Molecular cloning and characterization of the human anaphylatoxin C3a receptor.

In a human neutrophil cDNA library, an orphan G-protein-coupled receptor, HNFAG09, with 37% nucleotide identity to the C5a receptor (C5a-R, CD88) was identified. A novel feature of this gene, unlike C5a-R and other G-protein-coupled receptors, is the presence of an extraordinarily large predicted extracellular loop comprised of in excess of 160 amino acid residues between transmembrane domains 4 and 5. Northern blot analysis revealed that expression of mRNA for this receptor in human tissues, while similar, was distinct from C5a-R expression. Although there were differences in expression, transcripts for both receptors were detected in tissues throughout the body and the central nervous system. Mammalian cells stably expressing HNFAG09 specifically bound 125I-C3a and responded to a C3a carboxyl-terminal analogue synthetic peptide and to human C3a but not to rC5a with a robust calcium mobilization response. HNFAG09 encodes the human anaphylatoxin C3a receptor.

In a human neutrophil cDNA library, an orphan Gprotein-coupled receptor, HNFAG09, with 37% nucleotide identity to the C5a receptor (C5a-R, CD88) was identified. A novel feature of this gene, unlike C5a-R and other G-protein-coupled receptors, is the presence of an extraordinarily large predicted extracellular loop comprised of in excess of 160 amino acid residues between transmembrane domains 4 and 5. Northern blot analysis revealed that expression of mRNA for this receptor in human tissues, while similar, was distinct from C5a-R expression. Although there were differences in expression, transcripts for both receptors were detected in tissues throughout the body and the central nervous system. Mammalian cells stably expressing HNFAG09 specifically bound 125 I-C3a and responded to a C3a carboxyl-terminal analogue synthetic peptide and to human C3a but not to rC5a with a robust calcium mobilization response. HNFAG09 encodes the human anaphylatoxin C3a receptor.
During complement activation the 74 -77-amino acid anaphylatoxins C3a, C4a, and C5a are released. They are potent inflammatory mediators, inducing cellular degranulation, smooth muscle contraction, arachidonic acid metabolism, cytokine release, and cellular chemotaxis (reviewed in Refs. 1-3), and have been implicated in the pathogenesis of a number of inflammatory diseases (4,5).
Studies have demonstrated the presence of a C3a receptor (C3a-R) on guinea pig platelets, rat mast cells, human neutrophils, eosinophils, and platelets (3). A single class of high affinity C3a binding sites has been characterized on human neutrophils and differentiated U937 cells (6). Competition binding and functional desensitization studies are consistent with the presence of a receptor for C3a, which is distinct from the C5a-R (3,6). However, there is evidence that C3a and C4a may bind to the same receptor as the two anaphylatoxins cross-desensitize guinea pig ileal tissue (2,3), although other investigators using guinea pig macrophages indicate that there may be separate receptors (7). Functional activity of the C3a-R is sensitive to pertussis toxin, consistent with the binding site being composed of a G-protein-coupled receptor (6).
A complete understanding of the role of C3a in the pathogenesis of the inflammatory response has been hampered by the lack of the cloned receptor. In this report we describe the molecular cloning, stable expression in mammalian cells, and functional characterization of the human C3a receptor. This same receptor was recently independently cloned from an HL-60 library by low stringency screening with a fMet-Leu-Phe receptor probe and, lacking functional data, claimed to be an orphan receptor (AZ3B) (8). Mouse L cells expressing AZ3B failed to bind and respond to the agonists examined, although C3a was not tested (8).

EXPERIMENTAL PROCEDURES
Materials-The C3a carboxyl-terminal analogue synthetic peptide (WWGKKYRASKLGLAR) (9) was obtained from Bachem Bioscience, Inc., King of Prussia, PA. C3a was purchased from Advanced Research Technologies, San Diego, CA. Human rC5a was expressed in Escherichia coli and purified to homogeneity. Other agonists were obtained from Sigma.
cDNA Cloning-cDNA library construction and screening were carried out essentially as described (10), and DNA sequence was determined using a ABI sequencer (11). Expressed sequence tag analysis (11)(12)(13) of cDNA clones derived from a human neutrophil (lipopolysaccharide activated) cDNA library (oligo(dT)-primed and constructed in the Uni-ZAP XR vector (Stratagene)) identified a clone demonstrating significant homology (approximately 40% amino acid sequence identity) to the C5a-R (14,15). This cDNA clone contained an incomplete open reading frame and therefore was used to reprobe the neutrophil cDNA library to obtain a full-length cDNA. The alignment of HNFAG09 and the C5a-R was determined by the method of Needleman and Wunsch (21) using the Gap comparison program of the Wisconsin Package, version 8, September 1994, Genetics Computer Group, Madison, WI.
Northern Blot Analysis-Commercially prepared (Clontech, Palo Alto, CA) multiple tissue blots containing approximately 2 g of poly(A) mRNA per lane were sequentially hybridized with random primer 32 Plabeled cDNAs spanning the coding regions of C5a-R and HNFAG09. C5a-R was cloned via PCR 1 from differentiated U937 RNA. The final washing step was carried out twice in 0.5 ϫ SSC, 1% SDS at 65°C for 20 min.
Stable Expression in RBL-2H3 Cells-To prepare HNFAG09 for expression in mammalian cells, a 1.6-kb cDNA fragment was obtained by PCR amplification that encompassed the entire HNFAG09 open reading frame. This fragment was subcloned into KpnI/HindIII sites of the mammalian expression vector, pCDN (16). Oligonucleotide primers used for PCR amplification were 5Ј-GA AGT GGT ACC ATG GCG TC -3Ј and 5Ј-GC TCC AAG CTT TCA CAC AGT TG -3Ј (the translation start and stop codons are underlined). RBL-2H3 cells were electroporated with either HNFAG09 or C5a-R in the pCDN mammalian expression vector (16), exactly as described (17). Individual G418-resistant (400 g/ml) colonies were isolated and expanded. Clonal cell lines expressing either HNFAG09 or C5a-R were chosen for further functional and binding studies.
Binding Assay-C3a was radioiodinated using IODO-BEADS * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked (Pierce) to a specific activity of 100 Ci/mmol. Increasing concentrations of cold competitor were added to 1 ϫ 10 6 cells in the presence of 125 I-C3a (2.3 nM), and the assay was performed essentially as described (6).

RESULTS AND DISCUSSION
Expressed sequence tag analysis (11-13) of cDNA clones derived from a human neutrophil (lipopolysaccharide activated) cDNA library identified a clone demonstrating significant homology (approximately 40% amino acid sequence identity) to the C5a-R. This expressed sequence tag contained an incomplete open reading frame that therefore was used to reprobe the neutrophil cDNA library to obtain a 2040-base pair cDNA encoding a complete orphan G-protein-coupled receptor of 482 amino acids, which shared 37% nucleotide identity throughout the coding regions with the C5a-R (Fig. 1A). Although similar to the C5a-R, this cDNA contains two predicted extracellular N-linked glycosylation sites and an unusually large extracellular domain between transmembrane domains 4 and 5 comprised of over 160 amino acid residues (Fig. 1A). The majority of the identical residues between the C5a-R and HNFAG09 reside in the predicted transmembrane spanning domains and in the second intracellular loop (Fig. 1B).
By Northern blot analysis, expression of HNFAG09 in human tissues and cell lines is distinct from C5a-R expression. An ϳ2.2-kb C5a-R transcript was abundantly expressed in peripheral blood leukocytes (PBL), lung, spleen, heart, placenta, spinal cord, and throughout the brain. An ϳ2.1-kb HNFAG09 transcript was predominantly expressed in lung, spleen, ovary, placenta, small intestine, throughout the brain, and to a much lesser extent than C5a-R, in heart and PBL (Fig. 2). Although by Northern blot analysis the specific cells within the various tissues examined, which are expressing C5a-R and HNFAG09, cannot be determined, these data are suggestive that these receptors are abundantly expressed throughout the body. By fluorescent activated cell sorting using polyclonal antibodies generated to fusion proteins composed of glutathione S-transferase or maltose binding protein and the extracellular loop, this receptor has been shown to be expressed on several cell types, including U937, HL-60, PBL, and human neutrophils and monocytes (8).
Preliminary functional characterization in Xenopus laevis oocytes suggested that HNFAG09 encoded a human anaphylatoxin receptor. 2 To confirm these results in mammalian cells, this receptor was expressed in RBL-2H3 cells (19), a rat basophil cell line, which when transfected with an expression plasmid encoding the C5a-R expresses receptors that are functionally active (17). RBL-2H3 cells were stably transfected with mammalian expression plasmids encoding the C5a-R or HNFAG09, and Fura 2-loaded cells were tested for a C5a-or C3a-induced mobilization of intracellular Ca 2ϩ . C5a-R-but not FIG. 1. A, nucleotide and deduced amino acid sequence of HNFAG09. The predicted seven-membrane spanning domains of HNFAG09 are indicated by bold, and glycosylation sites are indicated by italics and underline. This nucleotide sequence has been submitted to GenBank; the accession number is U62027. B, predicted membrane topology of HNFAG09. Amino acid residues in common between C5a-R and HNFAG09 have been highlighted black; two predicted N-linked glycosylation sites, in the large extracellular loop and the amino terminus, are indicated by gray shading. The Human Anaphylatoxin C3a Receptor 20232 HNFAG09-expressing cells responded to rC5a (Fig. 3, A and B). A robust response to a C3a carboxyl-terminal analogue synthetic peptide (WWGKKYRASKLGLAR) (9) (EC 50 ϭ 3.9 nM) was detected in cells expressing HNFAG09, but no response was obtained for C5a-R-expressing cells (Fig. 3, D and C, respectively). Similarly, HNFAG09 but not C5a-R expressing RBL-2H3 cells also responded to native human C3a (EC 50 ϭ 0.3 nM; data not shown).
C3a was radioiodinated and used in whole cell binding assays to further characterize HNFAG09. Binding of 125 I-C3a to HNFAG09 expressing RBL-2H3 cells was competed by increasing concentrations of C3a (IC 50 ϭ 3.0 nM) and the C3a analogue synthetic peptide (IC 50 ϭ 155 nM) but not by rC5a (Fig. 3E). By saturation binding and Scatchard analysis a single class of C3a binding sites was identified with an estimated K d of 0.3 nM and a B max of 32,000 receptors/cell (data not shown). Curiously, a HEK 293 cell line stably expressing HNFAG09 mRNA by Northern blot neither bound nor responded to C3a (data not shown).
RBL-2H3 cells expressing HNFAG09 bind and respond to C3a and a C3a analogue synthetic peptide but not C5a. These data, along with the results of the tissue distribution analysis, are consistent with HNFAG09 (AZ3B) (8) encoding the human C3a receptor.
The demonstration that C5a-R (reviewed in Ref. 20) and C3a-R expression is not limited to myeloid cells but that they both are expressed in a variety of non-myeloid cells throughout the body and that they are abundantly expressed in the central nervous system is consistent with these receptors having a much greater role in the pathogenesis of inflammatory and autoimmune diseases than previously suspected. Now that the receptor for C3a has been identified, further studies to elucidate the role of C3a in immune function and disease will be facilitated.