Activation of Myeloid Cell-Specific Adhesion Class G Protein-Coupled Receptor EMR2 via Ligation-Induced Translocation and Interaction of Receptor Subunits in Lipid Raft Microdomains

Reagents and antibodies.

Unless otherwise specified, general reagents and antibodies were obtained from Sigma-Aldrich (St. Louis, MO). DNA and protein reagents were obtained from Invitrogen (Carlsbad, CA), Qiagen (Valencia, CA), Fermentas (Ontario, Canada), New England Biolabs (Massachusetts), or Amersham (GE Healthcare). The monoclonal antibodies (MAbs) used in the study include EMR2 stalk-specific 2A1 and anti-CD4 (1F6) from AbD Serotec (Kidlington, United Kingdom), anti-c-myc (9E10) from Invitrogen; anti-HA.11 (16B12) from Covance (New Jersey), anti-CD71 (H68.4) from Zymed Laboratories (San Francisco, CA), anti-β-actin (clone C4) and chicken anti-DAF (CD55) polyclonal antibodies from Chemicon (California), and anti-caveolin-1 (7C8) from Upstate (Lake Placid, NY). Fluorochrome-conjugated goat anti-mouse IgG was from Jackson Immunoresearch (West Grove, PA). Mouse IgG1 isotype control (clone 11711) was from R&D Systems (Minnesota). The plasmids containing PAR1 and CD4-HA-AATN cDNAs were kindly provided by Hua-Wen Fu and Shaun R. Coughlin (45) and Min Li (39), respectively.

Cell culture and transient transfection.

All cell culture media and supplements, including 10% heat inactivated fetal calf serum (FCS), 2 mM l-glutamine, 50 IU of penicillin/ml, and 50 μg of streptomycin/ml, were purchased from Invitrogen. CHO-K1 and HEK293T cells were cultured in Ham F-12 and Dulbecco modified Eagle medium, respectively. THP-1 cells were cultured in RPMI 1640 and differentiated to macrophage-like cells with 1 μg of phorbol 12-myristate 13-acetate/ml. Baf/3b stable cells were described previously (52). HT1080 stable cells were maintained in minimal essential medium containing 0.3 μg of G418/ml, nonessential amino acids, and 1 mM sodium pyruvate. Transient transfection of expression constructs was performed using Lipofectamine (Invitrogen) reagent as described previously (13). Cells were washed 6 h posttransfection and fed with fresh medium for 2 to 3 days for further analysis.

Construction of expression vectors.

All expression vectors were constructed on pcDNA3.1(+)/myc-His vector (Invitrogen) unless otherwise specified. EMR2-mFc, EMR2-WT-myc and EMR2-S518A-myc expression constructs have been described previously (13, 26). The retroviral expression constructs were generated using the pFB-Neo vector system (Stratagene, La Jolla, CA). The HA-EMR4-TM1 construct was made by cloning the ECD and the first TM region of EMR4 into a previously described pSecTaq2A/HA via the HindIII and BamHI sites. Constructs encoding the EMR2-PAR1 chimeric proteins were made by ligating the respective EMR2-ECD cDNA fragments with the full-length PAR1 cDNA using appropriate restriction sites. The EMR2-CD4 chimeric expression constructs were made by cloning the full-length CD4-HA-AATN fragment immediately downstream of the EMR2 stalk via the BamHI and ApaI sites.

Retroviral infection and selection of stable cell lines.

HEK293T packaging cells in 10-cm dishes were transfected with 3 μg each of the pFB-Neo expression construct, pVPack-VSV-G, and pVPack-GP vectors (Stratagene) with 25 μl of Lipofectamine in Opti-MEM as recommended by the supplier. Virus-containing supernatant was harvested 48 h posttransfection, to which a final concentration of 5 μg of Polybrene solution/ml was added. HT1080 cells (∼40 to 50% confluence) in six-well plates were infected with 1 ml of viral supernatant. Cells were spun for 90 min at 25°C at 600 × g, followed by the addition of 2 ml of fresh complete medium 3 h later. The infected cells were then incubated for an additional 24 h at 37°C before selection in medium containing 1 mg of G418/ml. G418-resistant cells were collected after ∼2 weeks of selection and confirmed by appropriate analysis.

Preparation of polyclonal antibodies again EMR2 β-subunit.

The peptide antigen (Ag), containing 12 amino acid residues (AKADTSKPSTVN) at the C-terminal end of the EMR2 β-subunit, was synthesized and conjugated to the keyhole limpet hemocyanin carrier protein. The Ag was mixed with complete adjuvant and injected subcutaneously into rabbits, followed by injection of the Ag-incomplete adjuvant mixture at the same site 1 month later. The immunized serum was collected periodically to test for its reactivity to the Ag by enzyme-linked immunosorbent assay (ELISA). The polyclonal antiserum was purified first by ammonium sulfate precipitation, followed by use of a Melon Gel IgG spin purification kit (Thermo Scientific, Fremont, CA) according to the standard procedures.

Cell chemotaxis assays.

Cell chemotaxis were carried out with a QCM Chemotaxis 96-well cell migration assay kit (Millipore, Bedford, MA). Briefly, the cells were serum starved for ca. 18 to 24 h, detached in phosphate-buffered saline (PBS)–2 mM EDTA, and resuspended at a concentration of 5 × 10⁵ cells/ml in serum-free medium–5% bovine serum albumin (BSA). Cells (100 μl/well) were placed in the upper chamber of a QCM chemotaxis plate equipped with an 8-μm-pore-size membrane. The bottom chambers were filled with culture medium with or without 10% FCS as chemoattractants. The plates were kept in a 37°C incubator for 5 h. Migrated cells were dissociated from the bottom of the filter membrane and stained with CyQuant GR dye. The fluorescence intensity was measured in a fluorescence plate reader (Molecular Devices) with a 480/520-nm filter set.

PFO treatment of EMR2-mFc-Dynabead complex and EMR2-expressing cells.

Protein A-conjugated Dynabeads (Dynal A.S., Oslo, Norway) were blocked with 1% BSA in Hanks balanced salt solution for 1 h and then mixed with EMR2-mFc containing conditioned medium for 2 h at 4°C. After extensive washes, beads were precipitated by magnet separation. The EMR2-mFc-dynabead complexes or cells expressing recombinant EMR2 proteins were incubated for 10 min on ice in perfluoro-octanoic acid (PFO)-containing PBS buffer (PFO solution). The supernatant was collected following centrifugation at 13,000 rpm at 4°C for 15 min, and the pellet was dissolved in radioimmunoprecipitation assay lysis buffer to obtain total cell lysate.

Flow cytometry analysis.

Transiently transfected cells were harvested and fixed in 2% paraformaldehyde-PBS at 4°C for 30 min. The cells were then blocked for 1 h in ice-cold blocking buffer (1% BSA, 5% serum of secondary antibody, and PBS) with or without 0.1% saponin. Then, the cells were incubated with the indicated primary antibody diluted in blocking buffer for 1 h before washing. The cells were incubated for 1 h with fluorochrome-conjugated second antibody in blocking buffer (1:200) and then washed three times by cold PBS and subjected to analysis by using a FACScan flow cytometer (BD Biosciences).

Antibody patching.

Transiently transfected cells cultured on coverslips (BD BioCoat poly-d-lysine; BD Biosciences) were blocked for 1 h in ice-cold blocking buffer (1% BSA, 5% normal serum of secondary antibody, and PBS). The cells were incubated with 2A1 MAb (5 μg/ml) for 30 min, followed by Alexa Fluor 647-conjugated anti-mouse IgG (1:200; Invitrogen) for 30 min on ice, with extensive washes in between by ice-cold blocking buffer. The cells were then transferred to a 22°C incubator for 15 min for patching and immediately fixed in 2% paraformaldehyde-PBS at 4°C before washing. For lipid raft patching, the cells were incubated sequentially with Alexa Fluor 647-conjugated cholera toxin β-subunit (10 μg/ml) and anti-cholera toxin β-subunit polyclonal antibodies (1:200; Invitrogen) for 30 min each on ice with extensive washes in between. The cells were then transferred to a 22°C incubator for 15 min for patching and immediately fixed in 2% paraformaldehyde-PBS at 4°C for 20 min. The fixed cells were then processed for 2A1 staining (5 μg/ml) in cold blocking buffer (PBS, 1% BSA, and 5% normal serum from a secondary antibody-producing host animal) using Alexa Fluor 488-conjugated anti-mouse IgG (1:200; Invitrogen). Coverslips were mounted onto slides with antifade mounting medium (ProLong Gold Antifade reagent with DAPI 4′,6′-diamidino-2-phenylindole [Invitrogen]). Fluorescence images were taken at 1,000-fold magnification using a confocal microscope (Zeiss LSM 510 Meta).

Immunoprecipitation.

Anti-myc antibody-conjugated (100 μl) agarose beads were washed twice with 500 μl of PBS and then incubated with 250 μl of blocking buffer (1% BSA in PBS) for 3 h at 4°C. Total cell lysates (∼150 μg) were precleared by incubating with unconjugated agarose beads, spun, and then mixed gently with the conjugated agarose beads overnight at 4°C with end-over-end rotation. After incubation, agarose beads were pelleted by centrifugation at 1,000 × g for 3 min at 4°C and washed five times each with 250 μl of blocking buffer. Finally, equal volume of 2× sodium dodecyl sulfate loading dye was added for subsequent Western blotting.

Lipid raft floatation.

All lipid raft separation procedures were carried out on ice. Cells (∼5 × 10⁶ cells) were washed three times with ice-cold PBS and then lysed in 150 μl of TNET buffer (25 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM EDTA, 1 mM EGTA, 1% Triton X-100, 10 mM NaF, 1 mM sodium pyrophosphate, 1 mM Na₃VO₄, 1× protease inhibitor cocktail) for 30 min. Total cell lysates were passed through a 30-gauge needle 20 times on ice and then centrifuged at 1,000 × g for 10 min at 4°C to collect the supernatant for protein quantification. Typically, 200 μl of cell lysate (∼500 μg of protein) was mixed with 400 μl of 60% Opti-Prep and then placed at the bottom of a 5-ml polyallomer ultracentrifuge tube (Beckman). The samples were overlaid sequentially with 3,400 μl of 30% and 200 μl of 5% ice-cold Opti-Prep medium diluted in TNET buffer and subjected to ultracentrifugation at 200,000 × g for 16 h at 4°C. After ultracentrifugation, seven equal fractions (∼600 μl/fraction) were collected from the top of the tube for Western blotting.

Thrombin treatment and cell surface protein biotinylation and immunoprecipitation.

Transfected cells were lifted from the plates, washed, and divided equally into two 1.5-ml Eppendorf tubes. The cells were incubated in ∼500 μl of thrombin digestion solution (10 mM HEPES–Opti-MEM) in the presence or absence of thrombin (200 nM, 23.5 U) for 30 min at 37°C before the supernatants were collected. The cells were subjected to fluorescence-activated cell sorting (FACS) analysis or cell lysis for Western blot analysis. For cell surface protein biotinylation, thrombin-treated cells were washed three times with PBS (pH 8.0). The cells were suspended in 820 μl of PBS (pH 8.0) at a concentration of ∼2.5 × 10⁶ cells/ml and mixed with 180 μl of ∼10 mM biotin solution to a final concentration of ∼2 mM. After incubation at room temperature for 30 min by end-over-end rotation, the reaction was stopped by washing the cells in 100 mM glycine solution for three times to quench and remove the excess biotin reagent. Cells were lysed in ice-cold TNET lysis buffer to collect cell lysate for immunoprecipitation using streptavidin-coated agarose beads.

Detection of cytokines by ELISA.

THP-1 cells or primary monocytes isolated from peripheral blood mononuclear cells were resuspended in RPMI complete medium at 1 × 10⁶ or 3 × 10⁴ cells/ml in the absence or presence of lipopolysaccharide (LPS). The cells (1 ml/well) were stimulated with soluble or immobilized 2A1 (precoated on plates) or IgG1 isotype control antibody in six-well plates for 24 h. For the lipid raft depletion experiments, the cells were pretreated with lovastatin (5 or 10 μM) for 12 h or filipin (0.1 μg/ml) for 30 min in serum-free RPMI medium. The cells were then spun and resuspended in serum-free RPMI medium with lovastatin (5 or 10 μM) or filipin (1 ng/ml) and stimulated with immobilized 2A1 or IgG1 isotype control antibody in six-well plates. The supernatant was collected and spun, and the concentration of the cytokines was determined by using a DuoSet ELISA development kit (R&D Systems, Minnesota) according to the manufacturer's protocol.

Statistical analysis.

Quantifications were based on at least three independent experiments. The data are shown as means ± the standard errors of the mean (SEM). Statistical analysis of data was performed by using the Student t test using Prism 5 software. The P values are indicated by asterisks in the figures.

GPS autoproteolysis is critical for EMR2-mediated cell migration.

A role for EMR2 in promoting cell migration was demonstrated previously (52). To investigate the functional involvement of GPS autoproteolysis in EMR2 receptor, we generated and compared the migratory ability of HT1080 cells stably expressing EMR2-WT, -S518A, or -TM1 receptor. The EMR2-S518A molecule represents the cleavage-deficient receptor whose GPS cleavage site was mutated to Ala, whereas EMR2-TM1 contains only the first TM domain (Fig. 1A). All stable cell lines show similar growth characteristics and express comparable levels of EMR2, as examined by flow cytometric and Western blotting (WB) analysis (Fig. 1B and C).

In agreement with the earlier results, cells expressing EMR2-WT are found to migrate much faster than the vector-only control and EMR2-TM1-expressing cells (Fig. 1D). Interestingly, EMR2-S518A stable cells also migrate slower than do EMR2-WT cells. This result is further confirmed using murine B-lymphoblast Baf/3b stable cell lines. Again, EMR2-WT-expressing cells show an enhanced migration toward CXCL12 than do control and EMR2-S518A stable cells (Fig. 1E). These data indicate that, in addition to the 7TM domain, GPS proteolytic modification is also critical for the cellular function of EMR2.

The extracellular α-subunit of EMR2 can independently self-anchor on the cell membrane.

Intrigued by the conflicting findings in the latrophilin 1 receptor and to provide a mechanistic insight into how GPS proteolysis modulates EMR2 function, we examined the ability of EMR2 α-subunit to self-anchor on the plasma membrane. EMR2-expressing cells and precipitated EMR2-mFc fusion proteins were treated with PFO, a weak surfactant that does not destroy protein-protein interaction (Fig. 2A). As shown in Fig. 2B, GPS-cleaved EMR2 α-subunit and mFc fragments remain associated even when treated with 1% PFO. However, the α-subunit of EMR2-WT-myc is dissociated efficiently from cell membrane by 0.2% PFO with no apparent effect on the β-subunit (Fig. 2C). Only at 1% PFO is the cell membrane damaged to solubilize α- and β-subunits simultaneously. Similarly, the control EMR2-S518A-myc is detected in the supernatant only after 1% PFO treatment (Fig. 2C). Finally, the α-subunit of endogenous EMR2 receptor in macrophage-like THP-1 cells is solubilized effectively by 1% PFO, again with no detectable β-subunit (Fig. 2D). These results are consistent with the report by Volynski et al. (48) and further indicate that α- and β-subunits associate tightly in soluble EMR2-mFc, but a proportion of the two subunits are separate on the cell membrane. This conclusion was reinforced by the subsequent antibody cross-linking experiment (Fig. 2E). While the α- and β-subunits of EMR2-WT-CFP in control cells show relatively homogeneous distribution on cell membrane, surface cross-linking of the α-subunit by 2A1 MAb produces distinct patches that do not fully overlap with the β-subunit. In contrast, colocalization of both tags is always found in cleavage-deficient EMR2-S518A-CFP before and after antibody cross-linking (Fig. 2E). These data indicate that EMR2 α-subunit is indeed able to self-anchor on the membrane.

If adhesion-GPCR subunits are independent molecular entities on the plasma membrane, one would expect the possibility of cross-interaction between the complementary subunits of different adhesion-GPCR molecules. Indeed, coimmunoprecipitation (co-IP) of complementary subunits has been demonstrated previously between latrophilin 1 and EMR2, as well as GPR56 (40). Additional co-IP experiments using EMR2-WT-myc and HA-EMR4-TM1 show that EMR4 α-subunit is readily co-IP with the EMR2 β-subunit from cells cotransfected with EMR2-WT-myc and HA-EMR4-TM1 constructs, but not those transfected only with the HA-EMR4-TM1 construct (Fig. 2F). Potential false-positive effects caused by β-subunit oligomerization are ruled out by using mixture of cell lysate from cells transfected singly with EMR2-WT-myc or HA-EMR4-TM1 (Fig. 2F). These data verify that the independent α-subunit can cross-interact with the complementary β-subunit of a different adhesion-GPCR.

GPS autoproteolysis is not required for the self-anchoring of EMR2 α-subunit on the membrane.

The GPS autoproteolytic reaction is highly similar to the one identified in Hh proteins (25, 36). Interestingly, the final hydrolytic step of Hh biosynthesis is known to be mediated by cholesterol to form a cholesterol adduct, Hh-Np. This modification allows the otherwise soluble Hh-Np to anchor on the plasma membrane, creating a concentration gradient important for the morphogenic function of Hh (25, 36).

To investigate whether a similar modification occurs during EMR2 maturation, we examined whether GPS autoproteolysis is absolutely required for the membrane association of EMR2 α-subunit. To this end, chimeric EMR2 proteins where the 7TM region of EMR2 is replaced by a protease-activated receptor 1 (PAR1) sequence were generated and analyzed (Fig. 3A). PAR1 is a classical GPCR that can be specifically cleaved and activated by its protease ligand, thrombin (9). We reason that if GPS autoproteolysis is necessary for the membrane association of EMR2 α-subunit, the ECD of GPS cleavage-deficient EMR2-S518A-PAR1 would not associate with the membrane and thrombin treatment should shed it completely from the surface.

Using specific MAbs against the ECD and 7TM region, we show that GPS proteolysis is efficient in EMR2-WT-PAR1 as in WT EMR2 (Fig. 3B). Next, we show the majority of EMR2-S518A-PAR1 is an ∼130-kDa uncleaved protein as expected (Fig. 3B, black arrow), even though a background cleavage event is detected (Fig. 3B, black arrowhead). After thrombin digestion, the ∼130-kDa band disappears completely (Fig. 3B, asterisk) and two fragments (∼70 and ∼40 kDa) representing cleaved EMR2-ECD and PAR1 are detected in the supernatant and cell lysate fractions, respectively (Fig. 3B, white arrowhead). In contrast, the same treatment fails to produce additional protein fragments in control EMR2-WT and EMR2-S518A samples (Fig. 3B). These results confirm that thrombin digestion is highly specific and efficient.

Unexpectedly, following thrombin digestion an ∼60-kDa fully cleaved ECD (α-subunit) is detected in total cell lysate of EMR2-S518A-PAR1-expressing cells (Fig. 3B). Likewise, flow cytometric analysis of the same cells indicates that ∼50% of cells stains positive for surface EMR2 after thrombin treatment (Fig. 3C). Again, surface EMR2 levels remain unchanged in cells expressing control EMR2 proteins under the same treatment (Fig. 3C and data not shown). The fact that a substantial fraction of α-subunit remains on the membrane after the complete digestion of EMR2-S518A-PAR1 by thrombin indicates that the membrane association of EMR2 α-subunit does not require GPS proteolysis. To further support this conclusion, EMR2-S518A-PAR1-transfected cells are fully digested by thrombin, surface biotinylated, followed by cell lysis and pulldown (PD) by streptavidin-agarose. WB analysis again shows the presence of a fully cleaved and membrane-bound α-subunit (Fig. 3D).

On the other hand, analysis of EMR2-WT-PAR1 protein shows that a proportion of EMR2 α-subunit seems to interact tightly with the 7TM β-subunit on cell membrane. As such, increased amount of α-subunit was detected in the supernatant of thrombin-treated EMR2-WT-PAR1-transfected cells compared to that of untreated cells (Fig. 3B, lane 3, white arrowhead). Likewise, an ∼66% reduction of surface α-subunit was found by flow cytometric analysis in the same cells after thrombin digestion (Fig. 3C and data not shown). Based on these data, we suggest that GPS autoproteolysis leads to the formation of two distinct EMR2 receptor complexes: one is a noncovalently linked α-β heterodimer, and the other consists of independent α- and β-subunits.

EMR2 subunits are differentially distributed on the membrane microdomains.

To investigate further the interaction characteristics of the two EMR2 receptor subunits with cell membrane, we decided to analyze their distribution patterns in membrane microdomains. For this purpose, EMR2-CD4 chimeric proteins are engineered in which the entire EMR2-ECD is fused with the full-length CD4 molecule (Fig. 4A). CD4 was used since it is a well-known lipid raft-associated protein that can be partially shed from the membrane (15, 35, 54). As shown in Fig. 4B, EMR2-WT-CD4 is efficiently processed to produce a cleaved α-subunit and a CD4 receptor, proving its suitability for the analysis of α-subunit distribution in membrane microdomains. WB analysis reveals that partial CD4 ectodomain shedding is apparent not only in the wild-type CD4 molecule but also in EMR2-WT-CD4 and EMR2-S518A-CD4 (Fig. 4B). Coshedding of EMR2-ECD with CD4 is detected in both chimeric receptors as expected, but is more obvious in cleavage-deficient EMR2-S518A-CD4 than in EMR2-WT-CD4. This suggests that GPS cleavage-competent EMR2-WT-CD4 probably produces more membrane-associated α-subunit.

Lipid raft flotation analysis confirms the localization of CD4 molecule in the raft as well as nonraft fractions as reported previously (15, 35). Surprisingly, unlike CD4, the α-subunit of EMR2-WT-CD4 is only detected in the nonraft fractions, whereas the noncleavable EMR2-S518A-CD4 is distributed in the raft and nonraft fractions similar to CD4 (Fig. 4C). The differential localization of EMR2 α-subunit and CD4 on membrane microdomains is confirmed alternatively by antibody patching of lipid rafts. Confocal microscopic analysis reveals that while EMR2-S518A-CD4 displays a strong raft-associated pattern, the α-subunit of EMR2-WT-CD4 shows only minimal colocalization with lipid rafts (Fig. 4D).

Since an increasing number of GPCRs are known to reside in the raft microdomains (7, 32), the membrane distribution of EMR2-WT-myc and EMR2-S518A-myc was examined. Surprisingly, while the α-subunit is again detected only in the nonraft fractions, EMR2 β-subunit is identified in both the raft and nonraft fractions (Fig. 5A and B). As expected, the distribution of EMR2-S518A-myc is the same as the β-subunit of EMR2-WT-myc. Most importantly, similar differential distribution patterns are identified for endogenous EMR2 α- and β-subunits in THP-1 cells (Fig. 5C) and primary monocytes (data not shown). Altogether, we conclude that the separate EMR2 α- and β-subunits are distributed differentially on membrane raft microdomains, while the noncovalently associated α-β heterodimer is mostly located at the nonraft regions.

Ligation of EMR2 α-subunit induces translocation to lipid rafts, colocalization with the β-subunit and subsequent generation of inflammatory cytokines.

We have previously shown that ligation of EMR2 by the α-subunit specific 2A1 MAb strongly potentiates neutrophil responses to inflammatory stimuli, suggesting that 2A1 is likely an EMR2 receptor agonist (5, 52). Is 2A1-induced EMR2 function caused by the interplay between the two separate subunits? We first investigated whether the ligation of α-subunit by 2A1 could modulate its distribution pattern on membrane microdomains. Interestingly, while the α-subunit of control cells stays in the nonraft fractions, the α-subunit of 2A1-treated cells shows a transient migration into the lipid raft fractions (data not shown). 2A1-induced α-subunit translocation to lipid rafts peaks at 2 to 5 min after 2A1 stimulation and gradually returns to its normal distribution pattern. In contrast, the distribution of EMR2 β-subunit shows no obvious changes under the same condition. The same results were obtained in different cell lines, including CHO-K1, HT-1080, and Baf/3b cells (data not shown), suggesting a genuine 2A1-induced effect.

We next investigated the functional consequence of 2A1-induced α-subunit translocation to lipid rafts. For this purpose, THP-1 cells and human primary monocytes were treated with or without 2A1 MAb, followed by the detection of proinflammatory cytokines. Although the myeloid cells responded positively to inflammatory stimuli such as LPS, no cytokine production in 2A1-treated cells was observed (data not shown). Nevertheless, a substantial response was noted in cells incubated with immobilized 2A1 MAb precoated on plates. Under these conditions, immobilized 2A1-induced lipid raft translocation of α-subunit is found to persist for as long as 3 h (Fig. 6A). The same results were obtained in primary monocytes (data not shown). Moreover, the production of proinflammatory cytokines including interleukin-8 (IL-8) and tumor necrosis factor alpha (TNF-α) was readily detected (Fig. 6B and C). Importantly, the effect of immobilized 2A1 on cytokine production is found to be specific in a dose- and time-dependent manner (Fig. 6B, C, and D). As such, no cytokine production is detected when plates are coated with an isotype-control IgG. Furthermore, IL-6 (Fig. 6B and C) and IL-10 (data not shown) are not produced by 2A1-treated cells, confirming the restricted specificity of immobilized 2A1-induced effect. The possibility of contamination in the system is ruled out since the incubation of soluble 2A1 with THP-1 cells did not induce any cytokine production. These results demonstrate for the first time that sustained ligation of EMR2 α-subunit in macrophages induces the production of specific inflammatory cytokines. Similarly, EMR2 ligation by membrane-coated 2A1 also enhances cell migration of stable HT1080 cells expressing EMR2-WT receptors (Fig. 6E).

Finally, we examined whether the EMR2 β-subunit in lipid rafts is relevant for the 2A1-induced cytokine production. To do this, lipid rafts are disrupted by treating cells with lovastatin or filipin, two commonly used reagents to manipulate the cholesterol levels in cell membranes (41). MβCD, a cholesterol-depleting raft disruptor, causes a high background of cytokine secretion in our macrophage culture system and so was not used further (41). WB analysis revealed effective lipid raft disturbance in lovastatin- and filipin-treated cells, as demonstrated by the apparent changes in the distribution patterns of raft markers such as CD55 (data not shown). Similarly, the localization pattern of EMR2 β-subunit was also shifted more toward the nonraft fractions in treated cells (Fig. 7A and B). Interestingly, when lovastatin- or filipin-treated cells are incubated with immobilized 2A1, the 2A1-induced α-subunit migration to lipid rafts is clearly retarded (Fig. 7A and B). More significantly, the production of IL-8 and TNF-α is reduced dose dependently in lovastatin- and filipin-treated THP-1 cells (Fig. 7C and D) and primary monocytes (data not shown) upon stimulation by immobilized 2A1. These results indicate that the independent EMR2 β-subunit in lipid rafts is critical for the 2A1-induced translocation of α-subunit to the raft fractions and is most likely responsible for the subsequent production of proinflammatory cytokines.