Isolation , Characterization , and Antimicrobial Properties of Bovine Oligosaccharide-binding Protein A MICROBICIDAL GRANULE PROTEIN OF EOSINOPHILS AND NEUTROPHILS *

Peptidoglycan recognition proteins (PGRPs) constitute a recently characterized family of pattern-recognition molecules that are conserved from insects to humans and are implicated in mammalian innate immunity. Here we report the isolation, characterization, cDNA cloning, and antimicrobial activities of a bovine PGRP ortholog termed bovine oligosaccharidebinding protein (bOBP). Milligram quantities of bOBP were purified from peripheral leukocytes, thus allowing for the characterization of the disulfide array and for determining the in vitro antimicrobial activities of the native protein. Of the tissues analyzed, bOBP mRNA was detected only in bone marrow where the protein is synthesized as a 190 amino acid precursor. The mature 169 amino acid protein is stored in the cytoplasmic granules of neutrophils and eosinophils but is absent from lymphocytes, monocytes, and platelets. bOBP was microbicidal for Gram-positive and Gram-negative bacteria and yeast at low micromolar concentrations. The finding that bOBP was microbicidal for organisms in which peptidoglycan is absent (Cryptococcus neoformans) or buried (Salmonella typhimurium) indicates that previous conclusions about the specificity of peptidoglycan recognition proteins must be reevaluated and suggests that other envelope components may mediate the antimicrobial action of PGRP family members.

Peptidoglycan recognition proteins (PGRPs) constitute a recently characterized family of pattern-recognition molecules that are conserved from insects to humans and are implicated in mammalian innate immunity. Here we report the isolation, characterization, cDNA cloning, and antimicrobial activities of a bovine PGRP ortholog termed bovine oligosaccharidebinding protein (bOBP). Milligram quantities of bOBP were purified from peripheral leukocytes, thus allowing for the characterization of the disulfide array and for determining the in vitro antimicrobial activities of the native protein. Of the tissues analyzed, bOBP mRNA was detected only in bone marrow where the protein is synthesized as a 190 amino acid precursor. The mature 169 amino acid protein is stored in the cytoplasmic granules of neutrophils and eosinophils but is absent from lymphocytes, monocytes, and platelets. bOBP was microbicidal for Gram-positive and Gram-negative bacteria and yeast at low micromolar concentrations. The finding that bOBP was microbicidal for organisms in which peptidoglycan is absent (Cryptococcus neoformans) or buried (Salmonella typhimurium) indicates that previous conclusions about the specificity of peptidoglycan recognition proteins must be reevaluated and suggests that other envelope components may mediate the antimicrobial action of PGRP family members.
Studies over the past two decades have revealed that antimicrobial innate immunity plays a critical role in host defense mediated in part by pattern-recognition proteins and microbicidal effector molecules that contain infection prior to an adaptive immune response (1). Pattern-recognition proteins distinguish invading microbes by identifying conserved microbial structures that are not expressed by host cells (2). Microbial killing is mediated by granulocytes and macrophages that produce oxygen-and nitrogen-derived toxins (3,4) as well as microbicidal proteins and peptides (1).
Among the pattern-recognition proteins of innate immunity are lipopolysaccharide (LPS) 1 -binding protein and bactericidal/ permeability-increasing protein, molecules that bind LPS by amino-terminal structures conserved between the two proteins (5,6). Bactericidal/permeability-increasing protein is potently bactericidal for Gram-negative bacteria (7,8) and neutralizes the inflammatory properties of LPS (9). In contrast, LPS-binding protein is not bactericidal and activates an inflammatory response in myeloid cells upon LPS binding (10,11). Toll-like receptors (TLRs) conserved from insects to humans (12) are key elements of innate immunity (13,14). More than ten mammalian TLRs have been identified, each of which may have pattern-recognition functions (15). For example, TLR2 recognizes Gram-positive bacteria by binding peptidoglycan or lipoteichoic acid (15,16), TLR4 identifies Gram-negative bacteria through LPS interactions (15,17,18), TLR5 recognizes flagellin (15,19), and TLR9 identifies bacterial (CpG)DNA (15,20). TLRligand complexes signal the induction of inflammatory cytokines and chemokines. Mannose-binding protein is a soluble pattern-recognition protein found in the serum of all mammals studied (21). A collectin, mannose-binding protein activates complement (21) and opsonizes microbes (22) by binding to carbohydrate moieties of bacterial and fungal cell envelopes. Humans deficient in circulating mannose-binding protein demonstrate increased vulnerability to a number of microbial pathogens (21,23).
Peptidoglycan recognition proteins (PGRPs) are recently characterized components of innate immunity, which are reported to bind specifically to peptidoglycan moieties. These studies were the basis for concluding that PGRPs selectively contribute to innate immunity by detecting Gram-positive bacteria (24,25). Silkworm PGRP, the first member of the family, was implicated in the innate immune response by its ability to trigger the prophenoloxidase cascade in the presence of peptidoglycan or its glycan moiety (26,27). Orthologous proteins have been found in other lepidopteran insects (28 -30) and in Drosophila (31) wherein PGRP gene expression is up-regulated in larvae inoculated with Gram-negative bacteria (25,28,31).
Mammalian PGRPs have been reported in humans (24,26,28), mice (26,28), rats (GenBank TM accession number AF154114), and camels (GenBank TM accession number AJ131676). PGRP mRNAs have been localized to human and mouse bone marrow and polymorphonuclear leukocytes (26, 28) but not to lymphocytes or monocytes (26). Genomic analysis has determined that there are four human genes (24) and at least eight Drosophila genes (31) that have significant PGRP homology. Three of the human genes and three of the Drosophila genes predict products with transmembrane domains (24,31).
In this study, we isolated several milligrams of a bovine PGRP, which we have termed oligosaccharide-binding protein because it was found to kill microorganisms in a peptidoglycanindependent manner. The precursor and mature protein structures, including the conformation of the tri-disulfide array, were determined. Furthermore, tissue expression analysis and immunolocalization studies were performed to ascertain the sites of biosynthesis and the storage of bOBP.

EXPERIMENTAL PROCEDURES
Bovine Granulocytes-Granulocytes were purified from fresh citrated bovine blood (dairy cattle) as described previously (32). Preparations contained an average of 1 ϫ 10 9 cells/liter of whole blood of which 93 Ϯ 3% were neutrophils and 4 Ϯ 1% were eosinophils. Some granulocyte preparations were treated with 2 mM diisopropylfluorophosphate (33) or protease inhibitor mixture tablets (Complete TM EDTA-free, Roche Molecular Biochemicals) prior to protein extraction. Eosinophilenriched populations of granulocytes were obtained by centrifugation over a Percoll continuous density gradient prepared according to the manufacturer's directions (Amersham Biosciences). 4 ϫ 10 8 bovine granulocytes suspended in Hanks' balanced salt solution plus Ca 2ϩ and Mg 2ϩ were layered over the gradient and centrifuged at 7500 ϫ g for 20 min at 4°C from which two distinct bands were formed. The less dense fraction, which contained ϳ35% eosinophils, was used for immunocytochemistry (see below). Cytoplasmic granule preparations were produced by nitrogen cavitation of granulocytes as described previously (34). The cavitate was centrifuged at 700 ϫ g at 4°C for 10 min, and the granule-containing supernatant was collected. Granules were harvested by centrifugation at 27,000 ϫ g at 4°C for 40 min.
Purification of bOBP-derived Peptide-13 (BDP-13)-Cytoplasmic granules derived from diisopropylfluorophosphate-treated leukocytes were extracted with 1 ml of ice-cold 10% acetic acid/2 ϫ 10 9 cell equivalents. After stirring on ice for 18 h, suspensions were clarified by centrifugation at 27,000 ϫ g at 4°C for 20 min. Supernatants were subjected to size exclusion chromatography on a Bio-Gel P-60 column as described previously (32). Low molecular mass components (Ͻ10 kDa) identified by SDS-PAGE were further resolved by C 18 reversed-phase high pressure liquid chromatography (RP)-HPLC, and fractions with antimicrobial activity were collected as described previously (32). BDP-13 among the low molecular weight antimicrobial peptides ob-tained was characterized by automated Edman sequencing on an Applied Biosystems 475A instrument configured with on-line phenylthiohydantoin amino acid analysis (32). The primary sequence was confirmed by comparison with the sequence predicted from the corresponding cDNA (see below) and by biochemical comparison with a synthetic congener (see below). Purified peptide was lyophilized, dissolved in 0.01% acetic acid at 100 -500 g/ml, and stored at Ϫ70°C.
Synthetic BDP-13 (5 mg) was conjugated to 5 mg of chicken ovalbumin using the glutaraldehyde method as described previously (37). The BDP-13-ovalbumin conjugate was suspended in phosphate-buffered saline and used to immunize two New Zealand White rabbits (Zymed Laboratories Inc., South San Francisco, CA). Serum samples were collected after 10 weeks when the specific antibody titer, as determined by enzyme-linked immunosorbent assay, was ϳ1:5000. IgG was prepared from antiserum by DEAE-Econo-Pac chromatography (Bio-Rad) per the manufacturer's instructions.
Isolation and Characterization of bOBP-Acid extracts of 1 ϫ 10 9 cell equivalents of bovine leukocytes or granule preparations were loaded onto a 10 ϫ 25 cm Delta Pak RP-HPLC C 18 cartridge equilibrated in 0.1% trifluoroacetic acid (solvent A) at a flow rate of 10 ml/min. A linear gradient of acetonitrile containing 0.1% trifluoroacetic acid (solvent B) was applied at 2%/min from 0 to 37%, 0.1%/min from 37 to 38%, and 0.167%/min from 38 to 40%. 10-ml fractions were collected, lyophilized, and resuspended in 0.01% acetic acid. Samples containing bOBP were identified by Dot Blot analysis (as described below) using anti-bOBP IgG (as described above), and pure bOBP was obtained following a second round of RP-HPLC. Purity was confirmed by SDS-PAGE, acidurea PAGE, and analytical RP-HPLC. Quantitation and composition analysis of bOBP was determined by amino acid analysis performed on 6 N HCl hydrolysates at 150°C for 2 h using the AccQTag system (Waters, Milford, MA) according to the manufacturer's directions. Subsequent quantification of bOBP was determined spectrophotometrically (1 mg/ml ϭ 1.28 A 280 ). Purified protein was lyophilized, dissolved in 0.01% acetic acid at 2 mg/ml, and stored at Ϫ70°C.
bOBP and trypsin digestion fragments of bOBP (see below) were analyzed by matrix-assisted laser desorption ionization time-of-flight mass spectroscopy (MALDI-TOF MS) on a VOYAGER DE-PRO™ mass Isolation, Characterization, and Antimicrobial Properties of bOBP spectrometer in the linear mode. Samples (ϳ3 pmol each) were mixed with equal volumes of ␣-cyano-4-hydroxycinnamic acid matrix prior to analysis (38).
Partial amino acid sequence was obtained by automated Edman degradation of cyanogen bromide cleavage fragments of bOBP on a Hewlett Packard 1100 instrument (UCI Biotechnology Research Facility, Irvine, CA) (32). The sequencer cartridge loaded with 300 pmol of bOBP was soaked in a solution of 10% cyanogen bromide in 90% formic acid for 2 h at room temperature. The cartridge was then washed with water, and the eluted material was diluted to Ͻ5% formic acid, reapplied to the cartridge, and sequenced. The internal sequences were compared with the amino acid sequence deduced from the bOBP cDNA (as described below).
Disulfide Connectivities-Two nanomoles of bOBP was digested for 42 h at 37°C with 1 g of TPCK-treated trypsin in 50 l of 0.1 M pyridine acetate, pH 6.6. The reaction was terminated by acidification with trifluoroacetic acid. Half of the digest was directly analyzed by RP-HPLC on a 4.6 ϫ 250-mm C 18  hydrochloride, 0.2 M Tris-HCl, 2 mM NaEDTA, pH 8.2, purged with nitrogen, and incubated at 50°C for 30 min. The sample was reduced with 390 M dithiothreitol for 4.5 h at 50°C and alkylated with 3-fold excess iodoacetamide for 10 min in the dark. Excess alkylating reagent was quenched with dithiothreitol. The reduced and alkylated tryptic digest was analyzed by RP-HPLC under the same conditions as the non-reduced digest. A comparison of the two chromatograms identified three disulfide-containing fractions in the untreated digest. These three fractions were characterized by MALDI-TOF MS and automated Edman degradation.
cDNA Cloning-Total RNA was isolated from bovine bone marrow by guanidium thiocyanate-phenol extraction (39). Total RNA (1 g) was used to synthesize cDNA, and 3Ј-RACE was performed according to the manufacturer's protocols (Invitrogen) using a degenerate gene-specific primer, 5Ј-CARCARTGGCCNCAYTAYMG-3Ј. PCR amplification was carried out using the following cycling parameters: 95°C for 1 min; 55°C for 1 min; and 72°C for 1 min for 35 cycles. Subsequently, 1 g of total RNA was subjected to cDNA cloning and 5Ј-RACE according to manufacturer's directions with the gene-specific primer 5Ј-TGGGATGGGTGGGGTGTGAGAAGACGGACAGGCCCTCACACGCGG-3Ј. PCR-amplified RACE products were subcloned and sequenced as described previously (37). Once the 5Ј-and 3Ј-ends of the BDP-13 cDNA were known, a PCR product corresponding to the full-length BDP-13 precursor sequence was generated and characterized by sequence analysis. Sequence data were analyzed using Geneworks 2.5.1 (IntelliGenetics, Mountain View, CA). Amino acid sequences were analyzed using BLAST and Swiss Protein data bases. Protein and peptide theoretical masses were calculated by the ExPASy PeptideMass program (40).
Western Blot and Dot Blot Analyses-Acid extracts of bovine leukocytes, extracts of leukocyte granules, and purified bOBP were lyophilized, boiled for 2 min in dithiothreitol-containing SDS-PAGE sample buffer, subjected to Tricine-SDS-PAGE (12.5% acrylamide), and transferred to 0.22-m nitrocellulose membranes (MSI, Westborough, MA) using the semi-dry electroblotting technique (41). For Dot Blot analyses, RP-HPLC fractions were subjected to rotoevaporation briefly to remove acetonitrile, and they were applied in 1-l aliquots to 0.22-m nitrocellulose membranes. Western blot and Dot Blot membranes were fixed in 0.2% glutaraldehyde, quenched with 1 M NH 4 Cl, and incubated with 1:1000 anti-bOBP IgG or anti-bOBP IgG preabsorbed with BDP-13. Immunoreactive bands or Dots were detected with 1:100,000 goat anti-rabbit IgG conjugated to horseradish peroxidase using Super-Signal West Pico substrate (Pierce). Blots were developed by exposure to x-ray film or using digital imaging (Alpha Innotech, San Leandro, CA).
Immunohistochemistry-Slides of eosinophil-enriched leukocytes were prepared on a cytocentrifuge, fixed in 4% paraformaldehyde for 10 min at 4°C, and stored at 4°C in 70% ethanol. Slides were rehydrated for 5 min in 1ϫ phosphate-buffered saline and treated with Fc receptor blocking reagent (Innovex BioSciences, Richmond, CA) for 20 min and then with 1% normal goat serum for 20 min. Following incubation with 1:40 anti-bOBP IgG for 60 min at room temperature, slides were developed using biotinylated goat anti-rabbit IgG and a Vectastain ABC glucose oxidase kit (Vector Laboratories, Burlingame, CA) according to the manufacturer's instructions. Nuclear Fast Red was used as counterstain. Negative control incubations were performed using preimmune IgG and anti-bOBP IgG preabsorbed with BDP-13.
Northern Blot Analysis-Total RNA from various bovine tissues was extracted and subjected to Northern blot analysis as described previously (42). Gel-purified bOBP cDNA probes were 32 P-labeled by random priming. A mouse GAPDH cDNA probe (Ambion, Austin, TX) was used as a control for RNA loading and integrity. The membrane was stripped and exposed to film to ensure the removal of bOBP probe prior to hybridization with GAPDH cDNA.

RESULTS
Isolation, Characterization, and Synthesis of BDP-13-Approximately 2 ϫ 10 10 cell equivalents of acid-solubilized granule protein was fractionated on a Bio-Gel P-60 column, and the antibacterial activity in pooled eluent fractions was evaluated by radial diffusion assay (32). One broad P-60 peak that contained neutrophil ␤-defensins (32) was further fractionated by RP-HPLC. BDP-13, later identified as the carboxyl-terminal tridecapeptide of bOBP, eluted on C 18 RP-HPLC as a single peak at 23 min (32). After a subsequent round of RP-HPLC under slightly different gradient conditions, pure BDP-13 was obtained which was homogeneous by analytical RP-HPLC and acid-urea PAGE (data not shown). Automated sequence analysis was performed providing unambiguous assignment of the primary sequence of BDP-13, YKIIQQWPHYRRV (Fig. 1, double-underlined sequence). As discussed below, the sequence was confirmed by comparison with the bOBP cDNA. Synthetic BDP-13 was used to produce anti-bOBP antibody (see "Experimental Procedures").
bOBP cDNA-The full-length bOBP cDNA was 688 nucleotides in length and predicted a 190-amino acid precursor protein (Fig. 1). Within the precursor, the sequence of the first 21 residues was characteristic of a signal peptide (45,46), and the predicted signal peptide cleavage site was confirmed by characterization of native bOBP (169 amino acids, see below). BDP-13 corresponded to the 13 carboxyl-terminal residues of bOBP.
A BLAST search of GenBank TM identified bOBP to be a member of the peptidoglycan recognition protein family of proteins. Eight proteins with greatest sequence similarity to bOBP included seven PGRPs and a butterfly molt protein (Fig. 2). Four cysteines are completely conserved, and six cysteines are conserved in the mammalian sequences. The identity of amino acid sequences to bOBP ranged from 37 to 74%.
Isolation and Characterization of bOBP-bOBP was purified from acid extracts of bovine granulocytes and of granule preparations by RP-HPLC fractionation and immunologic detection with anti-bOBP IgG (Fig. 3A). Purified bOBP was homogeneous by RP-HPLC (Fig. 3B), behaved as a single 19-kDa band on SDS-PAGE (Fig. 3C), and had a mass of 18779.6 by MALDI-TOF MS. The mass data are consistent with a polypeptide composed of the 169 carboxyl-terminal residues of the bOBP precursor with additional modifications including the formation of three disulfides and the cyclization of glutamine 22 to form pyroglutamic acid at the amino terminus of the mature protein (theoretical mass ϭ 18785.3). Consistent with this structure, the amino terminus was blocked as revealed by automated sequence analysis. The sequence of a cyanogen-bromide digestion fragment was determined to be GNYMHRVPPASALRAAQSL (Fig. 1, single underline). This sequence was identical to bOBP residues 127-145 predicted by the cDNA. The predicted pI of mature bOBP is 9.38 (47).
The average yield of bOBP from 1 ϫ 10 9 bovine granulocytes (ϳ400 ml of whole blood) was 2.9 mg. In contrast, no more than 2.125 g of BDP-13 was detected in an equivalent number of granulocytes, indicating that the major form of the bOBP gene product in leukocytes is the 169 amino acid molecule.
bOBP Disulfide Motif-RP-HPLC chromatograms of native and reduced/alkylated tryptic digests of bOBP revealed three peptides that were modified by reduction and alkylation, indicating the presence of disulfide bonds. Each of the corresponding peptides obtained from the native digest was analyzed by MALDI-TOF MS and six steps of Edman sequencing (Table I). An analysis of the masses and sequences obtained allowed for the unambiguous assignment of the cysteine connectivities shown in Fig. 4. The resultant motif of 1-6, 2-5, and 3-4 pairings is consistent with the pattern reported for Bombyx mori PGRP that contains two disulfides, those corresponding to the 1-6 and 3-4 pairings in bOBP.
Immunolocalization of bOBP-Immunohistochemical staining of bovine peripheral blood leukocytes with rabbit anti-bOBP IgG demonstrated that bOBP is expressed in eosinophils and neutrophils. Immunoreactivity was strongest in eosinophils, slightly less in neutrophils, but absent from lymphocytes and monocytes (Fig. 5). Extracts of purified lymphocytes and monocytes and of purified platelets were negative for bOBP when subjected to Western blot analysis (data not shown). The immunostaining of bovine neutrophils and eosinophils was punctate and cytoplasmic, consistent with granular storage of bOBP in these cells.
Distribution of bOBP mRNA-Because bOBP was isolated from granulocytes of circulating bovine blood, the detection of 0.9-kb bOBP mRNA in bone marrow extracts was not unexpected (Fig. 6). Six other tissues including liver and spleen (highly populated by myeloid elements), small intestine, kidney, lung, and trachea were analyzed and found to contain no detectable bOBP mRNA ( Fig. 6 and data not shown).
Antimicrobial Activities of BDP-13 and bOBP-The ability of bOBP and BDP-13 to inhibit the growth of microorganisms was determined in radial diffusion inhibition assays. Significant zones of inhibition were observed with 10 M (190 g/ml) bOBP against S. aureus 502a, L. monocytogenes EGD, and E. coli ML35 (Fig. 7). The fungus, C. neoformans 271a, was less sensitive to bOBP than bacteria, but an appreciable zone of clearing was apparent at 30 M (570 g/ml). The diameters of clearing were dose-dependent, demonstrating log-linear relationships in the range tested typical of antimicrobial peptides and proteins. At concentrations of 3-100 M, BDP-13 was in-active against bacteria and weakly inhibited growth of C. neoformans at 30 and 100 M. Neither bOBP nor BDP-13 inhibited the growth of a second yeast, C. albicans 16820, at concentrations of up to 100 M (data not shown).
Antimicrobial suspension assays were used to determine the microbicidal activities of bOBP. A 2-h incubation with 1.3 M (25 g/ml) bOBP induced at least three logs of killing of S. typhimurium SH6497, a polymyxin-resistant strain, and of L. monocytogenes 967 (Fig. 8). S. aureus 502a and L. monocytogenes 10403s were killed by at least three logs by 2.7 M bOBP (50 g/ml), whereas equivalent killing of S. typhimurium GALe required twice as much protein (Fig. 8). E. coli ML35, although inhibited in the diffusion assay (Fig. 7), was resistant to bOBP concentrations of up to 200 g/ml (10.7 M) (data not shown). bOBP killed more than two logs of C. neoformans 271a at 10.7 M (200 g/ml) (Fig. 8). DISCUSSION bOBP is a leukocyte-derived member of the PGRP gene family that is expressed at high levels in circulating neutrophils and eosinophils. The average yield of bOBP was ϳ7 mg/liter of adult bovine blood. The abundance of the protein and the relative ease of purification facilitated the structural and functional studies described herein.
bOBP is the first vertebrate PGRP for which the amino acid sequence (Fig. 2) and tri-disulfide structure (Fig. 4) have been determined. Mass spectroscopic analyses demonstrate that bOBP is not glycosylated and that the amino terminus is a pyroglutamyl residue. The cystine motif is consistent with that of the two-disulfide motif of B. mori PGRP (25). In all PGRPs, one disulfide of 1-6 bridges the amino and carboxyl ends of the chain, and a second disulfide of 3-4 forms a heptapeptide loop, the sequence of which is highly conserved particularly in mammals (Fig. 2). The third disulfide of 2-5 occurs only in mammalian PGRPs, introducing an additional degree of backbone constraint that is absent in the invertebrate orthologs.
An assignment of the PGRP nomenclature to this family of proteins was rationalized by studies indicating that natural PGRP from B mori bound to peptidoglycan (27) and that recombinant PGRP from Trichoplusia ni and from Mus musculus bound well to peptidoglycan and Gram-positive bacteria but poorly to Gram-negative bacteria (26,28). However, our studies FIG. 4. Disulfide connectivities of bOBP. Trypsin-cleavage fragments that contain disulfide bonds are highlighted. Sequences obtained by Edman degradation analysis are underlined (see Table I  demonstrate that Gram-positive and Gram-negative bacteria and at least one fungus are killed by bOBP in the 0.5-10 M range. Antimicrobial activity of a PGRP has been reported on one previous occasion (26), i.e. an analysis of recombinant mouse PGRP. Polyhistidine-tagged murine PGRP inhibited the growth of four Gram-positive bacteria, Bacillus megaterium, Staphylococcus hemolyticus, Staphylococcus warneri, and Staphylococcus capitis, but was not microbicidal (26). The difference between our findings and those published previously may be attributed to differences between the murine and bovine proteins and/or the conditions used in the antimicrobial assays.
We selected the term bOBP for this member of the PGRP family because of the fact that it kills some microorganisms that lack peptidoglycan (C. neoformans) or in which the peptidoglycan is obscured by LPS (e.g. S. typhimurium). Moreover, recombinant mouse PGRP reportedly binds LPS (26). Preliminary studies in our laboratory have shown that bOBP binds to each of the microorganisms that it kills, and it also binds to preparations of microbial antigens including lipopolysaccha-ride and lipoteichoic acid in addition to peptidoglycan. 2 Also, PGRP is reported to interact with the glycan moiety of peptidoglycan (26,27). Thus, we hypothesize that the binding of bOBP to microbial targets is mediated by oligosaccharide components present in each of these cell surface structures, and that the relative binding affinities are dependent on the composition and stereochemistry of the specific sugar residues.
Thus far, we have detected bOBP in eosinophils and neutrophils but not monocytes, lymphocytes (Fig. 5), or platelets (data not shown), and bOBP mRNA was detected only in bone marrow. Immunocytochemical analysis was suggestive of a granular location of bOBP in granulocytes (Fig. 5), and this subcellular localization was further supported by the purification of bOBP from cytoplasmic granule preparations of bovine leukocytes. Despite the stronger immunostaining of bOBP in eosinophils (Fig. 5), quantitative HPLC and immunoblotting assays demonstrated that neutrophils and eosinophils contain equal quantity of bOBP per cell (data not shown). Thus, the difference in immunostaining probably reflects differences in antigen presentation in the cytoplasm of the two granulocyte populations.
To our knowledge, bOBP is the only microbicidal granule protein other than bactericidal/permeability-increasing protein and eosinophil cationic protein that is expressed in eosinophils and neutrophils (48). Largely considered effectors of antiparasitic immunity, eosinophils are not recruited to sites of bacterial infection and they possess little bactericidal activity in vitro. This finding raises the possibility that bOBP may contribute to the antiparasitic activities of the eosinophil.
The data presented here provide further evidence for the role of PGRPs in mammalian innate immunity. Specifically, our studies demonstrated that (i) bOBP is bactericidal and fungicidal in vitro at low micromolar concentrations, (ii) bOBP is expressed in the granules of circulating neutrophils and eosinophils, and (iii) it is an ortholog of proteins implicated in the immune functions in insects (25,27,28). Although it is apparent that bOBP and other PGRPs recognize microbes by pattern recognition, additional studies are needed to identify the mo-FIG. 6. Northern blot of bovine tissues. Expression of bOBP mRNA in bovine tissues was determined by Northern blot analysis. 20 g of RNA from bovine bone marrow (BM), small intestine (SI), trachea (Tr), liver (L), and spleen (Sp) was resolved by electrophoresis in a standard 1% agarose, 6% formaldehyde gel. The membrane was sequentially hybridized under high stringent conditions with the 32 Plabeled full-length bOBP cDNA and mouse GAPDH cDNA.
FIG. 7. Antimicrobial activities of bOBP and BDP-13. Agar radial diffusion assays were performed using varied protein concentrations. Activity is expressed as the diameter of clearing of duplicate samples resulting from the application of 5 l of BDP-13 (q) or bOBP (E) against each organism at the indicated concentrations.