Dynamic Regulation of LFA-1 Activation and Neutrophil Arrest on Intercellular Adhesion Molecule 1 (ICAM-1) in Shear Flow*

Neutrophil recruitment during acute inflammation is triggered by G-protein-linked chemotactic receptors that in turn activate β2 integrin (CD18), deemed a critical step in facilitating cell capture and arrest under the shear force of blood flow. A conformational switch in the I domain allosteric site (IDAS) and in CD18 regulates LFA-1 affinity for endothelial ligands including intercellular adhesion molecule 1 (ICAM-1). We examined the dynamics of CD18 activation in terms of the efficiency of neutrophil capture of ICAM-1, and we correlated this with the membrane topography of 327C, an antibody that recognizes the active conformation of CD18 I-like domain. Adhesion increased in direct proportion to chemotactic stimulus rising 7-fold over a log range of interleukin-8 (IL-8). A threshold dose of ∼75 pm IL-8, corresponding to ligation of only ∼10–100 receptors, was sufficient to activate ∼20,000 CD18 and a rapid boost in the capture efficiency on ICAM-1. This was accompanied by a rapid redistribution of active LFA-1, but not Mac-1, into membrane patches, a necessary component for optimum adhesion efficiency. Shear-resistant arrest on a monolayer of ICAM-1 was reversed within minutes of chemotactic stimulation correlating with a shift from high to low affinity CD18 and dispersal of patches of active CD18. Mobility of active CD18 into high avidity patches was dependent on phosphatidylinositol 3-kinase activity and not F-actin polymerization. The data reveal that the number of chemotactic receptors bound and the topography and lifetime of high affinity LFA-1 tightly regulate the efficiency of neutrophil capture on ICAM-1.

Neutrophil recruitment during acute inflammation is triggered by G-protein-linked chemotactic receptors that in turn activate ␤ 2 integrin (CD18), deemed a critical step in facilitating cell capture and arrest under the shear force of blood flow. A conformational switch in the I domain allosteric site (IDAS) and in CD18 regulates LFA-1 affinity for endothelial ligands including intercellular adhesion molecule 1 (ICAM-1). We examined the dynamics of CD18 activation in terms of the efficiency of neutrophil capture of ICAM-1, and we correlated this with the membrane topography of 327C, an antibody that recognizes the active conformation of CD18 I-like domain. Adhesion increased in direct proportion to chemotactic stimulus rising 7-fold over a log range of interleukin-8 (IL-8). A threshold dose of ϳ75 pM IL-8, corresponding to ligation of only ϳ10 -100 receptors, was sufficient to activate ϳ20,000 CD18 and a rapid boost in the capture efficiency on ICAM-1. This was accompanied by a rapid redistribution of active LFA-1, but not Mac-1, into membrane patches, a necessary component for optimum adhesion efficiency. Shear-resistant arrest on a monolayer of ICAM-1 was reversed within minutes of chemotactic stimulation correlating with a shift from high to low affinity CD18 and dispersal of patches of active CD18. Mobility of active CD18 into high avidity patches was dependent on phosphatidylinositol 3-kinase activity and not F-actin polymerization. The data reveal that the number of chemotactic receptors bound and the topography and lifetime of high affinity LFA-1 tightly regulate the efficiency of neutrophil capture on ICAM-1.
Neutrophils are among the first cells to respond to acute inflammation through a multistep process of specific bond formation with adhesion molecules up-regulated on the surface of activated endothelium (1,2). Endothelial and leukocyte selectins function to capture leukocytes from the circulation by rapid bond formation. However, selectins form transient bonds that last a second or less in shear flow and only mediate cell capture and rolling on endothelium (3,4). Unlike selectins, ␤ 2 integrins (CD18) do not constitutively recognize ligand but require cellular activation to form stable shear-resistant bonds with endothelial ligands including ICAM-1. 1 Signaling of inte-grin activation from the "inside-out" of leukocytes is typically mediated through ligation of G-protein-coupled receptors (GPCR) such as the CXC chemokine receptors (CXCR1 and CXCR2) that bind interleukin-8 (IL-8) (5)(6)(7). Chemokine binding to GPCR initiates a cascade of pro-inflammatory cell responses. One of the earliest responses is intracellular calcium flux that signals cell shape change, degranulation, integrin activation, adhesion, and cell motility (8 -11). This cascade of cell responses that hinge on activation of CD18 is critical to neutrophil arrest on the endothelium as evidenced by leukocyte adhesion deficiency (LAD) patients who manifest a diminished or absent immune response at sites of inflammation (12,13).
Recently, it was shown (14,15) that leukocyte stimulation elicits a conformational shift in LFA-1 that correlates with an increase in affinity of binding to ICAM-1. An increase in avidity of LFA-1 on lymphocytes has been shown to support cell rolling on a cellular substrate of ICAM-1 in the shear field of a parallel plate flow chamber (16). Published data also indicate that firm adhesion of neutrophils to ICAM-1 involves a cooperative and sequential process of LFA-1 (CD11a/CD18)-dependent capture followed by Mac-1 (CD11b/CD18)-mediated stabilization (17). This occurs within seconds of chemotactic stimulation of cell suspensions sheared in a cone-plate viscometer over a range of stress typical of blood flow in the microcirculation (ϳ0.2-1.5 dynes/cm 2 ) (17). Studies in the microcirculation have corroborated the flow chamber data revealing that LFA-1 is sufficient for the transition from leukocyte rolling to arrest in the inflamed microcirculation of the mouse (18,19).
Leukocyte stimulation by chemokines triggers a high affinity conformation and an increase in the lateral mobility of CD18, resulting in formation of receptor clustering on the membrane (20). Both of these mechanisms can elicit the arrest of circulating lymphocytes (20). LFA-1 occurs at the I domain metal ion-dependent adhesion site (MIDAS) (21). An allosteric shift in LFA-1 to a conformation that supports high affinity ligand binding was recently shown to localize to a site spatially distinct from the MIDAS, which is denoted the IDAS (15,22). This allosteric shift in the LFA-1 heterodimer resulted in a 6-fold increase in the K d of binding of soluble LFA-1 to ICAM-1 (15). Binding of a unique CD18 mAb 240Q was also found to induce a high affinity conformation of LFA-1. A CD18 I-like domain (␤I) mAb, 327C, was shown to report on the active conformation induced by soluble ICAM-1, IDAS activating mutation or 240Q (14,15,22). The induction of the ␤I domain 327C epitope was blocked by a small molecule diarylsulfied inhibitor that binds to the IDAS (14). Thus the IDAS conformation switch is conformationally linked to CD18 ␤I domain conformation. In addition, regulation at the IDAS is dominant to cellular activation of LFA-1 (15). Taken together the data support a mechanism by which cellular activation of an LFA-1 high affinity state that supports cell adhesion is regulated by a conformational switch(s) at the IDAS and ␤I domain (22,23). Despite a wealth of evidence indicating that conformational activation of ␤ 2 integrin plays a central role in neutrophil recruitment during acute inflammation, the precise mechanisms that lead from GPCR signaling to shear-resistant arrest are not entirely known. For example, the relations between IL-8 receptor ligation, the rate of activation of CD18, and the dynamics of neutrophil adhesion have not been systematically studied. Yet, formation of bonds in sufficient numbers may provide a critical gatekeeper in regulating the efficiency of the transition from cell rolling to shear-resistant arrest, a prerequisite for transendothelial migration (24).
In the current study, we examined neutrophil capture and firm adhesion to ICAM-1 via LFA-1 and Mac-1 in response to stimulation with IL-8 and allosteric activation by mAb 240Q. To examine the process of collisional capture of neutrophils, a flow cytometric based assay was employed to detect adhesion of fluorescent microbeads coated with ICAM-1. The strength and stability of neutrophil adhesion to an ICAM-1-expressing cellular substrate was imaged in a parallel plate flow chamber as a function of applied shear stress. By correlating the dynamics of adhesion with the expression and membrane distribution of mAb 327C reporting on activated CD18, we show that IL-8 stimulation regulates neutrophil recruitment on ICAM-1 through changes in LFA-1 affinity and membrane mobility.

EXPERIMENTAL PROCEDURES
Isolation of Human Neutrophils and PBMC-Whole blood was obtained from healthy adult individuals by venipuncture into sterile syringes with heparin (10 units/ml of blood, Elkins-Sinn, Inc., Cherry Hill, NJ) using the University of California, Davis, approved human subjects protocol (Protocol Identification 993120). Neutrophils were isolated from whole blood using a one-step Ficoll-Paque density gradient (Mono-Poly Resolving medium, ICN, Aurora, OH, or PMN, Robbins Scientific Corp., Sunnyvale, CA) following the manufacturers' protocols. Peripheral blood mononuclear cells (PBMC) were isolated from whole blood using a one-step sodium metrizoate/Ficoll density gradient (Lympho Sep, ICN Biomedicals, Aurora, OH) following the manufacturer's protocol with the following modification. Blood was diluted 2-fold with PBS prior to layering on the separation media. Following isolation, neutrophils were washed once with HEPES buffer (10 mM KCl, 110 mM NaCl, 10 mM glucose, 1 mM MgCl 2 , and 30 mM HEPES, pH 7.4), and PBMC were washed 3 times with PBS by centrifugation at 300 ϫ g for 8 min. The PBMC or neutrophils were resuspended in HEPES buffer (ϳ2.5 ϫ 10 7 cells/ml). Endotoxin was removed from the buffer by affinity chromatography over polymyxin B-Sepharose (Pierce). Neutrophils and PBMC were maintained at room temperature in a calcium-free HEPES buffer. Neutrophils were assessed by trypan blue exclusion and deemed to be Ͼ98% viable. Following cell isolation, neutrophils were found to remain unactivated at room temperature for ϳ4 h after separation. The criteria was that incubation for 20 min at 37°C yielded less than 10% of Mac-1 up-regulation.
Parallel Plate Flow Chamber Assay-Neutrophil adhesion to ICAM-1 was measured in a parallel plate flow chamber that produces a uniform laminar flow field as described previously (29). Briefly, L cells expressing ICAM-1 were grown to confluence on 35 mm untreated glass cover slips, washed twice with Dulbecco's PBS containing 11 mM glucose, and mounted on the flow chamber as described previously (25). Neutrophils (1.5 ϫ 10 6 ) were incubated for 10 min at room temperature with 240Q at indicated concentrations, incubated for an additional 2 min at 37°C, then stimulated with IL-8 at indicated concentrations, and injected as a bolus into the flow chamber. Neutrophils were allowed to sediment to the monolayer in the absence of shear stress. Shear stress was initially set at 0.1 dyne/cm 2 for an additional minute to promote interactions with the monolayer substrate and then ramped to 4 dynes/cm 2 to assess the strength and stability of adhesion. Image sequences were captured (in a phase separated field of view) for 30 s (4 frames/s) throughout the period of applied shear using a Nikon TE200 inverted microscope equipped with a Plan Fluor 20ϫ objective and a cooled CCD camera (Dage-MTI, Michigan City, IN). Images were compiled directly into digital sequences using a frame grabber (Scion Corp., Frederick, MD) and then analyzed using Image Pro Plus 4.1 software (Media Cybernetics, Silver Spring, MD). The number of adherent neutrophils per field of view was quantitated by identifying their phase-bright appearance in the same focal plane as the monolayer (Fig. 1a). Adherent neutrophils were defined as those neutrophils that moved less than one cell diameter in 30 s. This fraction of arrested cells was normalized to the average of the total number of neutrophils, both adherent and rolling, present in the field of views of the first two time points (0.1 dyne/cm 2 time samples).
Preparation of Latex Beads Presenting ICAM-1 on Their Surface-To provide a quantitative measure of CD18-dependent capture of ICAM-1, fluorescent latex beads (1 m diameter, Fluospheres, Molecular Probes, Eugene, OR) coated with protein A were derivatized with recombinant human ICAM-1/Fc. Beads were washed three times with Dulbecco's PBS without calcium and magnesium (Invitrogen), resuspended in HEPES buffer, and incubated in a sonic bath for 1 h with a saturating concentration (15 g/ml) of ICAM-1/Fc to allow for both dispersion and coating of beads. In order to block nonspecific adhesion, blocked (Molecular Probes, Eugene, OR) was added to the bead solution (20% by volume) for 30 min with sonication. The density of ICAM-1/Fc on the surface of the beads has been determined previously (30) to contain ϳ25 binding sites/m 2 . Beads were added to the samples at 2 ϫ 10 7 beads/ml (20 beads per neutrophil).
Flow Cytometric Detection of ICAM-1-coated Latex Beads-To examine specific adhesion via the ␤ 2 integrins, neutrophils were mixed with fluorescent beads with ICAM-1/Fc derivatized to their surface. Sample volumes of 250 l (HEPES buffer, 1.5 mM CaCl 2 ) contained 1 ϫ 10 6 neutrophils/ml, 2 ϫ 10 7 beads/ml, LDS-751 (8 ng/ml, Molecular Probes, Eugene, OR) to label neutrophils, and a small magnetic stir bar. Adhesion blocking antibodies and/or 240Q were preincubated for 10 min with the cell suspension without beads. For samples stimulated with IL-8, stimulus was added just after the beads, and the suspension was then placed immediately on the sample injection port of the FACScan flow cytometer (BD PharMingen). Samples were maintained at 37°C within a mixing chamber with a magnetic motor as described previously (31). The magnetic motor coupled with the magnetic stir bar created a shear field (shear stress ϳ1.0 dyne/cm 2 ) within the test tube and initiated collisional interactions of relatively brief contact duration (ϳ25 ms). Neutrophil capture of ICAM-1-coated beads was continuously monitored at low flow. Neutrophils were distinguished by their characteristic forward and right angle light scatter properties and gated in order to exclude unbound beads. Neutrophil-bead adhesion was quantitated by green bead fluorescence on fluorescence histograms. Quantal increases in fluorescence appeared as peaks in the fluorescence histogram corresponding to populations of neutrophils binding increasing numbers of beads (31). To distinguish relative levels of bead capture within the stimulated neutrophil population, neutrophil-bead interactions were quantitated as the average number of beads per neutrophil according to Equation 1, where N represents the number of nonadherent neutrophils, and NB i represents the number of neutrophil-bead aggregates bound to between 1 and 6 or more beads.
Activation Epitope Recognized by mAb 327C-mAb 327C has been reported previously (15) to recognize the active conformation of CD18 associated with LFA-1 function in phorbol ester or T cell receptor cross-link-stimulated primary human lymphocytes. To determine the kinetics of CD18 activation, neutrophils and PBMC were mixed with 327C-Alexa and examined for 10 min with the flow cytometer under conditions of stimulation by soluble ICAM-1/Fc (150 g/ml), 240Q, or IL-8 using identical stimulation protocols as the bead adhesion assay. Additional experiments examining the transience of active CD18 expression after IL-8 stimulation were conducted where 327C-Alexa was added to unfixed samples at indicated times after stimulation by 1 nM IL-8. All samples were washed, resuspended, and analyzed on the FACScan flow cytometer. Mean fluorescence intensity (MFI) was determined for the neutrophil population at indicated time points. Receptor expression in terms of sites/cell was determined by comparing the MFI of bound 327C to Quantum Simply Cellular beads (FCSC, San Juan, Puerto Rico) which contain five bead sets with increasing numbers of goat anti-mouse-binding sites on their surface. From this analysis, a linear relation between MFI and receptor expression was determined for each directly conjugated antibody bound to cells and the calibration bead set.
Immunofluorescence-To observe colocalization of either LFA-1 or Mac-1 with 327C, neutrophils (1 ϫ 10 6 /ml) were stimulated with IL-8 or 240Q and incubated at room temperature for 2 min with biotinylated 327C (10 g/ml) and either anti-LFA-1-PE or anti-Mac-1 (2LPM19c-RPE) according to manufacturer's suggested staining procedure. Cells were fixed in 1% paraformaldehyde, and excess antibody was removed by centrifugation. Following this wash, cells were incubated with neutravidin-FITC for 20 min at room temperature. Cells were centrifuged following staining and resuspended in ice-cold HEPES buffer with 0.1% HSA followed by fixation with an equal volume of freshly prepared 2% paraformaldehyde to a final concentration of 1%. To determine the expression of active CD18 as a function of the time after IL-8 stimulation, neutrophils (1 ϫ 10 6 /ml) were stimulated at time 0 with 1 nM IL-8. Biotinylated 327C (10 g/ml) was added 2 min prior to each time point. At indicated times, 10 5 cells were removed and immediately fixed in HEPES buffer containing 1% paraformaldehyde and 0.1% HSA. Following fixation, neutrophils were incubated with neutravidin-FITC for 20 min at room temperature, washed, and resuspended in 1% paraformaldehyde. The labeled cells were imaged using a Nikon TE200 inverted microscope employing an oil immersion 60ϫ Plan-Apo objective and bandpass filters appropriate for FITC or PE labels.
Intracellular Calcium Flux-Neutrophil intercellular calcium flux was measured using Fluo-3-AM (Molecular Probes, Eugene, OR). Neutrophils (2 ϫ 10 6 cells) were added to a 1 M Fluo-3-AM solution in HEPES buffer with 0.1% Me 2 SO with no calcium or HSA and incubated at 37°C for 45 min in the dark to allow for dye loading. Neutrophils were washed and resuspended in indicator-free HEPES with calcium and HSA and then incubated at room temperature for 30 min in the dark. Neutrophils stimulated with 240Q or R15.7 were allowed to incubate with the noted concentrations of mAb for an additional 10 min at room temperature. Samples were incubated for 2 min at 37°C, maintained at 37°C with a water jacket, and read on the FACScan flow cytometer. After ϳ1 min of reading, the sample tube was removed from the FACScan; stimulant was added at the indicated concentrations (IL-8, 240Q, R15.7, or no additional stimulation), and the sample tube was replaced on the sample injection port of the FACScan and read for an additional ϳ3 min. The activated neutrophil population was gated according to the increase in Fluo-3 fluorescence (FL-1) due to intracellular calcium flux that corresponded to two standard deviations above unstimulated neutrophils.
Statistical Analysis-Data were independently collected for each experiment, and a mean and S.E. were calculated for all experiments. The Student's t test was used to determine statistical significance between mean values. Statistical analysis was performed using Prism 3 (Graph-Pad Software Inc., San Diego, CA). Values were considered significant when p Ͻ 0.05.

Neutrophil Arrest on ICAM-1 in a Parallel Plate Flow
Chamber-Neutrophils overcome the repulsive shear force of blood flow during the process of arrest on inflamed endothelium through the formation of sufficient numbers of CD18 bonds to ICAM-1. We directly observed the dynamics of neutrophil adhesion to a monolayer of cells expressing ICAM-1 in a parallel plate flow chamber at defined shear stress, and we quantitated these dynamics over the time course of activation with IL-8 and mAb 240Q. Neutrophil suspensions were stimulated and then injected into the flow chamber and allowed to settle onto the monolayer substrate at low shear for 1 min. Phase contrast microscopy of neutrophils interacting with the ICAM-1-expressing monolayer at low shear enabled clear delineation of those interacting that remained arrested for at least 30 s (Fig.  1a). Analysis of neutrophil capture efficiency, defined as the transition from collisional interactions such as cell rolling or tumbling motion to shear-resistant arrest, was limited to relatively few cells in a single field of view following injection in the flow chamber over the low shear interval (Fig. 1b). Ramping up the shear stress for the duration of the experiment allowed enumeration of a greater number of neutrophils that remained arrested as a function of time as measured on five separate fields of view. A significant increase in capture from the base-line level required stimulation above a threshold dose of IL-8 (ϳ0.1 nM) (Fig. 1b). Neutrophil arrest increased with both dose of stimulus and time at low shear, reaching a peak of ϳ75% of interacting neutrophils at 1 nM IL-8 stimulation, which did not increase at a 5 nM dose (data not shown). At the high shear stress a steady drop in the fraction of neutrophils remaining arrested was observed beyond the 1st min of stimulation over the dose range of IL-8. Even at a saturating dose of 5 nM IL-8, cell arrest approached base line by 5 min of flow.
In contrast to IL-8 stimulation, direct activation of ␤ 2 integrindependent adhesion through binding of mAb 240Q was less efficient at eliciting initial capture over a range of binding up to receptor saturation at 5 g/ml (data not shown). However, activation of just 5% of surface-expressed CD18, corresponding to 0.125 g/ml of 240Q bound to the IDAS, was sufficient to trigger adhesion at low flow and effected a steady increase in neutrophil arrest that was sustained for up to 6 min of flow at the high shear stress. Interestingly, higher doses of mAb 240Q up to saturation of binding to CD18 (5 g/ml) elicited neither a higher initial capture rate nor a greater extent of adhesion at 6 min than observed with 0.125 g/ml of 240Q (data not shown). We next examined neutrophil recruitment in response to costimulation with a low dose of IL-8 (50 pM) and 240Q (0.125 g/ml). Costimulation did not significantly boost the initial capture of cells under low flow conditions, as compared with either 240Q or IL-8 alone. However, the number of neutrophils remaining arrested on ICAM-1 after 5 min at high flow was significantly increased above that induced by 240Q at any dose (Fig. 1c). These flow chamber studies reveal an increase in the arrested fraction of neutrophils with increasing dose of IL-8, above a threshold required to initiate capture. Neutrophil arrest signaled by chemotactic factor was transient. However, stabilizing as few as 5% of surface CD18 in an active conformation sustained shear-resistant adhesion in virtually the entire cell population.
Dynamics of Neutrophil Capture of ICAM-1 under Fluid Shear-To define further the relationship between the dynamics of CD18 activation and the efficiency of ICAM-1 capture, we analyzed the collisional interactions between neutrophils and ICAM-1-coated fluorescent latex beads in real time by flow cytometry. Beads were coated with ICAM-1 at a site density commensurate with that expressed on inflamed endothelium (ϳ25 sites/m 2 ) and mixed together with neutrophils in a cytometry tube at physiological shear stress (ϳ1 dyne/cm 2 ). Under these conditions each neutrophil experience on the order of 10 collisions/s with beads just upstream of the flow cytometer providing continuous analysis at rates of up to 200 cells/s. Over the 1st min of IL-8 stimulation the rate of capture increased by 20-fold from the base-line adhesion of 3 beads per 100 neutrophils in unstimulated cells and up to 60 beads per 100 neutrophils at maximum response (Fig. 2a). In the absence of stimulation, this base-line adhesion remained constant over 10 min of shear. Stimulation with IL-8 (1 nM) elicited bead capture that increased by 4-fold over the 1st min, culminating in a 10-fold rise in the average number of ICAM-1 beads firmly adherent to neutrophils within 10 min. Direct activation of CD18 by addition of mAb 240Q also elicited capture of ICAM-1 beads but at one-half the rate induced by IL-8 stimulation.
We next examined the relative contributions of LFA-1 and Mac-1 to the capture of ICAM-1 beads in response to either chemotactic or allosteric activation. Treatment of neutrophils with anti-LFA-1 abrogated the initial rapid boost in ICAM-1 capture triggered by IL-8, while anti-Mac-1 inhibited the accumulation of beads only beyond 4 min of stimulation (Fig. 2b). Blocking with anti-CD18 completely abrogated bead adhesion. Together, the data indicate that LFA-1 is sufficient for capture of ICAM-1, whereas Mac-1 accounts for ϳ30% of ICAM-1 bead adhesion at later time points of IL-8 stimulation.
In contrast to GPCR signaling of adhesion, allosteric activa- Neutrophil arrest on ICAM-1 in a parallel plate flow chamber. Recruitment of activated neutrophils to an ICAM-1-expressing L-cell monolayer was assessed in the parallel plate flow chamber under defined shear stress. Neutrophils in suspension (10 6 /ml) were activated with either IL-8 or mAb 240Q and introduced as a bolus into the flow chamber. Following injection, neutrophils were allowed to sediment to the ICAM-1 substrate in the absence of flow. Shear stress was then ramped up to 0.1 dyne/cm 2 to facilitate interaction between neutrophils and the ICAM-1 substrate. To assess adhesion strength, shear stress was subsequently ramped up to 4 dynes/cm 2 (vertical dashed line) for an additional 5 min. a, video micrographs of representative fields of neutrophils interacting and arrested (phase white PMNs and dark L-cell monolayer) at 0.1 dyne/cm 2 (left) and 4 dynes/cm 2 (right). b, time course of the fraction of neutrophil arrested as a function of the dose of IL-8 or 240Q. Plotted on the ordinate is the ratio of arrested neutrophils, defined as those moving less than one cell diameter in 30 s, to the total number interacting with the monolayer at 1 min. c, costimulation of neutrophils with IL-8 (50 pM) and 240Q (0.125 g/ml). The percent of arrested neutrophils following costimulation was evaluated after 5 min of high flow. * denotes p Ͻ 0.05 compared with 240Q alone. Data represent mean PMN arrested Ϯ S.E. from 3 to 5 separate experiments. tion of CD18 at a saturating dose of 240Q elicited capture solely supported by LFA-1. Blocking Mac-1 function with mAb did not significantly reduce ICAM-1 bead capture, as compared with 240Q alone (Fig. 2c). ICAM-1 capture was entirely accounted for by activation of LFA-1, as blocking with anti-LFA-1 abrogated 240Q-induced adhesion. We confirmed that 240Q indeed bound to the Mac-1 receptor by testing binding to cell lines expressing Mac-1 in either the active or inactive conformation (data not shown). Additional studies were performed on neutrophils to assess Mac-1 function in response to stimulation with both 240Q and IL-8. Neutrophil capture of albumin-coated latex beads in the flow cytometry assay has been shown previously (32) to be mediated entirely by activation and binding of Mac-1, and was found to be blocked in the presence of mAb 240Q binding to CD18 (data not shown). These data confirmed that activated LFA-1 is sufficient for IL-8-stimulated or 240Qinduced capture of ICAM-1.
Relation between Stimulus Dose and Rate of Capture of ICAM-1-Recent studies (17,30) have shown that neutrophil adhesion elicited by chemotactic stimulation is most rapid over the initial seconds following addition of stimulus. Therefore, we computed the rate of ICAM-1 bead capture following the 1st min of applied shear and stimulation over a dose range of IL-8 or 240Q. A direct correlation was observed between the dose of stimulus and the rate of ICAM-1 bead capture above a threshold of 75 pM IL-8 necessary to induce significant bead adhesion. The sensitivity of this assay was exhibited by a detectable increase in rate and extent of adhesion as the IL-8 dose was incremented by just 100 pM (Fig. 3a). The effective concentration to induce half-maximum capture (EC 50 ) was 250 pM IL-8, which corresponds to ϳ25% IL-8 receptor occupancy.
Activation via the ␤I domain and the conformationally linked IDAS using mAb 240Q yielded significant adhesion at 0.125 g/ml, a dose that corresponds to ϳ5% site occupancy of surface-expressed CD18 (Fig. 3b). The rate of bead capture increased in direct proportion to binding of 240Q up to 15% CD18 occupancy. A plateau in the efficiency of ICAM-1 bead capture corresponded to occupancy of only ϳ30% CD18. At this dose of 240Q, IDAS-activated capture of ICAM-1 was less than onehalf that elicited by maximal IL-8 stimulation (0.31 versus 0.64 beads/PMN/min; p Ͻ 0.05). Commensurate with the parallel plate flow chamber data, these results confirm that inside-out signaling via ligation of IL-8 is more efficient at eliciting dosedependent neutrophil capture on ICAM-1 than via allosteric activation of CD18.
We examined further the ability of IL-8 to augment the efficiency of ICAM-1 bead capture over a range of CD18 site occupancy by 240Q (Fig. 3c). At a dose of IL-8 (100 pM) just above the threshold to induce adhesion, the rate of ICAM-1 bead capture was amplified most profoundly at low CD18 site occupancy of 240Q. With only ϳ3% IDAS activation of CD18, IL-8 amplified the rate of capture by over 3-fold above 240Q alone. However, bead capture induced by 240Q doses exceeding 15% occupancy of CD18 was not augmented by addition of IL-8. Taken together, the data demonstrate that capture efficiency induced via allosteric activation of CD18 is limited, yet cellular processes initiated through chemotactic signaling significantly amplify adhesion.
Intracellular Calcium Flux Signaled via IL-8 and 240Q-In light of recent evidence (33) linking an increase in intracellular calcium with subsequent surface redistribution of LFA-1 into high avidity clusters, we measured Ca 2ϩ flux in neutrophils stimulated with IL-8 and 240Q. In particular, we examined whether costimulation would result in a potentiation of the Ca 2ϩ flux, as was observed for neutrophil capture and arrest on ICAM-1. Neutrophils were loaded with the fluorescent calcium indicator Fluo-3, and intracellular Ca 2ϩ flux was detected by flow cytometry over the time course of stimulation (Fig. 4,  inset). Activation was quantitated as the neutrophil fraction that responded to stimulus with a rapid increase in fluorescence above the base-line value in unstimulated cells (Fig. 4a). Ca 2ϩ flux was found to be the most sensitive assay to the onset of activation, as a threshold dose of just 25 pM IL-8 elicited a significant signal in ϳ10% of the neutrophil population (Fig.  4a). The EC 50 of Ca 2ϩ flux was ϳ100 pM IL-8, almost 3-fold lower than measured for ICAM-1 adhesion. Maximal Ca 2ϩ flux corresponded to 1 nM IL-8, which elicited a response in 85% of neutrophils. These data verify the role of Ca 2ϩ flux as a faithful secondary messenger in regulating neutrophil adhesion signaled through ligation of chemotactic receptor.
Allosteric activation of neutrophils with mAb 240Q, but not a control anti-CD18 (mAb R15.7), at doses that saturate receptor binding failed to induce significant Ca 2ϩ flux (Fig. 4b). In contrast, costimulation with both 240Q (5% CD18 site occupancy) and IL-8 (100 nM) resulted in a 32% potentiation above IL-8 alone in signaling Ca 2ϩ flux. This boost was not observed when IL-8 was combined with mAb R15.7, which binds to CD18 with comparable affinity to 240Q. These data suggest that allosteric activation of CD18 as induced by mAb 240Q may also  Fig. 2) over the 1st min following stimulation (c). Costimulation of neutrophils with IL-8 (100 pM) and 240Q over a range of doses expressed in terms of the site occupancy of surface CD18. Plotted is the increase in rate of capture above activation with 240Q alone. Data are presented as the mean Ϯ S.E. from 3 to 8 separate experiments, and * denotes a significant increase (p Ͻ 0.05) in capture compared with activation by IL-8 or 240Q alone. function in receptor-mediated signaling from the outside-in.
Activation of CD18 Detected by Binding of Monoclonal Antibody 327C-Lymphocyte adhesion to ICAM-1 mediated by LFA-1 was recently shown to correlate with a shift in binding affinity regulated at the IDAS (15,22). Previous studies have demonstrated that 327C faithfully reports on the active conformation of LFA-1 high affinity. In contrast to the binding of mAb24 which also reports on CD18 activation, 327C binding to LFA-1 on lymphocytes does not require ICAM-1 binding at the I domain (14,34). We first examined expression of 327C on neutrophils and PBMC in response to soluble ICAM-1/Fc (150 g/ml). Consistent with Lupher et al. (15), ICAM-1/Fc induced a 4-fold increase in 327C binding from base-line-unstimulated mononuclear cells (Fig. 5). This represented a shift to an active conformation in ϳ50% of LFA-1 expressed on PBMC. In contrast, base-line expression of 327C was higher in purified neutrophils and was not up-regulated in response to incubation with ICAM-1/Fc. It appears that neutrophils do not undergo ligand-induced activation of LFA-1, suggesting that significant differences exist in the regulation of LFA-1 function between neutrophils and PBMC.
Next we used mAb 327C to elucidate the kinetics of CD18 activation in real time following the addition of IL-8 or mAb 240Q. Binding of fluorescently conjugated 327C to neutrophil CD18 was continuously measured by flow cytometry as cells were stimulated with either IL-8 or 240Q. In the absence of stimulus, neutrophils expressed a background level of 327Cbinding sites that did not increase significantly over 10 min of shear at 37°C (Fig. 6a). Stimulation elicited a rapid up-regulation of active CD18 sites in which the rate and extent was commensurate with the dose of IL-8. Similar to the kinetics of ICAM-1 capture, expression of the CD18 activation epitope was most rapid over the 1st min of stimulation regardless of IL-8 dose. Up-regulation of 327C increased 10-fold over a dose range of IL-8, from a base-line detection of ϳ10,000 sites up to 100,000 sites/neutrophil (Fig. 6b). The dose-response curve of CD18 activation was similar in shape to that observed for adhesion function, exhibiting a threshold dose of 100 pM and an EC 50 of ϳ300 pM IL-8 (Fig. 6b).
To assess the kinetics and extent of 327C up-regulation in response to allosteric activation of CD18, neutrophils were pretreated for 10 min at room temperature with 240Q and then continuously measured on the flow cytometer at 37°C. Preincubation with 240Q induced a higher base-line expression of 327C as compared with IL-8 stimulation. Up-regulation of active CD18 with incubation was most rapid over the 1st min of incubation and increased at a rate comparable with low dose IL-8 stimulation (Fig. 6a). As with adhesion to ICAM-1, IDAS activation of CD18 exhibited a 40% narrower dynamic range than for IL-8 stimulation, increasing from a base line of 15,000 sites up to only 60,000 sites (Fig. 6c).
This comparable difference in efficacy of CD18 activation via GPCR and the IDAS also correlated with a difference in ICAM-1 capture efficiency between the two modes of stimulation. For example, the threshold dose of 240Q (5% CD18 occupancy) that induced activation of ϳ23,000 CD18 sites/PMN corresponded to a rate of ICAM-1 capture of ϳ0.1 beads/PMN/ min. By comparison this number of active CD18 was elicited by ϳ150 pM IL-8, which corresponded to greater than 0.2 beads/ PMN/min. This trend also held at higher doses of stimulus, whereas 0.5 nM IL-8 and 240Q at 1 g/ml (20% CD18 occupancy) both elicit about the same number of active sites (ϳ40,000 sites); IL-8 activated a 3-fold greater rate of ICAM-1 capture. Additionally, although costimulation at low doses of IL-8 and 240Q elicited significant increases in adhesive function, no significant increases were detected for 100 pM IL-8 and 0.03-0.125 g/ml 240Q (data not shown).
Immunofluorescence Detection of Active CD18 on the Neutrophil Surface-The above data are consistent with the hypothesis that CD18-adhesive function in neutrophils is acutely dependent on factors in addition to a shift from a low to high affinity conformation. For example, it has been demonstrated  Fc (150 g/ml). LFA-1 was detected with a directly conjugated mAb, and expression of activated CD18 was reported by mAb 327C that recognizes a neoepitope previously shown to up-regulate in response to cell stimulation (14). Mean fluorescence intensities of antibody binding was converted to receptor site numbers using calibration beads with known receptor site densities as described under "Experimental Procedures." Data are presented as the mean Ϯ S.E. from three separate experiments. * denotes significant increase (p Ͻ 0.05) in activated CD18 compared with 327C alone. that lymphocyte adhesion to ICAM-1 may be regulated by an increase in the lateral diffusion of integrins that supports the formation of high avidity clusters of receptors (33). We employed two-color immunofluorescence to examine the surface topography of mAb 327C relative to LFA-1 or Mac-1 on neutrophils following stimulation with IL-8 and/or 240Q (Fig. 7 and Table I). In the resting state LFA-1 and Mac-1 were uniformly distributed on the cell surface as revealed by the circumferential red fluorescence on unstimulated neutrophils (Fig. 7, a and b). The absence of 327C green fluorescence indicates the absence of detectable levels of activated CD18. Stimulation with IL-8 induced a redistribution of LFA-1 into punctate clusters of red fluorescence circumferentially distributed on the plasma membrane (Fig. 7c). Image analysis revealed that 75% of the LFA-1 colocalized with active CD18 as indicated by clusters and larger patches of 327C-associated yellow fluorescence (Table I). Mac-1 expression increased with IL-8 stimulation but remained uniformly distributed, and only ϳ35% colocalized with clusters of 327C ( Fig. 7d and Table I). Activation of CD18 in the presence of 240Q again resulted in colocalization of membrane LFA-1 with 327C defined active CD18 (Fig. 7e). In contrast to IL-8 stimulation that elicited a wide distribution of active sites, 240Q induced tight clusters of yellow fluorescence recruiting a smaller fraction of activated LFA-1 (Table I). Surface distribution of Mac-1 sites was not altered in the presence of 240Q (Fig. 7f).
Kinetic Analysis of CD18 Topography following Activation by IL-8 -The dynamics of CD18 activation and surface distribution were examined by correlating 327C topography by immunofluorescence with the time course of expression by flow cytometry. 327C expression increased most rapidly over the initial 30 s of IL-8 addition, increasing from a base line of ϳ10,000 sites up to 70,000 sites/PMN (Fig. 8a). This peak in expression was maintained over the first minutes of stimulation before dropping back toward base line by 20 min, clearly revealing a shift in CD18 conformation back to the resting state. This time course of 327C expression would appear to be at odds with the kinetics depicted in Fig. 5, which increased to a peak level that was maintained over 10 min. This apparent discrepancy highlights an interesting feature of mAb 327C in that binding to CD18 stabilized the active conformation of CD18 under conditions in which it is present during stimulation and flow cytometric analysis.
The kinetics of CD18 activation and down-regulation correlated temporally with the decrease in firm adhesion to ICAM-1 under shear as observed in the parallel plate flow chamber. This prompted us to examine the dynamics of receptor redistribution immediately following activation with IL-8. Fluores-  cence microscopy of neutrophils fixed and labeled with 327C revealed the formation of both small punctate clusters (Ͻ1 m 2 ) and large caps (ϳ3 m 2 ) within 30 s of stimulation (Fig.  8b). Within 2 min of stimulation, these caps dissipated into numerous smaller proximal clusters. By 10 min the smaller clusters had decreased in number, and the overall surface fluorescence was diminished. In contrast, treatment with 240Q induced small clusters of active CD18 that remained intact for over 10 min of observation (data not shown). Taken together, the immunofluorescence revealed the rapid mobility of 327C into aggregates of active CD18, which formed and subsequently dissipated over a time course that mirrored the rise and fall of ICAM-1 capture and adhesion stability in the flow chamber. Inhibition of CD18 Mobility and Neutrophil Capture of ICAM-1 Beads-It has been reported recently (35) that chemotactically stimulated neutrophil adhesion to ICAM-1, but not expression of activated ␤ 2 integrin, is dependent upon PI(3)K activity. It has also been reported that chemokine-triggered PI(3)K activity regulates the rapid lateral mobility of LFA-1 in lymphocytes (20). Although the mechanism for this phenomenon is not clearly understood, it is thought that PI(3)K generates lipid secondary messengers including inositol 1,4,5-trisphosphate and Ca 2ϩ flux that in turn regulate the formation of complexes between integrins and the cytoskeleton (36). Based on these findings, we examined the effect of inhibitors of PI(3)K and F-actin polymerization on the surface distribution and adhesive function of active LFA-1 following stimulation. Treatment with wortmannin did not affect the expression of LFA-1 but significantly inhibited the extent of clustering and capping of active CD18 stimulated by IL-8 (Fig. 9a). In contrast, wortmannin had no effect on the formation of punctate clusters of 327C induced by 240Q (data not shown).
We next examined the effects of pharmacological inhibition on expression and function of CD18 assessed by fluorescence flow cytometry. Cytochalasin B was applied to disrupt the F-actin polymerization, whereas PI(3)K activity was specifically inhibited by pretreatment with wortmannin or LY294002. Blocking PI(3)K activity or F-actin formation resulted in nominal inhibition in expression of active CD18 induced by either IL-8 or 240Q (Fig. 9b). Interestingly, treatment with cytochalasin B augmented expression of 327C sites in response to either mode of CD18 activation. Capture of ICAM-1 beads in response to IL-8 stimulation was significantly diminished in the presence of either inhibitor of PI(3)K, but not by treatment with cytochalasin B (Fig. 9c). In contrast, ICAM-1 bead adhesion activated by 240Q was not diminished in the presence of PI(3)K inhibitors. Treatment with cytochalasin B did not significantly inhibit ICAM-1 bead capture stimulated through either mode of activation. The data are consistent with a mechanism in which signaling via IL-8 triggers conformational activation of LFA-1, which subsequently redistributes into patches through a PI(3)K-dependent process that is independ- 8. Redistribution of active CD18 following stimulation with IL-8. The expression and topography of active CD18 was quantitated on 327C-labeled samples over the time course of stimulation. a, expression dynamics of active CD18 were quantitated in terms of 327C site number following IL-8 (1 nM) stimulation. Neutrophil suspensions in the presence of 327C-Alexa were stimulated, washed, and analyzed for 327C-Alexa binding using flow cytometry. b, binding of 327C to active CD18 on stimulated neutrophils was imaged at specific time points following stimulation with IL-8 (1 nM). Samples were labeled with biotinylated mAb 327C and then rapidly fixed with 1% paraformaldehyde, secondarily labeled with avidin-FITC, and imaged by phase contrast fluorescence microscopy. Each image is representative of at least 15 separate observations. Data are presented as the mean site density of 327C/PMN Ϯ S.E. from three separate experiments.

FIG. 9. Neutrophil activation in presence of inhibitors of PI(3)K and actin polymerization.
The role of the cytoskeleton and PI(3)K signaling activity on activation of CD18 was examined using wortmannin, LY294002, or cytochalasin B in conjunction with IL-8 or 240Q stimulation. a, topography of LFA-1 and active CD18 on neutrophils and stimulated with IL-8. Neutrophils were fixed 2 min after IL-8 (1 nM) stimulation and compared with those preincubated with wortmannin (100 nM). Images are representative of 20 observations. b, conformational activation of CD18 in the presence of inhibitors of PI(3)K and actin polymerization. Expression of 327C was quantitated by flow cytometry as described under "Experimental Procedures." Neutrophils were preincubated with cytochalasin B (0.3 mg/ml), wortmannin (100 nM), or LY294002 (20 mM) and then stimulated with either IL-8 (1 nM) or 240Q (5 g/ml) for 10 min. c, ICAM-1 bead capture in the presence of inhibitors of actin and PI(3)K. Neutrophils were preincubated with inhibitors and stimulated as described above. Activation of CD18 is plotted as percentage of stimulated vehicle control samples (0.1% MeSO). Data are presented as the mean Ϯ S.E. from three separate experiments. * denotes significant change (p Ͻ 0.05) in CD18 function compared with vehicle control. ent of F-actin formation. These steps appear to be critical for optimum capture efficiency of ICAM-1 as allosteric activation of LFA-1 in the presence of 240Q was less efficient at mediating capture of ICAM-1 but was not affected by PI(3)K blockers. DISCUSSION Neutrophils constitute greater than 70% of circulating leukocytes and are the first to accumulate within inflamed microcirculation arresting within a few cell diameters of the site of transmigration (37). Recruitment of neutrophils is a tightly regulated process in which the efficiency of conversion from transient to shear-resistant adhesion is influenced by the extent of cellular stimulation, which in turn controls the coordinated expression of activated receptor and vascular ligand (38). Several lines of evidence indicate that the gatekeeper in neutrophil recruitment is binding of activated ␤ 2 integrins to ICAM-1 and ICAM-2 on endothelium (39). In the current study we investigated the following two modes of CD18 activation in order to elucidate the molecular dynamics of neutrophil capture in shear flow: stabilization of the active conformation through binding of mAb 240Q, and chemotactic signaling via the IL-8 receptors, CXCR1 and CXCR2. By using this strategy we demonstrated the following. 1) A shift in LFA-1 conformation is necessary and sufficient for neutrophil adhesion to surface-expressed ICAM-1 at physiological levels of shear stress.
2) IL-8 at 75 pM, corresponding to ligation of no more than ϳ100 receptors/PMN, is sufficient to trigger neutrophil capture to ICAM-1. 3) Capture efficiency and expression of activated CD18 increased in direct proportion to IL-8 over a log range in dose. 4) Allosteric activation of only 5% of surface LFA-1 via 240Q was sufficient to trigger adhesion to ICAM-1 but at half the efficiency stimulated by IL-8. 5) the boost in capture efficiency involves a rapid redistribution of activated LFA-1 into membrane clusters via a signaling pathway involving PI(3)K but not F-actin assembly. 6) Stable adhesion is reversible within minutes of IL-8 activation and involves a progressive shift from high to low affinity CD18, which could be reversed by 240Q stabilization of the IDAS.
The adhesion kinetics observed in the parallel plate flow chamber mirrored the initial boost and subsequent drop in expression of the activation epitope of CD18 defined by 327C. Allosteric activation of CD18 by binding of 240Q was less effective at mediating capture over the initial low shear interval in the flow chamber even at a saturating dose of 5 g/ml, corresponding to activation of ϳ30% of expressed CD18. However, increased numbers of neutrophils transitioned to stable adhesion with time of exposure to high shear stress. Occupancy of just 5% of surface CD18 with 240Q was sufficient to stabilize ϳ20,000 receptors per cell in an active state, resulting in arrest of ϳ60% of interacting neutrophils. Interestingly, the arrested fraction did not increase with dose of 240Q; however, costimulation with low dose IL-8 (0.05 nM) and 240Q (5% CD18 occupancy) resulted in nearly the entire neutrophil population remaining arrested for up to 10 min of high shear stress. This synergistic response with costimulation suggests that adhesion is regulated by a shift in CD18 conformation in addition to other cell responses elicited by GPCR stimulation. Supporting this was the observation that allosteric activation, and not simply binding of antibody to CD18, potentiated intracellular Ca 2ϩ flux stimulated by IL-8. This result suggests that conformational activation of CD18 itself may provide an important outside-in signal. Adhesion strengthening with costimulation was determined primarily by the number of LFA-1 remaining in the active conformation over time, as Mac-1 adhesion function was blocked in the presence of bound 240Q. In this regard it has recently been reported that LFA-1 provides an important cosignal in activation of neutrophils. For example, superoxide secretion is defective in LFA-1 knockout mice (40). Moreover, LFA-1 is necessary and sufficient for transmigration at sites of inflammation as indicated in studies of LFA-1 and Mac-1 knockout mice (41,42).
Lymphocyte adhesion via LFA-1 can be signaled by phorbol ester or treatment with divalent cations such as Mg 2ϩ (43), which was found to correlate with a shift in binding affinity regulated at the IDAS (15). Expression of 327C on lymphocytes correlated with the ligand-binding state of LFA-1 but was not dependent upon ICAM-1 binding (14). Consistent with previous reports, ICAM-1/Fc induced an active conformation in ϳ50% of available LFA-1 in PBMCs resulting in a 4-fold increase in binding of 327C (15). This was not observed in a donor-matched population of neutrophils that constitutively expressed higher levels of active CD18 but exhibited no activation in the presence of excess ICAM-1/Fc. Another difference between LFA-1mediated adhesion of neutrophils and PBMCs is the respective dependence on F-actin. We found that neutrophil capture of ICAM-1 beads was not affected in the presence of cytochalasin B. Others have reported (44) that lymphocyte adhesion to ICAM-1 is facilitated by disruption of F-actin. These provide the first data showing conformational activation of LFA-1 may be differentially regulated on neutrophils as compared with PBMCs.
Flow cytometry provided a real time assay of the rate of adhesion between neutrophils and fluorescent beads coated with recombinant ICAM-1 at a site density commensurate with that measured in inflamed venules (ϳ25 sites/m 2 ) (30). This assay proved to be twice as sensitive as the parallel plate flow chamber in detecting the threshold dose of IL-8 (75 pM) sufficient to activate neutrophil capture of ICAM-1. Incrementing IL-8 dose by as little as 100 pM elicited a detectable increase in the rate and extent of ICAM-1 bead capture. Adhesion function closely followed stimulus dose, increasing by 6-fold over a log range in IL-8 concentration from ϳ0.1 to 1 nM. Peak capture efficiency at 1 nM corresponded to the reported K d of IL-8 binding, ϳ50% receptor occupancy. A threshold dose of 0.1 nM IL-8 was also found to be sufficient to trigger binding of 327C, which reported on ϳ20,000 active CD18 (ϳ100 sites/m 2 ). This dose of stimulus also corresponded to the minimum activation that triggered capture of ICAM-1 beads. Moreover, the dose response between IL-8 stimulation and 327C expression was nearly identical to the curve describing ICAM-1 bead adhesion. An important difference was that maximal expression of 327C at a dose of 5 nM IL-8 exceeded by 5-fold that required to activate optimal ICAM-1 adhesion. Thus, maximal IL-8 stimulation corresponded to activation of ϳ100,000 CD18 or ϳ50% of total membrane-expressed ␤ 2 integrin. However, half-maximum capture of ICAM-1 corresponded to ϳ0.2 nM IL-8 and activation of only 30,000 CD18, of which no more than ϳ15,000 LFA-1 participated in adhesion. Thus, with increasing dose of IL-8 stimulus, up-regulation of activated CD18 exceeds by as much as 3-fold that necessary to trigger adhesion function. Taken together, the data reveal a wide dynamic range coupling stimulus and response such that small increments in fractional occupancy of IL-8 above the threshold dose elicit an exponential increase in signaling of adhesion.
In comparison to GPCR signaling, allosteric activation of ␤ 2 integrin was less potent at eliciting capture of ICAM-1. However, only 5% occupancy of CD18 by 240Q, an estimated ϳ5,000 LFA-1 receptors or ϳ25 sites/m 2 , was sufficient to trigger ICAM-1 adhesion above base line in the bead capture and parallel plate adhesion assays. Efficiency of bead capture increased with binding of 240Q but over a relatively narrow dose range and at one-half the efficiency signaled by IL-8 stimulus. A limitation in eliciting optimum adhesion via the IDAS corre-lated with a diminished capacity of 240Q to bind and activate greater than ϳ25% of membrane CD18, as compared with the 50% CD18 activation elicited by IL-8. Because 240Q activates ICAM-1 adhesion through LFA-1, and not Mac-1, this corresponds to activation of less than 15% of CD18 at maximum allosteric activation. Binding kinetics of 327C to stimulated neutrophil suspensions confirmed what was observed for ICAM-1 bead capture, that CD18 activation occurred most rapidly over the 1st min of stimulation regardless of IL-8 dose.
In contrast, binding of 327C to neutrophils maximally activated with 240Q simply revealed the kinetics of recognition for CD18 already converted to the active state. These dynamics highlight the extraordinary efficiency by which IL-8 binds its receptor and converts CD18 to the active conformation over the initial seconds of stimulation. Thus, conformational activation of CD18 did not appear to be the rate-limiting step in ICAM-1 capture in the presence of IL-8.
Immunofluorescence of ␤ 2 integrin expression on resting and activated neutrophils revealed little 327C binding to unstimulated cells. IL-8 stimulation of neutrophils transduced rapid activation of CD18 as adhesion was measured within seconds, and subsequent deactivation of sites and disaggregation of clusters occurred within minutes. This transience in the active conformation of CD18 could only be detected by 327C labeling of samples that were removed from the test tube over the time course of stimulation, as 327C in excess over the course of stimulation appeared to stabilize the active conformation (Figs. 8a versus 6a). After only 30 s of IL-8 stimulation, a 7-fold increase in CD18 activation was detected, and a majority of active sites redistributed into submicron clusters and larger patches several m 2 in area. Two-color fluorescence showed that ϳ75% of LFA-1 colocalized with active CD18, as compared with less than 40% of Mac-1, despite the fact that Mac-1 expression was up-regulated severalfold with chemotactic stimulation. After only 2 min of stimulation membrane patches began to disperse with only small clusters remaining, and by 10 min practically all clusters had dispersed, and the number of active receptors dropped by ϳ50%. This transience of both active CD18 and LFA-1 cluster formation corresponded temporally with the reversibility of firm adhesion in the parallel plate flow chamber. Transience in the avidity of LFA-1 has been reported previously (30) for adhesion to ICAM-1-expressing cells and in adhesion to ICAM-1/Fc on beads. In contrast, 240Q-treated neutrophils remained adherent at a shear stress of 4 dynes/cm 2 . Because 240Q was also shown to lock CD18 in an active conformation, this suggests that a conformational shift to low affinity was responsible for the transience in adhesion.
Treatment with 240Q resulted in formation of small punctate clusters of active CD18 by a mechanism that did not necessarily require outside-in signal transduction. Supporting this was the observation that binding of 240Q did not itself signal intercellular calcium flux. The possibility exists that cluster formation was a consequence of CD18 dimerization upon bivalent binding of 240Q. However, this is probably not the case as Fab fragments of 240Q have been reported (14) to be equally effective at conformational activation of CD18. Antibody-induced aggregation of LFA-1 into clusters is an unlikely mechanism as discrete clusters of 327C formed in response to binding of just 1-5% of CD18 receptors by 240Q. Clustering of LFA-1 has been reported in response to binding of its physiological ligand, ICAM-3, which activated strong LFA-1/ICAM-1 adhesion in T-cells (45).
Chemokine stimulation can initiate signaling via distinct pathways that trigger a high affinity binding state and an increase in the lateral mobility of LFA-1, and both mechanisms activate the arrest of circulating lymphocytes (20,33,46). Enhanced membrane mobility of LFA-1 can be mediated by calciumdependent activation of calpain, which dissociates LFA-1 from the cytoskeleton by proteolysis of talin links to the actin cytoskeleton (47,48). Rapid lateral mobility of LFA-1 has been shown to be a critical factor in the efficiency of chemokinetriggered lymphocyte arrest to ICAM-1, and this process requires PI(3)K activity (20). Less information has been available regarding the mechanism of neutrophil capture on ICAM-1; however, it was recently reported (35) that CD18-mediated adhesion is dependent upon PI(3)K activation. We found that two inhibitors of PI(3)K activity, wortmannin and LY294002, significantly diminished the efficiency of ICAM-1 bead capture stimulated by IL-8 but had no effect on adhesion activated via the ␤I domain or IDAS conformational shift. IL-8 stimulated a rapid clustering of conformationally active LFA-1 that was essential to optimum capture efficiency of ICAM-1. Blocking PI(3)K decreased, but did not eliminate, clustering of 327C while leaving intact the ability of CD18 to undergo conformational activation in response to IL-8 or 240Q. Depleting cytoplasmic F-actin by pretreatment with cytochalasin actually increased expression of active CD18. These data support a model of IL-8-stimulated neutrophil capture on ICAM-1 that is critically dependent on an increase in LFA-1 mobility over a time frame of less than 1 s. The transition to stable shearresistant adhesion was associated with the rapid formation (ϳseconds) of patches of active LFA-1. The involvement of PI(3)K on both mobility and patching of LFA-1 suggests that it may be involved in the initial release of cytoplasmic attachments of LFA-1 to ␣-actinin (47). Our data also show that Ca 2ϩ flux, a signal downstream of PI(3)K activation, is important in the amplification of the adhesion. Stimulation with only 25 pM IL-8, corresponding to as few as 10 GPCR receptors bound per neutrophil, elicited Ca 2ϩ flux in a significant fraction of cells. A remarkable finding was the boost in Ca 2ϩ flux in response to costimulation with 240Q and IL-8. The fact that 240Q alone was sufficient to induce small clusters of LFA-1 but not a Ca 2ϩ flux suggests that GPCR-signaled events are critical to optimal adhesion efficiency. Increased lateral or rotational diffusion and formation of clusters of active LFA-1 could serve to amplify the probability of ligand binding and promote efficient capture. Allosteric activation at the IDAS via 240Q also elicited distinct clusters of active CD18, but this redistribution was not dependent on PI(3)K, and capture of ICAM-1 was less than one-half as efficient as that elicited by GPCR signaling.
In this study, we examined how the efficiency of neutrophil capture and firm adhesion on ICAM-1 is dynamically regulated by GPCR-signaled alterations in LFA-1 affinity and membrane distribution. The implication for neutrophil recruitment at acute sites of inflammation is that, like selectin bonds, capture may be mediated by attachment of few activated LFA-1 receptors elicited by as few as 10 GPCR. This notion has been confirmed recently in the microcirculation of the mouse, and Dunne et al. (19) has found that LFA-1 supports slow rolling of leukocytes and the efficient transition to cell arrest. In addition, our study has implications for the use of small molecule inhibitors of the active conformation of LFA-1. Recent data (14,49,50) suggest that statins as well as other I domain inhibitors act as potent antagonists of adhesion function. Our results suggest that suppression of recruitment at the site of inflammation will require tight binding of such inhibitors and occupancy of greater than 90% of active CD18. We conclude that the efficiency of neutrophil capture is tightly regulated by the dose of G-protein-linked stimuli and its effect on the site density and subsequent mobility of active LFA-1. Clustering of LFA-1 to high avidity sites supports adhesion strengthening leading to neutrophil arrest, but these adhesive bonds exist for only minutes thereby facilitating cell motility and transendothelial migration.