Kindlin-3 Is Essential for the Resting α4β1 Integrin-mediated Firm Cell Adhesion under Shear Flow Conditions*

Integrin-mediated rolling and firm cell adhesion are two critical steps in leukocyte trafficking. Integrin α4β1 mediates a mixture of rolling and firm cell adhesion on vascular cell adhesion molecule-1 (VCAM-1) when in its resting state but only supports firm cell adhesion upon activation. The transition from rolling to firm cell adhesion is controlled by integrin activation. Kindlin-3 has been shown to bind to integrin β tails and trigger integrin activation via inside-out signaling. However, the role of kindlin-3 in regulating resting α4β1-mediated cell adhesion is not well characterized. Herein we demonstrate that kindlin-3 was required for the resting α4β1-mediated firm cell adhesion but not rolling adhesion. Knockdown of kindlin-3 significantly decreased the binding of kindlin-3 to β1 and down-regulated the binding affinity of the resting α4β1 to soluble VCAM-1. Notably, it converted the resting α4β1-mediated firm cell adhesion to rolling adhesion on VCAM-1 substrates, increased cell rolling velocity, and impaired the stability of cell adhesion. By contrast, firm cell adhesion mediated by Mn2+-activated α4β1 was barely affected by knockdown of kindlin-3. Structurally, lack of kindlin-3 led to a more bent conformation of the resting α4β1. Thus, kindlin-3 plays an important role in maintaining a proper conformation of the resting α4β1 to mediate both rolling and firm cell adhesion. Defective kindlin-3 binding to the resting α4β1 leads to a transition from firm to rolling cell adhesion on VCAM-1, implying its potential role in regulating the transition between integrin-mediated rolling and firm cell adhesion.

rins that mediate only firm cell adhesion upon activation, the resting integrin ␣4␤1 can mediate a mixture of rolling and firm leukocyte adhesion to vascular cell adhesion molecule-1 (VCAM-1) 3 but only supports firm cell adhesion postactivation, playing an important role in leukocyte trafficking and immune homeostasis (2)(3)(4)(5). Integrin-mediated cell adhesion is regulated by the dynamic shift between low and high affinity conformations of integrin for ligand binding (6). In the resting state, integrin has a low affinity bent conformation with the headpiece facing down toward the cell membrane; upon activation, integrin undergoes a series of conformational rearrangements and extends upward in a switchblade-like opening motion, leading to the increased integrin affinity (7)(8)(9). This process is commonly controlled by inside-out signals from the cytoplasm that are dependent on specific interactions between intracellular effector molecules, such as talin and kindlins, and the integrin cytoplasmic tail (10 -12). Binding of talin to integrin ␤ tails is a final common element of cellular signaling cascades that control integrin activation (13,14), and kindlins are thought to be coactivators (15)(16)(17). It is also reported that distinct kindlin-3 binding patterns can lead to distinct binding affinities of mucosal vascular addressin cell adhesion molecule-1 and VCAM-1 to integrin ␣4␤7 (18). In addition to inside-out signaling, extracellular metal ions can also regulate integrin affinity via a cluster of three divalent cation-binding sites in integrin ␤ I domain (19). Compared with the low affinity state in Ca 2ϩ /Mg 2ϩ , addition of Mn 2ϩ or removal of Ca 2ϩ strikingly increases the affinity and adhesiveness of almost all integrins (20 -22).
Although numerous studies have revealed the role of kindlin-3 in inside-out activation of integrin, little is known regarding its function in regulating the resting ␣4␤1-mediated cell adhesion. Herein we report that kindlin-3 was required for the resting ␣4␤1-mediated firm cell adhesion but not rolling adhesion under shear flow conditions. Silencing of kindlin-3 in K562 cells stably expressing human ␣4␤1 (K562-␣4␤1) significantly decreased the binding of kindlin-3 to ␤1 and thus down-regulated the binding affinity of the resting ␣4␤1 to soluble VCAM-1. Furthermore, knockdown of kindlin-3 converted the resting ␣4␤1-mediated firm cell adhesion to rolling adhesion on VCAM-1 substrates, increased cell rolling velocity, and impaired the stability of cell adhesion under flow. By contrast, firm cell adhesion mediated by Mn 2ϩ -activated ␣4␤1 was barely affected. Moreover, lack of kindlin-3 resulted in a more bent conformation of the resting ␣4␤1. Re-expression of knockdown-resistant wild-type (WT) kindlin-3 but not integrin binding-deficient kindlin-3 mutant could rescue the observed defects in integrin ␣4␤1-mediated cell adhesion and ␣4␤1 conformation in kindlin-3 knockdown cells, suggesting that the observed defects were due to the deficient kindlin-3/ integrin binding induced by kindlin-3 knockdown. Thus, kindlin-3 has an important role in maintaining a proper conformation of the resting ␣4␤1 and its ability to mediate firm cell adhesion before activation.
Antibodies and Reagents-Alexa Fluor 647-conjugated goat anti-mouse IgG, Cy3-conjugated goat anti-rat IgG, and Alexa Fluor 647-conjugated goat anti-human IgG were from Invitrogen. mAb to kindlin-3 was from Santa Cruz Biotechnology (N-12). mAb to ␤1 was from Abcam (EP1041Y). mAbs 9F10 and AIIB2 against human ␣4 and ␤1 integrin, respectively, were prepared from hybridomas (Developmental Studies Hybridoma Bank). mAb to CD45 was from Sino Biological (10086-H02H). mAb Act-1 against human ␤7 integrin was as described previously (33). Fab fragments were produced as described (34), and direct labeling of antibodies with Alexa Fluor 488 was performed using a protein labeling kit according to the manufacturer's instructions (Invitrogen). Human VCAM-1/Fc fusion protein containing Ig domains 1-7 of human VCAM-1 fused to the hinge and Fc region of human IgG1 was generated as described (35). Complete protease inhibitor mixture tablets were from Roche Applied Science.
Flow Cytometry-Flow cytometry was done as described (37). Cell surface expression of integrin ␣4␤1 on K562-␣4␤1 transfectants was determined by staining with mAbs 9F10 and AIIB2. Stained cells were then measured using a FACSCalibur (BD Biosciences) and analyzed using FlowJo software.
Soluble Ligand Binding Assay-The soluble ligand binding assay was performed as described (35,37). Briefly, 20 g/ml VCAM-1/Fc fusion protein was preincubated with Alexa Fluor 647-conjugated goat anti-human IgG in 50 l of Hepes-buffered saline (20 mM Hepes, pH 7.4) containing either 1 mM Ca 2ϩ /Mg 2ϩ or 1 mM Mn 2ϩ and then incubated with cells for 30 min at room temperature. Next, cells were washed twice, measured using a FACSCalibur, and analyzed using FlowJo software. As a control, cells were preincubated with 20 g/ml ␣4␤1 blocking mAb AIIB2 for 5 min at 37°C before addition of VCAM-1/Fc complexes.
Flow Chamber Assay-The flow chamber assay was performed as described (21,38). A polystyrene Petri dish was coated with a 5-mm diameter, 20-l spot of 5 g/ml purified VCAM-1/Fc in coating buffer (PBS, 10 mM NaHCO 3 , pH 9.0) for 1 h at 37°C followed by 2% BSA in coating buffer for 1 h at 37°C to block nonspecific binding sites. Cells were diluted to 1 ϫ 10 6 /ml in Buffer A (Hepes-buffered saline, 0.5% BSA) containing the indicated divalent cations immediately before infusion in the flow chamber. Cells were allowed to accumulate for 30 s at 0.3 dyne/cm 2 and 10 s at 0.4 dyne/cm 2 . Then shear stress was increased every 10 s from 1 dyne/cm 2 up to 32 dynes/cm 2 in 2-fold increments. The number of cells remaining bound at the end of each 10-s interval was determined.
Rolling velocity at each shear stress was calculated from the average distance traveled by rolling cells in 3 s. A velocity of 1 m/s, which corresponds to a movement of 1 ⁄ 2 cell diameter during the 3-s measurement interval, was the minimum velocity required to define a cell as rolling instead of firmly adherent. For integrin ␣4␤1 blocking, cells were preincubated with 20 g/ml AIIB2 for 5 min at 37°C.
Cell Detachment Assay-Cells were prepared as described in the flow chamber assay and then infused in the flow chamber. Cells were allowed to accumulate for 0.3 dyne/cm 2 and 10 s at 0.4 dyne/cm 2 . Then shear stress was increased every 10 s from 1 dyne/cm 2 up to 16 dynes/cm 2 in 2-fold increments. The cells remaining bound to VCAM-1 substrates (5 g/ml) at each wall shear stress were determined as a percentage of initial adherent cells at 1 dyne/cm 2 .
Fluorescence Resonance Energy Transfer (FRET) Assay-FRET was measured as described (33,39). For detecting the orientation of integrin ectodomain relative to cell membrane, cells were seeded on a poly-L-lysine (100 g/ml)-coated surface in serum-free DMEM with the indicated divalent cation and incubated for 30 min at 37°C. Adherent cells were fixed with 3.7% paraformaldehyde for 15 min at room temperature, and nonspecific sites were blocked by incubation with 10% serumrich medium for 10 min at room temperature. Then cells were stained with 20 g/ml Alexa Fluor 488-conjugated AIIB2 Fab, Alexa Fluor 488-conjugated anti-CD45 Fab, or Alexa Fluor 488-
Next, we examined the effect of kindlin-3 knockdown on the association of kindlin-3 with the resting ␤1 integrin in 1 mM Ca 2ϩ /Mg 2ϩ or with the activated ␤1 integrin in 1 mM Mn 2ϩ . A co-immunoprecipitation assay showed that knockdown of kindlin-3 significantly reduced the binding of kindlin-3 to both the resting and Mn 2ϩ -activated ␤1 integrins (Fig. 1C). As expected, re-expression of WT kindlin-3 in kindlin-3 knockdown cells restored the binding of kindlin-3 to ␤1 integrin to the level in control cells. However, re-expression of kindlin-3 W596A mutant did not rescue the kindlin-3 binding (Fig. 1C).
Kindlin-3 Is Essential for Firm Cell Adhesion Mediated by the Resting ␣4␤1-Integrin ␣4␤1 mediates a mixture of rolling and firm cell adhesion in shear flow on VCAM-1 substrates when in its resting state and only supports firm cell adhesion upon activation (2). We next investigated the role of kindlin-3 in regulating the cell adhesion mediated by ␣4␤1 pre-and postactivation. The adhesive behaviors of the K562-␣4␤1 transfectants in shear flow were characterized in a parallel wall flow chamber with human VCAM-1/Fc absorbed to its lower wall. The shear stress was incrementally increased, and the velocity of the cells remaining bound at each increment was determined (42). In 1 mM Ca 2ϩ /Mg 2ϩ , the control and luciferase shRNA-treated K562-␣4␤1 cells showed a mixture of about 30% of rolling events and 70% of firmly adherent events in the total adherent cells (Fig. 3, A and B). In contrast, kindlin-3 knockdown cells showed a similar number of adherent cells, but the percentage of firmly adherent cells decreased from 70 to 28% (Fig. 3, A and  B), indicating that reduced kindlin-3 expression results in a transition from firm adhesion to rolling adhesion mediated by the resting ␣4␤1. In addition, kindlin-3 knockdown cells showed significantly faster rolling compared with control cells (Fig. 3C). The addition of Mn 2ϩ strikingly increased the adhesiveness of K562-␣4␤1 cells to VCAM-1, leading to significantly increased adherent cells with nearly 100% firmly adherent events (Fig. 3, A and B). Knockdown of kindlin-3 led to a slight decrease in the number of adherent cells but did not affect the percentage of firmly adherent events. These data indicate that kindlin-3 is essential for the resting ␣4␤1-mediated firm cell adhesion, and reduced kindlin-3 expression converts the resting ␣4␤1-mediated firm cell adhesion to rolling adhesion. Moreover, re-expression of WT kindlin-3, but not kindlin-3 W596A mutant, in kindlin-3 knockdown cells efficiently rescued the defects in cell adhesion (Fig. 3), suggesting the essen-tial role of kindlin-3/␤1 interaction in ␣4␤1-mediated cell adhesion.

Kindlin-3 Knockdown Leads to a More Bent Conformation of
␣4␤1-Integrin activation is accompanied by global conformational rearrangements as the headpiece of integrin folds over its legs and faces down toward the membrane in the low affinity bend conformation and extends upward in a switchblade-like opening upon activation (7,43). We next used a FRET assay to study the effect of kindlin-3 knockdown on integrin conformation. To assess the orientation of integrin ␣4␤1 ectodomain relative to the plasma membrane, ␣4␤1 was labeled with Alexa Fluor 488-conjugated AIIB2 Fab fragment, which binds to the top of ␤1 I domain, as donor (44), and the plasma membrane was labeled with a lipophilic probe, FM4-64 FX, as acceptor (33,39). In 1 mM Ca 2ϩ /Mg 2ϩ , kindlin-3 knockdown cells showed higher FRET efficiency than the control and luciferase shRNAtreated cells, suggesting a more bent conformation of the resting ␣4␤1 when kindlin-3 was knocked down (Fig. 5A). Activation of integrin ␣4␤1 by 1 mM Mn 2ϩ significantly decreased the FRET efficiency, suggesting the extension of ␣4␤1 ectodomain (Fig. 5A). In addition, the FRET efficiency of kindlin-3 knockdown was higher than that of controls in Mn 2ϩ , suggesting that kindlin-3 knockdown also reduces the extension of Mn 2ϩ -activated ␣4␤1 to some degree (Fig. 5A). Re-expression of WT kindlin-3 in kindlin-3 knockdown cells fully abolished the integrin ␣4␤1 conformational change induced by kindlin-3 knockdown in 1 mM Ca 2ϩ /Mg 2ϩ and 1 mM Mn 2ϩ , whereas re-expression of kindlin-3 W596A mutant showed no rescue effect (Fig. 5A), suggesting that the observed defects are due to the kindlin-3 knockdown-induced deficient kindlin-3/␤1 integrin binding. Thus, the binding of kindlin-3 to ␤1 integrin is important for maintaining a proper conformation of ␣4␤1 in both resting and active states.
To further confirm that the observed regulation is specific for integrin, we also examined the effect of kindlin-3 knockdown on the conformation of integrin ␣4␤7 and CD45 as controls. Kindlin-3 expression level does not affect the cell surface expression of ␣4␤7 and CD45 (18,(45)(46)(47). To examine the orientation of ␣4␤7 ectodomain relative to the plasma mem-brane using the FRET system, K562 cells stably expressing human ␣4␤7 (K562-␣4␤7) was labeled with Alexa Fluor 488conjugated Act-1 Fab fragment, which binds to the top of ␤7 I domain (48), as donor. The FRET results showed that knockdown of kindlin-3 had similar effects on the global conformation of ␣4␤7 as observed in ␣4␤1 (Fig. 5B). By contrast, kindlin-3 knockdown did not change the FRET efficiency between CD45 and plasma membrane, indicating that kindlin-3 knockdown does not induce a global conformational change of CD45 (Fig. 5C).

Discussion
Kindlins serve as coactivators of integrins through binding to integrin ␤ tails to induce integrin activation (41,49). In addition to the major function of kindlins in integrin activation, we report that kindlin-3 has an important role in regulating the conformation and function of integrin ␣4␤1 in its resting state. Our data show that inhibition of kindlin-3 binding to ␣4␤1 by kindlin-3 silencing triggered a more bent conformation of the resting ␣4␤1, leading to a transition from firm cell adhesion to rolling adhesion, higher rolling velocity, and less stable interaction between the resting ␣4␤1 and VCAM-1.
Previous study shows that ␣4␤1 in kindlin-3-null lymphocytes retains intrinsic rolling adhesions to VCAM-1 and exhibits partial defects in chemokine-stimulated adhesiveness to VCAM-1 (31). Moreover, it has been reported that kindlin-3deficient lymphocytes, although deficient in optimal firm adhesions, still are able to use their residual integrin adhesiveness to enter tissues (50). Consistent with these results, our study provides additional information that inhibition of kindlin-3 binding to ␣4␤1 converted the resting ␣4␤1-mediated firm cell adhesion to rolling adhesion, allowing ␣4␤1 to support robust rolling cell adhesion before activation, and only partially affected firm cell adhesion mediated by the activated integrin (Fig. 3, A and B).
Unlike most integrins that only mediate firm cell adhesion upon activation, integrin ␣4␤1 mediates a mixture of rolling and firm cell adhesion in its resting state (3,5). Studies have shown that rolling and firm cell adhesion are two distinct phases of adhesion with a phase transition between them, interpreted directly by integrin conformational rearrangements that can be induced by intracellular effector proteins via inside-out signaling (21,38). A clasp formed by a salt bridge between the integrin ␣ and ␤ tails is crucial for maintaining integrins in the bent, inactive conformation. Forced separation of the clasp can trigger extension of ectodomains and conformational changes in the ligand-binding site, generating the activated integrin with the extended, high affinity conformation (51,52). Intracellular proteins that interact with integrin tails, such as talin, could induce conformational activation of the integrin by disrupting the integrin clasp (53,54). Kindlins serve as coactivators as they cooperate with talin to activate integrin (16,55). Our data show that disassociation of kindlin-3 with the resting ␣4␤1 triggered a more bent conformation of the resting ␣4␤1, resulting in the transition from firm cell adhesion to rolling adhesion, implying an important role of kindlin-3 in modulating the unfolding transition of integrin ␣4␤1, which is crucial  MAY 6, 2016 • VOLUME 291 • NUMBER 19

Kindlin-3 Regulates Resting ␣4␤1-mediated Firm Cell Adhesion
for the phase transition between cell rolling adhesion and firm adhesion.
Integrin affinity and avidity regulation are both important for integrin-mediated cell adhesion; they are distinct processes but mutually regulated and often occur at the same time (56 -58). Integrin affinity transition is associated with the conformational rearrangements of integrin molecules (7). By using an intramolecular FRET system, we found that inhibition of kindin-3 binding to ␤1 led to a more bent resting conformation of ␣4␤1 as well as ␣4␤7 (Fig. 5), suggesting the important role of kindlin-3 in triggering extended (high affinity) conformation of ␣4 integrins. The results are consistent with previous reports that kindlin-3 is required for the induction of the high affinity conformation of ␣L␤2 (31,59). Interestingly, kindlin-3 has been shown to have little effect on the affinity of purified monomeric ␣IIb␤3 integrin in a cell-free system (60). Moreover, kindlin-2 increases the multivalent ligand binding to integrin ␣IIb␤3 by promoting the clustering of ligand-occupied ␣IIb␤3 in non-hematopoietic cells (60). These data suggest that kindlins may promote integrin-ligand binding by clustering ␣IIb␤3 rather than inducing conformational activation of monomeric integrin. It is noteworthy that the reported distinct mechanisms of kindlin-3 in regulating integrin-ligand binding are  observed in different integrins, and some experiments use different kindlins. Furthermore, it has been reported that integrin ␤1 tails have higher binding affinity for kindlin-3 than ␤3 tails in a cell-free system (31,45). Thus, it is possible that kindlin-3 regulates the ligand binding of different integrins (␣4␤1 and ␣IIb␤3) via distinct mechanisms.
Clinically, loss of kindlin-3 expression accounts for the pathogenesis of leukocyte adhesion deficiency type III that is characterized by bleeding disorders and defective recruitment of leukocytes into sites of infection (28,29). Our finding suggests that leukocytes in these patients might have several functional deficiencies of integrin ␣4␤1, including lack of firm cell adhesion mediated by the resting ␣4␤1, less stable ␣4␤1/ VCAM-1 interactions, and higher rolling velocity besides the deficient activation of this integrin via inside-out signaling (26).
Taken together, kindlin-3 is crucial for maintaining a proper conformation of the resting ␣4␤1 and its ability to mediate firm cell adhesion before activation. Defective kindlin-3 binding to the resting ␣4␤1 leads to a transition from firm to rolling cell adhesion on VCAM-1, implying its critical role in regulating the transition between integrin-mediated rolling and firm cell adhesion.