Migration to Apoptotic “Find-me” Signals Is Mediated via the Phagocyte Receptor G2A*

Phagocytosis of apoptotic cells is fundamentally important throughout life, because non-cleared cells become secondarily necrotic and release intracellular contents, thus instigating inflammatory and autoimmune responses. Secreted “find-me” and exposed “eat-me” signals displayed by the dying cell in concert with the phagocyte receptors comprise the phagocytic synapse of apoptotic cell clearance. In this scenario, lysophospholipids (lysoPLs) are assumed to act as find-me signals for the attraction of phagocytes. However, both the identity of the lyso-PLs released from apoptotic cells and the nature of the phagocyte receptor are largely unknown. By a detailed analysis of the structural requirements we show here that lysophosphatidylcholine (lysoPC), but none of the lysoPC metabolites or other lysoPLs, represents the essential apoptotic attraction signal able to trigger a phagocyte chemotactic response. Furthermore, using RNA interference and expression studies, we demonstrate that the G-protein-coupled receptor G2A, unlike its relative GPR4, is involved in the chemotaxis of monocytic cells. Thus, our study identifies lysoPC and G2A as the crucial receptor/ligand system for the attraction of phagocytes to apoptotic cells and the prevention of autoimmunity.

The elimination of apoptotic cells is a critical end point of apoptosis and of fundamental importance for multicellular organisms. Non-cleared apoptotic cells otherwise become secondarily necrotic and release intracellular contents into the surroundings, thus instigating inflammatory reactions. For the timely and efficient removal of apoptotic cells, a network of interactions between the dying cell and the phagocyte has evolved that are displayed at the phagocytic synapse. Secreted find-me and eat-me signals exposed by the dying cell, together with bridging proteins and phagocyte receptors, comprise the central elements for removal of apoptotic cells and prevention of secondary necrosis (1). Growing evidence suggests that the defective clearance of apoptotic cells favors the onset of autoimmune diseases (2).
Lysophospholipids (lysoPLs), 4 like lysophosphatidylcholine (lysoPC), have been shown to function as eat-me signals recruiting complement proteins for recognition by the phagocyte (3). In addition, we could demonstrate that lysoPC also acts as a chemotactic find-me signal attracting the phagocyte to the apoptotic cell (4). During apoptosis lysoPC is generated by the calcium-independent phospholipase-A 2 that is activated by a caspase-3-dependent cleavage mechanism. Once released, lysoPC is assumed to bind to a phagocyte receptor, thereby triggering chemotaxis.
The nature of the phagocyte receptor involved in the attraction to apoptotic cells is currently unknown. Receptors involved in phospholipid signaling often exhibit significant promiscuity with many receptors recognizing more than one ligand and vice versa. For lysoPC different receptors have been proposed, including the G-protein-coupled receptor G2A (5) and its structural relative GPR4 (6). However, due to the failure to reproduce the original receptor binding data, the role of G2A and GPR4 as high affinity lysoPC receptors remains controversial (6,7). Their biological role is further complicated by the fact that both G24 and GPR4 have been suggested to bind other ligands such as oxidized fatty acids (oxFAs) as well as extracellular protons, which implicated a pH-sensing function of both receptors (8,9). As the receptor for phagocyte recruitment remains unknown, it is also unclear whether lysoPC represents the sole lipid find-me signal or whether additional phospholipids secreted during apoptosis are involved in chemotaxis. Finally, it is unknown whether lysoPC mediates its chemotactic into the lower chamber of 8-m pore ChemoTX plates (Neuroprobe Inc., Gaithersburg, MD). The filter was adjusted, the stained cell suspension was added on top, and the assay was incubated for 120 -180 min at 37°C. Subsequently, the transmigrated cells were collected by centrifugation and lysed in 100 l of lysis buffer (25 mM Tris-phosphate, pH 7.8, 2 mM EDTA, 10% glycerol, 1% Triton X-100). Green fluorescence was analyzed, and transmigration was calculated in percentage of total cells deployed (mean values Ϯ S.D. from triplicate experiments).
Measurement of Cell Viability-Cell viability tests were performed in triplicate as described previously (4).
Transfection with siRNA Oligonucleotides-Transfection with siRNA oligonucleotides was carried out twice (at day 0 and day 3) with the Gene Pulser II ϩ Capacity Extender II (Bio-Rad) and 0.4-cm gap cuvettes. 5 ϫ 10 6 THP-1 cells were electroporated with 2 M siRNA oligonucleotides in 500 l of Opti-MEM TM medium (Invitrogen) by a single pulse (800 F, 200 V, time constant 20 -30 ms). The cells were cultured for 3 days before electroporation was repeated. All following experiments were carried out at day 6.
Preparation of Membrane Protein Extracts and Immunoblotting-Expression of G2A was detected by immunoblotting of membrane protein extracts. Therefore, 1 ϫ 10 7 cells were collected by centrifugation, washed in cold phosphate-buffered saline, and resuspended in 1 ml of buffer A containing 20 mM HEPES, pH 7.5, 1.5 mM MgCl 2 , 10 mM KCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, and 10 g/ml aprotinin. The cells were homogenized in a Dounce homogenizer, and the homogenates were centrifuged at 1,000 ϫ g for 10 min at 4°C. The supernatant was centrifuged at 50,000 ϫ g for 2 h. The resulting membrane pellet was solubilized in 5% Triton X-100, 50 mM Tris-HCl, pH 7.6, and 150 mM NaCl containing 3 g/ml aprotinin, 3 g/ml leupeptin, 3 g/ml pepstatin A, and 2 mM phenylmethylsulfonyl fluoride. Subsequently, 100 -200 g of membrane protein extracts were applied to reducing 8 -15% SDS-PAGE and Western blot analysis as previously described (4).
qRT-PCR Analysis-The detection of G2A and GPR4 mRNA levels was performed by quantitative reverse transcription (qRT)-PCR analysis with an ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA) and qPCR Mastermix Plus (Eurogentec, Seraing, Belgium). The following probes and primers (MWG Biotech, Ebersberg, Germany) were used: G2A forward, 5Ј-TGT GCC CAA TGC TAC TGA AAA  AC-3Ј; G2A reverse, 5Ј-CGA CCA GGA CTA TCC TGC TCT  CT-3Ј; G2A probe FAM, 5Ј-TTA CAA TGG AAA CGC CAC  CCC AGT GAC-3Ј-TAMRA;  Total RNA from 2 ϫ 10 6 cells was extracted with the Nucleo-Spin RNA II kit (Macherey & Nagel, Dueren, Germany). 1 g of total RNA was reverse transcribed with 200 units of Superscript RT II TM reverse transcriptase (Invitrogen) in the presence 50 M random hexamers (Amersham Biosciences), 400 M dNTPs (Promega), and 1.6 units/l RNAsIn TM (Invitrogen). 40 -80 ng of the resulting cDNA were applied to the following qRT-PCR analyses (20-l final volume) with 200 nM primers and 100 nM probe. For the study of siRNA knockdown effects, relative quantification was performed employing the standard curve method. The results were normalized on ␦-aminolevulinate-synthase-1, and the cell population transfected with the control oligonucleotide was used as calibrator. The comparison of G2A and GPR4 expression in THP-1 cells was done with the (1ϩE) ϪC T -method (efficiency of amplification ϭ 0.96 (G2A) or 0.92 (GPR4)). All experiments were performed in duplicates and are presented as mean values.

RESULTS
We have previously found that apoptotic cells secrete find-me signals ensuring the recruitment of macrophages and their final engulfment (4). During this process, caspase-3-mediated activation of calcium-independent phospholipase-A 2 was shown to result in the release of the phospholipid lysoPC that is involved in the chemotactic response. Our previous study, however, could not exclude that lysoPC derivatives or additional lipid signals contribute to the attraction of macrophages. Furthermore, the identity of the corresponding phagocyte receptor remained elusive. To address these objectives, we therefore first performed a detailed characterization of the lysoPL agonist(s), which also allowed us to narrow down the nature of the involved lipid receptor.
Headgroup Specificity-The substrate and headgroup specificity of calcium-independent phospholipase-A 2 is relatively unknown, and first reports demonstrated little preference for PC (12). Thus, it was conceivable that apart from lysoPC other lysoPLs might be generated and involved in phagocyte attraction. To explore this possibility, we first investigated whether lysoPLs with different headgroups could compete with the chemoattractant signal in apoptotic culture supernatants (ACS). To neutralize the chemotactic gradient, we added lysoPC, lysophosphatidylserine (lysoPS), and lysophosphatidylethanolamine (lysoPE) to THP-1 monocytes in the upper chamber of a transmigration plate and applied ACS of MCF-7 casp3 cells to the lower chamber (diagram in Fig. 1A). Subsequent measurement of cell migration revealed that only lysoPC could efficiently block chemoattraction of THP-1 cells by ACS, whereas lysoPS and lysoPE had little effect (Fig. 1A). To confirm that this was not due to a potential cytotoxicity of the phospholipids, we performed a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium-bromide assay showing that the viability of THP-1 cells was not measurably altered by any lysoPL (Fig. 1B).
Next, we performed the reverse assay and tested whether the lysoPLs could directly induce THP-1 cell migration. When the lysoPLs were added to the lower transmigration chamber (diagram Fig. 1C), only lysoPC was able to attract THP-1 cells. From these data we can conclude that (i) a phosphocholine headgroup in the lysoPL released by apoptotic cells is required to stimulate cell migration and that (ii) phosphoserine or phos- phoethanolamine headgroups only weakly antagonize the apoptotic find-me signal.
Role of the Lipid Backbone-It had been shown that, in addition to PC, calcium-independent phospholipase-A 2 can also hydrolyze PAF to lysoPAF (12). We therefore investigated the structural requirement of the lipid backbone. Because of the similarity to lysoPC, we included PAF and lysosphingomyelin (lysoSM) in our experiments. With the same experimental design as in Fig. 1A, lysoPC, PAF, lysoPAF, and lysoSM were tested for their potential to compete with the lysoPL attraction signal in ACS. As can be seen in Fig. 2A, all lipids neutralized the apoptotic chemoattractant. Again, examination of THP-1 cell viability did not reveal any profound cytotoxic effect (Fig. 2B).
Because of their antagonistic effects, we next studied whether PAF, lysoPAF, and lysoSM also induce migration directly. Remarkably, when applied to the lower transmigration chamber only lysoPC, and to a lesser extent lysoSM, could stimulate THP-1 cell chemotaxis (Fig. 2C). From these results we presume that migration can be induced by glycerol or sphingosine lysoPLs, whereas antagonizing effects on lysoPL-induced migration can also be exerted by sn-1 etherlipids with a free sn-2 OH-or small acyl residue (see Fig. 2D).
Role of LysoPC Metabolites-Although lysoPC induced chemotaxis and neutralized the chemotactic activity in ACS, due to its inherent instability, we could not exclude that a lysoPC metabolite acts as the actual find-me signal. For lysoPC, different ways of metabolization are known (Fig. 3E). First, lysoPC can be reesterified to PC. Second, it can be hydrolyzed to lysophosphatidic acid (lysoPA) and a free choline residue by phospholipase D or to glycerophosphocholine and a free fatty acid Apoptotic "Find-me" Signals and G2A FEBRUARY 29, 2008 • VOLUME 283 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5299 (FA) by lysophospholipase A1. Therefore, we tested the metabolites PC, lysoPA, and glycerophosphocholine in the competition assay with ACS but could not detect any neutralizing potential (Fig. 3A) or cytotoxic effect (Fig. 3B). Next, we examined the role of the free FA. We deliberately focused on oxFAs because FA oxidation occurs during apoptosis (13) and chemo-  Fig. 1A). B, viability of THP-1 cells treated with different lysoPC derivatives (measured as in Fig. 1B). C, competition of different oxFAs with the chemoattractant in ACS (measured as in Fig. 1A). D, viability of THP-1 cells treated with different oxFAs (measured as in Fig. 1B). Error bars represent S.D. of triplicates. E, schematic diagram of putative metabolization steps for lysoPC. taxis to oxFAs in contrast to non-oxFAs has been reported (14).
Characterization of the Phagocyte Receptor Mediating the Recruitment to Apoptotic Cells-Our established agonist/antagonist profile of the chemotactic activity hinted at the possibility that the G-protein-coupled receptors G2A or GPR4 might be involved in the recruitment of phagocytes to ACS. Both receptors have originally been reported to bind lysoPC and, with a weaker affinity, lysoSM (5, 6); however, their functional role in mediating lysoPC biological effects has been obscured because of nonvalid receptor binding data (5-7). Furthermore, whether these receptors are involved in the attraction of monocytes to apoptotic cells was completely unknown.
To investigate a functional role of G2A and GPR4, we compared their expression levels in THP-1 cells by qRT-PCR. Fig.  4A shows that THP-1 cells expressed Ͼ5-fold more G2A than GPR4 mRNA. Furthermore, we found that, in contrast to THP-1 cells, the monocytic cell line U937 expressed considerably less G2A transcripts (Fig. 4B). The difference of G2A expression was even more evident on the protein level, as demonstrated by Western blot analysis that revealed no anti-G2A reactive band in U937 cells but a prominent expression in THP-1 cells (Fig. 4C).
In view of this different expression pattern, we compared the migration activity of THP-1 and "low level G2A" expressing U937 cells toward ACS and lysoPC. Interestingly, as depicted in Fig. 4D, both stimuli induced chemotaxis only in THP-1 cells but not in U937 cells. To verify that U937 cells had no general migration defect, supernatants of necrotic MCF-7 casp3 cells and recombinant stromal cell-derived factor-1␣ (SDF-1␣) were employed. Transmigration assays confirmed that these stimuli could induce chemotaxis also in U937 cells (Fig. 4D). In conclusion, only THP-1 cells expressing G2A show migration to ACS and lysoPC.
Knock Down of G2A, but Not GPR4 Expression, Results in Decreased Migration to ACS and LysoPC-To further investigate the functional role of the receptor in phagocyte chemotaxis to ACS and lysoPC, we employed an RNA interference approach in THP-1 cells. The knock down of G2A expression with two different siRNAs resulted in a strong reduction of migration toward ACS and purified lysoPC in comparison to control oligonucleotide-transfected cells (Fig. 5, A and B). As anticipated, chemotaxis to supernatants of necrotic MCF-7 casp3 cells and SDF-1␣ was not affected in G2A-silenced cells. This implicates that G2A plays a role in chemotaxis stimulated by lysoPLs released by apoptotic cells and purified lysoPC. The observation that migration was not fully abolished in G2A knockdown cells might be due to the incomplete suppression of G2A expression, because qRT-PCR revealed ϳ20% of residual G2A mRNA (Fig. 5C). Also, in Western blot analysis G2A protein was markedly reduced but still detectable (Fig. 5D). Additionally, other lysoPC receptors such as GPR4 might be responsible for the residual migration observed in G2A-silenced THP-1 cells. However, silencing of GPR4 expression with two different GPR4-specific siRNA oligonucleotides had no significant impact on the transmigration to ACS (Fig. 5E), even though the knock down was as efficient as in the case of G2A with ϳ20% of residual GPR4 mRNA (Fig. 5F). Thus, expression of G2A, but not GPR4, appears to be required for migration to apoptotic lysoPL attraction signals.
Expression of G2A Reconstitutes Migration to ACS and LysoPC-To corroborate these findings and to investigate whether G2A is sufficient for migration to lysoPL find-me signals, we employed retroviral transduction experiments with G2A in U937 cells that are unresponsive to ACS. Comparison of the low level G2A U937 cells transduced with the retroviral vector alone and U937 cells transduced with the G2A construct revealed that expression of G2A strongly restored the migration activity toward ACS and lysoPC (Fig. 6A). Again, expression of G2A did not affect the migration to purified SDF-1␣.
In addition, we compared the two U937 cell populations for migration toward different lysoPLs. Regardless of G2A expression, however, neither lysoPS nor lysoPE was able to trigger cell migration (Fig. 6B). Interestingly, expression of G2A reconstituted the migration of U937 cells not only to lysoPC but also to lysoSM (Fig. 6B). This observation is consistent with our previous finding that indeed both phospholipids might act as putative phagocyte receptor agonists (Fig. 2C).
To confirm our above established agonist/antagonist profile (Fig. 2D), we incubated G2A-transduced U937 cells with the different lysoPLs in a neutralization experiment. To this end, lysoPC was added to the lower and the other lysoPLs were added with the cells to the upper transmigration chamber. As shown in Fig. 6C, only lysoPC itself, PAF, lysoPAF, and lysoSM could neutralize lysoPC-stimulated migration of U937 G2A cells, supporting the data obtained with THP-1 cells and ACS (Figs. 1 and 2).
Finally, we analyzed migration of mouse J774A.1 macrophages with different G2A expression levels. Again, RNA interference and retroviral overexpression were used to inhibit and enhance G2A expression, respectively. As shown in supplemental Fig. S1, lysoPC could only induce chemotaxis in G2A high expressing J774A.1 macrophages, whereas the G2A-silenced cells showed no migration. Furthermore, the other lyso-PLs exerted only weak effects in G2A-expressing cells (supplemental Fig. S1). Taken together, these results suggest that in both human and murine cells lysoPC and lysoSM are the only lysoPLs inducing monocyte migration via G2A, an event that can be antagonized by PAF and lysoPAF.

DISCUSSION
The attraction of the phagocyte appears to be the initial step in the clearance of apoptotic cells. The players involved in this process, however, have been ill-defined. We showed previously that during apoptosis caspase-3-mediated cleavage leads to the activation of calcium-independent phospholipase-A 2 , thereby generating a lipid attraction signal for the recruitment of Apoptotic "Find-me" Signals and G2A FEBRUARY 29, 2008 • VOLUME 283 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 5301 phagocytes (4). Whether lysoPC was the sole attraction signal or whether other lysoPLs or metabolites were involved in this process remained unknown. Neither was the corresponding receptor on phagocyte site identified. In this study, we show that among the putative lysoPLs generated by calcium-independent phospholipase-A 2 during apoptosis only lysoPC rep-resents a major find-me signal that stimulates migration via the phagocyte receptor G2A both in human and murine cells.
G2A forms together with T cell death-associated gene 8 (TDAG8), ovarian cancer G-protein-coupled receptor 1 (OGR1), and GPR4 a subgroup of structurally related G-protein-coupled receptors that had been originally proposed to  E) ϪCT method, and G2A mRNA level was set as calibrator. The inset shows the primary amplification curves. B, relative quantitation of G2A mRNA levels in THP-1 and U937 cells. Total RNA of 2 ϫ 10 6 THP-1 or U937 cells was extracted and reverse transcribed, and 80 ng of the resulting cDNA were used for qRT-PCR as described under "Experimental Procedures." Relative G2A mRNA level in THP-1 cells was used as calibrator. C, detection of G2A on protein level by immunoblot analysis of THP-1 and U937 membrane protein extracts. Membrane proteins of 1 ϫ 10 7 THP-1 or U937 cells were extracted as described under "Experimental Procedures," and 130 g of the extract were applied to 8 -15% SDS-PAGE with consecutive anti-G2A immunoblot analysis. Detection of sodium-protonexchanger-1 (NHE-1) protein was used as loading control. Asterisks indicate unspecific bands of the anti-G2A antibody. D, migration of THP-1 and U937 cells to supernatants of healthy, apoptotic, or necrotic cells and lysoPC or SDF-1 ␣. Supernatants of MCF-7 casp3 cells and lysoPC or SDF-1 ␣ were applied to the lower transmigration chamber, and migration of THP-1 or U937 cells was assessed as described under "Experimental Procedures." Error bars represent S.D. of triplicates.
bind proinflammatory lipids (5,6,15,16). More recent studies have challenged the identification of lipid agonists for these receptors and have suggested that they function primarily as proton sensors (8,15,17). However, in contrast to OGR1, GPR4, and TDAG8, G2A lacks essential histidine residues that were previously shown to be important for pH sensing FIGURE 5. Knock down of G2A but not GPR4 expression results in decreased migration to ACS and lysoPC. A, influence of G2A knock down on migration to supernatants of healthy, apoptotic, and necrotic cells. Supernatants of MCF-7 casp3 cells were prepared and applied to the lower transmigration chamber. Migration of THP-1 cells electroporated with a control oligonucleotide or two different G2A-specific siRNA oligonucleotides was assessed as described under "Experimental Procedures." B, effect of G2A knock down on migration stimulated by lysoPC or SDF-1 ␣. 10 M lysoPC or 200 ng/ml SDF-1 were added to the lower chamber of a double chamber plate, and migration of siRNA-treated THP-1 cells (as in Fig. 5A) was assessed. C, knock down of G2A mRNA expression by siRNA oligonucleotides. G2A mRNA level was measured in siRNA-treated THP-1 cells as depicted under "Experimental Procedures." The G2A mRNA level in THP-1 cells treated with the control siRNA was set as 100% calibrator. D, knock down of G2A protein expression by siRNA oligonucleotides. G2A was detected by immunoblot analysis of siRNA-treated THP-1 membrane protein extracts. Membrane proteins of 1 ϫ 10 7 THP-1 cells were extracted as described under "Experimental Procedures," and 100 g of membrane protein were applied to 8 -15% SDS-PAGE with consecutive anti-G2A immunoblot analysis. Detection of sodium-proton-exchanger-1 (NHE-1) protein was used as loading control. E, effect of GPR4 knock down on migration to supernatants of healthy and apoptotic cells. Migration of THP-1 cells electroporated with a control oligonucleotide or two different GPR4-specific siRNA oligonucleotides was assessed as in Fig. 5A. F, knock down of GPR4 mRNA expression by siRNA oligonucleotides. GPR4 mRNA level was measured in siRNA-treated THP-1 cells as in Fig. 5C. Error bars represent S.D. of triplicates.

FIGURE 6. Expression of G2A reconstitutes migration to ACS and lysoPC.
A, migration of U937 G2A or U937 cells to supernatants of healthy and apoptotic cells or 10 M lysoPC or 200 ng/ml SDF-1 ␣, respectively, was measured as described under "Experimental Procedures." B, stimulation of U937 cell migration by different lysoPLs. LysoPC, lysoPS, lysoPE, and lysoSM were added at 5 M to the lower transmigration chamber, and chemotaxis assay with U937 G2A or U937 cells was performed as depicted under "Experimental Procedures." C, neutralization of lysoPC induced U937 G2A cell migration by different phospholipids. LysoPC was added at 5 M to the lower transmigration chamber, 20 M of lysoPC, lysoPS, lysoPE, PAF, lysoPAF, or lysoSM were added to the U937 G2A cells in the upper transmigration chamber, and chemotaxis of U937 G2A cells was assessed. Error bars represent S.D. of triplicates. (18). This and the observation that G2A-deficient thymocytes and splenocytes show no impaired response to extracellular protons (19) suggest that G2A exerts a distinct biological function.
Similar to the pH-sensing ability, the phospholipid binding function of the OGR1 family of receptors has been debated. The initial description of the lysoPC binding activity for both G2A and GPR4 has been retracted, mainly because of inconsistencies in these studies. Nevertheless, using RNA interference and retroviral overexpression, our experiments clearly show that G2A mediates the attraction of monocytes in response to lysoPC and ACS. Our data are strengthened by follow-up studies demonstrating a functional role of G2A in lysoPC-mediated chemotaxis of different cell types (10,18).
In addition to lysoPC and lysoSM, G2A has recently been reported to bind oxFAs (9). In this study the authors showed intracellular calcium mobilization, cAMP decline, and JNK activation induced by (9)-HODE and other oxFAs but found no detectable calcium flux by lysoPC in G2A-overexpressing cells (9), yet they failed to detect (9)-HODE-induced ERK phosphorylation, an event described to occur upon lysoPC-mediated G2A stimulation (20). Although the authors did not measure cell migration, in the present study we could not observe cell migration in response to (9)-HODE or other oxFAs (data not shown). Moreover, lysoPL-induced migration was not neutralized by oxFAs (Fig. 3C). From these data it can be concluded that oxFAs do not stimulate G2A-mediated monocyte migration but, rather, distinct signaling events. Ectopic expression studies with different G␣ subunits and partial sensitivity to pertussis toxin led Obinata et al. (9) to the conclusion that (9)-HODE-stimulated G2A signaling is G q -and G i -coupled. This, however, contrasts with other studies demonstrating a pertussis toxin insensitivity of lysoPC-induced and G2A-mediated migration. Nevertheless, it is conceivable that certain G2A agonists trigger different signaling cascades via dynamic G-protein interactions. For instance, oxFAs might induce calcium release and JNK activation via G q or G i interactions, whereas cell migration induced by lysoPC or lysoSM could be mediated by G q/11 /G 12/13 interaction and subsequent Rho and ERK activation.
So far, valid binding data are not available for either lysoPC or for oxFAs. It remains therefore unknown whether the interaction between G2A and its putative ligands is direct or indirect. Lipid mediators such as externalized phosphatidylserine and oxidized phospholipids as well as lysoPC play a role in the recognition of apoptotic cells. In the phagocytic synapse, these lipid signals do not act directly on the corresponding phagocyte receptors but indirectly via certain bridging proteins (1). In this context, complement recruitment has been described for lysoPC (3). Consequently, the interaction between soluble lysoPC and G2A could also be mediated via a bridging protein.
Alternatively, lysoPC-mediated G2A activation might be controlled by a different mechanism. In this regard, Wang et al. (20) showed that intracellular sequestration and surface exposition of G2A control the signaling responses toward lysoPC.
In our experiments employing RNA interference-mediated knock down of G2A expression, we could not completely inhibit monocyte migration to ACS. Although this might be due to an incomplete suppression of G2A, it is also conceivable that other receptors additionally contribute to monocyte migration. We could exclude a role of GPR4, because knock down of GPR4 expression had no effect on migration toward lysoPL attraction signals. This fits to the suggestion of GPR4 acting as a proton sensor rather than a lysoPL receptor (15,19), even although other biological effects of lysoPLs, for instance in endothelial cells, might be still mediated by GPR4 (21,22). Other receptors capable of lysoPC binding include scavenger receptors or the complement receptors 3 and 4, which recognize complement-bound lysoPC (23). Finally, the calreticulin/CD91 system has also been described to bind to complement proteins and collectins (24,25) and could thereby mediate lysoPC-induced migration. However, whether these receptor systems indeed mediate chemotaxis to lysoPC has not been studied so far.
The observation that G2A knock down did not completely abolish migration could also be explained by the existence of additional lysoPC-independent find-me signals in ACS. In view of the great variety of eat-me signals presented by the apoptotic cell, lysoPC might not be the only apoptotic find-me signal. In this context, S19 ribosomal protein, split human tyrosyl-tRNA synthetase, or thrombospondin-1 have been described (26 -28) as chemoattractant signals. Certainly, the lysoPC/G2A receptor system apparently plays a predominant role in the clearance of apoptotic cells. It is now widely accepted that defects in the removal of apoptotic cells can lead to autoimmunity (2). Interestingly, G2A knock-out mice develop an autoimmune syndrome (29) similar to mice with a deficiency in certain eat-me signals, such as MFG-E8 or C1q (30,31). This similar phenotype underscores the importance of both find-me and eat-me signals for the clearance of apoptotic cells.