Cloning and characterization of a novel mannose-binding protein of Acanthamoeba.

Acanthamoebae produce a painful, blinding infection of the cornea. The mannose-binding protein (MBP) of Acanthamoeba is thought to play a key role in the pathogenesis of the infection by mediating the adhesion of parasites to the host cells. We describe here the isolation and molecular cloning of Acanthamoeba MBP. The MBP was isolated by chromatography on the mannose affinity gel. Gel filtration experiments revealed that the Acanthamoeba lectin is a approximately 400-kDa protein that is constituted of multiple 130-kDa subunits. Cloning and sequencing experiments indicated that the Acanthamoeba MBP gene is composed of 6 exons and 5 introns that span 3.6 kb of the amoeba genome and that MBP cDNA codes for a precursor protein of 833 amino acids. That the cloned cDNA encodes authentic MBP was demonstrated by showing that: (i). recombinant MBP possesses mannose binding activity, and (ii). polyclonal antibodies prepared against Acanthamoeba MBP bound to the recombinant protein. Sequence analysis revealed that the MBP contains a large N-terminal extracellular domain, a transmembrane domain, and a short C-terminal cytoplasmic domain. Despite extensive BLAST searches using the MBP sequence, no significant matches were retrieved. The most striking feature of the Acanthamoeba MBP sequence is the presence of a cysteine-rich region containing 14 CXCXC motifs within the extracellular domain. In summary, we have isolated, cloned, and characterized a novel MBP from Acanthamoeba. Because the presence of antibodies to MBP in tears provides protection against infection, the availability of the MBP cDNA sequence and rMBP should help develop: (i). a tear-based test to identify individuals who are at risk of developing the keratitis and (ii). strategies to immunize high-risk individuals.

Acanthamoebae are causative agents of two distinct disease entities. In immunocompromised individuals, this protozoan parasite produces chronic granulomatous amoebic encephalitis and disseminating infections including dermatitis and pneumonitis (1,2). Granulomatous amoebic encephalitis is nearly always fatal because of the lack of an effective treatment. In immunocompetent individuals, Acanthamoebae provoke a debilitating vision-threatening corneal infection known as Acanthamoeba keratitis (AK) 1 (3)(4)(5). AK is characterized by intense pain and a slowly worsening clinical course. Contact lens wear is a major risk factor (3,6,7). In the developed world, ϳ85% of the cases are diagnosed in contact lens wearers (7). However, in the developing countries, where contact lens use is infrequent, the disease is more commonly found in non-contact lens wearers (8). The mechanism by which Acanthamoebae produce granulomatous amoebic encephalitis and AK has not been fully elucidated. It is generally accepted that the two major predisposing factors in the pathogenesis of AK are minor corneal trauma caused by contact lens wear or other noxious agents and exposure to contaminated solutions including lens care products and tap water (9,10). The adhesion of parasites to the host cells is clearly a critical first step in the pathogenesis of infection (11)(12)(13). Subsequent to adhesion, the parasites produce a potent cytopathic effect leading to target cell death (12, 14 -16). That the Acanthamoeba may adhere to host cells via a carbohydrate-binding protein has been suggested by studies demonstrating that: (i) the adhesion of Acanthamoeba to corneal epithelial cells in culture as well as to the surface of the corneal buttons can be inhibited by free methyl-␣-mannopyranoside (␣-Man) but not by a number of other sugars (13, 16 -18), (ii) Acanthamoebae bind to a neoglycoprotein, mannosylated-bovine serum albumin but not to galactose-bovine serum albumin (16), (iii) mannose-related saccharides that inhibit amoeba binding to corneal epithelial cells are also potent inhibitors of the amoeba-induced cytopathic effect (16). In addition, preliminary studies have shown that Acanthamoebae express a putative mannose-binding protein (MBP) of 136 kDa (18). These findings suggest that the adhesion of Acanthamoeba to the corneal surface is mediated by interactions between a mannose-specific lectin on the surface of the amoeba and mannose residues of glycoproteins of corneal epithelium, and that the mannose-mediated cross-talk between amoeba and corneal epithelial cells is a key component of the Acanthamoeba-induced cytopathic effect. In the present study, we demonstrate that Acanthamoebae express a ϳ400-kDa MBP that is constituted of multiple 130-kDa subunits; the MBP is highly specific for ␣-Man with little or no affinity for related sugars including mannitol and ␣-glucose. Cloning and sequenc-ing experiments revealed that Acanthamoeba MBP is (i) a novel protein with little homology to any known protein and (ii) a transmembrane protein with characteristics of a typical cell surface receptor.
Parasites and Epithelial Cells-An Acanthamoeba strain derived from an infected human cornea (MEEI 0184; Acanthamoeba castellanii based on morphological characteristics) was used throughout this study. The parasites were axenically cultured in a protease peptone/ yeast extract/glucose medium (19).
Isolation and Characterization of MBP of Acanthamoeba-To isolate MBP of Acanthamoeba, frozen cell pellets (3.5 ϫ 10 9 , Ͼ95% trophozoites) were washed in resuspension buffer (25 mM Tris-HCl, pH 7.2, 100 mM NaCl, 20 mM CaCl 2 , 1 mM phenylmethylsulfonyl fluoride) by centrifugation and were then homogenized using a Dounce homogenizer (20 strokes) in 20 ml of ice-cold extraction buffer (0.5% CHAPS ϩ 2 mM ␤-mercaptoethanol in resuspension buffer). The amoeba extracts were clarified by centrifugation (100,000 ϫ g, 1 h, 4°C) and were chromatographed on an ␣-Man affinity column (4°C, 60 min). The unbound components were removed by washing the column with the extraction buffer, and the bound components were eluted in 1.0-ml fractions with extraction buffer containing 150 mM ␣-Man. In some experiments, to confirm sugar binding specificity of the amoeba MBP, prior to elution with ␣-Man, the column was eluted with a number of unrelated sugars, including ␣-Gal, ␣-Glc, and mannitol. The fractions eluted with each sugar were then analyzed by SDS-PAGE (20) either directly or after concentrating by centrifugation in 10-kDa cut-off Centricon concentrators (Millipore, Bedford, MA). The affinity purified protein was extensively dialyzed against resuspension buffer and was then chromatographed on a Sephadex G-200 (Amersham Biosciences) column to estimate the approximate molecular weight based on the elution position of various standards including aldolase (158 kDa), catalase (232 kDa), ferritin (440 kDa), and thyroglobulin (669 kDa). The protein eluted from the gel filtration column was subsequently tested for binding to ␣-Man and ␣-Gal affinity matrices to further verify its sugar binding specificity.
N-terminal and Internal Peptides Sequencing of MBP-To determine N-terminal amino acid sequence, MBP (1 g) was electrophoresed in 8% SDS-polyacrylamide gels, blotted onto a polyvinylidene difluoride membrane (Applied Biosystems, Foster City, CA), and visualized by Ponceau S staining. The MBP band was cut out and sent to the Harvard Microchemistry Laboratory (Cambridge, MA) for N-terminal amino acid sequencing by automated Edman degradation. For internal peptide sequencing, 10 g of MBP was electrophoresed in an 8% SDS-polyacrylamide gel, stained with 0.25% (w/v) Coomassie Brilliant Blue R-250 in methanol/water/glacial acetic acid (4:5:1). The MBP band was cut out and sent to the Harvard Microchemistry Laboratory for digestion with trypsin and endoproteinase Asp-N (Asp-N), and analysis by matrixassisted laser desorption ionization time-of-flight mass spectrometry and sequencing of the peptides by automated Edman degradation. Amino acid sequence of the N-terminal and four internal peptides are shown in Table I. RT-PCR and Northern Blot Analysis-To synthesize PCR probes, total RNA was prepared using an RNAqueous kit (Ambion, Austin, TX). Reverse transcription was performed with 1 g of RNA, using 1 l of oligo(dT) (500 g/ml) and 200 units of Superscript II reverse tran-scriptase (RNase H Ϫ , Invitrogen) at 42°C for 60 min. One degenerate oligonucleotide sense primer and two antisense primers were synthesized based on the amino acid sequences of the N-terminal and two internal peptides using the Acanthamoeba codon bias (Table II). Amplification of Acanthamoeba cDNA by PCR was performed under the following cycle conditions: the first cycle at 94°C for 5 min, followed by 30 cycles of: 94°C (melting) for 1 min, 48°C (annealing) for 1 min, and 72°C for 2 min (extension), and the last cycle at 72°C for 10 min. The PCR products were fractionated on 1% agarose gels, DNA fragments were isolated from the gel using the Gel Extraction System (Qiagen) and were sequenced at the DNA sequencing facility at our institution. PCR with sense and antisense primers 1 and 2 generated fragments of ϳ300 and ϳ600 bp, respectively. The 300-bp product encoded the Nterminal and internal peptide I, and the 600-bp fragment encoded N-terminal and internal peptides 1 and 2. Based on the PCR product sequence of the 600-bp fragment, a set of nested primers (Table II) was designed using the Oligo 6.0 software (Molecular Biology Insights, Cascade, CO) to amplify a region of 585 bp. The 585-bp product was cloned into the pCR 2.1-TOPO vector using the TOPO TA Cloning Kit (Invitrogen), sequenced, and used as a probe for Northern blot analysis. A probe for Acanthamoeba actin (ϳ390 bp) was generated using the primers shown in Table II. For Northern blot analysis, poly(A) ϩ RNA was isolated from total cell RNA using the Oligotex mRNA mini kit (Qiagen). Aliquots of 3 g of mRNA were fractionated on 1% agarose gels containing formaldehyde and transferred onto a Hybond nylon membrane (Amersham Biosciences) as described elsewhere (21). The blots were prehybridized in a solution consisting of 6ϫ SSC, 0.5% SDS, 100 g/ml denatured fragmented salmon sperm DNA, and 50% formamide (4 h, 42°C). The MBP and actin PCR probes prepared as described in the previous section were labeled with [␣-32 P]dCTP using the Prime-It RT Random Primer labeling kit (Stratagene, La Jolla, CA). Unincorporated nucleotides were removed from the labeled probes using the Push Column Beta Shield device and NucTrap probe purification columns (Stratagene). The specific activity of the labeled probes was ϳ10 9 cpm/g of DNA. The probe was denatured at 100°C for 5 min and chilled on ice briefly and was then added to prehybridization buffer to the final concentration of 10 6 cpm/ml. Hybridization was performed overnight at 42°C. Blots were sequentially washed with 2ϫ SSC buffer containing 0.1% SDS (15 min, room temperature), 0.1ϫ SSC buffer containing 0.5% SDS (30 min, 37°C), 0.1ϫ SSC buffer containing 0.5% SDS (5 min, 55°C), and then subjected to autoradiography.
Cloning and Nucleotide Sequencing of Acanthamoeba MBP-To obtain the MBP DNA sequence, an A. castellanii genomic library constructed in EMBL3A (22) was screened using the 585-bp MBP probe. The genomic library was plated out on 20 15-cm plates, each containing ϳ50,000 plaques. The plaques were blotted onto nitrocellulose filters (Schleicher & Schuell), and the DNA was denatured, neutralized, and immobilized (23). The filters were prehybridized at 42°C for 4 h in a solution consisting of 20 mM Pipes, pH 6.5, 0.8 M NaCl, 0.5% SDS, 100 g/ml denatured fragmented salmon sperm DNA, and 50% formamide, and then hybridized with a 32 P-labeled MBP probe overnight at 42°C. Filters were then washed three times with 500 ml of 0.1ϫ SSC containing 0.1% SDS (15 min, 55°C). One positive MBP genomic clone was isolated and re-plated to assess its purity. The selected clone was analyzed by digestion with various restriction enzymes (EcoRI, BamHI, and SalI) and mapped by Southern blot analysis (see below). The MBP gene mapped to a ϳ9.5-kilobase BamHI fragment. The 9.5-kilobase fragment was subcloned into a pUC18 vector and sequenced. The sequence of the MBP gene was analyzed using Genscan software (MIT, Cambridge, MA) to predict the MBP-cDNA (24). Based on the predicted sequence, primers were designed using Oligo 6.0 software (Molecular Biology Insights) to amplify the full-length MBP-cDNA by RT-PCR. The full-length MBP-cDNA was subcloned into the pCR 2.1-TOPO vector using the TOPO TA Cloning Kit (Invitrogen) and sequenced.
DNA Preparations and Southern Blot Analysis-Bacteriophage DNA was isolated using the Lambda DNA extraction kit (Qiagen). Plasmid DNA was prepared using the Qiagen plasmid kit. Bacteriophage DNA and plasmid DNA preparations (1 g) were fractionated on 0.4 and 0.7% agarose gels containing ethidium bromide, respectively, and were transferred onto Hybond nylon membranes (Amersham Biosciences). The blots were prehybridized (4 h, 42°C) in a solution consisting of 6ϫ SSC, 0.5% SDS, 100 g/ml denatured fragmented salmon sperm DNA, and 50% formamide. The hybridization was performed using the 32 P-labeled MBP probe (specific activity ϳ10 9 cpm/g of DNA, 10 6 cpm/ml, overnight, 42°C). The blots were washed to a final stringency of 0.1ϫ SSC buffer containing 0.5% SDS at 60°C and were then subjected to autoradiography.

Cloning of Acanthamoeba Mannose-binding Protein
Computer Analyses-Genscan software (Maize data base) was used for analysis of the MBP gene structure to predict cDNA and to generate intron/exon maps. The intron/exon map generated by Genescan software was verified and corrected as necessary by comparison of the gene sequence with that of the full-length MBP cDNA sequence produced by RT-PCR. Similarity searches of the MBP cDNA and of deduced amino acid sequences to the published sequences (National Biomedical Research Foundation, Swiss-Prot and GenBank data bases) were done using BLAST and FASTA programs (25, 26). Similarity searches were also performed against sequences in the TIGR partial Acanthamoeba genomic data base. Putative functional domains of Acanthamoeba MBP were analyzed using various programs (see below) and modeled using Simple Molecular Architecture Research Tool (27). Signal P (28) was used to identify signal peptides, and the TMHMM program (29) was used to identify transmembrane, intracellular, and extracellular domains. PSORT was used to predict the MBP cellular localization. Unique structural features including specific motifs and domains were identified using Pfam (30) and PROSITE (31), respectively. NetNGlyc 1.0 and NetOGlyc 3.0 were used to identify putative N-and O-glycosylation sites (32,33). Protparam Program found on the ExPASy Proteomics tool web page was used to predict molecular weight and isoelectric point of the MBP.
Production of Polyclonal Antibodies Against Acanthamoeba MBP and Western Blot Analysis-Affinity purified Acanthamoeba MBP was shipped to Aves Laboratories Inc. (Tigard, OR) for preparation of polyclonal IgY antibody in chicken. One hen received 6 boosts of 20 g of MBP each at 2-week intervals. One week after the last injection, 6 eggs were collected over 2 weeks, and the IgY was purified from the egg yolk (anti-MBP IgY) using a proprietary method (Aves Laboratories Inc.). The purified preparation was Ͼ95% IgY as determined by Coomassie staining of SDS-PAGE gels. For comparison purposes, preimmune IgY was isolated from the egg yolk of the same hen collected prior to immunization. The antibody was tested for specificity by Western blot analysis (34). For this, affinity purified MBP and membrane extracts of parasites were electrophoresed on 8% SDS-polyacrylamide gels and transferred onto nitrocellulose filters using a Trans-Blot cell (Bio-Rad). The protein blots were treated overnight at room temperature with 5% dry milk in phosphate-buffered saline to block nonspecific binding sites and were then sequentially incubated with anti-MBP IgY (1:1000 dilution, 37°C, 1 h), horseradish peroxidase-linked anti-chicken IgY (Aves Laboratories, 1:5000 dilution, 37°C, 1 h), and a freshly prepared solution of diaminobenzidine-H 2 O 2 reagent (0.05% diaminobenzidine, 0.01% H 2 O 2 , 0.04% nickel chloride in Tris-HCl buffer, pH 7.2). The membranes were washed extensively with phosphate-buffered saline containing 0.1% Tween 20 after each incubation step. Control reactions were performed as described above, except that the blots were incubated with preimmune IgY instead of the anti-MBP IgY.
Expression of Recombinant MBP (rMBP)-To construct an expression plasmid with a C-terminal histidine tag, vector pQE-16 (Qiagen) was double digested (BamHI-HindIII) to remove the recombinant dihydrofolate reductase encoded by this vector. A DNA construct encoding MBP with BamHI-HindIII restriction sites and a C-terminal 6-histidine tag was obtained by PCR amplification (see Table II for details on PCR primers) and checked by DNA sequencing. The construct was digested with BamHI and HindIII, and the insert encoding the MBP open reading frame was inserted into the respective sites of pQE. The resultant plasmid was designated as pQE-MPB. In this construct, MBP was under the control of the T5 promoter, and the natural stop codon of MBP was exchanged with the C-terminal 6-histidine tag.
Competent Escherichia coli cells (M15 (pREP4), Qiagen) were transfected by heat shock with the plasmid pQE-MBP. In control experiments, the host cells were transfected either with the empty vector alone or with the original vector encoding dihydrofolate reductase, pQE-16. To express rMBP, transfected cells were grown at 37°C to an A 600 of 0.5-0.7 in 100 ml of LB media supplemented with kanamycin (25 g/ml) and ampicillin (100 g/ml). Expression was induced by addition of 1 mM isopropyl-␤-D-thiogalactopyranoside. Cells were lysed in BugBuster buffer (Novagen) containing 1 unit/ml benzonase nuclease (Novagen), 1 unit/ml recombinant lysozyme (Novagen), and 1 mM phenylmethylsulfonyl fluoride (20 min, room temperature), and the lysate was cleared by centrifugation (12,000 ϫ g, 20 min). Inclusion bodies were prepared from the pellet and resuspended using the Inclusion Bodies Solubilization kit (Novagen). Supernatant and resuspended inclusion bodies were chromatographed on a nickel-affinity column (1-ml Ni-NTA resin, Qiagen). The bound fraction was eluted with 0.3 M imidazole in 20 mM Tris-HCl, pH 7.9, 0.5 M NaCl, and analyzed by SDS-PAGE and silver staining. Recombinant proteins were tested for mannose binding activity and for reactivity with anti-Acanthamoeba MBP and a PentaHis monoclonal antibody (Qiagen).
Prior to testing for mannose binding activity, the rMBP was allowed to refold using the reagents provided in the Pro-Matrix Protein Refolding Kit (Pierce). Briefly, 100-g aliquots of rMBP were reduced by treatment with 10 mM ␤-mercaptoethanol (30 min, 37°C), desalted on a Bio-Gel P-2 column (Bio-Rad), and then incubated in a number of different buffer formulations and additives provided in the refolding kit. Following the incubation, insoluble material was removed by centrifugation (10,000 ϫ g for 10 min), and the supernatant was allowed to bind to ␣-Man-conjugated agarose beads (␣-Man gel) in the absence or presence of 0.1 M ␣-Man or 0.1 M ␤-lactose (2 h at 4°C). After incubation, the beads were washed three times by centrifugation in 0.1 M Tris buffer, pH 7.2, containing 120 mM NaCl and 20 mM CaCl 2 . An equal volume of SDS-PAGE sample buffer was added to the washed beads, the samples were boiled for 4 min and centrifuged; the supernatants were electrophoresed in 8% SDS-polyacrylamide gels, and the proteins in the gel were visualized by silver staining.

Acanthamoebae Express a MBP
Isolation of Acanthamoeba MBP-MBP was isolated by affinity chromatography from detergent extracts of Acanthamoeba parasites. The MBP in the parasite extracts was allowed to bind to an ␣-Man gel. The column was extensively washed with lysis buffer and then sequentially eluted with lysis buffer containing 0.1 M ␣-Glc, ␣-Gal, mannitol, and ␣-Man. Equal aliquots of each fraction were electrophoresed in SDS-polyacrylamide gels, and protein bands were visualized by silver staining. A 130-kDa component was detected in the fraction eluted with ␣-Man, but not in the fractions eluted with any other sugar tested (Fig. 1A). Gel filtration on Sephadex G-200 using various calibration standards (158 to 669 kDa, see "Experimental Procedures" for details) revealed that the apparent molecular weight of the MBP under non-denaturing conditions is ϳ400 kDa (not shown). This suggests that the MBP is an oligomer constituted of 130-kDa subunits. Upon rechromatography on sugar affinity columns, the ϳ400-kDa component eluted from the gel filtration column, bound to ␣-Man gel but not to ␣-Gal gel; again the protein that bound to the ␣-Man gel migrated as a 130-kDa component on SDS-polyacrylamide gels (Fig. 1B). These observations confirm that the 130-kDa component detected by SDS-PAGE is the major subunit of a ϳ400-kDa oligomer, which retains its mannose binding activity after

CCCAAGCTTGGGTTAGTGATGGTGATGGTGATGGAGGGTGTGCAGGTCAGCC
Cloning of Acanthamoeba Mannose-binding Protein mannose affinity isolation and gel filtration. ConA, a mannosebinding protein, and Ricinus communis agglutinin I, a galactose-binding protein, served as positive and negative controls for ␣-Man and ␣-Gal affinity columns, respectively (Fig. 1B). Acanthamoebae Express a 2.8-kb MBP Gene Transcript-Northern blot analysis was used to detect the MBP gene transcript. To synthesize a PCR probe for Northern blot analysis, affinity purified MBP was enzymatically digested with trypsin and Asp-N, and the N-terminal and four internal peptides of MBP were sequenced by automated Edman degradation. Longer peptides (12-18 amino acid residues) were sequenced after Asp-N digestion (Table I, internal 1 and 2). The MBP was relatively resistant to trypsin digestion. In repeated efforts, because of low yield of tryptic peptides, it was not possible to obtain an amino acid sequence of more than 5-7 amino acid residues (Table I, internal 3 and 4). Based on the amino acid sequence of the N-terminal and an internal peptide, a 585-bp PCR probe was obtained. For Northern blot analysis, the Acanthamoeba poly(A) ϩ RNA was electrophoresed on 1% formaldehyde-agarose gels in the presence of ethidium bromide (Fig. 2,  left panel), and the RNA blots were hybridized with the 585-bp MBP probe. A single mRNA transcript of ϳ2.8 kb was detected (Fig. 2, middle lane). Similar results were obtained from three different preparations of Acanthamoeba poly(A) ϩ RNA.
Cloning and Nucleotide Sequencing of Acanthamoeba MBP Gene-To obtain a genomic clone of MBP, an A. castellanii genomic library, constructed in EMBL3A, was screened using the 32 P-labeled 585-bp MBP probe. One positive MBP genomic clone was mapped to a ϳ9.5-kb BamHI fragment. The ϳ9.5-kb BamHI fragment was isolated and subcloned into a pUC18 vector, and the fragment encompassing the MBP gene was sequenced. The resulting sequence was analyzed with Genscan software (Maize data base, Massachusetts Institute of Technology) for gene structure and cDNA prediction. The predicted cDNA (not shown) was used to design primers to clone the full-length MBP cDNA by RT-PCR. The 2502-bp MBP cDNA obtained by RT-PCR was sequenced. Comparative analysis of full-length MBP cDNA and MBP gene in conjunction with the analyses of the intron/exon map generated by Genscan revealed that the Acanthamoeba MBP gene is composed of 6 exons and 5 introns that span 3.6 kb of the amoeba genome (Fig. 3) and that MBP cDNA codes for a precursor protein of 833 amino acids. The calculated peptide mass of mature MBP was 85 kDa, which is significantly less than the apparent molecular mass (130 kDa) detected by SDS-PAGE. All five peptide sequences of MBP obtained by Edman degradation (Table I)  cDNA (Fig. 4, underlined). Five targets with high probability of N-glycosylation and 3 targets for O-glycosylation were identified in the extracellular domain (Fig. 4, solid circles). In addition, a cysteine-rich domain (residues 282-701) was identified. In this domain, cysteine content was 16.4% and there were 14 CXCXC repeats (Fig. 4, shaded rectangles). Analysis of the deduced amino acid sequence (Fig. 5) using various programs described under "Experimental Procedures" revealed that residues 1-21 correspond to the MBP signal peptide, and that amoeba MBP is a transmembrane protein containing a 712amino acid long N-terminal extracellular domain (residues 22-733), a 22-amino acid long transmembrane domain (residues 734 -755), and a short, 78-amino acid long, C-terminal intracellular domain. BLAST searches of the MBP DNA sequence against NCBI data bases and the TIGR partial Acanthamoeba genomic data base retrieved no significant matches. However, the MBP sequence contains small regions that have limited homology to some of the known proteins including E-and P-selectin precursor proteins (Table III).

Expression and Biochemical Characterization of rMBP
MBP with a 6-histidine tag added to the C terminus was expressed in E. coli M15 (pREP4) cells. Cells transfected with the empty vector alone and pQE-16, which encodes for dihydrofolate reductase, were used as negative and positive controls, respectively. Dihydrofolate reductase was recovered in the soluble fraction (not shown), whereas rMBP was isolated from the inclusion bodies. The recombinant MBP was isolated by affinity chromatography on Ni-NTA-agarose and was tested for reactivity with anti-Acanthamoeba MBP IgY, and for binding to ␣-Man gel.
Polyclonal anti-MBP was produced in chicken against affinity purified Acanthamoeba MBP, and the purified IgY fraction was tested for specificity by Western blot analysis. The antibody reacted with a 130-kDa component present in the membrane extracts of parasites (Fig 6A). This 130-kDa component comigrated with affinity purified Acanthamoeba MBP which, as expected, also reacted with the antibody. Many intensely stained components seen in the Ponceau S-stained blot of the parasite extract did not react with anti-MBP (Fig. 6A), attest-ing to the specificity of the antibody. Also, no components were detected in the control blots incubated with preimmune IgY (not shown). In Western blot analysis, Ni-NTA affinity purified rMBP reacted with both PentaHis and anti-Acanthamoeba MBP antibodies (Fig. 6B). On SDS-polyacrylamide gels, rMBP migrated as a 110-kDa component that is significantly higher than the predicted mass of 85 kDa, but is identical to that of deglycosylated cellular MBP. 2 The recombinant protein exhibited mannose binding activity only after it was allowed to refold using the Pro-Matrix Protein Refolding kit. The refolding conditions required to restore the optimal mannose binding activity involved reducing the rMBP and incubating it for 72 h at 4°C in 1 ml of a buffer containing 1 M guanidine, 50 mM Tris, pH 8.2, 120 mM NaCl, 20 mM CaCl 2 , 0.8 mM KCl, 2 mM GSH, and 0.2 mM GSSG. The refolded rMBP prepared under these conditions bound to ␣-Man-gel (Fig. 6C, lane Buffer). The binding of the rMBP to the ␣-Man gel was inhibited by ␣-Man (Fig.  6C) but not by ␤-lactose (not shown). These findings confirm that the recombinant protein is authentic MBP.

DISCUSSION
In this study, we describe the isolation and molecular cloning of a novel MBP from A. castellanii. The MBP migrated as a single component of 130 kDa on SDS-polyacrylamide gels in reducing conditions but as a ϳ400-kDa component in nonreducing conditions on gel filtration columns. Affinity chromatography on ␣-Man affinity columns proved to be a very efficient and specific single-step method for the isolation of Acanthamoeba MBP. The protein was eluted from the mannose affinity column by ␣-Man but not by a number of other sugars, including mannitol and glucose. Also, the ϳ400-kDa component isolated by gel filtration of the ␣-Man affinity purified material, when rechromatographed on sugar-affinity columns, bound to ␣-Man gel but not to ␣-Gal gel; again, the protein that bound to the ␣-Man gel migrated as a 130-kDa component on SDS-polyacrylamide gels. These data provide evidence that the 130-kDa component detected by SDS-PAGE is the major sub-

FIG. 3. Acanthamoeba gene is constituted of 6 exons and 5 introns.
A genomic clone of MBP was sequenced, and the resulting sequence was analyzed with Genscan software (Maize data base, MIT) for cDNA prediction. The predicted cDNA was used to design primers to clone the MBP cDNA by RT-PCR. The MBP cDNA was sequenced and comparatively analyzed to generate an exon/intron map. This analysis revealed that the MBP gene has 6 exons comprising 2502 bp coding for a precursor protein of 833 amino acids. Sequences of the Acanthamoeba MBP gene (AY604039) and cDNA (AY604040) have been deposited in GenBank TM .
unit of a ϳ400-kDa oligomer that retains its mannose binding activity after purification on ␣-Man affinity and gel filtration columns. Our findings that the MBP migrate as a single component of 130-kDa on SDS-polyacrylamide gels in reducing conditions but as a ϳ400-kDa component in nonreducing conditions on gel filtration columns, suggest that the Acanthamoeba MBP is composed of at least three 130-kDa subunits. The multiple subunits may endow multivalent binding properties to the MBP. This is of significant interest because it is well established that high affinity multivalent binding capable of clustering the receptors is often required for the biological function of most lectins (35,36).
Cloning and sequencing of the Acanthamoeba MBP gene enabled us to predict the MBP cDNA sequence and design primers to clone the full-length MBP cDNA by RT-PCR. That the cDNA cloned by RT-PCR encodes authentic MBP is suggested by a number of criteria including: (i) the cloned cDNA encoded the peptides sequenced from the purified MBP isolated from membrane preparations of parasites; (ii) the antibody raised against purified MBP prepared from parasites specifically recognized a rMBP expressed in E. coli enzymatically deglycosylated MBP 2 is ϳ110 kDa. Therefore, the discrepancy between the predicted and apparent molecular weight appears to be at least in part because of the aberrant migration behavior of the protein on SDS-polyacrylamide gels.
Analysis of the deduced amino acid sequence of the MBP cDNA revealed that the architecture of the MBP protein is characteristic of a cell surface receptor (Type 1a, as deduced by PSORT). The N-terminal amino acids (residues 1-21) contain a hydrophobic stretch and code for a typical, cleavable signal peptide. The remaining sequence constitutes a mature protein of 812 amino acids that contains: (i) a large extracellular domain; (ii) a putative transmembrane region that resembles other known membrane anchors; and (iii) a cytoplasmic domain that, as in most known transmembrane proteins, is located at C terminus. The extracellular domain contains five putative N-glycosylation sites and three O-glycosylation sites.
Like the galactose-specific lectin of Entamoeba (37, 38), the Acanthamoeba MBP lacks sequence identity to well character- FIG. 6. The recombinant protein binds to anti-Acanthamoeba MBP and possesses mannose binding activity. A, specificity of anti-Acanthamoeba MBP. Polyclonal antibody was prepared against affinity purified Acanthamoeba MBP in chicken and the purified IgY fraction was tested for specificity by Western blot analysis. Aliquots of Acanthamoeba membrane extracts (5 g) and affinity purified Acanthamoeba MBP (0.1 g) were electrophoresed in SDS-polyacrylamide gels; proteins from the gels were transferred onto nitrocellulose filters and were visualized by staining with Ponceau S (left panel). After destaining, the same blots were processed for immunostaining with anti-MBP. Note that the antibody specifically reacted with a 130-kDa component in the membrane extracts of parasites as well to the affinity purified Acanthamoeba MBP (lane Amoeba MBP). B, the recombinant protein expressed in E. coli binds to anti-Acanthamoeba MBP. Recombinant MBP expressed in E. coli was isolated from inclusion bodies by affinity chromatography on a Ni-NTA-agarose column and was tested for reactivity with a PentaHis monoclonal antibody and polyclonal anti-Acanthamoeba MBP. Note that both antibodies reacted with a 110-kDa component. Ponceau S, blot stained with Ponceau S prior to staining with the PentaHis monoclonal antibody. A similar Ponceau S staining pattern was detected in the blot used for staining with anti-MBP. C, the recombinant protein possesses mannose binding activity. Reduced rMBP was refolded as described under "Experimental Procedures" and was incubated with mannose-conjugated agarose beads (␣-Man gel) in the absence or presence of 0.1 M ␣-Man. After incubation, the beads were rinsed, boiled in the SDS-PAGE sample buffer, and then centrifuged; the supernatants were electrophoresed in 8% SDSpolyacrylamide gels, and proteins in the gel were visualized by silver staining. Note that the recombinant protein bound to the ␣-Man gel (lane Buffer) and the binding was inhibited by ␣-Man (lane Buffer ϩ ␣-Man) but not ␤-lactose (not shown).
ized lectin carbohydrate recognition domains of C-type lectins, mammalian galectins, and the plant lectins. Despite extensive BLAST searches, no protein with known function shows significant homology to the deduced amino acid sequence of MBP. Many carbohydrate-binding proteins have been grouped into several distinct families, such as selectins, galectins, collectins, attractins, and polycystins (39,40). MBP does not fall into any of these families, as judged by alignment between MBP and the consensus sequences that characterize each family. However, the MBP sequence contains several regions that have limited homology (29 to 42%, Table II) to some of the known proteins including silk proteins, Balbiani ring protein 3 precursor, Entamoeba galactose-specific lectin, and E-and P-selectin precursor proteins. The most striking structural feature of the deduced amino acid sequence of the Acanthamoeba MBP is the presence of a cysteine-rich region containing 14 CXCXC motifs within the extracellular domain. The homology among MBP, silk, and Balbiani ring proteins is because of the presence of these CXCXC repeats. This conserved pattern CXCXC, where X can be any amino acid, is found in up to five copies in vascular endothelial growth factor C (41) and over 70 copies in silk and Balbiani ring proteins (42). Whereas the function of the CX-CXC motifs remains to be determined, a similar motif, CXXC, is well known for its redox function in many proteins (43). As regards the possible function of the cysteine-rich domain of the MBP, of relevance to the current study is the finding of Petri and colleagues (37) that disulfide bonds in the extracellular domain of Entamoeba lectin folds the lectin into a proteaseresistant conformation because the lectin was only sensitive to proteolytic attack upon reduction and alkylation of cysteine residues. As described earlier, despite repeated efforts, we were unable to obtain tryptic peptides of MBP in good yield for amino acid sequencing. Therefore, it appears that the Acanthamoeba MBP is also relatively resistant to trypsin. Because tears are rich in proteases (44,45), MBP resistance to proteolytic attack may be critical to the pathogenesis of Acanthamoeba infection.
Determination of the primary structure of MBP is an important first step in understanding the mechanism of Acanthamoeba keratitis. The availability of the cDNA sequence coding for Acanthamoeba MBP should provide the opportunity to design experiments for recombinant expression of selected regions of the MBP to identify carbohydrate recognition domains and other functionally relevant regions that can be used to: (i) develop a tear-based test to identify individuals who are at risk of developing the keratitis; and (ii) immunize high-risk individuals. Indeed, in a recent study, we found that tears of healthy individuals contain antibodies against Acanthamoeba MBP, and oral immunization with rMBP protects against Acanthamoeba keratitis in a hamster animal model and this protection correlates with the appearance of MBP-specific IgA in tears of immunized animals. 3 The MBP-specific IgA in tears is likely to provide protection by preventing the parasite binding to the corneal epithelium.