α_Mβ₂ Integrin Activation Prevents Alternative Activation of Human and Murine Macrophages and Impedes Foam Cell Formation

Abstract

Rationale

The alternative activation of monocytes by IL-13 and IL-4 is a significant component of the inflammatory response. The consequences of alternative activation in inflammatory diseases remain to be determined.

Objective

In this paper we explored how integrins, receptors important for monocyte migration to inflammatory sites, regulate IL-13-mediated monocyte activation. We focused on the analysis of two proteins, which are upregulated during the alternative activation and are important for the development of atherosclerosis - an oxidative enzyme 15-lipoxygenase (15-LO) and a scavenger receptor CD36.

Methods and Results

We found that adhesion of resting monocytes through β₂ integrins and inside-out activation of β₂ integrins by MCP-1 did not change IL-13-stimulated 15-LO upregulation; however, preincubation of monocytes with the antibody MEM48, which generates full activation of β₂ integrins, significantly inhibited 15-LO mRNA and protein expression. In contrast, activation of β₁ integrins had no effect on 15-LO expression. Analysis of integrin clustering through α_M, α_L, α_X and α_D subunits demonstrated the pivotal role for integrin α_Mβ₂ in inhibiting 15-LO expression. IL-13 treatment upregulates 15-LO-dependent CD36 expression on human monocytes, our studies showed that β₂ integrin activation and α_M integrin clustering significantly inhibited IL-13-dependent CD36 mRNA and protein expression as well as CD36-related foam cell formation. Moreover, IL-13 stimulation of α_M-deficient peritoneal macrophages demonstrated an upregulated level of 15-LO induction, CD36 expression and lipid accumulation as compared to wild type controls.

Conclusions

The adhesion of monocytes/macrophages through activated integrin α_Mβ₂ has a regulatory and potential athero-protective function during the alternative activation of macrophages.

Introduction

Monocytes/macrophages are critical components of host defense. During the development of the inflammatory response monocytes are recruited to the site of injury and migrate to the inflamed tissue, where they differentiate into macrophages ¹. A further role of macrophages in inflammation is determined by the type of activation program, which triggers specific gene expression and related cell functions. Classically activated M1 macrophages have long been known to be induced by IFN-γ, which initiates a harsh pro-inflammatory stimulus, whereas the alternatively activated M2 macrophages are induced by IL-4 and/or IL-13 and commonly believed to be associated with an anti-inflammatory response. However, recent data demonstrate both anti-inflammatory and pro-inflammatory effects for M2 macrophages on different stages of many diseases ².

Atherosclerosis is an inflammatory disease of the vessel wall characterized by monocyte infiltration in response to proatherogenic factors then the subsequent transformation of monocytes into lipid-laden macrophage foam cells, which accumulate in the subendothelial space³. The role of alternative activation of macrophages in the development of atherosclerosis is not clear. The experimental data obtained generally for IL-4-stimulated monocytes have demonstrated that exogenous IL-4 has protective effects, while knockout models for IL-4 components show the development of milder disease, indicating a proatherogenic role for endogenous IL-4 ⁴.

Monocyte activation and associated monocyte migration are complex and multifunctional processes, which involve simultaneous stimuli from different cell surface receptors. Integrins, adhesion receptors which guide monocyte migration, are good candidates for playing a role in regulating macrophage differentiation. During inflammation integrins are activated, which leads to the firm adhesion of monocytes to the endothelial cells, followed by monocyte transmigration through the endothelial monolayer and subsequent migration through the extracellular matrix (ECM) to the site of inflammation ⁵. The subfamily of β₂ integrins plays a leading role in this monocyte migration and the immune-inflammatory response. The group consists of four members, α_Lβ₂ (CD11a/CD18, LFA-1), α_Mβ₂ (CD11b/CD18, Mac-1), α_Xβ₂ (CD11c/CD18, p150,95) and α_Dβ₂ (CD11d/CD18), which have a common β₂ subunit and homologous α subunits ⁶. The role of β₂ integrins in atherogenesis is controversial. In particular, opposing roles for integrin α_X and α_M in the development of atherosclerosis have been recently defined. While α_X knockout mice on the background of ApoE-deficiency demonstrated decreased atherosclerosis ⁷, α_M knockout on the same background showed a progression of atherosclerosis ⁸, indicating that α_X has pro-atherogenic and α_M has anti-atherogenic functions; however, the mechanisms that initiate these opposite effects by two related integrins are not clear.

In this manuscript we studied how the alternative activation of monocytes mediated by IL-13 stimulation is regulated by β₂ integrins. We focused on the analysis of expression of the enzyme 15-lipoxygenase (15-LO), scavenger receptor CD36 and concomitant events related to the development of atherosclerosis. CD36 is a type II macrophage scavenger receptor, which is initially involved in the internalization of apoptotic cells, pathogens and modified low-density lipoproteins, but during atherogenesis CD36 is responsible for foam cell formation ⁹. The proatherogenic role of CD36 is well documented by in vitro and in vivo studies ^{10, 11}. 15-LO catalyzes hydroperoxidation of fatty acids, a reaction of potential relevance to inflammation, membrane remodeling, and atherosclerosis ¹². 15-LO is not expressed on circulating blood monocytes but is dramatically upregulated after IL-13 or IL-4 stimulation ^{13, 14}, providing a disease-relevant marker of alternative activation of monocytes.

In this study we report that IL-13-mediated induction of 15-LO is inhibited during β₂ integrin activation or clustering through α_M integrin. We also found that while IL-13 stimulation promotes the upregulation and surface expression of scavenger receptor CD36, a key protein in foam cell formation, the activation of β₂ integrin completely inhibited this effect. Moreover, β₂ integrin activation blocked CD36 related foam cell formation on monocyte-differentiated macrophages. Based on our results we suggest a regulatory athero-protective role of integrin α_Mβ₂ during IL-13-mediated alternative activation of macrophages.

Methods

Human peripheral blood monocytes were isolated using a Ficoll-Paque density gradient followed by adherence to bovine calf serum (BCS)-coated flasks as described earlier ¹⁵. α_M-knockout mice were generated in the laboratory of Dr. Christie Ballantyne (Baylor College of Medicine) ¹⁶. The experimental protocol for isolation of peritoneal macrophages was approved by the Cleveland Clinic Institutional Animal Care and Use Committee. Statistical analyses were performed using the Student’s t-test. An expanded Methods section is available in the Online Data Supplement at http://circres.ahajournals.org and includes information regarding the reagents and antibodies used in the study and detailed protocols for isolation of human monocytes, monocyte stimulation, adhesion assays, FACS analysis, cell sorting, western blotting, real-time quantitative RT-PCR, foam cell formation assays and analysis of α_M-deficient mouse peritoneal macrophages.

Results

Activation of β₂ integrins, but not β₁, integrins inhibits IL-13-mediated 15-Lipoxygenase expression in human monocytes

The dramatic upregulation of 15-lipoxygenase (15-LO) after monocyte stimulation with IL-4 and IL-13 is a well characterized hallmark during the alternative activation of macrophages ^{2, 13}. In this paper we studied which conditions can modify 15-LO induction and related events. Because integrins are important surface receptors that are involved in monocyte activation and migration, we tested the effect of integrin-dependent adhesion on IL-13-mediated 15-LO expression.

For this purpose we compared the induction of 15-LO in primary human monocytes incubated in non-adhesive polypropylene tubes or on cell culture plates. Although it has been shown that cell incubation in naked cell-culture plates initiates integrin-mediated adhesion ¹⁷, we also precoated some wells with fibrinogen, a plasma protein which can specifically interact with several monocyte integrins – α_Mβ₂, α_Xβ₂, α_Dβ₂ ^18–20. After 24 hours of incubation, cells were harvested, lysed and 15-LO expression was detected by immunoblotting. We found that all conditions evoked a similar level of 15-LO upregulation (Online Figure I), therefore adhesion through integrins on rested monocytes had no effect on IL-13-mediated 15-LO induction. However, the majority of integrins molecules on resting monocytes are expressed in the non-active conformation ²¹, therefore to mimic the conditions which exist during the monocyte transmigration to the inflammatory tissue we activated integrins β₂ and β₁, two major subfamilies of integrins expressed on monocytes. For these experiments monocytes were preincubated with anti-β₂ activation antibody (clone MEM48) and placed on plates precoated with fibrinogen or preincubated with anti-β₁ activation antibody (clone P4G11) and placed on plates precoated with fibronectin (a well characterized ligand for several β₁ integrins). The 15-LO protein expression was analyzed by immunoblotting at 24 hours after integrin activation (Fig. 1A, left panel). We found that IL-13 mediated 15-LO induction was dramatically inhibited in the presence of activation antibody against β₂ integrin, while stimulation with activation antibody against β₁ integrins had no effect on 15-LO expression. Tested in parallel, the β₂ blocking antibody also had no effect on 15-LO induction. Our analysis demonstrated that pretreatment with activation antibodies leads to a significant increase in monocyte adhesion, which was detected after 30 minutes (Fig. 1B), and to the marked changes in monocyte spreading, which was recorded after 24 hours (Fig. 1C). At the same time, preincubation with blocking β₁ and β₂ antibodies inhibited monocyte adhesion which confirmed the specificity of binding. Therefore, our data showed that activation of β₂ integrins, which leads to increased monocyte adhesion and cell spreading, inhibited IL-13-mediated 15-LO induction, but activation of β₁ integrins, which initiate similar events, failed to change this upregulation.

Figure 1. Induction and inhibition of 15-LO expression in primary human monocytes.

A. Activation of β₂ integrins inhibits IL-13 and IL-4 mediated 15-LO induction. Monocytes were preincubated with 10μg/ml of anti-β₂ function blocking antibody (IB4), anti-β₂ activation antibody (MEM48) or anti-β₁ activation antibody (P4G11) and then stimulated with IL-13 (left panel) or IL-4 (right panel). After 24 hours of incubation monocytes were harvested and 15-LO expression was analyzed by immunoblotting (upper panel). Equal loading was confirmed by immunoblotting with anti-β-tubulin antibody after membrane stripping (lower panel). Data are from a representative experiment of four with similar results. B. Monocyte adhesion to fibrinogen (Fg) and to fibronectin (Fn) is modified in the presence of blocking and activation antibodies. Grey Bars -Monocytes were preincubated with 10μg/ml of anti-β₂ function blocking antibody (IB4) or anti-β₂ activation antibody (MEM48) for 20 minutes and added to the plate with immobilized fibrinogen (10 μg/ml). White Bars - In parallel experiment monocytes were preincubated with 10μg/ml of anti-β₁ function blocking antibody (TDM29) or anti-β₁ activation antibody (P4G11) for 20 minutes and added to the plate with immobilized fibronectin (10 μg/ml). Adhesion assays were done as described in the Methods section. The experiment was repeated twice with a similar result. A representative experiment is shown. Data are expressed as a percentage of added cells and are the mean ± SD of triplicate wells for each experimental condition. C. Microscopic analysis of monocytes after adhesion in the presence of different antibodies. Monocytes were incubated on 6 well plates, precoated with fibrinogen or fibronectin, for 24 hours after IL-13 stimulation (original magnification x200). Data are from a representative experiment of four with similar results.

The alternative activation of monocytes can be initiated by stimulation with IL-4 and/or IL-13. To test the effect of β₂ integrin activation on the IL-4-mediated 15-LO induction we performed a similar experiment where monocytes pretreated with MEM48 were stimulated with IL-4. Our experiments demonstrated that activation of β₂ integrins blocked the 15-LO upregulation on IL-4 stimulated monocytes in the same manner (Fig. 1A, right panel). Therefore, activation of β₂ integrin inhibited 15-LO induction mediated by either IL-13 or IL-4.

Complete integrin activation is required for the suppression of 15-LO induction

Our results led us to a more detailed analysis of integrin activation and inhibition of 15-LO expression in alternatively activated monocytes. Circulating monocytes express integrins in a non-active conformation, which prevents monocyte adhesion to the endothelial cells and other activation-dependent functions. During the activation of the inflammatory cascade integrins are activated through an inside-out mechanism by a pro-inflammatory stimulus (MCP-1, TNF-α and etc.) on the surface of endothelium ²². This activation transforms integrins into the intermediate stage of activation, while subsequent interaction with the specific ligand (outside-in integrin activation) changes integrin conformation to the fully active stage ²³.

To explore how inside-out integrin signaling affects 15-LO induction we preincubated monocytes with MCP-1, a chemokine which induces inside-out activation of β₂ integrins ²⁴. To separate MCP-1 stimulation and follow integrin interaction with ligand, we performed these experiments in non-adherent polypropylene tubes. As shown above (Online Figure I), IL-13 treatment in polypropylene tubes evokes 15-LO induction. For these experiments monocytes were preincubated with 10 μg/ml of mAb MEM48 (anti-β₂ activation antibody) or with 10 nmol/L MCP-1. The level of integrin activation was detected with the conformation-dependent antibody against β₂ integrins mAb 24. This antibody can interact only with β₂ integrins expressed in the active conformation ^{25, 26}. Monocytes were analyzed by FACS after treatment with MEM48 and MCP-1. The right shift in fluorescence intensity indicates the more active conformation for β₂ integrins (Fig. 2A). Our results confirmed that the treatment with MCP-1 only slightly increased the activation stage of β₂ integrins, while incubation with mAb MEM48 significantly changed β₂ integrin conformation and transforms the integrin into its fully active stage (Fig. 2B).

Figure 2. Full β₂ integrin activation is required for inhibition of 15-LO induction.

The conformational stage of β₂ integrins on the surface of monocytes was quantified with the conformation-dependent antibody mAb 24. Monocytes were preincubated in polypropylene tubes for 5 minutes with mAb 24 conjugated to FITC, after which 10nmol/L MCP-1 or 10μg/ml MEM48 antibody was added to the cells. After 20 minutes of incubation cells were fixed with 10% formaldehyde and analyzed by flow cytometry. A. Representative FACS is shown. Thin solid line – incubation in the presence of EDTA, dashed line – in the presence of Mg²⁺, dotted line – treatment with MCP-1, Thick solid line –treatment with activation antibody MEM48. B. Mean fluorescence values are plotted based on three independent experiments. (* P< 0.01 compared to Mg²⁺ stimulated cells) C. Human monocytes were incubated with 10 nmol/L MCP-1 or 10 μg/ml MEM48 antibody for 5 minutes and IL-13 was added for an additional 24 hours. After incubation cells were harvested and 15-LO expression was analyzed by immunoblotting (upper panel). Equal protein loading was evaluated by stripping and reprobing with β-tubulin antibody (lower panel). Data are from a representative experiment of three with similar results.

The level of 15-LO in monocytes pretreated with mAb MEM48 or MCP-1 was analyzed after 24 hours. We found that while β₂ integrin activation in non-adherent conditions inhibited 15-LO induction, inside-out integrin activation with MCP-1 had no effect (Fig. 2C). Therefore, we concluded that conversion of β₂ integrins to the fully active conformation is required for the inhibition of IL-13-induced expression of 15-LO.

Integrin clustering through α_M integrin induces the inhibition of 15-LO induction in human monocytes

To obtain additional mechanistic insight into the role of β₂ integrins in the modification of monocyte responses after IL-13 treatment we tested the changes in 15-LO induction after clustering through β₂ integrin. For this purpose 5 μg/ml of β₂ antibody (clone IB4) was incubated with human monocytes in non-adherent polypropylene tubes. After 20 minutes of incubation, 2 μg/ml of goat-anti-mouse antibody was added for an additional 5 minutes followed by treatment with IL-13. As a control, the cells were preincubated in the presence of only the β₂ antibody for 20 minutes before IL-13 treatment. 15-LO expression was detected by immunoblotting after 24 hours of incubation (Fig. 3A). Our results demonstrated that clustering through β₂ integrins inhibited 15-LO induction similar to the effect observed after activation of β₂ integrin. The treatment with only the β₂ antibody had no effect.

Figure 3. Integrin clustering mediated by β₂ integrin AB or α_M integrin AB inhibits 15-LO induction.

Human monocytes were incubated in polypropylene tubes in the presence of anti-β₂ antibody (clone IB4) (A) or different anti-α integrin subunit antibodies (B) at 5 μg/ml. After 20 minutes normal mouse IgG (2 μg/ml) was added to cells to cluster the primary antibodies. After 5 minutes of incubation with normal mouse IgG IL-13 was added to cells for an additional 24 hours. As a control monocytes were incubated only in the presence of primary antibody (A, line 3). After incubation cells were harvested and 15-LO expression was analyzed by immunoblotting. Equal protein loading was evaluated by stripping and reprobing with β-tubulin antibody. C. The levels of β₂ integrin expression on the surface of primary human monocytes were analyzed with anti-integrin antibodies and evaluated by FACS. Mean fluorescence (M.F.) values are shown on the graphs. Data are from a representative experiment of three with similar results.

To obtain additional information about integrin receptors involved in suppression of 15-LO induction we performed clustering through different α subunits to determine which one was responsible for inhibition (Fig. 3B). Antibody against α_M, α_X, α_D and α_L integrins were preincubated with monocytes followed by incubation with goat anti-mouse antibody as described above. Analysis of 15-LO induction at 24 hours after IL-13 stimulation demonstrated that 15-LO expression was dramatically inhibited only in the presence of antibody against α_M integrin, while anti-α_L antibody had no effect and anti-α_X and anti-α_D antibodies showed very slight inhibition. This result demonstrated that integrin α_Lβ₂ is not involved in this process and integrin α_Mβ₂ is a major integrin receptor responsible for the inhibition of IL-13-induced 15-LO expression. Since the reported levels of integrin expression on monocytes are vary in the literature we tested the density of all β₂ integrins on primary human monocytes isolated for our studies (Fig. 3C). We found that all four integrin complexes, α_Mβ₂, α_Lβ₂, α_Xβ₂, α_Dβ₂, were expressed on the surface. While the expression of α_M and α_L integrins was moderate, the densities of α_X and α_D were lower but still readily detectable. Therefore, incubation with antibodies to all four subunits was able to induce integrin clustering, although only α_M integrin ligation evoked significant inhibition of 15-LO expression.

CD16⁻ and CD16⁺ monocyte subsets demonstrated a similar response to IL-13 stimulation and α_Mβ₂ integrin activation

We performed our prior experiments on the total population of monocytes isolated from peripheral blood. To better understand the specificity of 15-LO induction after IL-13 treatment we tested two major subsets of CD14⁺ human monocytes - CD16⁻ and CD16⁺. Corresponding to the published data the majority of freshly isolated monocytes were CD16 negative (Online Figure II, A). The amount of CD16 positive monocytes varied from 11% to 24% in different donors and additionally increased during monocyte incubation in tissue culture plates. Using double-colored FACS analysis we found that both CD16⁺ and CD16⁻ subsets expressed integrin α_M on the cell surface and α_M densities on both subsets are similar (Online Figure II, B). To test 15-LO induction we incubated the total monocyte population in the presence of IL-13 for 24 hours. Then cells were collected from the plates and labeled with PE-conjugated anti-CD16 antibody. The monocyte population was gated and separated on CD16⁺ and CD16⁻ subsets using fluorescence activated cell sorting (Online Figure II, C). Isolated subsets were lysed and analyzed by Western Blotting with 15-LO antibody. (Online Figure II, D). We found that both subsets induced 15-LO in response to IL-13 stimulation and this induction was inhibited by MEM48 on both CD16⁺ and CD16⁻ monocytes. Therefore, our result demonstrated that CD16⁻ and CD16⁺ monocytes have a similar mechanism of 15-LO induction and inhibition in response to IL-13 stimulation.

Activation of α_M integrin inhibits 15-LO mRNA induction

To better understand the mechanism of 15-LO inhibition by β₂ integrins we tested the changes in 15-LO mRNA expression after antibody treatment. Monocytes were incubated in the presence of mAb MEM48 followed by IL-13 stimulation as described previously. mRNA was isolated at different time points and analyzed by Quantitative Real Time RT-PCR. In agreement with published observations ²⁷ the level of 15-LO mRNA was gradually upregulated after IL-13 treatment, while cells pretreated with β₂-activated antibody demonstrated reduced 15-LO upregulation detectable even at 2 hours after IL-13 stimulation (Online Table I). After 24 hours, cells incubated with the β₂ integrin activation antibody or clustering through α_M integrin demonstrated significant inhibition of 15-LO mRNA expression (about 16 fold as compared with IL-13 stimulated cells) (Fig. 4). Therefore, this result confirmed our data obtained for 15-LO protein expression and suggested that inhibition occurs at the transcriptional level.

Figure 4. 15-LO mRNA expression is inhibited by α_M integrin activation.

Human monocytes were incubated in the presence of activating antibody to β₂ integrin or clustering through α_M integrin (as described in the Methods section) followed by stimulation with IL-13 for 24 hours. After incubation monocyte mRNA was isolated using a Qiagen Oligotex mRNA Midi Kit and levels of 15-LO mRNA were evaluated by quantitative real time RT-PCR. After normalization with GAPDH amplification, the fold induction of 15-LO mRNA expression after β₂ integrin activation or α_M integrin clustering was compared with the control. Data were collected from five independent experiments and are shown as means ± SE. Statistical analysis was performed using Student’s paired t-test, (* P< 0.05 compared to IL-13 stimulated cells).

The expression of mannose receptor during alternative activation of monocytes is inhibited after β₂ integrin activation

To better understand the role of β₂ integrin activation in alternative activation of macrophages we analyzed the change in the expression of the mannose receptor – a well established marker of alternative activation ²⁸. For this purpose monocytes were pretreated with the antibody MEM48 followed by activation with IL-13 as described above. After 5 days of incubation the expression of mannose receptor was tested by FACS (Online Figure III). Similar to published results we observed that the expression of mannose receptor in the presence of IL-13 was significantly increased (about 7 fold). At the same time preincubation with antibody MEM48 completely inhibited upregulation of the mannose receptor. Therefore, these data demonstrated that activation of β₂ integrin is not restricted to regulation of the 15-LO pathway, but rather has a common inhibitory effect on the alternative activation of macrophages.

Activation of a_M integrin inhibits IL-13 mediated CD36 expression

Recent data from our laboratory demonstrate that stimulation of human monocytes with IL-13 leads to the upregulation of scavenger receptor CD36 through a 15-LO dependent mechanism. Particularly we found that antisense oligonucleotides to 15-LO dramatically inhibited IL-13 stimulated CD36 expression (Bhattacharjee A. et al, manuscript in preparation). Based on the PPARγ dependent mechanism of CD36 expression we analyzed whether 15-LO induction is regulated by PPARγ agonist or antagonist. For this purpose human monocytes were treated with PPARγ agonist Rosiglitazone (10 μmol/L) or pretreated with PPARγ antagonist GW9662 (5μmol/L) and stimulated with IL-13. Monocytes were incubated for 24 hours and modification of 15-LO induction was analyzed by Western Blotting. We found no change in 15-LO expression after PPARγ agonist or antagonist treatment (Online Figure IV). Therefore, IL-13-mediated 15-LO induction is PPARγ independent.

Since the activation of β₂ integrin inhibits 15-LO expression we hypothesized that the same mechanism would suppress CD36 upregulation. To test this hypothesis, we analyzed CD36 mRNA expression 24 hours after β₂ integrin activation or α_M integrin clustering followed by IL-13 treatment (Fig. 5A). We found that the level of CD36 mRNA is upregulated about fourfold after IL-13 treatment and this upregulation is completely inhibited in the presence of either activation of β₂ integrin or α_M integrin clustering. For further analysis we tested the change in CD36 protein expression in human monocytes. Although the change in mRNA expression was detected at 24 hours, the change in CD36 protein expression required prolonged differentiation of monocytes to macrophages. The protein expression of CD36 remained at the same low level until 4 days after IL-13 stimulation. We could detect a considerable change in CD36 protein expression only after 5 days, when IL-13 treated cells showed significant upregulation of total CD36 protein as detected by Western blotting (Fig. 5B) and CD36 protein surface expression as detected by FACS (Fig. 5C). Similar to the CD36 mRNA analysis, CD36 protein expression was completely inhibited after β₂ integrin activation or clustering through α_M integrin. In summary, our data demonstrate on several levels that activation of α_Mβ₂ integrin on the surface of primary human monocytes inhibited IL-13 mediated expression of CD36.

Figure 5. CD36 expression is inhibited by α_M integrin activation.

A. Human monocytes were incubated as indicated in the Fig. 4 legend. After incubation the level of CD36 mRNA was evaluated by the real time quantitative RT-PCR after normalization with the GAPDH control. Data were collected from three independent experiments and are shown as means ± SE. Statistical analysis was performed using Student’s paired t-test, (** P< 0.01 compared to IL-13 stimulated cells). B. C. In a parallel experiment cells in the presence of β₂ integrin activating antibody (MEM48) or α_M integrin clustering antibody were incubated 5 days with IL-13 and then harvested. CD36 protein expression was evaluated by immunoblotting (B) or by FACS analysis (C). Legend for FACS data: thin solid line – unstained control (mean fluorescence (m.f.) − 23.5); dotted line – cells incubated without IL-13 (m.f. 155); thick solid line – cells incubated in the presence of IL-13 (m.f. 536); dashed line – cells incubated in the presence of IL-13 and activation antibody MEM48 (m.f. 118); cells incubated in the presence of IL-13 after clustering through α_M integrin (m.f. 92.6) demonstrated a effect similar to antibody MEM48 and result is not shown to simplify the graph. Data are from a representative experiment of three with similar results.

The foam cell formation is suppressed after α_M integrin activation

The uptake of oxidized LDL by differentiated macrophages through scavenger receptors, particularly CD36, leads to foam cell formation – the hallmark of atherosclerosis ¹¹. Based on our observation that CD36 expression is altered in our experiments we decided to analyze how foam cell formation would be affected after α_M integrin activation. We pretreated monocytes with mAb MEM48 followed by IL-13 stimulation as described above and differentiated cells for 5 days in DMEM with 10% lipoprotein-deficient serum (LPDS) media. To assess the role of CD36 in foam cell formation we exposed monocyte-differentiated macrophages to a fluorophore (DiI)-tagged form of NO₂ oxidized LDL (DiI-NO₂ LDL) for 30 minutes at 37°C and measured DiI-NO₂ LDL internalization by fluorescence microscopy (Fig. 6A). NO₂-oxidazed LDL is a preferential ligand for CD36 ²⁹. Our data demonstrate that uptake of oxidized LDL was increased more than 3 fold after IL-13 stimulation and significantly diminished after integrin activation (Fig. 6B).

Figure 6. α_M integrin activation reduced IL-13-mediated foam cell formation.

A. The uptake of oxLDL was assessed in human monocytes incubated in different conditions. Monocytes were plated in the presence or absence of mAb MEM48 followed by IL-13 stimulation and kept for 5 days in DMEM medium supplemented with 10% lipoprotein-deficient serum. One hour before the oxLDL treatment the media was exchanged to serum-free DMEM medium. Cells were incubated for 30 minutes in the presence of DiI-NO₂ oxLDL, washed, fixed, mounted on slides in the presence of DAPI and analyzed by fluorescence microscopy. Original magnification x400. Data are from a representative experiment of three that were performed. B. Five randomly selected fields from each slide were analyzed and DiI intensity was quantified using the computer-assisted image analysis software Image-Pro Plus. Data were normalized based on the total amount of cells in a particular field and plotted as the mean ± SD. Statistical analysis was performed using Student’s t-test (* P< 0.05; ** P<0.01). The experiment was repeated three times with the similar result.

Therefore, activation of β₂ integrins on alternatively activated M2 macrophages leads to functional changes in macrophage differentiation that initiate the inhibition of foam cell formation. Thus, our data demonstrate that α_Mβ₂ integrin activation may have an athero-protective effect during the development of atherosclerosis.

α_M-deficiency upregulates 15-LO induction, CD36 expression and foam cell formation on mouse peritoneal macrophages

To distinguish the role of integrin α_M in the inhibition of alternative activation we analyzed 15-LO induction in α_M-deficient mouse peritoneal macrophages. It has been shown previously that peritoneal macrophages express the upregulated level of 15-LO ³⁰. Wild type and α_M-deficient mice were injected with thioglycollate and after 72 hours peritoneal cells were isolated. All samples contained similar percentages of macrophages in the peritoneal exudate (around 80%) detected by FACS with F4/80 antibodies (data not shown). Macrophages were lysed and 15-LO expression was analyzed by Western Blotting. We found that endogenous stimulation led to slightly increased 15-LO induction in α_M-deficient mice (1.52 fold, P<0.05) (Online Figure V). To explore the effect of exogenous IL-13 we injected mouse IL-13 into the peritoneal cavity of wild type and α_M-deficient mice with pre-induced peritoneal inflammation. Mice were injected with thioglycollate and after 48 hours, when macrophages become the major subset of leukocytes in the peritoneum cavity, were injected intraperitoneally with IL-13 (100nmol/L). After an additional 24 hours, peritoneal cells were isolated, lysed and 15-LO induction was analyzed by Western Blotting. We found a significant increase in 15-LO induction in α_M-deficient macrophages (2.82 fold, P<0.05). (Fig. 7A,B). Therefore, these results demonstrate that integrin α_M regulates 15-LO expression and α_M-deficiency promotes the additional induction of 15-LO. To analyze the changes in CD36 expression in α_M-deficient macrophages in vitro, wild type and α_M-deficient mouse macrophages were isolated from the peritoneal cavity at 72 hours after thioglycollate injection, plated and mouse IL-13 was added to cells for 5 days of incubation. Both wild type and α_M-deficient macrophages strongly adhered to plastic and demonstrated a similar morphology and spreading. As a control we incubated macrophages in the presence of IL-13 in non-adhered polypropylene tubes. After incubation cells were harvested and analyzed by FACS. We found marked difference in the level of CD36 expression on non-adhered and adhered macrophages. The non-adhered wild type and α_M-deficient cells demonstrated a similar level of CD36 expression. Whereas in adhered macrophages the CD36 level was significantly reduced in wild type cells as compared with both α_M−/− adhered cells and wild type non-adhered cells (Fig. 7C). These results demonstrated that the strong adhesion via α_Mβ₂ suppressed IL-13-induced CD36 expression, while α_M-deficiency completely restored it. Moreover, because IL-13-mediated CD36 expression in the non-adhered condition is not α_M-dependent, the wild type and α_M-deficient cells displayed similar levels of expression. Therefore, these data demonstrated that strong adhesion through α_Mβ₂ integrin is critical for the inhibition of IL-13-mediated CD36 expression.

Figure 7. α_M -deficient mouse peritoneal macrophages demonstrated the enhanced level of 15-LO, CD36 and foam cell formation.

A. Wild type (n=4) and α_M-deficient mice (n=4) were injected with 0.5 ml of 4% thioglycollate. After 48 hours IL-13 was injected in the peritoneal cavity. Peritoneal cells were collected after an additional 24 hours. Cells were lysed and analyzed by Western Blotting with anti-15-LO antibody (upper panel). The 15-LO blot was stripped and reprobed with an antibody against β-tubulin (lower panel) to assess sample loading. B. Data were normalized and 15-LO protein expression was quantified using TotalLab TL120 Software (Nonlinear Dynamics, Durham, NC). Data were plotted as the mean ± SE. Statistical analysis was performed using Student’s t-test (* P< 0.05). C. In separate experiment peritoneal macrophages from wild type and α_M-deficient mice were isolated at 72 hours after thioglycollate injection. Macrophages were plated and stimulated with IL-13 for 5 days. Alternatively, macrophages were incubated in the presence of IL-13 in non-adhesive polypropylene tubes. After incubation the expression of CD36 by FACS was detected. The data represent the mean ± SE for non-adhered wild type mice (n=3); non-adhered α_M−/− mice (n=3); adhered wild type mice (n=6) and adhered α_M−/−mice (n=7). Statistical analysis was performed using Student’s t-test, (** P<0.01). D. Foam cells formation was evaluated on adherent peritoneal macrophages using DiI-NO₂ oxLDL as described above. Data were normalized based on the total amount of cells in a particular field and plotted as the mean ± SD. Statistical analysis was performed using Student’s t-test (** P<0.01).

To obtain additional evidence of regulation of CD36 expression by α_Mβ₂ we also tested foam cell formation in α_M-deficient macrophages incubated 5 days in the presence of IL-13 as described above. After treatment, cells were incubated for 30 min in the presence of DiI-NO₂ LDL and lipid uptake was analyzed. We observed a significant increase in foam cell formation by α_M-deficient macrophages, indicating that the increased CD36 actively functions as a scavenger receptor (Online Figure VI and Fig. 7D). Therefore, our data demonstrated that adhesion through α_Mβ₂ integrin is critical for the inhibition of IL-13-induced CD36 expression and foam cell formation verifying that integrin α_M is a major β₂ integrin responsible for the regulation of alternative M2 macrophage activation and the ability of these cells to accumulate lipids.

Discussion

IL-4 and IL-13 drive the alternative activation of macrophages to the M2 phenotype by expression of certain surface markers such as mannose receptor and also lead to the enhanced expression of 15-LO and CD36, well described participants of atherogenesis. Although IL-4 and IL-13-mediated alternative activation of macrophages has been historically attributed to anti-inflammatory stimuli, the analysis of major human diseases, including infection, inflammation, fibrosis and malignancy, demonstrate the two-edged nature of their function ². Corresponding to these observations, the role of alternatively activated macrophages in the pathogenesis of atherosclerosis is also complex, which is confirmed by numerous reports describing pro- and anti-atherogenic effects of critical components involved in the alternative activation of macrophages.

In particular, the role of IL-4 in atherosclerosis has been investigated in several mouse models. While the injection of exogenous IL-4 decreased aortic lesion formation in mice ³¹, IL-4 deficiency on the background of LDLR−/− or ApoE−/− leads to significant reduction of atherosclerosis ^{32, 33}. Taken together these data suggest that exogenous and endogenous IL-4 may have the opposite effects during the development of atherosclerosis. Based on the observations that effects of IL-4 and IL-13 are very similar because one chain (IL-4Rα) is common to both receptors ³⁴ we can hypothesize that IL-13 may have similar functions in atherogenesis.

Opposing effects on the development of atherosclerosis were also detected for 15-lipoxygenase (15-LO). Studies of 12/15-LO deficiency (mouse analog of human 15-LO) demonstrated that 12/15-LO gene disruption lead to a significant reduction of atherosclerotic lesion progression in ApoE−/− ³⁵ and LDLR−/− ³⁶ mouse models of atherosclerosis. In addition, mice overexpressing human 15-LO are more susceptible to the development of atherosclerotic lesions than littermate controls ³⁷. At odds with these observations, transgenic rabbits overexpressing human 15-LO in their macrophages developed significantly smaller atherosclerotic lesions when fed a western type diet ³⁸. In addition, a recently discovered polymorphism in the 15-LO promoter showed a tendency to be protective against atherosclerosis in a case–controlled study for coronary artery disease involving 498 Caucasian participants ³⁹. Therefore, 15-LO induction during IL-13 stimulation appears to have pro- or anti- atherogenic functions depending on selected models. However, in the context of upregulation of CD36 and augmentation of foam cell formation we consider that our data is consistent with studies that suggest a proatherogenic role of IL-4, IL-13 and 15-LO during alternative activation of monocytes.

Recent studies suggest a different role for two major subsets of CD14⁺ human monocytes CD16⁻ and CD16⁺. Based on the better described mouse analog – Gr1⁺/Ly6C^high and Gr1⁻/Ly6C^Low monocytes subsets, CD16 negative monocytes are referred to as “inflammatory monocytes”, whereas CD16 positive monocytes are believed to be “patrolling” monocytes, which crawl in the luminal side of endothelium to survey surrounding tissue and probably are responsible for the scavenging of lipids and dying cells ⁴⁰. Although, the expression of CD36 on CD16⁺ non-stimulated monocytes is slightly reduced ⁴¹, the uptake of CD36-dependent oxLDL is similar on CD16⁻ and CD16⁺ monocytes isolated from the healthy donors ⁴². In our study we tested how these two subsets respond to IL-13-mediated 15-LO induction. We found that both CD16⁺ and CD16⁻ monocytes induced 15-LO expression and this induction is inhibited by an activation antibody to α_Mβ₂ integrin. Based on our data we conclude that 15-LO pathways are similar in both monocytes subsets, however, considering observations that CD16⁺ monocytes are predisposed to become migratory dendritic cells ⁴¹, we cannot exclude that while CD16⁻ monocytes contributes to the macrophage foam cell accumulation in the atherosclerotic lesion, CD16⁺ monocytes capture oxLDL and emigrate from the lesion site to the lymphatics.

In this study we found that IL-13- and IL-4-mediated 15-LO induction and CD36 expression are significantly inhibited by activation of β₂ integrins. Integrin activation is an important step during monocyte migration to inflammatory sites. Our data demonstrate that only complete integrin activation (outside-in integrin activation) inhibits IL-13-mediated 15-LO induction, with MCP-1 treatment alone (inside-out integrin activation) insufficient to generate that signal (Fig. 2). Monoclonal antibody MEM48 was described as an activation antibody for β₂ integrins. The antibody interacts with cysteine rich repeats in the β₂ subunit stalk region and activates all four integrin complexes from the β₂ integrin subfamily ⁴³. Therefore, this antibody alone (as well as other known β₂ integrin activation antibodies) cannot specify which α subunit is involved in blocking IL-13 mediated functions. However, there are several lines of evidence that allow us to suggest that integrin α_Mβ₂ is primarily responsible for this effect on monocytes. First, although the fact that mAb MEM48 activates all β₂ integrins is well established, special analysis demonstrated that this antibody activates 90% of α_Lβ₂, about 70% of α_Mβ₂ and only 16% of α_Xβ₂ integrins on the cell surface ⁴⁴. Integrin α_Dβ₂ was not tested in that study. Therefore, integrin α_Xβ₂ is not activated by this antibody. Second, integrin α_Lβ₂ is a receptor for ICAM molecules and cannot recognize fibrinogen ⁴⁵, therefore integrin activation and related monocyte adhesion to fibrinogen cannot be mediated through α_Lβ₂ integrin. Third, clustering through different integrin subunits demonstrated that only antibodies to the β₂ subunit and α_M subunit dramatically inhibit 15-LO induction (Fig. 3). And finally, the analysis of 15-LO and CD36 expression on α_M-deficient mouse peritoneal macrophages demonstrated that strong adhesion through α_Mβ₂ is required for the blocking of these IL-13 mediated events. β₂ integrins on peritoneal macrophages are activated due to transmigration through the endothelium and the peritoneal tissue, therefore these macrophages do not require additional activation for strong adhesion and spreading. Therefore, summarizing we can propose that integrin α_Mβ₂ is the major β₂ integrin, which is responsible for the regulation of IL-13-mediated alternative activation of monocytes/macrophages.

Based on the inhibition of CD36 expression (Fig. 5) and foam cell formation (Fig. 6) our data suggest a potential athero-protective effect of integrin α_Mβ₂ in the development of atherosclerosis. The role of β₂ integrins in atherogenesis is not clear yet. It was previously suggested that β₂ integrins have proatherogenic functions ⁴⁶ and recent analysis of integrin α_X in atherosclerosis is consistent with this hypothesis. α_X-knockout mice crossbred with ApoE-deficient mice demonstrated a significant reduction of lesion formation compared to ApoE−/− control⁷. The examination of α_M-knockout on an LDLR−/− background demonstrated no difference in the development of atherosclerosis ⁴⁷. Surprisingly, the recent analysis of α_M–deficiency on ApoE−/− mice showed an increase of lesion formation, suggesting that α_M has an athero-protective effect on the development of atherosclerosis ⁸. It has been shown that CD36 expression is upregulated by oxidized LDL and IL-4 by a common signaling pathway via the transcriptional factor PPARγ. Our data demonstrated that IL-13-mediated CD36 expression is also PPARγ dependent. (Bhattacharjee A. et al, manuscript in preparation). Based on the common signaling pathways for oxLDL and IL-13 enhancement of CD36 expression we hypothesize that integrin α_M may inhibit oxLDL-stimulated CD36 expression by similar mechanisms. Indeed, the analysis of lipid deposition in peritoneal macrophages isolated from α_M/ApoE-deficient mice showed a marked increase in foam cell formation 8. And this result is consistent with our finding that IL-13-stimulated peritoneal macrophages demonstrated an increased level of foam cell formation in α_M-deficient cells (Fig. 7D). Therefore, α_M activation is required for controlling CD36 expression on macrophages, diminishing lipid deposition and potentially reducing lesion size.

In summary, although alternatively activated M2 monocyte/macrophages are often considered to be anti-inflammatory in nature, the IL-13-induced 15-LO/CD36/foam cell axis suggests that this delineation is not so simple and the role of β₂ integrin activation in inhibiting 15-LO, CD36 and foam cell formation is a novel and unexpected finding, which demonstrates a potential regulatory role for the integrin α_Mβ₂ on alternative activation of macrophages during atherogenesis.

Non-standard Abbreviations and Acronyms

15-LO

15-lipoxygenase

ECM

Extracellular matrix

BCS

bovine calf serum

DiI

1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate

OxLDL

oxidized low density lipoprotein