Molecular analysis of the bovine anaphylatoxin C5a receptor

Abstract

Recruitment of phagocytes to inflammatory sites involves the coordinated action of several chemoattractants, including the anaphylatoxin C5a. While the C5a receptor (C5aR) has been well characterized in humans and rodents, little is known about the bovine C5aR. Here, we report cloning of bovine C5R1, the gene encoding bovine C5aR. We also analyzed genomic sequence upstream of the C5R1 translation start site. Although the bovine C5aR amino acid sequence was well conserved among species, significant differences in conserved features were found, including major differences in the N terminus, intracellular loop 3, and transmembrane domain VII. Analysis of C5aR expression by flow cytometry and confocal microscopy demonstrated high levels of C5aR on all bovine neutrophils and a subset of bovine monocytes. C5aR was not expressed on resting or activated bovine lymphocytes, although C5aR message was present in these cells. C5aR was also expressed on a small subset of bovine mammary epithelial cells. Pharmacological analysis of bovine C5aR-mediated responses showed that bovine C5a and C5a_desArg both induced dose-dependent calcium fluxes and chemotaxis in bovine neutrophils, with similar efficacy for both agonists. Treatment of bovine neutrophils with C5a or C5a_desArg resulted in homologous desensitization of bovine C5aR and cross-desensitization to interleukin 8 (IL-8) and platelet-activating factor (PAF); whereas, treatment with IL-8 or PAF did not cross-desensitize the cells to C5a or C5a_desArg. Overall, these studies provide important information regarding distinct structural and functional features that may contribute to the unique pharmacological properties of bovine C5aR.

INTRODUCTION

Although multiple factors contribute to host defense against infection, the first line of defense generally involves the innate immune system and initiation of an inflammatory response [1]. Specific defense mechanisms include neutralization of toxins, phagocytosis and killing of bacteria, and direct killing of the pathogen by cytotoxic cells [2]. Among these mechanisms, the complement system plays a central role in the inflammatory process and microbial killing [3]. Complement contributes to the host response by opsonizing pathogens, recruiting phagocytes, and priming or activating cells for pathogen ingestion and killing [4]. Activation of the complement system results in production of proinflammatory fragments C3a, C4a, and C5a, which are anaphylatoxins [5]. These anaphylatoxins represent endogenous danger signals that induce inflammatory responses and modulate innate immune cell functions [3, 6]. Although anaphylatoxins play an important role in host defense, they are also involved in a variety of acute and chronic inflammatory diseases. For example, sepsis is associated with excessive production of C5a, leading to compromised innate immune functions [7].

C5a is a 74-77 amino acid peptide, depending on species, which is cleaved from C5 during complement activation [8]. In humans and rats, C5a is glycosylated; whereas, the bovine, porcine, and murine forms lack this post-translational modification [9, 10]. Under normal conditions, the C-terminal arginine of C5a is cleaved by serum carboxypeptidase N, resulting in the formation of C5a_desArg [11]. While C5a is a potent granulocyte activator, the effect of C5a_desArg varies among species. For example, C5a_desArg is 20-50 times less active than C5a for human neutrophils [12] and 10 times less active for teleost granulocytes [13]. In contrast, both C5a and C5a_desArg are equally potent or only slightly less potent activators of bovine and rat neutrophils, respectively [9, 10]. The reasons for these differences are currently unknown but may relate to differences in ligand and/or receptor structure.

The effects of C5a are mediated through binding to a specific G protein-coupled C5a receptor designated as C5aR [14]. C5aR is one of several peptide chemoattractant receptors that are coupled to Pertussis toxin-sensitive G_i-like G proteins and, in humans, includes N-formyl peptide receptors and interleukin 8 (IL-8) receptors [15]. In human phagocytes, the N-formyl peptide receptor plays an important role in inflammation, coordinating with the C5a and IL-8 receptors and cross-desensitizing their respective signaling pathways [16]. In contrast, bovine cells do not respond to N-formyl peptides and apparently do not express the corresponding receptor [17]. Thus, in the absence of a receptor for N-formyl peptides, C5aR may play a more dominant role in modulating inflammatory responses in the bovine system. In support of this idea, C5a is the first inflammatory stimulus to be induced during bovine mastitis, followed temporally by IL-8, tumor necrosis factor α (TNF-α), and IL-1β [18, 19]. Thus, we propose that a better understanding of the bovine C5aR would not only provide clues to its unique pharmacological responses but would also lead to a better understanding of the role of the C5a/C5aR system in bovine inflammatory responses.

In the present study, we cloned the bovine C5aR, evaluated expression, and characterized pharmacological responses to C5a and C5a_desArg, including an analysis of C5aR cross-desensitization. Molecular and biochemical characterization of this key bovine receptor provides an important advance in our understanding of its role in host defense processes in cattle.

MATERIALS AND METHODS

Materials

Histopaque 1077 and 1119 and Escherichia coli K-235 lipopolysaccharide (LPS) were purchased from Sigma Chemical (St. Louis, MO, USA). TRIzol reagent and Hank’s balanced salt solution without (HBSS⁻) and with (HBSS⁺) Ca²⁺ and Mg²⁺ were from Invitrogen (Carlsbad, CA, USA). DNA-free DNase was from Ambion (Austin, TX, USA). Pfu turbo DNA polymerase was from Stratagene (La Jolla, CA, USA). Human recombinant IL-8 was from CalBiochem (San Diego, CA, USA). Human recombinant TNF-α was from Fitzgerald Industries International (Concord, MA, USA). Goat anti-mouse IgG (H⁺L) Fab fragment-fluorescent label (FITC or PE) was obtained from Jackson ImmunoResearch (West Grove, PA, USA). All of the oligonucleotide PCR primers were obtained from Integrated DNA Technologies (Coralville, IA, USA).

Animal care and blood collection

All animal use was conducted in accordance with a protocol approved by the Institutional Animal Care and Use Committee at Montana State University. Holstein bull (Bos taurus) calves, aged 4 to 9 mo, were obtained from local private herds and housed in our approved large animal care facility. Blood was obtained by jugular venous puncture and collected into 20-ml siliconized Vacutainer tubes containing 5 mM EDTA.

Leukocyte isolation

Total bovine leukocytes were isolated from anticoagulated bovine blood by hypotonic lysis of erythrocytes, followed by washing with Dulbecco’s phosphate-buffered saline (DPBS; GIBCO BRL, Grand Island, NY), as described previously [20]. Purified leukocytes were then centrifuged at 800 g for 10 min, and the pelleted cells were used to isolate total RNA or resuspended in HBSS⁻ at a concentration of 10⁷/ml for flow cytometric analysis.

For neutrophil purification, bovine blood was centrifuged, and buffy coats were collected. After hypotonic lysis of red blood cells, the leukocytes were separated on two-step Histopaque gradients, as described previously [21]. Isolated neutrophils were washed twice and resuspended in HBSS⁻. Neutrophil preparations were routinely >95% pure, as determined by Wright stain and light microscopy, and >98% viable, as determined by Trypan blue exclusion.

Peripheral blood mononuclear cells (PBMCs) were collected from the gradient interface between Histopaque and buffer. In some experiments, cells were washed with HBSS and analyzed by flow cytometry as described below. For analysis of activated lymphocytes, the cells were resuspended in RPMI 1640 medium supplemented with 2 mM L-glutamine, nonessential amino acids, 1% Penn/Strep solution (Invitrogen), and 10% fetal bovine serum (FBS) and aliquotted into 6-well Costar tissue culture plates at 1-2×10⁶ cells/well. Control, untreated cells, or cells treated with 2 μg/ml phytohemagglutinin (PHA) were incubated at 37°C and 10% CO₂. At 1, 2, 3, and 4 days, cells were collected and analyzed by flow cytometry, as described below.

Bovine PBMCs were fractionated by high-speed cell sorting on a Vantage SE cell sorter (BD Biosciences, San Jose, CA, USA) to obtain >95% pure peripheral blood αβ and γδ T cell populations, as described by Graff et al. [22].

Cell culture

The established bovine mammary epithelial cell line (MAC-T) [23] was obtained from Nexia Biotechnologies, Inc. (Montreal, Quebec, Canada) and cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% heat-inactivated FBS, 2 mM L-glutamine, 1% Pen/Strep solution, and bovine insulin (10 μg/ml) at 37°C and 5% CO₂. Confluent cells were removed from tissue culture flasks by gently scraping pipetting to disrupt clumps and filtered through a Nitex membrane. The cells were analyzed by flow cytometry as described below.

The SV40-transformed bovine macrophage cell line (BoMac) [24] was cultured in RPMI-1640 (GIBCO BRL, Gaithersburg, MD, USA) supplemented with 5% heat-inactivated FBS and 1% Penn/Strep solution at 37°C and 5% CO₂.

Cloning and sequencing bovine C5aR

Bovine C5R1 (also known as C5aR1) was cloned from a bovine bone marrow cDNA library. Primer sequences used for the PCR were based on transmembrane domain (TM)-III (forward primer: 5′-TCATCCTGCTCAACATGTAC-3′) and TM-IV (reverse primer: 5′-TGGTAGGGCAACCAGAAGAT-3′) of the human C5a receptor [25]. The conditions for the C5R1 PCR reaction were 95°C for 1 min, 6 cycles of 95°C for 1 min, 50°C for 1 min, 72°C for 1 min, followed by 30 cycles of 95°C for 1 min, 55°C for 30 s, and 72°C for 1 min. The PCR product was purified on 1% agarose gels, TA cloned into pGEM-Teasy (Promega, Madison, WI, USA), and sequenced using an ABI 377 DNA Sequencer (ABI, Foster City, CA, USA). The sequence obtained from this 450-bp PCR fragment was then used in the design of primers to obtain the 5′ and 3′ sequences by rapid amplification of cDNA ends (RACE; SMART RACE cDNA Amplification Kit, Clontech, Palo Alto, CA, USA).

Bovine bone marrow RNA was extracted using TRIzol reagent and treated with DNA-free DNase prior to cDNA synthesis. 5′- and 3′-RACE-ready cDNA was synthesized with 3.6 μg of bovine bone marrow RNA, as described by the manufacturer, and 2.5 μl of the RACE-ready cDNA was used in the RACE reaction. The 5′-RACE gene-specific primer sequence was 5′-TGGTAGGGCAACCAGAAGATAAAGAAGCT-3′, and the 3′-RACE gene-specific primer sequence was 5′-CGATGTGTGTCGTCGCATACGGTAA-3′. Because distinct bands were not observed after separation of the PCR products on 1% agarose gels, a nested PCR reaction was performed using a nested gene-specific primer and the nested universal primer provided in the SMART RACE kit. The final products were then purified on agarose gels, TA cloned into pGEM-Teasy, and at least five clones from three independent PCR experiments were sequenced to obtain the complete bovine C5aR sequence, which was deposited in GenBank (see text for accession numbers).

On the basis of our RACE sequences, full-length bovine C5R1 was cloned by RT-PCR from bovine bone marrow total RNA using a gene-specific sense primer spanning the start site (5′-GATGGACTCCATGGCTCTCA-3′) and an antisense primer from the 3′-untranslated region (5′-GAAGCAGGACAAGTCCGAGAA-3′). Reverse transcription was performed using the SuperScript III One-Step RT-PCR System with Platinum Taq (Invitrogen), and the PCR was performed using Pfu turbo DNA polymerase (Stratagene, La Jolla, CA, USA), according to the manufacturer’s instructions. The PCR product was gel purified, TA cloned into pGEM-Teasy, and sequenced in both directions to confirm no errors were introduced by PCR. The full-length C5R1 cDNA was then subcloned into pcDNA 3.1 (Invitrogen) and sequenced again to verify correct orientation and error-free sequence.

The genomic sequence of bovine C5R1 was obtained using its cDNA sequence as a query for the bovine genome database. The genomic sequence was confirmed to contain the C5R1 coding sequence and analyzed for a TATA box using Promoter Scan [26]. The genomic sequence upstream of the bovine C5R1 translation initiation codon was analyzed for putative transcription factor binding sites using the Transcription Element Search System (TESS) [Schug, J. and Overton, G.C. (1997) TESS: Transcription Element Search Software on the Web, Technical Report CBIL-TR-1997-1001-v0.0, Computational Biology and Informatics Laboratory, School of Medicine, University of Pennsylvania, URL: http://www.cbil.upenn.edu/tess].

Semi-quantitative RT-PCR

Total RNA extracted from bovine leukocytes and cell lines using an RNeasy kit (Qiagen, Valencia, CA, USA), and semiquantitative PCR was performed using the SuperScript III one-Step RT-PCR System with Platinum Taq (Invitrogen), as suggested by the manufacturer. Approximately 120-200 ng of RNA were routinely used in each reaction. The conditions for C5R1 RT-PCR reactions were 50°C for 30 min, 94°C for 2 min, and 25-35 cycles of (94°C for 15 s, 55°C for 10 s, 68°C for 30 s), as indicated. Conditions for 18S PCR reactions were: 45°C for 30 min, 94°C for 2 min, and 12-17 cycles of (94°C for 15 s, 50°C for 10 s, 68°C for 30 s), as indicated. Sequences of PCR primer sets were as follows: bovine C5R1, forward 5′-GAGGATCTGGGAGAGACCTAAAGATGG-3′ and reverse 5′-AGGATGAGAGAGGGCAGGAT-3′; and bovine 18S, forward 5′-TCGAGGCCCTGTAATTGGAA-3′ and reverse 5′-CCCAAGATCCAACTACGAGCTT-3′. Ten microliters of each reaction were separated on 1.5% agarose gels and stained with ethidium bromide for visualization. Identity of the PCR products was confirmed by DNA sequencing using an ABI 310 Genetic Analyzer.

Alignment of C5aR protein sequences and phylogenetic tree construction

The amino acid sequences of all known C5aR were aligned with AlignX (Vector NTI Advance, Invitrogen) using the Clustal W algorithm [27]. The phylogenetic tree constructed using AlignX was based on a sequence distance method that used the Neighbor Joining (NJ) algorithm of Saitou and Nei [28].

Preparation of anti-bovine C5aR antibodies

A synthetic peptide corresponding to residues 8-26 of the bovine C5aR was synthesized with an added cysteine at the C terminus (NH₂-TPDYSDYDKWTSNPDVLVDC-COOH) and coupled to keyhole limpet hemocyanin (KLH) (Macromolecular Resources, Fort Collins, CO, USA). The peptide-KLH conjugate was used to immunize rabbits for polyclonal antibodies and mice for monoclonal antibodies. Monoclonal antibodies were obtained through fusion of the hyperimmune splenocytes with the myeloma line Sp2/0, as described previously [29]. The fusion was screened by ELISA against C5aR peptide conjugated to ovalbumin and by flow cytometry with bovine neutrophils. Positive clones were subcloned by limiting dilution and rescreened by flow cytometry and immunohistochemistry, and one clone (3D5) was selected. The isotype of 3D5 was found to be IgG1_κ using an Isostrip antibody isotyping kit (Roche Diagnostics, Indianapolis, IN, USA). Monoclonal 3D5 and polyclonal antibodies were purified from hybridoma supernatant and antiserum, respectively, using Protein G Sepharose 4 Fast Flow beads (GE Healthcare, Piscataway, NJ, USA).

Transient expression of bovine C5aR in Drosophila melanogaster S2 cells

Bovine C5aR cDNA was subcloned into the inducible pMT/V5-His C expression vector (Invitrogen, Carlsbad, CA, USA), and the resulting plasmid was sequenced to confirm orientation and absence of mutations. S2 cells were transiently transfected with empty vector or vector containing bovine C5aR following the manufacturer’s instructions. At log phase, 500-μM copper sulfate was added to induce protein expression, and the cells were harvested after induction for 24 h. The cells were washed, lysed with cell lysis buffer, and lysates were analyzed by immunoblotting, as described below.

Preparation of bovine C5a and C5a_desArg

Bovine C5a and C5a_desArg were purified from yeast-activated bovine serum, as described by Gennaro et al. [9] and Rainard et al [30], respectively, with slight modifications. Bovine serum was incubated for 45 min (C5a preparation) or 120 min (C5a_desArg preparation) at 37°C with 10 g/l boiled and washed yeast in the presence (C5a preparation) or absence (C5a_desArg preparation) of 1 mM DL-2-mercaptomethyl-3-guanidinoethylthiopropanoic acid (Calbiochem, San Diego, CA, USA), which inhibits serum carboxypeptidase N. The activated serum was adjusted to pH 5.5 (for C5a) or 6.0 (for C5a_desArg) with 1.0 M sodium acetate, centrifuged to remove the yeast (2500 g for 15 min), and mixed overnight at 4°C with Sephadex C-25 (Pharmacia, Piscataway, NJ, USA) equilibrated in buffer A (150 mM NaCl, 50 mM sodium acetate, pH 6.0) at 100 ml of settled slurry/l of serum. The resin was washed extensively with buffer A, loaded into an Econo-Column (Bio-Rad, Richmond, CA, USA), and eluted with 800 mM NaCl, 50 mM sodium acetate, pH 6.0 at 1 ml/min, while monitoring A₂₈₀ and collecting 1-ml fractions. Fractions were analyzed for their ability to elicit a calcium flux in bovine neutrophils (described below) and those with highest activity were combined, diluted 1:10 with H₂O, applied to a MonoS column equilibrated with 100 mM sodium formate, pH 5.0, washed with the same buffer, and eluted with a gradient of 0-1 M NaCl in 100 mM sodium formate, pH 5.0 at 1 ml/min while collecting 1-ml fractions. Fractions were analyzed for protein content using the BCA method (Pierce, Rockford, IL, USA), for their ability to induce a calcium flux, and for C5a/ C5a_desArg by immunoblotting with monoclonal anti-C5a/C5a_desArg [30]. The peak fractions were aliquotted and stored at –70°C.

Preparation of FITC-C5a_desArg

Bovine C5a_desArg was labeled with fluorescein isothiocyanate (FITC) following methods described by Van Epps and Chenoweth [31]. Briefly, purified bovine C5a_desArg was diluted in 0.1 M sodium phosphate, pH 7.0, concentrated to 1.6 mg/ml using Microcon YM-10 concentrators (Millipore, Bedford, MA, USA), and incubated for 45 min at 25°C with a three-fold molar excess of FITC. The reaction was terminated by gel filtration, and the labeled C5a_desArg was lyophilized. Stock solutions of FITC-C5a_desArg were prepared by dissolving in H₂O, and protein content was determined to estimate molar concentration for use in flow cytometric analysis.

Flow cytometry

One- and two-color flow cytometric analyses were performed, as described previously [32], with modifications. For single-color analysis of C5aR expression on total leukocytes, bovine blood was diluted (1:4) with ice-cold DPBS containing 5% heat-inactivated fetal calf serum and 7.5 mM azide (DPBS⁺), and the cells were pelleted by centrifugation. In some cases, cells were activated prior to leukocyte isolation by treatment of whole blood for 30 min at 37°C with 50 ng/ml TNF-α, 100 ng/ml IL-8, or 20 μg/ml LPS. The reactions were quenched by a 1:5 dilution into ice-cold DPBS⁺. After hypotonic lysis of erythrocytes, the leukocytes were incubated for 1 h at 4°C with primary monoclonal antibody 3D5, monoclonal anti-bovine neutrophil antibody BN15.6 [32], or monoclonal anti-bovine CD11b/CD18 antibody MM12a (VMRD, Inc., Pullman, WA, USA). After labeling, the cells were washed and resuspended in DPBS⁺-containing Alexa 488-conjugated goat anti-mouse IgG (1:1000 dilution). Control cells were stained with secondary antibody only in the primary incubation. After incubation for 1 h at 4°C, the cells were washed, resuspended in DPBS⁺, and analyzed by flow cytometry with a FACScan or FACSCalibur with CellQuest software (BD Biosciences, San Jose, CA, USA).

For two-color flow cytometric analysis, cells isolated as above were incubated with antibody 3D5 for 1 h on ice, washed with DPBS⁺, incubated for 1 h on ice with Alexa 488-conjugated goat anti-mouse IgG (1:1,000 dilution), washed with DPBS⁺, and incubated with Alexa 594-conjugated antibody BN15.6 or FITC-conjugated mAb BN1.80 for 1 h on ice. BN15.6 labeling was performed with an Alexa Fluor 594 Monoclonal Antibody Labeling Kit (Molecular Probes, Eugene, OR, USA). The double-labeled cells were washed, resuspended in DPBS⁺, and analyzed by flow cytometry. A total of 10,000 events were collected for all samples.

For flow cytometric two-color analysis of MAC-T cells, cells were labeled with antibody 3D5 (1.5 μg/ml) for 1 h on ice, washed with DPBS⁺, incubated for 1 h on ice with phycoerythrin (PE)-conjugated goat anti-mouse IgG (1:250 dilution), washed with DPBS⁺, and incubated with 50 nM FITC-C5a_desArg for 1 h on ice. The double-labeled cells were washed, resuspended in DPBS⁺, and analyzed by flow cytometry. A total of 10,000 events were collected for all samples.

To confirm specificity, mAb 3D5 was preincubated for 1 h at 37°C with excess (∼100-fold) C5aR peptide antigen conjugated to ovalbumin or ovalbumin alone (control) and then used for staining, as described above.

Analysis of calcium flux

Changes in intracellular Ca²⁺ flux [Ca²⁺]_i were measured with a FlexStation II scanning fluorometer after labeling the cells with a FLIPR 3 Calcium Assay Kit (Molecular Devices, Sunnyvale, CA), as described previously [33]. Briefly, purified bovine neutrophils were resuspended in HBSS⁻ and loaded with FLIPR Calcium 3 dye following the manufacturer’s protocol. After dye loading, Ca²⁺ was added to the cell suspension (1.26 mM final), and 100 μl of cell suspension were aliquotted into the wells of a black flat-bottommed microtiter plate (2×10⁵ cells/well). After monitoring background fluorescence (emission/excitation), agonists [IL-8, C5a, C5a-_desArg, and platelet-activating factor (PAF)] were automatically pipetted from the compound source plate, and fluorescence was monitored for 4-8 min with readings every 5 s. The maximum change in fluorescence, expressed in arbitrary units, over baseline was used to determine agonist response. Median effective concentrations (EC₅₀) were determined by nonlinear regression analysis of the dose-response curves generated using GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA).

Desensitization assays were performed with the Flexstation and FLIPR 3 assay system, as described above. Initially, dose-response curves were generated for C5a, C5a_desArg, IL-8, and PAF to determine EC₁₀₀ concentrations for bovine neutrophils, as described above. The EC₁₀₀ concentrations were found to be 1 μM for C5a, C5a_desArg, and IL-8, and 1 nM for PAF. Using these concentrations, we first pretreated the cells with control buffer or EC₁₀₀ concentrations of C5a, C5a_desArg, IL-8, or PAF for 4 min at 37°C. Agonist-pretreated cells were then treated with a range of concentrations of the same agonist to evaluate homologous desensitization or a different agonist to evaluate heterologous cross-desensitization (12-14 different concentrations of each agonist were tested). Buffer-pretreated cells served as controls to evaluate these responses in the absence of agonist pretreatment. Fluorescence was monitored for an additional 4 min after the second addition, and EC₅₀ values were determined by nonlinear regression analysis of the dose-response curves generated for the second agonist added. The data are plotted as relative fluorescence induced by the EC₅₀ concentration determined for a given agent in cells pretreated with buffer or agonist.

Chemotaxis assay

Bovine neutrophil chemotaxis was analyzed in 96-well ChemoTx chemotaxis chambers (Neuroprobe, Gaithersburg, MD, USA). Lower wells were loaded with 30 μl of HBSS⁺ containing 0.1% bovine serum albumin (BSA) and the indicated concentrations of agonist. Neutrophils suspended in HBSS⁺ containing 0.1% BSA (2 × 10⁶ cells/ml) were added to the upper wells and allowed to migrate through a 5.0-μm pore polycarbonate membrane filter for 60 min at 37°C and 5% CO₂. After incubation, the upper wells were gently aspirated from the filter, 2.5 mM EDTA was added to the upper wells for 5 min, and the wells were rinsed again to remove remaining cells. The plates were centrifuged (800 g) for 5 min to dislodge any cells adherent to the underside of the filter, and the filter was removed. The number of migrated cells was determined by measuring ATP in lysates of transmigrated cells using a luminescence-based assay (CellTiter-Glo; Promega). Luminescence measurements were converted to absolute cell numbers by comparison of the values to standard curves created by aliquotting known numbers of neutrophils into the lower wells of a ChemoTx plate and processing then identically to the test sample wells (average r²>0.99 for standard curves). The results are expressed as the number of migrated cells minus background response to control medium.

Immunocytochemistry

Aliquots (10 ml) of bovine blood were incubated for 15 min in a shaking water bath at 37°C without or with 20 μg/ml LPS. The tubes were then placed on ice, 40 ml of ice-cold DBPS⁺ was added, and the cells were pelleted by centrifugation. After lysis of erythrocytes, the cells were resuspended in DPBS⁺, aliquotted into Eppendorf tubes, and stained with mAb 3D5 or mAb 3D5 and mAb BN15.6, as described above under flow cytometry. In some experiments, cells were also labeled with the Vybrant Lipid Raft Labeling Kit (Molecular Probes) to label lipid rafts with Alexa Fluor 594-conjugated cholera toxin B, following the kit protocols. Controls included unstained cells and cells stained with secondary antibody alone. After staining, the cells were washed with DPBS⁺, fixed with 1% paraformaldehyde in DPBS⁺, and analyzed by confocal microscopy using a Zeiss LSM 510 Meta confocal microscope (Carl Zeiss, Inc., Thornwood, NY, USA).

Electrophoresis and immunoblot analyses

SDS-PAGE using 7-18% polyacrylamide gradient gels and Western blot analysis were performed as described previously [34]. Transfers were blotted with primary antibody for 3 h at 25°C, followed by alkaline phosphatase- or horseradish peroxidase-conjugated goat anti-rabbit secondary antibodies (Bio-Rad, Hercules, CA, USA). Blots were developed using a Bio-Rad alkaline phosphatase development kit or chemiluminescence, respectively.

RESULTS

Cloning and sequencing of bovine C5aR cDNA

To clone bovine C5R1, oligonucleotide primers were synthesized to TM3 and TM5 of the human C5aR, as these domains are functionally conserved between many G protein-coupled receptors (GPCR) [35]. These primers were used to amplify a 450-bp fragment of the bovine C5R1 cDNA from a bovine bone marrow cDNA library, and this fragment provided sequence information for completion of the bovine C5R1 cDNA cloning by 5′- and 3′-RACE. The final cloned cDNA insert was 1478 nucleotides long, containing an open reading frame of 1056 bases that encoded 352 amino acids (GenBank accession no. AY540054). The predicted molecular mass and pI of bovine C5aR are 38.99 kDa and 9.06, respectively, and the predicted hydrophobic regions of bovine C5aR, based on Kyte-Doolittle hydropathy analysis [36], were consistent with the presence of seven transmembrane domains [35] (data not shown). A comparative alignment between the bovine C5aR and C5aR sequences from other species demonstrated the bovine sequence was very similar to sheep and pig C5aR (93 and 74% identity, respectively), somewhat less similar to human and primate C5aR (70–71%) and rodents (61–64%), and the least similar to trout C5aR (35%) (Fig. 1 and Supplemental Table S1). Phylogenetic analysis of bovine C5aR showed that it falls within a clade containing equine, porcine, and canine C5aR (Fig. 2A). This group was distinct from the clades containing human and primate C5aR, rodent C5aR, and teleost C5aR. While the bovine C5a receptor is fairly well conserved among species, we noted some unique and significant differences in this species that may contribute to functional differences (Fig. 1). For example, sequence of the bovine C5aR N-terminus is quite divergent from that of other species, and bovine C5aR lacks a site for N-linked glycosylation, which is present in the N-terminus of human (Asn5) and many other species′ C5aR.

Fig. 1.

Alignment of bovine C5aR with other known C5aR amino acid sequences. Sequences were obtained from NCBI and were aligned using AlignX, which is based on the Clustal W algorithm (Vector NTI Advance, Invitrogen, Carlsbad, CA). Residues identical to bovine C5aR are shaded in gray, and residues identical in all species’ C5aR are also indicated in bold. The seven putative transmembrane (TM) domains are indicated by boxes. N-terminal aspartic acid residues important for ligand binding (*), conserved cysteine residues involved in intramolecular disulfide bridging (♦), conserved phosphorylation sites in the C-terminus (▾), and residues known to be important for human C5aR structure/function but not conserved in bovine C5aR (•) are indicated (see Discussion for further details). Accession numbers for the sequences used are as follows: bovine, AY540054; sheep, AAG12475; dog, CAA46690; pig, ABP01833 and AAG12474; human, CAA40530; gorilla, CAA66317; orangutan, CAA66316; rhesus, P79188; rabbit, AAF13030; guinea pig, AAC40074; gerbil, AAP50850; mouse, AAP50849; rat, NP_446071; and trout, AAR12188.

Fig. 2.

Phylogenetic analysis of the bovine C5aR and genomic organization of bovine C5R1. (A) Dendrogram showing relative sequence similarities between all known C5aR sequences. Calculated distances from the closest branch are shown in parentheses. (B) Genomic organization of bovine C5R1 showing the size of all exons and introns. Putative splicing is shown.

Using our cloned bovine C5R1 cDNA sequence, we screened the bovine genome database, which was released subsequent to the cloning, and obtained the C5R1 genomic sequence, which is located on chromosome 18, locus C5R1. The genomic sequence is characterized by a three-exon structure, with the initiating methionine in exon 1 and the remainder of the coding sequence in exon 2, whereas exon 3 contains 3′-untranslated sequence and poly-A signal (Fig. 2B). On the basis of the sequence of 12 different 5′-RACE products, it appears that exon 1 is ∼55-57 bp long, with the translational start site located at the 3′ end. One kilobase of genomic sequence upstream of the bovine C5R1 translation start site was also analyzed for the presence of cis regulatory elements to gain insight into potential regulatory factors for this gene. This region did not contain a TATA box, which is present in the human and murine C5R1 promoters but did contain consensus binding sites for CCAAT binding factor, transcription factor II-D and II-I (TF-II-D, TF-II-I), Sp1, and Core binding factor (see Supplemental Figure S1). In addition, consensus binding sites were present for transcription factors important in myeloid cell function and inflammatory responses, including serum response factor (SRF), Elk-1, CCAAT/enhancer binding protein α (C/EBPα), activating protein-1 (AP-1), PU.1, GATA-1, NF-κB, and interleukin 6 responsive element binding protein (IL-6 RE-BP) (see Supplemental Figure S1).

Evaluation of C5aR expression

To evaluate expression of bovine C5aR, we prepared monoclonal and polyclonal antibodies against a peptide corresponding to residues 8-26 of bovine C5aR. Rabbit polyclonal antibody R4394 recognized a protein with average M_r of ∼37 kDa by immunoblotting, and staining was completely blocked by the antigenic peptide to verify specificity (Fig. 3A). Although anti-C5aR monoclonal antibody 3D5 did not work for immunoblotting, it stained native C5aR on bovine leukocytes. Flow cytometric analysis of bovine leukocytes stained with mAb 3D5 demonstrated that C5aR was expressed on all bovine neutrophils (Fig. 3B). In confirmation of staining specificity, inclusion of excess ovalbumin-antigenic peptide conjugate completely blocked staining of bovine neutrophils by mAb 3D5 in flow cytometric analysis, whereas, no loss in staining was observed when ovalbumin alone was added (Fig. 3B). Likewise, analysis of mAb 3D5 staining by confocal microscopy showed mAb 3D5 stained bovine leukocyte plasma membranes, as indicated by colocalization with cholera toxin B (CTB), which stains plasma membrane lipid rafts, and addition of excess ovalbumin-antigenic peptide conjugate completely blocked staining by mAb 3D5 but had no effect on staining with CTB (Supplemental Figure S2). Furthermore, analysis of S2 cells transiently transfected with bovine C5aR showed that mAb 3D5 stained only cells transfected with C5aR (confirmed by immunoblotting), while no staining was found in cells transfected with the empty pMT/V5-His vector (Supplemental Figure S3). Thus, mAb 3D5 is specific for bovine c5aR and does not appear to nonspecifically stain any other antigens on intact cells or in cell lysates. In addition, bovine C5aR may be localized in lipid raft regions; however, this specific issue will require further analysis in future studies.

Fig. 3.

Immunoblot and Flow cytometric analysis of bovine neutrophil C5aR. (A) Bovine neutrophil lysates were separated by SDS-PAGE, transferred to nitrocellulose, and immunoblotted with anti-bovine C5aR polyclonal antibody R5767. Specificity of antibody staining was confirmed using 20 μg/ml of the antigenic peptide as a blocking agent (+ peptide). The locations of prestained molecular weight standards (STD) are indicated. A representative blot is shown from 3 independent experiments. (B) Isolated bovine blood leukocytes were stained with mAb 3D5, followed by an Alexa 488-conjugated goat anti-mouse IgG. Fluorescence histograms are shown for control cells stained with secondary antibody alone (2° Control) and cells stained with mAb 3D5. To confirm specificity, mAb 3D5 was preincubated for 1 h at 37°C with excess (∼100-fold) C5aR peptide antigen conjugated to ovalbumin (3D5+Ova-peptide) or control ovalbumin alone (3D5+Ova) and then used for flow cytometry. A representative experiment is shown from 3 independent experiments using different animals.

Although resting bovine neutrophils expressed high levels of C5aR on their cell surface, we considered whether this expression could be up-regulated even more by activation in vitro. Interestingly, C5aR expression only changed modestly when the cells were activated in vitro with a variety of agents, including LPS (Fig. 4), TNF-α, or IL-8 (data not shown). However, activation-induced changes in C5aR expression were often inconsistent between cells isolated from different calves or from the same calf on different days (average change in C5aR expression was 1.1±0.1-fold with a range from 0.1-fold decrease to a maximum of 1.3-fold increase from nine different experiments with five different animals). In contrast, activation consistently resulted in significantly increased staining for mAb BN15.6 (average of 12.1±3.4-fold increase from five different experiments), a surface antigen that is up-regulated on activated bovine neutrophils [32], and mAb MM12a (average of 1.4±0.1-fold increase from 5 different experiments), which represents activation-dependent up-regulation of bovine CD11b/CD18 expression (Fig. 4).

Fig. 4.

Analysis of changes in C5aR expression in activated bovine neutrophils. Bovine whole blood was treated with 20 μg/ml LPS for 30 min at 37°C, as described. After lysis of the red blood cells, the leukocytes were stained with mAbs 3D5, BN15.6, or MM12a (CD11b/CD18), followed by Alexa 488-conjugated goat anti-mouse IgG. Mean fluorescence intensity measured for gated neutrophils, which were identified by their distinctive forward and side light scatter profiles, is shown. Control cells were stained with secondary antibody alone (2° Control). The data are expressed as mean ± se of five independent experiments using different animals. Statistically significant differences (*, P<0.02; **, P<.004) between control and LPS-treated cells are indicated.

Two-color staining of activated bovine neutrophils with mAbs 3D5 and BN15.6 confirmed that C5aR was expressed on all bovine neutrophils, as indicated by costaining of all mAb BN15.6-positive cells with mAb 3D5 (Fig. 5A and Supplemental Figure S4), and this finding was confirmed by confocal microscopy (Supplemental Figure S5). In comparison, only low-level staining for C5aR was observed on bovine monocytes, as demonstrated by two-color staining of PBMC preparations with mAbs 3D5 and BN1.80, which is specific for bovine monocytes (Fig. 5B and Supplemental Figure S6). No staining for mAb 3D5 was present on bovine lymphocytes, indicating this receptor is not expressed on the surface of resting bovine lymphocytes (Supplemental Figure S6), although RT-PCR analysis of bovine αβ and γδ T cells purified by cell sorting [22] showed that the C5aR message was also present in these cells (Fig. 6). Nevertheless, treatment of bovine lymphocytes with PHA for up to 96 h still did not induce C5aR protein expression (data not shown).

Fig. 5.

Two-color flow cytometric analysis of C5aR expression on bovine neutrophils and monocytes. (A) Bovine blood leukocytes isolated from LPS-treated bovine blood were analyzed by two-color flow cytometry using Alexa Fluor 594-conjugated mAb BN15.6 (bovine neutrophils) vs. mAb 3D5 indirectly stained with a Alexa 488-conjugated goat anti-mouse IgG (bovine C5aR). Plots for gated neutrophils, which were identified by their distinctive forward and side light scatter profiles, are shown. The upper level of background staining seen with negative controls is indicated by the quadrant markers. Staining of mAb 3D5 and mAb BN15.6 alone are shown in Supplemental Figure S4. The percentage of gated cells present in each quadrant is indicated. (B) Purified bovine peripheral blood mononuclear cells (lymphocytes and monocytes) were analyzed by two-color flow cytometry using FITC-conjugated mAb BN1.80 (bovine monocytes) vs. mAb 3D5 indirectly stained with Alexa 488-conjugated goat anti-mouse IgG (bovine C5aR). The upper level of background staining seen with negative controls is indicated by the quadrant markers. Staining of mAb 3D5 and mAb BN1.80 alone are shown in Supplemental Figure S6. The % of gated cells present in each quadrant is indicated. For both panels, a representative experiment is shown from 3 independent experiments using different animals.

Fig. 6.

Semi-quantitative RT-PCR analysis of bovine C5aR message in bovine leukocytes and cell lines. Total RNA extracted from purified bovine neutrophils, αβ T cells, γδ T cells, MAC-T cells, and BoMac cells was subjected to RT-PCR for the indicated number of cycles, and the products were separated on agarose gels and stained, as described. Control samples contained no input RNA. Identity of the PCR products was confirmed by DNA sequencing. Representative gels from 2–3 independent experiments.

Because of the importance of C5a in bovine mammary gland defense and indirect evidence, suggesting mammary epithelial cells respond to C5a (reviewed in [4]), we evaluated expression of C5aR message and protein in cultured bovine mammary epithelial cells (MAC-T cells). C5aR message was detected in MAC-T cells, although expression was very low compared with that found in bovine leukocytes (Fig. 6). Likewise, very low expression was also observed in a bovine macrophage cell line (BoMac) [24] (Fig. 6). Analysis of MAC-T cells by flow cytometry showed that a small subpopulation of MAC-T cells (∼10-12%) stained with mAb 3D5 or FITC-labeled C5a_desArg, confirming the presence of C5aR protein on these cells (Supplemental Figure S7). Almost all MAC-T cells stained with mAb 3D5 were colabeled with FITC-labeled C5a_desArg, confirming that antibody staining was specific for C5aR on these cells (Supplemental Figure S7).

Pharmacological features of the bovine C5aR

Previously, Gennaro et al. [9] reported that bovine C5a and C5a_desArg were equally effective in eliciting bovine neutrophil chemotaxis and specific granule release. To further investigate this issue, we evaluated calcium flux in bovine neutrophils treated with purified bovine C5a and C5a_desArg and found that both induced a dose-dependent increase in [Ca²⁺]_i (Fig. 7). Although the dose-response curves varied from calf to calf and day to day with the same calves, the average EC₅₀ concentration for C5a (69 nM; n=5; range 9-150 nM) was only about half that of C5a_desArg (160 nM; n=5; range 16-330 nM). The ability of C5a and C5a_desArg to induce a similar [Ca²⁺]_i flux is consistent with the similar efficacy of these agonists to induce downstream responses in bovine neutrophils, such as degranulation and chemotaxis.

Fig. 7.

Analysis of calcium mobilization in bovine neutrophils activated with C5a and C5a_desArg. Purified bovine neutrophils were loaded with FLIPR Calcium 3 dye, treated with the indicated doses of C5a (○) or C5a_desArg (▪), and Ca²⁺ was monitored, as described. The data are expressed as mean ± se, n = 3. A representative experiment is shown from at least 4 independent experiments using different animals.

To confirm previous observations regarding comparable efficacy of bovine C5a and C5a_desArg in stimulating bovine neutrophil chemotaxis, we evaluated chemotactic responses to these two agonists, as well as human C5a for comparison. As shown in Fig. 8A, bovine C5a and C5a_desArg induced dose-dependent chemotaxis of bovine neutrophils, and the responses were quite similar. Note, however, that the response to C5a was slightly higher at lower concentrations of chemoattractant. In contrast, human C5a failed to stimulate chemotaxis of bovine neutrophils across the entire concentration range tested (Fig. 8B).

Fig. 8.

Analysis of bovine neutrophil chemotaxis toward C5a and C5a_desArg. Bovine neutrophil chemotaxis was analyzed in 96-well chemotaxis chambers, as described. (A) Comparison of responses toward bovine C5a and C5a_desArg. (B) Comparison of responses toward bovine and human C5a. The results are expressed as number of migrated cells minus background response to control medium and represent the mean ± se; n = 3. A representative experiment is shown from 3 independent experiments using different animals.

To evaluate the interplay between bovine C5aR and other inflammatory receptors, we evaluated desensitization and cross-desensitization of bovine C5aR, as described by Richardson et al. [37] for human neutrophils. Treatment of bovine neutrophils with EC₁₀₀ doses of C5a or C5a_desArg (not shown) resulted in homologous desensitization to subsequent treatment with C5a or C5a_desArg (Fig. 9A, B). Likewise, treatment of bovine neutrophils with EC₁₀₀ doses of C5a or C5a_desArg (not shown) resulted in cross-desensitization to subsequent treatment with IL-8 and platelet-activating factor (Fig. 9C, D). Conversely, treatment of bovine neutrophils with EC₁₀₀ doses of IL-8 did not desensitize the cells to subsequent activation by C5a or C5a_desArg (Fig. 9E, F). These results are consistent with those reported for human neutrophil chemoattractant receptor cross-desensitization [37, 38].

Fig. 9.

Analysis of C5aR desensitization and cross-desensitization by IL-8 and PAF. Purified bovine neutrophils were loaded with FLIPR Calcium 3 dye and pretreated with control buffer (open or hatched bars) or EC₁₀₀ concentrations of the indicated agonists (solid bars) for 4 min at 37°C. The cells were then treated with control buffer (Control) or a range of concentrations of C5a, C5a_desArg, IL-8, or PAF to determine their EC₅₀ values in control and pretreated cells. Relative fluorescence induced by the EC₅₀ concentration determined for a given agent in cells pretreated with buffer (open or hatched bars) or agonist (solid bars) is shown. The data are expressed mean ± se of triplicate samples, and a representative experiment is shown from 3 independent experiments using different animals. Statistically significant differences (*, P<0.001) between agonist alone and agonist after desensitization are indicated.

DISCUSSION

The C5a/C5aR system has been heavily characterized in humans and rodent model systems (e.g., see [6, 39]); however, much less is known about this system in other species. In bovines specifically, very little is known about this system outside of the various responses observed when bovine leukocytes are exposed to C5a. Note, however, C5a plays a significant role in defense of the bovine mammary gland against infection [4] and may be an important inflammatory mediator in bovine mastitis [40, 41]. Thus, it is essential to expand our understanding of the bovine C5a/C5aR system in order to address issues related to inflammatory tissue damage associated with mammary inflammation and mastitis.

In the present study, we report cloning and characterization of the bovine C5aR. The bovine C5R1 gene is characterized by a three-exon structure; whereas, human and murine C5R1 genes contain two exons. In all three cases, the initiating methionine is separated from the rest of the coding sequence by a large intron, which is similar to other chemoattractant receptor genes [39]. Note, however, that the chromosome location varies widely between species (e.g., human, murine, and canine C5R1 genes are located on chromosomes 19, 7, and 1, respectively). Analysis of genomic sequence upstream of the bovine C5R1 ATG start site demonstrated the presence of consensus binding sites for several key myeloid transcription factors. For example, four sites were found for the myeloid-specific transcription factor PU.1 [42], suggesting an important role for this factor in regulating C5aR expression. A consensus site for C/EBPα is also present, and it has been shown that C/EBPα induces PU.1 and interacts with AP-1 and NF-κB during myeloid cell development [43]. The presence of response elements for IL-6 and interferon are consistent with the role of these cytokines in regulating differentiation and activation of myeloid cells (e.g., [44, 45]). Finally, the presence of an NF-κB binding site suggests a role for this important inflammation-associated transcription factor [46] in regulating bovine C5aR expression. Further analysis is needed to determine the functionality of these putative transcription factor binding sites with respect to C5R1 regulation.

Comparison of the bovine C5aR amino acid sequence with that of the other species’ C5aR showed an average of 67.5% homology. Phylogenetic analysis showed ovine and porcine C5aR exhibited the highest degree of homology (93 and 76%, respectively) and were closely grouped with bovine C5aR. Although lower in homology, canine C5aR was also grouped in the same clade as bovine C5aR. It is interesting that bovine, porcine, ovine, and canine leukocytes do not respond to N-formyl peptides and apparently do not express receptors for these peptides [17, 47,48,49,50,51,52]. Therefore, these species express only one of the two end target chemoattractant receptors that are found in most other species [53, 54]. The close homology of these species’ C5aR may reflect evolution of uniquely regulated C5aR systems to compensate for the absence of the N-formyl peptide receptor pathway.

Analysis of the bovine C5aR sequence showed that this receptor retained many of the amino acids reported to be absolutely required for its structure and ligand binding; however, there are distinct differences that may provide clues relative to various structural and functional features of these proteins. For example, the N-terminal sequence is quite distinct from that of other species and lacks a putative site for N-linked glycosylation (N-X-T/S), which is present in many other C5aR. Nevertheless, aspartic acid residues reported to be important for tethering of C5a to the receptor are present in the bovine C5aR (Asp10, 15, and 27). Because the extracellular N-terminal domain has been shown to contain one of the C5a binding sites, the high divergence in this region among C5aR suggests the possibility that this domain may confer species specificity for C5a binding. Indeed, human C5a failed to stimulate bovine neutrophil chemotaxis. Previously, Postma et al. [55] showed that residues 10-18 within the human C5aR N-terminus represented a binding site for the chemotaxis inhibitor protein Staphylococcus aureus (CHIPS) and that mutations in residues Asp10, Gly12, Asp15, and Asp18 significantly reduced CHIPS binding. In comparison, Asp18 is substituted by Trp18 in the bovine C5aR. Since mutation of only one of these residues completely abolishes CHIPS binding [55], it is likely that bovine phagocytes are not susceptible to CHIPS, and this may facilitate bovine responses against S. aureus infection. However, further studies are necessary to evaluate this issue.

In general, the transmembrane (TM) regions are fairly well conserved between bovine C5aR and other species’ C5aR, whereas more divergence is present in the connecting loops (Fig. 1). Among the notable differences, bovine C5aR contains an insertion of two basic residues in the N-terminus before TM1 (His32, Arg33), and these residues are absent from most other species’ C5aR. In intracellular loop 3, two highly conserved threonine residues are replaced by alanine (Ala231, Ala240), and these residues have been shown to be important for G protein coupling and activation of phospholipase C [56]. Note that both of these residues are also alanines in ovine C5aR. The structure of extracellular loop 3 has been shown to be important for activating G proteins, and mutation of human C5aR Pro270 resulted in receptors that have substantially reduced C5a binding. In bovine C5aR, the corresponding residue is replaced by valine (Val272). This residue is also divergent in ovine, porcine, and canine C5aR, although such substitutions are seen in a few other C5aR as well. In TMVII, a highly conserved serine is replaced by alanine (Ala285) in bovine C5aR, and this substitution is also present in ovine and porcine C5aR. Overall, it appears that differences in a number of key individual amino acid residues contribute to the unique pharmacological properties of bovine C5aR, and many of these changes are shared by closely related C5aR from other species.

Previous studies showed that both C5a and C5a_desArg are similarly effective at activating bovine neutrophil chemotaxis and degranulation [9], whereas, human C5a_desArg is 10- to 100-fold less potent than human C5a in various biological assay systems (reviewed in [39]). Here, we found that bovine C5a and C5a_desArg were also both potent activators of Ca²⁺ mobilization in bovine neutrophils, although C5a_desArg was slightly less potent than C5a (∼2-fold). Likewise, bovine C5a and C5a_desArg stimulated comparable chemotactic responses. Since chemotaxis and certain other neutrophil responses do not require a rise in [Ca²⁺]_i (reviewed in [57]), the twofold difference in efficacy of C5a and C5a_desArg for inducing Ca²⁺ mobilization is not reflected in the chemotactic responses. The reason for a similar efficacy of these two agonists in bovine cells is still not clear. Bovine C5a is not glycosylated [9], and it has been suggested that the lack of C5a glycosylation enhances potency of the desArg form [58]. For example, deglycosylation of human C5a_desArg enhanced chemotactic activity 12- to 15-fold, while no change in activity was observed when human C5a was deglycosylated, and the authors concluded that the presence of glycosylation decreases C5a potency only when there is a concurrent loss of the C-terminal arginine and that the presence of this charged residue overcomes any inhibitory effects of the sugar [58]. On the other hand, rat C5a_desArg, which is glycosylated, is only fourfold less potent than rat C5a and 50-fold more potent than human C5a_desArg [10]. Thus, the C5a glycosylation status likely plays a role but does not fully explain the similar efficacy of bovine C5a and C5a_desArg. The absence of N-terminal glycosylation of bovine C5aR may also contribute to its agonist binding characteristics; although, previous studies on the human C5aR showed that receptor deglycosylation had little effect on C5a binding [59]. This system is not completely analogous, however, since both bovine C5aR and C5a are not glycosylated. Thus, it is still possible that binding in the complete absence of glycosylation on both ligand and receptor, as is the case for bovine C5a and C5aR, could be different than when one or both moieties are glycosylated. In addition to glycosylation, several residues in human C5aR have been shown to be involved in C5a/C5a_desArg binding, including Asp82, Pro103, Gly105, Arg175, Glu199, Arg206, and Asp282 (reviewed in [39]). These residues are all conserved in bovine C5aR, with the exception of Pro103, which is a serine in the bovine receptor. Mutation of Pro103 to Tyr in human C5aR resulted in a higher affinity and a supra-maximal response to C5a but decreased maximal response to C5a_desArg [60]. Thus, the high potency of bovine C5a_desArg cannot be explained by changes in any of the individual residues currently mapped for the C5aR ligand binding site. Regardless of the specific mechanisms, both C5a and C5a_desArg can be considered as effective inflammatory mediators in the bovine system, and this may compensate, in part, for the absence of an N-formyl peptide system.

The C5aR was highly expressed on all bovine neutrophils, and activation of these cells ex vivo with various stimuli only slightly increased the level of C5aR detected. Thus, resting bovine neutrophils appear to be fully responsive to the presence of C5a/C5a_desArg, and priming is not necessary to up-regulate receptor expression for a rapid response. Note also that studies on human neutrophils have shown that only a portion of the total C5aR expressed are required to stimulate a full functional response [61]. While this issue has not been evaluated in bovine neutrophils, a similar response pattern is likely. Overall, the pattern of C5aR expression would suggest bovine cells are highly sensitive to the presence of C5a/C5a_desArg, and this enhanced sensitivity may reflect compensation for absence of N-formyl peptide chemoattractant receptors. Note, however, that these studies do not rule out the possibility that the level of bovine neutrophil C5aR expression could be modulated in vivo during infection, and further work is clearly warranted to address this issue.

In addition to neutrophils, C5aR was expressed on ∼60% of bovine monocytes, which is similar to that observed in human monocytes [31, 62]. Expression of C5aR was not detected on resting or activated bovine T lymphocytes, although it is clear that the message for this receptor is expressed in both αβ and γδ T cells. These findings are consistent with those of Soruri et al. [63] who showed murine T and B lymphocytes did not express C5aR. In contrast, there are several reports that indicate human lymphocyte subsets may express low levels of C5aR, which can be up-regulated by mitogens [64, 65]. Note, however, that these results are controversial, and it has been suggested that these data need to be re-evaluated because of reagent and/or technical issues [39, 63, 66]. Thus, it appears that the C5aR protein is not normally expressed on bovine lymphocytes; however, our findings do not rule out the possibility that expression can be induced under specific conditions that have not yet been evaluated. C5aR was also expressed on a small subpopulation (∼12%) of bovine MAC-T cells, which is consistent with the presence of very low levels of C5aR message in these cells. The expression of C5aR on epithelial cells is also controversial, and it has been suggested that C5aR may be induced in cultured primary cells while expression of C5aR is absent on the corresponding cells in vivo [67, 68]. Thus, further work is clearly necessary to determine the physiological relevance of MAC-T staining to bovine mammary epithelial cell function in vivo.

Recruitment of phagocytes to sites of infection involves the coordinated action of a number of chemoattractants (reviewed in [69]). In the past decade, a number of groups have shown that this process involves interplay between chemoattractant receptors and that a signaling hierarchy determines direction of migration in complex chemotactic gradients [53, 54, 70]. This hierarchy involves preferential migration toward end-target chemoattractants (fMet-Leu-Phe and C5a), which can override responses to intermediary host-derived chemoattractants, such as IL-8, PAF, and leukotriene B₄ [53, 54, 70]. While bovines lack one of the primary end target chemoattractant pathways, the remaining hierarchy seems to be intact. Here, we show that the bovine C5aR undergoes homologous desensitization by C5a and C5a_desArg, but is not cross-desensitized by IL-8. On the other hand, exposure of cells to C5a/C5a_desArg results in unidirectional heterologous desensitization of IL-8 and PAF receptors, which is similar to that found in human neutrophils [71,72,73]. Because bovine cells contain only one of the two primary end-target chemoattractant systems, the C5a/C5aR must play a more dominant role in regulating bovine phagocyte recruitment and activation. Thus, bovine neutrophils can still respond to C5a/C5a_desArg even after pre-exposure to IL-8 or PAF. This would also be important in intramammary infection, where both IL-8 and C5a are generated. Indeed, low levels of C5a and the absence of subsequent cytokine production have been reported to contribute to the establishment of chronic mastitis due to S. aureus infection [19].

In summary, cloning and molecular analysis of bovine C5aR demonstrated that this receptor is highly expressed on bovine neutrophils, moderately expressed on bovine monocytes, and not expressed on bovine lymphocytes. Although the bovine C5aR sequence is homologous to that of other species’ C5aR, unique differences are present that may contribute to the observed functional differences, such as the similar efficacy of C5a and C5a_desArg to induce intracellular Ca²⁺ flux and chemotaxis. Finally, activation of bovine C5aR unidirectionally cross-desensitized bovine neutrophils to IL-8 and PAF, which is consistent with the primary importance of the C5a/C5aR system in a species lacking N-formyl peptide receptors.