Interleukin-1β Causes Acute Lung Injury via αvβ5 and αvβ6 Integrin–Dependent Mechanisms

Acute lung injury (ALI) is a devastating clinical syndrome in critically ill patients with an overall mortality rate of 30% to 40%.¹ The syndrome is characterized by alveolar epithelial and lung endothelial injury leading to increased permeability across the alveolar–capillary barrier, pulmonary edema, and acute respiratory failure.² Despite an improved understanding of the pathogenesis of ALI in recent years, the molecular steps regulating the development of increased lung endothelial and epithelial permeability remain poorly understood, and no specific pharmacological therapies are currently available.

During the early phase of ALI, a variety of inflammatory mediators are released into the distal air spaces.² Among those, interleukin (IL)-1β has been shown to be among the most biologically active cytokines in the lungs early after the onset of ALI.^3–5 Furthermore, IL-1β stimulates the production of a variety of chemokines (eg, IL-8, monocyte chemotactic protein [MCP]-1, and macrophage inflammatory protein [MIP]-1α)⁶ involved in epithelial wound repair^7,8 and is a potent inducer of lung fibrosis.^9,10 It has been previously shown in rats that lung vascular permeability increases after short-term exposure of IL-1α and IL-1β protein when given intratracheally.^11,12 However, the exact mechanism by which IL-1β causes the increased lung vascular permeability is unknown.

Therefore, the objective of this study was to determine the mechanisms of IL-1β–mediated increase in lung vascular permeability and pulmonary edema following transient overexpression of this cytokine in the lungs by adenoviral gene transfer. Our results demonstrate a critical role for the αvβ5/β6 integrins in mediating the IL-1β–induced ALI and indicate that these integrins could be a potentially attractive therapeutic target in ALI.

Materials and Methods

Reagents and antibodies (Abs), recombinant adenovirus expressing human (h)IL-1β (Ad-hIL-1β) and its control vector (Ad-Empty), details on the animal studies (including animals, experimental protocol, measurement of lung endothelial permeability to protein and pulmonary edema, bronchoalveolar lavage [BAL] fluid and cell count, and measurement of IL-1β) and in vitro studies (including cell culture, measurement of transepithelial and transendothelial albumin fluxes, bioassay for transforming growth factor [TGF]-β activation, RhoA activation assay, Western blot analysis for detecting phospho-Smad2, immunocytochemistry for visualization of actin stress fibers, immunoprecipitation for detecting phospho–β-catenin, and cell viability assay) are described in the online data supplement, available at http://circres.ahajournals.org.

Statistical Analysis

All data are summarized as means±SEM. One-way ANOVA and the Fisher’s exact t test were used to compare experimental with control groups. A probability value of <0.05 was considered statistically significant.

Results

IL-1β Increases Lung Vascular Permeability via an αvβ6- and TGF-β–Dependent Mechanisms

Transient overexpression of IL-1β in lungs of mice by adenoviral gene transfer caused an increase in hIL-1β concentrations in BAL fluid at days 4 and 7 after instillation of the adenovirus (Figure 1A) that was associated at day 7 with an increased lung endothelial permeability to protein (extravascular plasma equivalents) and the development of pulmonary edema (excess lung water) (Figure 1B and 1C). Over time, the permeability changes decreased and returned back to baseline after 28 days (Figure 1B and 1C). We found that the hIL-1β–mediated increase in lung vascular permeability was associated with a significant increase in total cell and neutrophil BAL count, with a maximal effect 7 days after Ad-hIL-1β intratracheal instillation (total BAL cell count: 18±3×10⁴ cells [Ad-Empty] versus 262±42×10⁴ cells [Ad-hIL-1β], P<0.05; BAL neutrophil count: 5±4% [Ad-Empty] versus 86±3% [Ad-hIL-1β], P<0.05).

• Download figure
• Download PowerPoint

We have previously found that the increased lung vascular permeability caused by lipopolysaccharide is TGF-β–dependent and that αvβ6 integrin is critical for the local activation of TGF-β.¹³ Furthermore, the receptors for bacterial products such as Gram-negative–derived lipopolysaccharide belong to the IL-1 receptor (IL-1R)/Toll-like receptor superfamily and involved in host defense and inflammation. Thus, we hypothesized that IL-1β may increase lung vascular permeability by mechanisms comparable with lipopolysaccharide.¹⁴ We treated mice with blocking Ab to the αvβ6 integrin and/or soluble chimeric receptor TGF-β type II. Inhibition of the αvβ6 integrin and/or TGF-β in the presence of IL-1β significantly reduced the increase in lung vascular protein permeability by 50% and the quantity of pulmonary edema by 30%. (Figure 1D and 1E). When the blocking Ab to the αvβ6 integrin and a soluble chimeric receptor TGF-β type II were administered together in mice subsequently instilled with Ad-hIL-1β, there was no additional inhibition of the IL-1β–induced increase in lung vascular permeability (Figure 1D and 1E). Taken together, these results indicate that the adenoviral gene transfer of IL-1β caused an increase in lung vascular permeability that is partially mediated by αvβ6 integrin and TGF-β signaling.

IL-1β Increases Transepithelial Albumin Flux in Alveolar Epithelial Type II Cells via an αvβ6-Dependent TGF-β Activation

To better understand the mechanism by which IL-1β causes the development of protein-rich alveolar edema, we determined whether IL-1β might exert its effect on increasing lung epithelial permeability by activating TGF-β via αvβ6 integrin–dependent mechanism. IL-1β increased epithelial permeability in polarized rat ATII cells and this effect was completely blocked by Ab against the integrin αvβ6 or TGF-β (Figure 2). The next series of experiments were designed to determine whether IL-1β would activate TGF-β in rat ATII cells and whether other cytokines and chemokines present in the airspace during the early phase of ALI would also activate TGF-β. Rat ATII cells were stimulated with IL-1β, tumor necrosis factor-α, interferon-γ, IL-11, cytokine-induced neutrophil chemoattractant-1, KC (keratinocyte-derived chemokine, MCP-1, MIP-2, MIP-3β, or exodus-2 after being cultured in coculture system with mink lung epithelial reporter cells to measure active TGF-β. Among all mediators tested, only IL-1β augmented the luciferase activity in these reporter cells, thus indicating activation of TGF-β by IL-1β in ATII cells (Figure 3A). TGF-β activation was prevented by blocking the αvβ6 integrin or TGF-β, as shown by the luciferase activity in the reporter cells (Figure 3B) or by the inhibition of the RhoA-activated kinase (ROCK) (immediate downstream effector of RhoA) (Figure 3C). Furthermore, Smad2 phosphorylation in ATII cells, a protein that is activated by the binding of TGF-β to its receptors and used as a marker of the activation of the TGF-β cell signaling was prevented by blocking the αvβ6 integrin or TGF-β (Figure 3D).¹⁵ In addition to its activation of TGF-β via an αvβ6 integrin–dependent mechanism, IL-1β also increased by 40% the expression of TGF-β mRNA measured by RT-PCR in rat ATII cells. Finally, IL-1β activated RhoA signaling in rat ATII cells, an effect that was not inhibited by blocking the αvβ6 integrin, suggesting that RhoA signaling was upstream of the αvβ6 integrin in the IL-1β signaling leading to TGF-β activation (Figure 3E). Alamar Blue assay indicated that none of the mediators at the concentrations used in these experiments significantly increased cell death (data not shown). Taken together, these data demonstrate that IL-1β increases lung epithelial permeability via an RhoA/αvβ6-dependent TGF-β activation and that IL-1β is the only mediator among a series of inflammatory mediators important in ALI that activates TGF-β signaling.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

IL-1β Increases Transendothelial Albumin Flux in Pulmonary Artery Endothelial Cells via αvβ5 Integrin– and RhoA-Dependent Mechanisms

The results of our in vivo experiments indicate that the inhibition of the TGF-β signaling inhibits only 50% of the increase in vascular permeability induced by adenoviral IL-1β gene transfer to the lungs. We next determined whether IL-1β could directly increase the permeability across bovine pulmonary arterial endothelial cell (BPAEC) monolayers. We found that IL-1β induced a dose- and time-dependent increase in endothelial permeability, an effect that was completely blocked by its receptor antagonist IL-1RA (Figure 4A and 4B).

• Download figure
• Download PowerPoint

We previously identified the αvβ5 integrin as among the central regulators of the lung endothelial paracellular permeability.¹⁶ To determine the role of αvβ5 integrin in mediating IL-1β–induced increase in lung endothelial permeability, monolayers of BPAECs were pretreated with αvβ5 blocking or isotype control Ab and subsequently stimulated with IL-1β. We found that blocking the αvβ5 integrin function completely inhibited the IL-1β–mediated increased permeability across BPAEC monolayers (Figure 5A) and actin polymerization (Figure 5B). These in vitro results were confirmed by a last series of in vivo experiments in which mice receiving Ad-hIL-1β intratracheally were treated with αvβ5 blocking or isotype control Ab. IL-1β–mediated increase in vascular permeability was markedly attenuated in mice treated with αvβ5 blocking Ab (Figure 5C and 5D). In the next series of experiments, we determined that the simultaneous pretreatment with specific Abs to αvβ5 and αvβ6 integrin provided an additive protective effect against the IL-1β–mediated increase in vascular permeability (Figure 5C and 5D), demonstrating the critical role of these integrins in mediating the increase in lung vascular permeability induced by IL-1β.

• Download figure
• Download PowerPoint

It has been reported previously that the activation of the small GTPase RhoA is important for the control of the endothelial barrier function by regulating the actin cytoskeletal organization and integrity of intercellular endothelial junctions.¹⁷ Thus, we examined whether this small GTPase was involved in the cell signaling induced by IL-1β in BPAECs. We found that IL-1β activated RhoA within 10 minutes after stimulation in BPAECs (Figure 6A). Interestingly, this effect was not inhibited by blocking the αvβ5 integrin, suggesting that RhoA signaling was upstream of the αvβ5 integrin in the IL-1β signaling (Figure 6A). Furthermore, the increase in endothelial albumin flux by IL-1β was completely blocked by the inhibition of ROCK (the immediate downstream effector of RhoA) (Figure 6B). To further determine the mechanisms of IL-1β–mediated increase in protein permeability, we examined whether exposure of BPAECs to IL-1β would affect the actin cytoskeleton. IL-1β stimulation induced the formation of actin stress fibers that was blocked by exposing BPAECs to a ROCK inhibitor before IL-1β stimulation (Figure 6C). Finally, Alamar blue assay indicated that none of the mediators at the concentrations used in these experiments significantly increased cell death (data not shown).

• Download figure
• Download PowerPoint

Previous studies have identified that inflammatory mediators, such as thrombin, vascular endothelial growth factor, and TGF-β, cause myosin phosphorylation, actin polymerization, and disruption of the adherens junctions by phosphorylation of its components, such as β-catenin, and subsequent degradation by the ubiquitin–proteasome system.^16,18 In the present study, we found that IL-1β causes a phosphorylation of β-catenin in BPAECs (Figure 7A and 7C) and that this phosphorylation was absent by pretreating the endothelial cells with a ROCK inhibitor or an Ab to the αvβ5 integrin (Figure 7A and 7C). Furthermore, IL-1β causes the formation of paracellular gaps between lung endothelial cells that was largely prevented by a pretreatment with a ROCK inhibitor or an Ab to the αvβ5 integrin (Figure 7B and 7D). Finally, the formation of paracellular gaps was associated with the decrease in the expression of β-catenin at the cell membrane, indicating a disassembly of the adherens junction complex (Figure 7B and 7D). Taken together, these results indicate that IL-1β disrupts the adherens junctions and causes the contraction of endothelial cells via a RhoA- and αvβ5-dependent mechanisms that lead to increased paracellular permeability.

• Download figure
• Download PowerPoint

Discussion

Previous clinical studies have reported that IL-1β is among the most biologically active cytokines in the airspace of patients with ALI.^3–5 IL-1β has been shown to inhibit fluid transport across the distal lung epithelium,¹⁹ to cause surfactant abnormalities,²⁰ and to increase protein permeability across the alveolar–capillary barrier.²¹ However, the exact mechanism by which IL-1β causes an increase in lung epithelial and endothelial permeability is unknown and was examined here following the transient overexpression of IL-1β in the lungs by adenoviral gene transfer. The effect of IL-1β on lung epithelial and endothelial permeability has been examined in previously published studies demonstrating that the increase in permeability across the alveolar–capillary barrier following the instillation of IL-1α protein into the airspace was blocked by pretreatment with TGF-β blocking Ab.²² Furthermore, we recently reported that the increase in lung epithelial and endothelial permeability caused by a clinically relevant experimental model of ventilator-induced lung injury was TGF-β–dependent.¹⁶ However, these studies did not provide any information about a possible causal relationship between IL-1β and TGF-β in mediating lung injury caused by ventilator-induced lung injury. Thus, this is the first study to describe that pretreatment with a soluble chimeric TGF-β type II receptor attenuates the increase in lung epithelial and endothelial permeability induced by adenoviral gene transfer of IL-1β into the lungs, establishing a direct causal relationship between the release of IL-1β within the airspace, the activation of TGF-β in the lung, and the increase in epithelial and endothelial permeability induced by IL-1β signaling.

What are the mechanisms of TGF-β activation in the lungs? TGF-β is synthesized as a protein complex including the active mature TGF-β (carboxy-terminal fragment) and a disulfide-linked homodimer of the amino-terminal fragment (latent-associated peptide).²³ This complex is further bound to members of another protein family termed TGF-β latent binding proteins. This complex is secreted as an inactive complex that is extracellularly cross-linked to fibronectin, so that latent inactive TGF-β can be stored in large quantity in the extracellular matrix of the lungs.²³ Several mechanisms activate TGF-β outside of the cells. First, mild protein denaturation induced by changes in temperature or pH or exposure to oxidants or ionizing radiation.²⁴ Second, several proteases, such as plasmins, can release free and active TGF-β, although the in vivo importance of this activation is unknown. Third, 2 additional mechanisms of TGF-β activation have probably more in vivo relevance: interaction of TGF-β with the secreted protein thrombospondin-1, leading to its activation,²⁵ and TGF-β activation by the integrin αvβ6 (Figure 8).²⁶ IL-1 receptor type I (IL-1RI) (the receptor for IL-1β) belongs to the IL-1R/Toll-like receptor superfamily that includes the receptors for bacterial products such as Gram-negative bacteria–derived endotoxin.¹⁴ We have previously shown that exposure to Escherichia coli endotoxin caused the in vivo activation of TGF-β that was responsible for the increase in lung epithelial and endothelial permeability induced by this bacterial product.¹³ Thus, we tested whether IL-1β would cause an increase in lung permeability across the alveolar–capillary barrier via a αvβ6-dependent mechanism. We found that inhibition of the αvβ6 integrin by treatment with blocking Ab to this integrin significantly decreased the IL-1β–induced increase in lung epithelial and endothelial permeability. Interestingly, when mice were treated with blocking Ab to the αvβ6 integrin and a soluble chimeric TGF-β type II receptor, there was no further inhibition of the IL-1β–mediated lung edema, indicating that TGF-β may be activated primarily by the αvβ6 integrin.

• Download figure
• Download PowerPoint

Does IL-1β activate TGF-β via the αvβ6 integrin pathway? It has recently been shown that protease-activated receptor-1 agonists activate TGF-β in an αvβ6 integrin–dependent manner via a RhoA-dependent mechanism.¹⁵ We determined that IL-1β was the only mediator among several, such as tumor necrosis factor-α, IL-8, MIP-1, or MCP-2, that are important in the early phase of ALI, to cause the activation TGF-β via RhoA and αvβ6 integrin–dependent mechanisms. It has previously been shown that the IL-1 receptor forms a multiprotein complex that can directly activate the small GTPase RhoA and causes actin stress fiber formation in epithelial cells.²⁷ Because we have previously found that binding to the actin cytoskeleton is required for αvβ6 integrin to allow a change in its conformation that will expose the active form of TGF-β to its receptor,²⁶ it explains why the activation of RhoA is necessary for IL-1β to activate the αvβ6 integrin/TGF-β pathway. Taken together, these results indicate that IL-1β causes the development of alveolar edema via the activation of the RhoA/αvβ6 integrin/TGF-β pathway.

Our results indicate that the in vivo inhibition of the αvβ6 integrin/TGF-β pathway prevented only half of the IL-1β–mediated increase in lung epithelial and endothelial permeability. We hypothesized that the release of IL-1β within the airspace may also directly cause the development of pulmonary edema via binding to its receptors on lung endothelial cells. The IL-1β–mediated increase in lung endothelial permeability was RhoA-dependent, but not src-dependent (data not shown), as we have previously reported for other important mediators of ALI.¹⁶ Passage of solutes through the endothelial barrier is thought to occur via transcellular pathway or through receptor-activated transcytosis.²⁸ The relative contribution of these pathways remains incompletely understood. However, it has been suggested that the formation of actin stress fibers and of gaps between cells as a consequence of imbalanced competition among cytoskeletal, adhesive cell–cell, and cell–matrix forces leads to an increased flux of solutes and protein via the paracellular pathway.²⁸ In our experimental model, IL-1β induced the formation of actin stress fibers and paracellular gaps and caused the disassembly the adherens junction complex. In addition, these results were confirmed by the next series of experiments that showed a critical role for the αvβ5 integrin in controlling the increase in lung endothelial permeability as well as the formation of actin stress fibers mediated by IL-1β. Although our studies did not directly distinguish between paracellular and transcellular pathways, these results suggest that the paracellular pathway plays a role in the IL-1β–mediated increase in lung epithelial and endothelial permeability.

A limitation of this series of studies is that we used cells from various species as well as cells derived from proximal pulmonary macrovascular endothelium. Indeed, microvascular lung endothelial cells would be more relevant to study pulmonary capillary leak, because several studies have shown significant physiological differences between lung cells from microvascular and macrovascular beds.^29,30 To address this question, we performed additional series of in vivo experiments in mice by blocking either the αvβ5 or both the αvβ5 and αvβ6 integrins. The results showed that the IL-1β–mediated development of pulmonary edema was attenuated in mice treated with a blocking Ab to the αvβ5 integrin and further reduced in mice treated with blocking Abs to both the αvβ5 and αvβ6 integrins. Thus, taken together, these results demonstrate the consistency of the IL-1β signaling pathway across species and support a critical role for both integrins in mediating the increase in lung endothelial paracellular permeability induced by exposure to IL-1β.

In summary, these studies demonstrate for the first time a critical and sequential role for the αvβ5/β6 integrins in mediating development of alveolar edema caused by IL-1β, which is among the most biologically important inflammatory mediators in the airspace of patients with early ALI.^3–5 We specifically found that the inhibition of αvβ6 integrin/TGF-β signaling completely blocked the IL-1β–mediated protein permeability across the distal lung epithelium and prevented, in large part, the development of pulmonary edema following transient overexpression of IL-1β in the lungs by adenoviral gene transfer. Furthermore, IL-1β increased the protein permeability across lung endothelial cell monolayers via RhoA- and αvβ5 integrin–dependent mechanisms. Finally, the combined inhibition of the αvβ5 and αvβ6 integrins considerably reduced the lung injury associated with the adenoviral gene transfer of IL-1β into the lungs. The findings reported here have potential clinical relevance. Indeed, transient blockade of both integrins by humanized Ab or small RGD peptides may provide new causal therapies for pulmonary edema from ALI, a substantial cause of morbidity and mortality in critically ill patients that is currently largely untreatable.

Footnotes