IL-4 is known to induce recruitment of eosinophils and mononuclear leukocytes. In vitro this occurs in part by selective expression of VCAM-1, the ligand for the α4 integrin. The objective of this study was to determine the molecular mechanisms that underlie IL-4-induced leukocyte recruitment in vivo. Mice received an intrascrotal injection of IL-4 (100 ng). Twenty-four hours later, leukocyte rolling, adhesion, and emigration in cremasteric postcapillary venules were examined via intravital microscopy, and expression of VCAM-1 and P- and E-selectin was quantitated using a radiolabeled mAb technique. IL-4 increased VCAM-1 expression, but P-selectin and E-selectin remained at constitutive levels. IL-4 induced significant increases in leukocyte adhesion and emigration, with 50% of the emigrated cells being eosinophils and the remainder being mononuclear leukocytes. Leukocyte rolling in IL-4-treated mice was >95% inhibitable using an anti-P-selectin Ab. However, IL-4-induced leukocyte recruitment was unaltered in mice treated chronically with P-selectin Ab or mice deficient in either P-selectin or P- and E-selectin, suggesting that the residual rolling supported all of the IL-4-induced recruitment. In IL-4-treated mice following P-selectin blockade, tethering and rolling were not dependent on L-selectin, but were abolished by α4 integrin blockade. These findings show that the α4 integrin can initiate leukocyte-endothelial cell interactions in the absence of selectins under shear conditions in vivo, and that the absence of selectins does not affect recruitment of eosinophils and mononuclear cells to IL-4-treated tissue.
The proinflammatory capabilities of IL-4 are worthy of detailed investigation, as there is convincing evidence implicating this cytokine as a crucial factor in allergic disease. High levels of IL-4 are observed in conditions such as allergic rhinitis, atopic dermatitis, and asthma, and are closely associated with recruitment of eosinophils and CD4+ T cells (1, 2, 3). In animal models of allergy and asthma, mice deficient in IL-4 have reduced pulmonary eosinophil recruitment and absence of airway hyperresponsiveness in response to aerosol Ag challenge (4, 5), suggesting a key role for IL-4 in eosinophil recruitment. Consistent with these observations is the fact that IL-4 injection induces recruitment of eosinophils and mononuclear leukocytes in a variety of species (6, 7, 8, 9). In two of these studies, IL-4 has been also shown to induce expression of the endothelial adhesion molecule VCAM-1, but whether VCAM-1 was responsible for the leukocyte recruitment remained unclear.
Leukocyte recruitment is a multistep cascade that involves initial selectin-dependent leukocyte tethering (attachment) to endothelium, which in vivo can occur at shear stresses between 2 and 15 dynes/cm2, followed by selectin-dependent leukocyte rolling and finally integrin-dependent leukocyte adhesion. From this simplistic model, one might predict that the integrin pathway, α4 integrin/VCAM-1, would mediate firm adhesion. However, this is far from incontestable, as in vitro the α4 integrin/VCAM-1 pathway has been shown to mediate leukocyte tethering, rolling, and adhesion, albeit at the lower end (0.7–2 dynes/cm2) of physiologic shear (10, 11). Other in vitro studies have proposed initial selectin-dependent leukocyte tethering and rolling, followed by subsequent α4 integrin/VCAM-1-dependent leukocyte adhesion (12, 13, 14). This group of in vitro studies is more consistent with in vivo data from two inflammatory models showing that selectins were absolutely required for the initial attachment or tethering of leukocytes to endothelium, whereas the α4 integrin/VCAM-1 pathway was shown to play a role in the subsequent rolling and adhesion (15, 16).
To date no one has examined whether a mediator such as IL-4, which presumably induces expression of high levels of VCAM-1 and only limited amounts, if any, of the selectins, induces leukocyte recruitment via the α4 integrin/VCAM-1 pathway exclusively or whether selectins are also involved. The adhesive mechanisms underlying IL-4-induced leukocyte recruitment have generally been derived from in vitro experiments. Initial studies using human umbilical vein endothelium have consistently reported that IL-4 induced expression of VCAM-1 independent of E-selectin, P-selectin, or other adhesion molecules, and this was sufficient to tether eosinophils and mononuclear leukocytes via the α4 integrin/VCAM-1 pathway, albeit at the lower range of physiologic shear (17, 18, 19, 20). Others have proposed that selectin pathways could be up-regulated on IL-4-stimulated HUVEC, which were then required for initial tethering before VCAM-1-dependent leukocyte adhesion (12, 21). Only one in vivo study to date has demonstrated a role for the α4 integrin/VCAM-1 pathway in IL-4-induced leukocyte recruitment (9). Sanz et al. used elicited, indium-labeled eosinophils to demonstrate that these cells infiltrated IL-4-stimulated skin and 60% of this recruitment was inhibited with an α4 integrin Ab. A number of key questions have been raised by this study. First, the importance (if any) of selectins in IL-4-induced leukocyte recruitment needs investigation, and second, the mechanism by which α4 integrin/VCAM-1 pathway contributes to IL-4-induced tethering/rolling, adhesion, or all mechanisms of leukocyte recruitment also requires attention.
In this study, we made use of intravital microscopy that permits direct visualization of the leukocyte recruitment cascade, to systematically elucidate the mechanisms of IL-4-induced leukocyte recruitment in the microcirculation. The results reveal that in wild-type mice, baseline levels of P-selectin supported ∼95–98% of all rolling cells following IL-4 treatment. However, chronic immunoneutralization of P-selectin with Ab or P-selectin gene deletion (or even P-selectin/E-selectin gene deletion) had no impact on the ultimate leukocyte accumulation in tissue. The α4 integrin was entirely responsible for the few tethering, rolling, and adhering cells in E/P-selectin-deficient mice that underlie the eosinophil and mononuclear cell recruitment in IL-4-treated tissue. These findings also demonstrate that in peripheral microvessels IL-4 treatment allows the α4 integrin to mediate the entire leukocyte-endothelial cell cascade under physiological shear conditions in vivo.
Materials and Methods
Male C57BL/6 wild-type mice and P-selectin-deficient mice on the C57BL/6 background were obtained from The Jackson Laboratory (Bar Harbor, ME). Mice deficient in both P-selectin and E-selectin (P-sel/E-sel−/−) were generously provided by Dr. Dan Bullard (University of Alabama, Birmingham). These mice were generated on a mixed background of 129/Sv × C57BL/6 (22), and have been backcrossed onto a C57BL/6 background for six generations. All mice were used between 6–10 wk of age.
Abs and cytokines
The Abs used in this study were: RB40.34, an anti-murine P-selectin mAb (PharMingen, San Diego, CA); R1-2, an anti-murine α4 integrin mAb (PharMingen); 9C10, a mAb that recognizes the murine α4 integrin, but does not block function (PharMingen); MK-1.9.1, a rat IgG1 against murine VCAM-1 (Bayer Laboratories, West Haven, CT); 9A9, an anti-murine E-selectin Ab (generously provided by Dr. Barry Wolitzky, Hoffman LaRoche Pharmaceuticals, Nutley, NJ); and P-23, a murine IgG1 against human P-selectin (Pharmacia & Upjohn, Kalamazoo, MI). Murine IL-4 was purchased from R&D Systems (Minneapolis, MN).
The mouse cremaster preparation was used to study the behavior of leukocytes in the microcirculation (23). Mice were anesthetized by i.p. injection of a mixture of xylazine hydrochloride (10 mg/kg; MTC Pharmaceuticals, Cambridge, Ontario, Canada) and ketamine hydrochloride (200 mg/kg; Rogar/STB, London, Ontario, Canada). The jugular vein was cannulated and used to administer additional anesthetic and Abs. The cremaster muscle was dissected free of tissues and exteriorized onto an optically clear viewing pedestal. The muscle was cut longitudinally with a cautery and held flat against the pedestal by attaching silk sutures to the corners of the tissue. The muscle was then superfused with bicarbonate-buffered saline.
An intravital microscope (Axioskop; Carl Zeiss Canada, Don Mills, Ontario, Canada) with a ×25 objective lens (Wetzlar L25/0.35; E. Leitz, Munich, Germany) and a ×10 eyepiece was used to examine the cremasteric microcirculation. A video camera (Panasonic 5100 HS, Osaka, Japan) was used to project the images onto a monitor, and the images were recorded for playback analysis using a videocassette recorder. Single unbranched cremasteric venules (25–40 μm in diameter) were selected, and to minimize variability, the same section of cremasteric venule was observed throughout the experiment. The number of rolling and adherent leukocytes was determined off-line during video playback analysis. Rolling leukocytes were defined as those cells moving at a velocity less than that of erythrocytes within a given vessel. Leukocyte rolling velocity was determined by measuring the time required for a leukocyte to roll along a 100 μm length of venule. Rolling velocity was determined for 20 leukocytes at each time interval. Leukocytes were considered adherent to the venular endothelium if they remained stationary for 30 s or longer. Leukocyte emigration was defined as the number of extravascular leukocytes per microscopic field of view (×25 objective lens), and was determined by averaging the data derived from four to five fields adjacent to postcapillary venules. Leukocyte tethering was quantitated as the number of new leukocyte-endothelium interactions initiated over a 1-min period within a 100 μm length of venule. Venular diameter (Dv) was measured on-line using a video caliper (Microcirculation Research Institute, Texas A&M University, College Station, TX). Centerline RBC velocity (VRBC) was also measured on-line using an optical Doppler velocimeter (Microcirculation Research Institute), and mean RBC velocity (Vmean) was determined as VRBC/1.6. Venular wall shear rate (γ) was calculated based on the Newtonian definition: γ = 8(Vmean/Dv) (24).
Transmission electron microscopy
Cremaster specimens for transmission electron microscopy (TEM) were rapidly cut into 2-mm3 blocks and fixed by immersion in freshly prepared 4% glutaraldehyde in Millonig’s buffer (all materials for electron microscopy were supplied by Electron Microscopy Sciences, Cedar Lane Laboratories, Hornby, Ontario, Canada) (pH 7.25) overnight. Following washing in buffer and 1 h of postfixation in 1% osmium tetroxide in distilled water, the tissue was prestained in 2% uranyl acetate, then dehydrated through graded ethanol, embedded in Quetol resin, and polymerized at 60°C for 48 h. Areas of tissue containing extravascular leukocytes were identified by light microscopy in toluidine blue-stained 1-μm sections cut from randomly selected blocks. Thin sections (60–90 nm) of these areas were prepared, stained with 2% uranyl acetate and 0.35% mM lead citrate, and examined in a Hitachi H7000 electron microscope. Approximately 20 extravascular leukocytes were identified and classified per animal.
Quantitation of expression of VCAM-1, P-selectin, and E-selectin
Expression of the adhesion molecules VCAM-1, P-selectin, and E-selectin was quantified using a modified dual-radiolabeled Ab technique (25, 26). The Abs MK1.9.1 (against VCAM-1), RB40.34 (against P-selectin), 9A9 (against E-selectin), and P-23 (a murine IgG1 against human P-selectin) were labeled with either 125I (MK1.9.1, RB40.34, 9A9) or 131I (P-23) using the iodogen method, as previously described (25, 26). P-23 was used to detect nonspecific binding in the murine system. To measure VCAM-1, mice were injected with 10 μg 125I-labeled anti-VCAM-1 (MK1.9.1), 20 μg unlabeled anti-VCAM-1 (MK1.9.1), and a variable dose of 131I-labeled nonbinding Ab (P-23) calculated to achieve a total injected 131I activity of 400,000–600,000 cpm (total volume 200 μl). This Ab combination was chosen after pilot experiments, conducted over a range of doses of unlabeled MK1.9.1, showed that this protocol ensured receptor saturation under stimulated conditions.
To study P-selectin or E-selectin, animals were injected i.v. with a mixture of either 10 μg 125I-anti-labeled P-selectin (RB40.34) or 10 μg 125I-anti-labeled E-selectin (9A9), respectively, and a variable dose of 131I-labeled P-23, as above. The Abs were allowed to circulate for 5 min, then the animals were heparinized. A blood sample was obtained from a carotid artery catheter, then the mice were exsanguinated by blood withdrawal through the carotid artery catheter and simultaneous i.v. infusion with bicarbonate-buffered saline. The cremaster muscles were harvested and weighed, along with lungs, heart, and areas of skeletal muscle for control purposes. Both 131I and 125I activity were measured in plasma and tissue samples. VCAM-1 and P-selectin expression were calculated per gram of tissue, by subtracting the accumulated activity of the nonbinding Ab (131I-labeled P-23) from the accumulated activity of the binding Ab (125I-labeled MK1.9.1, 125I-labeled RB40.34, or 125I-labeled 9A9). Data for VCAM-1, P-selectin, and E-selectin expression were represented as the percentage of the injected dose of Ab per gram of tissue. We have previously demonstrated that this approach provides reliable quantitative values of adhesion molecule expression, that radiolabeled binding Ab can be displaced specifically with sufficient amounts of unlabeled Ab, and that in the case of P-selectin, values not different from zero are obtained in P-selectin-deficient mice (25, 26).
In initial experiments, a range of doses of IL-4 (1, 10, 100 ng) in 200 μl sterile saline was injected under the scrotal skin adjacent to the cremaster muscle. Twenty-four hours later, leukocyte accumulation in the cremaster muscle was assessed using intravital microscopy. Significant leukocyte accumulation was observed at 100 ng of IL-4, so all additional experiments were performed using this dose. For these and all subsequent IL-4 experiments, including those using gene-deficient mice, control animals were injected with the same volume of normal saline and examined 24 h later. To further characterize the effects of 100 ng IL-4, the expression of 1) VCAM-1, 2) P-selectin, and 3) E-selectin in the cremaster muscle was quantitated in three separate groups of mice 24 h after IL-4 treatment.
The aim of the next series of experiments was to examine the effects of IL-4 treatment on leukocyte trafficking. At 24-h IL-4 treatment, mice were prepared for intravital microscopy, and leukocyte rolling, adhesion, and emigration were examined in the cremasteric microvasculature. In addition, cremaster muscles from some of these mice were fixed for electron microscopy to identify the types of leukocytes recruited.
In subsequent experiments using intravital microscopy, IL-4-treated mice were treated with Abs against the α4 integrin (R1-2, blocking Ab, 75 μg/mouse, i.v., or 9C10, nonblocking Ab, 75 μg/mouse, i.v.) or P-selectin (RB40.34, 20 μg/mouse, i.v.), or both, and the effects on leukocyte trafficking were recorded. To characterize the role of P-selectin over the entire 24-h IL-4 treatment period, additional mice were treated with two doses of 50 μg RB40.34 i.p., 0 and 12 h after IL-4 injection. Pilot experiments determined that this treatment protocol blocked leukocyte rolling in the unstimulated cremaster muscle (known to be P-selectin dependent) over a 24-h period. Examination of postcapillary venules in the exteriorized but otherwise untreated cremaster 12 h after RB40.34 administration revealed no rolling leukocytes. This Ab treatment protocol did not affect circulating leukocyte counts. To further characterize the role of P-selectin, a group of P-selectin-deficient mice was also treated with IL-4 and examined 24 h later.
As inhibition or absence of P-selectin could not completely inhibit IL-4-induced leukocyte rolling (only 95% effective) and subsequent leukocyte recruitment, the aim of the final series of experiments was to determine the ability of IL-4 treatment to induce leukocyte recruitment in the complete absence of the endothelial selectins. P-sel/E-sel−/− mice were treated with 100 ng IL-4 and examined using intravital microscopy 24 h later. These mice were treated with an Ab against the α4 integrin (R1-2, 75 μg/mouse, i.v.) during the microscopic observation.
Circulating leukocyte counts
At the end of each experiment, whole blood was drawn via cardiac puncture. Total leukocyte counts were performed, using a Bright-line hemocytometer (Hausser Scientific, Horsham, PA).
All data are displayed as mean ± SEM. Normally distributed data were analyzed using Student’s t test, or when nonparametric analysis was appropriate, the Mann-Whitney test was used. A value of p < 0.05 was deemed significant.
IL-4 induces eosinophil and mononuclear leukocyte recruitment
In intitial experiments, we used intravital microscopy to examine leukocyte trafficking in the IL-4-treated cremaster muscle. To determine a dose of IL-4 that was effective at inducing leukocyte recruitment, wild-type mice were treated with 1, 10, and 100 ng of IL-4. At 100 ng, but not 1 or 10 ng, IL-4 induced a significant increase in leukocyte adhesion (≈15 cells/100 μm) (Fig. 1). Furthermore, IL-4 at 100 ng induced a profound accumulation of leukocytes in the extravascular tissue (≈50 cells/microscopic field) (Fig. 1). The numbers of both adherent and extravascular leukocytes observed following IL-4 treatment were comparable with that observed in mouse cremasteric venules in an allergen model of inflammation 24 h after challenge (P. Kubes, unpublished observations), indicating that IL-4 on its own was extremely effective at inducing leukocyte recruitment. All additional experiments were performed at this dose. IL-4 caused a reduction in leukocyte rolling velocity of greater than 50% (≈65 μm/s in saline treated vs ≈30 μm/s in IL-4 treated), but leukocyte rolling flux was not affected (Fig. 2).
Electron-microscopic analysis of the IL-4-treated tissue was used to determine the type of leukocytes recruited by IL-4 (Table I). These data showed that ∼50% of the cells recruited were eosinophils (Fig. 3), with the remaining cells consisting of monocytes/macrophages and lymphocytes. Fig. 3 is an electron micrograph demonstrating the level of resolution used to assess the profile of leukocyte recruitment in IL-4-treated mice. Neutrophils were not seen in IL-4-treated tissues. In addition, within postcapillary venules, 60–70% of the leukocytes adherent on the vascular wall were eosinophils.
IL-4 increases expression of VCAM-1, but not P-selectin or E-selectin
Analysis of the effect of IL-4 on expression of VCAM-1 in the cremaster muscle revealed that after 24 h of IL-4 treatment, VCAM-1 expression was significantly increased (Fig. 4). In contrast, P-selectin expression did not change from constitutive levels. E-selectin expression was negligible in both untreated and IL-4-treated mice. These data indicate that in this model, leukocyte recruitment was occurring within a microvasculature expressing high levels of VCAM-1, only constitutive levels of P-selectin, and minimal levels of E-selectin. Previous work from our laboratory has demonstrated that the constitutive levels of P-selectin are indeed significantly higher than the degree of P-selectin Ab binding in P-selectin-deficient mice. It is also worth noting that the nonbinding Ab (P-23), infused with the binding Ab in each experiment, accounts for alterations in blood flow or microvascular permeability that may occur with IL-4 administration.
A role for P-selectin and the α4 integrin following IL-4 treatment
We next determined the adhesion molecules responsible for the leukocyte recruitment induced by IL-4. As IL-4 induced a significant increase in VCAM-1, we first examined the importance of the ligand of VCAM-1, the α4 integrin. After 24 h of IL-4 treatment, acute Ab blockade of the α4 integrin did not affect the number of rolling leukocytes (Fig. 5,A). As results from in vitro experiments suggest that the α4 integrin is able to mediate leukocyte tethering and rolling more effectively at the low end of physiological shear rates, we examined the effectiveness of α4 integrin blockade relative to shear rate within postcapillary venules. However, despite the shear rates in the venules examined ranging from ≈160–600 s−1, blockade of the α4 integrin was not observed to reduce leukocyte rolling flux significantly, regardless of shear rate (Fig. 5,B). In contrast, inhibition of the α4 integrin did cause a significant (43%) increase in leukocyte rolling velocity (Fig. 5 C).
As we had shown that P-selectin is expressed at constitutive levels in the IL-4-treated murine cremaster, and that this level of P-selectin expression is sufficient to mediate all leukocyte rolling in cremasteric venules of untreated mice (27, 28), animals were treated with an anti-P-selectin Ab 24 h after IL-4 treatment. P-selectin blockade immediately reduced leukocyte rolling to 2–3 cells/min, suggesting that P-selectin was critical for leukocyte rolling induced by IL-4 (Fig. 6). In every animal, some rolling cells were noted, an event not seen when P-selectin Ab is administered under baseline conditions in which rolling is completely inhibited (27, 28). Fig. 6 (lower panel, note change in scale) illustrates that the residual leukocyte rolling was inhibitable with a function-blocking Ab against the α4 integrin (R1-2), but not with an Ab that binds the α4 integrin but does not affect function (9C10). Neither the anti-α4-integrin Abs nor the anti-P-selectin Ab affected circulating leukocyte counts (data not shown).
P-selectin is not required for IL-4-induced leukocyte recruitment
The almost complete absence of leukocyte rolling in mice treated acutely with anti-P-selectin Ab suggested that P-selectin was critical for the overall IL-4-induced leukocyte recruitment. Therefore, to determine the importance of P-selectin over the 24-h course of IL-4, we next treated wild-type mice with IL-4 and inhibited P-selectin throughout the 24-h period via chronic Ab treatment. In these mice, leukocyte rolling was only 2–3 cells/min 24 h after IL-4 administration (Fig. 7,A), reproducing the findings from acute P-selectin blockade (Fig. 6). Treatment with additional P-selectin Ab did not affect leukocyte rolling, indicating that P-selectin remained completely immunoneutralized at the end of the 24-h Ab protocol (data not shown). Despite the almost complete blockade of leukocyte rolling, a significant amount of leukocyte adhesion was observed in P-selectin Ab-treated mice (Fig. 7,B). Most importantly, the number of leukocytes recruited to the extravascular tissue was similar to IL-4-treated animals not given anti-P-selectin Ab; more than 50 cells/field had emigrated out of the vasculature (Fig. 7,C). The chronic P-selectin Ab experiments were entirely confirmed in P-selectin-deficient mice (Fig. 7). Although rolling was dramatically reduced in these mice and adhesion was also 50% reduced, the ultimate emigration was not affected. In accordance with previous experiments, saline-injected mice in which P-selectin was inhibited or absent showed negligible (<1 cell every 4 min) leukocyte rolling and minimal levels of adhesion and emigration.
To exclude a role for E-selectin, which has been shown to have overlapping roles with P-selectin (16), P-sel/E-sel−/− mice were used. It should be noted that in these mice, circulating leukocyte counts were elevated significantly above those of wild-type mice (Table II), as previously reported (16). Nevertheless, the number of rolling cells in IL-4-treated P-sel/E-sel−/− mice was ∼2–3 cells/min, i.e., not different from wild-type mice treated with IL-4 and P-selectin Ab. Most importantly, similar to results with chronic anti-P-selectin treatment or in P-selectin-deficient mice, leukocyte emigration in IL-4-treated P-sel/E-sel−/− mice did not differ from IL-4-treated wild-type mice (Fig. 8). Electron-microscopic analysis of the emigrated leukocytes in IL-4-treated P-sel/E-sel−/− mice revealed that similar percentages of leukocyte types were present in the extravascular tissue, when compared with wild-type mice (Table I). Saline-treated P-sel/E-sel−/− mice had negligible (<1 cell every 5 min) leukocyte rolling, and minimal adhesion and emigration.
The α4 integrin is sufficient for IL-4-induced leukocyte rolling
In IL-4-treated mice, leukocyte rolling velocity was not different when P-selectin Ab was added or when the experiments were performed in P-sel/E-sel−/− mice (Fig. 9,A). However, in both wild-type mice treated chronically with P-selectin Ab, and P-sel/E-sel−/− mice, this low level of leukocyte rolling was completely dependent on the α4-integrin, as an Ab against this molecule blocked rolling (Fig. 9 B). In contrast, following P-selectin blockade, treatment of wild-type mice with the selectin-binding polysaccharide, fucoidan, did not inhibit the residual leukocyte rolling (data not shown). This finding indicated that the third selectin, L-selectin, was not important in mediating this rolling. This approach has been used previously to exclude a role for L-selectin in leukocyte recruitment in P-sel/E-sel−/− mice (29). Together these data show that the α4 integrin was responsible for sustaining leukocyte rolling, independent of P-, E-, and L-selectin.
Leukocyte tethering (i.e., initiation of leukocyte rolling) in IL-4-treated P-sel/E-sel−/− mice was examined to determine whether the α4 integrin was also responsible for initiating leukocyte rolling in these mice. Leukocyte tethering was observed in 16 of 28 venules examined, at an average rate of 1 tether/min/100 μm. Tethering was observed in venules exhibiting shear rates ranging from 156–611 s−1, and was blocked by an Ab against the α4 integrin. These findings suggest for the first time that the α4 integrin can initiate tethering in vivo independent of endothelial selectins.
Analysis of hemodynamic parameters in postcapillary venules of saline-treated mice and mice treated with 100 ng IL-4 showed that IL-4 treatment did not significantly alter the diameters of the venules examined or the venular shear rate relative to saline-treated controls (Table II). In addition, local IL-4 administration did not affect systemic circulating leukocyte counts (Table II).
In this in vivo study, we report that IL-4 induced significant increases in leukocyte adhesion and emigration, a reduction in leukocyte rolling velocity, and a significant increase in VCAM-1 expression. Although P-selectin expression was not increased by IL-4 administration, constitutive levels were responsible for ∼68 of 70 rolling cells/min in cremasteric microvessels. However, these 68 rolling cells/min do not appear to be crucial to the recruitment process, as chronic inhibition of P-selectin, or P-selectin gene deletion did not impact upon the ultimate number of emigrated leukocytes. It is clear that a P-selectin-independent mechanism of leukocyte recruitment was responsible for the 2 of 70 rolling cells/min that ultimately accounted for all of the leukocyte recruitment. Our data also reveal that the few cells rolling were not dependent on E-selectin, as the rolling persisted in the P-sel/E-sel−/− mice, in accordance with our observation that E-selectin expression was negligible following IL-4 treatment. Additionally, we ruled out L-selectin, as the addition of an L-selectin inhibitor (fucoidan) to mice following P-selectin blockade also did not impact on the residual 2 rolling cells induced by IL-4. However, an α4 integrin Ab inhibited all of the P-selectin-independent rolling. This demonstrates that following IL-4 treatment, the α4 integrin is able to induce leukocyte rolling independently of the three selectins.
Although at first glance one could argue that leukocyte rolling was not affected by IL-4, this is not the case. The 24-h exposure of the microcirculation to IL-4 did not change the flux of rolling leukocytes, but did reduce leukocyte rolling velocity by almost 50%. This is not a trivial amount inasmuch as reducing the rolling velocity by half doubles the transit time (time for a cell to traverse a given length of vessel) and therefore doubles the number of cells within the microvasculature interacting with venules at any given time. When the α4 integrin Ab was given to wild-type mice treated with IL-4, a significant reduction in leukocyte rolling was not observed, but this is not surprising and in fact consistent with only a few cells rolling via the α4 integrin per minute. What is noteworthy is that addition of the α4 integrin Ab increased leukocyte rolling velocity by 50% back to near control values. Of course, this reduces transit time of rolling cells and reduces the number of cells interacting with venules within the microvasculature. Finally, it is worthwhile noting that the number of leukocytes recruited to the cremaster 24 h after Ag challenge is approximately the same as that induced by IL-4 treatment, but in the Ag challenge model leukocyte rolling velocity is not reduced (unpublished observations). Unlike in the IL-4 experiments, blocking the endothelial selectins (for 8 h) in the Ag model is sufficient to entirely inhibit leukocyte emigration (16). Therefore, the different endothelial phenotype induced by IL-4 treatment allows leukocytes to be recruited independently of selectin function.
An unexpected finding based on previous work is that IL-4 treatment allows for leukocytes to use the α4 integrin to not only roll, but also to tether to endothelium in vivo. Tethering (initial capture or attachment) and rolling are distinct adhesive mechanisms inasmuch as rolling can occur in vitro at shear stress, where tethering does not. For example, Alon et al. demonstrated that VCAM-1 could support rolling via α4 integrin up to very high shear stresses in vitro (7 dynes/cm2), but tethering would not occur at 2 dynes/cm2, which is a shear that supports selectin-dependent tethering (11). These data suggest a requirement for selectins as tethering molecules for α4 integrin-dependent rolling to occur. Moreover, there is a lack of evidence that the α4 integrin can mediate tethering of leukocytes in vivo. In a model of Ag-induced inflammation of the cremaster microcirculation, intravital microscopy revealed that ∼50% of the leukocyte rolling was α4 integrin dependent (16). However, the rolling was entirely inhibited by an Ab against P-selectin, demonstrating an absolute requirement for this selectin. In the P-selectin-deficient mouse, the α4 integrin supported 100% of the rolling, but again all rolling could be inhibited by an E-selectin Ab. From that work, it was concluded that α4 integrin-mediated leukocyte/endothelial cell interactions required selectins for tethering and/or rolling. In the present study, IL-4 was able to induce α4 integrin-dependent leukocyte tethering and rolling in the absence of selectin function over a broad range of physiologic shear stresses in vivo. An obvious difference between the IL-4 and Ag model is that in the latter, P-selectin expression is greatly increased in response to Ag (27), but VCAM-1 expression remains near control levels (our unpublished results). This different profile of adhesion molecule expression induced by each of these treatments provides a potential explanation for the absence of a role for selectins in IL-4-induced inflammation, but a prominent role in Ag-induced inflammation.
It is tempting to conclude that the increased VCAM-1 expression was supporting the α4 integrin-dependent leukocyte recruitment. However, there is some recent evidence that IL-4-induced up-regulation of VCAM-1 does not necessarily correlate with a functional role for this molecule in leukocyte recruitment. In a model of IL-4-induced eosinophil accumulation, Larbi et al. (30) compared the functional roles of the α4 integrin and VCAM-1 in the skin and pleural cavity. They observed that while recruitment was dependent on the α4 integrin in both tissues, VCAM-1 blockade reduced recruitment in the skin, but not in the pleural cavity. These observations suggest that despite elevated expression of VCAM-1, in some tissues the α4 integrin is able to use a ligand other than VCAM-1 to mediate leukocyte recruitment. The role of VCAM-1 in the IL-4-stimulated murine cremaster remains to be fully characterized; however, our preliminary results are equivocal. Whereas VCAM-1 antisense oligonucleotides blocked 50% of the IL-4-induced leukocyte recruitment, an anti-VCAM-1 mAb (M/K-2), which binds to a single domain of VCAM-1, had no effect on the α4 integrin-dependent rolling (M. Hickey and P. Kubes, unpublished observations). The latter observation may be due to the ability of the α4 integrin to bind to multiple domains on VCAM-1 (19).
Our data also suggest that P-selectin-mediated leukocyte rolling following IL-4 treatment is of limited importance to subsequent leukocyte recruitment. Although the great majority of leukocytes rolled in IL-4-treated postcapillary venules via P-selectin, leukocyte recruitment was unaffected in mice treated with P-selectin Ab, P-selectin-deficient mice, or P-sel/E-sel−/− mice. This raises the possibility that the leukocytes that use P-selectin for rolling are not destined to undergo firm adhesion and emigration in an IL-4-treated microvasculature and preventing rolling of these cells does not impact upon the ultimate emigration. The data also suggest that the few remaining leukocytes interacting with the endothelium by tethering and rolling via α4 integrin are the population that are recruited to this tissue. Although it is impossible to determine the type of leukocyte rolling (using intravital microscopy), we previously performed in vitro experiments using whole blood that revealed that 90% of cells tethered by P-selectin were neutrophils, whereas VCAM-1 selectively tethered mononuclear leukocytes and eosinophils (31). It is therefore conceivable that with the presence of both P-selectin and VCAM-1, all leukocytes are tethered to the venular wall and downstream activation events select for rolling eosinophils and mononuclear leukocytes, but not rolling neutrophils. As 1–3% of all circulating leukocytes are eosinophils, but ∼50% of the emigrated leukocytes were eosinophils, the results lend credence to a very highly selective process of recruitment of eosinophils, but not neutrophils at the endothelial cell surface. Specific chemokines induced by IL-4 such as eotaxin (32, 33) are likely to underlie the very selective recruitment process induced by IL-4.
A final issue that needs to be raised is the less than uniform data regarding IL-4-induced VCAM-1 expression in vitro. Initial work using human umbilical vein endothelium has consistently reported that IL-4 induced expression of VCAM-1 independently of P-selectin, E-selectin, or other adhesion molecules (17, 18, 19). Use of microvascular endothelium from dermal (34) or intestinal origin revealed no up-regulation of VCAM-1 following IL-4 treatment (35), but VCAM-1 expression was observed on nasal polyp-derived microvascular endothelium in response to IL-4 (36). Cultured microvascular endothelium requires passaging to increase numbers of cells, but passaging of cells may result in the loss of P-selectin (37) and perhaps inappropriate alteration of other adhesion molecules including VCAM-1. Other investigators have argued that primary cultures of HUVEC may not always reflect expression of adhesion molecules on microvascular endothelium. In the present study using the dual radiolabeling technique, wherein the nonbinding Ab accounts for such factors as 1) nonspecific binding, 2) microvascular permeability changes, and 3) alterations in blood flow, we reveal that VCAM-1 does indeed increase in microvessels in vivo in response to IL-4 (wherein no culture artifact was introduced) and that P-selectin and E-selectin remain at constitutive levels. The constitutive levels of P-selectin are small but functional as these P-selectin values are significantly higher than those generated in P-selectin-deficient animals, which reflect complete absence of P-selectin (25). In addition, the P-selectin is present regardless of whether surgery is performed on the cremaster muscle (27), suggesting constitutive expression of P-selectin and constitutive rolling in this microvascular bed.
In summary, to date in vitro work has revealed an important role for α4 integrin-dependent leukocyte rolling and adhesion and at lower shear forces for the initial tethering process. However, in vitro experiments may differ from the in vivo setting in several critical aspects. First, it is unknown whether the site densities of VCAM-1 protein used, either in purified form or expressed by tranfected cells, correspond to the amount of protein expressed by microvascular endothelial cells under inflammatory conditions in vivo. Second, whether the shear stresses at which α4 integrin-dependent leukocyte tethering has been shown to occur in vitro overlap with the lower end of the physiological range or differ from those present in vivo in postcapillary venules also remains unclear. Our data have for the first time demonstrated that the cytokine IL-4 does indeed alter the endothelial phenotype in microvessels in vivo, to a sufficient extent to induce α4 integrin-dependent leukocyte recruitment across the physiologic range of shear forces in vivo to allow for significant leukocyte recruitment independent of selectin function.
We thank Dr. Dan Bullard, University of Alabama at Birmingham, for generously providing the P-sel/E-sel−/− mice, and Lesley Marshall for performing the electron microscopy.
This study was supported by grants from the Medical Research Council (MRC) of Canada (to P.K.) and the National Institutes of Health (PO1 DK43785) (D.N.G.). M.J.H. is supported by an MRC/Canadian Association for Gastroenterology/Astra fellowship. P.K. is an Alberta Heritage Foundation for Medical Research senior scholar and an MRC scientist.