Published online before print September 22, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
is required for chemokine-induced, sustained adhesion under flow in vivo
,1
,
,2
Departments of
* Molecular Physiology and Biophysics and
Biomedical Engineering, and
Robert M. Berne Cardiovascular Research Center, University of Virginia, Charlottesville, Virginia, USA; and
Department of Genetics and Developmental Biology, University of Connecticut Health Center, Farmington, Connecticut, USA
2 Correspondence: University of Virginia, Robert M. Berne Cardiovascular Research Center, MR5 Bldg., Room 1013, P.O. Box 801394, Charlottesville, VA 22908-1394, USA. E-mail: klausley{at}virginia.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
can be activated directly by the β dimer of heterotrimeric G proteins coupled to CXCR2. Here, we used in vivo and ex vivo intravital microscopy models to test the role of PI-3K in leukocyte arrest. PI-3K null mice showed an 80% decrease in CXCL1-induced leukocyte adhesion in venules of the exteriorized mouse cremaster muscle. In wild-type mice, rolling leukocytes showed rapid and sustained adhesion, but in PI-3K–/– mice, adhesion was not triggered at all or was transient, suggesting that absence of PI-3K interferes with integrin bond strengthening. Wild-type mice reconstituted with PI-3K null bone marrow showed a 50% decrease in CXCL1-induced leukocyte adhesion. In a blood-perfused micro-flow chamber, leukocytes from PI-3K–/– mice showed a defect in adhesion on a P-selectin/ICAM-1/CXCL1 substrate, indicating that leukocyte PI-3K was required for adhesion. The adhesion defect in PI-3K–/– mice was as severe as that in mice lacking LFA-1, the major integrin responsible for neutrophil adhesion. We conclude that the isoform of PI-3K must be functional in leukocytes to allow efficient adhesion from rolling in response to chemokine stimulation.
Key Words: signal transduction • cell trafficking • inflammation
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Arrest chemokines bind to heptahelical G protein-coupled receptors to activate downstream signaling pathways that induce firm adhesion. Each G protein is composed of 
, β, and subunits [10
]. Upon chemokine binding, the G subunit dissociates from the Gβ subunit to initiate specific signaling pathways [10
]. However, signaling mechanisms downstream of the receptor involved in chemokine-induced leukocyte adhesion have yet to be defined clearly. In human monocytes, CXCL1-induced adhesion was blocked by the PI-3K inhibitors wortmannin and LY294002 [9
]. Furthermore, these inhibitors also blocked CXCL12-, CCL19-, and CCL21-induced lymphocyte adhesion [11
]. These data suggest that chemokine receptor signaling through PI-3K plays a role in leukocyte adhesion but do not identify the isoform responsible for arrest.
Leukocyte arrest from rolling requires coordinated signaling between the chemokine receptor and the integrin LFA-1(Lβ2) [12
]. LFA-1 is naturally in a folded, bent conformation with low affinity for its ligands such as ICAM-1 [13
]. Upon cell activation, the cytoplasmic tails of L and β2 move apart and induce the extended conformation by inside-out signaling [13
]. The extended conformation can bind the ligand, which in turn, leads to further integrin activation [14
]. LFA-1 binding is stabilized by post-receptor occupancy events such as clustering [11
, 15
].
PI-3K is a family of enzymes that phosphorylates the 3-hydroxyl group of the inositol head group of D-myo-phosphatidylinositol (PI) and its derivative PI 4,5-bisphosphate [16
]. PI-3K, present in multiple isoforms, can be divided into three classes [17
], categorized based on subunit structure, substrate specificity, and regulation [18
]. PI-3K is a Class IB PI-3K composed of a p110 catalytic subunit and a 101-kDa regulatory subunit, p101 [19
, 20
]. This PI-3K isoform has been shown to signal downstream of the Gβ subunit of chemokine receptors [21
].
Previous research about PI-3K has focused on its role in cell activation, superoxide production, and migration. This work was aided by the generation of PI-3K knockout (KO) mice lacking the p110 catalytic subunit [22
23
24
]. PI-3K was found to be necessary for PI 3,4,5-trisphosphate production, protein kinase B/Akt activation, and superoxide production in fMLP and C5a-treated neutrophils isolated from PI-3K–/– mice [22
23
24
]. Furthermore, these neutrophils had impaired migration in response to fMLP [22
23
24
]. Several reports have identified a role for PI-3K in chemoattractant-directed migration of other leukocyte subsets. Dendritic cells (DC) from PI-3K–/– mice have been shown to have an impaired migratory response to the chemokines CCL3, CCL5, and CCL19 compared with wild-type DC [25
]. Macrophage migration to MCP-1 and CCL5 was also impaired in these mice [26
, 27
]. The possible clinical significance of the role of PI-3K in inflammation is demonstrated by a significant beneficial effect of pharmacological inhibition of PI-3K in experimental arthritis [28
]
An inability of leukocyte subsets to migrate into tissue may in part be a result of an inability to adhere to the vessel wall. Blocking PI-3K by pharmalogical inhibitors was shown to block mobility but not affinity changes of integrins on lymphocytes [11
], effectively blocking lymphocyte arrest to a low-density ICAM-1 substrate. However, these data did not identify the isoform of PI-3K involved. Gerszten et al. [29
] demonstrated that PI-3K was required for MCP-1-stimulated arrest of human monocytes to E-selectin-transfected HUVEC. Lysates from these monocytes were shown to have increased PI-3K activity [29
]. However, it was not determined if this increased activity mediates human monocyte arrest. In vivo studies using PI-3K–/– mice have demonstrated a role for PI-3K in platelet aggregation in a carotid artery injury model [30
] and granulocyte tethering to endothelial cells in TNF--treated cremaster venules [31
]. However, no studies have addressed whether leukocyte PI-3K is required for leukocyte adhesion.
This study was undertaken to explore the role of leukocyte PI-3K in adhesion in vivo and in an ex vivo, blood-perfused micro-flow chamber with controlled substrates. We discovered a profound adhesion defect in PI-3K–/– mice, similar in severity to that seen in LFA-1–/– mice.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mice
All mouse experiments were performed under a protocol approved by the Animal Care and Use Committee of the University of Virginia (Charlottesville). Animals were housed in a barrier facility under specific pathogen-free conditions, and all experiments were performed on 6- to 20-week-old mice that appeared healthy. Breeder pairs of PI-3K KO mice were provided by Dr. Dianqing Wu at the University of Connecticut (Farmington) [24
]. Mice were bred, and deletion of the p110 subunit of PI-3K was determined by PCR. Data from PI-3K null mice were compared with data from littermate-control, wild-type mice. In experiments using the PI-3K inhibitor LY294002, mice were treated for 1 h with 100 mg/kg LY294002 or the vehicle control ethanol followed by intravital microscopy. LFA-1–/– mice were a kind gift of Dr. Christie Ballantyne [32
].
Surgical preparation
Mice were anesthetized with an i.p. injection of ketamine hydrochloride (125 mg/kg, Sanofi Winthrop Pharmaceuticals, New York, NY), atropine sulfate (0.025 mg/kg, Fujisawa USA, Inc., Deerfield, IL), and xylazine (12.5 mg/kg, TranquiVed, Phoenix Scientific, Inc., St. Joseph, MO) and placed on a heating pad maintained at 37°C. Polyethylene (PE) 90 tubing [internal diameter (ID) 0.86 mm, outer diameter (OD) 1.27 mm] was used to intubate the trachea of the mouse to allow unobstructed breathing, and PE 10 tubing (ID 0.28 mm, OD 0.61 mm) was used to cannulate the carotid artery. The cannula in the carotid artery was flushed with 10 units/ml heparin sodium in saline before exteriorization of the cremaster. A 50 µl aliquot of blood was drawn from the PE 10 tubing connected to the carotid artery, and differential leukocyte counts were measured from this aliquot using an automated blood cell counter (Hemavet 850FS, CDC Technologies, Oxford, CT).
Intravital microscopy
The cremaster muscle was prepared for intravital microscopy as described previously [33
, 34
]. The cremaster muscle, epididymis, and testis were exteriorized and pinned to the surgical stage. An incision was made along the length of the cremaster so that the muscle could be splayed. The epididymis and testis were pinned to the side to expose the cremaster muscle microcirculation. During surgery, the cremaster was superfused with bicarbonate-buffered saline (131.9 mM NaCl, 4.7 mM KCl, 18 mM NaHCO3, 1.2 mM MgCl2, and 2.0 mM CaCl2·2H2O) and equilibrated with 5% CO2 in N2. In this model, >95% of all adherent and rolling leukocytes are neutrophils [35
].
Brightfield intravital microscopy was performed using a Zeiss intravital microscope with a salt water (SW) 40/0.75 saline immersion objective. Three venules per mouse with diameters between 20 µm and 50 µm were recorded using a charged-coupled device (CCD) camera (Model VE-1000CD, Dage-MTI) on a Panasonic S-VHS recorder. Adherent cells were counted in the vessel 1 min before and 1 min after injection of 600 ng CXCL1 or 0.1 µM fMLP i.v. Only the cells that were adherent at the 1 min time-point were included. The number of adherent cells was divided by the surface area of the vessel wall seen in the field of view (FOV). Surface area (S) was calculated for each vessel, using S = 
*d*lv, where d is the diameter, and lv is the length of the vessel. Rolling flux was measured as the number of cells that rolled past a line perpendicular to the vessel axis per minute. To compare leukocyte-rolling flux in venules of different diameters, rolling flux was normalized to an average vessel width of 50 µm.
Blood-perfused micro-flow chamber
Micro-flow chambers were constructed from 20 x 200 µm rectangular glass capillaries (VitroCom, Mountain Lake, NJ) as described [36
]. Rectangular glass capillaries (30 mm), cut with a capillary cutting stone (Hampton Research, Aliso Viejo, CA), were placed between two plastic microscope coverslips, allowing 
5 mm of the capillary to extend beyond the edge of the coverslips. The slide covers were attached to a glass microscope slide using clear fingernail polish. P-selectin (18 µg/ml), ICAM-1 (15 µg/ml), and CXCL1 (15 µg/ml) were diluted in PBS and allowed to adsorb onto the glass surface for 2 h at room temperature alone or as mixtures. Heparinized PE 50 tubing (ID 0.58 mm, OD 0.965 mm, Becton Dickinson, Sparks, MD), 5 cm in length, was inserted over the end of one side of the capillary tubing, and the joint was sealed with epoxy gel resin (Loctite, Manco Inc., Avon, OH). The chambers were washed with saline and incubated with 10% casein (Pierce Chemicals, Dallas, TX) in PBS for 1 h at room temperature. The free end of the chamber was inserted into a piece of saline-filled PE 50 tubing, and the joint was glued with epoxy gel resin. The saline-filled PE 50 tubing was used to control the wall shear stress of the blood flow. The upstream PE 50 tubing was connected to the PE 10 inserted into the carotid artery of the mouse. Microscopy was conducted using a Zeiss Axioskop (Carl Zeiss, Inc., Thornwood, NY) with a saline immersion objective (SW 20/0.5). Images were recorded with a 3-CCD color video camera (Model DXC-390, Sony Corp., Japan) connected to a Panasonic S-VHS recorder.
Blood was allowed to perfuse through the flow chamber for 6 min. After the 6th min, one FOV was recorded from the center of each capillary chamber for 1 min. Arrested cells were defined as those cells that were adherent for 30 s.
Bone marrow transplants
Male, wild-type mice were irradiated with a split dose of 1200 rads over 3 h. The mice were then reconstituted with 5 x 106 bone marrow cells from female PI-3K–/– mice by lateral tail vein injections. Reconstitution was verified by RT-PCR using primers for the y-chromosome [37
]. The chimeric mice were used after 6 weeks.
Statistics
Data are presented as the mean ± SEM of three to 10 independent experiments. An ANOVA followed by a Tukey’s multiple comparisons test was used with the null hypothesis rejected at P< 0.05.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (38K): [in a new window] |
Figure 1. Rolling and adhesion in cremaster muscle postcapillary venules of PI-3K–/– and LY294002-treated mice. (A) Adherent leukocytes in LY294002-treated mice (open bars) and vehicle control-treated mice (solid bars) after injection of 600 ng CXCL1. (B) Adherent leukocytes or (C) rolling flux in PI-3K–/– mice (open bars) and littermate control wild-type mice (solid bars) after injection of 600 ng CXCL1. Data presented are the mean ± SEM from at least 10 vessels from three mice. *, P < 0.05, compared with vessels prior to CXCL1 injection; #, P < 0.05, compared with littermate control wild-type mice. (D) Representative pictures of cremaster muscle postcapillary venules in wild-type and PI-3K–/– mice before and 1 min after CXCL1 injection.
|
has been shown to signal downstream of G protein-coupled chemokine receptors [21
], we next determined if PI-3K is required for leukocyte adhesion in the cremaster muscle venules. CXCL1 induced leukocyte adhesion in wild-type mice (279±50 adherent cells per mm2 before to 583±98 after; Fig. 1B
); however, in PI-3K null mice, CXCL1-induced adhesion was reduced drastically (66±27 adherent cells per mm2 before to 162±55 after; Fig. 1B
). Representative video micrographs of wild-type and PI-3K–/– mice before and after CXCL1 treatment are shown in Figure 1D
. It is important to note that trauma-induced adhesion in PI-3K null mice was significantly less than in wild-type mice, suggesting that PI-3K signaling is important for this adhesion, even without chemokines added exogenously. Furthermore, this defect in adhesion is likely a result of signaling, as neutrophils from wild-type and PI-3K do not differ in adhesion molecule and chemokine expression (data not shown). As expected, rolling flux was decreased in wild-type mice upon chemokine injection. This effect was less pronounced in PI-3K–/– mice (Fig. 1C)
. A decrease in rolling is often seen when more cells adhere [39
]. These data indicate that PI-3K–/– is required for CXCL1-dependent leukocyte adhesion.
PI-3K-deficient leukocytes adhere transiently but cannot remain attached
In wild-type mice, leukocytes adhered rapidly to the vessel wall after injection of CXCL1 and remained attached over time (Fig. 2A
). However, in PI-3K–/– mice, leukocytes did not attach or attached transiently (Fig. 2B)
. This was illustrated further when individual leukocytes were tracked within a given vessel. Individual leukocytes within a vessel of a CXCL1-treated wild-type mouse adhered rapidly and remained attached to the vessel (Fig. 2C)
. In contrast, individual leukocytes within a vessel of a CXCL1-treated PI-3K null mouse adhered rapidly but only 50% remained firmly attached over time (Fig. 2D)
. This suggests that firm adhesion is defective in these mice.
 View larger version (23K): [in a new window] |
Figure 2. Tracking of leukocytes in postcapillary venules of littermate-control wild-type and PI-3K–/– mice after injection of CXCL1. The change in the number of adherent leukocytes in postcapillary venules of (A) littermate-control wild-type mice and (B) PI-3K–/– mice tracked every 5 s. Time (0) begins when soluble CXCL1 was injected into the carotid artery of each mouse. Each symbol represents one venule. Length-of-time adherent leukocytes remained bound to the endothelium in one representative postcapillary venule of (C) a wild-type and (D) a PI-3K–/– mouse. Horizontal lines indicate no displacement = adhesion.
|
is required for adhesion in an auto-perfused micro-flow chamber. Blood-perfused micro-flow chambers [36
] were coated with P-selectin (20 µg/ml) and ICAM (15 µg/ml), with or without CXCL1 (15 µg/ml), and connected to the carotid artery of PI-3K–/– or wild-type mice. This combination of adhesion molecules and chemokines has been demonstrated previously to mediate leukocyte adhesion [41
]. In matched littermate wild-type control mice, CXCL1-coated flow chambers increased leukocyte adhesion significantly compared with flow chambers coated without CXCL1. However, in PI-3K null mice, CXCL1 did not induce leukocyte adhesion (Fig. 3B
), although the number of cells rolling on P-selectin was similar to wild-type mice (Fig. 3A)
. To ensure that rolling and adhesion were not a result of nonspecific interactions, control chambers were coated with P-selectin, ICAM-1, P-selectin and CXCL1 or ICAM-1 and CXCL1. As expected, chambers with P-selectin supported rolling but not adhesion, and chambers coated with ICAM-1 did not support rolling or adhesion (Fig. 3A
and 3B)
. Therefore, these data indicate that leukocyte PI-3K is required for chemokine-triggered adhesion in flow chambers with a defined substrate.
 View larger version (19K): [in a new window] |
Figure 3. The number of rolling and adherent leukocytes from wild-type and PI-3K–/– mice in a blood-perfused micro-flow chamber. (A) Rolling and (B) adherent leukocytes in an auto-perfused flow chamber (2.5 dynes/cm2) coated with P-selectin (20 µg/ml) and ICAM-1 (15 µg/ml), with or without CXCL1 (15 µg/ml). Data presented are the mean ± SEM from at least nine flow chambers. *, P < 0.05, compared with flow chambers coated with P-selectin and ICAM-1 alone.
|
-deficient leukocytes show an adhesion defect to PI-3K expressing endothelial cells in response to CXCL1 in postcapillary venules of the cremaster muscle is required for CXCL1-induced adhesion. However, in vivo, we found a reduced number of adherent leukocytes in vessels prior to CXCL1 stimulation. To address if endothelial cell PI-3K plays a role in trauma-induced adhesion, bone marrow from PI-3K null mice was transplanted into lethally irradiated wild-type mice. Leukocyte adherence in response to CXCL1 was examined in postcapillary venules of the cremaster muscle 6 weeks later. Complete bone marrow reconstitution was verified by RT-PCR (data not shown). CXCL1 induced leukocyte adhesion in wild-type mice (279±50 adherent cells per mm2 before to 583±98 after); however, in wild-type mice with PI-3K-deficient leukocytes, CXCL1-induced adhesion was reduced significantly (86±34 adherent cells per mm2 before to 288±74 after; Fig. 4A
). This reduction in adhesion is statistically significant from wild-type mice but not as severe as in PI-3K null mice, suggesting that there may also be an endothelial cell component to CXCL1-induced adhesion. Similar to what was seen in PI-3K–/– mice, baseline leukocyte adhesion was also reduced significantly from 279 ± 50 to 86 ± 34 cells per mm2. It is interesting that the reduction in leukocyte rolling upon CXCL1 injection was similar in PI-3K chimeric mice (66±11 before compared with 25±5 after) and wild-type mice (62±7 compared with 29±7; Fig. 4B
), suggesting that leukocyte PI-3K is not required for reduced rolling after chemokine exposure. Taken together, these data show that leukocyte PI-3K signaling is important for trauma- and CXCL1-induced adhesion.
 View larger version (16K): [in a new window] |
Figure 4. Rolling and adhesion in cremaster muscle postcapillary venules of wild-type mice and wild-type mice transplanted with bone marrow from PI-3K–/– mice. (A) Adherent leukocytes or (B) rolling flux (cells/min in a 50 µm wide vessel) of leukocytes in wild-type (solid bars) or irradiated, wild-type mice transplanted with bone marrow from PI-3K–/– mice (open bars) after injection of 600 ng CXCL1. Data presented are the mean ± SEM from at least nine vessels from three mice. *, P < 0.05, compared with vessels prior to CXCL1 injection; #, P < 0.05, compared with wild-type mice.
|
null mice. Therefore, leukocyte adhesion in wild-type and LFA-1 null mice [32
] was analyzed by intravital microscopy. Similar to the PI-3K null mice, LFA-1 null mice had decreased trauma-induced (148±49 compared with 279±50) and CXCL1-induced adhesion significantly (213±70 compared with 583±98; Fig. 5A
); however, LFA-1 deficiency did not totally abolish adhesion in these mice, suggesting that LFA-1-independent mechanisms for adhesion exist. In most LFA-1–/– mice, adherent cells did not accumulate over time (Fig. 5B)
, underscoring the importance of LFA-1 in CXCL1-induced adhesion. To test whether LFA-1 is involved in the residual leukocyte adhesion in PI-3K–/– mice, we treated these mice with a function-blocking mAb to LFA-1. This treatment reduced neutrophil adhesion to baseline (Fig. 5C)
.
 View larger version (16K): [in a new window] |
Figure 5. Rolling and adhesion in cremaster muscle postcapillary venules of LFA-1–/– and wild-type mice. (A) Adherent leukocytes in LFA-1–/– (open bars) and wild-type mice (solid bars) after injection of 600 ng CXCL1. Data presented are the mean ± SEM from at least 10 vessels from three mice. *, P < 0.05, compared with vessels prior to CXCL1 injection; #, P < 0.05, compared with wild-type mice. (B) The change in the number of adherent leukocytes in postcapillary venules of LFA-1–/– mice tracked every 5 s. Time (0) begins when soluble CXCL1 was injected into the carotid artery of each mouse. Each symbol represents one venule. (C) A blocking antibody to LFA-1 was injected into P13K–/– mice (1 µg/g body weight; open bars). After 30 min when the antibody had time to bind, and the effects of injected CXCL1 were gone, another 600 ng CXCL1 was administered. The number of leukocytes adhered 1 min before and 1 min after was recorded in the same vessel visualized previously. Data presented are the mean ± SEM from five vessels from five mice. *, P < 0.05, compared with vessels prior to CXCL1 injection.
|
null mice cannot induce leukocyte adhesion in postcapillary venules of the cremaster muscle in response to fMLP null mice have a defect in CXCL1-induced adhesion, we tested whether PI-3K may be a signaling molecule for leukocyte adhesion triggered by other chemoattractants. To address this, leukocyte attachment in PI-3K null mice was analyzed using intravital microscopy. fMLP induced leukocyte adhesion in wild-type mice (232±68 adherent cells per mm2 before to 633±126 after; Fig. 6A
); however, in PI-3K null mice, fMLP-induced adhesion was reduced drastically (51±17 adherent cells per mm2 before to 92±27 after; Fig. 6A
). Examination of leukocyte attachment showed that increasing numbers of leukocytes in wild-type mice adhered firmly after fMLP injection (Fig. 6B)
. In contrast, leukocytes in vessels of fMLP-treated PI-3K–/– mice only attached transiently (Fig. 6C)
. These data suggest that PI-3K is also required for fMLP-induced leukocyte adhesion.
 View larger version (17K): [in a new window] |
Figure 6. Rolling and adhesion in cremaster muscle postcapillary venules of PI-3K–/– and wild-type mice treated with fMLP. (A) Adherent leukocytes in PI-3K–/– mice (open bars) and wild-type mice (solid bars) after injection of 0.1 µM fMLP. Data presented are the mean ± SEM from at least six vessels from three mice. *, P < 0.05, compared with vessels prior to fMLP injection; #, P < 0.05, compared with wild-type mice. The change in the number of adherent leukocytes in postcapillary venules of (B) wild-type mice or (C) PI-3K–/– mice tracked every 5 s. Time (0) begins when soluble fMLP was injected into the carotid artery of each mouse. Each symbol represents one venule.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
has been implicated previously in neutrophil migration and superoxide production [22
23
24
], its role in adhesion of rolling neutrophils has not been investigated. Furthermore, as chemokine signaling through G protein-coupled receptors induces leukocyte adhesion, and PI-3K is activated downstream of the β subunit of G protein-coupled receptors, we hypothesized that PI-3K-deficient mice would have a defect in chemokine-induced leukocyte adhesion. Using intravital microscopy to examine leukocyte adhesion to the microvessels of the cremaster muscle, we demonstrate that a lack of PI-3K in mice blocked trauma-, CXCL1-, and fMLP-induced adherence of leukocytes to the vessel wall. Furthermore, using a blood-perfused micro-flow chamber and bone marrow transfer experiments, we demonstrate that the inability of the leukocytes to adhere to the vessel wall is a result of a signaling defect within the leukocyte.
Leukocyte adhesion occurs by multiple sequential events. A leukocyte must first be captured from free flow, roll, and then loosely attach to the endothelium. After loosely attaching, the leukocyte strengthens its adhesion by outside-in signals that mediate cytoskeletal rearrangements, resulting in integrin clustering. As rolling leukocytes in PI-3K–/– and LY294002-treated mice show transient adhesion that can be blocked by anti-LFA-1 antibodies, it is unlikely that PI-3K is involved in LFA-1 activation by conformational change. Rather, the role of leukocyte PI-3K seems to be involved in strengthening the adhesion and keeping the leukocytes in place under flow. The transient adhesion seen in PI-3K–/– mice is similar to what has been shown for Vav1/3 null mice, which have impaired activation of Rho family GTPases [43
]. In PI-3K and Vav KO mice, the adhesion defect is likely a result of the lack of cytoskeletal rearrangements and thus, inhibition of integrin mobility. Ma et al. [44
] demonstrated that transfected COS cells induced fMLP-stimulated actin cytoskeleton restructuring through PI-3K and Vav signaling downstream of the β subunit of the G protein-coupled fMLP receptor.
Studies using different densities of the LFA-1 ligand ICAM-1 also confirm a role for PI-3K in interin clustering. Constantin and co-workers [11
] demonstrated that inhibition of PI-3K blocked lymphocyte adhesion on low-site densities of ICAM-1; however, high-site densities of ICAM-1 could overcome the requirement for PI-3K. These data are in agreement with the data presented here in neutrophils. The present dataset identifies PI-3K as the isoform responsible for this effect. In our studies, the PI-3K–/– mice are not treated with proinflammatory cytokines, and consequently, ICAM-1 expression is at baseline levels, and PI-3K is required for firm adhesion. Given these data, it is likely that the leukocytes in the PI-3K null mice cannot adhere firmly to the endothelial cells because of their inability to redistribute integins properly on the cell surface.
A phenotype similar to that described here in PI-3K–/– mice was described recently in mice lacking the src family members hck and fgr [44
]. Although all chemoattractant-induced neutrophil function tested was normal in these mice, including arrest on fibrinogen, adherent hck–/– fgr–/– neutrophils failed to show sustained adhesion. However, the time course of adhesion failure was different: hck–/– fgr–/– neutrophils detached after 4–6 min of chemoattractant activation. PI-3K–/– neutrophils detached within 1–2 min of chemokine. This difference suggests that PI-3K may be upstream of hck and fgr signaling. Indeed, hck and fgr are known to be involved in postadhesion, β2-integrin-induced, outside-in signaling [45
]. Conclusive comparison of the roles of PI-3K and src kinases in adhesion will require side-by-side testing.
Although little is known about endothelial cell function in leukocyte arrest, some data suggest that the PI-3K family may play a role. It has been shown previously that endothelial PI-3K
and PI-3K are required for leukocyte capture in TNF--stimulated microvessels of the cremaster muscle of mice [31
, 40
]. In these studies, the authors failed to find a role for leukocyte PI-3K. Our flow chamber data and bone marrow transplant experiments show unequivocally that PI-3K null leukocytes have a significant and previously unidentified adhesion defect. However, the adhesion defect in chimeric mice is not as dramatic as in PI-3K–/– mice, which confirms a role of endothelial cell PI-3K as proposed by Puri et al. [40
]. Leukocyte PI-3K and PI-3K are not required for P- and E-selectin-dependent capture in flow chambers [40
], which agrees with our flow chamber data that PI-3K-deficient leukocytes roll equally as well as wild-type leukocytes.
In summary, this report furthers our understanding of signaling mechanisms required for leukocyte adhesion. This paper identifies a key role of leukocyte PI-3K in mediating CXCL1- and fMLP-induced leukocyte adhesion in vivo. Future studies will be important to determine if this is a universal signaling pathway for adhesion of all leukocyte subsets and if all arrest chemokines signal through PI-3K.
Received March 29, 2006; revised July 21, 2006; accepted August 15, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
4 integrin affinity for vascular cell adhesion molecule 1 which mediates arrest: an early step in the process of emigration J. Exp. Med. 193,1149-1158Lβ2 J. Cell Biol. 167,1241-1253 sensitivity of a PI3K is dependent upon a tightly associated adaptor, p101 Cell 89,105-114[CrossRef][Medline] subunits Cell 77,83-93[CrossRef][Medline] in thymocyte development, T cell activation, and neutrophil migration Science 287,1040-1046 in inflammation Science 287,1049-1053 in chemoattractant-mediated signal transduction Science 287,1046-1049-deficient mice EMBO J. 23,3505-3515[CrossRef][Medline] in macrophage migration to MCP-1 and CSF-1 Exp. Cell Res. 290,120-131[CrossRef][Medline], Rac, and PAK signaling in chemokine-induced macrophage migration J. Biol. Chem. 279,43273-43284 suppresses joint inflammation and damage in mouse models of rheumatoid arthritis Nat. Med. 11,936-943[Medline] in platelet activation Blood 106,110-117 in promoting neutrophil trafficking into inflamed tissue Blood 103,3448-3456 activity in neutrophil trafficking Blood 106,150-157, a Rac guanosine exchange factor, and Rac Mol. Cell. Biol. 18,4744-4751This article has been cited by other articles:
 |
|
 |
|
  B. Sarraj, S. Massberg, Y. Li, A. Kasorn, K. Subramanian, F. Loison, L. E. Silberstein, U. von Andrian, and H. R. Luo Myeloid-Specific Deletion of Tumor Suppressor PTEN Augments Neutrophil Transendothelial Migration during Inflammation J. Immunol., June 1, 2009; 182(11): 7190 - 7200. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Gavard, X. Hou, Y. Qu, A. Masedunskas, D. Martin, R. Weigert, X. Li, and J. S. Gutkind A Role for a CXCR2/Phosphatidylinositol 3-Kinase {gamma} Signaling Axis in Acute and Chronic Vascular Permeability Mol. Cell. Biol., May 1, 2009; 29(9): 2469 - 2480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Mayer, A. Kiessling, J. Ott, M. B. Schaefer, M. Hecker, I. Henneke, R. Schulz, A. Gunther, J. Wang, L. Wu, et al. Acute Lung Injury Is Reduced in fat-1 Mice Endogenously Synthesizing n-3 Fatty Acids Am. J. Respir. Crit. Care Med., March 15, 2009; 179(6): 474 - 483. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Zarbock, C. L. Abram, M. Hundt, A. Altman, C. A. Lowell, and K. Ley PSGL-1 engagement by E-selectin signals through Src kinase Fgr and ITAM adapters DAP12 and FcR{gamma} to induce slow leukocyte rolling J. Exp. Med., September 29, 2008; 205(10): 2339 - 2347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Thomas, J. S. Mitchell, C. C. DeNucci, A. L. Martin, and Y. Shimizu The p110{gamma} isoform of phosphatidylinositol 3-kinase regulates migration of effector CD4 T lymphocytes into peripheral inflammatory sites J. Leukoc. Biol., September 1, 2008; 84(3): 814 - 823. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Frommhold, A. Ludwig, M. G. Bixel, A. Zarbock, I. Babushkina, M. Weissinger, S. Cauwenberghs, L. G. Ellies, J. D. Marth, A. G. Beck-Sickinger, et al. Sialyltransferase ST3Gal-IV controls CXCR2-mediated firm leukocyte arrest during inflammation J. Exp. Med., June 9, 2008; 205(6): 1435 - 1446. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Kerfoot, G. Andonegui, C. S. Bonder, and L. Liu Exogenous stromal cell-derived factor-1 induces modest leukocyte recruitment in vivo Am J Physiol Heart Circ Physiol, June 1, 2008; 294(6): H2524 - H2534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Chavakis, G. Carmona, C. Urbich, S. Gottig, R. Henschler, J. M. Penninger, A. M. Zeiher, T. Chavakis, and S. Dimmeler Phosphatidylinositol-3-Kinase-{gamma} Is Integral to Homing Functions of Progenitor Cells Circ. Res., April 25, 2008; 102(8): 942 - 949. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Heit, L. Liu, P. Colarusso, K. D. Puri, and P. Kubes PI3K accelerates, but is not required for, neutrophil chemotaxis to fMLP J. Cell Sci., January 15, 2008; 121(2): 205 - 214. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Zarbock and K. Ley Mechanisms and Consequences of Neutrophil Interaction with the Endothelium Am. J. Pathol., January 1, 2008; 172(1): 1 - 7. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Pinho, R. de Castro Russo, F. A. Amaral, L. P. de Sousa, M. M. Barsante, D. G. de Souza, J. C. Alves-Filho, D. C. Cara, J. S. Hayflick, C. Rommel, et al. Tissue- and Stimulus-Dependent Role of Phosphatidylinositol 3-Kinase Isoforms for Neutrophil Recruitment Induced by Chemoattractants In Vivo J. Immunol., December 1, 2007; 179(11): 7891 - 7898. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Zarbock, T. L. Deem, T. L. Burcin, and K. Ley G{alpha}i2 is required for chemokine-induced neutrophil arrest Blood, November 15, 2007; 110(10): 3773 - 3779. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Liu, K. D. Puri, J. M. Penninger, and P. Kubes Leukocyte PI3K{gamma} and PI3K{delta} have temporally distinct roles for leukocyte recruitment in vivo Blood, August 15, 2007; 110(4): 1191 - 1198. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
