Beta-1 integrin-mediated adhesion may be initiated by multiple incomplete bonds, thus accounting for the functional importance of receptor clustering

Abstract

The regulation of cell integrin receptors involves modulation of membrane expression, shift between different affinity states, and topographical redistribution on the cell membrane. Here we attempted to assess quantitatively the functional importance of receptor clustering. We studied beta-1 integrin-mediated attachment of THP-1 cells to fibronectin-coated surfaces under low shear flow. Cells displayed multiple binding events with a half-life of the order of 1 s. The duration of binding events after the first second after arrest was quantitatively accounted for by a model assuming the existence of a short-time intermediate binding state with 3.6 s(-1) dissociation rate and 1.3 s(-1) transition frequency toward a more stable state. Cell binding to surfaces coated with lower fibronectin densities was concluded to be mediated by single molecular interactions, whereas multiple bonds were formed <1 s after contact with higher fibronectin surface densities. Cell treatment with microfilament inhibitors or a neutral antiintegrin antibody decreased bond number without changing aforementioned kinetic parameters whereas a function enhancing antibody increased the rate of bond formation and/or the lifetime of intermediate state. Receptor aggregation was induced by treating cells with neutral antiintegrin antibody and antiimmunoglobulin antibodies. A semiquantitative confocal microscopy study suggested that this treatment increased between 40% and 100% the average number of integrin receptors located in a volume of approximately 0.045 microm(3) surrounding each integrin. This aggregation induced up to 2.7-fold increase of the average number of bonds. Flow cytometric analysis of fluorescent ligand binding showed that THP-1 cells displayed low-affinity beta-1 integrins with a dissociation constant in the micromolar range. It is concluded that the initial step of cell adhesion was mediated by multiple incomplete bonds rather than a single equilibrium-state ligand receptor association. This interpretation accounts for the functional importance of integrin clustering.

FIGURE 1

Motion of THP-1 cells along fibronectin-coated surfaces under flow. Monocytic THP-1 cells were driven along fibronectin-coated surfaces with a low wall shear rate of ∼4 s⁻¹. A typical trajectory is shown: periods of fairly uniform motion (A, B, and C, with average velocities of 20.4 μm/s, 18.6 μm/s, and 16.8 μm/s, respectively) are separated by arrests of varying duration (1 and 2 are transient arrests lasting 1.03 s and 0.95 s, respectively, 3 is a durable arrest (only partially shown in the figure)).

FIGURE 2

Duration of cell-fibronectin association. Monocytic THP-1 cells moving along fibronectin-coated surfaces displayed binding events with a wide range of durations. This figure shows a typical detachment curve obtained after recording 448 arrests on 1226 control cells interacting with surfaces coated with 10 μg/ml fibronectin. Experimental values are shown as crosses. Three domains could be defined: i), initial horizontal part that usually displayed downward concavity; ii), linear region where a corrected number of arrests could be obtained by extrapolation of the linear regression line (thick line); and iii), curved region revealing delayed bond stabilization (between 0.45 and 1 s after arrest).

FIGURE 3

Summary of experimental data. A total number of 8582 cell trajectories were monitored and 1776 binding events were followed for at least 5 s each to determine their duration. Detachment curves obtained under various experimental conditions are shown together with a selection of theoretical curves. To allow optimal comparison, curves were normalized by setting at 100 the number of bound cells at time 0.15 s after arrest. (A) Control cells interacted with surfaces coated with 10 μg/ml (⋄), 1 μg/ml (crosses), or 0.1 μg/ml (▵) fibronectin. The thin line was obtained with a three-parameter model (k_off = 3.6 s⁻¹, k_t = 1.3 s⁻¹, k_2m = 0.19 s⁻¹). The thick line was obtained with a five-parameter model derived from the previous one by allowing bond formation with kinetic rate k_on = 1.44 s⁻¹ and average initial number of bonds 2.16. (B) Control cells interacting with surfaces coated with 0.1 μg/ml fibronectin. The thin line was obtained with the same three-parameter model as shown in A, the thick line was obtained with the simplest two-parameter model (single intermediate complex with off rate k_off = 3.6 s⁻¹ and transition rate to more stable conformation k_t = 1.3 s⁻¹). The dotted line was obtained with another three-parameter model, assuming that attachment might be mediated by two complexes with dissociation rates of 5.1 s⁻¹ (66%) or 0.137 s⁻¹ (34%). (C) Cells were treated with neutral K20 antibodies without (⋄) or with (▵) aggregation with anti-mouse immunoglobulins before studying interaction with surfaces coated with 10 μg/ml fibronectin. Theoretical curves were obtained with immediate bond-formation model, using previously determined parameters k_off = 3.6 s⁻¹ and k_t = 1.3 s⁻¹ with an average initial bond number of 1.52 (K20 only) or 2.47 (aggregated receptors). (D) Cells were treated with neutral K20 antibodies without (⋄) or with (▵) aggregation with anti-mouse immunoglobulins before studying interaction with surfaces coated with 1 μg/ml fibronectin. Theoretical curves were obtain with immediate bond-formation model, using previously determined parameters with an average initial bond number of 1 (K20 only) or 2.72 (aggregated receptors). (E) Cells were treated with 12G10 function-enhancing antibodies before studying interaction with surfaces coated with 10 μg/ml (⋄) or 1 μg/ml (crosses) fibronectin. The theoretical curve was calculated assuming single bonds with two kinetic constants k_off = 0.92 s⁻¹ and k_t = 1.3 s⁻¹. (F) Cells were treated with cytochalasin D (⋄) or latrunculin A (crosses) before interacting with surfaces coated with 10 μg/ml fibronectin . The theoretical curve was calculated with the same model as used in Fig. 3 B (single bond three-parameter model).

FIGURE 4

Interaction of control THP-1 cells with surfaces treated with lower fibronectin concentration. The interaction of control THP-1 cells with surfaces coated with 0.1 μg/ml fibronectin was studied. Triangles represent experimental values obtained after monitoring 1495 individual cells (180 arrests). The thick line was obtained with a single two-state bond model (k_off = 3.6 s⁻¹ and k_t = 1.3 s⁻¹). The thin line represents a theoretical curve obtained with multiple one-state bond model (k_off = 3.18 s⁻¹, k_on = 4.77 s⁻¹).

FIGURE 5

Interaction of control THP-1 cells with surfaces treated with higher fibronectin concentration. The interaction of control THP-1 cells with surfaces coated with 10 μg/ml fibronectin was studied. Diamonds represent experimental values obtained after monitoring 1226 individual cells (448 arrests). The thick line was obtained with the immediate bond-formation model average initial bond number: 2.64 (k_off = 3.6 s⁻¹ and k_t = 1.3 s⁻¹). The thin line represents a theoretical curve obtained with continuous bond-formation model (k_off = 3.6 s⁻¹, k_−t = 1.3 s⁻¹, k_on = 3.96 s⁻¹).

FIGURE 6

Absolute calibration of a confocal microscope. Serial dilutions of fluorescent antibodies were examined as parallelepipedic sheets, yielding fields of uniform brightness. The relationship between intensity and antibody concentration was fairly linear, as shown on a representative experiment. Correlation coefficient is 0.99; the regression line equation is: Intensity = 0.195 × concentration (μg/ml) + 2.06.

FIGURE 7

Spatial fluorescence distribution of fluorescent antibody solutions. Solutions of phosphate buffer (thin line), standard fluorescent antibodies (dotted line), or deaggregated fluorescent antibodies (thick line) were studied with confocal microscopy; the histogram of intensity distribution in a representative sample of 65,536 pixels is shown in each case.

FIGURE 8

Spatial resolution of a confocal microscope. A nondeaggregated solution of fluorescent immunoglobulin molecules was observed with confocal microscopy, and the fluorescence of individual pixels is shown on a representative area. Values are expressed with a 16-level scale, and pixel with intensity <1 (i.e., <16 on a 256-level scale) are not shown for the sake of clarity. Obviously, there is no light spread around the brightest pixels. Bar is 2.5 μm.

FIGURE 9

Influence of section plane and fluorescence distribution on the median fluorescence index. Monocytic THP-1 cells were labeled in suspension with fluorescent anti-CD29 monoclonal antibodies without (A, B) or with (C, D) additional anti-mouse immunoglobulin polyclonal (Fab′)₂. They were then fixed and examined with confocal microscopy and representative sections corresponding to a diametral plane (B, D) or closer to the cell boundary (A, C) are shown. The mean fluorescence aggregation index was calculated as described to yield a semiquantitative estimate of receptor aggregation. Values obtained on these representative examples were respectively 6 (A), 16 (B), 12 (C), and 43 (D). Bar length is 5 μm.

FIGURE 10

Validity of the linear approximation for nonspecific ligand binding. Monocytic THP-1 cells were incubated with an extensive concentration range of fluorescent peptide ligand specific for VLA-4 or VLA-5. Cell fluorescence F was determined with flow cytometry for each ligand concentration [L]. The dependence of [L]/F on [L] is shown after mere subtraction of cell autofluorescence (▪) or subtraction of autofluorescence and nonspecific binding assumed to be linearly dependent on [L] (crosses). The latter curve is linear, in accordance with Eq. 13.