Sulfation of Tyrosine 174 in the Human C3a Receptor Is Essential for Binding of C3a Anaphylatoxin*

The complement anaphylatoxin C3a and its cellular seven-transmembrane segment receptor, C3aR, are implicated in a variety of pathological inflammatory processes. C3aR is a G-protein-coupled receptor with an exceptionally large second extracellular loop of 172 amino acids. Previously reported deletion studies have shown that at least part of this region plays a critical role in binding C3a. Our data now demonstrate that five tyrosines in the second extracellular loop of the C3aR are posttranslationally modified by the addition of sulfate. Blocking sulfation by mutation of tyrosine to phenylalanine at positions 184, 188, 317, and/or 318 does not affect ligand binding or signal transduction. However, when tyrosine 174 is mutated to phenylalanine, binding of native C3a is completely blocked. This variant efficiently mobilizes calcium in response to synthetic C3a agonist peptides, but not to native C3a. This finding is consistent with a two-site model of ligand association typical of many peptide ligand-receptor interactions and identifies sulfotyrosine 174 as the critical C3a docking site. Tyrosine sulfation in the amino-terminal extracellular domain has been shown to be important in several other seven-transmembrane segment receptors. Our data now demonstrate that tyrosine sulfate in other extracellular domains can function for ligand interactions as well.

The human C3a anaphylatoxin is a 77-amino acid protein generated by proteolysis of C3 during activation of the complement cascade (1,2). C3a is a potent inflammatory mediator, inducing smooth muscle contraction, increased vascular permeability, arachidonic acid metabolism, cytokine release, cellular degranulation, and chemotaxis (3)(4)(5)(6)(7)(8)(9). The C3a receptor, C3aR, 1 is a G-protein-coupled, seven-transmembrane segment (7TMS) protein with an unusually large second extracellular loop (ECL2) between the fourth and fifth transmembrane domains (10 -12). Despite significant sequence homology with the receptor for the C5a anaphylatoxin, C5aR, the C3aR is devoid of the highly acidic and tyrosine-rich amino terminus that participates in its association with C5a (12)(13)(14). The unusual structural feature represented by the large ECL2 in the C3aR suggested that it might be involved in this ligand-receptor interaction. Indeed, several studies with chimeras and point mutants have shown that ECL2 is at least part of the binding site for C3a (12,15,16).
Tyrosine sulfation is a posttranslational modification occurring on a number of secreted proteins, including complement C4, coagulation factor VIII, cholecystokinin, and gastrin (17)(18)(19)(20). This modification tends to occur in acidic regions of proteins, usually those containing multiple tyrosines (21,22). Several chemokine receptors, including CCR5, CXCR4, CX3CR1, and CCR2b, have been shown to be sulfated on tyrosines in their amino-terminal domains, and at least some of this sulfation is critical for ligand binding (23)(24)(25)(26). For example, sulfotyrosine in the amino-terminal sequence of CCR5 is required for binding of the natural ligands MIP-1␣, MIP-1␤, and RANTES (regulated on activation normal T cell expressed and secreted) and certain HIV-1 GP120/CD4 complexes as well as the binding and entry of the CCR5-using strains of HIV-1 (23). Amino-terminal sulfation of tyrosines in the C5a receptor also contributes to formation of the docking (but not the signaling) site for the C5a anaphylatoxin (14). Thus, tyrosine sulfation is a critical modification for conferring the natural function of a number of 7TMS receptors.
To date, all the reported functional sulfotyrosines identified in 7TMS receptors have occurred in the amino-terminal extracellular domains. In contrast to the C5aR, which has an acidic and tyrosine-rich motif at its amino terminus, the C3aR does not (11,12). There are nine predicted tyrosines in the fulllength human C3aR, seven of which are located at positions 174, 184, 188, 255, 306, 317, and 318, all in ECL2, and some of these are flanked by acidic amino acids. The two remaining tyrosines are in the sixth and seventh transmembrane sequences. The multiplicity of tyrosines, coupled with the unusually large ECL2, raised the possibility that tyrosine sulfation may occur in this domain and that this modification may contribute to the binding of C3a.
Here we demonstrate that five of the seven tyrosines in ECL2 of the C3aR are sulfated, but only one, tyrosine 174, is essential for binding and signaling with native C3a. The C3aR variant in which tyrosine 174 was mutated to phenylalanine mobilized calcium in response to 8-and 15-residue synthetic peptides containing the carboxyl terminus of C3a, consistent with a two-site model of ligand association. These data not only define the precise structural requirement for the C3a docking site but also demonstrate that tyrosine sulfation can occur at sites distinct from the amino-terminal regions of 7TMS receptors.

EXPERIMENTAL PROCEDURES
Cells, Plasmids, Antibodies, and Peptides-HEK 293T and Cf2Th cells were obtained form the American Tissue Type Culture Collection (CRL115544 and CRL 1430, respectively). 293 GP packaging cells were purchased from BD Biosciences. Cells were maintained in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, penicillin, and streptomycin. An expression plasmid encoding the human C3aR with the addition of a ten-amino acid tag at its amino terminus (Myc tag) and a nine-amino acid extension at its carboxyl terminus (C9 tag) was generated by PCR amplification of human genomic DNA and subcloned into the pcDNA 3.1 expression vector (Invitrogen). All the C3aR variants in which one or more tyrosines were mutated to phenylalanine (see Table I) were made by the PCR-based QuikChange method (Stratagene) and confirmed by sequencing the entire reading frame. The antibody 1D4, which recognizes the C9 tag, and 9E10, which binds the Myc tag, were provided by the National Cell Culture Center (Minneapolis, MN). The C3aR agonist peptides, containing 8 and 15 residues (AAALGLAR and WWGKKYRASKLGLAR, respectively) (27), were synthesized to 95% purity by New England Peptide. Lyophilized peptides were dissolved in water at 10 g/l and diluted into cell suspensions. The wild type C3aR coding sequence containing both amino-and carboxyl-terminal tags and the variants C3aR YFFFF and FYYYY were cloned into the retroviral vector pQCXIX (BD Biosciences) for transduction into Cf2 cells. The cDNAs for human tyrosyl protein sulfotransferases (TPSTs) 1 and 2 were amplified by PCR from U87 human astroglioma cell cDNA and subcloned into pcDNA3.1 (28). Small hairpin RNA (shRNA) constructs targeting nucleotides 259 -276 of TPST1 and nucleotides 73-94 of TPST2 were generated and subcloned into pBluescript under the control of the murine U6 promoter (29).
Labeling and Immunoprecipitation of C3aR and C3aR Variants-HEK 293T cells were transfected with plasmids encoding C3aR or the C3aR variants using calcium phosphate (14). One day later, cells were washed twice with phosphate-buffered saline (PBS) and subcultured into three separate aliquots. One of each of the flasks was labeled with [ 35 S]cysteine and [ 35 S]methionine ([ 35 S]-Express, PerkinElmer Life Sciences), one was labeled with [ 35 S]sulfate (PerkinElmer Life Sciences) overnight, and the third was retained for binding and flow cytometry. Cells were treated with 3 mg/ml tunicamycin (Sigma-Aldrich) to inhibit N-glycosylation 5 h prior to and throughout the labeling period (14). For immunoprecipitation, labeled cells were harvested and lysed in 1% N-dodecyl-␤-D-maltoside (Anatrace) in PBS containing a protease inhibitor mixture (Sigma-Aldrich and Roche Applied Science) and 0.2 mM phenylmethylsulfonyl fluoride (Sigma-Aldrich). Cell debris was removed by centrifugation at 18,000 ϫ g for 5 min at 4°C, and the supernatants were immunoprecipitated in the presence of the anti-C9tag antibody, 1D4, covalently cross-linked to protein A-Sepharose (Amersham Biosciences). Immunoprecipitates were washed twice with 1% N-dodecyl-␤-D-maltoside in PBS containing 0.5% SDS and once with PBS and then eluted with SDS sample buffer under reducing conditions by heating at 55°C for 10 min and analyzed by autoradiography following electrophoresis on 12% SDS Tris-glycine polyacrylamide gels (Invitrogen).
O-linked carbohydrates were enzymatically removed from immunoprecipitated C3a receptors using O-glycanase (Glyko) and neuraminidase. N-linked carbohydrates were removed with PNGase F (New England Biolabs), as described by the manufacturer under non-denaturing conditions, eluted with SDS sample buffer, and analyzed as above on 12% SDS gels.
Binding of C3a to Cells Expressing C3aR and C3aR Variants-Binding experiments were performed using HEK 293T cells transfected with wild type C3aR or C3aR variants. Two days after transfection, cells were detached with 5 mM EDTA in PBS, washed with Dulbecco's modified Eagle's medium, counted, and resuspended in binding buffer (20 mM HEPES, pH 7.4, 125 mM NaCl, 1 mM MgCl 2 , 1 mM CaCl 2 , 5 mM KCl, 0.5 mM glucose, 0.2% bovine serum albumin, and 0.02% sodium azide) at 2 ϫ 10 6 /ml. An aliquot of cells was subjected to flow cytometric analysis to control for the relative expression levels of the receptors. Briefly, 5 ϫ 10 5 cells were incubated with 0.5 g/ml anti-Myc antibody 9E10 followed by fluorescein isothiocyanate-conjugated goat antimouse IgG (Santa Cruz Biotechnology), fixed in 2% paraformaldehyde, and analyzed using a FACS-Star cytometer and CellQuest software (BD Biosciences). For binding, duplicate aliquots of 50 l were incubated with 0.1 nM 125 I-C3a (PerkinElmer Life Sciences) and 0 -200 nM unlabeled C3a (Advanced Research Technologies) for 30 min at 37°C in a final volume of 100 l. Cells were centrifuged, washed once with binding buffer, and bound C3a was determined by ␥-counting. Binding data were subjected to nonlinear regression analysis using Prism software (GraphPad).
For some experiments, the degree of sulfation was modified by cotransfecting the C3aR variant YFFFF with plasmids encoding TPST1 and 2 to increase sulfation or with shRNA constructs directed against TPST1 and 2 to inhibit sulfation. Controls included co-transfection of the C3aR variant with pcDNA3.1 without insert.
Calcium Mobilization Mediated by the C3aR and C3aR Variants-The transduction vector pQCXIX (BD Biosciences) encoding the wild type C3aR or C3aR variants FYYYY and YFFFF (see Table I) were co-transfected into 293 GP packaging cells (BD Biosciences) with plasmids encoding the vesicular stomatitis virus G protein (VSV-G) envelope and Gag/Pol using calcium phosphate. Two days later, the culture supernatants were collected, and virus particles were concentrated by centrifugation at 15,000 ϫ g for 75 min. This concentrated virus was resuspended in 0.5-ml cell culture media and used to infect Cf2 cells, which were plated at 3ϫ 10 5 cells per well in 12-well plates 1 day before. One day later, the infected Cf2 cells were transferred to flasks and grown in Dulbecco's modified Eagle's medium for at least 24 h. Receptor expression was quantified by flow cytometric analysis as described above. Cells were harvested, counted, and incubated with the indicator dye Fura-2-AM (Molecular Probes) for 1 h at 37°C in 20 mM HEPES, pH 7.4, 1 mM CaCl 2 , 1 mM MgCl 2 , 125 mM NaCl, 5 mM KCl, 0.5 mM glucose, and 0.2% bovine serum albumin. Cells were washed twice and resuspended in the same buffer at 1ϫ 10 6 /ml. Changes in intracellular calcium concentration in response to C3a or C3a synthetic peptide agonists were determined fluorometrically at 37°C by monitoring the emission at 510 nm and the excitation at 340 and 380 nm as a function of a time. Responses were quantified as the peak of the fluorescence ratio of 340/380 nm. Measurements were made using cells from at least three independent infections.

RESULTS
Tyrosines in the ECL2 of the C3aR Are Sulfated-Previous work has shown that the sulfation of tyrosines on the extracellular amino-terminal domain of the C5a anaphylatoxin receptor is a critically important posttranslational modification for ligand recognition (14). Given the similarities of the C3a and C5a anaphylatoxins and their receptors, we hypothesized that a similar posttranslational modification might be important for C3aR function as well. The C3aR has no sulfatable tyrosines in its extracellular amino-terminal sequence; however, seven are evident in the unusually large ECL2 consisting of 172 amino acids (12). In addition, the tyrosines at positions 184, 188, 317, and 318 are flanked by acidic residues characteristic of preferred sites for sulfation (21,22). Studies with chimeric C3a/ C5a receptors and several ECL2 deletion mutants have shown that ECL2, particularly its amino acids adjacent to the fourth and fifth transmembrane sequences, is essential for C3a binding (15,16). We therefore tested the hypothesis that one or more of the tyrosines in this region is sulfated and plays a role in C3a binding and signal transduction. Tyrosines at positions 174, 184, 188, 317, and 318 in ECL2 were individually and multiply mutated to phenylalanine, as indicated in Table I. Constructs were transfected into mammalian cells and tested for expression, binding, and signal transduction.
To facilitate the determination of relative levels of cell surface expression, expression plasmids were engineered to encode the wild type or mutant C3aRs fused in-frame with an aminoterminal Myc tag, which is recognized by the antibody 9E10. Immunoprecipitations were accomplished using the antibody 1D4, which is specific for the in-frame C9-tag at the carboxyl terminus, corresponding to the nine carboxyl-terminal amino acids of rhodopsin. Control experiments demonstrated that these modifications had no detectable effect on binding or signal transduction compared with native C3aR lacking these tags (not shown). Transfected HEK 293T cells were assessed for surface expression of wild type C3aR or C3aR variants by flow cytometry using the anti-Myc tag antibody and mean fluorescence compared for each experiment. Simultaneously, cells from the same transfection were labeled with [ 35 S]cysteine and [ 35 S]methionine or [ 35 S]sulfate and lysed, and C3aRs were immunoprecipitated with the anti-C9 antibody, 1D4, cross-linked to Protein A-Sepharose beads. To ensure selection of sulfation exclusively on tyrosines, tunicamycin was also included in the culture medium (10 -12).
As shown in Fig. 1, cells expressing the wild type C3a receptor and grown in the absence of tunicamycin reveal a predominant form migrating with an apparent molecular mass of ϳ90 kDa that is labeled with both [ 35  Because a previous report indicated an important role for sequences adjacent to transmembrane sequences 4 and 5 (15), we focused first on the tyrosines in these regions and generated the mutant receptors outlined in Table I. As shown in Fig. 2, only when all five tyrosines at positions 174, 184, 188, 317, and 318 in the ECL2 were changed to phenylalanine was the ability to incorporate [ 35 S]sulfate lost, suggesting that tyrosines at 255 and 306 may not be sulfated. As we have observed previously with other 7TMS receptors (14,23), when multiple tyrosines are mutated to phenylalanine in the C3aR, particularly the YFFFF and FFFFF variants (see Table I), their electrophoretic mobility is increased, likely due to alterations in SDS binding.
C3a Binding to Wild Type C3aR and C3aR Variants-To examine the functional importance of the sulfated tyrosines in the ECL2, we next compared the ability of 125 I-C3a to associate with wild type C3aR and C3aR variants in which one or more of the five sulfated tyrosines in ECL2 was replaced with phenylalanine. Transfected HEK 293T cells expressing similar levels of wild type C3aR or C3aR variants, as indicated by their fluorescence intensity of anti-Myc-tagged antibody staining, were tested in competition binding experiments with 125 I-C3a and increasing concentrations of unlabeled C3a as indicated.
As shown in Fig. 3, HEK 293T cells expressing the C3aR variants YFYYY, YYYFF, and YFFFF bound C3a with affinities similar to the wild type receptor. Uniquely, binding to C3aR FYYYY was not detected. Cell surface expression of this mutant receptor was essentially the same as the wild type receptor.
Calcium Mobilization through C3aR and Its Mutants-In the case of the C5a receptor, intact C5a was unable to induce a calcium flux when two of the three amino-terminal tyrosines were mutated to phenylalanine, although a synthetic peptide agonist could still activate the same cells (13). These data were interpreted as further supporting the two-site model for ligand-receptor interactions (30). Synthetic peptide agonists corresponding to the carboxyl-terminal 8 or 15 amino acids of C3a have also been described (27), and we sought to study them in parallel experiments.
In 293T cells the C3a receptor, like the C5a receptor, requires co-transfection with G␣16 to elicit a calcium flux (31). Because we have no convenient method to monitor expression  of G␣16, we used Cf2 cells, which appear to endogenously express an intact Ca 2ϩ signaling pathway for most G-proteincoupled receptors. These cells are relatively inefficiently transfected, so we packaged the wild type C3aR or C3aR variant expressing transduction vectors in murine leukemia viral particles pseudo-typed with the vesicular stomatitis virus G protein and used them to transduce Cf2 cells. Cells expressing wild type C3aR or its variants, C3aR YFFFF, which binds native C3a normally, and C3aR FYYYY, which does not, were tested for the ability to mobilize calcium in response to native C3a or the two synthetic C3a agonist peptides. As shown in Fig.  4, cells expressing wild type C3aR or C3aR YFFFF induced effective calcium signals in response to 100 nM native C3a, whereas the FYYYY variant failed to do so. In contrast, all three receptors, wild type C3aR, C3aR FYYYY, and C3aR YFFFF, responded almost equally well to both of the synthetic C3a agonists. Thus, the lack of calcium mobilization by the FYYYY variant is not due to a global alteration of receptor structure. The data further suggest that the C3a receptor, like the C5a receptor (14,30), presents distinct docking and activation sites for a ligand. Enzymatic Modulation of Sulfation-One possible explanation for the lack of C3a binding to C3aR FYYYY is the absence of sulfate on tyrosine 174. An alternative explanation could be that the expression of phenylalanine at this site altered receptor conformation due to the increase in hydrophobicity resulting from the absence of the hydroxyl group. We addressed these possibilities by altering the extent of C3aR sulfation. Tyrosine sulfation is accomplished by tyrosine protein sulfotransferases 1 and 2. The activity of these enzymes can be modified by transfection, either with plasmids encoding the native enzymes to boost sulfation or with constructs encoding shRNAs target-ing the TPSTs to inhibit them (28). To try to increase C3aR sulfation, we co-transfected the C3aR variant YFFFF with plasmids encoding TPST1 and 2. Cells transfected with C3aR YFFFF alone or with TPST1 and 2 were assessed for receptor expression levels, binding, and extent of sulfation. As shown in Fig. 5, no change was observed for ligand binding or sulfation when cells were co-transfected with TPST1 and 2, suggesting that either the C3aR variant is already maximally sulfated in these cells or that the sulfate moiety is not critical.
To test the possibility that C3aR YFFFF was maximally sulfated under these conditions, we transfected cells with C3aR YFFFF alone or in combination with shRNA constructs to target both TPST1 and 2. These cells were tested for C3aR expression and ligand binding as before. As shown in Fig. 6, although C3aR expression levels were quite similar, maximal binding was significantly reduced with no change in affinity. Consistent with the critical nature of this sulfotyrosine residue, incorporation of [ 35 S]sulfate was commensurately reduced (Fig.  6). Thus sulfation of tyrosine 174 on the C3a receptor is a critical determinant for recognition of native C3a. DISCUSSION The molecular cloning of the C3a receptor places this molecule in the superfamily of G-protein coupled 7TMS receptors (10 -12). Despite a number of similarities, the C3a receptor does not have a highly acidic tyrosine-rich amino-terminal  Table I for identification of the variant receptors). Cells were split and labeled with [ 35 S]-cysteine and [ 35 S]methionine or [ 35 S]sulfate in the presence of tunicamycin, and C3a receptors were purified and analyzed as described for Fig. 1. B, the C3aR variants, FYYYY, YFFFF, and FFFF, were transfected and analyzed as described for panel A. Only when all five tyrosines were mutated to phenylalanine was labeling with [ 35 S]sulfate substantially reduced.
FIG. 3. Sulfotyrosine 174 in the ECL2 of the C3a receptor forms the binding site for C3a. A, HEK 293T cells were transfected with plasmids encoding wild type C3aR (YYYYY) or the C3aR variants YFYYY or YYYFF. Cells were incubated with 0.1 nM 125 I-C3a and the indicated concentrations of unlabeled C3a as described under "Experimental Procedures." Cells were washed, and bound 125 I-C3a was determined by ␥-counting. Aliquots of the same transfections were analyzed for C3aR expression levels by flow cytometry using the anti-Myc-tagged antibody 9E10. Each point was determined in duplicate, and data are expressed as the percentage of specific binding determined from at least three independent experiments. B, cells were treated in a manner identical to that described for panel A, except that ligand binding to the C3aR variants YFFFF and FYYYY are compared with wild type C3aR. sequence like the C5a receptor, which we have recently shown to be involved in its interaction with C5a (14). The predicted extracellular amino-terminal sequence of the C3a receptor is relatively short (ϳ21 amino acids) and has been shown not to participate in C3a binding (15,16). Instead, the C3aR has a large ECL2 that contains seven potentially sulfatable tyrosines. Here, we have shown that five of the seven tyrosines in this region of the C3a receptor are sulfated and that sulfotyrosine 174 alone plays a critical role in the binding and signaling of native C3a. The absence of sulfotyrosine 174 does not alter transduction in response to either of the two synthetic C3a peptide agonists tested, suggesting first, that this tyrosinemutated C3aR variant is not completely inactivated. Second, the data lend support to a two-site model for ligand-receptor interactions similar to C5a and C5aR interactions as well as interactions between a number of other peptide ligand receptor pairs (30,32). The observation that native C3a does not signal in C3aR FYYYY-expressing cells while agonist peptide responses are not altered suggests that the "docking" step induces a conformational change in the ligand, exposing a structure that can subsequently activate the cell.
Sulfation of tyrosines is a posttranslational modification occurring late in the trans-Golgi network (33), and estimates indicate that up to 1% of all tyrosines in eukaryotic proteins are sulfated (34). This modification is most appreciated on a large number of secreted proteins and in the extracellular domains of membrane-bound proteins (21,33). More recently, we and others have demonstrated the existence of sulfated tyrosines in amino-terminal sequences of several chemoattractant receptors that play a critical role in receptor function (23)(24)(25)(26). In the case of the chemokine receptors CCR5, CXCR4, CCR2b, and CX3CR1, it is not clear whether distinct docking and activation FIG. 4. The C3aR variant FYYYY does not mobilize calcium in response to native C3a but responds robustly to C3a synthetic peptide agonists. Cf2Th cells expressing wild type C3aR or the C3aR variants, FYYYY and YFFFF, were generated by transduction as described under "Experimental Procedures." Cells were loaded with the calcium indicator dye Fura-2 and assessed for their ability to mobilize calcium in response to the C3a or C3a peptides, which were added at the time points indicated by the arrows. Calcium flux is displayed as the fluorescence ratio of 340 nm to 380 nm. Receptor expression was determined by staining with the anti-Myc-tagged antibody, and the mean fluorescence levels were 79.3 for wild type C3aR, 68 for C3aR YFFFF, and 78.9 for C3aR YFFFF; for mock-transfected cells the value was 7.7. The data shown are representative of three independent experiments. sites exist, because small synthetic peptide agonists have not been identified. In contrast, the C5a anaphylatoxin receptor has been shown to contain sulfotyrosines in its amino terminus that are involved in formation of the docking site but not the activation site (14).
In the C3a receptor, tyrosines at positions 184, 188, 317, and 318 are adjacent to or in the proximity of acidic amino acids, a feature that is believed to favor tyrosine sulfation, and our data are consistent with this prediction. Tyrosine 174, however, which is surrounded only by basic and neutral residues, is also sulfated. This appears to contradict the predicted substrate requirement for TPSTs and suggests that the substrate specificity may be considerably broader (35). To ensure that the loss of C3a binding commensurate with the mutation of tyrosine 174 to phenylalanine reflected the loss of sulfate and not a change in receptor conformation, we attempted first to increase the extent of sulfation by co-transfecting the C3aR variant YFFFF with TPST1 and 2. An increase in sulfation without a change in the level of receptor expression should have increased the binding of ligand if the sulfate was important and the receptor less than maximally sulfated. Our data suggest this receptor variant is maximally sulfated, because no change was evident with overexpression of TPST; but when the cotransfections were performed using shRNA constructs targeting TPST1 and 2, both sulfation and binding were inhibited, with essentially no change in receptor expression. The observation of only partial reduction in C3a binding is reflective of the enzymatic activity of TPSTs synthesized prior to shRNA transfection.
Tyrosine sulfation has been identified as a key mediator of protein-protein interactions involved in leukocyte adhesion, hemostasis, and chemokine signaling (36). Sulfate is a charged and highly polarizable moiety that is likely to contribute significant free energy to the binding. Here we have shown that tyrosine at 174 is not only sulfated but that this sulfated moiety is critically important for C3a binding. This result is consistent with the previous report using deletion mutants of the C3aR, wherein deletion of amino acids from 198 to 308 in ECL2 (as much as 65% of this region) did not affect C3a binding or calcium mobilization (15). The removal of residues 174 to 183 resulted in a total loss of receptor function. Our findings demonstrate the importance of sulfotyrosine at 174 for C3a receptor function. It is curious that additional tyrosines in ECL2 are sulfated but have no apparent role in C3a-C3aR interactions. That finding, coupled with the unusual size of this receptor domain, suggests the possibility of additional ligands. Potential candidates might include other products of C3 that result from complement activation. Alternatively, this region may associate with another cell surface protein possibly modulating signal transduction.
Taken together, our data support a two-site model in which the binding/docking site for the native ligand is separable from the activation/effector site. In other characterized receptors with peptide ligands, the docking/binding site resides in the acidic amino acid and tyrosine-rich amino termini, and the activation site includes ECL2 and transmembrane helices. In the case of the C3aR, the docking site also resides in ECL2 but is clearly separable from the activation site. Even though five tyrosines in this region of the C3aR are sulfated, only the tyrosine at 174 is involved in C3a binding. Receptors were immunoprecipitated and analyzed as described for Fig. 1. B, cells from the same transfection were tested for binding of 125 I-C3a as described for Fig. 3 following flow cytometric analyses of receptor expression levels. Only transfections with similar expression levels were used. Neither sulfation nor receptor binding is completely inhibited, as shRNAs cannot inactivate TPSTs already synthesized.