INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Heartwater, or cowdriosis, a rickettsial disease of domestic and wild ruminants, is caused by the etiologic agent Cowdria ruminantium, which is transmitted by ticks belonging to the genus Amblyomma. Heartwater is responsible for mortality levels as high as 90% in susceptible ruminants and has been reported throughout sub-Saharan Africa and on several Caribbean islands (22). Animals surviving the disease may have persistent infections; thus, they are capable of serving as reservoirs for transmission to ticks (1).

Major antigenic protein 1 (MAP1) of C. ruminantium was found to be antigenically conserved in nine strains of C. ruminantium from Africa and the Caribbean (12) and was used in a competitive enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies (13). However, further studies revealed that MAP1 (complete, recombinant, or truncated) also reacts with sera from ruminants from heartwater-free areas (14, 15, 18) and with several members of the genus Ehrlichia (10). Furthermore, significant variability in nucleotide sequences of MAP1 among four African and two Caribbean strains ranged from 0.6 to 14.0% (20). These nucleic acid differences translated to amino acid substitutions, deletions, or insertions at three hypervariable regions of the gene (20). Demonstration of homology in MAP1 coding sequences between C. ruminantium and Ehrlichia spp. and divergence in hypervariable regions between strains suggest that MAP1 may not be a suitable candidate for specific serodiagnosis of heartwater. Thus, current serological tests for cowdriosis lack specificity, i.e., false-positive sera react with the whole C. ruminantium organism and the reported immunoreactive MAP1 (8).

The gene encoding another major antigenic protein (MAP2) was isolated, cloned, and expressed in Escherichia coli (17). Sera obtained from cattle, sheep, and goats infected with C. ruminantium reacted with this recombinant MAP2 (17). MAP2 is 55.5% identical at the amino acid level to major surface protein 5 (MSP5; 19 kDa) (17, 25) of Anaplasma marginale. A competitive ELISA using a monoclonal antibody against MSP5 has been evaluated and shown to be sensitive and specific for the diagnosis of Anaplasma infections in acutely or persistently infected cattle (16).

In this study, we examine the potential value of MAP2 as an antigen for the serodiagnosis of heartwater. First, we report the cloning and sequencing of map2 genes from several C. ruminantium strains of differing geographic origins. Second, we determine if, as with MAP1, false-positive sera react with MAP2. Finally, we determine whether MAP2-like proteins are present in Ehrlichia canis and Ehrlichia chaffeensis (causative agent of human monocytic ehrlichiosis), which are closely related to C. ruminantium according to 16S rRNA studies (6, 24), to evaluate whether epitope-specific tests are required for diagnosis of these rickettsiae.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Origins of C. ruminantium strains, E. canis, and E. chaffeensis. Three strains of C. ruminantium from Zimbabwe (Crystal Springs, Highway, and Palm River [4]), one strain from Sudan, Um Banein (11), and the Antigua strain (3) from the Caribbean were grown in bovine aortic endothelial cells (2, 4). E. canis (Oklahoma strain) and E. chaffeensis (Arkansas isolate) were grown in the canine macrophage cell line DH82 in minimum essential medium containing 12.5% fetal bovine serum and 200 mM L-glutamine (7). Organisms were harvested from the culture following the complete lysis of host cells, and rickettsial genomic DNA was isolated as previously described (5, 20).

Amplification of the map2 genes of C. ruminantium strains of different geographic origins and of the map2-like genes of E. chaffeensis and E. canis. Primers AB249 and AB251, corresponding to flanking regions of the map2 gene (open reading frame [ORF] 1) of C. ruminantium clone pF5.2 (17), were synthesized. Primers for PCR amplification of E. canis and E. chaffeensis were selected by using information from one of three sources: (i) sequences determined after alignment of conserved regions of the map2 gene of C. ruminantium and the msp5 gene of A. marginale, (ii) sequences obtained from newly sequenced regions of the map2-like genes of E. canis or E. chaffeensis, and (iii) primers developed to sequence the pF5.2 clone of the Crystal Springs strain of C. ruminantium (17). Primers (Table 1) were synthesized by the Oligonucleotide Synthesis Laboratory (Interdisciplinary Center for Biotechnology Research, University of Florida, Gainesville) or National Biosciences, Inc. (Plymouth, Minn.). These primers were used in a PCR assay to amplify DNA of the strains of C. ruminantium, E. canis, and E. chaffeensis. Briefly, target DNA (1 ng) was amplified in a mixture of 0.4 mM deoxynucleoside triphosphates, 0.5 µM (each) primer, 20 mM MgCl₂, 100 mM KCl, 200 mM Tris-HCl (pH 8.2), 60 mM (NH₄)₂SO₄, 1% Triton X-100, 100 µg of nuclease-free bovine serum albumin/ml, and 2.5 U of native Pfu DNA polymerase. PCR assays, for the amplification of map2 and map2-like genes of C. ruminantium, E. canis, and E. chaffeensis, were performed at 94°C for 5 min, followed by 30 cycles of varying denaturing, annealing, and extension temperatures and times (Table 2). A final extension step at 72°C was performed for 10 min in all reactions. The amplicons were analyzed by gel electrophoresis on a 0.8% agarose gel in 1× TBE buffer (89 mM Tris, 89 mM boric acid, 2 mM disodium EDTA). PCR products were purified by using the QIAquick Spin PCR purification kit (Qiagen, Inc., Chatsworth, Calif.) as described by the manufacturer. The DNA was eluted in 50 µl of 10 mM Tris-HCl, pH 8.3. 

Cloning and restriction enzyme analysis. Amplicons were concentrated to 20 µl in a Speed Vac concentrator (Savant Instruments, Inc., Farmingdale, N.Y.) and ligated to EcoRV-digested pBluescript SK(+) (Stratagene, La Jolla, Calif.) at 16°C in the presence of T4 DNA ligase at an insert/vector molar ratio of 8:1. The recombinant plasmid was used to transform competent E. coli XL1-Blue cells, and transformants were grown on Luria-Bertani agar plates in the presence of ampicillin (50 µg/ml), 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) (50 mg/ml), and isopropyl--D-thiogalactopyranoside (IPTG) (0.2 mg/ml). Positive colonies were selected, inoculated into terrific broth (21) containing ampicillin (250 µg/ml), and incubated overnight at 37°C with vigorous shaking. Plasmid DNA from positive colonies was extracted by the boiling prep method (9). DNA was reconstituted in Tris-EDTA buffer (pH 8.0) containing 20 µg of DNase-free RNase/ml and was analyzed on a 0.8% agarose gel. Recombinant clones containing the map2 genes of different strains of C. ruminantium were digested with seven restriction enzymes (AccI, HincII, EcoRV, EcoRI, XbaI, BamHI, and HindIII) to compare their various sizes to those obtained for the reported Crystal Springs strain (GenBank accession no. G289922). The resulting restriction enzyme-digested DNA was analyzed on a 0.8% agarose gel.

DNA sequencing and sequence data analysis. The DNA insert in pBluescript SK(+) was sequenced as double-stranded DNA by using Sequenase (United States Biochemical) as recommended by the manufacturer. T3 and T7 plasmid-specific promoter primers were used in initial reactions, and then new oligonucleotides were synthesized based on the sequences obtained ("primer walking"). Insert DNA was completely sequenced on both strands. DNA sequences were analyzed with the PCGENE (Intelligenetics) and Genetics Computer Group (University of Wisconsin) programs. Database searches were performed with the BLAST program of the National Center for Biotechnology Information (Bethesda, Md.).

C. ruminantium recombinant MAP2 production and purification. The 627-bp ORF encoding MAP2 of C. ruminantium was subcloned into the expression vector pFLAG (International Biotechnologies, Inc., New Haven, Conn.) and transformed into E. coli for expression of recombinant protein (17). An E. coli colony expressing MAP2 was selected and expanded in L broth (1% tryptone-0.5% yeast extract-1% NaCl), induced with 0.1 mM IPTG, and purified by monoclonal antibody affinity chromatography, as described by the manufacturer of the pFLAG vector system. The protein concentration was determined, and the size and purity were analyzed on a 12% acrylamide gel. The antigenicity of the protein was analyzed by using an indirect ELISA and hyperimmune sheep serum.

Antisera. Positive sera were collected from sheep born and reared under Amblyomma-free conditions in Zimbabwe and experimentally infected with the Welgevonden strain of C. ruminantium (17). False-positive sera were obtained from sheep (numbers 056, 060, 062, and 066) born and raised in regions of Zimbabwe where neither Amblyomma ticks nor the disease has ever been observed. Animals were confirmed free of infection by tick pickup, PCR, and challenge (18). Negative sera were obtained from sheep born and reared under Amblyomma-free conditions in Zimbabwe. Normal and anti-E. canis dog sera (obtained from the Veterinary Medical Teaching Hospital at the University of Florida) were confirmed positive or negative by an indirect fluorescent antibody test.

Indirect ELISA. ELISA plates were coated by using isolated recombinant MAP2 (4 µg/ml) in 0.14 M NaCl-20 mM Na₂HPO₄-3 mM KH₂PO₄ (pH 7.2) containing 0.02% (wt/vol) sodium azide (phosphate-buffered saline [PBS]-azide) at 50 µl per well and incubated overnight at 4°C. After five washes with buffer containing PBS-azide and 0.05% (vol/vol) Tween-20, the wells were blocked for 60 min at room temperature with 1.0% (wt/vol) bovine serum albumin in PBS-azide. The plates were washed five times and incubated for 60 min with sera or supernatant diluted in PBS-azide (50 µl per well). The plates were washed (five times) and incubated at room temperature for a further 60 min in the presence of rabbit anti-species-specific immunoglobulin G conjugated to alkaline phosphatase (1:1,000; 50 µl per well). After another five washes, the substrate, p-nitrophenylphosphate (1 mg/ml; Sigma), in 0.16 M NaHCO₃-0.14 M Na₂CO₃-0.02 M MgCl₂ (pH 9.6) was added at 50 µl per well to each well and incubated for 30 min at room temperature. Absorbance was read at 405 nm.

Immunoblot of C. ruminantium and E. canis antigens. C. ruminantium (Fig. 5a, lanes A) or E. canis (Fig. 5a, lanes B) lysates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. Membranes were blocked with Tris-buffered saline (TBS) (0.1 M Tris HCl-0.9% NaCl [pH 8.0])-0.25% gelatin, then reacted with an anti-C. ruminantium antiserum (from sheep 483; Crystal Springs strain), anti-recombinant MAP2 antiserum (from sheep 177), or anti-E. canis antiserum (dog Mukanya; E. canis Oklahoma). Membranes were washed three times in TBS containing 0.25% Tween 20, reacted with peroxidase-labeled protein G (Zymed) for 2 h, washed three times as described above, and developed by incubation with 4CN peroxidase substrate (Kirkegaard and Perry, Gaithersburg, Md.).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Analysis of the map2 genes of C. ruminantium. Five geographically differing strains of C. ruminantium were examined in order to determine the conservation of MAP2. The map2 genes from the four strains were amplified with primers AB249 and AB251 (Table 1), resulting in an amplicon of approximately 0.74 kb (Fig. 1). The map2 genes from the strains were cloned by PCR cloning methods, and restriction enzyme analysis was performed and compared with that for the previously reported map2 gene of the Crystal Springs strain. Restriction map analysis of the map2 genes demonstrated similarities between Palm River, Um Banein, and Antigua and similarities between Crystal Springs and Highway (Fig. 2). EcoRV, XbaI, and HincII restriction sites were found in all strains analyzed. The AccI restriction site was observed in the Crystal Springs and Highway strains only. Likewise, the Palm River, Um Banein, and Antigua strains had identical restriction enzyme maps and contained an additional HincII site.

View larger version (80K): [in this window] [in a new window]   FIG. 1.   Analysis of PCR products of the five strains of C. ruminantium amplified by using primers AB249 and AB251. PCR amplicons of Antigua (lane A), Crystal Springs (lane C), Highway (lane H), Palm River (lane P), and Um Banein (lane U) are shown.

View larger version (19K): [in this window] [in a new window]   FIG. 2.   Restriction maps of the map2 genes from C. ruminantium strains. Maps are drawn left to right, 5' to 3' with respect to the MAP2 coding sequence. Restriction enzymes used for mapping are shown. The restriction map of the Crystal Springs strain map2 gene is presented for comparison.

View larger version (48K): [in this window] [in a new window]   FIG. 3.   Comparison of the coding nucleotide sequences of MAP2 and MAP2 homologs of C. ruminantium strains and Ehrlichia species. The MAP2 coding sequences of the four strains of C. ruminantium and the two species of Ehrlichia were determined for both strands and were aligned with the MAP2 sequence from the Highway strain by using the Genetics Computer Group programs. The complete nucleotide sequence from the Highway strain is presented. The sequences for other strains and species are presented only when they differ from the Highway sequence. A dot indicates identity with the Highway sequence. The underlined area is the predicted ribosomal binding site of the revised MAP2 nucleic acid sequence of C. ruminantium.

View larger version (36K): [in this window] [in a new window]   FIG. 4.   Comparison of the MAP2 and MAP2-like coding sequences at the amino acid level. The nucleotide sequences for five strains of C. ruminantium and two Ehrlichia species were translated. The complete sequence for the Highway strain is presented. The sequences for other strains and the Ehrlichia species are presented only when they differ from the Highway sequence. A dot indicates identity with the Highway sequence.

Analysis of the map2-like genes of E. chaffeensis and E. canis. The map2-like genes were amplified by using several sets of PCR primers (Tables 1 and 2), cloned, and sequenced. The nucleotide and translated amino acid sequences were compared with the sequences of the C. ruminantium strains (Fig. 3 and 4). When the amino acid sequences are aligned, Ehrlichia proteins begin 4 amino acids downstream from the starting methionine of the ORF of MAP2 of C. ruminantium (Fig. 4). Although MAP2-related polypeptides in E. chaffeensis and E. canis are highly homologous to C. ruminantium MAP2, they were distinguishable by numerous substitutions throughout the polypeptide chain. Overall, the MAP2 homologs of E. canis and E. chaffeensis were 83.4 and 84.4% identical, respectively, to MAP2 of C. ruminantium (Crystal Springs strain). These data show that MAP2 homologs are present in two Ehrlichia spp., E. canis and E. chaffeensis, but that there is variability throughout the nucleic acid and amino acid sequences that may allow the development of epitope-specific serodiagnostic assays.

Immunoassays demonstrating antigenic similarity between C. ruminantium and E. canis. C. ruminantium antiserum (from sheep 483) recognized an approximately 21-kDa protein in C. ruminantium and E. canis (Fig. 5a). More specifically, anti-recombinant MAP2 serum (from sheep 177) recognized MAP2 in C. ruminantium (lane A) and a MAP2 homolog in E. canis (lane B). Anti-E. canis serum also recognized MAP2 of C. ruminantium and E. canis (Fig. 5). These data confirm the sequence data and show that a MAP2 homolog which shares epitopes with MAP2 of C. ruminantium is expressed in E. canis.

View larger version (17K): [in this window] [in a new window]   FIG. 5.   (a) Immunoblot demonstrating antigenic similarity between C. ruminantium and E. canis. Lysates of C. ruminantium (lanes A) or E. canis (lanes B) were separated by SDS-PAGE and transferred to nitrocellulose membranes. Membranes were reacted with anti-C. ruminantium antiserum (sheep 483), anti-rMAP2 antiserum (sheep 177), or anti-E. canis antiserum (Mukanya). (b) Indirect ELISA using dog sera. Serum from a dog infected with E. canis (positive) or from a healthy dog (negative) was tested in an indirect ELISA against recombinant MAP2.

Indirect ELISA for heartwater diagnosis using MAP2. Sera from sheep were tested to determine whether recombinant MAP2 is recognized by both anti-C. ruminantium and false-positive sera, as observed with MAP1 (18). Sera obtained from five animals experimentally infected with C. ruminantium tested positive at dilutions as low as 1:3,200 (Fig. 6). False-positive sera obtained from four animals from heartwater-free regions of Zimbabwe (18) reacted in a pattern similar to that of true-positive sera (Fig. 6). Immunoblot-negative sera (n = 3) consistently gave optical density values lower than the true-positive and false-positive sera at similar dilutions down to 1:12,800. Hence, an indirect ELISA using the complete recombinant MAP2 does not discriminate between true- and false-positive sera.

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Indirect ELISA using sheep sera. Sera obtained from animals experimentally infected with C. ruminantium (positive), sera obtained from animals from a heartwater-free region but reactive to C. ruminantium antigens in an immunoblot (false positive), and sera obtained from animals with no previous exposure and negative in an immunoblot to C. ruminantium (negative) were tested in an indirect ELISA against recombinant MAP2.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

The isolation of genes encoding immunoreactive proteins of C. ruminantium may enable the development of diagnostic tests with improved sensitivity and specificity. The immunoreactive protein must be conserved among C. ruminantium strains and yet sufficiently different from the homologs present in related species. MAP1 has been tested in various assays (10, 13) and shown to cross-react with anti-Ehrlichia sera and with sera from known C. ruminantium-free animals (18). A truncated MAP1 protein, MAP1B, has improved specificity but still reacts with some anti-Ehrlichia sera (14, 15, 23). Recent data showing (i) the presence of three hypervariable regions in MAP1 in different strains of C. ruminantium (20), (ii) recognition of linear peptides representing these hypervariable regions (20) by sera from animals infected with C. ruminantium (19, 20), and (iii) the presence in E. canis and E. chaffeensis of MAP1 homologs which have sequences very similar to that of MAP1 of C. ruminantium, except in the hypervariable regions, suggest that MAP1 may not be an ideal diagnostic antigen.

MAP2 is an immunoreactive protein consistently recognized by sera from sheep, goats, and cattle infected with C. ruminantium (17). MAP2 is 55.5% identical to MSP5 of A. marginale. MSP5 is conserved in all the different isolates of A. marginale tested, and an epitope-specific assay using MSP5, which consistently identifies cattle both acutely and chronically infected with A. marginale, has been developed (16, 25). Examination of the MAP2 of five geographically diverse strains of C. ruminantium revealed a highly conserved protein with amino acid substitutions at only three positions (amino acid variability between 0 and 1.44%). This differs from the sequenced MAP1 of several strains, which was found to contain up to 10.0% variability in amino acid sequence (20). MAP2 meets two criteria for a suitable diagnostic antigen, immunoreactivity and conservation. However, further data showed reactivity between recombinant MAP2 and sera from known heartwater-free animals. Thus, like MAP1, MAP2 lacked specificity for a diagnostic test.

Because MAP2 reacted with false-positive sera, it was essential that we examine other closely related rickettsiae for the presence of MAP2 homologs in order to evaluate whether there is a basis for screening of anti-MAP2 monoclonal antibodies or the use of specific MAP2 regions to develop an epitope-specific assay. The existence of a MAP2 homolog in E. canis was demonstrated by immunoblotting total antigens of C. ruminantium and E. canis with an antiserum specific to recombinant MAP2.

The genes encoding MAP2 homologs in E. canis and E. chaffeensis were amplified and sequenced, and ORFs similar to the map2 gene were translated into amino acids. The resulting MAP2 homologs had amino acid similarities of 83.41% (for E. canis) and 84.39% (for E. chaffeensis) with MAP2 of C. ruminantium. MAP2 homologs in Ehrlichia species may hinder the development of adequate diagnosis for heartwater using MAP2. Although similarities exist between MAP2 homologs of C. ruminantium and Ehrlichia species, numerous sequence differences throughout the proteins were observed. The numerous substitutions found throughout MAP2 homologs in the different species suggest that it may be possible to define unique epitopes suitable for the diagnosis of either C. ruminantium or ehrlichial infections.

The initial discrepancy between the MAP2 homologs of C. ruminantium and Ehrlichia appears to be the first methionine. The second methionine of C. ruminantium (4 amino acids from the first methionine) aligns with the first methionine of the Ehrlichia species, suggesting that it is probably the initiator methionine in MAP2. This is further supported by alignment of MAP2 with MSP5 of A. marginale; MSP5 also begins at the second methionine of C. ruminantium. There is a predicted ribosomal binding site 5' to the second methionine but 3' to the first methionine (Fig. 3). Thus, it appears that the reported amino acid sequence of MAP2 (17) may be incorrect and should actually be shorter by 4 amino acids at the N terminus.

Like MSP5 of A. marginale, therefore, MAP2 may be useful in the diagnosis of animals infected with C. ruminantium if a similarly unique epitope can be defined. MAP2 is highly conserved among geographically different isolates and is recognized by sera from C. ruminantium-infected sheep, goats, and cattle. However, MAP2 is recognized antigenically by anti-E. canis antiserum, and false-positive sera and MAP2 homologs were identified from E. canis and E. chaffeensis; thus MAP2 as a whole protein lacks specificity for heartwater diagnosis. From this study, it was determined that development of a serodiagnostic assay using MAP2 may require the use of truncated peptides or monoclonal antibodies specific to unique sequences of MAP2 of C. ruminantium. Further, the availability of the sequences of MAP2 homologs in E. chaffeensis and E. canis may aid in the development of specific diagnostic tests for these species.