The Transcription Factors AP-1 and Ets Are Regulators of C3a Receptor Expression*

The anaphylatoxin C3a is a proinflammatory mediator generated during complement activation. The tight control of C3a receptor (C3aR) expression is crucial for the regulation of anaphylatoxin-mediated effects. Key factors regulating constitutive expression of the C3aR in the mast cell line HMC-1 and receptor induction by dibutyryl-cAMP in monomyeloblastic U937 cells were determined by functional characterization of the C3aR promoter. Nucleotides -18 to -285 upstream of the translational start site proved to be critical for promoter activity in HMC-1 cells. Binding sites for the transcription factors AP-1 and Ets could be located. Overexpressed c-Jun/c-Fos (AP-1) and Ets-1 led synergistically to increased promoter activity that was substantially reduced by site-directed mutagenesis of the corresponding elements within the C3aR promoter. In HMC-1 cells, Ets interacted directly with the predicted binding motif of the C3aR promoter as determined by electromobility shift assays. AP-1 binding to the C3aR promoter was augmented during C3aR induction in U937 cells. A retroviral gene transfer system was used to express a dominant negative mutant of Ets-1 in these cells. The resulting cells failed to up-regulate the C3aR after stimulation with dibutyryl-cAMP and showed decreased AP-1 binding, suggesting that Ets acts here indirectly. Thus, it was established that Ets and the AP-1 element mediates dibutyryl-cAMP induction of C3aR promoter activity, hence providing a mechanistic explanation of dibutyryl-cAMP-dependent up-regulation of C3aR expression. In conclusion, this study demonstrates an important role of AP-1 and a member of the Ets family in the transcriptional regulation of C3aR expression, a prerequisite for the ability of C3a to participate in immunomodulation and inflammation.

The complement system is activated in a variety of diseases (1)(2)(3) and leads to the generation of the anaphylatoxic peptide C3a. 2 This proinflammatory mediator binds specifically to its heptahelical receptor (C3aR) (4,5) which is expressed on various cells such as leukocytes, endothelial and broncho-epithelial cells, smooth muscle cells of the lung, and even neurons (4,6,7).
The best experimental evidence for the involvement of C3a in diseases has been obtained in animal models using C3aR knock-out mice, guinea pigs with a natural C3aR defect, or specific C3aR inhibitors (8 -12). Intriguingly, C3a and its receptor act partially protective, partially destructive. In endotoxin shock, C3a seems to be protective; the mortality rate of C3aR Ϫ/Ϫ mice was much higher than that of their wild type littermates, investigated by a cecal ligation puncture sepsis model (12). Moreover, C3a participates in liver regeneration after partial hepatectomy or toxic injury (13,14).
As demonstrated in an adjuvant-induced arthritis rat model, C3a aggravates the inflammatory reaction (8). In asthma, C3a is deleterious because C3aR Ϫ/Ϫ -mice are protected from ovalbumin-driven airway hyper-responsiveness and early phase bronchoconstriction (15). In man C3a might play a similar role because its concentration is augmented in bronchoalveolar lavage fluid of asthmatics after allergen provocation (16). Mast cells and eosinophils contribute to certain inflammatory responses. The proinflammatory mediator C3a is a chemotaxin for these cells, suggesting their anaphylatoxin-mediated recruitment into inflamed tissue (17)(18)(19). As recently demonstrated, mast cells are additionally involved in autoimmune diseases (20,21). Here, anaphylatoxins are also of importance.
C3a exhibits various biological functions in leukocytes, mast cells, and related cell lines expressing the C3aR. Mediated by a pertussis-toxin sensitive G-protein, it leads to the increase of free cytosolic Ca 2ϩ , cell activation, and degranulation (17,18). Surface expression of adhesion molecules (22) and cytokine release are also modulated by C3a (23)(24)(25)(26). Obviously, the tightly controlled expression of the C3aR is crucial for the regulation of anaphylatoxin-mediated effects.
As an example for such a regulation in situ, C3aR expression cannot be detected in normal brain sections although a strong C3aR immunohistochemical staining is evident in areas of inflammation. In meningitis, the C3aR is abundantly expressed by reactive astrocytes, microglia, and infiltrating macrophages and neutrophils. In multiple sclerosis, a strong staining can be additionally detected on smooth muscle cells surrounding blood vessels (27). Finally, in experimental allergic encephalitis the C3aR was elevated on microglia, infiltrating monocytes/ macrophages, and astrocytes (28). C3aR regulation can also be observed in other tissues: In ovalbumin-or lipopolysaccharide-challenged mice, C3aR expression increased on bronchial smooth muscle cells (29). In patients suffering from atopic dermatitis or allergic contact dermatitis, C3aR expression can be detected in CD4 ϩ and CD8 ϩ T-cell clones, cell types that normally do not express this receptor (6,30). Corresponding to the regulation in situ, the C3aR can be induced in cell culture by Bt 2 cAMP or interferon-␥ on monomyeloblastic cell lines such as U937 (31).
The aim of the current study was to obtain a deeper understanding of the regulation of the human C3aR, in particular, its transcriptional control, which has never been investigated before. Constitutive expression of the C3aR in the mast cell line HMC-1 (32) and receptor induction by Bt 2 cAMP (and TPA) in monomyeloblastic U937 cells (33) were investigated. The transcription initiation site was identified. The essential part of the C3aR promoter was narrowed down by 5Ј and 3Ј deletion analysis. Detailed characterization by various methods revealed a crucial role for the transcription factors activator protein-1 (AP-1) and E26-transformation specific (Ets) and their corresponding binding sites within the C3aR promoter. Ets seems to act directly on the C3aR promoter but also indirectly when receptor expression is induced.

MATERIALS AND METHODS
All reagents were obtained from Sigma if not otherwise indicated. Cell Lines and Cell Culture Conditions-The human mast cell line HMC-1 (kindly provided by J. H. Butterfield) and human monomyeloblastic U937 cells (ATCC CRL 1593) were cultured in RPMI 1640 medium (Biochrom AG, Berlin, Germany) at 37°C, 5% CO 2 . The human embryonic kidney cell line HEK293 (ATCC CRL 1573) (4) was cultured in Dulbecco's modified Eagle's medium/Ham's F-12 (Biochrom AG). Both media were supplemented with 10% (v/v) fetal calf serum (Sigma).
Transient Transfection of HMC-1 Cells and Reporter Gene Assay-Human C3aR promoter fragments were generated by PCR using the pDP1 plasmid as template (34) and subcloned in between the NheI and HindIII site of pSEAP2-E (Clontech, Heidelberg, Germany). Primer sequences can be obtained from the corresponding author. Point mutations were introduced using the QuikChange mutagenesis protocol (Stratagene, La Jolla, CA): TATAA to GCTAA for the TATA box, CAGGAAG to CAAGAAG for the Ets binding site, ATGACTC to ACAACTC for AP-1. Putative transcription factor binding sites were identified with MatInspector Version 6.2. All constructs were sequenced for verification. HMC-1 cells were transiently transfected in 6-well plates at 1 ϫ 10E6 cells/well with the C3aR promoter constructs either alone or in combination with Ets-1 from chicken, human c-Jun, c-Fos, and Ha-Ras (generous gifts from Dr. A. Nordheim and Dr. B. Luescher), 200 ng each. Effectene transfection reagent (Qiagen, Hilden, Germany) was used according to the manufacturer's instruction. The transfection efficiency was normalized by cotransfection with pCMV-␤-galactosidase (Clontech). Cells were harvested and centrifuged 72 h post-transfection. Secreted alkaline phosphatase (SEAP) was determined in the supernatant; ␤-galactosidase was determined in the cell-lysate. Great EscAPE SEAP TM detection Kit from Clontech and High Sensitivity ␤-GAL Assay Kit TM were used as described in the manufacturer's protocol.
Transient Transfection of HEK293 Cells-HEK cells were transiently transfected with Lipofectamine TM according to the manufacturer's instructions (Invitrogen). Mutant of Ets-1 and Its Transduction into U937 Cells-A retroviral gene transfer system was applied  to generate U937 cells expressing a dominant negative mutant of Ets-1.  It consists of amino acids 306 -441 of chicken Ets-1 (accession number  M22462) and is, thus, lacking the transactivation domain (35).

Generation of a Dominant Negative
RNA Preparation and cDNA Synthesis-Total RNA was isolated from cell cultures using TRIzol TM reagent (Invitrogen) according to the manufacturer's protocol. Five g of total RNA was transcribed by SuperScript II RNase H Ϫ reverse transcriptase from Invitrogen using an oligo(dT) primer (Stratagene).
Detection of the Transcription Initiation Site by RT-PCR-cDNA (2 of 60 l) was generated using a reverse antisense primer (CTGCATCT-TCAGGCCAGC) hybridizing to the coding region of the C3aR for cDNA synthesis. The resulting cDNA was applied as template in PCR reactions, combining this reverse antisense primer with a whole set of sense primers that anneal to the C3aR 5Ј-UTR. The amplified genomic DNA fragments could be excluded because they would additionally contain the 6-kilobase intron. The construct pSEAP2-E308-hC3aR-234, artificially combining part of the promoter region (308 bp) with part of the coding region of the C3aR, was used as a positive control. The primer positions upstream from the start codon were: Nuclease Protection Assay-3, 10, and 30 g of RNA were used in the nuclease protection assay (Ambion Inc., Austin, TX). Thirty g of yeast RNA served as the negative control. A probe representing part of exon 1 of the C3aR (position Ϫ444 to Ϫ26) was generated by PCR using a [␣-35 S]dATP-labeled oligonucleotide (GTCCTGTCTGGTCCCCACA-CTTAG; Ϫ85 to Ϫ26) as the reverse antisense primer and a biotinylated sense primer (Ϫ444 to Ϫ425). Genomic sequence (ligated into the plasmid DP1) served as the template. The sense primer was removed using streptavidin-coupled Dynabeads (Dynal/Invitrogen). The RNA was hybridized overnight at 43°C followed by nuclease treatment according to the manufacturer's instruction. After digestion, the protected fragments were separated in a 6% denaturing polyacrylamide gel followed by 3 weeks of exposure of the x-ray film (Kodak X-Omat Ar35 ϫ 43).
C3aR Real-time RT-PCR-TaqMan real-time RT-PCR was performed using a C3aR-specific TaqMan gene expression assay (HS00377780-m1) in combination with the TaqMan Universal PCR Master Mix and the 7000 Sequence Detection System for analysis (all Applied Biosystems, Weiterstadt, Germany). For normalization, the housekeeping gene RPS9 was used. Standard curves were generated using serially diluted cDNA. For each sample, the -fold induction was calculated as the ratio of the normalized volume equivalent of the transfected cells to the normalized volume equivalent of the corresponding mock-transfected cells.
Competitive 125 I-Labeled C3a and C5a Binding Studies-Competitive binding studies with human 125 I-labeled C3a or C5a (PerkinElmer Life Sciences) as tracer and increasing concentrations of cold C3a or C5a were performed as previously described (6).
Fura2 Assay for the Determination of Free Cytosolic Ca 2ϩ in Bt 2 cAMP-induced U937 Cells-This assay was performed exactly as previously described (6,31).
Binding reactions were performed in 20 mM HEPES, 50 mM NaCl, 0.1 mM EDTA, 2% Ficoll, 1 g of poly(dI-dC), and 1 mM dithiothreitol. For competition, 5-7.5 g of nuclear extract were preincubated for 20 min at room temperature with an excess of nonlabeled oligonucleotide: 5-, 50-, or 500-fold molar concentrations as indicated in the figure legends. After the addition of 100,000 cpm of the 32 P-radiolabeled oligonucleotide, the mixture was incubated at room temperature for 20 min. The DNA-nuclear factor complexes were resolved on a 1 ϫ Tris-buffered EDTA, 7% polyacrylamide gel pre-electrophoresed at room temperature for 60 min at 100 V. The gel was dried and then analyzed on a phosphorimaging plate. Fig. 1 and 2a) of the C3aR promoter, total RNA from HMC-1 cells was analyzed by nuclease protection assay. A single band at position Ϫ185 relative to the start codon became visible in the radiography (Fig. 1, upper panel), indicating the (major) transcription initiation site.

C3aR Transcription Initiation Site Determined by Nuclease Protection Assay and RT-PCR-To localize the transcription initiation site (TIS in
The RNA was additionally analyzed by RT-PCR, combining a whole set of sense primers annealing to the C3aR 5Ј-UTR with one common reverse antisense primer hybridizing to the coding region of the C3aR (Fig. 1, lower panel). Strong amplified fragments (P1-P4 in Fig. 1) became apparent using primer combinations with sense primers binding within 185 bp of the 5Ј-UTR relative to the start codon. Primer combinations with sense primers annealing further upstream resulted in weaker but still visible PCR products (P5-P10 in Fig. 1), suggesting an additional minor transcription start site located upstream of position Ϫ280. A vector containing the C3aR 5Ј-UTR (cloned from genomic DNA) in combination with part of the coding region of the C3aR served as positive control. As expected, amplified fragments resulted with all primer pairs (not shown).
Identification of Regulatory Regions by 5Ј and 3Ј Deletion of the C3aR 5Ј-UTR in HMC-1 Cells-To gain further insights into the modes of regulation of the C3aR, the minimal transcriptional unit of the C3aR promoter was mapped, and key regulatory elements within this promoter directing basal gene expression were identified. HMC-1 cells constitutively expressing the C3aR (and thus, exhibiting all necessary cofactors) were used in this deletion analysis.
The gene of the human C3aR is located on chromosome 12p13. A single 6-kilobase intron is present 17 bp upstream of the ATG initiation codon of exon 2 (34). A 2-kilobase region of the C3aR 5Ј-UTR located directly upstream of the intron was analyzed by SEAP reporter gene assay (Fig. 2a). For this purpose, a whole set of overlapping fragments of the C3aR 5Ј-UTR was generated by PCR and cloned into the pSEAP2-Enhancer vector, as indicated in the corresponding figures (Fig. 2, b-e). In all figures and throughout the paper, all nucleotide positions are given relative to the ATG start codon. The names of the constructs indicate the lengths of the analyzed fragments. In all experiments pCMV-␤ expressing ␤-galactosidase was cotransfected and used for normalization. All reporter gene experiments were performed at least n ϭ 3 times and repeated with an independent clone of each construct, leading essentially to the same results.
The long C3aR reporter gene constructs showed elevated SEAP activity compared with the pSEAP2-E control vector without insert. Progressive 5Ј deletions of the C3aR 5Ј-UTR from position Ϫ1921 to position Ϫ285 (with a common 3Ј-end at position Ϫ18) did not significantly affect SEAP expression in HMC-1 cells (Fig. 2b). Additionally, a whole set of similar constructs with a 5Ј-end in between position Ϫ1921 and Ϫ285 was analyzed, demonstrating similar promoter activity (data not shown). Thus, although in principle possible, it is unlikely that there are positive and negative regulatory elements within this region whose effects may cancel each other. Hence, the region upstream of position Ϫ285 does not most likely contribute to promoter activity in HMC-1 cells. However, different results might be obtained in other cell lines or if other stimuli are present.
In subsequent experiments the pSEAP2-E267 construct representing nucleotide Ϫ285 to Ϫ18 of the C3aR 5Ј-UTR was usually used as internal standard. For better comparison, the promoter activity of this construct was defined as 100% (usually, the corresponding SEAP activity was in the range of 50,000 -200,000 cps, and the activity of ␤-galactosidase used for normalization was in the range of 0.15-0.25 A 595 nm ). Further truncation of the 5Ј-end down to position Ϫ180 represented by the construct pSEAP2-E162 yielded a drastic reduction of promoter activity (Fig. 2c). Therefore, the region distinguishing the constructs 267 and 162, i.e. position Ϫ285 to Ϫ180, must exhibit positive regulatory promoter function.
To localize and identify positive regulatory elements positioned in this region, bioinformatic analysis using the MatInspector Professional Software (Version 6.2) was performed. It led to the prediction of consensus transcription factor binding sites, including a putative AP-1  DECEMBER 23, 2005 • VOLUME 280 • NUMBER 51 element centered at position Ϫ225 and an adjacent Ets site centered at position Ϫ207. Hence, one additional construct was generated excluding the AP-1 site but still including the Ets binding motif (pSEAP2-E197). Its activity was drastically reduced. However, promoter activity of this fragment containing the putative Ets site was still higher than those of pSEAP2-E162 (Fig. 2c). These data suggest that the AP-1 and the Ets binding motifs play a prominent role in C3aR regulation.

AP-1 and Ets as Regulators of C3a Receptor Expression
Promoter activity also decreased due to stepwise 3Ј deletion (pSEAP2-E-206, -142, and -125 in Fig. 2d). Thus, the region from position Ϫ160 to Ϫ18 also exhibits positive regulatory elements and partic-ipates in constitutive C3aR expression. The construct pSEAP2-E162 had no promoter activity (Fig. 2c). Hence, most likely the positive regulatory elements spanning Ϫ160 to Ϫ18 can only act in the context of additional promoter regions further upstream, present in pSEAP2-E-267, -206, -142, or -125, respectively.
TATA-less Promoter-The promoters of G-protein-coupled receptors are often TATA-less. Two putative TATA boxes are located close to the major transcription initiation site in the C3aR promoter (position Ϫ185; see above). To clarify experimentally whether these TATAA sequences represent active boxes or not, reporter gene constructs with single and combined point mutations of the two sites (Ϫ351 and Ϫ275, respectively) were generated within pSEAP2-E1511. The more upstream located TATAA sequence was analyzed to exclude redundancy. The wild type construct pSEAP2-E1511, which was mutated here, exhibited similar reporter gene activity as the other long constructs such as pSEAP2-E1607 (data not shown). The three mutants showed unaltered promoter activity in HMC-1 cells as compared with the wild type fragment, indicating that both TATAA sequences are not needed for C3aR transcription and expression (Fig. 2e).
Characterization of AP-1 and Ets Binding Sites within the C3aR Promoter Determined in HMC-1 Cells-In the following experiments the role of the above mentioned putative cis-acting elements for AP-1 and Ets was elucidated. First, we overexpressed the corresponding transcription factors in HMC-1 cells by transient transfection and analyzed them by SEAP reporter gene assay. SEAP expression in the absence of exogenous transcription factors and Ha-Ras was defined as 100%. The small GTP-binding protein Ha-Ras, which is needed for the activation of AP-1 and Ets, was partially coexpressed, as indicated in Fig. 3, a-c. Moreover, the AP-1 and Ets binding motifs in the reporter gene construct pSEAP2-E267 were disrupted by site-directed mutagenesis.
Simultaneous coexpression of the pSEAP2-E267 wild type promoter construct with c-Jun, c-Fos, and Ha-Ras yielded a strong, more than 10-fold increase of SEAP expression (Fig. 3a). In the same setting, the construct with the mutated AP-1 binding site showed only a slight increase of reporter gene activity. The individual transcription factor subunits of AP-1 (c-Jun and c-Fos in this study), just as Ha-Ras alone, increased the promoter activity ϳ3-fold. c-Jun. but not c-Fos. can form homodimers. Accordingly, the combined overexpression of c-Jun and Ha-Ras yielded a 5-fold enhancement of promoter activity. Hence, these data clearly show that c-Jun/c-Fos (or other members of the AP-1 family) can induce C3aR promoter activity and demonstrate the position of the consensus AP-1 element.
Similar experiments were performed for the Ets cis-acting element within the C3aR promoter. Therefore, Ets-1 was overexpressed as a prototypical member of the large Ets transcription factor family exhibiting the unique winged helix-turn-helix motif known as the Ets domain. As depicted in Fig. 3b, coexpression of Ets-1 and Ha-Ras strongly elevated promoter activity of the pSEAP2-E267 wild type construct. As expected, expression of Ets-1 or Ha-Ras alone was much less effective. The corresponding reporter gene construct with a mutation in the Ets binding motif showed decreased basal activity, which only modestly increased in the presence of overexpressed Ha-Ras and/or Ets-1. These data demonstrate and localize in HMC-1 cells a cis-acting element for an Ets transcription factor within the C3aR promoter.
To check whether AP-1 and Ets cooperate in C3aR regulation, HMC-1 cells were transfected with the pSEAP2-E267 wild type construct. The simultaneous cotransfection of c-Jun, c-Fos, and Ets-1 (in the presence of Ha-Ras) yielded higher SEAP activity than the activity induced by AP-1 or Ets-1 alone (Fig. 3c). For a detailed analysis of this cooperation, the mutants of the Ets and of the AP-1 binding motif, respectively, were additionally examined. Similar to other figures depicting results of reporter gene assays, 100% reporter activity is also defined here by the SEAP expression of the pSEAP2-E267 wild type construct. As demonstrated above, overexpressed Ets-1 did not increase promoter activity of the Ets binding site mutant, and overexpression of AP-1 did not result in higher promoter activity of the mutant with the destroyed AP-1 motif. Intriguingly, overexpression of Ets-1 caused only a slight increase in promoter activity of the AP-1 binding site mutant and vice versa. Overexpressed AP-1 resulted in an augmentation of enzyme activity, which was much smaller than that which can be observed on the wild type C3aR reporter gene construct.
Hence, we have identified a consensus AP-1 transcription factor binding site centered at position Ϫ225 and a consensus Ets binding element centered at position Ϫ207. Our data strongly suggest that the two corresponding transcription factors cooperate in the activation of the C3aR promoter.
Up-regulation of C3aR Transcription in HEK293 Cells Transiently Transfected with AP-1 and Ets-1-We wanted to clarify whether AP-1 and Ets do not only up-regulate promoter activity but demonstrate that  DECEMBER 23, 2005 • VOLUME 280 • NUMBER 51 they can actually increase C3aR transcription. Hence, C3aR-mRNA was quantified when the two transcription factors were coexpressed in cells that normally do not express the C3aR. In these experiments on HEK293 cells, Ha-Ras was always cotransfected. On day 3, mRNA was isolated. The level of C3aR-mRNA was determined by TaqMan realtime RT-PCR; the housekeeping gene RPS9 was used for normalization.

AP-1 and Ets as Regulators of C3a Receptor Expression
Compared with HEK293 cells, which only expressed Ha-Ras, a significant increase of C3aR-mRNA was observed when Ets-1 was additionally transfected or when Ets-1 and AP-1 were additionally coexpressed (Fig. 4). Moreover, the increase of C3aR-mRNA was significantly higher when AP-1 and Ets-1 were cotransfected as compared with the expression of only one of these two transcription factors. Additionally, in accordance with the data obtained in the reporter gene assays, an upregulatory effect of AP-1 on C3aR transcription could only be seen when Ets-1 was simultaneously expressed. Thus, AP-1 and Ets can up-regulate the level of C3aR-mRNA, i.e. C3aR transcription.
Investigation of C3aR Induction and Signaling in U937 Cells Expressing a Dominant Negative Mutant of Ets-1-To elucidate the role of Ets in C3aR induction, the up-regulation of the anaphylatoxin receptor was investigated when this transcription factor was functionally inactivated by a dominant negative mutant (Ets-1DN) lacking the transactivation domain. A retroviral gene transfer and expression system (RetroX, Clontech) had to be used for that purpose because U937 cells cannot be transiently transfected. Using the retroviral expression vector pQCXIN, Ets-1DN under the control of a CMV-promoter, and the neomycin resistance gene were cotranslated in mammalian cells via the internal ribosomal entry site (IRES) from a bicistronic message (Fig. 5a). Control virus was obtained from the same construct, lacking Ets-1DN. After selection with G418, a transduction/expression efficiency of Ͼ90% could be reached in this system. U937 cells were chosen for this investigation as the model cell line for C3aR induction, and Bt 2 cAMP was used as the primary inducer of anaphylatoxin receptors on these cells. C3aR expression was determined directly and indirectly by various methods: specific 125 I-labeled C3a binding, C3aR-dependent Ca 2ϩ signaling, and real-time RT-PCR. The expression of Ets-1DN was confirmed by EMSA as demonstrated in the following section (see the left two lanes of Fig. 7a).
As expected, 3 days of Bt 2 cAMP treatment led to a strong increase of C3aR expression in control vector-transduced U937 cells. In contrast, in three independent competitive binding studies, specific 125 I-labeled C3a binding was almost completely diminished in Bt 2 cAMP-treated U937 cells stably expressing Ets-1DN: The residual binding was only slightly higher than the minimal one, which could be observed in non-treated cells (Fig. 5b). Corresponding to that, only a slight increase of Ca 2ϩ signaling became apparent in the Bt 2 cAMP-treated U937-Ets-1DN cells when 40 nM C3a were applied (Fig. 5d). Control cells transduced with pQCXIN and induced with Bt 2 cAMP showed a normal C3a response in this assay.
The increase of C3aR-mRNA determined by TaqMan real-time RT-PCR after 72 h of induction with Bt 2 cAMP is in good accordance with the results of the binding studies and the functional Ca 2ϩ assay (demonstrating an inhibition of C3aR up-regulation). Expression of the dominant negative mutant (Ets-1DN) strongly diminished C3aR transcription in U937 cells in 3 experiments from ϳ15-fold down to ϳ2-fold (data not shown).
To check how specific the effects of the Ets-1 dominant negative mutant on receptor regulation are, the C5a receptor (C5aR, CD88) and fMLP receptor were also analyzed (37). These G-protein-coupled receptors are closely related to each other and to the C3aR, exhibiting overlapping functions. Both receptors can also be induced by Bt 2 cAMP on U937 cells. Their activation causes a strong Ca 2ϩ response. In contrast to the observations on the C3aR, there was almost no difference in the Ca 2ϩ response of Bt 2 cAMP-induced U937 cells due to the expression of Ets-1DN applying C5a or fMLP as stimulus (Fig. 5, e and f). In accordance with that, 125 I-labeled C5a binding of Bt 2 cAMP-treated U937 cells decreased only partially in the presence of Ets-1DN (Fig. 5c). Most likely, the remaining number of C5aR binding sites is still sufficient to reach an almost normal Ca 2ϩ response.
Hence, our results clearly show that induction of C3aR expression in U937 cells by Bt 2 cAMP is strictly dependent on Ets. Additionally, our data suggest that the regulation of the C5aR or the fMLP receptor is at least partially different from that of the C3aR, i.e. the induction of these receptors is less dependent on the Ets transcription factor family.
Binding of AP-1 to the C3aR Promoter, Determined by Electromobility Shift Assay-To confirm the presence of nuclear/transcription factors capable of binding to the AP-1 site centered at position Ϫ225, EMSAs were carried out using nuclear extracts prepared from U937 cells and a radiolabeled double-stranded oligonucleotide probe spanning nucleotides Ϫ213 to Ϫ233 of the C3aR promoter (AP-1-C3aRwt). Additionally, a commercial double-stranded AP-1 control oligonucleotide served also as radiolabeled probe (AP-1wt). For competition the same nonlabeled oligonucleotides and their derivatives with specifically mutated AP-1 binding motifs were coincubated (AP-1-C3aRmut and AP-1mut).
As depicted in Fig. 6a, incubation of the radiolabeled AP-1-C3aRwt probe with nuclear extracts from native (0 h) and induced U937 cells (2, 6, 24, or 48 h of Bt 2 cAMP) resulted in the formation of radiolabeled nuclear factor-DNA complexes. At least three bands could be distinguished, most likely representing the binding of dimers consisting of different monomers of the Fos, Jun, or activating transcription factor subfamilies. This phenomenon is well known for AP-1 binding (38). The nuclear factor-DNA complexes could be efficiently competed by an excess of nonlabeled double-stranded AP-1wt. The formation of the radiolabeled complexes could not be competed by the oligonucleotide AP-1mut, demonstrating the specificity of the observed AP-1 binding to the C3aR promoter (left part of Fig. 6a). In three independent experiments, the signal intensity of the radiolabeled nuclear factor-DNA complexes clearly increased during the time course of the experiment, as  , was incubated with a radiolabeled double-stranded oligonucleotide representing part of the C3aR promoter, in particular an AP-1 binding motif centered at position Ϫ225 ( 32 P-labeled AP-1-C3aRwt). Nuclear factor-DNA complex formation was competed with an excess of nonlabeled double-stranded oligonucleotides as indicated: AP-1wt, AP-1mut (both of them 500-fold), AP-1-C3aRwt, and AP-1-C3aRmut (5-, 50-, and 500-fold) for a, and AP-1-C3aRwt as well as AP-1-C3aRmut (500-fold) for b. For a, the shift of radioactivity (i.e. binding of AP-1 to the labeled probe) was measured using phosphorimaging and quantified by computer analysis. After subtraction of the background, the relative band intensities without competition for 0, 2, 6, 24, and 48 h were 1, 3.3, 7.2, 8.5, and 6.1. c, nuclear extract from HMC-1 cells was incubated either with 32 P-labeled AP-1-C3aRwt or with a radiolabeled consensus AP-1 oligonucleotide (AP-1wt) as probe. Nuclear factor-DNA complex formation was competed with a 500-fold excess of nonlabeled double-stranded oligonucleotides as indicated: AP-1wt, AP-1mut, AP-1-C3aRwt. In all three EMSA, the oligonucleotides exhibiting the AP-1 motif (AP-1wt and/or AP-1-C3aRwt) could compete efficiently. The oligonucleotide exhibiting a mutated AP-1 consensus motif could not compete, demonstrating the specificity of the interaction. The results of one of two independent experiments are depicted.  Ets-1DN). a, the dominant negative mutant of Ets-1 was ligated downstream from a CMV promoter into a NotI restriction site. In eukaryotic cells, a neomycin resistance gene for G418 selection was cotranslated from a bicistronic message via the internal ribosomal entry site (IRES). ϩ indicates the packaging signal. The long terminal repeats (LTRs) are derived from hybrids of CMV/murine sarcoma virus (MSV) or the Moloney murine leukemia virus (MoMuLV), respectively. b-f, human monomyeloblastic U937 cells that were either transduced with pQCXIN control vector or pQCXIN-Ets-1DN were incubated for 3 days with 500 M Bt 2 cAMP or buffer, respectively. B ϩ c, competitive binding curves were obtained applying tracer amounts of 125 I-labeled C3a or 125 I-labeled C5a and competing with increasing amounts of nonlabeled C3a or C5a, respectively. d-f, the increase of free cytosolic Ca 2ϩ as response to 40 nM C3a, 20 nM C5a, or 1 M fMLP was determined using the fluorescence indicator Fura2. The result of one representative of n ϭ 3 independent experiments (and duplicates (b ϩ c) and quadruplicates (d-f) within the experiment, respectively) with similar outcome is depicted. DECEMBER 23, 2005 • VOLUME 280 • NUMBER 51 determined by phosphorimaging and subsequent computer analysis (Quantity One 4.5.0 software, Bio-Rad).

AP-1 and Ets as Regulators of C3a Receptor Expression
To further characterize the interaction of the C3aR promoter with AP-1, the radiolabeled AP-1-C3aRwt probe was also competed with the corresponding nonlabeled oligonucleotide and its mutated counterpart using nuclear extracts from U937 cells stimulated for 48 h with Bt 2 cAMP (right part of Fig. 6a). As expected, the radiolabeled nuclear factor-DNA complexes could be specifically competed by the nonlabeled AP-1-C3aRwt oligonucleotide but not by the otherwise identical oligonucleotide harboring the mutated AP-1 site (AP-1-C3aRmut).
Similar results were obtained using the radiolabeled AP-1 probe (AP-1wt) and, as competitor, AP-1wt, AP-1mut, AP-1-C3aRwt, or an irrelevant double-stranded oligonucleotide serving as negative control. The results confirmed the up-regulation of a nuclear factor binding to the AP-1 cis-acting element in U937 cells due to Bt 2 cAMP treatment and the ability of this factor to bind specifically to the C3aR promoter (data not shown).
In HMC-1 cells, which constitutively express the C3aR, a similar nuclear factor-DNA complex resulted when the radiolabeled AP-1-C3aRwt probe was incubated with nuclear extract. This complex could be specifically competed by the nonlabeled AP-1wt oligonucleotide (and the nonlabeled AP-1-C3aRwt oligonucleotide) but not by AP-1mut (left part of Fig. 6c). The same pattern resulted when AP-1wt served as the radiolabeled probe (right part of Fig. 6c). Intriguingly, in HMC-1 cells, the nuclear factor-DNA complex consisted of one prominent band that showed a different electrophoretic mobility compared with the prominent bands observed in U937 cells. This observation suggests that different members of the AP-1 family bind to the C3aR promoter in the context of HMC-1 or U937 cells, respectively.
AP-1 was also found as a transcription factor that mediates gene induction by the phorbol ester tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA or PMA) and, hence, the alternative name TPA response element for the recognition site of AP-1. TPA is an activator of protein kinase C. Moreover, it induces up-regulation of C3aR-mRNA in U937 cells as demonstrated by Northern blot analysis (27). As expected, when radiolabeled AP-1-C3aRwt was applied as probe, one main nuclear factor-DNA complex could be detected in the EMSA using nuclear extract (abbreviated as NE in the figure) from TPA induced U937 cells (right part of Fig. 6b). The complex could be efficiently competed by an excess of nonlabeled AP-1-C3aRwt oligonucleotide but not with the corresponding oligonucleotide exhibiting a mutation in its AP-1 site (AP-1-C3aRmut). Using the same amount of nuclear extract from native U937 cells, such a complex was not detectable (left part of Fig. 6b).
As described above, C3aR induction was suppressed in U937 cells by the transduced Ets-1DN. To check whether the up-regulation of functional AP-1 is indirectly affected by the dominant negative mutant of Ets-1, EMSAs were performed incubating radiolabeled AP-1-C3aRwt as probe with nuclear extract from the corresponding U937 cells. In this experiment two different modes of induction were compared; 48 h of incubation with Bt 2 cAMP or TPA, respectively. Due to Ets-1DN expression the signal intensity of the AP-1-DNA complex was diminished by ϳ50% measured using phosphorimaging and quantified by computer analysis in two different experiments in nuclear extract obtained from U937 cells induced by Bt 2 cAMP. In contrast, the strength of the signal was not at all affected using nuclear extract obtained from U937 cells after TPA induction (data not shown).
Hence, the EMSA data show that the up-regulation of the C3aR in U937 cells induced by Bt 2 cAMP correlates with an increase of functional AP-1 with its maximum ϳ24 h after induction. Similarly, an up-regulation of functional AP-1 can also be observed due to the phorbol ester TPA, which is an inductor of C3aR transcription. Moreover, we have confirmed the presence of nuclear factors that specifically bind to the consensus AP-1 cis-acting element within the C3aR promoter in induced U937 and native HMC-1 cells. Stably expressing Ets-1DN, it could additionally be shown that the up-regulation of AP-1 and its binding to the C3aR promoter in Bt 2 cAMP-induced but not in TPA-induced U937 cells is partially dependent on Ets activation. Because the applied oligonucleotide probe contained only the AP-1 motif but not the Ets motif, this must be an indirect effect.
Binding of Ets to the C3aR Promoter in U937 Cells Determined by EMSA-To prove the expression of Ets-1DN in U937 cells, EMSAs were performed using a defined radiolabeled double-stranded oligonucleotide harboring an Ets-1 consensus sequence (Ets-1wt) as the positive control. This oligonucleotide binds preferentially Ets-1 (36). Incubation of this probe with nuclear extract from non-induced U937-Ets-1DN cells resulted in the formation of a prominent band (bold arrow in Fig.  7a) and two weak bands with lower electromobility (the two thin arrows in Fig. 7a). In the nuclear extracts from the non-induced U937-pQCXIN negative control cells, the prominent band was lacking, indicating that this band represents the Ets-1DN-DNA complex (left two lanes). The relative high mobility of this complex is in good accordance with the smaller size of Ets-1DN as compared with Ets-1 or other members of this transcription factor family. Thus, these data confirm a high expression of the dominant negative Ets-1 mutant in transduced U937 cells.
In contrast to the strong band caused by Ets-1DN, the intensity of the weak bands was the same in U937 cells transduced with pQCXIN-Ets-1DN or with the negative control virus (left two lanes of Fig. 7a). These nuclear factor-DNA complexes could reflect the participation of endogenous transcription factors. Indeed, the Ets-1DN-DNA complex as well as the endogenous nuclear factor-DNA complexes bound to the radiolabeled double-stranded Ets-1wt probe could be efficiently competed by a 50-fold excess of nonlabeled double-stranded Ets-1wt oligonucleotide. A double-stranded oligonucleotide probe spanning nucleotides Ϫ196 to Ϫ219 of the C3aR promoter including the predicted Ets site (Ets-C3aRwt) could also compete the three complexes. However, a 500fold excess of nonlabeled Ets-C3aRwt oligonucleotide was needed to achieve complete competition. Proving specificity, an irrelevant negative control oligonucleotide (NC) as well as the corresponding oligonucleotide with a specific mutation within the putative Ets binding element (Ets-C3aRmut) did not compete at all (Fig. 7a).
AP-1 was up-regulated during Bt 2 cAMP induction of C3aR expression. To check whether Ets-1 is similarly up-regulated during C3aR induction, nuclear extracts of U937 cells induced with Bt 2 cAMP (2, 6, 24, or 48 h) were compared with those of buffer-treated cells. In contrast to the results obtained for AP-1, no increase of the radiolabeled nuclear factor-DNA complex could be observed using Ets-1wt as probe (data not shown). Thus, even in U937 cells induced with Bt 2 cAMP, only a small amount of nuclear factor interacting with the Ets-1 consensus binding motif is present. This factor, most likely Ets-1, specifically interacts with the C3aR promoter; however, with relative low affinity. Other members of the Ets family might interact preferentially and with higher affinity with the Ets site in the C3aR promoter. Those factors might be up-regulated in U937 cells during induction instead.
We also tried the reverse experiment, i.e. an EMSA with radiolabeled double-stranded Ets-C3aRwt oligonucleotide as probe. In native or induced U937 cells, almost no nuclear extract-DNA complex binding to this part of the C3aR promoter could be detected (data not shown).

Binding of Ets to the C3aR Promoter in HMC-1 Cells Determined by
EMSA-In HMC-1 cells, different results (compared with U937) were obtained in the EMSA for Ets. Using nuclear extracts of these cells and the double-stranded Ets-C3aRwt oligonucleotide as the radiolabeled probe, one nuclear factor-DNA complex could be detected. It could be effectively competed with an excess of the same nonlabeled oligonucleotide or the Ets-1 consensus oligonucleotide. As proof of specificity, it could neither be competed with an oligonucleotide representing the C3aR promoter but harboring a completely mutated Ets site (Ets-C3aRmut2) nor with the negative control oligonucleotide (NC) (Fig. 7b).
Using nuclear extracts of HMC-1 cells and the double-stranded Ets-1wt oligonucleotide as radiolabeled probe, two strong nuclear factor-DNA complexes became visible. Both bands could be competed with a 50-fold excess of Ets-1wt oligonucleotide or a 500-fold excess of Ets-C3aRwt. However, the complexes could not be competed with an irrelevant oligonucleotide (NC) or the Ets-C3aRmut oligonucleotide (data not shown).
In conclusion, a nuclear factor is present in HMC-1 cells, which binds specifically to the Ets motif in the C3aR promoter. Additionally, an Ets transcription factor can be detected in the nuclear extract of HMC-1 cells, which binds to the Ets-1 motif within the Ets-1 consensus oligonucleotide. Similar to U937 cells, this (endogenous) factor binds with higher affinity to the consensus Ets-1 motif than to the Ets site within the C3aR promoter. The signal intensity of the complex caused by the Ets transcription factor and the C3aR-Ets oligonucleotide is rather weak, suggesting relative low expression of the Ets factor which binds to the C3aR promoter in HMC-1 cells (this is in accordance with the relative low reporter gene activity obtained in HMC-1 cells in the absence of overexpressed Ets-1). Taken together, it could be demonstrated that HMC-1 cells, but not U937 cells, express a transcription factor that interacts with the Ets site within the examined part (Ϫ219 to Ϫ196) of the C3aR promoter.

DISCUSSION
The importance of the regulation of the C3a/C3aR system is indicated by its control at various levels. The activation of the complement system itself, including the generation of the anaphylatoxic peptides, is already tightly controlled. Equally important is the ability to regulate the responsiveness of cells to C3a by the modulation of C3aR surface expression. Under physiological conditions a variety of cells do not express the C3aR. In inflammation, the C3aR is strongly up-regulated. On the other hand, overstimulation and permanent activation of C3aRexpressing cells is restricted by rapid receptor internalization. C3aR internalization has recently been analyzed by us (39,40). In the present study we focused on the transcriptional control of C3aR expression.
Our data obtained with mutated reporter gene constructs indicate that the human C3aR promoter is TATA-less, as already postulated for the murine receptor (11). We identified by nuclease protection assay and RT-PCR a major (Ϫ185 upstream of the ATG start codon) and a minor (ϳϪ280) C3aR transcription initiation site in HMC-1 cells. Multiple start sites are a typical feature of TATA-less promoters, which can often be found in G-protein-coupled receptors, such as the D1A dopamine receptor (41,42). Deletion analysis by reporter gene assays indicated positive regulatory promoter regions in close proximity to the major transcription initiation site. Bioinformatic analysis predicted the presence of various cis-acting elements, among them binding sites for the transcription factors AP-1 and Ets. It should be noted that additional factors participate most likely in the regulation of the C3aR, as indicated by the results of our 3Ј deletion analysis.
For a more detailed investigation, we had to focus on promising candidates. AP-1 and Ets were selected for the following reasons. 1) The results of our 5Ј deletion analysis (Ϫ285 to Ϫ180; Fig. 2c) were in good accordance with the location of the predicted binding motifs. Additionally, the two putative binding sites were in close proximity to each other (a feature frequently observed for these two cis-acting elements), sup-  1 cells (b). a, nuclear extract (NE) from U937 cells either transduced with a retroviral control vector (pQCXIN) or with a retroviral vector expressing a dominant negative mutant of Ets-1 (Ets-1DN) was incubated with a radiolabeled double-stranded oligonucleotide representing an Ets-1 consensus motif ( 32 P-labeled Ets-1wt) as probe. Nuclear factor-DNA complexes were competed with a 5-, 50-, and 500-fold excess of nonlabeled double-stranded oligonucleotides as indicated: Ets-1wt, representing an Ets-1 consensus binding motif (identical to the probe); NC, an irrelevant negative control oligonucleotide; Ets-C3aRwt, part of the C3aR promoter including an Ets binding motif centered at position Ϫ207; Ets-C3aRmut representing the same region of the C3aR promoter; however, with mutated Ets binding motif. The two thin arrows at the right side of the figure indicate the position of the endogenous nuclear factor(s) which specifically bind to the Ets binding element. The bold arrow indicates the position of the highly expressed dominant negative mutant Ets-1DN. b, nuclear extract from HMC-1 cells was incubated with 32 P-labeled Ets-C3aRwt as probe. Nuclear factor-DNA complexes were competed with a 500-fold excess of nonlabeled double-stranded oligonucleotide, as indicated: Ets-C3aRwt, Ets-1wt, Ets-C3aRmut2 (completely mutated Ets site), and nuclear extract. In both EMSAs the oligonucleotides exhibiting the Ets motif (Ets-1wt and Ets-C3aRwt) could compete efficiently. The oligonucleotide exhibiting a mutated Ets consensus motif could not compete, demonstrating the specificity of the interaction. The results of one of two independent experiments are depicted. DECEMBER 23, 2005 • VOLUME 280 • NUMBER 51 porting the credibility of the prediction (38,43,44). 2) The two transcriptional regulators were also selected because of their biological features; AP-1 plays a central role in cell biology and can be activated by a variety of stimuli such as cytokines, growth factors, serum, and oncogenes (for review, see Refs. 45 and 46). It also plays a critical role in the assembly of the preinitiation complex within TATA-less promoters (47). The Ets transcription factor family is implicated in cellular proliferation, differentiation, migration, apoptosis, and cell-cell-interaction. Ets family members are upstream effectors of signal transduction pathways. Their function is controlled by phosphorylation. All members of the Ets family are defined by a highly conserved DNA binding domain that interacts with the core consensus sequence GGA(A/T) (48). Ets-1 is one of the best characterized factors of this family. Its expression pattern partially resembles that of the C3aR. This receptor is constitutively expressed in mature mast cells which also express Ets-1 (49). Ets-1 regulates the maturation and survival of B-and T-lymphocytes (50); C3aR can also be found on the activated lymphocytes. Ets-1 is a central regulator for proliferating vascular endothelial cells. Thus, it is controlling angiogenesis in inflammation caused by rheumatoid arthritis (51), a disease where anaphylatoxins play a prominent role. Our aim was to clarify whether the C3aR expression is also influenced by the regulatory key players AP-1 and Ets.

AP-1 and Ets as Regulators of C3a Receptor Expression
The results of the reporter gene assay confirmed the assumption that AP-1 and Ets-1 can interact with the predicted binding motifs of the C3aR promoter in HMC-1 cells augmenting its activity. As expected, higher SEAP activity was achieved when Ha-Ras was coexpressed. This suggests that the Ras-mitogen-activated protein kinase pathway might be involved in C3aR regulation. Overexpressed AP-1 or Ets-1 was not active on reporter gene constructs exhibiting a deletion in the corresponding binding motif. Furthermore, compared with the wild type C3aR promoter, the activity of overexpressed Ets-1 was diminished when the AP-1 site was missing and vice versa. The highest SEAP production was measured when AP-1 and Ets-1 were simultaneously coexpressed. In conclusion, the observations obtained in HMC-1 cells strongly suggest that the two transcription factors can cooperate directly in C3aR up-regulation. Moreover, C3aR-mRNA levels increased after coexpression of AP-1 and Ets-1 in HEK293 cells, proving that these transcription factors up-regulate C3aR transcription.
The ability of nuclear factors to bind to the consensus AP-1 element within the C3aR promoter was demonstrated by EMSAs. AP-1 was up-regulated in U937 cells induced with Bt 2 cAMP, a treatment that leads simultaneously to the expression of the C3aR. Using the part of the C3aR promoter harboring the AP-1 site or a commercial AP-1 consensus oligonucleotide as radiolabeled probe, identical patterns of nuclear factor-DNA complexes were observed. They could be efficiently competed by an excess of nonlabeled doubled-stranded AP-1 consensus or C3aR promoter oligonucleotide but not with the corresponding mutated oligonucleotides. Hence, it was established that AP-1 interacts specifically with the C3aR promoter. Our data confirm the exact location of the AP-1 cis-acting element within this promoter. Moreover, the results strongly suggest that AP-1 is responsible for C3aR up-regulation in U937 cells.
Similar results were obtained with nuclear extracts from HMC-1 cells. Unlike U937 cells, only one AP-1-DNA complex with a different electrophoretic mobility became apparent. AP-1 is a homo-or heterodimer composed of members of the Jun, Fos, and activating transcription factor families. Differential expression of AP-1 proteins modulates its activity. Thus, most likely AP-1 complexes consisting of different subunits of this transcription factor family interact with the C3aR promoter in HMC-1 or U937 cells.
To further elucidate the role of Ets in C3aR-regulation, the dominant negative mutant of Ets-1, Ets-1DN, was transduced by a retroviral gene transfer and expression system into U937 cells. The mutant abrogated Bt 2 cAMP-induced C3aR up-regulation on the mRNA and protein levels as well as function in these cells. In contrast, C5aR expression and function (as well as the function of the closely related fMLP receptor) were only slightly diminished. Thus, our data indicate that the dominant negative Ets-1 mutant suppresses preferentially and with relative specificity Bt 2 cAMP-induced C3aR up-regulation. It should be noted that this experiment does not show whether the affected Ets transcription factor acts directly or indirectly on the C3aR promoter. However, it clearly proves that a member of the Ets family plays a central role in C3aR induction in monocytes-like U937 cells. However, this endogenous regulator inhibited by Ets-1DN is probably not Ets-1 itself because this factor was not up-regulated during receptor induction by Bt 2 cAMP. Furthermore, the EMSA of Bt 2 cAMP-induced U937 cells using the Ets-1 consensus oligonucleotide as probe resulted (independent of Ets-1DN expression) in two weak bands representing nuclear factor-DNA complexes. The affinity of the cis-acting element within the C3aR promoter to Ets-1 was just sufficient to compete in a 500-fold excess with the radiolabeled Ets-1 consensus probe. No bands were detected in the reverse experiment when the C3aR promoter oligonucleotide harboring the Ets site served, now in much smaller amounts, as probe, indicating that the affinity of this factor might be too low to obtain a band shift under these conditions. These results indicate that Ets-1 is a possible but not the ideal binding partner for the analyzed Ets site within the C3aR promoter. This is in accordance with the observation that highly overexpressed Ets-1 can activate the C3aR promoter, as demonstrated by reporter gene assay. The Ets family consists of more than 50 different members (52,53), making it extremely difficult to identify the postulated superior/additional binding partner(s) in U973 cells.
As described above, our data indicate that AP-1 plays a central role in the Bt 2 cAMP-induced up-regulation of the C3aR in U937 cells. Ets-1DN can block C3aR up-regulation almost completely. Therefore, we checked whether the activation of AP-1 is influenced by the inhibition of endogenous Ets factors that are functionally suppressed by the dominant negative mutant. And indeed, the binding of AP-1 to the C3aR promoter in Bt 2 cAMP-stimulated U937 cells was reduced (by ϳ50%) due to Ets-1DN expression as demonstrated by EMSA, indicating an indirect effect of Ets transcription factors in C3aR regulation. However, it was also possible to show in HMC-1 cells a direct effect of Ets on the C3aR promoter; EMSA confirmed the presence of transcription factors capable of specific binding to the Ets motif centered at position Ϫ207 of the C3aR promoter. These factors might cooperate with AP-1 in the human mast cell line, as demonstrated by reporter gene assays. The transcription factors that bound to the AP-1 motif within the C3aR promoter showed a different electromobility in HMC-1 and U937 cells, indicating that different members of the AP-1 family participate in the two cell types. This might explain that the up-regulation of the C3aR in U937 cells can occur in the absence of Ets, indicating that considerable differences in the details of C3aR regulation exist in constitutively expressing and inducible cells.
In conclusion, AP-1 and members of the Ets transcription factor family play a critical role in the transcriptional regulation of the C3aR. AP-1 interacts directly with the C3aR promoter. Depending on the cell type, a direct interaction of Ets with the C3aR promoter could be shown. Additionally, Ets can modify AP-1 activation and, thus, also indirectly effects C3aR expression. An essential component of the inflammatory response is the regulation of anaphylatoxin receptors such as the C3aR.