The JI PBL Intereron Source
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
Right arrow Citation Map
Services
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
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Fillon, S.
Right arrow Articles by Tuomanen, E. I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Fillon, S.
Right arrow Articles by Tuomanen, E. I.
The Journal of Immunology, 2006, 177: 6182-6191.
Copyright © 2006 by The American Association of Immunologists, Inc.

Platelet-Activating Factor Receptor and Innate Immunity: Uptake of Gram-Positive Bacterial Cell Wall into Host Cells and Cell-Specific Pathophysiology

Sophie Fillon1,*, Konstantinos Soulis1,*, Surender Rajasekaran1,{dagger}, Heather Benedict-Hamilton{dagger}, Jana N. Radin*, Carlos J. Orihuela*, Karim C. El Kasmi*, Gopal Murti{ddagger}, Deepak Kaushal§, M. Waleed Gaber, Joerg R. Weber||, Peter J. Murray* and Elaine I. Tuomanen2,*

* Department of Infectious Diseases, {dagger} Division of Critical Care Medicine, {ddagger} Department of Molecular Biotechnology, and § Hartwell Center for Bioinformatics and Biotechnology, St. Jude Children’s Research Hospital, Memphis TN 38105; Department of Biomedical Engineering, University of Tennessee, Memphis, TN 38163; and || Department of Neurology, Charite-Universitaetsmedizin, Berlin, Germany


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
The current model of innate immune recognition of Gram-positive bacteria suggests that the bacterial cell wall interacts with host recognition proteins such as TLRs and Nod proteins. We describe an additional recognition system mediated by the platelet-activating factor receptor (PAFr) and directed to the pathogen-associated molecular pattern phosphorylcholine that results in the uptake of bacterial components into host cells. Intravascular choline-containing cell walls bound to endothelial cells and caused rapid lethality in wild-type, Tlr2–/–, and Nod2–/– mice but not in Pafr–/– mice. The cell wall exited the vasculature into the heart and brain, accumulating within endothelial cells, cardiomyocytes, and neurons in a PAFr-dependent way. Physiological consequences of the cell wall/PAFr interaction were cell specific, being noninflammatory in endothelial cells and neurons but causing a rapid loss of cardiomyocyte contractility that contributed to death. Thus, PAFr shepherds phosphorylcholine-containing bacterial components such as the cell wall into host cells from where the response ranges from quiescence to severe pathophysiology.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Proinflammatory bacterial components are at least partially responsible for causing the clinical features of sepsis, a syndrome that causes >100,000 deaths each year in the United States (1). In the case of Gram-positive infection, a key bacterial element recognized by the innate immune system is the cell wall, a complex network of peptidoglycan covalently linked to teichoic acids, proteins, and lipoproteins (2). Host peptidoglycan recognition proteins bind pathogen-associated molecular patterns (PAMPs)3 and degrade the cell wall (3, 4), releasing a wide variety of smaller proinflammatory fragments. Extracellular Gram-positive cell wall components, lipoproteins, and lipopeptides interact with membrane-bound TLRs, particularly TLR2 (5, 6, 7), and induce a proinflammatory response. Nod proteins are proposed to function as intracellular surveillance proteins for the recognition of subcomponents of peptidoglycan (8, 9). Nod2-deficient macrophages respond to TLR agonists but are refractory to muramyl dipeptide stimulation (10, 11, 12, 13, 14). How peptidoglycan might reach the cytoplasm is unclear.

Streptococcus pneumoniae, a leading cause of pneumonia, sepsis, and meningitis (15), has served as an important model organism for determining the innate immune response to a Gram-positive bacterial cell wall because not only are the detailed chemical structures of its cell wall components well characterized (reviewed by Ref. 2), but these structures have also been linked to specific pathologies in disease (16, 17, 18, 19, 20). Cell wall fragments are released in vivo during pneumococcal growth and especially during antibiotic-induced bacterial death and are likely to contribute to inflammation and damage during infection.

For the progression of invasive infection, pneumococci traverse human cells by binding to the platelet-activating factor (PAF) receptor (PAFr) (21, 22), a widespread, G protein-coupled receptor (GPCR) that naturally recognizes the phosphorylcholine determinant on the eukaryotic proinflammatory chemokine PAF. Like PAF, the pneumococcal cell wall displays phosphorylcholine that binds to the PAFr, initiating bacterial uptake. Recently, uptake via PAFr has also been shown for several other respiratory pathogens that display phosphorylcholine on their surfaces (23, 24, 25). Ligation of PAFr induces pleiotropic effector responses that are highly dependent on the cell context, in part due to coupling to a variety of pathways (26, 27). Two signal transduction cascades that have been studied extensively are the G protein-coupled phosphorylation of protein kinases, which leads to chemotaxis, and the G protein-independent PAFr interaction with the scaffold protein beta-arrestin-1, which results in receptor recycling by endocytosis (27, 28, 29, 30, 31, 32, 33). The beta-arrestin-dependent pathway is responsible for pneumococcal uptake via PAFr (21, 34), but no direct PAF-like G protein-stimulating activity has been described for the bacteria (35). The absence of PAFr modifies the progression of pneumococcal disease in several experimental models (34, 36, 37) while responses to endotoxin challenge are unchanged (38). We sought to determine the trafficking and inflammatory consequences of the interaction of the phosphorylcholine-containing cell wall with the PAFr at the level of cellular physiology as well as in animal models.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Preparation of bacterial components

S. pneumoniae serotype 4, strain TIGR4 (39), or unencapsulated strain R6 were grown on tryptic soy agar (Difco) supplemented with 3% defibrinated sheep blood or in defined semisynthetic casein liquid medium supplemented with 0.5% yeast extract. Radiolabeling of S. pneumoniae strain R6 or its cell wall was performed by growth in (C+Y) medium with reduced choline (1 µg/ml) and supplemented with 2 µCi/ml [methyl-3H]choline (Amersham Bioscience) (16, 40). Highly purified cell wall was prepared as described previously (16). Briefly, bacteria were boiled in SDS, mechanically broken by shaking with acid-washed glass beads, and treated sequentially with DNase, RNase, trypsin, LiCl, EDTA, and acetone. Cell wall was confirmed to be free of protein by analysis for non-cell wall amino acids using mass spectrometry. The absence of contaminating endotoxin was confirmed by Limulus test (Associates of Cape Cod). As determined by electron microscopy, the majority of cell wall particles were <30 nm in size, a value within the size range capable of passing nuclear pores (41, 42). Ethanolamine-containing cell wall was prepared by replacing choline with 20 µg/ml ethanolamine (Sigma-Aldrich) in chemically defined growth medium (43). Peptidoglycan free of teichoic acid was prepared from insoluble cell wall by treatment with 48% hydrofluoric acid (44). Cell wall was directly labeled with 1 mg/ml FITC (Sigma-Aldrich) solution in carbonate buffer (pH 9.2) for 1 h at room temperature (RT) in the dark and washed twice with PBS containing Ca2+ and Mg2+ (45).

Intravital fluorescence microscopy

Forty wild-type vs 17 Pafr–/– littermates (34), 8 Tlr2–/–mice (The Jackson Laboratory), and 4 Nod2–/– mice (10) vs 6 wild-type C57BL/6 littermates 8–10 wk old were maintained in biosafety level 1 and level 2 facilities at the St. Jude Children’s Research Hospital Animal Facility (Memphis, TN). All experimental procedures were done with mice anesthetized with either inhaled isoflurane (Baxter Healthcare) at 2.5% or MKX (1 ml of ketamine (Fort Dodge Laboratories) at 100 mg/ml, 5 ml of xylazine (Miles Laboratories) at 100 mg/ml, and 21 ml of PBS). MKX was administered by i.p. injection at a dose of 50 µl per 10 g of body weight. All experiments were done in compliance with National Institutes of Health and institutional guidelines.

For cranial window studies, animals were placed in a stereotaxic frame (Kopf Instruments) and their body temperatures were maintained at ~37°C by a silicon heating mat. As described previously (46), the scalp and underlying tissue over the parietal cortex were removed on one side of the midsagittal suture using a low-speed dental drill with irrigation, creating a rectangular cranial window extending from the bregma to the lambdoid sutures. The dura mater was punctured and excised. Using cyanoacrylate glue, a 0.5 x 0.5-cm glass plate was fixed to the bone surrounding the cranial window. After surgery, mice were allowed to recover for 7–14 days before data collection was initiated. Intravital microscopy was used to visualize the attachment of a FITC-labeled cell wall to the microvessels of the brain (46). Mice with cranial windows were anesthetized, immobilized on a stereotaxic frame, and placed under an industrial scale microscope (model MM-11; Nikon) with a camera assembly and bright field and fluorescent light sources. Baseline pictures of the cerebral microvasculature were taken. Mice were injected retro-orbitally or by tail vein with a bolus of FITC-labeled cell wall equivalent to 5 x 107 CFU of pneumococci, and video images and still pictures of the injection, using an excitation wavelength of 490 nm and emission wavelength of 520 nm, were captured using MetaMorph software (Universal Imaging). Images were captured at 1, 3, and 5 h after injection. Cell wall fragments attached to the microvasculature were subsequently quantitated (46). For experiments with cytidine diphosphate (CDP)-choline, mice were given CDP-choline (500 mg/kg; Biomol) 16 and 1 h before i.v. treatment with 5 x 107 cell wall equivalents. For some mice, CDP-choline was given 10 min after cell wall attachment to the brain vasculature was visualized (n = 3). For experiments with PAFr antagonist (CV-6209, 1 mg/kg; Biomol), mice (n = 6) were treated i.v. at 16 and 1 h before cell wall challenge or received a single dose 10 min after challenge with 5 x 107 cell wall equivalents (n = 3). Six hours after challenge, animals were sacrificed for histopathology.

To determine more precisely the distribution of cell wall, mice were sacrificed and the brain and heart were excised and submerged in 10% buffered formalin (Globe Scientific). The organs were then embedded in paraffin, cut into 5-µm-thick sections, and viewed at x100 magnification using fluorescence or confocal multiphoton microscopy or stained with H&E for light microscopy.

To follow cell wall distribution from the cerebrospinal fluid, cell wall (2 x 107 bacterial equivalents) was injected into the lumbar subarachnoid space of 4-wk-old C57BL/6 mice as described (n = 3) (47). Three and 6 h later, animals were sacrificed, brains were fixed, and 5-µm sections were stained using anti-choline TEPC-15 Ab (Sigma-Aldrich) and FITC-labeled secondary anti-IgA Ab (Sigma-Aldrich) as described previously (4). Distribution of cell wall was determined by fluorescence microscopy.

To view the intracellular distribution of cell wall, deparaffinized brain sections of animals challenged with FITC-labeled cell wall were blocked with 1x TBS, 0.5% BSA, 0.5% Triton X-100, and 4% goat serum for 30 min. For brain, neurons were detected using the neuronal marker NeuN (1/500; Chemicon) in blocking solution overnight at 4°C in a humid chamber. Brain sections were washed twice with TBS for 5 min and incubated with the secondary Ab, Alexa Fluor 594 goat anti-mouse IgG (1/50; Molecular Probes) for 1 h at RT, rinsed twice in TBS, and stained with 4',6-diamidino-3-phenylindole (DAPI; a nuclear stain) and viewed at x100 in a Zeiss 510 NLO confocal-multiphoton microscope. The FITC label was viewed in the confocal mode using an argon laser, and the DAPI label was observed in the multiphoton mode with the infrared laser set at the appropriate wavelength and filter combination for viewing DAPI. The images were collected as Z series, and maximal projections were derived using the Zeiss software. Heart sections were deparaffinized and coverslipped with Vectashield Hard Set mounting medium with DAPI. No staining was done other than with DAPI nuclear stain.

Langendorff isolated heart

All experiments were performed in compliance with National Institutes of Health and institutional guidelines according to the method of Gao et al. (48). Three- to 4-wk-old Sprague Dawley rats (four per group; Harlan Breeders) were anesthetized i.m. with a combination of ketamine (35 mg/kg) and xylazine (5 mg/kg). Once the animal was completely unresponsive, a midsternal incision was made and the heart was excised and placed immediately in cold Krebs-Henseleit solution consisting of 118 mM NaCl, 4.75 mM KCl, 1.19 mM KH2PO4, 1.19 mM MgSO4·7H2O, 1.8 mM CaCl2·2H2O, 25 mM NaHCO3, and 11 mM glucose. Within 30 s the hearts were mounted onto a Langendorff perfusion apparatus (Radnotti Glass Technology), perfused with Krebs-Henseleit solution at 37°C, and oxygenated with 95% oxygen and 5% carbon dioxide with the pH adjusted to 7.35–7.45 at a constant pressure of 60 mm Hg with flow maintained at 8 cc/min. After the hearts were mounted, an incision was made into the left atrium and a microtip SPR-524 1.25 F Millar catheter (AD Instruments) was passed into the left ventricle. The Millar catheter was calibrated before every experiment. The heart was then suspended in a warm bath and the end diastolic pressure was adjusted to between 5 and 8 mm Hg. After a 20-min equilibration period the peak systolic pressure and end diastolic pressure were measured at 5-min intervals. Left ventricular developed pressure (LVdevP) was derived by subtracting left ventricular end diastolic pressure from left ventricular peak systolic pressure. The hearts were exposed to various interventions after equilibration. Choline (n = 4) or ethanolamine cell wall (n = 4) was administered through a side arm just above the heart. Controls (n = 4) received saline alone. The rate of infusion was kept at 1.25% of coronary flow using a programmable WPI 200 infusion pump (World Precision Instruments) and adjusted to keep the final pneumococcal cell wall concentration of 106 CFU equivalents per milliliter constant. For some experiments, rats (n = 4) were treated continuously with the PAFr antagonist BN52021 (10 µM; BIOMOL) beginning 30 min before cell wall infusion. The amplitude contractility and LVdevP data were assessed using one-way ANOVA and the Tukey honestly significant difference test.

To follow the cell wall into cardiac muscle in vivo, 3- to 4-wk-old Sprague Dawley rats (Harlan; n = 3) were injected i.v. with 2 x 108 FITC cell wall and sacrificed 3 h later. Hearts were excised and placed in 10% formalin (Globe Scientific) For PAF antagonist studies, rats received 0.1 mg of CV6209 (Biomol) i.v. at 24 and 1 h before administering the cell wall. For all animals, heart tissue was deparaffinized using EZ-DeWax (BioGenex) and coverslipped using Vectashield mounting medium with DAPI (Vector Laboratories). Slides were then viewed at x40 magnification on a Zeiss 510 NLO confocal multiphoton microscope. The FITC label was viewed in the confocal mode using an argon laser, and the DAPI label was observed in the multiphoton mode with the infrared laser set at the appropriate wavelength and filter combination for viewing DAPI. The images were collected as Z series, and maximal projections were derived using the Zeiss software. The amount of cell wall present was determined by manually counting the particles present in each image.

Eukaryotic cell culture and transfection

The rat brain capillary endothelial cell line rBCEC6 (49) or A549 lung epithelial cell line (American Type Culture Collection) was grown to confluence (1.5 x 108 cells in 150-mm tissue culture plates) in humidified 5% CO2 at 37°C. Cells were activated with 10 ng/ml TNF-{alpha} (Research Diagnostics) for 2 h and then challenged with 1 x 107 bacteria per milliliter or cell wall at 5 x 105 bacterial equivalents per milliliter. For some studies, radiolabeled cell wall was used and nuclear and cytoplasmic fractions were analyzed for cell wall uptake (as described below). Both cell lines were demonstrated to express PAFr and beta-arrestin (data not shown). Pertussis toxin (1 µg/ml; Sigma-Aldrich) was added for 16–20 h to define the role of the Gi protein as described by Zhang et al. (50). To determine cell survival, cultures were exposed to 5 x 107 cell equivalents of cell wall for 12, 24, and 48 h and then washed in annexin buffer, stained with FITC-annexin V (Roche) and propidium iodide according to the manufacturer’s instructions, and subjected to FACS analysis with a FACSCalibur flow cytometer and CellQuest software (BD Biosciences).

Primary neurons from embryonic day 15 embryos of wild-type or pafr–/– mice were isolated as described (51). Primary cerebral microvascular endothelial cells were isolated from 6- to 12-wk-old pafr–/– mice or wild-type littermates according to published methods (52) with the following modifications. Briefly, the cortices were homogenized in PBS, pelleted, and resuspended in collagenase/dispase (1 mg/ml; Roche) for 30 min at 37°C with stirring. The digest was centrifuged, washed twice with PBS, resuspended in 30 ml of 15% dextran, and centrifuged at 4500 x g for 30 min at 4°C. The top layer of myelin was discarded and the pellet was again resuspended in collagenase/dispase and digested for 30 min at 37°C with stirring. The digest was pelleted, washed with PBS, resuspended in EGM endothelial cell growth medium (EGM; Cambrex), and cells were plated on collagen-coated plates. FITC and radiolabeled cell wall challenges were performed as for RBCEC6 cells.

COS-1 cells (American Type Culture Collection) were grown to monolayers of 3 x 106 cells and transfected using the FuGENE kit (Roche) with 0.4 µg of expression vector DNA for PAFr-GFP, mutant PAFr D289A-GFP, and 0.8 µg of DNA for beta-arrestin-1 and mutant beta-arrestin V53D (31, 34). Successful transfection with PAFr-GFP plasmid was monitored by fluorescence microscopy.

Imaging of intracellular cell wall

All cell lines were seeded on glass chamber slides (Nalge Nunc International), and at 30% confluence the cells were exposed to a 0.5–1 x 108 CFU/ml equivalent of FITC-labeled cell wall for 2 h. Cells were washed with PBS, fixed with 4% paraformaldehyde (Electron Microscopy Sciences) in PBS at RT for 20 min, and then rinsed and permeabilized with 1% Triton X-100 (Sigma-Aldrich) at RT. For some experiments, internalized FITC-labeled cell wall was detected after staining the eukaryotic cells for 10 min with 50 µg/ml propidium iodide (Sigma-Aldrich).

For the colocalization of cell wall with PAFr or the nuclear envelope marker lamin, samples were blocked with 1% BSA (Sigma-Aldrich) in PBS at RT followed by goat Ab to mouse PAFr (1/100 or 1/250; Santa Cruz Biotechnology) or human lamin (1/100; Santa Cruz Biotechnology) overnight at 4°C. Cells were washed with blocking buffer and incubated with the secondary Ab Alexa Fluor 594 rabbit anti-goat IgG (1/100 or 1/200; Molecular Probes) at RT. Cells were washed twice with PBS and mounted in a medium containing the DNA-specific dye TO-PRO-3 (Molecular Probes) to stain the nucleus blue. Primary endothelial cells were stained with monoclonal anti-CD31 (1/50; Sigma-Aldrich) and the secondary Ab, Alexa Fluor 594 goat anti-mouse IgG (1/200; Molecular Probes). The samples were examined in a Leica TCS NT/SP confocal laser scanning microscope equipped with argon (488 nm), krypton (568 nm), and helium-neon (633 nm) lasers. The three lasers permitted the imaging of FITC (green; emission at 518 nm), Texas Red (red; emission at 570 nm), and TO-PRO-3 (far red; emission at 661 nm, pseudocolored blue) fluorochromes respectively. The samples were examined with x100 Plan Apochromat, numerical aperture 1.4, oil immersion objective. Scanning was performed in X, Y, and Z planes at a laser power of 100% on all lasers. When collecting the Z series, the step size was set at 0.5 µm. The three channel images and an overlay image were recorded using the Leica PowerScan software and rescaled and {gamma} corrected with the Adobe Photoshop. Three-dimensional reconstruction was done using the Leica Power Scan software. Orthogonal sections of the image stack were obtained using the sectioning feature of the Leica LCS software. The sectional views in the XZ and YZ planes enabled unambiguous localization of a given object in a specific cellular compartment.

Effect of cell wall on myocyte contractility in vitro

HL-1 atrial myocytes (a gift from Dr. W. Claycomb, Louisiana State University, New Orleans, LA) were maintained in Claycomb’s medium (JRH Biosciences) supplemented with 10% FBS (JRH Biosciences), 2 mM L-glutamine (Invitrogen Life Technologies), and 0.1 mM norepinephrine (Sigma-Aldrich). Contractility studies were performed with a video edge motion detector (Crescent Electronics) interfaced to a standard closed circuit television video camera (Javelin Electronics) attached to a Nikon Diaphot microscope (x40 magnification; Nikon Ph3 DL objective). Output was interfaced to a Windograf 980 chart recorder (Gould) and images were recorded using a video signal recorder (Sony). HL-1 cells are spontaneously contractile when confluent and did not require pacing. Edge detection of beating was recorded at 60 Hz on myocyte monolayers at RT in Claycomb’s medium supplemented with 10% FBS (JRH Biosciences), 2 mM L-glutamine (Invitrogen Life Technologies), and 0.1 mM norepinephrine (Sigma-Aldrich). Cells were allowed to contract for 5 min to obtain a baseline amplitude and then challenged with choline (n = 4), ethanolamine cell wall (n = 4) (5 x 107 CFU equivalents per milliliter in 1 ml of supplemented Claycomb’s medium), or medium alone (n = 3). Contractility was recorded for 1 h. Cells were exposed to 10 µM BN52021 PAFr antagonist or PLC inhibitor U73122 (10 µM; BIOMOL) for 1 h and then to cell wall, and contractility was recorded for 1 h (n = 4).

Preparation of cellular extracts

Nuclear and cytosolic fractions of eukaryotic cells were prepared as described previously (53). Briefly, whole cells were lysed by mechanical disruption through a 25-gauge needle in hypotonic buffer and centrifuged to pellet the crude nuclei and membrane fraction; the supernatant was harvested as the cytosolic fraction. The pellet was resuspended in high salt buffer, disrupted mechanically again, and centrifuged at 25,000 x g to pellet nuclear debris; the supernatant was harvested as the nuclear extract. Extracts were measured for radioactivity in their entirety. Protein concentrations were determined using the Bio-Rad protein concentration reagent as described by the manufacturer. Absence of cross-contamination of cytoplasmic and nuclear fractions was confirmed (data not shown). For some experiments, whole or fractionated endothelial cell extracts were subjected to SDS PAGE and analyzed by Western blotting to detect phosphorylated ERK-1/ERK-2 kinases using the PhosphoPlus p44/42 MAPK (Thr202/Tyr204) (54).

Cytokine ELISA

Primary neurons, rBCEC6 cells, A549 epithelial cells, or HL-1 cardiomyocytes were seeded in duplicate in 24-well plates at 250,000 cells/well in 1 ml of medium. The next day the medium was changed and 100 µl of supernatant was recovered at 0, 2, 4, 6, and 24 h after stimulation with 5 x 106 or 1 x 107/ml bacterial equivalents of cell wall or 100 ng/ml LPS from Escherichia coli serotype O111:B4 (Sigma-Aldrich) or 10 µg/ml Pam3Cys. Supernatant was stored at –20°C until assayed. To determine cytokines in supernatants, microtiter plates (Maxisorb; Nalge Nunc International) were coated with the respective anti-mouse capture Abs (50 µl/well) against IL-1beta (4 µg/ml) and TNF-{alpha} (3 µg/ml; e-Bioscience) in coating buffer (0.1 M Na2CO3H2O and 0.1 M NaHCO3) overnight at 4°C. The next day, plates were washed with ELISA washing buffer (1% Tween 20, 1 mM Tris-base, and 154 mM NaCl), blocked with 300 µl/well ELISA buffer (10% FCS in PBS), and incubated for 2 h at RT. After washing, 50 µl per well of a 1/2 dilution in ELISA-buffer of cell supernatant and duplicate serial 2-fold dilutions of the respective standards (50 ng/ml starting concentration) were added and incubated overnight at 4°C. Plates were then washed three times, and biotin-conjugated detection Abs for IL-1 (300 ng/ml; BD Pharmingen) and TNF-{alpha} (1.6 µg/ml, eBioscience) were diluted in ELISA buffer (50 µl/well) and added for 45 min at RT. After washing the plates three times, streptavidin peroxidase (50 µl/well; Sigma-Aldrich) diluted 1/1000 in ELISA buffer was added for 30 min at RT. Plates were soaked in washing buffer for 1 min and washed five times, the tetramethylbenzidine peroxidase substrate system (Kirkegaard & Perry Laboratories) was applied at 50 µl/well, reactions were stopped with 10% phosphoric acid (50 µl/well), and absorbance was measured at 405 nm in an ELISA reader (Molecular Dynamics). Cytokine concentrations were calculated according to the respective standard curves.

Microarray analysis

The response of rBCEC6 cells and primary rat neurons to cell wall was interrogated using rat whole genome-spanning 60-mer oligonucleotide microarrays representing >20,000 characterized rat transcripts (Agilent Technologies). Triplicate RNA samples were isolated from three independent biological experiments exposing rBCEC6 cells and neurons to TNF for 2 h and then treatment with or without 1 x 107 CFU/ml choline cell wall for 135 min or 5 x 106 CFU/ml Lyt4-4 heat-killed bacteria (1 h at 70°C). RNA was isolated from the monolayer using Qiagen RNeasy Midi kit with on-column DNase digestion following the manufacturer’s protocol. RNA samples were labeled in triplicate with Cy-3 (untreated) and Cy-5 (treated) monoreactive dyes (Amersham Biosciences) using reverse transcription-based indirect labeling (<http://www.hartwellcenter.org/bio_services/fungen/cDNA.php#PROTOCOLS>). Labeled products were hybridized overnight, washed, and scanned using a GenePix 4000B scanner (Axon Instruments). The text data files were prefiltered to remove unreliable data and normalized using Spotfire DecisionSite for Functional Genomics version 7.2 (Spotfire) to remove dye bias. Fold change of expression levels and significance of differential gene expression were calculated using Spotfire. Genes were considered to be differentially expressed based on both the fold change and the p value of significance as calculated by the pairwise t test. Raw data are available at the Gene Expression Omnibus repository (GSE5545). The Ingenuity Pathways (Ingenuity Systems) analysis application was used to visualize and explore the results of microarray analysis.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
Cell wall accumulates on the vascular endothelium and kills mice

Pneumococci use PAFr to cross from one body compartment to another, i.e., from lung to blood for the progression of pneumonia to bacteremia or from blood to brain for the development of meningitis. Indeed, PAFr-deficient mice show improved survival in a pneumococcal sepsis model (50 vs 0% survival of Pafr–/– vs wild type at day 8). To measure the contribution of the PAFr to host cell binding and the uptake of cell wall, we chose a whole animal imaging model to track FITC-labeled cell wall. In this model, the blood-borne cell wall represents material that would be released from lysing pneumococci as a result of antibiotic therapy. Protein-free, teichoic acid-bearing cell wall was inoculated into the blood stream of wild-type mice. We observed that wild-type mice rapidly became moribund, with most mice (17 of 19) dying within 2 h of injection. The distribution of FITC-labeled cell wall in the vascular compartment was determined by continuously imaging cerebral vessels through a cranial window. In the first 3 min after injection, rapid streaming of FITC-labeled cell wall was seen along with blood inside the vessels. By 1 h, sustained binding of cell wall to the microvasculature had occurred (Fig. 1A). In contrast to wild-type mice, Pafr–/– mice had virtually no cell wall attached to microvessels up to 5 h after FITC-labeled cell wall injection, and most of the mice survived (Fig. 1A). Mice deficient in two other cell wall recognition proteins, TLR2 and Nod2, showed abundant cell wall margination and low survival rates similar to those of control mice (Fig. 1B). These data suggested that in this model PAFr, but not TLR2 or Nod2, played a predominant role in the trafficking of cell wall and the associated lethality.


Figure 1
View larger version (75K):
[in this window]
[in a new window]
  		FIGURE 1. Imaging of intravascular cell wall trafficking via cranial windows in wild-type and receptor-deficient mice. A and B, FITC-labeled cell wall (CW) equivalent to 5 x 107 CFU was injected i.v. into 8- to 10-wk-old C57BL/6 (wild-type) or Pafr–/–, Tlr2–/–, and Nod2–/– mice or their littermate controls. At 1 h after injection, images of the microvasculature were acquired through a cranial window with a fluorescence microscope (excitation at 490 nm and emission at 520 nm) and a magnification of x40. Fragments outside the vasculature were in areas devoid of blood flow (full movie is available at <http://www.stjuderesearch.org/data/sepsis>; user name is sepsis, and password is _sepsis_). An image taken 1 h post-PBS injection is shown as a baseline control. Quantification identified an average of 208 ± 84 cell wall particles bound per field of vision in wild-type mice (n = 8) compared with 26 ± 12 particles for Pafr–/– mice (n = 5; p = 0.0009). C, For some experiments, peptidoglycan (PG) or ethanolamine (EA) cell wall was substituted for choline cell wall as indicated. Survival at 5 h is shown as a percentage (number of animals).

 
We next conducted experiments to document the specific requirement for phosphorylcholine for the binding of bacterial elements to PAFr. Cell wall bearing phosphorylethanolamine in place of phosphorylcholine failed to bind to cerebral vessels, and all of the animals survived i.v. challenge (Fig. 1C). Similarly, peptidoglycan free of choline-bearing teichoic acid failed to bind cerebral vessels or induce mortality (Fig. 1C).

Cell wall exits the vasculature and enters nonphagocytic eukaryotic cells

Examination of in vivo cranial window images revealed that cell wall was not only attached to the endothelium of the vessel wall but was also visible outside the vasculature within the brain parenchyma (Fig. 1A). Translocation across cell barriers also occurred when cell wall was instilled directly into the cerebrospinal fluid from where it accumulated within endothelial cells of the choroid plexus (Fig. 2A) and within the brain parenchyma contiguous to the ventricles (Fig. 2B). Localization of extravascular cell wall in brain sections from animals studied by cranial window showed that cell wall was present within neurons (Fig. 2C) and cardiomyocytes (Fig. 2D). Thus, cell wall was distributed beyond the vascular space and entered nonphagocytic parenchymal cells of the brain and heart.


Figure 2
View larger version (70K):
[in this window]
[in a new window]
  		FIGURE 2. Distribution of cell wall in organs. A, Upon injection into the cerebrospinal fluid of a mouse, cell wall (2 x 107 bacterial equivalents) localized within the cells of the choroid plexus as detected at 3 h after injection by fluorescent-labeled anti-choline TEPC 15 Ab (original magnification x20; inset shows negative control stained after PBS injection). B, FITC-labeled cell wall injected as in A was taken up into the ependymal epithelium and the underlying brain parenchyma. Central black area is the cerebrospinal fluid ventricle. C and D, Injection of mice i.v. with 5 x 107 cell wall equivalents. FITC-labeled cell wall localized within neurons (shown in C at 5 h after injection; red is the NeuN neuronal marker) and cardiomyocytes (shown in D at 1 h after injection). Original magnification is x100. E and F, Cell wall (green) localized to the nucleus (blue) by confocal multiphoton imaging of neuron (E) and cardiomyocyte (F); images taken at 5 h after injection. G, Change in uptake of wall into nuclei of neurons in receptor deficient animals; image was taken at 1 h after injection (movies showing localization of cell wall in the nucleus can be found at <http://www.stjuderesearch.org/data/cellwall/>; user name is cellwall, and password is _cellwall_). For all panels, distribution of cell wall did not differ between 1, 3, or 5 h after injection, and images shown are representative.

 
Analysis of in vivo tissue sections indicated that cell wall within neurons and cardiomyocytes localized within the cytoplasm and the nucleus (Fig. 2, E and F). In contrast to control mice, virtually no cell wall was found in the cytoplasm or nucleus of neurons of Pafr–/– animals (Fig. 2G) but was present in neurons of Tlr2–/– and Nod2–/– mice (Fig. 2G). These data suggested that cell wall could enter the cytoplasm and even the nucleus of nonphagocytic endothelial cells, neurons, and cardiomyocytes in a PAFr-dependent manner. We next initiated a series of experiments to determine the consequences to the host cell of indwelling Gram-positive cell walls.

Uptake of cell wall into host cells in vitro

More malleable in vitro culture systems of rBCEC6 brain endothelial cells, primary neurons, and A549 lung epithelial cells were used to analyze the kinetics of cell wall uptake and distribution. Living or heat-killed bacteria bearing radiolabeled cell wall or purified radiolabeled cell wall were added to cell monolayers, and the 3H signal was measured in the cytosolic and nuclear fractions at 45-min intervals. Cytoplasmic and nuclear localization of the cell wall label was detected for living (Fig. 3A) and heat-killed bacteria (Fig. 3B). Purified cell wall showed the greatest intracellular accumulation, peaking at 135 min (Fig. 3C). Unlabeled wall strongly competed with the uptake of labeled wall in a dose-dependent way, supporting specificity of the uptake process (data not shown). Approximately 1% of the total radiolabel added was internalized into the cells, with 0.7% in the nucleus and 0.4% in the cytoplasm. Neurons accumulated 10-fold less total radiolabel than endothelial cells, whereas virtually no radiolabel was detected in A549 epithelial cells (data not shown). These data indicated that the uptake of cell wall varied by cell type and involved both cytoplasmic and nuclear compartments. The PAFr was essential for uptake, because Pafr–/–neurons showed a 60 ± 2% decrease in the accumulation of radiolabeled cell wall in the cytoplasm as compared with wild-type neurons and a 70 ± 8% decrease for the nucleus (n = 4).


Figure 3
View larger version (34K):
[in this window]
[in a new window]
  		FIGURE 3. Quantification of cell wall uptake into cytoplasm and nucleus. S. pneumoniae bacteria or purified cell wall labeled with [methyl-3H]choline (0.00125 cpm per bacterium or 0.168 cpm per cell wall equivalent) were incubated with monolayers of 1.5 x 108 rBCEC6 cells for the times indicated. Cytosolic fractions ({circ}) and nuclear extracts ({blacksquare}) were prepared and radioactivity was counted and expressed as counts per minute per 600-cm2 monolayer. A, Living bacteria R6 (1 x 107 CFU/ml). B, Heat-killed bacteria (1 x 107 CFU/ml). C, Purified cell wall (5 x 105 bacterial equivalents/ml). Mean ± SD of 3–5 experiments each. D, Over 135 min, FITC-labeled cell wall was distributed sparsely in the cytoplasm of cultured rBCEC6 cells and accumulated in the nucleus. Two nuclei are shown with membrane marked by anti-lamin Ab (red). FITC solution alone or FITC-zymosan failed to show intracellular fluorescence (data not shown). E, To further confirm the intranuclear location of the fluorescent cell wall foci, the nucleus was stained with TO-PRO-3 (pseudo-colored blue) and orthogonal sectioning on the foci was performed within a 2-µm area in XZ and YZ planes. FITC-labeled cell wall component is located in the nucleus (in crosshairs) as evidenced by the cyan color (green foci colocalized with blue nuclear dye). The horizontal (XZ) section of the image stack of the nucleus is shown at the bottom and the vertical section (YZ) on the right side of the image.

 
Consistent with in vivo images and in vitro subcellular fractionation, confocal microscopy of rBCEC6 endothelial cells exposed to cell wall in vitro demonstrated cell wall in the cytoplasm and also within the nuclear envelope (Fig. 3D). In >50 rBCEC6 cells examined by orthogonal sectioning (Fig. 3E), 1–5 fluorescent foci localized to the nucleus with occasional cytoplasmic foci found adjacent to the nucleus (blue), and these cyan foci were at or near the nuclear periphery. Similar localization in the nucleus and cytoplasm was observed using primary neurons (data not shown).

Functional PAFr can be detected on the eukaryotic cell surface and in perinuclear and intranuclear regions (55, 56, 57). All internalized cell wall fragments colocalized with PAFr in rBCEC6 cells (Fig. 4, A–C), cardiomyocytes (Fig. 4D), and primary neurons (data not shown). To further substantiate the role of the PAFr in cell wall translocation, endothelial cells isolated from wild-type and Pafr–/– mice were challenged with cell wall in vitro. FITC-labeled cell wall was readily detected in wild-type endothelial cells (Fig. 4E) but only rarely in Pafr–/– cells (Fig. 4F). beta-Arrestin-1 is required for nuclear localization of GPCR (29, 57) and, consistent with nuclear colocalization of PAFr and cell wall, functional beta-arrestin-1 was essential for cell wall uptake in a COS cell transfection model (Fig. 4G). PAFr can activate multiple signaling cascades with and without G protein signaling but, characteristically, beta-arrestin-mediated signaling through PAFr leads to the activation of ERK1/2 and p38 MAP kinases without G protein participation (27, 29, 30, 32, 33, 57, 58). Consistent with this finding, rBCEC6 cells, primary neurons, and HL-1 cardiomyocytes challenged with cell wall in the presence of the G protein inhibitor pertussis toxin did not show a decrease in the amount of uptake of cell wall (data not shown) or a change in phosphorylation of ERK1/2 (Fig. 5A). In contrast, although virtually no uptake of cell wall was detected, lung epithelial cells showed strongly increased ERK phosphorylation at 60 and 120 min after cell wall challenge, virtually all of which was pertussis toxin sensitive. Thus, indwelling cell wall caused no G protein-mediated ERK phosphorylation in three cell types, whereas the cell showing no uptake demonstrated strong phosphorylation. These results point to the PAFr-beta-arrestin-1 pathway as being responsible for cell wall uptake into the cytoplasm and nucleus of host cells.


Figure 4
View larger version (55K):
[in this window]
[in a new window]
  		FIGURE 4. Interaction of cell wall with PAFr. A–C, Colocalization of FITC-labeled choline cell wall (green in A) and Alexa Fluor 594-labeled PAF receptor (red in B) in the nucleus (blue in C) as determined by the merger of all three colors in the overlay image (green plus red plus blue equals white in C). Controls in which cells were treated with the secondary Ab alone failed to label the nuclear foci. Orthogonal sections show that the cell wall fragment and the PAFr are within the nuclear envelope (crosshairs) (representative of >50 cells in two experiments). D, FITC-labeled cell wall (green) was visualized in cardiomyocytes colocalized with PAFr (red) as indicated by merged colors (yellow). E, FITC-labeled cell wall was visualized in the cytoplasm and nucleus (blue) of wild-type primary murine brain-derived microvascular endothelial cells labeled with CD31 marker (red). F, Only the rare FITC-labeled cell wall was observed in brain-derived microvascular endothelial cells from Pafr–/– mice. G, COS-1 cells (3–4 x 106) were transfected with PAFr (P) and beta-arrestin-1 (A) or with combinations of wild-type and nonfunctional mutant receptors as indicated (mP, mutant PAFr; mA, mutant arrestin). Nontransfected cells (NT) were used as a negative control (100% = 177 ± 85 cpm per 600-cm2 monolayer). Cells were challenged with radiolabeled cell wall (5 x 105 bacterial equivalents/ml) and uptake was quantitated at 135 min (mean ± SD of three experiments; *, p ≤ 0.019 from t test).

 

Figure 5
View larger version (37K):
[in this window]
[in a new window]
  		FIGURE 5. A, Time course of phosphorylation of ERK-1/2 MAPKs in rBCEC6 endothelial cells, neurons, A549 epithelial cells, and HL-1 cardiomyocytes exposed to 1 x 107 bacterial equivalents per milliliter of cell wall (representative of 3–4 experiments for each cell type). B, The mean (± SEM) percentage change from baseline in amplitude contractions was calculated at 15 min intervals for HL-1 mouse atrial myocyte cell cultures exposed to cell wall (•), cell wall plus the BN52021 PAFr antagonist ({triangleup}), ethanolamine cell wall ({blacktriangledown}), or no treatment ({circ}). *, Significant difference from control at p < 0.05. Experiment was performed four times.

 
Effect of cell wall/PAFr interaction on pathophysiology in vitro and in vivo

Extracellular cell wall has been described as inducing a variety of cellular responses, including induction of apoptosis (59, 60, 61). However, apoptosis was not induced by the uptake of cell wall into neurons, A549 epithelial cells, cardiomyocytes, and endothelial cells, because annexin V staining of both control untreated cells and cell wall-challenged cells remained below 10% of each cell population over 48 h (p not significantly different; data not shown). In phagocytic cells, extracellular cell wall interactions with TLR2 activate NF-{kappa}B for cytokine production (62). Similarly, Nod proteins are proposed to form an intracellular equivalent of the TLR system, recognizing peptidoglycan fragments and leading to the activation of NF-{kappa}B (44, 63, 64). According to this hypothesis, we should expect to observe the induction of a range of NF-{kappa}B target genes in cells that have an indwelling cell wall. We would also anticipate that animals with high levels of cell wall taken up by the PAFr-dependent process should develop an inflammatory response. To test these hypotheses, we treated rat endothelial cells and neurons with cell wall over time and performed microarray analyses using rat whole genome-spanning 60-mer oligonucleotide arrays. Surprisingly, even though the cultured cells contained cell wall throughout the cytoplasm and nucleus, no widespread changes in gene expression were observed and, specifically, no significant changes were detected in classic NF-{kappa}B target genes including Tnf-{alpha}, Il-1beta, and I-{kappa}B (raw data are available at the Gene Expression Omnibus site as cited in Materials and Methods). These data suggested that an indwelling cell wall was not a strong stimulant of a typical inflammatory response and that it appeared to have a "neutral" effect on gene expression. These data were further confirmed by measuring IL-1beta or TNF-{alpha} in the supernatants of primary neurons, rBCEC6 cells, A549 cells, or HL-1 cardiomyocytes at 0, 2, 4, 6, or 24 h after cell wall stimulation. In these experiments, no secretion of these cytokines was observed (data not shown).

Whereas neurons and endothelial cells remained healthy and quiescent after cell wall uptake, cardiomyocytes observed to contain indwelling cell wall in vivo showed a dramatic response. HL-1 atrial myocytes challenged with cell wall remained viable but exhibited a rapid decrease in contractility as measured by edge detection. The amplitude of contraction of cell wall-challenged cells was decreased as compared with control cells and with cells exposed to ethanolamine cell wall at all of the assessed time points (Fig. 5B). In two of the five experiments the effect was so strong that no amplitude of contraction could be detected at 60 min, a phenomenon not observed in any control experiment. Contractility was preserved in cardiomyocytes pretreated with the PAFr antagonist (Fig. 5B) or phospholipase C (PLC) inhibitor (beat amplitude of cells challenged with cell wall plus PLC inhibitor remained at ≥98% of PLC inhibitor alone for 60 min; p < 0.05 by Tukey comparison of means). These in vitro results indicated that the cell wall/PAFr interaction in heart was associated with significant harmful effects on cellular physiology as compared with neurons or endothelial cells. The significance of these derangements was further investigated in vivo.

Initiation of antibiotic therapy in the context of sepsis rapidly releases cell wall fragments in the bloodstream, an effect that is well known to correlate with transient impairment of cardiac function (65, 66, 67). Such detrimental effects of cell wall uptake on cardiac physiology were demonstrable in the mouse model both in terms of survival and cardiac dysfunction. Treatment of wild-type mice with PAFr antagonist CV-6209 for 16 h before cell wall challenge prevented cell wall attachment to vessels and death (Fig. 6A). Treatment of mice with the antagonist immediately after cell wall administration failed to displace cell wall bound to microvessels (data not shown) or prevent lethality (three of three mice died within 1 h). The protective benefit of interference with PAFr extended to competitive inhibition of cell wall binding by CDP-choline. Pretreatment with CDP-choline greatly diminished cell wall binding to the endothelium and increased survival from 10 to 50% after cell wall challenge (Fig. 6B). Rats treated with PAFr antagonist accumulated significantly less cell wall in cardiomyocytes (2 ± 2.3 particles per x40 field vs untreated control 6.3 ± 0.4 particles per x40 field). Similarly, rats infused with an ethanolamine cell wall that lacks the choline determinant for binding PAFr demonstrated significantly less uptake of cell wall into the heart (0.9 ± 0.2 vs control 6.3 ± 0.4 particles per x40 field).


Figure 6
View larger version (66K):
[in this window]
[in a new window]
  		FIGURE 6. Amelioration of cell wall-induced pathophysiology by inhibition of PAFr. A, Distribution of i.v. injected FITC-labeled cell wall (5 x 107 bacterial equivalents) was viewed at 1 h after injection through a cranial window in wild-type mice treated with saline and PAFr antagonist CV6209 at 16 and 1 h before cell wall challenge or CDP-choline at 16 and 1 h before cell wall challenge. Survival is indicated as a percentage of the total number of mice per treatment group. B, The mean (± SEM) percentage of baseline LVdevP for isolated rat heart preparations exposed to cell wall ({circ}), ethanolamine cell wall ({blacktriangledown}), pretreatment of rats with BN52021 PAFr antagonist before cell wall challenge ({triangledown}), or no treatment (•). *, Significant difference between challenge with cell wall vs untreated control (p < 0.05); **, significant difference between hearts challenged with cell wall ± PAFr antagonist (p < 0.05); +, significant difference between PAFr antagonist treated and untreated control hearts. Experiment was performed four times.

 
The Langendorff isolated heart model was used to link the pathophysiology seen in rats more directly to the effects of cell wall on cardiac function. The continuous infusion of cell wall was associated with a decrease in LVdevP in all experiments (Fig. 6B). The effect was significant at 20 min and remained so throughout the remainder of the experiment. There was no difference in the LVdevP in isolated heart preparations receiving ethanolamine cell wall or the controls (Fig. 6B). In the isolated heart preparations treated with PAFr antagonist, LVdevP remained at control levels until 55 min with a modest decrease thereafter (Fig. 6B). The efficacy of the PAFr antagonist and the requirement for choline on the cell wall for bioactivity further indicated that cell wall/PAFr interactions depressed contractility of the heart.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
We found that PAFr is a physiologically important innate immune receptor that recognizes pneumococcal cell wall by virtue of the PAMP associated with phosphorylcholine-containing bacterial components. Phosphorylcholine has been detected on a variety of respiratory pathogens on cell surface components as diverse as LPS, proteins, or cell wall and is known to direct these bacteria to the PAFr (21, 23, 24, 25). The amount of phosphorylcholine on a bacterial surface is regulated in a phase-variable fashion, indicating that bacteria devote energy to modulating this interaction in the context of the infected host (68). A variety of bacterial elements bearing the PAMP phosphorylcholine become accessible to the PAFr when released from a pathogen, for instance by cell surface turnover or antibiotic-induced lysis. We determined that PAFr supports binding of circulating cell wall to endothelia, entry of cell wall into organs, and, within organs, specific uptake into nonphagocytic cells such as neurons and cardiomyocytes. This egress of cell wall out of the vascular space indicates that phosphorylcholine-containing material distributes itself widely and comes to reside within organs, an event that correlates with the death of the animal. Survival is promoted by the absence of PAFr in mice, the removal of phosphorylcholine from cell wall, or the treatment of mice with CDP-choline or PAFr antagonists that block cell wall margination on the vascular endothelium.

A striking feature of cell wall uptake was the progression of distribution into the nucleus. PAFr is one of several GPCRs that demonstrates both a cell surface distribution and a perinuclear distribution independent of the binding of ligand (55, 56, 69). The scaffold protein beta-arrestin associates with GPCRs to enable their nuclear distribution (57). Consistent with these findings, we demonstrated that the beta-arrestin/PAFr association is required for the localization of cell wall within the nucleus of eukaryotic cells. Access of cell wall to nuclear material presents the interesting possibility of novel mechanisms for modulation of gene expression.

The binding of PAFr by macromolecular cell wall adds a new dimension to the process of cell wall recognition by nonimmune cells such as endothelial cells and neurons (Fig. 7). Bacterial peptidoglycan degradation products are proposed to interact with Nod1 and Nod2 in the cytoplasm of cells leading to cellular activation via NF-{kappa}B (70, 71). In models of Helicobacter infection, peptidoglycan components have been demonstrated to be injected into host cells by a bacterial type IV secretion system (70, 71). Our results show that Gram-positive bacterial cell wall enters not only the cytoplasm but also the nucleus of nonphagocytic endothelial cells, cardiomyocytes, and neurons. However, this uptake pathway differs significantly from that for the peptidoglycan fragments from Gram-negative bacteria in several ways. First, rather than their own injection systems, the bacteria use host cell machinery, i.e., PAFr and beta arrestin. Second, large, teichoicated cell wall components 10–30 nm in size enter cells. Third, entry extends beyond the cytoplasm to the nucleus. Fourth, uptake is cell specific, being prominent in endothelial cells, cardiomyocytes, and neurons, but not in epithelial cells. Because the Gram-positive cell wall used here is a fragmented teichoic acid-bearing peptidoglycan, including the muramyl dipeptide subcomponent recognized by Nod2, we expected to observe a robust inflammatory response in all cell types carrying an indwelling cell wall. Such data would be in keeping with the concept that Nod proteins recognize cell wall within the host cell and activate an inflammatory response (63, 64). In contrast, we observed that most types of host cells were refractory to significant changes in gene expression, including an absence of induction of a classic-type NF-{kappa}B response. The lack of induction of an inflammatory response to indwelling cell wall is consistent with inhibition of NF-{kappa}B activation in response to some GPCRs by beta-arrestin (72). These findings argue that although cell wall uptake robustly activates macrophages and dendritic cells, it is neutral in most nonimmune cell types. Recent findings from microarray and proteomic experiments using invasive Shigella in epithelial cells (73) and muramyl dipeptide-stimulated 293T cells overexpressing Nod2 (74) support the notion that Nod2 does not necessarily activate groups of well-recognized NF-{kappa}B target genes in nonphagocytic cells.


Figure 7
View larger version (24K):
[in this window]
[in a new window]
  		FIGURE 7. Schematic comparison of the proposed PAFr-cell wall innate recognition system to known cell wall receptors and the host responses they trigger. Cho, phosphorylcholine; M, N-acetylmuramic acid; G, N-acetylglucosamine.

 
The apparent innocuous nature of cell wall fragments residing in endothelial cells and neurons left open the question of why systemic cell wall administration leads to the death of mice. In this regard, we established that cell wall causes PAFr-dependent cardiotoxicity and that this effect is cell specific. It is well known that different cell types exhibit unique physiological responses to ligation of PAFr based on the activation of distinct signaling pathways (32, 33, 58). Consistent with this fact, cell wall uptake invoked G protein-independent signaling in neurons, cardiomyocytes, and endothelial cells, whereas lung epithelial cells demonstrated G protein-dependent signaling and lack of cell wall uptake. The heart is known to regulate expression of the Pafr transcript in a manner distinct from that other cell types and, in agreement with our findings concerning the cell wall, PAF induces G protein-independent signaling in cardiomyocytes (75) that results in the activation of PLC{gamma} (33). The ability of a PLC inhibitor to preserve cardiomyocyte contractility suggests this approach could serve as a targeted therapy to prevent cardiac dysfunction in the context of treatment of sepsis involving circulating phosphorylcholine-containing bacterial surface components.

PAFr appears to represent an innate PAMP recognition system for phosphorylcholine that is the first one known to lead to uptake of Gram-positive cell wall into host cells. It has been generally believed that cell wall recognition is proinflammatory, but the lack of response of endothelial cells and neurons to cell wall/PAFr interactions introduces the concept that, in some circumstances, recognition of cell walls is quiescent. This may enhance immune evasion during infection and promote latency or chronicity. Despite the tolerance of cell wall in some organs, however, it appears that the heart is particularly susceptible to cell wall-induced pathophysiology and that cardiac dysfunction is a strong contributor to acute mortality in the context of circulating bacterial debris.


   Acknowledgments
 
We thank G. Gao, J. Sublett, S. H. Smith-Sielicki, D. Johnson, and P. Austin for excellent technical assistance and Dr. C. Obert, T. Chin, and S. Bhuvanendran for methodological advice and equipment. Dr. Robert F. Tamburro provided excellent statistical advice for the cardiac studies.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1 S.F., K.S., and S.R. contributed equally to this work. Back

2 Address correspondence and reprint requests to Dr. Elaine I. Tuomanen, Department of Infectious Diseases, St. Jude Children’s Research Hospital, 332 North Lauderdale Street, Memphis, TN 38105. E-mail address: elaine.tuomanen{at}stjude.org Back

3 Abbreviations used in this paper: PAMP, pathogen-associated molecular patterns; CDP, cytidine diphosphate; DAPI, 4',6-diamidino-3-phenylindole; GPCR, G protein-coupled receptor; LVdevP, left ventricular developed pressure; PAF, platelet-activating factor; PAFr, PAF receptor; PLC, phospholipase C; RT, room temperature; TMB, tetramethylbenzidine. Back

Received for publication April 26, 2006. Accepted for publication July 25, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Sprung, C., P. N. Peduzzi, C. Shatney, R. Schein, M. Wilson, J. Sheagren, L. Hinshaw. 1990. Impact of encephalopathy on mortality in the sepsis syndrome. Crit. Care Med. 18: 801-806. [Medline]
  2. Weber, J., P. Moreillon, E. Tuomanen. 2003. Innate sensors for Gram positive bacteria. Curr. Opin. Immunol. 15: 408-415. [Medline]
  3. Dziarski, R.. 2004. Peptidoglycan recognition proteins. Mol. Immunol. 40: 877-886. [Medline]
  4. Weber, J., D. Freyer, D. Pfeil, C. Kuster, N. Schroeder, E. Tuomanen, R. Schumann. 2003. Recognition of pneumococcal peptidoglycan, an expanded role for lipopolysaccharide binding protein. Immunity 19: 269-279. [Medline]
  5. Barton, G., R. Medzhitov. 2002. Toll-like receptors and their ligands. Curr. Top. Microbiol. Immunol. 270: 81-92. [Medline]
  6. Yoshimura, A., E. Lien, R. Ingalls, E. Tuomanen, R. Dziarski, D. Golenbock. 1999. Recognition of Gram positive bacterial cell wall components by the innate immune system occurs via toll-like receptor 2. J. Immunol. 63: 1-5.
  7. Aliprantis, A., R. Yang, M. Mark, S. Suggett, B. Devaux, J. Radolf, G. Klimpel, P. Godowski, A. Zychlinsky. 1999. Cell activation and apoptosis by bacterial lipoproteins through toll-like receptor 2. Science 285: 736-739. [Abstract/Free Full Text]
  8. Girardin, S., P. Sansonetti, D. Philpott. 2002. Intracellular vs extracellular recognition of pathogens — common concepts in mammals and flies. Trends Microbiol. 10: 193-199. [Medline]
  9. Inohara, N., G. Nunez. 2003. NODs: intracellular proteins involved in inflammation and apoptosis. Nat. Rev. Immunol. 3: 371-382. [Medline]
  10. Pauleau, A., P. Murray. 2003. Role of Nod2 in the response of macrophages to toll like receptor agonists. Mol. Cell. Biol. 23: 7531-7539. [Abstract/Free Full Text]
  11. Kobayashi, K., M. Chamaillard, Y. Ogura, O. Henegariu, N. Inohara, G. Nunez, R. Flavell. 2005. Nod2-dependent regulation of innate and adaptive immunity in the intestinal tract. Science 307: 731-734. [Abstract/Free Full Text]
  12. Maeda, S., L. Hsu, H. Liu, L. Bankston, M. Iimura, M. Kagnoff, L. Eckmann, M. Karin. 2005. Nod2 mutation in Crohn’s disease potentiates NF-{kappa}B activity and IL-1beta processing. Science 307: 734-738. [Abstract/Free Full Text]
  13. Magalhaes, J., D. Philpott, M. Nahori, M. Jehanno, J. Fritz, L. Bourhis, J. Viala, J. Hugot, M. Giovannini, J. Bertin, et al 2005. Murine Nod1 but not its human orthologue mediates innate immune detection of tracheal cytotoxin. EMBO Rep. 6: 1201-1207. [Medline]
  14. Mariathasan, S., D. Weiss, K. Newton, J. McBride, K. O’Rourke, M. Roose-Girma, W. Lee, Y. Weinrauch, D. Monack, V. Dixit. 2006. Cryopyrin activates the inflammasome in response to toxins and ATP. Nature 440: 228-232. [Medline]
  15. Schuchat, A., K. Robinson, J. Wenger, L. Harrison, M. Farley, A. Reingold, L. Lefkowitz, B. Perkins. 1997. Bacterial meningitis in the United States. N. Eng. J. Med. 337: 970-976. [Abstract/Free Full Text]
  16. Tuomanen, E., H. Liu, B. Hengstler, O. Zak, A. Tomasz. 1985. The induction of meningeal inflammation by components of the pneumococcal cell wall. J. Infect. Dis. 151: 859-868. [Medline]
  17. Tuomanen, E., R. Rich, O. Zak. 1987. Induction of pulmonary inflammation by components of the pneumococcal cell surface. Amer. Rev. Respir. Dis. 135: 869-874. [Medline]
  18. Burroughs, M., E. Rozdzinski, S. Geelen, E. Tuomanen. 1993. A structure-activity relationship for induction of meningeal inflammation by muramyl peptides. J. Clin. Invest. 92: 297-302. [Medline]
  19. Majcherczyk, P., H. Langen, D. Heumann, M. Fountoulakis, M. Glauser, P. Moreillon. 1999. Digestion of Streptococcus pneumoniae cell wall with its major peptidoglycan hydrolase releases branched stem peptides carrying proinflammatory activity. J. Biol. Chem. 274: 12537-12543. [Abstract/Free Full Text]
  20. Majcherczyk, P., E. Rubli, D. Heumann, M. Glauser, P. Moreillon. 2003. Teichoic acids are not required for S. pneumoniae and S. aureus cell walls to trigger the release of TNF. Infect. Immun. 71: 3707-3713. [Abstract/Free Full Text]
  21. Cundell, D., N. Gerard, C. Gerard, I. Idanpaan-Heikkila, E. Tuomanen. 1995. Streptococcus pneumoniae anchors to activated eukaryotic cells by the receptor for platelet activating factor. Nature 377: 435-438. [Medline]
  22. Ring, A., J. Weiser, E. Tuomanen. 1998. Pneumococcal penetration of the blood brain barrier: molecular analysis of a novel re-entry path. J. Clin. Invest. 102: 347-360. [Medline]
  23. Weiser, J., M. Shchepetov, S. Chong. 1997. Decoration of lipopolysaccharide with phosphorylcholine: a phase-variable characteristic of Haemophilus influenzae. Infect. Immun. 65: 943-950. [Abstract]
  24. Weiser, J. B., J. B. Goldberg, N. Pan, L. Wilson, M. Virji. 1998. The phosphorylcholine epitope undergoes phase variation on a 43-kilodalton protein in Pseudomonas aeruginosa and on pili of Neisseria meningitidis and Neisseria gonorrhoeae. Infect. Immun. 66: 4263-4267. [Abstract/Free Full Text]
  25. Swords, W., M. Ketterer, J. Shao, C. Campbell, J. Weiser, M. Apicella. 2001. Binding of the nontypeable Haemophilus influenzae lipooligosaccharide to the PAF receptor initiates host cell signalling. Cell Microbiol. 3: 525-536. [Medline]
  26. Zimmerman, G., T. McIntyre, S. Prescott, D. Stafforini. 2002. The platelet-activating factor signaling system and its regulators in syndromes of inflammation and thrombosis. Crit. Care Med. 30: S294-S301. [Medline]
  27. Honda, Z., S. Ishii, T. Shimizu. 2002. Platelet activating factor receptor. J. Biochem. 131: 773-779. [Abstract/Free Full Text]
  28. Ihida, K., D. Predescu, R. Czekay, G. Palade. 1999. Platelet activating factor receptor is found in a large endosomal compartment in human umbilical vein endothelial cells. J. Cell Sci. 112: 285-295. [Abstract]
  29. Tohgo, A., E. Choy, D. Gesty-Palmer, K. Pierce, S. Laports, R. Oakley, M. Caron, R. Lefkowitz, L. Luttrell. 2003. The stability of the G-protein coupled receptor-beta-arrestin interaction determines the mechanism and functional consequence of ERK activation. J. Biol. Chem. 278: 6258-6267. [Abstract/Free Full Text]
  30. DeFea, K., J. Zalexsky, M. Thoma, O. Dery, R. Mullins, N. Bunnett. 2000. beta-Arrestin-dependent endocytosis of proteinase-activated receptor 2 is required for intracellular targeting of activated ERK1/2. J. Cell Biol. 148: 1267-1281. [Abstract/Free Full Text]
  31. Chen, A., D. Dupre, C. Le Gouill, M. Rola-Pleszczynski, J. Stankova. 2002. Agonist-induced internalization of platelet-activating factor receptor is dependent on arrestins but independent of G-protein activation. J. Biol. Chem. 227: 7356-7362.
  32. Kato, M., H. Kimura, Y. Motegi, A. Tachibana, H. Minakami, A. Morkawa, H. Kita. 2002. Platelet-activating factor activates two distinct effector pathways in human eosinophils. J. Immunol. 169: 5252-5259. [Abstract/Free Full Text]
  33. Deo, D., N. Bazan, J. Hunt. 2004. Activation of platelet-activating factor receptor-coupled G{alpha}q leads to stimulation of Src and focal adhesion kinase via two separate pathways in human umbilical vein endothelial cells. J. Biol. Chem. 279: 3497-3508. [Abstract/Free Full Text]
  34. Radin, J., C. Orihuela, G. Murti, C. Guglielmo, P. Murray, E. Tuomanen. 2005. beta-arrestin 1 determines the traffic pattern of PAFr-mediated endocytosis of Streptococcus pneumoniae. Infect. Immun. 73: 7827-7835. [Abstract/Free Full Text]
  35. Cabellos, C., D. E. MacIntyre, M. Forrest, M. Burroughs, S. Prasad, E. Tuomanen. 1992. Differing roles of platelet-activating factor during inflammation of the lung and subarachnoid space. J. Clin. Invest. 90: 612-618. [Medline]
  36. McCullers, J., J. Rehg. 2005. Lethal synergism between influenza virus and Streptococcus pneumoniae: characterization of a mouse model and the role of platelet-activating factor receptor. J. Infect. Dis. 186: 341-350.
  37. Rijneveld, A., S. Weijer, S. Florquin, P. Speelman, T. Shimizu, S. Ishii, T. van der Poll. 2004. Improved host defense against pneumococcal pneumonia in platelet-activating factor receptor-deficient mice. J. Infect. Dis. 189: 711-716. [Medline]
  38. Ishii, S., T. Kuawki, T. Nagase, K. Maki, F. Tashiro, S. Sunaga, W. Cao, K. Kume, Y. Fukuchi, K. Ikuta, et al 1998. Impaired anaphylactic responses with intact sensitivity to endotoxin in mice lacking a platelet activating factor receptor. J. Exp. Med. 187: 1779-1788. [Abstract/Free Full Text]
  39. Tettelin, H., K. Nelson, I. Paulsen, J. Eisen, T. Read, S. Peterson, J. Heidelberg, R. DeBoy, D. Haft, R. Dodson, et al 2001. Complete genome sequence of virulent isolates of Streptococcus pneumoniae. Science 293: 498-506. [Abstract/Free Full Text]
  40. Tuomanen, E. I., A. Tomasz, B. Hengstler, O. Zak. 1985. The relative role of bacterial cell wall and capsule in the induction of inflammation in pneumococcal meningitis. J. Infect. Dis. 151: 535-540. [Medline]
  41. Pante, N., M. Kann. 2002. Nuclear pore complex is able to transport macromolecules with diameters of about 39 nm. Mol. Biol. Cell 13: 425-434.