Effects of JAM-A deficiency or blocking antibodies on neutrophil migration and lung injury in a murine model of ALI

Abstract

Transmigration of neutrophils (PMNs) from the vasculature into inflamed tissues, mediated by interactions between PMNs and adhesion molecules on endothelial cells, is an essential aspect of inflammation. The crucial adhesion molecules include junctional adhesion molecule (JAM)-A. Investigation of the role of this molecule in models of inflammatory disease has been limited, however, and results in different disease models have varied. No previous study has addressed JAM-A in lung disease or effects on oxidant stress and proinflammatory cytokines. We use JAM-A knockout mice and blocking antibodies to investigate the role of JAM-A in a murine model of acute lung injury (ALI). With either experimental system, we find that absence of JAM-A activity significantly reduces migration of PMNs into the alveolar space, with a resulting decrease in oxidative stress. However, there is no reduction in whole lung activity of PMN-associated myeloperoxidase, presumably reflecting the histologically observed retention of PMNs in lung tissue. Activity of these retained PMNs may account for our failure to find significant change in markers of lung oxidative stress or cytokine and chemokine levels in plasma, lung, and bronchoalveolar lavage fluid. We likewise see no JAM-A-related changes in markers of capillary permeability or lung injury. A similar lack of congruence between effects on PMN migration and tissue injury has been reported in other disease models and for other adhesion molecules in models of ALI. Our results thus confirm the crucial role of JAM-A in PMN transmigration but demonstrate that transmigration is not essential for other aspects of inflammation or for lung injury in ALI.

a variety of pulmonary and extrapulmonary insults can produce acute lung injury (ALI) (23). Pathophysiology of this disease is complex, but the earliest phases consistently feature severe neutrophil (PMN)-rich alveolar inflammation and resulting pulmonary injury (18). A major aspect of this injury is capillary leakage and pulmonary edema associated with hypoxemia and respiratory failure (17). Since there is currently no effective pharmacotherapy, the need for identification of novel therapeutic targets is urgent.

Junctional adhesion molecule (JAM)-A is a member of the immunoglobulin superfamily expressed on endothelial and epithelial cells, leukocytes, and platelets (14, 25). On endothelial and epithelial surfaces, it colocalizes with tight junctions (16) and appears to be required for their formation (13, 15) but is not itself directly part of the tight junction structure (7). Tight junctions are the main determinants of vascular and mucosal permeability, and genetic deficiency of JAM-A has been shown to increase permeability of mouse intestinal mucosa and epithelial cell monolayers to dextran and other solutes (11).

Adhesion molecules also play essential roles in migration of PMNs and monocytes from the vasculature into inflamed tissues. Woodfin and colleagues (26) showed that migration of PMNs from mouse cremasteric venules in response to IL-1β sequentially requires ICAM-1, JAM-A, and platelet endothelial cell adhesion molecule (PECAM)-1. Knockout of any one of these three molecules blocks IL-1β-induced PMN migration, but at distinctly different locations within the endothelium. Interestingly, if transmigration is induced by TNF-α, rather than IL-1β, these molecules are required only if direct PMN activation by TNF-α is blocked by deletion of their TNF-α receptors—that is, if TNF-α can act only on the endothelial cells. These results imply that the mechanism of PMN transmigration depends on the activating stimulus. This dependence on the specific stimulus involved may account for the varying results obtained in modestly different animal models of disease. For example, evidence suggests that JAM-A is required for PMN transmigration in cytokine-induced meningitis (3), but not in meningitis induced by Listeria monocytogenes or lymphocytic choriomeningitis virus (12).

The mechanism by which PMNs interact with endothelial JAM-A has not been fully elucidated. Binding between JAM-A and the leukocyte-surface integrin α_Lβ₂, also known as lymphocyte function-associated antigen-1, has been demonstrated (6, 19) and provides a plausible mechanism. However, interactions between PMN and endothelial JAM-A are also possible. JAM-A molecules homodimerize (10), and dimerization of molecules on different cells has been shown to be important for maintenance of epithelial barrier function (15). The possibility that PMN JAM-A might be important for transmigration is supported by the observation that JAM-A-deficient (JAM-A^−/−) PMNs show reduced transendothelial migration in inflamed peritoneum and myocardial ischemia-reperfusion injury (2). Surprisingly, however, deletion of JAM-A on endothelial cells, rather than PMNs, had no effect in this model.

Although there have been several investigations of the roles of other adhesion molecules in models of ALI (1, 4, 8, 21), the role of JAM-A has not been addressed. Employing a combination of anti-JAM-A antibodies and mice genetically deficient in JAM-A, we utilize a murine model to determine the role of this adhesion molecule in LPS-induced PMN migration, inflammation, and lung injury.

MATERIALS AND METHODS

Animals.

Female C57BL/6 (wild-type, JAM-A^+/+) mice were obtained from Jackson Laboratories (Bar Harbor, ME). JAM-A^−/− mice were generated as described previously (20) and backcrossed to a pure C57BL/6 genetic background. Disruption of the JAM-A gene was confirmed via PCR using primers designed to specifically detect hetero- and homozygous mice. Studies were conducted on female mice at 6–8 wk of age (20–25 g body wt). All studies were performed according to protocols reviewed and approved by the Atlanta Veterans Affairs Medical Center Institutional Animal Care and Use Committee.

Cells.

Human pulmonary artery endothelial cells (Lifeline Cell Technology, Walkersville, MD) were obtained at passage 3 and used at passages 5–8. Mouse lung endothelial cells were isolated as described previously (5). Endothelial identity was confirmed by cobblestone morphology, immunofluorescence staining, and Western blotting for ICAM-2, PECAM-1, vascular endothelial-cadherin, and negative staining for fibroblast-specific protein 1/S100A4. Cells were cultured in VascuLife Basal Medium (Lifeline Cell Technology) supplemented with 2% FBS, 10 mM l-glutamine, 0.2% endothelial cell growth supplement, 5 ng/ml recombinant human EGF, 1 μg/ml hydrocortisone hemisuccinate, 0.75 U/ml heparin sulfate (Lifeline Cell Technology), 10,000 U/ml penicillin, and 10,000 μg/ml streptomycin (HyClone, Logan, UT) at 37°C in a humidified atmosphere of 5% CO₂-95% air in T-75 tissue culture flasks, plates, or dishes coated with 2% gelatin and grown to confluent monolayers.

LPS administration and specimen collection.

Mice were anesthetized with an intraperitoneal injection of ketamine (90 mg/kg) and xylazine (10 mg/kg), and a tracheotomy was performed. ALI was then induced by intratracheal injection of 50 μg of endotoxin (LPS) prepared from Escherichia coli O111:B6 (Sigma-Aldrich, St. Louis, MO). After a further 6 h, the lungs were excised for analysis. Bronchoalveolar lavage (BAL) fluid (BALF) and plasma were obtained at the same time.

Antibody treatment.

To confirm the effects of JAM-A deficiency on leukocyte recruitment during endotoxin-induced ALI, an additional set of experiments was carried out in which we either blocked JAM-A in C57BL/6 mice using anti-JAM-A monoclonal antibody (R & D Systems, Minneapolis, MN) or, as a control, infused an isotype-matched control antibody (rat IgG2b, R & D Systems).

Immunofluorescence staining and confocal imaging.

Cells cultured on 2% gelatin-coated glass-bottom dishes (MatTek, Ashland, MA) were washed twice with PBS and subsequently fixed in 10% neutral buffered formalin for 15 min at 37°C. Cells were then permeabilized with Target Retrieval Solution (Dako, Carpinteria, CA) for 10 min at 95°C, allowed to cool to room temperature, and blocked with 1% BSA in PBS containing 0.05% Tween 20 (PBST) at 37°C for 1 h. After they were washed, cells were incubated with one of the following primary antibodies, diluted to 1:50 in PBST-1% BSA, at 37°C for 1 h: rabbit anti-mouse JAM-A (H-80, Santa Cruz Biotechnology, Santa Cruz, CA) or mouse anti-human F11 (BD Pharmingen, San Diego, CA). After they were washed with PBST, cells were incubated with the respective secondary antibodies, rhodamine-conjugated donkey anti-rabbit and FITC-conjugated goat anti-mouse (Jackson ImmunoResearch, West Grove, PA), diluted to 1:50 in PBST-1% BSA, at 37°C for 1 h and washed three times with PBST. Coverslips were retrieved and mounted on glass slides with Vectashield mounting medium with 4′,6-diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA). The slides were viewed by a confocal microscope (Fluoview FV1000, Olympus, Center Valley, PA) using a ×60 fluorescence lens along with Fluoview confocal software (FV10-ASW version 1.7, Olympus).

Lung histopathology and immunostaining for JAM-A.

The lungs were inflated and fixed with 10% neutral formalin overnight at room temperature. Lung tissue was dehydrated with increasing ethanol concentrations and then embedded in paraffin. Paraffin sections (5 μm thick) were stained with hematoxylin-eosin. For immunostaining, paraffin sections were incubated at 60°C and then allowed to cool to room temperature. The sections were deparaffinized in xylene and subsequently rehydrated in a series of graded alcohols. Sections were then permeabilized with Target Retrieval Solution for 10 min at 95°C and allowed to cool to room temperature. Endogenous peroxidase was blocked with 3% H₂O₂ for 10 min, after which the cells were incubated with normal serum (Vector Laboratories) for 30 min to block endogenous biotin and then for 1 h with rabbit anti-mouse JAM-A antibody (H-80). After the sections were washed with PBST, a biotinylated anti-rabbit IgG (Vector Elite Rabbit Kit, Vector Laboratories) was added for 30 min; then the sections were incubated with avidin-biotin complex (Vector Laboratories) for 30 min and stained with diaminobenzidine solution (Vector Laboratories) for 10 min. All slides were counterstained with hematoxylin (Vector Laboratories). Control sections were incubated with an isotype-matched goat IgG (R & D Systems). As an additional negative control, tissue sections were stained using the same procedure used for the actual test but without the primary antibody.

Western blotting.

Total protein extracts were prepared by lysing the cells and homogenizing the lung tissue in 500 μl of ice-cold radioimmunoprecipitation assay buffer supplemented with Halt protease inhibitor cocktail (Pierce Biotechnology, Rockford, IL). Extracts were incubated for 20 min at 4°C and then centrifuged at 14,000 g for 15 min. The supernatants were collected and stored at −80°C. Samples were mixed with sample buffer, separated on a 10% SDS-polyacrylamide gel, and electroblotted onto a polyvinylidene difluoride membrane. The membrane was blocked with blocking solution [3% BSA in Tris-buffered saline + Tween 20 (TBST)] for 1 h at room temperature. Blots were then incubated overnight at 4°C with primary antibodies as used for immunofluorescence staining or with α-tubulin antibody (Santa Cruz Biotechnology). The membrane was then washed in TBST and incubated with secondary antibodies consisting of donkey anti-mouse infrared dye (IRDye) 680 (red; LI-COR, Lincoln, NE) and goat anti-rabbit IRDye 780 (green; LI-COR), diluted to 1:5,000, for 1 h at room temperature. The infrared signal was detected using an Odyssey infrared imager (LI-COR).

PMN counts in blood, BALF, and lung tissue.

PMNs in blood and different compartments of lung were counted as described previously with slight modifications (22). Blood was collected by cardiac puncture and lysed with RBC lysis buffer (Sigma) to remove RBCs and centrifuged at 500 g at 4°C for 5 min. Pelleted cells were resuspended in 1 ml of PBS. After blood was collected, nonadherent PMNs were flushed from the pulmonary vasculature by injection of 10 ml of PBS into the right ventricle. For collection of BALF, the lung was flushed six times with 1 ml of PBS containing 0.1 mM EDTA via a tracheal cannula. The pooled BALF was centrifuged at 500 g at 4°C for 5 min. Pelleted cells were resuspended in 1 ml of PBS. After BAL, lungs were harvested, minced, and digested with 1 mg/ml collagenase/dispase (Roche, Indianapolis, IN), 100 U/ml hyaluronidase type I-s, and 100 U/ml DNase (Sigma) at 37°C for 30 min. Lung digest was passed through a 70-μm cell strainer (BD Falcon, Bedford, MA), and the resulting cell suspension was centrifuged for 10 min at 500 g. The pellet was treated with RBC lysis buffer and centrifuged again. The resulting pellet was resuspended in buffer. In each instance, total cell number was determined by counting on a hemocytometer, and a differential cell count was performed by cytospin staining with Diff-Quik (Siemens, Newark, DE).

Measurement of myeloperoxidase activity.

As an index of PMN accumulation, BALF- and tissue-associated myeloperoxidase (MPO) activities were determined. Frozen lung tissues were thawed, weighed, homogenized, and sonicated on ice in radioimmunoprecipitation assay buffer. After centrifugation at 10,000 g at 4°C for 20 min, the supernatant was collected. This supernatant and the cell-free supernatant from BALF were each used for determination of MPO activity by a commercially available fluorometric assay kit (catalog no. 700160, Cayman Chemical, Ann Arbor, MI).

BALF protein and lung wet:dry weight ratio.

Increase in BALF protein concentration was taken as a measure of increased permeability of alveolar-capillary barriers. Total protein concentration in the supernatant following BALF centrifugation was determined using the BCA Protein Assay Kit (Pierce, Rockford, IL). As an index of lung edema, the amount of extravascular lung water was calculated. The lower lobe of the right lung was ligated and excised, and the wet weight was recorded. The lung was then placed in an incubator at 60°C for 24 h to obtain the dry weight. The wet:dry ratio was calculated by dividing the wet weight by the dry weight.

Measurement of oxidant stress.

H₂O₂ production in lung tissue or BALF was determined using the Amplex red H₂O₂ assay kit (Molecular Probes, Eugene, OR). The concentrations of nitrate and malondialdehyde (MDA) in lung homogenates and BALF were measured using commercially available colorimetric assay kits (Cayman Chemical).

Measurements of BALF, plasma, and lung cytokine levels.

Levels of TNF-α, KC, and macrophage inflammatory protein 2 were measured using ELISA kits (R & D Systems) according to the manufacturer's instructions.

Statistical analysis.

Values are means ± SD. Differences between groups were analyzed using ANOVA followed by Bonferroni's multiple comparison test using GraphPad Prism 5.03 software (GraphPad Software, La Jolla, CA). P < 0.05 was considered significant.

RESULTS

JAM-A expression in the lung.

Although the lung is highly susceptible to PMN-associated inflammatory conditions, the presence of JAM-A in lung tissues has not previously been demonstrated. To confirm the presence of JAM-A in lungs of JAM-A^+/+ mice, sections were immunostained with a rabbit antibody to the COOH terminus of this molecule followed by a biotinylated anti-rabbit IgG antibody and a standard biotin-based development system. JAM-A was observed on lung endothelial cells and adherent cells of the myeloid lineage following induction of ALI by LPS treatment (Fig. 1A, left), and no expression was seen in JAM-A^−/− mice (Fig. 1A, right). We also used Western blotting to confirm the presence and expression of JAM-A in lysates from mouse lungs, isolated mouse lung endothelial cells, and human pulmonary artery endothelial cells treated with PBS or LPS (Fig. 1B). Western blotting likewise confirmed the absence of JAM-A in whole lung and isolated pulmonary endothelial cells of JAM-A^−/− mice. In all cases, expression of JAM-A was not altered by treatment with LPS. Immunofluorescence microscopy localized the presence of JAM-A in lung endothelial cells of wild-type mice (Fig. 1C, left) and human pulmonary artery endothelial cells (Fig. 1C, right), together with its absence in JAM-A^−/− mice (Fig. 1C, middle).

Fig. 1.

Junction adhesion molecule (JAM)-A expression in the lung. A: Lungs of JAM-A^+/+ and JAM-A^−/− mice were explanted en bloc 6 h after intratracheal injection of LPS (50 μg). Paraffin sections were prepared and immunostained, and sections were examined microscopically. Endothelial cells (spindle-shaped nuclei indicated by arrow) demonstrated positive staining for JAM-A (brown) in JAM-A^+/+, but not JAM-A^−/−, mice (n = 3 per group). JAM-A^+/+ section also demonstrated positive JAM-A staining for an adherent nucleated cell (arrowhead). V, vascular lumen. B: Western blot analysis for JAM-A was performed following 6 h of PBS or LPS treatment on whole lung and lung endothelial cells (MLEC) from JAM-A^+/+ and JAM-A^−/− mice (n = 6 per group), as well as human pulmonary artery endothelial cells (HPAEC, n = 3) treated with PBS or LPS for 6 h. Blots were stripped and reprobed with α-tubulin antibody. C: immunofluorescence microscopy for JAM-A was performed on MLEC (n = 6 per group; red) and HPAEC (n = 3; green). Blots and images are representative of 2 independent experiments.

Role of JAM-A in vascular-alveolar PMN transmigration.

A crucial role for JAM-A in migration of PMNs across endothelial and epithelial cell layers has been reported in a variety systems but has not been investigated in the lung. We employed knockout mice and antibody-mediated blockade of JAM-A to determine the role of this molecule in LPS-induced ALI. Controls for JAM-A knockout mice were wild-type mice, with parallel studies in which vehicle was administered in both strains. We found that intratracheal LPS treatment led to significantly lower total cell count (Fig. 2, A and D) and number of PMNs (Fig. 2, G and J) in BALF of knockout or antibody-treated mice than in BALF from wild-type mice without anti-JAM-A blockade. The total cell count (Fig. 2, B and E) and number of PMNs (Fig. 2, H, K, and R) in whole blood did not differ between knockout or antibody-treated mice and wild-type or IgG2b-treated mice following intratracheal LPS treatment. In contrast, the total cell count (Fig. 2, C and F) and number of PMNs (Fig. 2, I, L, and S) in blood-free lung digests following intratracheal LPS treatment were significantly higher in knockout or antibody-treated mice than in wild-type mice without anti-JAM-A blockade. BALF MPO, an enzyme found in association with PMNs, was likewise lower when JAM-A was deficient or inactivated (Fig. 2, M and N), and reduced PMN transmigration into alveolar spaces was further confirmed by cytospin observations on BALF (Fig. 2Q). However, there was no significant reduction of MPO activity in whole lung under these conditions (Fig. 2, O and P). In all cases, markers of PMN infiltration were lower following treatment with vehicle, rather than LPS, but there were no significant differences between knockout and wild-type mice.

Fig. 2.

JAM-A deletion or blockade reduces transendothelial migration of neutrophils (PMNs) in LPS-induced lung injury. In A–C and G–I, acute lung injury (ALI) was induced in JAM-A^+/+ and JAM-A^−/− mice by intratracheal injection of LPS (50 μg). In D–F and J–L, JAM-A^+/+ mice were injected with 100 μg of an anti-JAM-A or an isotype-matched control (rat IgG2b) antibody 30 min prior to LPS. In each case, after a further 6 h, samples were obtained. A–F: total cell number in bronchoalveolar lavage fluid (BALF), blood, and lung digests determined by hemocytometer counting. G–L: PMN number in BALF, blood, and lung digests determined by cytospin. M–P: myeloperoxidase (MPO) activity in BALF and whole lung. Q–S: microscopic examination of differentially stained cells in BALF, blood, and lung digests obtained following the indicated treatments. E-I, endothelial-interstitial. Data are representative of 3 independent experiments, with n = 6–8 mice/group. ***P < 0.001. ns, Not significant.

Impact of JAM-A deficiency or blockade on LPS-induced lung inflammation.

Activated PMNs generate reactive oxygen species and secrete chemoattractants for other leukocytes (24) that play important roles in inflammation. PMN infiltration into the alveolar space is therefore considered a critical feature of pulmonary inflammation. As we had observed reduced PMN infiltration into air spaces of JAM-A^−/− mice and following blockade of JAM-A by targeted antibodies, we next investigated these two aspects of inflammation. However, despite the accompanying reduction in PMN transmigration, neither knockout nor antibody-mediated inactivation of JAM-A significantly affected whole lung production of H₂O₂ (Fig. 3, A and B), a reactive oxygen species produced primarily by PMNs and macrophages; levels of MDA (Fig. 3, E and F), a lipid oxidation product considered a marker for overall oxidative stress, or nitrate (Fig. 3, I and J), a metabolic end product of NO and other reactive nitrogen species. Likewise, no differences were seen in BALF, plasma, or lung homogenate levels of the proinflammatory cytokine TNF-α (Fig. 4, A–F) or the chemokines KC (Fig. 4, G–L) and macrophage inflammatory protein 2 (Fig. 4, M–R). However, JAM-A inactivation did reduce markers of oxidant stress in BALF (Fig. 3, C, D, G, H, K, and L).

Fig. 3.

JAM-A deletion or blockade decreases LPS-induced inflammatory markers in BALF, but not in lung. In A, C, E, G, I, and K, ALI was induced in JAM-A^+/+ and JAM-A^−/− mice by intratracheal injection of LPS (50 μg). In B, D, F, H, J, and L, JAM-A^+/+ mice were injected with 100 μg of an anti-JAM-A or an isotype-matched control (rat IgG2b) antibody 30 min prior to LPS. In each case, after a further 6 h, samples were obtained. A–D: H₂O₂ production in lung and BALF. E–H: malonaldehyde (MDA)-to-protein ratio in lung and BALF. I–L: nitrate concentration in lung and BALF. Data are representative of 3 independent experiments, with n = 6–8 mice per group. *P < 0.05; ***P < 0.001.

Fig. 4.

JAM-A deletion or blockade has no effect on LPS-induced cytokine or chemokine production. In A–C, G–I, and M–O, ALI was induced in JAM-A^+/+ and JAM-A^−/− mice by intratracheal injection of LPS (50 μg). In D–F, J–L, and P–R, JAM-A^+/+ mice were injected with 100 μg of an anti-JAM-A or an isotype-matched control (rat IgG2b) antibody 30 min prior to LPS. In each case, after a further 6 h, samples were obtained. A–F: TNF-α in BALF, plasma, and lung homogenate. G–L: KC in BALF, plasma, and lung homogenate. M–R: macrophage inflammatory protein 2 (MIP-2) in BALF, plasma, and lung homogenate. Data are representative of 3 independent experiments, with n = 6–8 mice per group. ***P < 0.001.

Impact of JAM-A deficiency or blockade on capillary permeability and severity of LPS-induced lung injury.

Increased capillary permeability and resulting pulmonary edema are crucial features of ALI that are often the proximate causes of hypoxemia and morbidity. Discrepant results between effects of JAM-A deficiency or blockade on PMN infiltration into lung air spaces, on the one hand, and other measures of lung inflammation, on the other, raise the question of how these aspects of lung injury might be affected. The question is particularly cogent, because association of JAM-A with the tight junctions responsible for regulating capillary permeability suggests that JAM-A deficiency itself might conceivably increase capillary permeability. We find, however, that JAM-A deficiency or blockade does not alter the protein concentration of BALF (Fig. 5, A and B) or the lung wet:dry weight ratio (Fig. 5, C and D), a measure of edema, in response to LPS. There are likewise no differences between wild-type and knockout mice under baseline conditions. However, direct examination of the lung following LPS treatment reveals greater numbers of PMNs in the lungs of wild-type mice not treated with active antibody (Fig. 5E) than in lungs of JAM-A^−/− mice or mice treated with anti-JAM-A antibodies (Fig. 5F). Especially notable is the presence of PMNs in the alveolar spaces of control mice and their absence from the alveolar spaces of knockout or antibody-treated mice. Conversely, PMNs are seen adhering to the vascular walls of knockout mice (Fig. 5F, inset in 1st image and 2nd, 3rd, and 5th images).

Fig. 5.

JAM-A deletion or blockade has no effect on LPS-induced capillary leakage and lung edema. In A and C, ALI was induced in JAM-A^+/+ and JAM-A^−/− mice by intratracheal injection of LPS (50 μg). In B and D, JAM-A^+/+ mice were injected with 100 μg of an anti-JAM-A or an isotype-matched control (rat IgG2b) antibody 30 min prior to LPS. In each case, after a further 6 h, samples were obtained. A and B: protein concentration in BALF. C and D: lung wet:dry weight ratio. E and F: lung histology following hematoxylin-eosin staining. In 3 images at left, ALI was induced as described in A. In 2 images at right, mice were treated as described in B. A, alveolar space; V, vascular lumen. Arrow identifies PMN. Data are representative of 3 independent experiments, with n = 6–8 mice per group. ***P < 0.001.

DISCUSSION

This appears to be not only the first study of the role of JAM-A in ALI, but the first in any disease model to investigate inflammatory markers other than PMN transmigration. We find that knockout of JAM-A or blockade of its activity by systemic administration of anti-JAM-A antibodies significantly reduces the ability of PMNs to cross the vascular-alveolar wall in response to LPS, with a concomitant reduction in BALF oxidative stress. However, there is no reduction in whole lung markers of oxidative stress, including the lung MDA-to-protein ratio, or in plasma, lung, or BALF levels of proinflammatory cytokines or chemokines. Markers of capillary permeability and lung injury are likewise unaffected by JAM-A deletion or blockade.

Although this apparent discrepancy between PMN transmigration and other markers of lung inflammation and injury may appear somewhat surprising, there have been similar reports in connection with adhesion molecule deletion or blockade in other disease models. Perkowski and colleagues (21) found that while antibody blockade of ICAM-1 or PECAM-1 reduced PMN infiltration in hyperoxia-induced ALI, as did genetic deficiency of PECAM-1, none of these factors reduced BALF protein or lung wet:dry weight ratio. Even depletion of circulating PMNs had no effect on lung injury in this model. In hepatic ischemia-reperfusion, JAM-A deficiency or blockade actually increased tissue injury in the face of reduced PMN infiltration (9). By contrast, antibody-mediated blockade of JAM-A in a meningitis model induced by IL-1β + TNF-α reduced not only PMN migration, but also albumin extravasation, into cerebrospinal fluid (3). On the other hand, there was no effect on PMN migration and either no effect or an actual increase in severity of clinical disease in meningitis induced by L. monocytogenes or lymphocytic choriomeningitis virus (12). There are similar inconsistencies between results for P-selectin and ICAM-1 mutants in murine models of sepsis-induced ALI (8) and ALI induced by cobra venom factor (4). In the sepsis model, deletion of either or both genes reduced PMN recruitment and capillary permeability in lung and liver. In the cobra venom factor model, while antibodies to P-selectin or ICAM-1 reduced these effects, deletion of either gene did not. Indeed, migration and capillary permeability were actually increased in the double mutant.

A plausible explanation for our observation that inhibition of PMN infiltration into the alveolar spaces does not reduce whole lung oxidant stress or injury may be that the PMNs appear to become trapped in the lung tissue, adhering to or entering the perivascular tissue but failing to leave it. This is supported by the absence of any change in whole lung MPO activity and by microscopic observation following hematoxylin-eosin staining. Others have similarly observed apparently trapped PMNs associated with JAM-A deficiency in mouse cremasteric venules (26) or in other tissues (2). Our results thus support the plausible suggestion that PMNs lodged within the alveolar-vascular wall have inflammatory effects similar to those of PMNs actually within the alveolar space and call into question the frequent assumption that blocking PMN transmigration into alveolar spaces is therapeutically valuable in all cases. It appears that in many diseases the therapeutically relevant event could be prevention of PMN attachment to the vascular endothelium.

Although inactivation of JAM-A or other adhesion molecules by genetic deletion and blocking antibodies might be expected to give equivalent results, as observed in our studies, this has not always been the case. Doerschuk and colleagues (4) suggest several possible explanations for the differing effects of antibodies to P-selectin and ICAM-1, on the one hand, and genetic deficiency of these molecules, on the other. These include antibodies blocking PMN adhesion to the endothelium or otherwise interfering with the injury-associated responses of the endothelium or PMNs. Alternatively, they suggest that life-long deficiency of specific adhesion molecules could lead to upregulation of alternative pathways. Perkowski and colleagues (21), who similarly found that PMN migration was blocked by antibodies to ICAM-1, but not by genetic deficiency of this molecule, suggest that their genetically deficient mice may have had small amounts of an alternatively spliced form of ICAM-1. Our results demonstrate that none of these events occur with anti-JAM-A antibodies or knockout.

Although not all studies agree, most evidence indicates that JAM-A, like ICAM-1 and PECAM-1, is essential for PMN migration across the vascular wall and, hence, for recruitment of PMNs into the inflamed tissue. Our study adds ALI to the list of conditions for which this has been demonstrated and supports the conclusion that these JAM-A effects are general throughout all types of PMN-rich inflammation. However, we find that, in ALI, PMNs are retained on or within the vascular wall and have the same effect on lung inflammation and injury as if they had actually entered the alveolar spaces. Studies by others in different diseases and investigating different adhesion molecules involved in PMN migration have given varying results, with some suggesting that blocking PMN transmigration reduces tissue injury and others that it does not. These differing outcomes raise the possibility that effects of adhesion molecule inactivation may be specific to a given disease or even a given model of that disease. By showing that not only tissue injury, but also inflammation, may fail to track PMN transmigration, our results emphasize the complexity of the PMN-inflammation-injury relationship and the need for specific investigation in each disease of interest.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grant HL-093196 (R. C. Reddy).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

S.P.L. and A.T.R. performed the experiments; S.P.L., A.T.R., and R.C.R. analyzed the data; S.P.L., A.T.R., U.P.N., and R.C.R. interpreted the results of the experiments; S.P.L. and A.T.R. prepared the figures; S.P.L., A.T.R., and R.C.R. drafted the manuscript; S.P.L., A.T.R., M.U.N., U.P.N., and R.C.R. edited and revised the manuscript; S.P.L., A.T.R., M.U.N., U.P.N., and R.C.R. approved the final version of the manuscript; A.T.R. and R.C.R. are responsible for conception and design of the research.