INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: monocyte; lung; inflammation; recruitment; PKH

Acute pulmonary inflammation is characterized by the accumulation of both neutrophils and monocytes within the lung parenchyma and alveolar compartment. Essential steps in the leukocyte migratory process from the pulmonary vascular bed into the alveolar air space include passage across the anatomically highly specialized microvascular endothelial cell barrier, the extracellular matrix (ECM) of both endothelial and epithelial cells, and across the tight alveolar-epithelial cell barrier (1). Because of their considerable capacity to elaborate proinflammatory cytokines, reactive oxygen species, and proteolytic enzymes, monocytes may actively contribute to both acute and chronic inflammatory lung diseases (2, 3). Recent findings by our group demonstrated that the transendothelial/epithelial recruitment process of monocytes per se was not sufficient to evoke a major inflammatory response involving this type of cell, including priming for enhanced responsiveness to lipopolysaccharide (LPS) (4, 5). This finding supports the concept that monocytes may have important effector cell functions in the alveolar compartment even in the absence of acute lung inflammation, with expansion of the resident alveolar macrophage (rAM) pool most probably being of major importance.

The adhesion molecule interactions involved in the multistep process of leukocyte emigration toward the alveolar air spaces are well characterized for the neutrophil, but are only poorly defined for monocyte trafficking (6). To migrate across endothelium in vitro, monocytes utilize sequential interactions of members of the selectin family (L-selectin, CD62L), the ₂-integrins (CD11a/CD18, _L2, lymphocyte function associated antigen-[LFA]-1; CD11b/CD18, _M2, Mac-1), and ₁-integrins (CD49d/CD29, very late activation antigen [VLA]-4; CD49e/ CD29, VLA-5) on the monocyte surface with members of the immunoglobulin superfamily (CD54, intercellular adhesion molecule [ICAM]-1; CD106, vascular cell adhesion molecule [VCAM]-1) on the endothelial cell surface. In addition, adhesion proteins expressed in the ECM of the interstitial compartment appear to be involved in monocyte migration (10). Moreover, we recently found that monocyte transmigration across the alveolar-epithelial barrier in vitro also involves homotypic interactions via CD47 (integrin-associated protein, [IAP]) (11).

Clinical strategies to selectively block monocyte emigration in inflamed lungs are hampered by insufficient understanding of adhesive interactions involved in monocyte trafficking across pulmonary endothelial/alveolar epithelial barrier cells in vivo. Since in vitro studies of monocyte transmigration can never perfectly mimic the highly specialized pulmonary capillary endothelial/alveolar epithelial barrier, and may therefore only partly reflect in vivo conditions of leukocyte trafficking, we established a mouse model to investigate transendothelial and transepithelial leukocyte migration into the alveolar compartment in vivo. A major forward step in this was the development of a fluorescence-activated cell sorter (FACS)-based technology, which allows the separation and quantification of monocytes newly recruited to the alveolar air spaces from differentiated rAM, comigrating neutrophils, and lymphocytes in the bronchoalveolar lavage fluid of mice (4). Using this methodologic approach, we found that the combined intratracheal administration of the monocyte chemoattractant JE/monocyte chemotactic protein (MCP)-1 and low dose Escherichia coli LPS induced a self-limiting pulmonary inflammatory response, characterized by tightly controlled sequelae of both neutrophil and monocyte emigration into the alveolar space, thus allowing us to investigate leukocyte recruitment patterns in mice under defined experimental conditions (5). We used this model to evaluate the effect of various function-blocking monoclonal antibodies (mAb) on the alveolar accumulation of monocytes in mice challenged intratracheally with JE/MCP-1 in either the absence (noninflammatory conditions) or presence (acute inflammatory conditions) of E. coli LPS. We report here that the alveolar monocyte accumulation in response to JE/MCP-1 depends largely on the ₂-integrin CD11b/CD18 and CD54 and on the ₁-integrin CD49d but not on CD106, whereas amplified monocyte recruitment on cochallenge with JE/MCP-1 and LPS depends additionally on CD11a and CD106. This pattern of adhesion molecule involvement differs from that for comigrating neutrophils, which is additionally influenced by the type of chemokine presented to the alveolar surface.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Female BALB/c mice (weight: 18 to 21 g) were purchased from Charles River Laboratories (Sulzfeld, Germany).

Reagents

The red fluorescent dye PKH26-PCL and diluent B solution were purchased from Zynaxis (Malvern, CA; distributed by Sigma, Deisenhofen, Germany). Murine JE, the homologue of the human MCP-1 gene product (JE/MCP-1) (12) was purchased from R&D Systems (Wiesbaden, Germany), as was murine MIP-1. The recombinant JE/MCP-1 and recombinant MIP-1 used in the study, as well as the PKH26-PCL and diluent B solution, were LPS-free as analyzed with the COATEST (Chromogenix, Mölndal, Sweden) amoebocyte lysate assay (detection limit < 10 pg/ml).

Treatment of Animals

For discrimination of newly recruited monocytes from differentiated, rAM, we prelabeled the rAM pool with the red fluorescent dye PKH26 in vivo in a manner essentially as described recently in detail (4, 5).

Treatment Protocol

We evaluated molecular pathways involved in monocyte and/or neutrophil emigration into the alveolar air space of mice in three treatment groups, consisting of: (1) mice given JE/MCP-1 in the absence of E. coli LPS (50 µg JE/MCP-1/mouse); (2) mice given E. coli LPS (10 ng/mouse); and (3) mice given JE/MCP-1 (50 µg/mouse) in the presence of E. coli LPS (10 ng/mouse). The intratracheal instillation of JE/ MCP-1 in mice in treatment Group I, eliciting monocyte recruitment into the alveolar air space, was recently shown to not provoke any significant alveolar inflammation, as judged by: (1) absence of the proinflammatory cytokines tumor necrosis factor (TNF)-, interleukin (IL)-6, and IL-8 in bronchoalveolar lavage fluid (BALF), corresponding to sham-operated control mice; (2) absence of alveolar accumulating neutrophils; and (3) lack of lung barrier dysfunction. In contrast, these inflammatory features were partly induced in mice of treatment Group II (treatment with LPS alone) and were fully present in mice of treatment Group III on intratracheal instillation of JE/MCP-1 in the presence of E. coli LPS (5).

At 24 h after labeling of the rAM pool by PKH26, we anesthetized BALB/c mice and instilled LPS alone, murine rJE/MCP-1 (50 µg/ mouse) alone, or JE/MCP-1 (50 µg/mouse) plus E. coli LPS (10 ng/ mouse) intratracheally into the animals' lungs, using a 26-gauge Abbocath (Abbott, Wiesbaden, Germany) placed into the trachea. At 15 min before intratracheal instillation of the respective reagents, mice received intravenous injections of respective function-blocking mAb preparations via the tail vein (5 mg/kg mAb in a total volume of 100 µl of phosphate-buffered saline [pH 7.4, in 0.1% mouse serum]). None of the mAb treatments caused neutropenia, monocytopenia, or lymphopenia as evaluated by sampling of ethylene diamine tetraacetic acid-anticoagulated peripheral blood from appropriately treated animals and subsequent analysis in a Coulter cell counter (Coulter Instruments, Krefeld, Germany). Mice were allowed to recover from anesthesia and then returned to their cages with free access to food and water.

mAb

The panel of function-blocking mAb used in the study was recently evaluated in various experimental inflammatory mouse models. Rat antimouse CD11a (LFA-1, clone M17/4, isotype IgG_2a [13, 14]), rat antimouse CD11b (Mac-1, clone M1/70, isotype IgG_2b [9, 15]), and rat antimouse CD106 (VCAM-1, clone M/K-2, isotype IgG_2a [16, 17]) were obtained from Pharmingen (Weiterstadt, Germany). The goat antimouse CD106 polyclonal antibody (C-19) used for immunohistochemistry was purchased from Santa Cruz Biotechnology (Heidelberg, Germany), and the secondary alkaline phosphatase-labeled rabbit antigoat F(ab)₂ was from Biotrend (Cologne, Germany). Rat antimouse CD49d (VLA-4, clone PS/2, isotype IgG_2b [16-18]) and rat antimouse F4/80 (isotype IgG_2b) were purchased from Serotec (Munich, Germany). Rat antimouse CD54 (ICAM-1, clone YN1/1, isotype IgG_2b [7, 19-21]) was a generous gift from Dr. Britta Engelhardt of the Max-Planck-Institute, Bad Nauheim, Germany. Preparation of hamster antimouse CD18 mAb (clone 2E6; isotype IgG_2a) was recently described (5). Nonimmune rat and hamster IgG (Sigma, Deisenhofen, Germany), as well as rat isotype IgG_2a and IgG_2b, were used as control antibodies (Pharmingen). In selected experiments, mAb anti-F4/80, directed at a monocyte/macrophage specific cell surface antigen (22), was used as an additional control to monitor nonspecific inhibition of monocyte traffic caused by antibody binding to the monocyte surface in vivo. The VectorRed Substrate Kit was purchased from Vector Laboratories (Burlingame, CA). All other biochemicals were obtained from Merck (Darmstadt, Germany) and Sigma.

Bronchoalveolar Lavage

The bronchoalveolar lavage (BAL) procedure used in the study, as well as the determination of distribution patterns of leukocytes recovered from BALF, were recently described in detail (4, 5). Viability of BAL cells was routinely analyzed by trypan blue dye exclusion.

Flow Cytometry

A FACStar Plus flow cytometer (Becton Dickinson, San Jose, CA) was used throughout the study (23). Discrimination and quantification of the leukocyte populations contained in BALF from mice of the various treatment groups was performed as recently outlined in detail (4).

Immunohistochemistry

Ten-micrometer-thick lung sections were stained with anti-CD106 antibodies, using alkaline phosphatase-based immunohistochemistry, as recently described (24). We evaluated endothelial cells of different anatomic segments of the pulmonary vasculature (25). The immunohistochemical signal intensity was evaluated through use of a grading system with four Grades (Grades 0 to 3), in which Grade 0 represented no staining; Grade 1 represented weak staining; Grade 2 represented moderate staining; and Grade 3 represented strong staining.

Statistics

The study data are expressed as mean ± SEM. Calculation of significant differences between groups was done by one-way analysis of variance. A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of Adhesion Function-Blocking mAb on the Alveolar Accumulation of Monocytes in BALB/c Mice

The intravenous injection of nonimmune rat IgG or isotype-matched control IgG (5 mg/kg) into BALB/c mice subsequently challenged with intraalveolar deposition of JE/MCP-1 in the absence (Figure 1A) or presence (Figure 1B) of E. coli LPS did not affect the alveolar accumulation of monocytes. Similarly, no inhibition of alveolar monocyte accumulation was observed in mice receiving intravenous injections of the monocyte/macrophage-specific mAb F4/80, demonstrating that antibody binding to the monocyte surface per se did not affect the transendothelial/epithelial migration process (Figure 1A). The intravenous administration of monoclonal anti-CD11b and anti-CD18 antibodies strongly inhibited alveolar monocyte accumulation by 92% and 95%, respectively, whereas monoclonal anti-CD11a antibody blocked the alveolar monocyte accumulation in response to JE/MCP-1 by only ~ 50% (Figure 1A). Blocking of the ₂-integrin ligand CD54 (ICAM-1) by intravenous injection of the monoclonal anti-CD54 antibody YN1/1 inhibited alveolar monocyte emigration by ~ 80% (Figure 1A). Blockade of the ₁-integrin CD49d (VLA-4) inhibited the accumulation of monocytes within the alveolar air space of mice by > 90%. In contrast, intravenous injection of monoclonal anti-CD106 (VCAM-1) antibody into mice receiving intratracheal instillation of JE/MCP-1 did not inhibit the monocytic emigration process above control levels (Figure 1A).

View larger version (23K): [in this window] [in a new window]   Figure 1.   Effect of function-blocking mAb on the alveolar accumulation of monocytes in response to intratracheal JE/MCP-1 in the absence or presence of E. coli LPS. Mice were pretreated with isotype-matched control IgG or function-blocking mAb, as indicated, followed by instillation of JE/ MCP-1 (50 µg/mouse) in the absence (A) or presence (B) of E. coli LPS. At 24 h, mice were killed and identification and quantification of alveolar monocytes was performed. Values are given as percent inhibition compared with mice receiving vehicle alone. Alv-Mo = alveolar recruited monocytes. Values are given as mean ± SEM (n = 5 each). *p < 0.05, **p < 0.01 compared with JE/MCP-1- treated mice.

In mice given JE/MCP-1 plus E. coli LPS, blockade of either CD11b or CD18 adhesion pathways effectively inhibited alveolar monocyte accumulation within the alveolar air spaces (90% and 94% inhibition; Figure 1B). Importantly, however, intravenous administration of monoclonal anti-CD11a antibody to BALB/c mice under these conditions also blocked alveolar monocyte accumulation by 90%, (p < 0.05 as compared with treatment with JE/MCP-1 plus E. coli LPS; Figure 1B). Blockade of the ₂-integrin ligand CD54 (ICAM-1) by intravenous administration of mAb YN1/1 also strongly inhibited alveolar monocyte recruitment (Figure 1B). Furthermore, monocyte accumulation within the alveolar compartment was found to be highly dependent on functional CD49d, since blockade of 4 significantly inhibited the monocytic emigration process by 85% (Figure 1B). Under the condition of combined JE/MCP-1 plus LPS challenge, monoclonal anti-CD106 antibody was also found to significantly block alveolar monocyte accumulation (65% inhibition; p < 0.05 as compared with JE/MCP-1 plus E. coli LPS; Figure 1B).

Effect of Intratracheal Instillation of JE/MCP-1 or JE/MCP-1 Plus E. coli LPS on Endothelial VCAM-1 Expression in the Pulmonary Vasculature

In lung sections of control mice, expression of VCAM-1 (CD106) was not detectable on endothelium of the pulmonary vasculature. In mice challenged with intratracheal instillations of either low dose LPS or JE/MCP-1, endothelial expression of CD106 was nearly undetectable in the vessels of the pulmonary vasculature (Figure 2). In contrast, intratracheal instillation of JE/MCP-1 plus E. coli LPS provoked a marked and highly significant upregulation of CD106 on endothelial cells throughout the vessels at 6 h, which declined by 24 h (Figure 2, and other data not shown in detail).

View larger version (25K): [in this window] [in a new window]   Figure 2.   Effect of intratracheal instillation of LPS, JE/ MCP-1, or JE/MCP-1 plus E. coli LPS on endothelial VCAM-1 expression in mouse lungs. Mice were either sham-treated or challenged intratracheally with LPS, JE/ MCP-1, or JE/MCP-1 plus E. coli LPS. In lung sections from sham-treated mice (control), no VCAM-1 expression was detectable. In mice challenged wih JE/MCP-1 plus E. coli LPS, strongly and significantly increased VCAM-1 expression was observed on large, muscular vessel endothelium (black bars), partly muscular vessel endothelium (dashed bars), and pre- and postcapillary vessel endothelium (white bars) at 6 h as compared with mice treated with either LPS alone or JE/MCP-1 alone. The VCAM-1 expression rapidly declined at 24 h (not shown). Data are given as mean ± SEM (n = 5 per treatment group). ***p < 0.001 versus all other groups.

Effect of Adhesion Function-Blocking mAb on the Alveolar Accumulation of Neutrophils in Response to Low Dose LPS Alone or JE/MCP-1 Plus E. coli LPS

Neutrophil trafficking toward the alveolar compartment of mice in response to E. coli LPS alone was partly inhibited in mice pretreated with monoclonal anti-CD11a antibody (~ 60% inhibition) and strongly inhibited in mice pretreated with monoclonal anti-CD11b, anti-CD18, or anti-CD54 antibody (Figure 3A). In contrast, both the monoclonal anti-CD49d and anti-CD106 antibodies only weakly inhibited the alveolar neutrophil trafficking in response to LPS treatment alone (Figure 3A).

View larger version (23K): [in this window] [in a new window]   Figure 3.   Effect of function-blocking mAb on alveolar accumulation of neutrophils in response to intratracheal JE/MCP-1 in the presence of E. coli LPS. Mice were pretreated with isotype-matched control IgG or function-blocking mAb, as indicated, followed by instillation of E. coli LPS (10 ng/mouse, A) or JE/MCP-1 (50 µg/ mouse) in the presence of E. coli LPS (B). At 24 h, mice were killed and identification and quantification of alveolar recruited neutrophils was performed. Values are given as percent inhibition compared with mice receiving vehicle alone. PMN = polymorphonuclear neutrophils. Values are given as mean ± SEM (n = 5 per treatment group). *p < 0.05 versus challenge with LPS alone.

Neutrophil trafficking toward the alveolar compartment of mice challenged with combined JE/MCP-1 plus E. coli LPS was also strongly inhibited by monoclonal anti-CD11b or anti-CD18 antibody (Figure 3B). Surprisingly, however, intravenous injection of the anti-CD11a antibody did not affect the alveolar accumulation of neutrophils elicited by combined JE/ MCP-1 plus LPS (< 10% inhibition; Figure 3B). In contrast, alveolar neutrophil accumulation was strongly inhibited when mice were pretreated with the function-blocking monoclonal anti-CD54 antibody (clone YN1/1; 90% inhibition; Figure 3B). As observed with LPS challenge alone, pretreatment of mice with the monoclonal anti-CD49d or anti-CD106 antibody only weakly affected the alveolar neutrophil accumulation elicited by combined JE/MCP-1 plus LPS (Figure 3B).

Effect of Anti-CD11a on the Alveolar Accumulation of Neutrophils in BALB/c Mice in Response to Intratracheal MIP-1 Plus E. coli LPS

To further investigate whether the unexpected lack of contribution of CD11a to the alveolar neutrophil accumulation in response to combined JE/MCP-1 plus LPS was specifically related to the chemokine stimulus employed, we either pretreated mice with control IgG or pretreated them with monoclonal anti-CD11a antibody followed by intratracheal instillation of MIP-1 or LPS, or of MIP-1 combined with LPS instead of JE/MCP-1 (Figure 4). The C-C chemokine MIP-1 was recently shown to recruit neutrophils but not monocytes into the alveolar air spaces of intact mice (4). Of note was that we observed an increase of ~ 50% in alveolar neutrophil recruitment in response to MIP-1 as compared with that for LPS alone, with an even further increase when MIP-1 combined with LPS was instilled into the lungs of mice (Figure 4). Interestingly, the alveolar accumulation of neutrophils induced by MIP-1 or MIP-1 plus E. coli LPS was nearly completely blocked in mice pretreated with monoclonal anti-CD11a antibody, suggesting that the CD11a-independent recruitment of neutrophils into the alveolar air space of mice in response to JE/MCP-1 plus E. coli LPS was a chemokine-related effect (Figure 4).

View larger version (13K): [in this window] [in a new window]   Figure 4.   Effect of monoclonal anti-CD11a antibody on alveolar accumulation of neutrophils in response to alveolar deposition of MIP-1 or MIP-1 plus E. coli LPS. Mice received instillation of vehicle plus intravenous injection of isotype-matched control IgG (CL), or received instillation of MIP-1 in the absence or presence of E. coli LPS, as indicated. In some experiments, mice were pretreated with monoclonal anti-CD11a antibody (5 mg/kg) followed by intracheal instillation of MIP-1 (50 µg/mouse) in the absence or presence of E. coli LPS (10 ng/mouse). At 24 h later, mice were killed and accumulation of PMN in BALF was evaluated. *p < 0.05 versus corresponding treatments in the absence of monoclonal anti-CD11a antibody (n = 3 per treatment group). PMN = polymorphonuclear neutrophil.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study with intact mice, alveolar monocyte recruitment elicited by intratracheal instillation of JE/MCP-1 in the absence of E. coli LPS was strictly dependent on CD11b/ CD18, CD54, and CD49d, and partly dependent on CD11a, but not dependent on CD106. In response to JE/MCP-1 plus E. coli LPS, we observed additional engagement of CD11a and CD106 for enhanced monocyte recruitment to the alveolar space. Neutrophils comigrating under these conditions were found to utilize CD11b, CD18, and CD54 but not CD49d or CD106 to enter the alveolar air space of mice. Surprisingly, CD11a was not involved in neutrophil recruitment provoked by JE/MCP-1 plus E. coli LPS. In contrast, alveolar deposition of the C-C chemokine MIP-1 instead of JE/MCP-1 in both the absence and the presence of E. coli LPS induced a strictly CD11a-dependent neutrophil recruitment.

The accumulation of monocytes within the alveolar air space of mice in response to the C-C chemokine JE/MCP-1 in either the absence or the presence of low doses of E. coli LPS was blocked by > 90% when mice were pretreated with monoclonal anti-CD18 antibody, suggesting a CD18-dependent monocyte recruitment pathway under the experimental conditions chosen for our study. Strict dependency on ₂-integrin adhesion pathways has been described for neutrophil extravasation out of the pulmonary capillary bed and toward the alveolar compartment of rabbits and of mice in response to E. coli LPS, whereas ₂-integrin-independent emigration was observed in response to Streptococcus pneumoniae toxins or hydrochloric acid, demonstrating that the use of leukocyte adhesion pathways can vary with the stimulus being investigated (1, 8). With regarding the -chains of ₂-integrin, blockade of CD11a was much less effective than blockade of CD11b in inhibiting alveolar monocyte accumulation in mice challenged with intratracheal JE/MCP-1 in the absence of E. coli LPS. In contrast, monocyte recruitment driven by JE/MCP-1 plus E. coli LPS was strictly dependent on both CD11a and CD11b. Such a differential contribution of CD11a to monocyte trafficking under noninflammatory versus inflammatory experimental conditions was most recently demonstrated in an in vitro system of monocyte transmigration across unstimulated as opposed to TNF--activated human primary type I alveolar epithelial cells, and may at least in part be explained by increased ₂-integrin avidity on monocytes, as well as by upregulation of CD54 (ICAM-1) on capillary endothelial/alveolar epithelial cells under inflammatory conditions (11, 26).

In contrast to monocytes, comigrating neutrophils utilized CD11b but not CD11a to enter the alveolar compartment of mice in response to JE/MCP-1 plus E. coli LPS. To investigate whether this CD11a-independence was related to the chemokine employed, we instilled the C-C chemokine MIP-1, which was recently shown to recruit neutrophils into the lungs of BALB/c mice (4), together with E. coli LPS, into mice pretreated with monoclonal anti-CD11a antibody. Notably, the alveolar neutrophil accumulation in response to MIP-1 in either the absence or the presence of E. coli LPS was completely blocked by pretreatment of mice with monoclonal anti-CD11a antibody. These data collectively demonstrate that neutrophils may migrate out of the pulmonary capillary bed and into the alveolar compartment of mice via either partly CD11a-dependent (LPS alone) or completely CD11a-dependent (MIP-1 alone or MIP-1 combined with LPS), as well as via CD11a-independent adhesion pathways (JE/MCP-1 combined with LPS). Interestingly, the neutrophil accumulation within the alveolar air spaces of mice in response to E. coli LPS alone was recently suggested to occur in a CD11a- and CD11b-dependent manner, since inhibiting the function of either ₂-integrin -chain was sufficient to block the accumulation process (9). The most likely explanation for the partial discrepancy found for the role of CD11a in our study and that of Qin and colleagues (9) is that the relative contribution of CD11a and CD11b is strongly affected by the stimulus used to elicit the inflammatory response, in analogy to the aforementioned CD18-dependent and -independent neutrophil recruitment processes (8). This hypothesis is further supported by data from Issekutz and Issekutz (6), who demonstrated that in the rat, CD11a plays a major role in polymorphonuclear leukocyte migration to sites of C5a-induced but not to sites of IL-1- or LPS-induced dermal inflammatory reactions (6), although different mechanisms of leukocyte emigration out of the systemic as opposed to the pulmonary circulation must also be taken into account.

As anticipated, blockade of CD54 (ICAM-1), which is expressed on pulmonary endothelial as well as on type I and type II alveolar epithelial cells (11, 20) and which serves as the counterreceptor of the ₂-integrins CD11a/CD18 and CD11b/ CD18, was also found to effectively inhibit the alveolar accumulation of monocytes and comigrating neutrophils. Interestingly, the monoclonal anti-CD54 antibody YN1/1 used in our study was found to block both CD11a-dependent monocyte emigration and CD11a-independent neutrophil emigration into the alveolar air spaces of mice in response to JE/MCP-1 plus E. coli LPS. The YN1/1 antibody is assumed to bind to the LFA-1-binding domain I of CD54 (27), thus blocking LFA-1-dependent leukocyte adhesion (28, 29). Experimental data currently at hand, however, suggest that the YN1/1 antibody may affect both CD11a/CD18 (LFA-1) and CD11b/CD18 (Mac-1) interactions with CD54, possibly by provoking steric and/or conformational alterations of the CD54 molecule.

We also evaluated the role of CD49d and CD106 on the alveolar accumulation of monocytes in response to JE/MCP-1 in the absence and presence of E. coli LPS. CD49d is known to participate in both cell-ECM and cell-cell adhesive interactions (30). In particular, CD49d mediates adhesion to an alternatively spliced domain (CS-1) within the Hep II region of fibronectin, but also interacts with the cell surface molecule CD106 (31, 32). Interestingly, intravenous injection of the mAb PS/2, directed against the ₄ chain, into mice challenged with intratracheal JE/MCP-1 in the absence of E. coli LPS reduced alveolar monocyte accumulation by ~ 90%, whereas under the same experimental conditions, administration of monoclonal anti-CD106 antibody was without effect. In this context, our immunohistochemical analysis revealed that CD106 was not expressed on pulmonary endothelium under baseline conditions and was also hardly detectable in the pulmonary vascular bed of mice challenged with either low doses of LPS or JE/MCP-1, both of which findings partially corresponded to previous reports (16, 17, 33). Thus, it appears reasonable that under noninflammatory conditions, monocytic CD49d primarily interacts with ligands other than CD106, such as components of the ECM, to promote monocyte-ECM interactions. However, when mice were challenged with intratracheal JE/MCP-1 in the presence of E. coli LPS, blockade of either CD49d or CD106 effectively inhibited alveolar monocyte accumulation, suggesting that engagement of ₄₁ in interactions with both CD106 and ECM components participates in the transendothelial/epithelial migration of monocytes under these inflammatory conditions. Interestingly, intratracheal challenge of mice with JE/MCP-1 plus E. coli LPS provoked a marked and highly significant upregulation of CD106 expression in the pulmonary vascular bed. In contrast to the case with monocytes, the emigration of comigrating neutrophils in response to both LPS alone and JE/MCP-1 combined with LPS challenge was only slightly affected by either monoclonal anti-CD49d or anti-CD106 antibody, indicating that CD49d is less relevant for neutrophil transendothelial and transepithelial passage in response to JE/ MCP-1 plus E. coli LPS.

In a previous study of acute lung inflammation, the accumulation of ⁵¹Cr-labeled monocytes and ¹¹¹In-labeled neutrophils from donor rats within the lung alveoli of recipient rats challenged intratracheally with E. coli LPS (100 µg/rat) was found to be inhibited only when both ₄ and ₂-integrin pathways were blocked simultaneously, but not by pretreatment of recipient rats with either monoclonal anti-CD49d or anti-CD18 antibody (34). Although this study's findings were in contrast to ours, in which blockade of either pathway was sufficient to effectively inhibit alveolar monocyte accumulation in response to JE/MCP-1 plus E. coli LPS, stimulus- and species-dependencies must be considered for explaining this difference. Our observation that alveolar neutrophil recruitment was CD11a-independent in response to JE/MCP-1 plus E. coli LPS but CD11a-dependent in response to MIP-1 plus E. coli LPS clearly supports the notion that different inflammatory stimuli engage different pathways of adhesive interactions to promote leukocyte trafficking. In this regard, Ridger and colleagues most recently demonstrated that mAbs to CD49d were ineffective in blocking neutrophil migration in response to LPS-induced pulmonary inflammation, whereas monoclonal anti-CD49d antibody was effective in blocking neutrophil responses to the C-X-C chemokine KC (35).

In conclusion, in an intact mouse model allowing detailed phenotyping of leukocyte traffic into the alveolar space, we characterized molecular pathways engaged by monocytes and neutrophils for transendothelial and transepithelial passage in response to chemotactic challenge in the presence and absence of lung inflammation. The emigration process of monocytes recruited in response to JE/MCP-1 was primarily dependent on CD11b, CD18, CD54, and CD49d but not on CD106. In the presence of E. coli LPS, increased engagement of CD11a and CD54, and additionally of CD106 was noted in the process of alveolar monocyte emigration. Contrarily to monocytes, comigrating neutrophils utilized a substantially CD11a-independent adhesion pathway to enter the alveolar air space of mice in response to JE/MCP-1 combined with LPS, but challenge with different chemokines is linked with different adhesion molecule requirements for this type of leukocytes.