Dynamic shifts in LFA-1 affinity regulate neutrophil rolling, arrest, and transmigration on inflamed endothelium

Abstract

Polymorphonuclear leukocyte (PMN) recruitment to vascular endothelium during acute inflammation involves cooperation between selectins, G-proteins, and beta2-integrins. LFA-1 (CD11a/CD18) affinity correlates with specific adhesion functions because a shift from low to intermediate affinity supports rolling on ICAM-1, whereas high affinity is associated with shear-resistant leukocyte arrest. We imaged PMN adhesion on cytokine-inflamed endothelium in a parallel-plate flow chamber to define the dynamics of beta2-integrin function during recruitment and transmigration. After arrest on inflamed endothelium, high-affinity LFA-1 aligned along the uropod-pseudopod major axis, which was essential for efficient neutrophil polarization and subsequent transmigration. An allosteric small molecule inhibitor targeted to the I-domain stabilized LFA-1 in an intermediate-affinity conformation, which supported neutrophil rolling but inhibited cell polarization and abrogated transmigration. We conclude that a shift in LFA-1 from intermediate to high affinity during the transition from rolling to arrest provides the contact-mediated signaling and guidance necessary for PMN transmigration on inflamed endothelium.

Figure 1.

Allosteric inhibition of LFA-1 binding to ICAM-1. (A) Neutrophils were preincubated with Fc fragments (all samples), anti-Mac-1 (2LPM19c), and IC487475, as indicated. ICAM-1/IgG-Alexa-488 was then added to neutrophil suspensions and was followed immediately by activation with mAb 240Q. Cells were fixed and analyzed by flow cytometry. Fluorescence histograms are representative of the average population response from 4 separate experiments. (B) Data represent the mean ± standard error of the mean (SEM) from 4 separate experiments, as described in panel A. ^*Significance compared with mAb 240Q + anti-Mac-1 (P < .05). ^**Significance compared with mAb 240Q (P < .05). (C) Isolated human neutrophils were incubated over a dose range with IC487475 or TS1/22. Suspensions were then activated with IL-8 and immediately analyzed by FACScan flow cytometry for binding of ICAM-1/IgG-coated latex beads (approximately 25 sites/μm²). Data are representative of the average response from 4 separate experiments.

Figure 2.

I-domain allosteric regulation of LFA-1 extension during neutrophil activation. (A-B) Isolated human neutrophils were preincubated with 1 μM IC487475, 1 μM BIRT or vehicle, and the CD18 extension reporter mAb KIM127. Suspensions were then activated with 1 nM IL-8 and were incubated with FITC-conjugated secondary goat anti-mouse F(ab′)₂ before fixation and analysis by flow cytometry. Fluorescence histograms (A) are representative of the average population response from 6 separate experiments. Bar graph data (B) represent the mean ± SEM from 6 separate experiments. ^*Significance in KIM127 binding between BIRT and both IC487475 (P < .01) and vehicle-treated PMNs (P < .05). (C) rLFA-1 heterodimer was derivatized to latex microspheres and incubated with allosteric inhibitors IC487475, BIRT, and lovastatin. rLFA-1 beads were then incubated with MgCl₂ and mAb R3.1, followed by secondary labeling with fluorescent goat anti-mouse F(ab′)₂ and flow cytometry. Data are given as mean ± SEM for 4 separate experiments and are presented as percentage inhibition relative to maximal binding by vehicle-treated rLFA-1 beads. ^*Significant inhibition of R3.1 binding compared with vehicle (P < .001).

Figure 3.

Microkinetics of neutrophil rolling is regulated by LFA-1. (A) The instantaneous velocity of rolling neutrophils was quantitated as the distanced traversed by individual neutrophil centroids every 0.250 second for the length of time the neutrophil remained in the field of view. Data are representative of the average response of 30 individual neutrophils from 6 to 8 separate experiments. (B) LFA-1 activation regulates neutrophil rolling dynamics. The average rolling velocity and standard deviation from the mean velocity for neutrophils on IL-1β activated HUVECs were quantitated at 0.250-second intervals for the length of time an individual rolling neutrophil remained in the field of view. Data represent mean ± SEM for 30 neutrophils in 4 to 6 separate experiments. ^**Significance in velocity between IC487475 and both TS1/22 (P < .01) and BIRT (P < .001). #Significance in deviation between IC487475 and both TS1/22 (P < .01) and BIRT (P < .001).

Figure 4.

Kinetics of neutrophil rolling, arrest, and transmigration on inflamed endothelium. Isolated human neutrophils were incubated with IC487475, BIRT, TS1/22, 2LPM19c, or vehicle, as indicated. Neutrophil suspensions were then perfused into a parallel-plate flow chamber over IL-1β activated or unstimulated control HUVEC monolayers at a shear stress of 2 dyne/cm². (A) The number of rolling neutrophils, defined as those moving more than 1 cell diameter in 30 seconds, was quantitated in a field of view every minute throughout the interaction. Data represent the mean ± SEM from 5 to 10 separate experiments. (B) The number of adherent neutrophils, defined as those moving less than 1 cell diameter in 30 seconds, was quantitated in a field of view every minute throughout the interaction. Data represent the mean ± SEM from 5 to 10 separate experiments. (C) The number of transmigrated neutrophils was quantitated in a field of view every minute throughout the interaction. Data represent the mean ± SEM from 5 to 10 separate experiments.

Figure 5.

Kinetics of neutrophil shape change on HUVECs and L cells expressing E-selectin and ICAM-1. The membrane aspect ratio was quantitated as a measure of cell shape change for individual neutrophils immediately after arrest on IL-1β-activated HUVECs in the parallel-plate flow chamber (A) or L-E/I in the microfluid flow channel device (B-C) at 2 dyne/cm² shear stress. Neutrophils were preincubated with IC487475, BIRT, TS1/22, or vehicle, as indicated, then perfused in the respective flow chamber. On L-E/I, IL-8 was perfused into the microflow channel after neutrophils reached maximum adhesion, 3 minutes after injection. (A-B) Data are presented as the fraction of all arrested neutrophils in the field of view exhibiting a membrane aspect ratio greater than 1.4, a value deemed to be a threshold in membrane skewness based on empirical observations. Data represent the mean ± SEM from 4 separate experiments. (C) Images are 20 × phase contrast and are representative of adherent neutrophils on L-E/I under the conditions described in panel B.

Figure 6.

Allosteric inhibition of LFA-1 diminishes persistence of migration. Isolated human neutrophils were incubated with IC487475 or vehicle, then perfused into a parallel-plate flow chamber over IL-1β-activated HUVECs at a shear stress of 2 dyne/cm². After arrest, the centroid of migrating neutrophils was tracked at 5-second intervals for 30 seconds. Arrows outline the trajectory of migrating neutrophils, with the arrowhead positioned at cell centroid at the conclusion of each time segment. Micrographs are representative of approximately 100 individual neutrophil observations over 4 to 6 separate experiments.

Figure 7.

β₂-Integrin activation and redistribution on polarized and migrating neutrophils. Neutrophils were preincubated with IC487475, BIRT, nonblocking anti-LFA-1-Alexa 546 (TS2/4), nonblocking anti-Mac-1-PE (ICRF44), or mAb 327C-Alexa 488, as indicated. Cell suspensions were then infused into the parallel-plate flow chamber at a shear stress of 2 dyne/cm² over a monolayer of IL-1β-activated HUVECs. (A) On neutrophil rolling to arrest, real-time detection of β₂-integrin distribution was visualized by sequential collection of phase-contrast and immunofluorescent images every second for 90 seconds using a 60 × oil immersion objective. Micrographs are presented as phase-contrast (gray) and deconvolved immunofluorescence (red) overlays representative of approximately 100 individual neutrophil observations in 4 to 6 separate experiments. (B) Fractional distribution of β₂-integrin on polarized neutrophils after 90 seconds was determined by quantitating pixel intensity of the fluorescence signal at the leading projection and uropod. Data are presented as mean ± SEM and are representative of the β₂-integrin distribution of approximately 10 cells in 4 to 6 separate experiments. ^*Significance compared with active β₂-integrin and LFA-1 at the leading projection (P < .01). #Significance compared with active CD18 and LFA-1 at the uropod (P < .01).