertussis toxin (PTX) exerts pulmonary toxicity remain unknown, but may involve its ability to ADP-ribosylate-specific G-proteins. We determined that PTX directly and reproducibly reduced lung endothelial and epithelial cell barrier function in vitro and in vivo assessed by decreases in transmonolayer electrical resistance (TER) and isolated perfused lung preparations. Alterations in lung permeability began ~30 min after PTX and were dependent on intrinsic ADP-ribosyltransferase activity, as neither the cell binding ß-oligomer subunit or a genetically engineered PTX mutant (devoid of ADP-ribosyltransferase activity) altered TER. PTX-induced barrier dysfunction was associated with mild increases in F-actin stress fiber formation and causally linked to p38 MAP kinase activities. PTX-mediated p38 MAP kinase activation did not involve either p42/p44 ERK, p60src, Rho family of GTPases, or phosphatidylinositol-3' kinase pathways. PTX-mediated decreases in TER were temporally linked to phosphorylation of the actin binding proteins Hsp27 and caldesmon, known substrates for the Ser/Thr kinase MAPKAP2, whose activity is regulated by p38 MAP kinase. In addition to defining novel signaling pathways involved in PTX-induced respiratory pathophysiology, these data suggest that the direct cell-activating effects of PTX be carefully considered as a potential limitation to its use as a tool in signal transduction analysis.—Garcia, J. G. N., Wang, P., Schaphorst, K. L., Becker, P. M., Borbiev, T., Liu, F., Birukova, A., Jacobs, K., Bogatcheva, N., Verin, A. D. Critical involvement of p38 map kinase in pertussis toxin-induced cytoskeletal reorganization and lung permeability.


Key Words: ADP-ribosyltransferase • ß-oligomer • endothelial cell • transendothelial electrical resistance • HSP 27 • caldesmon


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
BORDETELLA PERTUSSIS is the causative agent of whooping cough and remains a killer of young children and infants in unvaccinated populations (1) . One product of this infection, pertussis toxin (PTX), exerts a virulence factor effect by catalyzing the ADP-ribosylation of guanine binding G-proteins, a finding that has led to its wide use as a tool for examination of G-protein-regulated cellular signaling pathways. The PTX holotoxin is comprised of a ß-oligomer binding subunit and a S1 ADP-ribosyltransferase fragment whose targets include an arginine residue present on the {alpha} subunit of Gi and Go subclasses of heterotrimeric G-proteins. As S1-mediated ADP-ribosylation effectively uncouples the G{alpha} protein from its coupled receptor (GPCR), disruption of a signaling pathway by PTX is presumptive evidence of G-protein-mediated regulation. Direct effects of PTX on cellular function, i.e., in the absence of exogenous ligand/GPCR stimulation, have also been noted in numerous cell systems, suggesting potential tonic regulation by PTX of diverse second messenger signaling cascades such as Ca2+ mobilization (2) , cAMP synthesis, diacylglycerol generation (3) , tyrosine phosphorylation (4 , 5) , and Ser/Thr phosphatase activity (6) .

The exact mechanism by which PTX produces respiratory toxicity in humans is unknown but clearly involves a breach of lung cellular barriers with upper and lower airway edema formation. Intraperitoneally administered PTX directly increases lung weight gain in rodents consistent with lung cell activation and vascular barrier dysfunction (7 , 8) , although the exact mechanism for this edemagenic response was not identified. We previously noted PTX to potently and directly stimulate increases in albumin permeability across confluent bovine pulmonary artery endothelium in vitro (9 , 10) without increases in cytosolic Ca2+, myosin light chain kinase activity (11 12 13) , or membrane phospholipases such as phosphatidylinositol-specific phospholipase C or phospholipase D (14 15 16) . However, there was strong biochemical evidence of PTX-induced protein kinase C activation (10) ; PKC inhibition attenuated the extent of PTX-induced endothelial cell barrier dysfunction (10) and reduced the lung weight gain in mice (8) . Although these results suggest PTX may represent a novel model of endothelial cell permeability, the exact PTX targets and downstream effectors responsible for evoking changes in the endothelial cell cytoskeleton, gap formation, and permeability remain to be defined. The mitogen-activated MAP kinase family of serine/threonine protein kinases, a major signal system by which eukaryotic cells transduce extracellular signals to elicit intracellular responses may be important participants in these responses (17 , 18) . We recently reported that PTX is a rapid and robust stimulus of p42/p44 MAP kinase (ERK) activity in endothelium, but ADP-ribosyltransferase activity is not required for ERK activation (19) . The p38 MAP kinases have been shown to target several cytoskeletal proteins and, by inference, to participate in cell motility, apoptosis, stress responses, and inflammation (20 , 21) . Whether PTX alters lung barrier properties via MAP kinase activities has not been previously addressed.

Using measurements of human lung endothelial cell monolayer electrical resistance, a highly sensitive index of paracellular gap formation and barrier dysfunction and the isolated perfused ferret lung model, we have confirmed in vitro and in vivo our prior observations that PTX mediates lung barrier dysfunction. Neither the PTX ß-oligomer nor a holotoxin mutant lacking ADP-ribosyltransferase activity altered transmonolayer resistance, indicating that the permeability response depends entirely on G-protein modification. We further show that p38 MAP kinase activation is involved in PTX-induced barrier disruption, actin cytoskeletal rearrangement, and increased phosphorylation of the actin binding proteins Hsp27 and caldesmon, known MAP kinase targets. Together, these studies provide mechanistic information on the pathogenesis of direct PTX-induced lung toxicity and increases in vascular permeability. Given the extent of direct cellular activation evoked by PTX, a cautious approach should be considered when using PTX to examine selective G-protein-linked signal transduction pathways.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Reagents
Unless specified, reagents were obtained from Sigma Chemical Co. (St. Louis, MO). Phosphate buffered-saline (PBS), Hanks’ balanced salt solution without phenol red, DMEM media, M199 media, OPTI-MEM, phosphate-free DMEM, antibiotic and antimycotic solution (10,000 units/mL penicillin, 10 mg/mL streptomycin, and 25 µg/mL amphotericin B), and nonessential amino acids were purchased from Life Technologies (Gaithersburg, MD). HYQ RPMI 1640 media was obtained from Hyclone (Logan, UT). Endothelial cell growth supplement was purchased from Upstate Biotechnology (Lake Placid, NY). Heparin was obtained from Pharmacia & Upjohn Company (Kalamazoo, MI). Pertussis holotoxin, A protomer and ß-oligomer of PTX, the p38 MAP kinase inhibitor SB 203580, the phosphoinositide 3' kinase inhibitor LY 294002, the p60src kinase inhibitor PP-2, protease inhibitor mixtures, and Omnisorb were purchased from Calbiochem (La Jolla, CA). The MEK inhibitor UO 126 was purchased from Promega (Madison, WI). Phospho-p38 MAP kinase-specific and phospho-ATF-2-specific antibodies, ATF-2 fusion protein, immobilized phospho-p38 MAP kinase monoclonal antibody, and p38 MAP kinase antibody were purchased from New England Biolabs (Beverly, MA). Anti L-caldesmon antibodies were purchased from Chemicon (Temecula, CA), anti pan-ERK antibodies were purchased from Transduction Laboratory (Lexington, KY), and anti-Hsp 27 monoclonal antibodies were purchased from StressGen (Victoria, BC, Canada). Protein G Sepharose was purchased from Pharmacia Biotech (Piscataway, NJ). Protein concentration was determined using BCA protein assay reagent (Pierce, Rockford, IL). The S1 PTX holotoxin mutant was kindly provided by Dr. Rappuoli (Sienna, Italy) and dominant/negative p38 adenoviral constructs by Dr. William Gerthoffer (Reno, NV). All the radioactive reagents were purchased from NEN (Boston, MA).

Cell cultures
Bovine pulmonary artery endothelial cells (BPAEC) were obtained frozen at 16 passages from American Type Tissue Culture Collection (Rockville, MD; CCL 209) and used at passages 19–24 (11) . BPAEC were cultured in DMEM supplemented with 20% (v/v) fetal bovine serum (FBS), 0.1% endothelial cell growth supplement, 1% antibiotic and antimycotic solution, and 1% nonessential amino acids. Human umbilical vein endothelial cells (HUVEC) were purchased from Clonetics (Walkersville, MD) at passage 1 and used at passages 5–6. HUVEC were cultured in M199 supplemented with 20% (v/v) FBS, 0.15% endothelial cell growth supplement, 1% antibiotic and antimycotic solution, and 0.2% heparin. Cells were cultured in HYQ RPMI 1640 media supplemented with 10% FBS, 1% antibiotic and antimycotic solution, and 1% nonessential amino acids. All cell cultures were maintained at 37°C in a humidified atmosphere of 5% CO2-95% air and grew to contact-inhibited monolayers. Cells from each primary flask were detached with 0.05% trypsin, resuspended in fresh culture media, and passaged into appropriate size flasks or dishes.

Measurement of endothelial cell electrical resistance
Lung endothelial and epithelial cellular barrier properties were measured using the highly sensitive biophysical assay (22 23 24) with an electrical cell substrate impedance sensing system (Applied Biophysics, Troy, NY). Cells were cultured on several small gold electrodes (10-4 cm2) and culture media was used as the electrolyte. The total electrical resistance was measured dynamically across the monolayer and was determined by the combined resistance between the basal surface of the cell and the electrode, reflective of focal adhesion, and the resistance between cells. Changes in electrical resistance represent alterations in cell–cell adhesion and/or cell–matrix adhesion. As cells adhere and spread out on the microelectrode, the transmonolayer electrical resistance (TER) increases (maximal at confluence) whereas cell retraction, rounding, or loss of adhesion is reflected by a decrease in TER (25) . The small gold electrodes and the larger counter electrodes (1 cm2) are connected to a phase-sensitive lock-in amplifier (5301A; EG&G Instruments Corp., Princeton, NJ) with a built-in differential preamplifier (5316A; EG&G Instruments). A 1 V, 4000 Hz AC signal was supplied through a 1-M{Omega} resistor to approximate a constant current source. Voltage and phase data were stored and processed with a Pentium 100 MHZ computer that controlled the output of the amplifier and relay switches to different electrodes. Experiments were conducted only on wells that achieved >1000 ohms (10 microelectrodes per well) of steady-state resistance. Resistance was expressed by the in-phase voltage (proportional to the resistance) that was normalized to the initial voltage and expressed as a fraction of the normalized resistance value as described previously (23) . For some experiments, total resistance was resolved into components reflecting resistance to current flow beneath the cell layer ({alpha}) and resistance to current flow between adjacent cells (Rb) using the method of Giaever and Keese, which models the endothelial monolayer mathematically (26) . Thus, changes in {alpha} reflect alterations in the net state of cell–matrix adhesion whereas changes in Rb reflect alterations in the integrity of cell–cell adhesion.

MAP kinase activation assay
Endothelial or epithelial cell monolayers in 35 mm dishes grown to 100% confluence were challenged with agonists. After stimulation, cells were rinsed once with PBS, lysed with 150 µL of boiling lysis buffer (10 mM Tris-HCl, pH 7.4, 1% SDS, 1 mM sodium orthovanadate), heated to boiling for 5 min, and centrifuged for 5 min. The protein concentration of the resulting supernatant was determined using BCA Protein Assay Reagent. A common feature of the MAP kinase isoforms is the phosphorylation of threonine and tyrosine residues by a dual specificity Ser/Thr MAPK kinase. The p38 MAP kinase activity present in these samples was assessed by Western immunoblotting analysis with phospho-p38 MAP kinase-specific antibody as recently described (19) . To assess the total amount of p38 MAP kinase, the blot was stripped using Re-blot Western blot recycling kit (Chemicon, Temecula, CA) and subjected to Western immunoblotting analysis with pan-p38 MAP kinase antibody. The p38 MAP kinase activity was also measured by immune complex kinase assay using GST-ATF-2 as a substrate. After stimulation, cells from 35 mm dishes were rinsed once with ice-cold PBS and lysed with 150 µL of immunoprecipitation buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 0.5% Nonidet P-40, 20 mM NaF, 0.2 mM sodium orthovanadate) supplemented with 0.5% protease inhibitor mixtures for 30 min at 4°C. The cells were then scraped off the dishes, homogenized by passing through a 26 gauge syringe three times, and centrifuged for 10 min. Soluble cell lysates (100 µL) containing ~100 µg of total proteins were incubated with immobilized phospho-p38 MAP kinase monoclonal antibody overnight at 4°C. The immunoprecipitates were washed three times with immunoprecipitation buffer and three times with kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM ß-glycerol-phosphate, 2 mM DTT, 0.1 mM sodium orthovanadate, 10 mM MgCl2). The pellets were suspended in 45 µL of kinase buffer containing 200 µM ATP and 2 µg ATF-2 fusion protein and incubated for 30 min at 30°C. The reaction was terminated by adding 15 µL of boiling 4x Laemmli sample buffer. The samples were boiled for 5 min and centrifuged for 5 min. The p38 MAP kinase activity within the resulting supernatant was analyzed by Western immunoblotting with phospho-ATF-2-specific antibody.

Infection of endothelial cells with adenovirus encoding dominant negative p38 MAP kinase
Adenoviruses expressing the p38{alpha} MAPK dominant negative (TGY->AGF) mutant and a replication-deficient control virus were a generous gift from Dr. W. T. Gerthoffer (Reno, NV). Cultured endothelial cells at 70–80% confluency were exposed to either recombinant or control adenovirus at 10–40 moi for 1 h in DMEM (Gibco BRL, Grand Island, NY) containing 2% FBS. The virus-containing medium was replaced with virus-free DMEM with 10% FBS and endothelial cells were analyzed after 30 h.

ADP-ribosylation of cell proteins
Bacterial toxin-catalyzed ADP-ribosylation of lung endothelial or epithelial cell homogenates was measured by incorporation of [32P]-NAD (10–20 µCi/mL in ribosylation mixture) as previously described (14 , 15) . Pertussis toxin was used at a final concentration of 300 ng/mL and preactivated with 20 mM dithiothreitol. ADP-ribosylated proteins were separated by 1-D SDS-PAGE gels, located by autoradiography, and identity was confirmed by Western immunoblotting analysis.

HSP27 and caldesmon immunoprecipitation
Confluent endothelial cell monolayers in 35 mm dishes were labeled with [32P]-orthophosphate (0.25 mCi/plate) for 2.5 h in phosphate-free DMEM. After challenge with PTX, cells were rinsed twice with 1 mL media and twice with 1 mL PBS. Then 100 µL of cold immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, 1 mM EGTA, 0.2 mM sodium orthovanadate, 0.5% Nonidet P-40) supplemented with 0.5% protease inhibitor mixtures was scraped off the dishes; cells were incubated with 170 µL of cold immunoprecipitation buffer and 30 µL of Omnisorb for 30 min at 4 °C and centrifuged for 15 min. The resulting supernatant was incubated with 1 µL of monoclonal anti-caldesmon antibody (Chemicon, Temecula, CA) or monoclonal Hsp 27 antibody (StressGen Biotechnologies, Victoria, BC, Canada) for 1 h at 4°C and with 30 µL of Omnisorb for 30 min at 4°C with constant agitation. After centrifugation (1 min), pellets were washed with 1 mL cold immunoprecipitation buffer, resuspended in 100 µL 2x Laemmli sample buffer, and boiled for 5 min. After centrifugation (5 min), supernatants were loaded onto a SDS-PAGE gel, transferred to nitrocellulose membrane, and exposed to X-Omat film (Kodak, Rochester, NY).

Cytoskeletal visualization by immunofluorescent microscopy
Cultured cells grown on glass coverslips were fixed in 3.7% formaldehyde solution in PBS for 10 min at 4°C and washed three times with PBS. The cells were permeabilized with 0.2% Triton X-100 in PBST for 5 min, washed three times with PBS, and blocked with 2% BSA in PBST for 20 min. Incubation with mouse anti-caldesmon antibody or mouse anti HA antibody diluted 1:100 with blocking solution was performed for 1 h at room temperature. After three washes with PBS, cells were incubated with anti-mouse secondary antibody (1:300) conjugated with fluorescent dye Alexa 488 (Molecular Probes, Eugene, OR) for 1 h at room temperature. Actin filaments were visualized by staining cells with Texas Red-conjugated phalloidin (Molecular Probes) for 1 h at room temperature. The coverslips were mounted and analyzed using Nikon video imagine system consisting of a phase contrast inverted microscope equipped with objectives and filters for immunofluorescence and connected to a digital camera and image processor. The images were recorded and saved on a Pentium II PC as TIFF files compatible with Adobe Photoshop 4.0.

Isolated perfused ferret lung preparation
Commercially available ferrets were anesthetized with pentobarbital (50 mg/kg IP), ventilated (20 breaths/min, tidal volume 12 mL/kg) via tracheostomy with warmed, humidified gas containing 28% O2, then rapidly exsanguinated as described previously (27) . After exsanguination, the ventilatory gas mixture was switched to 16% O2-5% CO2, ventilatory rate adjusted to 10 breaths/min, and end-expiratory pressure of 3 mm Hg was added. Lungs were isolated by insertion of cannulas into the left atrium via the left ventricle and the pulmonary artery via the right ventricle, then residual blood was flushed from the lungs with physiological salt solution (PSS) containing 5 mM dextrose, 3 g/dL porcine albumin, and 2 g/dL Ficoll (27) . Isolated lungs were perfused at constant flow (50 mL·kg-1·min-1) for 60 min with PSS (control, n=3) or PSS containing 300 ng/mL PTX (n=3). Pulmonary arterial (Ppa), left atrial (Pla), and airway (Paw) pressures were continuously monitored (Grass Model 7) with Statham P50 transducers referenced to the left atrium. Glucose concentration and pH were monitored throughout the perfusion period and did not differ between preparations.

After 60 min of extracorporeal perfusion, the pulmonary vasculature was filled with PSS containing washed ferret red blood cells (Hct 20%); pulmonary arterial and left atrial cannulas were connected to a common reservoir containing the same solution. The reservoir was pressurized to 30 mm Hg for 20 min, then samples were rapidly withdrawn from the left atrial cannula for measurement of red blood cell and albumin concentration (27) . Osmotic reflection coefficient for albumin (95 alb) was calculated iteratively from the rate of change of albumin concentration relative to red blood cell concentration (27) .


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Pertussis toxin induces endothelial and epithelial cell barrier dysfunction via ADP-ribosylation
We have previously shown PTX to produce significant increases in lung endothelial cell macromolecular permeability (9 , 10) . To extend these findings, we assessed PTX-induced endothelial cell barrier dysfunction using real-time measurements of TER (28) as we have noted with other permeability-producing agents (22 , 23) . After a ~20 min lag time in which electrical resistance did not change, PTX (at concentrations >= 100 ng/mL) consistently induced significant declines in TER that remained sustained for several hours in bovine pulmonary artery (Fig. 1 A), and human umbilical vein endothelial cells (Fig. 1B ). Similar responses were noted in cultured human lung epithelium challenged with PTX (data not shown). This dramatic PTX-mediated reduction in TER was attributable to reduced cell–cell adhesion (Rb) without declines in cell–matrix adhesion ({alpha}) (Fig. 1C ). PTX-induced endothelial cell permeability was also confirmed in vivo using isolated perfused ferret lungs (Fig. 2 ) where the estimated osmotic reflection coefficient for albumin ({sigma}alb) in control lungs was 0.70, a value similar to that obtained for uninjured isolated ferret lungs (27) . In contrast, after 60 min of perfusion with PTX, {sigma}alb decreased to a mean value of 0.32 (Fig. 2) in association with significant lung weight gain and consistent with PTX-mediated increased vascular permeability and endothelial cell barrier dysfunction. Pulmonary artery pressures (Ppa, Paw) remained constant throughout the initial 30 min of PTX perfusion, but gradually increased during the final 30 min of PTX perfusion and airway with obvious edema fluid formation in the tracheal cannula.



View larger version (22K):
[in this window]
[in a new window]
  		Figure 1. Figure 1.

Effect of pertussis toxin on electrical resistance across endothelial cell monolayers. Transmonolayer electrical resistance (TER) across confluent bovine pulmonary artery endothelium (A, C) or human umbilical vein endothelium (B) was monitored for up to 4 h. Cells were incubated in complete media for 20 min to stabilize basal TER, then treated with either vehicle (PBS) or PTX (1–1000 ng/mL). Shown are results from representative experiments (n=3), where in each case PTX dramatically decreased TER values beginning ~30 min after challenge. A) Inset: Absolute resistance (expressed in ohms) across endothelium grown on 10 gold microelectrodes per well after vehicle (open bars) and PTX (300 ng/mL) challenge (solid bars). C) Representative experiment (n=3) designed to resolve the total electrical resistance into components reflecting resistance to current flow beneath the cell layer ({alpha}), i.e., net state of cell–matrix adhesion, or resistance to current flow between adjacent cells (Rb) (26) . PTX produces a substantial decrease in Rb whereas {alpha} remains relatively unchanged or improved, indicating that alterations in the integrity of cell–cell adhesion are the primary pathway of PTX-mediated TER reductions.



View larger version (80K):
[in this window]
[in a new window]
  		Figure 2. Effects of pertussis toxin on the osmotic reflection coefficient for albumin ({sigma}alb) in isolated perfused lungs. Shown is the mean osmotic reflection coefficient for albumin ({sigma}alb) in isolated ferret lungs measured after 60 min of perfusion with either PTX (300 ng/mL) or vehicle control (PBS). Values represent the mean ± SE for 3 experiments in each group. *P < 0.01 vs. control. These studies demonstrate that ({sigma}alb) decreased significantly after PTX, consistent with increased pulmonary vascular protein permeability mediated by PTX.

The onset of PTX-induced barrier dysfunction occurred in a temporal sequence comparable to the time course of PTX-mediated G{alpha} ADP-ribosylation in human and bovine endothelium we noted earlier (14 , 15 , 19) . To further explore this linkage, endothelial cell lysates were retrieved after defined durations of PTX exposure (15–120 min) and used to determine maximal in vitro PTX-induced ADP-ribosylation. These experiments confirmed that PTX catalyzes the ADP-ribosylation of 40 kDa G{alpha} proteins in endothelial cells (Fig. 3 ) and that maximal ADP-ribosylation occurs after 120 min of PTX exposure. More important, the earliest detectable PTX-induced ADP-ribosyltransferase activity temporally coincided with the PTX-mediated reduction in TER that began after 30 min (Fig. 1) . Two complementary strategies were used to further examine whether ADP-ribosyltransferase activity is necessary for PTX-induced endothelial cell barrier dysfunction. In contrast to the substantial decline in TER produced by the native wild-type PTX holotoxin, Fig. 4 demonstrates that neither the purified S1 subunit (A protomer), the purified S2–6 subunit complex (ß-oligomer), nor the addition of a genetically engineered PTX mutant protein with site-directed mutations (two) to completely eliminate S1 ADP-ribosyltransferase activity (29) altered bovine endothelial cell (Fig. 4A ) or human endothelial cell (Fig. 4B ). These results strongly indicate that the deleterious PTX effects on endothelial barrier properties are completely dependent on ADP-ribosylation of G{alpha}.



View larger version (37K):
[in this window]
[in a new window]
  		Figure 3. Time course of pertussis toxin-mediated ADP-ribosylation of endothelial cell heterotrimeric G-proteins. Endothelial cell monolayers in 100 mm dishes were incubated with pertussis toxin (300 ng/mL), lysed, and homogenized, followed by centrifugation (100,000 g). Pellets were collected and 60 µg of protein was used for an in vitro ADP-ribosylation reaction catalyzed by either vehicle (lane 1) or by DTT-activated PTX (lanes 2–7) as previously described (14 , 15) . The reactions were carried out in the presence of 5 mCi of [32P]-NAD at 30°C for 30 min. Proteins were then TCA precipitated and electrophoresed on 15% SDS-PAGE. The resulting gel was dried and exposed to Kodak X-OMAT film for 5 h. These results indicated PTX-induced ADP-ribosylation begins well after 30 min of PTX challenge.



View larger version (22K):
[in this window]
[in a new window]
  		Figure 4. Effect of pertussis toxin components on endothelial cell barrier function. Transmonolayer electrical resistance across bovine pulmonary artery endothelium (A) or human umbilical vein endothelial cells (B) was monitored for up to 4 h. Cells were incubated in complete media for ~20 min to stabilize basal electrical resistance, then treated with either vehicle control (PBS), PTX holotoxin (300 ng/mL), the S1 PTX holotoxin mutant (300 ng/mL) (which fails to exhibit ADP-ribosyltransferase activity due to site-directed double mutations), the purified A protomer subunit representing the ADP-ribosyltransferase component (80 ng/mL), or the purified ß-oligomer subunit representing the cell binding component (220 ng/mL). Whereas the intact PTX holotoxin significantly reduced electrical resistance across endothelial cell monolayers, neither the purified PTX subunits nor the S1 PTX mutant increased transmonolayer electrical resistance.

Pertussis toxin induces significant actin rearrangement in endothelium
We and others have clearly demonstrated the critical involvement of actin cytoskeletal rearrangement in endothelial cell barrier regulation (12 , 13 , 53) . Human and bovine endothelium were next challenged with PTX and stained to visualize polymerized F-actin by immunofluorescent microscopy. Compared to vehicle controls (Fig. 5 A, panel 1) PTX-induced endothelial cell activation is accompanied by the modest development of stress fibers that span the cell after 30 min of PTX challenge (Fig. 5A , panel 2) whereas the S1 PTX mutant holotoxin, which does not exhibit ADP-ribosyltransferase activity, failed to induce stress fiber formation (Fig. 5A , panel 3). Quantification of actin stress fiber formation induced by PTX verified the increase in F-actin beginning at 15 min. (Fig. 5B ). These results suggest that PTX-induced cytoskeletal rearrangement, known to be essential in barrier regulation (13 , 30) , is dependent on ADP-ribosyltransferase activity.



View larger version (76K):
[in this window]
[in a new window]
  		Figure 5. Effect of pertussis toxin and the S1 PTX mutant on endothelial cell actin microfilament reorganization. A) Confluent endothelial cells grown on glass coverslips were left in control conditions (first panel) or challenged with 300 ng/mL of PTX holotoxin (second panel) or 300 ng/mL of the S1 PTX mutant devoid of ADP-ribosyltransferase activity (third panel) for 30 min. After treatment, the cells were fixed, permeabilized and stained for F-actin with Texas red phalloidin. These studies show that compared to controls, the PTX holotoxin resulted in modest but highly significant actin stress fiber formation, which are consistent with contractile phenotype. These changes were not observed after challenge with the S1 PTX mutant. These results indicate that PTX-induced G{alpha} ADP-ribosylation is essential to subsequent stress fiber formation and intercellular gap formation, a process sustained for up to 3 h (data not shown). B) This box graph depicts the quantitation of the actin stress fibers after pertussis toxin and S1 mutant treatment over 1 h. Adobe Photoshop software was used to determine the extent of polymerized actin accumulation from a total of three experiments and expressed as the percentage of control F-actin SEM

Pertussis toxin induces rapid p38 MAP kinase activation
PTX-induced endothelial cell permeability proceeds in a PKC-dependent manner (10) and PKC is known to enhance several important effector systems, including members of the MAP kinase family. We have recently demonstrated that PTX produces robust activation of p42/p44 ERK MAP kinase in bovine endothelium in a PKC-dependent manner (19) , and Fig. 6 A demonstrates that pertussis toxin holotoxin increases p38 MAP kinase activity in a dose-dependent fashion. However, whereas the S1 PTX mutant and the ß-oligomer increase ERK activity (19) , neither reagent increases p38 MAP kinase activation (Fig. 6B ). This was evidenced by detection of phosphorylated p38 MAP kinase; peak p38 activity was noted 60 min after PTX, persisting for up to 180 min after PTX challenge, a time course consistent with PTX-mediated G-protein ADP-ribosylation (Fig. 3) and alterations in transmonolayer electrical resistance (Fig. 1) . PTX was a far more potent stimulus for p38 MAP kinase activation than other permeability agents such as the PKC-activating phorbol esters or the procoagulant serine protease thrombin (Fig. 6C ). PTX-stimulated p38 MAP kinase activation was unaffected by either inhibition of ß{gamma}G-protein subunit signaling produced by transfection with a ßARK minigene (data not shown) or by pharmacologic inhibition of phosphatidylinositol-3' kinase activity or p60src activity (Fig. 6D ). To determine whether p38 MAP kinase activation may be a consequence of cytoskeletal rearrangement, we assessed the effect of several pharmacologic inhibitors of actin microfilament and microtubule rearrangement or disruption. The p38 MAP kinase activation evoked by PTX was unaffected by complementary strategies to decrease Rho and Rho kinase activation (Fig. 7 A, B). Furthermore, microfilament disruption with cytochalasin B or microtubule stabilization with taxol, also failed to alter PTX-induced p38 MAPK activation (Fig. 7B, C ).



View larger version (35K):
[in this window]
[in a new window]
  		Figure 6. Effect of pertussis toxin on p38 MAP kinase activities in endothelial cells. A) Bovine endothelial cells were challenged with the indicated concentrations of PTX. The p38 MAP kinase activity was measured either by Western immunoblotting analysis with phospho-specific p38 MAP kinase antibody or by assessing phosphorylation of substrate ATF-2 after MAP kinase assay as described in Material and Methods. The same blot was stripped and reblotted with p38 MAP kinase antibody to show the total amount of p38 MAP kinase. p38 assays had similar results with dose-dependent PTX-medicated p38 MAP kinase activation. B) The effect of PTX holotoxin (300 ng/mL), the ß-oligomer (220 ng/mL), or the S1 PTX mutant devoid of ADP-ribosyltransferase activity (300 ng/mL) for the periods indicated. These studies demonstrate that PTX rapidly and significantly increased p38 MAP kinase activity in a time-dependent manner. We recently demonstrated that PTX induced p42/p44 ERK activity in a time- and close-dependent manner (19) . In contrast, ADP-ribosyltransferase activity is required for PTX-induced p38 MAP kinase activation. C) Endothelial cells were exposed to either PTX (300 ng/mL), PMA (100 nM), or thrombin (100 nM), and p38 MAP kinase activity was assessed by Western immunoblotting with phospho-specific p38 MAP kinase antibody. PTX was a far more potent agonist for p38 MAP kinase activation than either thrombin or PMA. D) The effect of p60src and PI-3' kinase on pertussis toxin-induced p38 MAP kinase activation in endothelial cells. Confluent bovine pulmonary artery endothelial cells were preincubated with either vehicle (0.1% DMSO), p60src inhibitor PP-2 (1 µM) or PI-3' kinase inhibitor LY 294002 (25 µM) for 1 h and challenged with either vehicle (PBS) or PTX (300 ng/mL) for 30 min. The p38 MAP kinase activity was assessed by Western immunoblotting with phospho-specific p38 MAP kinase antibody. The PTX-mediated p38 MAP kinase activation was not altered by PI-3' kinase or p60src inhibition.



View larger version (37K):
[in this window]
[in a new window]
  		Figure 7. Effect of microfilament and microtubule alteration on PTX-mediated p38 MAP kinase activation. A) Confluent bovine pulmonary artery endothelial cells were either permeabilized by lipofectamine treatment in the presence or absence of the specific Rho inhibitor C3 exoenzyme (2.5 µg/mL) (54) or without lipofectamine treatment, then stimulated with PTX (300 ng/mL) for 2 h. B) Bovine endothelial cells were preincubated with either vehicle or Rho kinase inhibitor Y-27632 (10 µM) for 1 h, then challenged with either vehicle (PBS) or PTX (300 ng/mL) for 2 h. C) Bovine endothelial cells were preincubated with either vehicle (0.1% DMSO), the microfilament-disrupting agent cytochalasin B (10 µg/mL), or the microtubule-stabilizing agent taxol (10 µM) for 30 min, then challenged with PTX (300 ng/mL) for 2 h. The p38 MAP kinase activity was assessed by Western immunoblotting with phospho-specific p38 MAP kinase antibody. The total amount of p38 MAP kinase was shown by Western immunoblotting with p38 MAP kinase antibody. These data demonstrate that PTX-mediated p38 MAP kinase activation was not altered by Rho, Rho kinase, microfilament disruption, or microtubule stabilization.

Pertussis toxin-induced cytoskeletal rearrangement and barrier dysfunction involves p38 MAP kinase activation
Temporal kinetic analysis suggested that PTX-induced ERK activation peaks (19) well before the onset of PTX-mediated reductions in TER (Fig. 1) . In contrast, p38 MAP kinase activity more closely coincides with the maximal decreases in TER across confluent endothelial cell monolayers and persists for several hours. Having carefully assessed the regulation of p38 MAPK activity, we directly assessed whether ERK and p38 MAP kinase activities represent relevant PTX downstream effectors involved in lung vascular barrier regulation by using specific pharmacologic inhibitors UO 126 (ERK kinase inhibitor) and SB 203580 (p38 MAP kinase inhibitor) (33) before PTX challenge. ERK inhibition failed to attenuate the PTX-mediated reduction in electrical resistance (Fig. 8 A). The p38 MAP kinase inhibitor SB 203580 significantly inhibited PTX-induced barrier dysfunction (Fig. 8B, C ), suggesting that p38 MAP kinase activity, but not ERK, is an important participant in the barrier dysfunction observed after PTX. These results were further confirmed in endothelial cells infected with either control adenoviral vector or a plasmid encoding dominant/negative p38 MAP kinase (Fig. 8D ). Pretreatment of human endothelium with SB 203580 also abolished PTX-induced stress fiber formation (Fig. 9 ). These results are again consistent with the notion that PTX-mediated p38 MAP kinase activation is intimately involved in subsequent stress fiber formation and barrier dysfunction after PTX challenge.



View larger version (37K):
[in this window]
[in a new window]
  		Figure 8. Effect of MAP kinase inhibition on pertussis toxin-induced endothelial cell barrier dysfunction. TER across confluent bovine pulmonary artery endothelium was monitored for 2.5 h. Cells were incubated in complete media for ~20 min to stabilize basal electrical resistance, treated with either vehicle (0.1% DMSO), the p38 MAP kinase inhibitor SB203580 (20 µM, B), or the p42 ERK inhibitor U0126 (10 µM, A) for 1 h, then challenged with either vehicle (PBS) or PTX (300 ng/mL). Shown are representative experiments (n=3) demonstrating the significant attenuation of PTX-mediated barrier dysfunction by p38 MAP but not ERK inhibition. Inset: Confluent bovine pulmonary artery endothelial cells were preincubated with either vehicle (0.1% DMSO) SB203580 (20 µM) for 1 h and treated with vehicle (PBS) or PTX (300 ng/mL) for 1 h (A, inset) or 5 min (B, inset). p38 MAP kinase activity was then measured by immune complex kinase assay using ATF-2 as a substrate and p42/p44 ERK activity was assessed by Western immunoblotting with phospho-specific ERK antibody. Shown are the partial inhibition of p38 MAP kinase activity by SB 203580 and the complete inhibition of ERK activity by UO126. C) Sum of 4 independent experiments where SB 203580 (20 µM) alleviated the PTX-mediated declines in TER. D) Bovine pulmonary artery endothelial cells were infected with either dominant/negative p38 MAP kinase or a control vector as described in Materials and Methods. Cells were then transferred to wells containing gold microelectrode and treated with either vehicle (PBS) or PTX (300 ng/mL). In contrast to the effect of empty vector infection, overexpression of the dominant/negative p38 MAP kinase protein (dominant/negative p38) produced total inhibition the PTX TER response. Together, these studies demonstrate that inhibition of p38 MAP kinase but not ERK significantly attenuates the PTX-induced decreases in TER, supporting an important role for p38 MAP kinase but not ERK in PTX-induced endothelial cell barrier dysfunction.



View larger version (194K):
[in this window]
[in a new window]
  		Figure 9. Effect of p38 MAP kinase inhibition on pertussis toxin-induced actin reorganization and stress fiber formation. Bovine pulmonary artery endothelium grown on glass coverslips was pretreated with either vehicle (0.1% DMSO, A, C) or SB 203580 (20 µM), B, D) for 1 h, followed by treatment with vehicle (PBS, A, B) or PTX (300 ng/mL, C, D) for 1 h. The endothelial cell monolayers were then fixed, permeabilized, and stained with Texas red phalloidin for F-actin. Shown is the prominent attenuation of PTX-induced actin rearrangement by p38 MAP kinase inhibition.

Pertussis toxin increases caldesmon and Hsp 27 phosphorylation in a p38 MAP kinase- dependent manner
Phosphorylation of Hsp 27 and caldesmon, two actin binding cytoskeletal proteins and known p38 MAP kinase targets, has been postulated as molecular switches that may facilitate actomyosin contraction without an increase in myosin light chain phosphorylation (34 , 35) . We next examined whether caldesmon or Hsp 27 phosphorylation is temporally linked to intercellular gap formation and actin cytoskeleton remodeling as shown in PKC-dependent models of endothelial cell permeability (34) . Immunoprecipitation of caldesmon from PTX-challenged bovine endothelium revealed a substantial increase in caldesmon phosphorylation from 10 min through 60 min (Fig. 10 A), with the increase in caldesmon phosphorylation (60 min) attenuated by p38 MAP kinase inhibition (Fig. 10B ). Similarly, Hsp27 immunoprecipitation after PTX revealed increased phosphorylation, which was p38 MAPK dependent (Fig. 1C ). Consistent with its actin binding properties, immunofluorescent studies demonstrated that caldesmon spatially colocalized with cortical actin under basal conditions but demonstrated marked translocation to newly formed stress fibers after PTX challenge (Fig. 11 ). Taken together, these results suggest that PTX reduces cell–cell adhesion and junctional integrity by significant p38 MAP kinase-dependent cytoskeletal rearrangement, leading to paracellular gap formation and barrier dysfunction, and may involve the stimulated phosphorylation of specific cytoskeletal targets such as Hsp27 and caldesmon.



View larger version (33K):
[in this window]
[in a new window]
  		Figure 10. Potential role of caldesmon and Hsp27 phosphorylation on pertussis toxin bovine pulmonary artery endothelium. Shown is the time-dependent effect of PTX (300 µM, 0–60 min) on caldesmon phosphorylation in bovine pulmonary artery endothelial cells preloaded with [32P]-orthophosphate (0.5 mCi/mL in phosphate-free DMEM) for 2 h at 37°C and subjected to caldesmon immunoprecipitation as described in Materials and Methods (A). In a similar fashion, caldesmon phosphorylation was assessed in bovine pulmonary artery endothelial cells pretreated with either vehicle (0.1% DMSO), or SB 203580 (20 µM) for 30 min, followed by PTX stimulation (B). The position of caldesmon was determined by subsequent staining of the same membrane with anti-caldesmon antibody. C) 32P-labeled bovine pulmonary artery endothelial cells were pretreated with either vehicle (0.1% DMSO) or SB203580 (20 µM) for 60 min, treated with either vehicle or PTX (300 nM) for 60 min, and the level of Hsp 27 phosphorylation was determined in the immunoprecipitates as described in Materials and Methods. To ensure equal protein loading, radioactive nitrocellulose membrane was subsequently stained with monoclonal Hsp 27 antibody. PTX-induced endothelial cell stimulation resulted in a significant increase in caldesmon and Hsp 27 phosphorylation in intact bovine endothelium that can be attenuated by p38 MAP kinase inhibition.



View larger version (127K):
[in this window]
[in a new window]
  		Figure 11. Effect of pertussis toxin on actin and caldesmon immunofluorescent colocalization. Bovine pulmonary artery endothelium grown on glass coverslips was challenged with either vehicle control (PBS, A, C) or PTX (300 ng/mL, B, D) for 1 h. The cells were then fixed, permeabilized, and stained for F-actin with Texas red phalloidin (A, B) or, for nonmuscle caldesmon, with anti-caldesmon antibody (C, D). Shown is the colocalization of caldesmon with F-actin initially in the dense peripheral band and subsequently as a component of the newly formed actin stress fibers.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Bordetella pertussis infection produces a clinical syndrome characterized by persistent mucosal edema of the upper and lower respiratory tract that contributes to the morbidity and mortality of the disease. Like many clinically important bacterial toxin-initiated disorders, the biochemical and molecular basis for the key pathogenetic feature of PTX infection is incompletely resolved, but the involvement of alterations in lung vascular and airway permeability is clear (7 , 8) . In this study, we have explored potential signaling pathways by which PTX increases lung permeability by focusing on mechanisms of endothelial cell activation and barrier dysfunction. Consistent with our earlier reports (9 , 10) , PTX proved a potent agonist for disassembly of endothelial cell junctions, resulting in intercellular gap formation, significant increases in macromolecule permeability, and decreases in transmonolayer electrical resistance. We previously noted that PTX-induced endothelial cell activation and barrier dysfunction were not associated with increases in cytosolic Ca2+ or in myosin light chain kinase-dependent contraction (10) . Although the PTX targets that alter endothelial cell barrier properties are largely unknown (12 , 36) , we have further demonstrated the ability of PTX to significantly impair endothelial cell barriers in vitro and in vivo, with ADP-ribosylation and p38 MAP kinase activation required for this response.

The early upstream signaling events by which PTX induces endothelial cell p38 MAP kinase activation are unknown. However, our data are entirely consistent with a striking divergence in the requirements for PTX-mediated ERK and p38 MAP kinase activities. Gi{alpha} ADP-ribosylation after PTX treatment ensued well after maximal ERK activation (5 min) (19) and actually coincided with significant down-regulation of ERK activities. In contrast, the time course of p38 MAP kinase activation was temporally linked to the level of Gi ADP-ribosylation as well as PTX-mediated alterations in TER. These data strongly argue that the function of the single S1 A protomer subunit, which contains the ADP-ribosyltransferase but cannot efficiently enter the cell, is required for the full effect of PTX on endothelial permeability. This notion was rigorously confirmed with the S1 mutant PTX holotoxin and the purified ß-oligomer, which are devoid of ADP-ribosylation activity and failed to reproduce the native wild-type holotoxin effects on either TER or p38 MAP kinase activation, although stimuli were proficient in inducing ERK activation (19) . The ß-oligomer, which allows the toxin to bind target cells through as yet uncharacterized surface receptors, contains five subunits and dramatically increases the efficiency of S1 entry (29 , 37 , 38) . Carbohydrate moieties have been reported as crucial components of ß-oligomer binding sites, and ADP-ribosylation-independent activation of signaling pathways by the ß-oligomer has been reported involving PKC, Ca2+, and tyrosine kinases (3 , 5 , 6 , 39) . Integrin receptors (CD11/CD18) have been suggested as PTX binding sites (40) , but it is unknown whether integrin ligation by PTX occurs on the endothelial or epithelial cell surface. Current studies are under way to explore this interesting notion.

The postreceptor events that link PTX-evoked extracellular signals with the p38 MAP kinase pathway have not been defined. Recently, overexpression of the pp125 focal adhesion kinase-related tyrosine kinase (known as RAFTK or PYK2) was found to lead to ERK and p38 MAP kinase activation whereas a dominant/negative PYK2 mutant reduced p38 MAP kinase activation elicited by cytotoxic agents such as methyl methane sulfonate (41) . This alternate mode of p38 MAP kinase activation is unlikely to be involved in PTX signaling, however, as PYK2 is only activated by extracellular signals that increase intracellular Ca2+ and pertussis toxin challenge in endothelium does not alter intracellular levels of Ca2+ (10 , 15) . Although not reported here, our prior results evaluating PTX-mediated p42/44 ERK activation (19) are consistent with the notion that upstream signaling involving Ras GTPase is unlikely to play a key role in p38 MAP kinase activation after PTX challenge. In these studies, PTX-mediated ERK activation was attenuated by either MEK or PKC inhibition but without measurable increases in Ras or Raf-1 activities (19) . The S1 PTX mutant and the ß-oligomer were comparable to the PTX holotoxin in their induction of ERK activity, again suggesting that toxin binding is sufficient to activate ERK in a Raf-1- and Ras-independent manner.

The five known p38 MAP kinase isoforms ({alpha}, ß, ß2, {gamma}, and {delta}) share a common TGY motif (T180 and Y182 phosphorylation is associated with enhanced enzymatic activity) and p38 MAP kinase is a well-recognized participant in regulating inflammatory responses, including cytokine secretion and apoptosis (31 , 32) . Osmotic stress, endotoxin, DNA-damaging agents, Rho family GTPases (Rac, cdc 42, p21-associated kinase), and proinflammatory cytokines (interleukin 1) and tumor necrosis factor all have the capacity to activate p38 MAP kinase (42) . Recent studies indicate the p38 MAPK{alpha} isoform is the predominant isoform present in endothelial cells (43) and is specifically inhibited by the pyridenyl imidazole compound SB 203580 (33) . Evaluation of signaling events potentially involved in PTX-induced p38 MAP kinase activation failed to define a significant role for Ras, p60src, phosphatidylinositol-3' kinase activities, or Rho kinase (Fig. 7A ) (19) . Despite reports that Rac GTPase mediates p38 MAP kinase activation, our preliminary data suggest, that Rac GTPases are also unlikely to be an important intermediate in the PTX-mediated p38 MAP kinase activation pathway (data not shown), as PTX did not reproducibly activate Rac. This suggests that MKK4, a Rac GTPase-sensitive, p38 MAPK-activating kinase, is unlikely to be involved; MKK3 or MKK6, which are not regulated by Rac GTPase (31) , are more likely to participate in PTX signaling in the endothelium. Since ADP-ribosylation of a Gi or Go heterotrimeric protein appears to be essential, we speculate the existence of a PTX-sensitive Gi{alpha} protein that regulates p38 MAP kinase through an as yet undefined mechanism. We further speculate that G-protein inactivation/uncoupling by PTX results in the release of tonic inhibition of MKK-regulated p38 MAP kinase activity.

We also noted that the onset of p38 MAP kinase activity began as ERK activation was rapidly reduced, consistent with the notion that direct cross-talk between different MAP kinase pathways may exist (44) . Diperoxovanadate-induced p38 MAP kinase activation was reported to exert an inhibitory effect on MEK-1 activity in baboon aortic smooth muscle cells, which was not due to direct phosphorylation (45 , 46) . Although we did not rigorously evaluate the extent of direct cross-talk between ERK and p38 MAP kinase in PTX-stimulated endothelium, this is unlikely to be a consequence of PTX-mediated ADP-ribosylation, as the S1 mutant devoid of ADP-ribosylation transferase activity) stimulated ERK but failed to significantly increase p38 MAP kinase activity.

We found that PTX was a potent stimulus for p38 MAPK-dependent Ser/Thr phosphorylation and attempted to define potential downstream targets that may be involved in PTX-mediated lung toxicity and barrier dysfunction. Analysis of electrical resistance vectors after PTX (Fig. 1D ) indicated that disassembly of cell–cell junctions is likely the primary pathway of barrier compromise. Subsequent studies revealed paracellular gaps and a profound actin cytoskeletal rearrangement and stress fiber formation (Fig. 5) , findings consistent with increased tension development. Recently, p42/p44 ERK MAP kinase activation was reported to increase the activity of the smooth muscle myosin light chain kinase (MLCK) isoform, providing a potential linkage to rapid tension development (47 , 48) . We previously cloned a unique nonmuscle endothelial cell MLCK isoform (49) and found several consensus sites for Ser/Thr phosphorylation catalyzed by proline-directed kinases such as ERK and p38 MAP kinases (50) . Despite the inherent attractiveness of this hypothesis, we consistently failed to identify an increase in endothelial (10) or epithelial cell MLCK activity after PTX and have yet to identify a role for either ERK or p38 MAP kinase in endothelial cell MLCK regulation (J. G. N. Garcia and A. D. Verin, unpublished observations). Appealing alternate p38 MAPK targets include the 27 kDa heat shock protein Hsp27 and the 77 kDa nonmuscle caldesmon, two actin binding proteins that have been proposed to function as molecular switches in the regulation of nonmuscle and smooth muscle contraction (34 , 35) . We have previously shown PKC and MAP kinase activation with either phorbol esters or thrombin to increase endothelial cell caldesmon phosphorylation (24 , 34) , and have now demonstrated that PTX also produces significant caldesmon phosphorylation catalyzed by p38 MAP kinases. Caldesmon was spatially localized with filamentous actin staining throughout PTX stimulation, initially as a component of the dense cortical actin ring and later in association with newly formed stress fibers. Although our experiments do not allow us to firmly establish a causal relationship between Hsp27 and caldesmon phosphorylation and PTX-induced barrier dysfunction, these events occur in a p38 MAP kinase-dependent manner. Activation of p38 MAP kinase is critical to subsequent PTX-mediated physiological derangement as SB 203580 attenuated Hsp27 and caldesmon phosphorylation as well as PTX-induced cytoskeletal rearrangement and barrier dysfunction. MAP kinases, Hsp27 and caldesmon have been found in close association with focal adhesions, strongly implicating p38 MAP kinase as a critical participant in transducing the cytoskeletal signals that result in barrier dysfunction (data not shown).

In addition to defining the need for careful assessment of direct PTX effects in future studies of signal transduction pathways, we have explored upstream signaling events and downstream targets involved in PTX-induced lung endothelial cell activation and barrier regulation. We have now identified the requirement for a signaling cascade involving ADP-ribosylation of Gi{alpha} and p38 MAP kinase activities for the full permeability-producing response. Evaluation of potential p38 MAP kinase targets that produce endothelial and epithelial cell cytoskeletal rearrangement leading to paracellular gap formation and permeability yield two actin binding proteins, Hsp27 and caldesmon, whose phosphorylation represent potential molecular switches for endothelial cell contractile regulation. However, we further speculate the existence of a PTX-sensitive barrier protective Gi{alpha} protein that regulates lung cellular barrier function with tonic suppression of p38 MAP kinase activity. ADP-ribosylation of this barrier regulatory Gi{alpha} protein by PTX results in p38 MAP kinase activation, which via subsequent actin cytoskeletal rearrangement represents an essential component by which PTX signals increased lung permeability and subsequent respiratory pathophysiology in humans. Future studies using PTX in the examination of G-protein-linked signal transduction pathways should acknowledge this potent direct effect of PTX on cell activation.


   ACKNOWLEDGMENTS
 
This work was supported by National Institutes of Health grants HL 58064; HL 03666, and HL 60628, the Dr. David Marine Endowment, and the Kirby Fund. The authors gratefully acknowledge Lakshmi Natarajan, Mary Ann Booth, and Steve Durbin for superb technical assistance and Ellen G. Reather for expert manuscript preparation.

Received for publication November 27, 2001. Revision received March 6, 2002.
   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Kaslow, H. R., Burns, D. L. (1992) Pertussis toxin and target eukaryotic cells: binding, entry and activation. FASEB J 6,2684-2690[Abstract]
  2. Sternad, C. F., Carchman, R. A. (1987) Human T lymphocyte mitogenesis in response to the ß-oligomer of pertussis toxin is associated with an early elevation of cytosolic calcium concentrations. FEBS Lett. 225,16-20[CrossRef][Medline]
  3. Thom, R. E., Casnellie, J. E. (1989) Pertussis toxin activates protein kinase C and a tyrosine protein kinase in the human T cell line Jurkat. FEBS Lett. 244,181-184[CrossRef][Medline]
  4. Müller-Wieland, D., Morris, F., White, F., Behnke, B., Gebhart, A., Neumann, S., Krone, W., Kahn, C. R. (1991) Pertussis toxin inhibits autophosphorylation and activation of the insulin receptor kinase. Biochem. Biophys. Res. Commun. 181,1479-1489[CrossRef][Medline]
  5. Wong, W. S., Luk, J. M. (1997) Signaling mechanisms of pertussis toxin-induced myelomonocytic cell adhesion: role of tyrosine phosphorylation. Biochem. Biophys. Res. Commun. 236,479-483[CrossRef][Medline]
  6. Chen, F., Ngoc-Diep, V., Wagner, P. D. (1998) Pertussis toxin modification of PC12 cells inhibits a protein phosphatase 2A-like phosphatase. J. Neurochem. 71,248-257[Medline]
  7. Clerch, L. B., Neithardt, G., Spencer, U., Melendez, J. A., Massaro, G. D., Massaro, D. (1994) Pertussis toxin treatment alters manganese superoxide dismutase activity in lung: evidence for lung oxygen toxicity in air-breathing rats. J. Clin. Invest. 93,2482-2489[Medline]
  8. Tsan, M. F., Cao, X., White, J. E., Sacco, J., Lee, C. Y. (1999) Pertussis toxin-induced lung edema. Role of manganese superoxide dismutase and protein kinase C. Am. J. Respir. Cell. Mol. Biol. 20,465-473[Abstract/Free Full Text]
  9. Patterson, C. E., Garcia, J. G. N. (1994) Regulation of thrombin-induced endothelial cell activation by bacterial toxins. Blood Coagul. Fibrinolysis 5,63-72[Medline]
  10. Patterson, C. E., Stasek, J. E., Schaphorst, K. L., Davis, H. W., Garcia, J. G. N. (1995) Mechanisms of pertussis toxin induced barrier dysfunction in bovine pulmonary artery endothelial cell monolayers. Am. J. Physiol. 286,L926-L934
  11. Garcia, J. G. N., Davis, H. W., Patterson, C. E. (1995) Regulation of endothelial cell gap formation and barrier dysfunction: role of myosin light chain phosphorylation. J. Cell. Physiol. 163,510-522[CrossRef][Medline]
  12. Lum, H., Malik, A. B. (1994) Regulation of vascular endothelial barrier function. Am. J. Physiol. 267,L223-L241[Abstract/Free Full Text]
  13. Dudek, S. M., Garcia, J. G. N. (2001) Cytoskeletal regulation of pulmonary vascular permeability. J. Cell. Physiol. 91,1487-1500
  14. Garcia, J. G. N., Painter, R. G., Fenton, J. W., English, D., Callahan, K. S. (1990) Thrombin-induced prostacyclin biosynthesis in human endothelium: Role of guanine nucleotide regulatory proteins in stimulus/coupling responses. J. Cell. Physiol. 42,186-193
  15. Garcia, J. G. N., Dominguez, J., English, D. (1991) Sodium fluoride induces phosphoinositide hydrolysis, Ca2+ mobilization, and prostacyclin synthesis in cultured human endothelium: Further evidence for regulation by a pertussis toxin-insensitive guanine nucleotide-binding protein. Am. J. Respir. Cell. Mol. Biol. 5,113-124[Medline]
  16. Garcia, J. G. N., Fenton, J. W., Natarajan, V. (1992) Thrombin stimulation of human endothelial cell phospholipase D activity. Regulation of phospholipase C protein kinase C and cyclic adenosine 3', 5'-monophosphate. Blood 79,2056-2067[Abstract/Free Full Text]
  17. Errede, B., Cade, R. M., Yashar, B. M., Kamada, Y., Levin, Y. D. E., Irie, K., Matsumoto, K. (1995) Dynamics and organization of MAP kinase signal pathways. Mol. Reprod. Dev. 42,477-485[CrossRef][Medline]
  18. Seger, R., Krebs, E. G. (1995) The MAPK signaling cascade. FASEB J 9,726-735[Abstract]
  19. Garcia, J. G. N., Wang, P., Liu, F., Hershenson, M. B., Borbiev, T., Verin, A. D. (2001) Pertussis toxin directly activates endothelial cell p42/p44 MAP kinases via novel signaling pathway. Am. J. Physiol. 280,C1233-C1241[Abstract/Free Full Text]
  20. Leppa, S., Bohmann, D. (1999) Diverse function of JNK signaling and c-Jun in stress response and apoptosis. Oncogene 18,6158-6162[CrossRef][Medline]
  21. Wang, X. S., Diener, K., Manthey, C. L., Wang, S., Rosenzweig, B., Bray, J., Delaney, J., Cole, C. M., Chan-Hui, P. Y., Mantlo, N, et al (1997) Molecular cloning and characterization of a novel p38 mitogen-activated protein kinase. J. Biol. Chem. 272,23668-23674[Abstract/Free Full Text]
  22. Schaphorst, K. L., Pavalko, F., Patterson, C. E., Garcia, J. G. N. (1997) Thrombin-mediated barrier dysfunction in endothelium. Role of adhesion protein phosphorylation. Am. J. Respir. Cell Mol. Biol. 17,441-455
  23. Garcia, J. G. N., Schaphorst, K. L., Shi, S., Verin, A. D., Hart, C. M., Callahan, K. S., Patterson, C. E. (1997) Mechanisms of ionomycin-induced endothelial cell barrier dysfunction. Am. J. Physiol. 17,L172-L184
  24. Verin, A. D., Liu, F., Bogatcheva, N., Borbiev, T., Hershenson, M. B., Wang, P., Garcia, J. G. N. (2000) Role of Ras-dependent ERK activation in PKC-dependent endothelial cell barrier dysfunction. Am. J. Physiol. 279,L360-L370
  25. Giaever, I., Keese, C. R. (1993) A morphological biosensor for mammalian cells. Nature (London) 366,591-592[CrossRef][Medline]
  26. Giaever, I., Keese, C. R. (1991) Micromotion of mammalian cells measured electrically. Proc. Natl. Acad. Sci. USA 88,7896-7900[Abstract/Free Full Text]
  27. Becker, P. M., Pearse, D. B., Permutt, S., Sylvester, J. T. (1992) Separate effects of ischemia and reperfusion on vascular permeability in ventilated ferret lungs. J. Cell. Physiol. 73,2616-2622
  28. Tiruppathi, C., Malik, A. B., Del Vecchio, P. J., Keese, C. R., Giaever, I. (1992) Electrical method for detection of endothelial cell shape change in real time: Assessment of endothelial barrier function. Proc. Natl. Acad. Sci. USA 89,7919-7923[Abstract/Free Full Text]
  29. Pizza, M., Covacci, A., Bartoloni, M., Perugine, L., Nencione, M., De Magistris, M. T., Villa, L., Noci, D., Magnetic, R., Bug noli, M., et al (1989) Mutants of pertussis toxin suitable for vaccine development. Science 246,497-500[Abstract/Free Full Text]
  30. Garcia, J. G. N., Birmboim, A. S., Bizios, R., Del Vecchio, P. J., Fenton, J. W., Malik, A. B. (1986) Thrombin-induced increases in albumin clearance across cultured endothelial monolayers. J. Cell. Physiol. 128,96-104[CrossRef][Medline]
  31. Moriguchi, T., Toyoshima, F., Masuyama, N., Hanafusa, H., Gotoh, Y., Nishida, E. (1997) A novel SAPK/JNK kinase, MKK7, stimulated by TNF alpha and cellular stresses. EMBO J 16,7045-7053[CrossRef][Medline]
  32. Yuasa, T., Ohno, S., Kehrl, J. H., Kyriakis, J. M. (1998) Tumor necrosis factor signaling to stress-activated protein kinase (SAPK)/Jun NH2-terminal kinase (JNK) and p38. Germinal center kinase couples TRAF2 to mitogen-activated protein kinase/ERK kinase 1 and SAPK while receptor interacting protein associates with a mitogen-activated protein kinases kinase upstream of MKK6 and p38. J. Biol. Chem. 273,22681-22692[Abstract/Free Full Text]
  33. Gum, R. J., McLaughlin, M. M., Kumar, S., Wang, Z., Bower, M. J., Lee, C. J., Adams, J. L., Livi, G. P., Goldsmith, E. M., Young, P. R. (1998) Acquisition of sensitivity of stress-activated protein kinases to the p38 inhibitor, SB 203580, by alteration of one or more amino acids within the ATP-binding pocket. J. Biol. Chem. 273,15605-15610[Abstract/Free Full Text]
  34. Stasek, J. E., Patterson, C. E., Garcia, J. G. N. (1992) Protein kinase C phosphorylates caldesmon77 and vimentin and enhances albumin permeability across cultured bovine pulmonary artery endothelial cell monolayers. J. Cell. Physiol. 153,62-75[CrossRef][Medline]
  35. Sobue, K., Sellers, J. R. (1991) Caldesmon, a novel regulatory protein in smooth muscle and non-muscle actomyosin systems. J. Biol. Chem. 266,12115-12118[Free Full Text]
  36. Vuong, P. T., Malik, A. B., Nagpala, P. G., Lum, H. (1998) Protein kinase C beta modulates thrombin-induced Ca2+ signaling and endothe