Function from Activated Neutrophils Receptor 3 Protects Endothelial

The Lectin-Like Domain of Complement Receptor 3 Protects Endothelial Barrier Function from Activated Neutrophils

This information is current as Vassiliki L. Tsikitis, Nicole A. Morin, Elizabeth O. of September 29, 2021. Harrington, Jorge E. Albina and Jonathan S. Reichner J Immunol 2004; 173:1284-1291; ; doi: 10.4049/jimmunol.173.2.1284 http://www.jimmunol.org/content/173/2/1284 Downloaded from

References This article cites 49 articles, 25 of which you can access for free at: http://www.jimmunol.org/content/173/2/1284.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 29, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2004 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

The Lectin-Like Domain of Complement Receptor 3 Protects Endothelial Barrier Function from Activated Neutrophils1

Vassiliki L. Tsikitis,* Nicole A. Morin,† Elizabeth O. Harrington,† Jorge E. Albina,* and Jonathan S. Reichner2*

The adhesion of neutrophils to endothelial cells is a central event leading to diapedesis and involves the binding of the I-domain ␤ ␤ of 2 integrins (CD11/CD18) to endothelial ICAMs. In addition to the I-domain, the 2 integrin complement receptor 3 (CR3) (CD11b/CD18) contains a lectin-like domain (LLD) that can alter leukocyte functions such as chemotaxis and cytotoxicity. The present study demonstrates that, in contrast to the CR3 I-domain, Ab blockade of the CR3 LLD has no role in mediating neutrophil-induced loss of endothelial barrier function. However, activation of CR3 with the LLD agonist ␤-glucan protects the barrier function of endothelial cells in the presence of activated neutrophils and reduces transendothelial migration without affecting adhesion of the neutrophils to the endothelium. The LLD site-specific mAb VIM12 obviates ␤-glucan protection while Downloaded from activation of the LLD by VIM12 cross-linking mimics the ␤-glucan response by both preserving endothelial barrier function and reducing neutrophil transendothelial migration. ␤-glucan has no direct effect on endothelial cell function in the absence of activated neutrophils. These findings demonstrate that signaling through the CR3 LLD prevents neutrophil-induced loss of endothelial barrier function and reduces diapedesis. This suggests that the LLD may be a suitable target for oligosaccharide-based anti- inflammatory therapeutics. The Journal of Immunology, 2004, 173: 1284–1291. http://www.jimmunol.org/ ntegrins are heterodimeric molecules that mediate cell-cell mediate a multitude of key functions in host defense. A region in and cell-extracellular matrix interactions (1). Leukocyte in- the extracellular domain of CR3 referred to as the inserted or I- tegrins of the ␤ family (CD11/CD18) are necessary for the Ͼ I 2 domain has been shown to bind 30 ligands including iC3b, fi- migration of leukocytes into sites of inflammation or infection as brinogen, and ICAMs 1, 2, and 3 (9, 10). The I-domain can also demonstrated by the recurrent, life-threatening infections suffered bind a number of extracellular matrix proteins including fibronec- by patients lacking expression of the CD18 subunit (leukocyte ad- tin, laminin, collagen, vitronectin, thrombospondin, and Cyr61 hesion deficiency) (2). Conversely, excessive neutrophil adhesion (11Ð14). This ability of the I-domain of CR3 to bind diverse li- ␤ to the vascular endothelium mediated by the 2 integrins has been gands has recently been attributed to a consensus binding site by guest on September 29, 2021 implicated in diseases such as myocardial infarction, ischemia, and within CD11b (Lys245-Arg261) (15). reperfusion injury (3Ð5). In addition to the I-domain, CR3 contains a unique lectin-like 3 Complement receptor 3 (CD11b/CD18) (CR3), one of four domain (LLD), which permits binding of microbial polysaccha- ␤ members of the 2-integrin family, is a multifunctional adhesion rides such as ␤1,3-linked glucose polymers (viz ␤-glucan). This ␤ molecule in which a common 2 (CD18) subunit is noncovalently aspect of CR3 mediates the recognition of several pathogens such ␣ bound to the M subunit (CD11b). A number of studies have as Candida albicans, Leishmania sp., Bordetella pertussis, and shown that CR3 plays a role in neutrophil adhesion and transmi- Pneumocystis carinii (16Ð18). Furthermore, the LLD participates gration through an endothelial cell layer, and in migration through in the formation of transmembrane signaling complexes with GPI- the extracellular matrix (6). CR3 also mediates leukocyte phago- anchored glycoproteins such as CD16b, CD14, and CD87 enabling cytosis of opsonized and unopsonized pathogens, homotypic ag- CR3 to mediate cytoskeleton-dependent functions including gregation, and adhesion-dependent respiratory burst (7, 8). The phagocytosis, degranulation, and adhesion (19Ð21). Taken to- extensive repertoire of CR3 ligands renders this integrin able to gether, these findings underscore the functional significance of the LLD in host defense against infection and injury. *Department of Surgery, Rhode Island Hospital and Brown University Medical Dual ligation of the I-domain and the LLD by their correspond- School, Providence, RI 02903; and †Pulmonary Vascular Biology Research Labora- ing ligands has been shown to affect CR3-dependent functions tory, Providence Veterans Affairs Medical Center, Department of Medicine, Brown University Medical School, Providence, RI 02912 differently than ligation of either site alone. Co-occupancy of the Received for publication December 19, 2003. Accepted for publication May 5, 2004. I-domain with the extracellular matrix protein fibronectin, and of the LLD with ␤-glucan, resulted in an increase in the chemotactic 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 capacity of neutrophils toward fMLP (22). Furthermore, tumor with 18 U.S.C. Section 1734 solely to indicate this fact. cells, otherwise resistant to NK cells, became susceptible to killing 1 This work was supported by National Institutes of Health Grants GM-066194 (to when opsonized with the CR3 I-domain ligand iC3b and treated J.S.R.), GM-42859 (to J.E.A.), and HL-067795 (to E.O.H.), the Carter Family Char- ␤ itable Trust (Armand D. Versaci Research Scholar in Surgical Sciences Award (to with the LLD ligand -glucan (23). The current investigation V.L.T.)), and allocations to the Department of Surgery by Rhode Island Hospital. sought to determine the effect of LLD activation on neutrophil: 2 Address correspondence and reprint requests to Dr. Jonathan S. Reichner, Depart- endothelial cell interactions. The rationale for this study is that the ment of Surgery, NAB-219, Rhode Island Hospital, 593 Eddy Street, Providence, RI I-domain of CR3 will be ligated by endothelial ICAM and thereby 02903. E-mail address: [email protected] mediate the interaction between the neutrophil and the endothelial 3 Abbreviations used in this paper: CR3, complement receptor 3; LLD, lectin-like domain; PMN, polymorphonuclear neutrophil; RPAEC, rat pulmonary aortic endo- cell (9, 10). The hypothesis to be tested is that coligation of the thelial cell; uPAR, urokinase-type plasminogen activator receptor. I-domain with ICAM and the LLD with ␤-glucan would alter this

CR3-dependent intercellular interaction. Findings to be presented Measurement of electrical resistance show that neutrophils, activated with fMLP, impaired the barrier Transendothelial electrical resistance, an index of endothelial cell barrier function of intact endothelial monolayers, which was preserved by function, was measured in real time using an electric cell-substrate imped- ␤-glucan or by Ab cross-linking of the CR3 LLD. Ligation of the ance sensor system (Applied Biophysics, Troy, NY) (24). For resistance CR3 LLD significantly reduced polymorphonuclear neutrophil measurements, endothelial cells were plated on sterile eight-chambered ␮ (PMN) transmigration. Additional findings show that ␤-glucan did gold-plated electrode arrays precoated with collagen type I (30 g/ml) and grown to confluence. The electrode arrays were then mounted on the elec- not affect the adhesion of neutrophils to endothelial cells. These tric cell-substrate impedance sensor system within an incubator (37¡C, 5%

findings suggest that activation of the CR3 LLD prevents endo- CO2) and connected to its recorder device. Confluence of each endothelial thelial barrier dysfunction in the presence of activated neutrophils monolayer was established before use from the resistive portion of the and may thereby attenuate tissue damage during autoimmune or impedance tracing. Only monolayers, showing a resistance of 1Ð1.3 kilo- ohms were used in experiments (25). Naive neutrophils (6 ϫ 105 cells per hyperinflammatory processes. This work identifies the CR3 LLD well in 200 ␮l of DMEM) or neutrophils pretreated with fMLP (10Ϫ6 M) as a potential target site for the development of oligosaccharide- for 20 min on ice, were added to each well of the array. Thrombin (0.8 based therapeutics with anti-inflammatory applications. U/ml), which causes instant retraction of the endothelial monolayer, was used as a positive control. Monolayer resistance was recorded over3hin 1-min intervals after PMN addition. Values are reported as percentage Materials and Methods change of resistance normalized to initial baseline resistance (observed Reagents resistance/initial resistance ϫ 100). Highly purified, pharmaceutical grade, soluble ␤-glucan (BETAFECTIN, Permeability assays Ϯ 150,000 20,000 m.w.) was obtained from Biopolymer Engineering Permeability assays were performed using a two-compartment chamber Downloaded from (Eagan, MN). Rat-tail collagen type I and fibronectin were purchased from ␮ ␣ system obtained from Costar (Acton, MA). This system contained 0.4- m BD Biosciences (Bedford, MA). Rat TNF- was obtained from BioSource pore polycarbonate transwell supports. HUVEC were plated to confluence International (Camarillo, CA). Dextran, thrombin, fMLP, Evans blue dye- in the rehydrated transwell supports at a density of 105 cells per well and conjugated albumin, and endotoxin-free BSA were purchased from Sigma- permitted to adhere overnight in complete medium. The following day, the Aldrich (St. Louis, MO). PBS and HBSS were obtained from Invitrogen medium was changed in both chambers. Thrombin and HRP (50 ␮M HRP) Life Technologies (Grand Island, NY). DMEM was purchased from In- were then placed in the upper chamber. The appearance of HRP in the vitrogen Life Technologies and FBS was obtained from HyClone (Logan, ␮

lower chamber was determined by retrieving 20- l aliquots of medium http://www.jimmunol.org/ UT). CFSE was purchased from Molecular Probes (Eugene, OR). All re- Ͻ from the lower compartment every 20 min, over 2 h. The HRP concentra- agents used contained 0.1 pg/ml endotoxin, as determined by Limulus tion was determined using a spectrophotometric assay. The assay was con- amoebocyte lysate screening (BioWhittaker, Walkersville, MD). Some re- ducted by incubating the sample in a solution containing 0.4 mg/ml-1 O- agents were treated with Detoxi-Gel (Pierce, Rockford, IL) to remove re- phenylenediamine HCl and 0.012% H2O2 in 0.5M phosphate-citrate buffer, sidual endotoxin. The CD11b-specific Ab VIM12 was purchased from pH 5.0, for 30 min at 25¡C. The reaction was stopped with HCl, a final Caltag Laboratories (Burlingame, CA). The F(abЈ) of goat anti-mouse IgG 2 concentration of 0.55 M, and the absorbance was taken at 492 nm. The data and human HLA class I Ab were purchased from Sigma-Aldrich. The CR3 are presented as the number of moles of HRP that have diffused to the I-domain specific Ab LM2.1 was obtained from Bender MedSystems (San lower chamber over time. Bruno, CA) and the CD18-specific Ab TS1/18 was obtained from Pierce (Woburn, MA). CBRM1/23 was a gift from Dr. T. Springer (Center for Neutrophil adhesion

Blood Research and Harvard University Medical School, Boston, MA). by guest on September 29, 2021 Neutrophils (5 ϫ 106 cells/ml) were fluoresceinated using CFSE at 1.0 ␮g/ml by incubating at room temperature for 30 min in the dark. Cells were Isolation and activation of human neutrophils washed and resuspended to 3 ϫ 106 cells/ml in DMEM and kept on ice Human neutrophils were isolated from heparinized venous blood of healthy until use. Viability of fluorescent PMNs was monitored by trypan blue dye volunteers by gradient centrifugation on Ficoll-Hypaque (Sigma-Aldrich), exclusion and cells were Ͼ85% viable at time of use. Polystyrene 96-well followed by erythrocyte sedimentation with 3% dextran (500,000 m.w.). tissue culture plates (BD Biosciences, Franklin Lakes, NJ) were coated The leukocyte-rich supernatant underwent hypotonic lysis of residual with collagen type I (30 ␮g/ml) for 1 h at 37¡C and then blocked with BSA erythrocytes. Cells were resuspended in ice-cold DMEM and counted. (10 mg/ml, endotoxin-free) for 1 h. After rinsing twice with sterile PBS, 5 PMN purity and viability were consistently Ͼ90% as determined by trypan RPAEC (10 cells/well) were plated and grown to confluence which was blue dye exclusion. Where indicated, neutrophils were treated with fMLP confirmed microscopically. Endothelial cells were stimulated, or not, with (10Ϫ6 M) for 20 min on ice and then ␤-glucan (10 ␮g/ml) was added as a rat TNF-␣ (50 ng/ml) for 6 h and then washed three times with sterile PBS. second treatment agent for 20 min at room temperature. When HUVECs were used, plates were precoated with fibronectin (30 ␮g/ml) and then blocked with BSA as described above. HUVECs were stimulated with human TNF-␣ (20 ng/ml) for 6 h before the addition of Ab treatment of human neutrophils neutrophils. Fluorescent neutrophils (3 ϫ 105 cells) were added to each Neutrophils (3 ϫ 106 cells/ml), were incubated with no Ab, LM2.1 (20 well and incubated for 45 min at 37¡Cin5%CO2. Nonadherent cells were ␮g/ml), TS1/18 (10 ␮g/ml), VIM12 (5 ␮g/106 cells), CBRM1/23 ascitic removed by washing twice with serum-free DMEM and fluorescence was fluid (10 ␮l/ml), and anti-HLA class I IgG (5 ␮g/106 cells) for 30 min on quantified in a FL 500 Microplate Fluorescence Reader (Bio-Tek Instru- ice and then, where indicated, with ␤-glucan for 20 min at room temper- ments, Winooski, VT) with excitation/emission wavelength settings of ature. Anti-HLA class I IgG was used to control nonspecific effects of Ab 485/530 nm and a sensitivity setting of 40Ð44. All determinations were binding to cell surface Ags. For cross-linking of the LLD, purified neutro- conducted in six-well replicates and repeated three times. phils (3 ϫ 106 cells/ml) were treated with VIM12 as described above, Ј Neutrophil transmigration assay washed, and then incubated with goat anti-mouse F(ab )2 (1/20 dilution) for 20 min on ice. All neutrophil groups were warmed to room temperature Assays were performed as described in Taooka et al. (26). HUVECs or for 20 min before addition onto the endothelial monolayers. RPAECs were plated onto type I collagen polycarbonate inserts (transwell inserts, 6.5-mm diameter, 3-␮m pore for 24-well plates; VWR, Boston, 5 5 Endothelial cell culture MA). HUVECs (10 cells) and RPAECs (2 ϫ 10 cells) were grown in 200 ␮l of cell-specific complete medium as described above. Additional com- HUVEC (Cambrex, Walkersville MD), were grown in endothelial cell plete medium was added to the lower chamber of each well and endothelial growth medium containing 2% FBS, 10 ng/ml human recombinant epider- cells were grown to confluence over 72 h. Evans blue dye-conjugated al- mal growth factor, 50 mg/ml gentamicin, 50 ng/ml amphotericin B, 12 bumin was used as a marker for macromolecular permeability of the en- mg/ml bovine brain extract, and 1 mg/ml hydrocortisone as provided by the dothelial cell monolayers before the start of the experiment as described in manufacturer. Cells were used between passages 5 and 9. Rat pulmonary Gautam et al. (27). Dye exclusion was determined by spectrophotometric Ͻ artery endothelial cells (RPAEC) and rat lung microvascular endothelial reading at OD550. Monolayers allowing 10% leakage of the dye relative cells were obtained from Dr. P. Lee (Yale University School of Medicine, to maximal leakage (thrombin ϫ 100%) were accepted for further study. New Haven, CT) and grown in DMEM with 20% FBS supplemented with PMN (106 cells in 200 ␮l of DMEM) activated with fMLP (10Ϫ9 M) and antibiotics. Cells were used between passages 17 and 24. then treated, or not, with ␤-glucan for 40 min at room temperature, were 1286 CR3 LECTIN-LIKE SITE PROTECTS ENDOTHELIAL BARRIER FUNCTION added to the upper chamber of the transwells. The lower chamber was replaced with 500 ␮l of serum-free DMEM or serum-free DMEM with Ϫ6 10 M fMLP. After3hat37¡Cin5%CO2 in air, medium from the lower chambers was collected and rinsed several times and the migrated cells were counted in a hemacytometer. All determinations were conducted in quadruplicate and repeated three times. Results are presented as percentage of cells migrated.

Data presentation and statistical analysis Data are presented are mean Ϯ SEM unless otherwise stated, with n being the number of independent experiments. Statistical assessments were made using the Student t test or ANOVA with Newman-Keuls posthoc analysis as appropriate and statistical signiﬁcance set at p Ͻ 0.05.

Results ␤-glucan protects the barrier function of endothelial monolayers in the presence of activated neutrophils Electrical resistance across conﬂuent RPAEC monolayers was measured in the absence or presence of PMNs over a 3-h period.

The change of the resistance shown correlates with macromolec- Downloaded from ular leakage and it is used here as an index of endothelial monolayer barrier function (Fig. 1). PMNs activated with fMLP caused a greater drop in transendothelial resistance (33 Ϯ 4%) than unstimulated PMNs, suggesting an increase in endothelial monolayer permeability. In contrast, when fMLP-activated PMNs were pre- ␤

treated with -glucan, the integrity of the endothelial monolayer http://www.jimmunol.org/ was preserved with a drop in the endothelial monolayer resistance of 8.2 Ϯ 3%. The extent of endothelial monolayer permeability caused by fMLP-activated neutrophils exceeds that of thrombin, a benchmark for instantaneous disruption of endothelial cell barrier function. ␤-glucan did not affect thrombin-induced endothelial barrier dysfunction (data not shown). FIGURE 1. ␤-glucan diminishes endothelial barrier dysfunction in- Ϫ6 Coincubation of PMNs with fMLP (10 M) and dextran (10 duced by activated PMNs. a, Representative tracing of electrical resistance ␮g/ml), a polyglucose polymer of similar size ␤-glucan, did not across RPAEC monolayers measured in the absence (medium) or presence Ϫ6 prevent endothelial barrier dysfunction suggesting that the effect of of untreated PMNs, PMNs treated with fMLP (10 M) alone, PMNs by guest on September 29, 2021 Ϫ ␤-glucan is not due to a nonspecific response to polysaccharide treated with fMLP (10 6 M) and ␤-glucan (10 ␮g/ml), or thrombin (0.8 (data not shown). Moreover, supernatants from naive or fMLP- U/ml) as detailed in Materials and Methods. Arrow indicates the time of Ϯ activated neutrophils that were pretreated with or without ␤-glucan addition of media, cells, or thrombin. b, Values are mean SEM of the percent change of normalized resistance at 60 min of PMN addition from did not alter endothelial monolayer resistance indicating the re- six independent experiments. Supernatants from PMNs activated with quirement of neutrophils for the observed increase in endothelial fMLPϪ (Sup/fMLP) or with fMLP and ␤-glucan (Sup/fMLP ϩ ␤-glucan) monolayer permeability (Fig. 1b). Direct addition of fMLP and caused no change on the normalized resistance recorded. The presence of ␤-glucan into the medium of endothelial cells was also without either ␤-glucan alone (10 ␮g/ml) or fMLP (10Ϫ6 M) alone in the media p Ͻ 0.01 vs fMLP-treated ,ء .effect on recorded resistance. caused no change in normalized resistance To further confirm that the presence of neutrophils is necessary PMNs. for endothelial cell retraction, neutrophils were fixed with 2% paraformaldehyde and added onto the endothelial monolayer in the presence or absence of ␤-glucan. Retraction was observed in both conditions suggesting that ␤-glucan only regulates the interaction layers. FMLP-activated neutrophils caused the same decrease in between viable neutrophils and endothelial monolayers (data not relative resistance (40 Ϯ 5.2%) of monolayers of activated shown). HUVEC (Fig. 2a) as well as RPAEC and RLMVC (not shown). This drop in normalized resistance was reduced to 20 Ϯ 4.8% by ␤ -glucan prevents barrier dysfunction by activated neutrophils 10 ␮g/ml ␤-glucan and to 9 Ϯ 2.8% by 50 ␮g/ml. Thus, the barrier ␣ on different TNF- -activated endothelial monolayers protective effect of ␤-glucan was independent of the activation Experiments were performed to determine whether the ␤-glucan state of the endothelial monolayers and was seen with three dif- protection of endothelial barrier function is maintained when both ferent endothelial cell lines. PMN and endothelial cells are activated and when different endo- To determine whether the effect of ␤-glucan is demonstrable on thelial cell lines are used. Confluent RPAEC or HUVEC mono- a model of endothelial paracellular permeability, PMNs were ac- layers were treated with TNF-␣ before the addition of fMLP-activated with fMLP (10Ϫ6 M) and placed together with HRP on tivated PMNs. To validate that the monolayers were activated by HUVEC monolayers grown on permeable inserts (Fig. 2b). Acti- TNF-␣, PMN:endothelial cell adhesion assays were performed, as vated PMNs caused a significant increase in the permeability of the described in Materials and Methods. There was a 2.7-fold increase endothelial monolayer to an extent comparable to thrombin. In- in the adhesion of fMLP-activated neutrophils to TNF-␣-treated clusion of ␤-glucan had a protective effect on monolayer perme- endothelial cells as compared with untreated monolayers (4700 Ϯ ability as shown by a decrease in the flux of HRP, consistent with 199 vs 1730 Ϯ 148 fluorescent units, respectively). Similar results the protection measured by transcellular electrical resistance were obtained with either RPAEC or HUVEC endothelial mono- shown above. The Journal of Immunology 1287

FIGURE 3. ␤-glucan does not alter PMN adhesion to endothelial cells. RPAEC monolayers were activated, or not, with TNF-␣ as described in

Materials and Methods. Fluorescent naive (unstimulated) or fMLP-acti- Downloaded from vated (10Ϫ6 M) PMNs were added to endothelial cells in the presence or absence of ␤-glucan (10 ␮g/ml). Adhesion was determined after incubation for 45 min. Data are presented as mean Ϯ SEM from six independent experiments. http://www.jimmunol.org/ These results suggest that ␤-glucan did not signiﬁcantly alter adhesion of resting or activated neutrophils to resting or activated endothelial cells. Neutrophil adhesion and transmigration can occur with or with- FIGURE 2. ␤-glucan preserves barrier function of TNF-␣-treated en- out increased vascular permeability (28), thus we next sought to dothelium in the presence of activated PMNs. a, HUVEC monolayers were determine whether the effect of ␤-glucan was germane to neutro- ␣ incubated with TNF- (20 ng/ml) for6hat37¡C. fMLP-activated PMN phil extravasation and migration across endothelial monolayers. ␣ were added to the TNF- activated HUVEC in the presence of 10 or 50 RPAECs were grown and activated with TNF-␣ as described in

␮g/ml ␤-glucan. Transendothelial resistance was measured at 60 min and by guest on September 29, 2021 Ͻ Materials and Methods and migration of activated neutrophils was ء Ϯ presented as mean SEM from three independent experiments. , p ␤ 0.01 vs fMLP treatment alone for both ␤-glucan concentrations. b, ␤-glu- measured. The presence of -glucan resulted in neutrophil trans- Ϯ Ϯ can preserves paracellular permeability of HUVEC monolayers in the pres- migration of 31 1.2% as compared with 60 8% for neutrophils ence of fMLP-activated PMNs. Transendothelial ﬂux of HRP is shown at in the absence of ␤-glucan (Fig. 4). This decrease in neutrophil 1 h. Data are mean Ϯ SD from representative of three identical experi- .p Ͻ 0.01 vs fMLP-activated PMNs ,ء .ments

␤-glucan does not alter PMN:endothelial cell adhesion and decreases neutrophil transmigration To investigate whether the role of ␤-glucan on barrier function is based on altering adhesive interaction between neutrophil and endothelial cells, adhesion assays were performed as described in the Materials and Methods. Results in Fig. 3 show the relative adhesion of fMLP-activated neutrophils vs fMLP ϩ ␤-glucan cotreated neutrophils: 4890 Ϯ 750 vs 4786 Ϯ 586 fluorescent units on resting endothelium, respectively. Experiments also compared neutrophil adhesion to TNF-␣-activated endothelium and show similar adhesion of fMLP-activated neutrophils vs fMLP ϩ ␤-glucan cotreated neutrophils: 5810 Ϯ 516 vs 4774 Ϯ 593 fluorescent units, respectively ( p Ͼ 0.05, t test). In this series of experiments, endothelial cells were activated with TNF-␣ 6 h prior to the adhesion assay as described in Materials and Methods. Adhesion assays were also performed where endothelial cells were pretreated with ␤ the same concentration of TNF-␣ for 18 h with similar results: FIGURE 4. Pretreatment of activated PMNs with -glucan decreased diapedesis. Monolayers of RPAECs were activated with TNF-␣ on colla- (adhesion of fMLP-activated neutrophils vs fMLP ϩ ␤-glucan co- Ϯ Ϯ gen precoated semipermeable membranes as detailed in Materials and treated neutrophils: 9280 406 vs 9704 1000 fluorescent units, Methods. The percentage of neutrophils pretreated with fMLP (10Ϫ9 M) respectively). Fig. 3 also includes results obtained from nonacti- that migrated through an intact monolayer toward an fMLP gradient (10Ϫ6 vated, naive neutrophils added to resting and activated endothe- M) in the presence or absence of ␤-glucan (50 ␮g/ml) was determined after lium and while the extent of adhesion is less than seen with acti- 1hat37¡C. Data are expressed as means Ϯ SEM from three independent .p Ͻ 0.05 vs no ␤-glucan ,ء ;vated neutrophils, the addition of ␤-glucan was without affect. experiments 1288 CR3 LECTIN-LIKE SITE PROTECTS ENDOTHELIAL BARRIER FUNCTION transmigration induced by ␤-glucan directly correlates with its in- does not play a role in regulating neutrophil-induced endothelial hibitory effect on neutrophil-dependent endothelial barrier dys- barrier dysfunction. function. Identical results were obtained using resting RPAEC as Transmigration experiments were performed to determine well as with resting and activated HUVEC (data not shown). whether Ab blockade of the CR3 LLD on neutrophils affects their migratory capacity through an intact RPAEC monolayer. As seen ␤ Ab blocking of the CR3 LLD obviates the protective role of - in Fig. 5b, ␤-glucan decreased the migration of neutrophils treated glucan of endothelial barrier function with control anti-HLA Ab but not of neutrophils treated with the To test the hypothesis that the prevention of endothelial barrier CR3 LLD-blocking Ab VIM12 (24 Ϯ 7% vs 64 Ϯ 17% transmi- dysfunction is mediated by ligation of the neutrophil CR3 LLD, grated cells, respectively). These results suggest that ␤-glucan in- Ab-blocking experiments were performed as described in Materi- hibits neutrophil diapedesis through a mechanism requiring its als and Methods. Results in Fig. 5a using RPAEC monolayers binding to the neutrophil CR3 LLD. show that neutrophils treated with VIM12 or CBRM1/23 in the presence of ␤-glucan caused a 45 Ϯ 2.4% and 48 Ϯ 7.7% drop in VIM12 activation of CR3 LLD mimicked the ␤-glucan response resistance, respectively, thereby blocking the protection afforded by preserving endothelial cell barrier function and reducing by ␤-glucan. Neutrophils treated with the cell surface control Ab neutrophil transmigration against HLA class I molecules did not alter the protection by To verify that signaling through the LLD of neutrophil CR3 is ␤ -glucan. Activated neutrophils treated with LM2.1, an Ab spe- sufficient to protect endothelial barrier function, the effect of site- cific to the CR3 I-domain, were not able to cause significant en- specific Ab cross-linking was examined. Cross-linking an anti- Ϯ Downloaded from dothelial cell retraction (5 1.3%) due to the ability of the Ab to integrin Ab with a secondary Ab leads to clustering and activation prevent neutrophil recognition of endothelial ICAM. Activated of the target integrin (29). As shown in Fig. 6a, blocking CR3 with PMNs, pretreated with TS1/18, a CD18-specific Ab, also blocked VIM12 did not alter the ability of activated PMNs to impair en- PMN-induced endothelial barrier dysfunction (data not shown). dothelial barrier function. However, cross-linking VIM12 with The above findings suggest that the I-domain of CR3 mediates goat anti-mouse F(abЈ) Ab to activate the CR3 LLD resulted in PMN-dependent endothelial cell retraction whereas the LLD alone 2 maintenance of barrier function. Cross-linking anti-HLA class I mAb had no effect. Fig. 6b demonstrates that VIM12 cross-linked http://www.jimmunol.org/ activated PMNs caused a 10 Ϯ 2.5% decrease in resistance vs HLA cross-linked activated PMNs which caused a 29 Ϯ 4% decrease. Cross-linking CR3 was able to mimic the protective role of ␤-glucan on the endothelial barrier function thereby suggesting that signaling through the LLD of CR3 can preserve the integrity of the endothelial monolayer in the presence of activated neutrophils.

Transmigration assays were performed to determine the migra- by guest on September 29, 2021 tory capacity of LLD cross-linked neutrophils through an intact endothelial monolayer. Fig. 6c shows that VIM12 cross-linking signiﬁcantly diminished neutrophil transmigration as compared with HLA cross-linking. Thus, cross-linking of CR3 was able to mimic the ␤-glucan effects on both maintenance of barrier function and reduction of neutrophil transmigration.

Discussion The regulated migration of leukocytes through an endothelial cell barrier is a universal aspect of host defense and inflammation. This process is mediated by leukocyte integrins such as CR3, the im- portance of which is recognized by the reduced ability of cells to extravasate under conditions where CR3 function is blocked or is genetically inactive (2, 30). Conversely, excessive integrin-mediated adhesion to vascular endothelium has been implicated in disease states such as myocardial infarction and ischemia/reperfusion injury (3Ð5). Therefore, the ability to regulate the function of CR3 may offer opportunities to intervene therapeutically in clinical cir- cumstances wherein excessive inflammation contributes to patho- genesis, rather than resolution of the disease. CR3 contains two distinct ligand-binding sites: the I-domain and the LLD. Studies FIGURE 5. Ab blocking of the CR3 LLD blocks the ␤-glucan effect on from this and other laboratories suggest that dual ligation of the endothelial barrier function and PMN transmigration. a, fMLP-activated lectin-like domain and the I-domain may alter neutrophil function Ϫ6 (10 M) PMNs were pretreated with Abs and added onto confluent mono- differently than ligation of either site alone (23). The current study layers of RPAECs in the presence or absence of ␤-glucan (10 ␮g/ml). Data tested and confirmed the hypothesis that LLD ligation would affect p Ͻ:#,ء ;are shown as mean Ϯ SEM from six independent experiments 0.01 vs control anti-HLA with ␤-glucan. b, Percentage of PMNs migrated CR3 activity as noted by alterations in neutrophil:endothelial cell toward fMLP or PBS after3hat37¡C is shown. PMNs were pretreated, or interactions. Agonists of the CR3 LLD prevented loss of endothe- not, with fMLP (10Ϫ9 M) and Ab before transmigration. Data are ex- lial barrier function caused by activated neutrophils and identified -p Ͻ 0.01 the LLD as a novel target site for the development of anti ,ء ;pressed as mean Ϯ SEM from three independent experiments vs VIM12 with ␤-glucan. inflammatory therapeutics. The Journal of Immunology 1289

plasmic domain permitted neutrophil adhesion to engineered Chinese hamster ovary cells, but abrogated transmigration. The authors concluded that neutrophil migration is not a direct consequence of enhanced adhesion mediated by ICAM-1, but requires ICAM signaling. Whether occupancy of LLD alters ICAM signaling in endothelial cells remains to be determined. This mechanism may not be applicable to epithelial cell targets that are also sensitive to PMN-induced disruption of barrier function. Fucoidin binding to the CR3 I-domain was shown to alter adhesive interactions between PMN and epithelial cells (33), an observation that was not found in our present studies. ␤ 2 integrins have been shown to be required for neutrophil: endothelial adhesion and transmigration in several in vitro experimental systems (32, 34, 35). Of direct relevance to the physio- ␤ logical response to inﬂammation, the role of the 2 integrin CR3 in mediating ICAM-1 binding is more pronounced in chemokine- activated neutrophils than in resting cells, which rely on LFA-1 for their binding onto the endothelial cells (36). Recent ﬁndings from

Luscinskas and coworkers (37) have shown that maintenance of Downloaded from endothelial barrier function does not necessarily equate with reduced neutrophil transmigration. Studies showed that overexpres- sion of vascular endothelial-cadherin in HUVECs was sufﬁcient to increase endothelial barrier function without altering neutrophil migration. Therefore, ﬁnding that ␤-glucan maintained barrier

function of endothelial cells in the presence of activated neutro- http://www.jimmunol.org/ phils could not predict whether diapedesis would be similarly af- fected. Thus, we determined experimentally the effect of the CR3- ligated LLD on neutrophil transmigration. Results indicate that the presence of ␤-glucan decreased extravasation and migration of activated neutrophils through an intact endothelial monolayer. Spe- cifically, Fig. 4 shows a 50% reduction in transmigration of neutrophils in the presence of ␤-glucan which directly correlated with FIGURE 6. Ab cross-linking of the CR3 LLD mimicked the ␤-glucan the extent of protection of barrier function shown in Fig. 1. This effect by preserving endothelial barrier function and decreasing PMN di- observation was repeated with the VIM12 cross-linking experi- by guest on September 29, 2021 apedesis. fMLP-activated (10Ϫ6 M) PMNs were preincubated, or not, with ments, thereby confirming that the decrease in neutrophil transmi- VIM12 and subsequently cross-linked (x-link) with goat anti-mouse gration is mediated through signaling of the CR3 LLD. Ј F(ab )2 as described in Materials and Methods. a, Representative tracing of Work by Ross and colleagues (38, 39) showed that binding of normalized transendothelial electrical resistance across RPAEC monolay- the CR3 LLD by ␤-glucan induces the high affinity metal ion- ers in response to fMLP-activated PMN addition either in the presence of dependent activation site reporter epitope which is recognized by ␤-glucan or with VIM12 cross-linked. b, Values are mean Ϯ SEM of the the CBRM1/5 activation reporter mAb. However, despite the high percent change of normalized resistance at 60 min of PMN addition from ␤ p Ͻ 0.01 vs HLA x-link. c, Transendo- affinity conversion, -glucan inhibited adhesion of neutrophils to ,ء ;six independent experiments thelial migration of fMLP-activated (10Ϫ9 M) PMNs that were cross- immobilized recombinant ICAM. These authors propose that CR3 linked, or not, with VIM12 or HLA Ab. Transmigration was measured must form complexes with uPAR for clustering to take place lead- relative to migration of naive PMNs toward fMLP and data are expressed ing to adhesion. Binding of ␤-glucan to the same lectin-like site as -p Ͻ 0.01 vs HLA would be occupied by urokinase-type plasminogen activator re ,ء ;as mean Ϯ SEM from three independent experiments x-link. ceptor (uPAR) prevents complex formation and thereby CR3 binding to ICAM. Findings in this report show a lack of a significant effect of ␤-glucan on neutrophil adhesion to endothelial cells. The present studies demonstrate that the protective effect of en- What the two studies have in common is the finding that ␤-glucan dothelial barrier function by ␤-glucan is due to its action on the does not promote excessive leukocyte adhesion to ICAM or to neutrophil, and not on the endothelial cell. The relevance of this endothelial cells despite maintaining CR3 in a high affinity con- finding is heightened by a recent report demonstrating the capacity formation. In that regard, the activation state of CR3 would not be of endothelial cells to directly bind ␤-glucan (31), although the expected to impose an impediment to neutrophil diapedesis. functional significance of this observation is not yet defined. LLD- An important physiological function of the CR3 LLD lies in its specific Ab blocking and activation experiments, performed on ability to form membrane complexes with GPI-anchored receptors neutrophils before addition to endothelial monolayers, were found such as Fc␥RIIIB (CD16b), uPAR (CD87), or LPS (CD14). Be- to regulate barrier function. These experiments demonstrate that cause these receptors do not transit the plasma membrane, signal- ␤-glucan does not exert a direct effect on endothelial cells in the ing functions are accomplished as a result of partnering with CR3 context of protecting barrier function from neutrophil-induced via the LLD. Occupancy of LLD with CD87 was shown by Ross disruption. and colleagues (40) to activate CR3 into a high affinity state for We demonstrated that LLD ligation can uncouple neutrophil ICAM binding. In these studies, soluble ␤-glucan dissociated CR3 adhesion from maintenance of barrier function and transmigration. from CD87 and reduced the affinity of CR3 for ICAM. The authors The mechanism underlying this novel finding is not yet known, predicted that use of small m.w. polysaccharides might cause a however Sans et al. (32) found that deletion of the ICAM cyto- steric blockade of the I-domain, thus preventing PMN diapedesis 1290 CR3 LECTIN-LIKE SITE PROTECTS ENDOTHELIAL BARRIER FUNCTION and collateral tissue damage. Current studies, using multiple en- tants were diluted 1/16 which approximates incubation with 6 ϫ dothelial cell lines, did not demonstrate alterations in adhesive 106 cells/ml. We used 3 ϫ 106 PMNs/ml supernatant as a control interactions between ␤-glucan-treated activated PMNs and the en- for our experimental system to match conditions in which perme- dothelium, but did in fact support that prediction. Our findings ability measurements were obtained in the presence of the same indicate that the LLD does not mediate PMN:endothelial cell ad- number of effector PMNs. Therefore, it is quite possible that we hesion because activated neutrophils pretreated with blocking Abs did not condition our media with a sufficient number of cells to against the lectin site still caused endothelial barrier dysfunction. reproduce the aforementioned mechanism of barrier function pro- In comparison, neutrophils pretreated with a blocking Ab to the tection. Finally, addition of fixed neutrophils caused retraction of CR3 I-domain maintained the integrity of the endothelial mono- the endothelial monolayer which was not altered by ␤-glucan, sig- layer presumably by obviating ICAM recognition. These findings nifying that protection of the endothelial barrier function occurs correlate with previous studies that binding of soluble ␤-glucan through active signaling of the LLD and therefore requires viable can induce the high affinity conformation of the CR3 I-domain effector cells. without promoting neutrophil spreading and adhesion (39). The preparation of ␤-glucan used in our current studies has al- Protection of the integrity of the monolayer via LLD cross-link- ready been shown to be well-tolerated by patients in clinical trials ing confirms that signaling events through the CR3 lectin site alter when used for an anti-infectious indication (51). It is suggested PMN CR3-dependent functions normally associated with occu- that ␤-glucan or structurally related lectin site agonists may be pancy of the I-domain. This finding supports and extends previous useful in the design of therapeutic agents indicated for attenuation reports from this laboratory that demonstrated a conversion of ran- of excessive inflammation. dom to directed neutrophil migration when extracellular matrix is Downloaded from supplemented with ␤-glucan (22). A series of Ab blocking exper- Acknowledgments iments attributed this shift in migration to corecognition of fi- We thank Dr. Anne Pereira (Department of Pathology, University of Okla- bronectin by the CR3 I-domain and ␤-glucan by the LLD. Dual homa Health Sciences Center, Oklahoma City, OK) for measurements of ligation of CR3 by ␤-glucan and the I-domain ligand iC3b has also CAP37 and for helpful advice. been shown to result in NK cell cytolysis of iC3b-opsonized tumor targets that would otherwise resist this mechanism of cell-medi- References http://www.jimmunol.org/ ated cytotoxicity (39). This suggests that regulation of CR3 I-do- 1. Hynes, R. O. 1992. Integrins: versatility, modulation, and signaling in cell ad- main function through the LLD is not restricted to neutrophils and hesion. Cell 69:11. 2. Bunting, M., E. S. Harris, T. M. McIntyre, S. M. Prescott, and G. A. Zimmerman. may find therapeutic applicability in heightening host defense 2002. Leukocyte adhesion deficiency syndromes: adhesion and tethering defects ␤ against tumor cells. involving 2 integrins and selectin ligands. Curr. Opin. Hematol. 9:30. The nature of the ␤-glucan receptor has been of great interest 3. Vedder, N. B., R. K. Winn, C. L. Rice, E. Y. Chi, K. E. Arfors, and J. M. Harlan. 1988. A monoclonal antibody to the adherence-promoting leukocyte glycopro- over the years and several strong candidates have been reported, in tein, CD18, reduces organ injury and improves survival from hemorrhagic shock ␤ and resuscitation in rabbits. J. Clin. Invest. 81:939. addition to the 2 integrin CR3 (41Ð43). For example, dectin-1, a ␤ 4. Mileski, W. J., R. K. Winn, N. B. Vedder, T. H. Pohlman, J. M. Harlan, and newly described -glucan receptor has been reported to be respon- C. L. Rice. 1990. Inhibition of CD18-dependent neutrophil adherence reduces sible for the recognition of unopsonized zymosan (44Ð46) and organ injury after hemorrhagic shock in primates. Surgery 108:206. by guest on September 29, 2021 shown to be expressed on cells of macrophage and neutrophil lin- 5. Simpson, P. J., R. F. Todd, 3rd, J. K. Mickelson, J. C. Fantone, K. P. Gallagher, ␤ K. A. Lee, Y. Tamura, M. Cronin, and B. R. Lucchesi. 1990. Sustained limitation eage. In our experimental system, the mechanism of -glucan pro- of myocardial reperfusion injury by a monoclonal antibody that alters leukocyte tection of the endothelial monolayer was found to be completely function. Circulation 81:226. accountable by its binding to the LLD of CR3. This suggests that 6. Gallin John, S. R. 1999. Inflammation Basic Principles and Clinical Correlates. Lippincott Williams and Wilkins, Philadelphia. the preservation of endothelial barrier function by ␤-glucan treat- 7. Lu, H., C. W. Smith, J. Perrard, D. Bullard, L. Tang, S. B. Shappell, ment of activated neutrophils may not involve alternate receptors. M. L. Entman, A. L. Beaudet, and C. M. Ballantyne. 1997. LFA-1 is sufficient in mediating neutrophil emigration in Mac-1-deficient mice. J. Clin. Invest. Various PMN-derived soluble factors have been implicated to 99:1340. cause changes in vascular monolayer permeability including aden- 8. Coxon, A., P. Rieu, F. J. Barkalow, S. Askari, A. H. Sharpe, U. H. von Andrian, ␤ osine, oxidants, lysosomal enzymes, and cationic proteins such as M. A. Arnaout, and T. N. Mayadas. 1996. A novel role for the 2 integrin ␤ CD11b/CD18 in neutrophil apoptosis: a homeostatic mechanism in inflammation. CAP37 (47, 48). We have identified the mechanism of the -glu- Immunity 5:653. can protection of the endothelium to be secondary neither to inhi- 9. Diamond, M. S., J. Garcia-Aguilar, J. K. Bickford, A. L. Corbi, and bition of adhesion nor to a paracrine event. Activated PMN super- T. A. Springer. 1993. The I domain is a major recognition site on the leukocyte integrin Mac-1 (CD11b/CD18) for four distinct adhesion ligands. J. Cell Biol. natants, treated with ␤-glucan or not, added onto endothelial 120:1031. monolayers did not cause endothelial barrier dysfunction (Fig. 1b), 10. Diamond, M. S., and T. A. Springer. 1993. A subpopulation of Mac-1 (CD11b/ indicating that direct interaction of neutrophils with the endothelial CD18) molecules mediates neutrophil adhesion to ICAM-1 and fibrinogen. J. Cell Biol. 120:545. cells is necessary for loss of barrier function. Levels of CAP37 11. Miller, J. L., and A. Castella. 1982. Platelet-type von Willebrand’s disease: char- (heparin-binding protein) known to cause increased vascular acterization of a new bleeding disorder. Blood 60:790. 12. Walzog, B., D. Schuppan, C. Heimpel, A. Hafezi-Moghadam, P. Gaehtgens, and monolayer permeability were the same among fMLP-activated K. Ley. 1995. The leukocyte integrin Mac-1 (CD11b/CD18) contributes to bind- neutrophils that were treated with or without ␤-glucan (data not ing of human granulocytes to collagen. Exp. Cell Res. 218:28. shown). Following activation, neutrophils release 5Ј-AMP that can 13. Nathan, C., S. Srimal, C. Farber, E. Sanchez, L. Kabbash, A. Asch, J. Gailit, and Ј S. D. Wright. 1989. Cytokine-induced respiratory burst of human neutrophils: be metabolized to adenosine by 5 -ectonucleotidase (CD73) ex- dependence on extracellular matrix proteins and CD11/CD18 integrins. J. Cell pressed on the endothelial cell surface resulting in increased bar- Biol. 109:1341. rier function in a paracrine loop (49). In contrast, our studies did 14. Schober, J. M., N. Chen, T. M. Grzeszkiewicz, I. Jovanovic, E. E. Emeson, T. P. Ugarova, R. D. Ye, L. F. Lau, and S. C. Lam. 2002. Identification of integrin ␣ ␤ not detect a gain in barrier function when supernatant was collect M 2 as an adhesion receptor on peripheral blood monocytes for Cyr61 (CCN1) from fMLP neutrophils either in the presence or absence of ␤-glu- and connective tissue growth factor (CCN2): immediate-early gene products expressed in atherosclerotic lesions. Blood 99:4457. can. An explanation for this apparent discrepancy may lie in a 15. Yakubenko, V. P., V. K. Lishko, S. C. Lam, and T. P. Ugarova. 2002. A mo- ␣ ␤ technical difference between the experimental protocols. Lennon et lecular basis for integrin M 2 ligand binding promiscuity. J. Biol. Chem. al. (50) used supernatants from 108 fMLP-treated PMNs/ml to see 277:48635. Ј 16. Russell, D. G., and S. D. Wright. 1988. Complement receptor type 3 (CR3) binds an increase in barrier function due to 5 -AMP derived from the to an Arg-Gly-Asp-containing region of the major surface glycoprotein, gp63, of activated PMNs. This affect was no longer evident when superna- Leishmania promastigotes. J. Exp. Med. 168:279. The Journal of Immunology 1291

17. Relman, D., E. Tuomanen, S. Falkow, D. T. Golenbock, K. Saukkonen, and molecule-1 inhibit chemoattractant-stimulated neutrophil transendothelial migra- S. D. Wright. 1990. Recognition of a bacterial adhesion by an integrin: macro- tion in vitro. Blood 78:2089. ␣ ␤ phage CR3 ( M 2, CD11b/CD18) binds filamentous hemagglutinin of Bordetella 35. Issekutz, A. C., D. Rowter, and T. A. Springer. 1999. Role of ICAM-1 and pertussis. Cell 61:1375. ICAM-2 and alternate CD11/CD18 ligands in neutrophil transendothelial migra- 18. Forsyth, C. B., and H. L. Mathews. 1996. Lymphocytes utilize CD11b/CD18 for tion. J. Leukocyte Biol. 65:117. adhesion to Candida albicans. Cell. Immunol. 170:91. 36. Smith, C. W., S. D. Marlin, R. Rothlein, C. Toman, and D. C. Anderson. 1989. 19. Ross, G. D., J. A. Cain, B. L. Myones, S. L. Newman, and P. J. Lachmann. 1987. Cooperative interactions of LFA-1 and Mac-1 with intercellular adhesion mole- Specificity of membrane complement receptor type three (CR3) for ␤-glucans. cule-1 in facilitating adherence and transendothelial migration of human neutro- Complement 4:61. phils in vitro. J. Clin. Invest. 83:2008. 20. Ross, G. D., V. Vetvicka, J. Yan, Y. Xia, and J. Vetvickova. 1999. Therapeutic 37. Shaw, S. K., P. S. Bamba, B. N. Perkins, and F. W. Luscinskas. 2001. Real-time intervention with complement and ␤-glucan in cancer. Immunopharmacology imaging of vascular endothelial-cadherin during leukocyte transmigration across 42:61. endothelium. J. Immunol. 167:2323. 21. Sitrin, R. G., R. F. Todd, 3rd, H. R. Petty, T. G. Brock, E. Albrecht, and 38. Ross, G. D. 2000. Regulation of the adhesion versus cytotoxic functions of the ␣ ␤ M. R. Gyetko. 1996. The urokinase receptor (CD87) facilitates CD11b/CD18- Mac-1/CR3/ M 2-integrin glycoprotein. Crit. Rev. Immunol. 20:197. mediated adhesion of human monocytes. J. Clin. Invest. 97:1942. 39. Vetvicka, V., B. P. Thornton, and G. D. Ross. 1996. Soluble ␤-glucan polysac- 22. Harler, M. B., E. Wakshull, E. J. Filardo, J. E. Albina, and J. S. Reichner. 1999. charide binding to the lectin site of neutrophil or natural killer cell complement ␤ ␤ Promotion of neutrophil chemotaxis through differential regulation of 1 and 2 receptor type 3 (CD11b/CD18) generates a primed state of the receptor capable integrins. J. Immunol. 162:6792. of mediating cytotoxicity of iC3b-opsonized target cells. J. Clin. Invest. 98:50. 23. Vetvicka, V., B. P. Thornton, T. J. Wieman, and G. D. Ross. 1997. Targeting of 40. Xia, Y., G. Borland, J. Huang, I. F. Mizukami, H. R. Petty, R. F. Todd, 3rd, and natural killer cells to mammary carcinoma via naturally occurring tumor cell- G. D. Ross. 2002. Function of the lectin domain of Mac-1/complement receptor bound iC3b and ␤-glucan-primed CR3 (CD11b/CD18). J. Immunol. 159:599. type 3 (CD11b/CD18) in regulating neutrophil adhesion. J. Immunol. 169:6417. 24. Wegener, J., C. R. Keese, and I. Giaever. 2000. Electric cell-substrate impedance 41. Brown, G. D., and S. Gordon. 2001. Immune recognition: a new receptor for sensing (ECIS) as a noninvasive means to monitor the kinetics of cell spreading ␤-glucans. Nature 413:36. to artificial surfaces. Exp. Cell Res. 259:158. 42. Czop, J. K., and J. Kay. 1991. Isolation and characterization of ␤-glucan recep- 25. Keese, C. R., K. Bhawe, J. Wegener, and I. Giaever. 2002. Real-time impedance tors on human mononuclear phagocytes. J. Exp. Med. 173:1511. assay to follow the invasive activities of metastatic cells in culture. BioTech- 43. Zimmerman, J. W., J. Lindermuth, P. A. Fish, G. P. Palace, T. T. Stevenson, and Downloaded from niques 33:842. D. E. DeMong. 1998. A novel carbohydrate-glycosphingolipid interaction be- ␣ ␤ ␤ 26. Taooka, Y., J. Chen, T. Yednock, and D. Sheppard. 1999. The integrin 9 1 tween a -(1Ð3)-glucan immunomodulator, PGG-glucan, and lactosylceramide of mediates adhesion to activated endothelial cells and transendothelial neutrophil human leukocytes. J. Biol. Chem. 273:22014. migration through interaction with vascular cell adhesion molecule-1. J. Cell 44. Brown, G. D., P. R. Taylor, D. M. Reid, J. A. Willment, D. L. Williams, Biol. 145:413. L. Martinez-Pomares, S. Y. Wong, and S. Gordon. 2002. Dectin-1 is a major ␤ ␤ 27. Gautam, N., H. Herwald, P. Hedqvist, and L. Lindbom. 2000. Signaling via 2 -glucan receptor on macrophages. J. Exp. Med. 196:407. integrins triggers neutrophil-dependent alteration in endothelial barrier function. 45. Willment, J. A., S. Gordon, and G. D. Brown. 2001. Characterization of the ␤ J. Exp. Med. 191:1829. human -glucan receptor and its alternatively spliced isoforms. J. Biol. Chem. http://www.jimmunol.org/ 28. Gautam, N., P. Hedqvist, and L. Lindbom. 1998. Kinetics of leukocyte-induced 276:43818. changes in endothelial barrier function. Br. J. Pharmacol. 125:1109. 46. Taylor, P. R., G. D. Brown, D. M. Reid, J. A. Willment, L. Martinez-Pomares, 29. Wary K., D. M., A. Mariotti, and F. Giancotti. 1999. Integrin Protocols. Humana S. Gordon, and S. Y. Wong. 2002. The ␤-glucan receptor, dectin-1, is predom- Press, Totowa. inantly expressed on the surface of cells of the monocyte/macrophage and neu- 30. Ross, G. D., R. A. Thompson, M. J. Walport, T. A. Springer, J. V. Watson, trophil lineages. J. Immunol. 169:3876. R. H. Ward, J. Lida, S. L. Newman, R. A. Harrison, and P. J. Lachmann. 1985. 47. Gautam, N., A. M. Olofsson, H. Herwald, L. F. Iversen, E. Lundgren-Akerlund, Characterization of patients with an increased susceptibility to bacterial infections P. Hedqvist, K. E. Arfors, H. Flodgaard, and L. Lindbom. 2001. Heparin-binding and a genetic deficiency of leukocyte membrane complement receptor type 3 and protein (HBP/CAP37): a missing link in neutrophil-evoked alteration of vascular the related membrane antigen LFA-1. Blood. 66:882. permeability. Nat. Med. 7:1123. 31. Lowe, E. P., D. Wei, P. J. Rice, C. Li, J. Kalbfleisch, I. W. Browder, and 48. Dallegri, F., and L. Ottonello. 1997. Tissue injury in neutrophilic inflammation. D. L. Williams. 2002. Human vascular endothelial cells express pattern recog- Inflamm. Res. 46:382.

nition receptors for fungal glucans which stimulates nuclear factor ␬B activation 49. Narravula, S., P. F. Lennon, B. U. Mueller, and S. P. Colgan. 2000. Regulation by guest on September 29, 2021 and interleukin 8 production. Winner of the Best Paper Award from the Gold of endothelial CD73 by adenosine: paracrine pathway for enhanced endothelial Medal Forum. Am. Surg. 68:508. barrier function. J. Immunol. 165:5262. 32. Sans, E., E. Delachanal, and A. Duperray. 2001. Analysis of the roles of ICAM-1 50. Lennon, P. F., C. T. Taylor, G. L. Stahl, and S. P. Colgan. 1998. Neutrophil- in neutrophil transmigration using a reconstituted mammalian cell expression derived 5Ј-adenosine monophosphate promotes endothelial barrier function via model: implication of ICAM-1 cytoplasmic domain and Rho-dependent signaling CD73-mediated conversion to adenosine and endothelial A2B receptor activa- pathway. J. Immunol. 166:544. tion. J. Exp. Med. 188:1433. 33. Zen, K., Y. Liu, D. Cairo, and C. A. Parkos. 2002. CD11b/CD18-dependent 51. Babineau, T. J., A. Hackford, A. Kenler, B. Bistrian, R. A. Forse, P. G. Fairchild, interactions of neutrophils with intestinal epithelium are mediated by fucosylated S. Heard, M. Keroack, P. Caushaj, and P. Benotti. 1994. A phase II multicenter, proteoglycans. J. Immunol. 169:5270. double-blind, randomized, placebo-controlled study of three dosages of an im- 34. Furie, M. B., M. C. Tancinco, and C. W. Smith. 1991. Monoclonal antibodies to munomodulator (PGG-glucan) in high-risk surgical patients. Arch. Surg. leukocyte integrins CD11a/CD18 and CD11b/CD18 or intercellular adhesion 129:1204.