Identification of Novel Antigenic Proteins in a Complex Anaplasma marginale Outer Membrane Immunogen by Mass Spectrometry and Genomic Mappin
http://www.100md.com
感染与免疫杂志 2005年第12期
Department of Veterinary Microbiology and Pathology
Department of Chemistry, Washington State University, Pullman, Washington 99164
ABSTRACT
Immunization with purified Anaplasma marginale outer membranes induces complete protection against infection that is associated with CD4+ T-lymphocyte-mediated gamma interferon secretion and immunoglobulin G2 (IgG2) antibody titers. However, knowledge of the composition of the outer membrane immunogen is limited. Recent sequencing and annotation of the A. marginale genome predicts at least 62 outer membrane proteins (OMP), enabling a proteomic and genomic approach for identification of novel OMP by use of IgG serum antibody from outer membrane vaccinates. Outer membrane proteins were separated by two-dimensional electrophoresis, and proteins recognized by total IgG and IgG2 in immune sera of outer membrane-vaccinated cattle were detected by immunoblotting. Immunoreactive protein spots were excised and subjected to liquid chromatography-tandem mass spectrometry. A database search of the A. marginale genome identified 24 antigenic proteins that were predicted to be outer membrane, inner membrane, or membrane-associated proteins. These included the previously characterized surface-exposed outer membrane proteins MSP2, operon associated gene 2 (OpAG2), MSP3, and MSP5 as well as recently identified appendage-associated proteins. Among the 21 newly described antigenic proteins, 14 are annotated in the A. marginale genome and include type IV secretion system proteins, elongation factor Tu, and members of the MSP2 superfamily. The identification of these novel antigenic proteins markedly expands current understanding of the composition of the protective immunogen and provides new candidates for vaccine development.
INTRODUCTION
Bacteria of the genera Anaplasma and Ehrlichia cause acute infection in immunologically nave hosts and are major causes of tick-borne disease in animals and humans. Studies with both bovine and murine models support the induction of a type 1 immune response leading to antigen-specific CD4+ T-lymphocyte proliferation, gamma interferon (IFN-) production, and immunoglobulin G2 (IgG2) antibody directed against the pathogen surface as being necessary for protection (10, 16, 25, 26, 34, 38, 61). However, the critical targets of this protective immune response have not been definitively identified for any of the pathogens in these two genera.
In several studies with Anaplasma marginale, immunization of cattle with purified outer membranes has been shown to induce protection against high levels of rickettsemia and clinical disease, and in one study, complete protection against infection was attained (16, 58). This protection is associated with high IgG2 titers against the immunodominant major surface protein 2 (MSP2) (16) and CD4+ T-lymphocyte proliferative responses to MSP1, MSP2, and MSP3 (16) as well as MSP4 (W. C. Brown, unpublished observations). However, none of these individual proteins has induced protection equivalent to that of the complex outer membrane immunogen (1, 16, 43, 44, 48, 49, 58). This discrepancy in ability to induce protection suggests at least two possible explanations. The first is that the critical outer membrane protein (OMP) antigens have not yet been identified. The limited repertoire identified to date has been biased by the methods used, primarily surface chemistry labeling and screening with panels of monoclonal antibodies (MAb) (45, 48, 50). However, the isolation of IFN--secreting CD4+ T-lymphocyte clones that responded specifically to outer membranes but did not recognize any of the previously identified MSPs (MSP1 to MSP5) indicates the presence of additional T-cell epitopes on as-yet-unidentified OMPs (17). The second possibility is that a combination of OMPs, alone or in a membrane context, is needed to induce complete protection. Addressing either possibility requires a more comprehensive knowledge of the composition of the protective outer membrane immunogen.
The recent availability of the complete genome sequence of the St. Maries strain of A. marginale allows new approaches to identify previously unknown OMPs in the complex outer membrane immunogen. In the present work, we report a combined immunologic, proteomic, and genomic approach for identification of novel antigenic OMP. The separation of the complex immunogen by using two-dimensional (2D) electrophoresis followed by detection of proteins bound by total IgG and IgG2 from outer membrane-immunized cattle restricted the identification to antigens relevant to the induction of an immune response. This approach also permitted identification of subdominant antigens that were weakly reactive with immune sera. Immunoreactive proteins were excised and subjected to liquid chromatography-tandem mass spectrometry (LC-MS-MS) for definitive identification by mapping to the annotated genome.
MATERIALS AND METHODS
Preparation of A. marginale outer membranes. Two splenectomized calves, C942b1 and C986b1, were inoculated intravenously with 1.8 ml cryopreserved blood stabilate mixed with 4.0 ml normal bovine serum. The stabilate was prepared from calf C533b1 infected with the St. Maries, Idaho, strain of A. marginale (20). On day 21 postinfection, when the levels of rickettsemia reached 23% and 24% for calves C942b1 and C986b1, respectively, fresh blood was obtained from each calf and cryopreserved in liquid nitrogen as previously described (20) for preparation of outer membranes to be used for immunization and fractionation. A. marginale organisms were isolated from thawed infected erythrocytes as previously described, with minor modifications (47). Briefly, infected erythrocytes were washed at least five times with phosphate-buffered saline (PBS) and centrifuged at 30,000 x g for 30 min after each wash. Infected erythrocytes were resuspended in PBS containing Complete Mini protease inhibitors by following the manufacturer's instructions (Roche, Penzberg, Germany). A. marginale organisms released from infected erythrocytes by sonication were centrifuged at 30,000 x g for 20 min and resuspended in 20% sucrose in 10 mM HEPES containing 50 μg per ml each of DNase I and RNase A. Outer membranes were prepared by sucrose density gradient centrifugation as previously described (58). Isolated A. marginale was disrupted by sonication, centrifuged at 1,500 x g for 15 min to pellet residual organisms, and layered onto the top of a sucrose step density gradient containing 2 ml of 32, 38, 44, 48, and 52% sucrose. Following centrifugation at 82,000 x g for 20 h at 4°C, three bands were visualized at 1.15 g per cm3, 1.17 g per cm3, and 1.22 g per cm3. The bands were collected and resuspended in ice-cold 10 mM HEPES buffer and centrifuged at 177,000 x g at 4°C for 1 h. This was repeated twice, and after a final wash, outer membrane pellets were either resuspended in PBS with 5 μg per ml gentamicin for use in immunization or processed through a ReadyPrep 2D cleanup kit (Bio-Rad, Hercules, Calif.) prior to 2D electrophoresis.
Outer membrane fractionation and immunoblotting. Infected erythrocytes, A.marginale organisms, and purified outer membrane fractions were solubilized and denatured by being boiled for 5 min in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (0.025 mM Tris-HCl [pH 6.8], 2% SDS, 15% glycerol, 2.5% -mercaptoethanol, 0.02% bromophenol blue), and serial dilutions ranging from 0.008 to 1 μg protein were applied to a 4 to 20% gel. Following electrophoresis, proteins were transferred to nitrocellulose membranes by using a Criterion blotter (Bio-Rad). Membranes were then blocked overnight in I-Block blocking reagent (Applied Biosystems, Bedford, Mass.) containing 0.5% Tween 20 and probed for 1 h with either MSP5-specific MAb ANAF16C1 (60) or anti-erythrocyte MAb ANA8A, diluted to 2 μg per ml. Immunoreactivity was detected using goat anti-mouse secondary antibody from Western-Star chemiluminescent immunoblot detection systems (Applied Biosystems, Foster City, Calif.) according to the manufacturer's specifications.
Immunoblotting was performed similarly to detect MSP1a in the outer membrane fraction. Either 5 μg uninfected erythrocyte membranes or 1.25 and 2.5 μg outer membranes were electrophoresed, transferred to nitrocellulose, reacted with 2 μg per ml MAb ANA22B1 specific for MSP1a (4), and developed as described above.
Immunization. Three 7-month-old neutered male Holstein calves, designated 04B90, 04B91, and 04B92, with bovine lymphocyte antigen (BoLA) DRB3 haplotypes common within Holstein-Friesian breeds (54), were selected. DRB3 alleles were defined by PCR restriction fragment length polymorphism analysis of exon 2 (59). Sequencing of the BoLA DQA genes was performed as previously described (51). The nomenclature of bovine class II genes can be found at http://www.projects.roslin.ac.uk/bola and http://www.ebi.ac.uk/ipd/mhc/bola. BoLA DRB3 and DQA haplotypes for the calves in this study are as follows: for calf 04B90, DRB3 1101/1501, DQA 10011/2206, and DQB 10011/22021; for calf 04B91, DRB3 1201/2703, DQA 12011/2201, and DQB 0101/22031; and for calf 04B92, DRB3 0201/1201, DQA 0203, and DQB 12011/2201.
Calves were immunized four times with outer membranes from the 1.22-g-per-cm3 fraction of the sucrose density gradient described above. Briefly, each calf received a subcutaneous inoculation at weeks 0, 2, 4, and 8 of 60 μg total protein of membranes resuspended in 1.3 ml PBS containing 6 mg saponin (16, 58). The antibody titer following immunization was determined by immunoblotting using pre- and postimmunization sera as previously described, with minor modifications (16). Briefly, 150 μg of whole bacteria or uninfected erythrocytes (negative control) was solubilized in SDS-PAGE sample buffer, electrophoresed in a 10 to 20% Tris-HCl Criterion precast gel (Bio-Rad), and then transferred to a nitrocellulose membrane. After membranes were incubated overnight at 4°C in I-Block reagent, preimmune and immune sera from calves 04B90, 04B91, and 04B92 were diluted to 1:200, 1:1,000, and 1:5,000 in I-Block reagent and incubated with the membranes for 1 h, followed by a 1-h incubation with a 1:3,000 dilution of horseradish peroxidase (HRP)-conjugated goat anti-bovine IgG antibody (Kirkegaard and Perry, Gaithersburg, Md.). Immunoblots were washed with I-Block reagent with 0.5% Tween 20 for 1 h, and antigen-antibody binding was detected by chemiluminescence using an ECL Western blotting detection system (Amersham Biosciences, Piscataway, N.J.) according to the manufacturer's instructions.
2D electrophoresis and immunoblotting. Purified outer membranes in the 1.22-g-per-cm3 density band from sucrose density centrifugation as described above were processed with a ReadyPrep 2D cleanup kit (Bio-Rad) and solubilized in 5 M urea, 2 M thiourea, 4% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 0.5% Bio-Lyte 3/10 ampholytes (Bio-Rad), 2 mM tributyl phosphine reducing agent, and 2% 3-(decyldimethyl-ammonio) propanesulfonate (SB3-10). Protein determinations were performed using a Quick Start Bradford (10a) protein assay (Bio-Rad) according to manufacturer specifications. Isoelectric focusing (IEF) was performed using a Protean IEF cell system (Bio-Rad) and 11-cm immobilized pH gradient (IPG) strips, pH 5 to 8, according to manufacturer specifications. Briefly, strips were rehydrated with 150 μg of solubilized outer membranes and focused for 30,000 V · h. The second dimension was performed on 10 to 20% SDS-PAGE gradient gels (Bio-Rad) using a Criterion cell system (Bio-Rad), per the manufacturer's instructions. Protein gels were stained with SYPRO Ruby (Bio-Rad) or transferred to a nitrocellulose membrane by using a Criterion blotter (Bio-Rad), blocked overnight in I-Block reagent with 0.5% Tween 20, and probed with immune serum from calf 04B90, 04B91, or 04B92 diluted at 1:200 in I-Block reagent. Antigen binding was detected with a 1:3,000 dilution of HRP-conjugated goat anti-bovine IgG antibody or HPR-conjugated sheep anti-bovine IgG2 antibody (Serotec, Raleigh, N.C.) followed by chemiluminescence as described above. The 2D gel and 2D immunoblot images were overlaid with aid from the 2D imaging software PDQuest (Bio-Rad), and corresponding spots were manually picked and prepared for LC-MS-MS.
In-gel trypsin digestion, LC-MS-MS, and A. marginale database search. For LC-MS-MS analysis, IgG-reactive protein spots were excised from SYPRO Ruby-stained gels and processed as previously described, with minor modifications (55). Individual protein spots were destained overnight in 10% methanol and 7% acetic acid at room temperature and then in 200 μl of high-pressure-liquid-chromatography-grade acetonitrile (ACN) (CH3CN) for two 5-min washes. ACN was completely evaporated by vacuum centrifugation, and gel pieces were reduced in 10 mM dithiothreitol for 30 min, alkylated in 50 mM iodoacetamide for 30 min, dehydrated with two 5-min incubations in ACN, rehydrated for 10 min in 100 mM ammonium bicarbonate, dehydrated again with two 5-min incubations in ACN, and completely dried by vacuum centrifugation. Gel pieces were rehydrated for 10 min in 20 ng per μl ice-cold sequencing-grade modified porcine trypsin (Promega, Madison, Wisc.) solution (solubilized in 50 mM ammonium bicarbonate) and overnight in 50 mM ammonium bicarbonate at 37°C with continuous shaking. Following centrifugation, the protein was extracted from individual gel spots for 10 min with 60 μl 5% formic acid in 50% ACN and pooled with the corresponding overnight supernatant. The extraction was repeated and pooled with the previous extract, and samples were evaporated by vacuum centrifugation until <25 μl remained. Samples were then subjected to LC-MS-MS analysis by use of an Esquire HCT electrospray ion trap (Bruker Daltonics, Billerica, Mass.) and an LC Packings (Dionex, Sunnyvale, Calif.) Ultimate nano-high-pressure-liquid-chromatography system with a Famos microautosampler and a Swichos microcolumn switching module (Dionex). The trap column was a C18 PepMap, 300 μm by 1 mm, 5 μm (LC Packings), and the nanoanalytical column was a C18 PepMap, 75 μm by 150 mm, 3 μm, 100 (LC Packings). For separation, 0.1% TFA with 90% ACN as solvent A was used. The column was rinsed with 5% solvent A for 3 min, and separation was performed at a flow rate of 250 nl per min by using a binary gradient starting at 20% solvent A and rising to 70% in 42 min. After elution, the column was rinsed with 90% solvent A for 10 min and subsequently equilibrated with 5% solvent A for 10 min. Nanospray tips were made in house by etching fused silica capillaries (20 μm by 360 μm) with hydrofluoric acid. Data were acquired in the Bruker Auto(MS2) mode, using a repeated cycle of a single MS survey scan followed by two MS-MS scans of parent ions determined from a dynamically updated list. Batches of MS-MS fragment ion lists were submitted to our local MASCOT server (www.matrixscience.com) for comparison to the protein sequences of the St. Maries strain of A. marginale. Protein identifications were based on one or more tryptic peptides yielding MASCOT ion scores greater than 16, corresponding to a probability of 95% or greater that the peptide match is not a random event.
In silico analysis of identified proteins. To predict the cellular locations of immunostimulatory proteins, posttranslation and topology prediction programs were used, including a signal peptide prediction algorithm, SignalP (http://www.cbs.dtu.dk/services/SignalP/) (9), and a transmembrane prediction algorithm, transmembrane hidden Markov model (TMHMM) (http://www.cbs.dtu.dk/services/TMHMM-2.0/) (56). Furthermore, protein similarity searches were performed for protein products by using Basic Local Alignment Search Tool (BLAST) (http://www.ncbi.nlm.nih.gov/BLAST/) and the public NCBI database. As a guideline, proteins with BLAST values of <1e-20 were considered to be similar to proteins in other organisms. Where applicable, Vector NTI suite 9 (Invitrogen)was used to predict pIs.
RESULTS
Evaluation of isolated A. marginale outer membranes. Immunoblot analysis of the different sucrose density gradient bands of sonicated A. marginale reacted with MAb specific for outer membrane protein MSP5 (60) revealed that, as described previously (58), the 1.22-g-per-cm3 band was the most highly enriched for outer membrane proteins (data not shown). Immunoblot analysis of infected erythrocytes, whole bacteria, and the outer membrane-enriched fraction from the 1.22-g-per-cm3 sucrose density gradient band, using MAb specific for A. marginale MSP5, revealed a 125-fold enrichment of MSP5 in outer membranes compared to infected erythrocytes and a 25-fold enrichment of MSP5 in outer membranes compared to A. marginale whole bacteria (Fig. 1A). Immunoblot analysis of the same antigens by use of a bovine erythrocyte-specific MAb that detects an undefined high-molecular-weight protein indicated there was a 25-fold decrease in erythrocyte contamination compared to infected erythrocytes and to whole A. marginale (Fig. 1B).
To further verify the presence of outer membrane proteins in the fraction to be used to immunize cattle and to perform 2D electrophoresis, immunoblotting was performed to detect MSP1a (Fig. 2). MSP1a is present in purified outer membranes but not in membranes from uninfected erythrocytes.
Seroconversion of outer membrane vaccinates. Three calves with common major histocompatibility complex class II haplotypes were selected for this study based on DRB3 allelic frequencies that are expressed in a majority of Holstein and Friesian cattle (54). To evaluate antibody responses after immunization with outer membranes, immunoblotting of A. marginale with serially diluted pre- and postimmunization sera from the three calves was performed. Preimmune sera were unreactive with A. marginale antigens at all dilutions. However, postimmunization sera diluted 1:5,000 from all three animals reacted with A. marginale antigens ranging from approximately 15 to 250 kDa (data not shown). As a negative control, immunoblotting was performed similarly, using uninfected bovine erythrocytes, and pre- and postimmunization sera from all three animals were negative at all dilutions.
Identification of novel antigenic OMP. To identify antigenic OMP, 2D electrophoresis and immunoblotting were performed with sera from the outer membrane vaccinates. Immunoblots were developed with anti-IgG (Fig. 3A to C) or anti-IgG2 (data not shown) antibody. Spots of interest were defined by overlay analysis of the stained 2D gels and immunoblots. A total of 75 spots (Fig. 3D), which coincided with the image overlay analyses, were excised from the gels, trypsinized, and processed by LC-MS-MS to obtain molecular masses of peptide fragments. Mass spectral data from the 75 spots were searched against the annotated A. marginale St. Maries strain genome database, yielding strong identification from 38 spots which contained a total of 24 proteins (Table 1). Protein scores in Table 1 are the sums of nonredundant individual ion scores of identified peptides, as reported by MASCOT. The table also reports, for each protein, the number of distinct peptides identified by the MS-MS analysis. Multiple proteins were identified in spots 17, 22, 49, 73, and 74. Of the 24 proteins identified by LC-MS-MS, MSP2, MSP3, and MSP5 have previously been described as immunogenic in cattle (2, 3, 13, 14, 16, 17, 39, 48, 49, 58). Of the 21 remaining proteins, 15 (AM075, AM097, AM127, AM197, AM254, AM387, AM854, AM956, AM1096, AM1142, AM1164, AM1220, AM1223, AM1314, and AM1315) were recognized by IgG antibodies from all three animals, and 5 (AM072, AM529, AM878, AM879, and AM880) of the 21 proteins were recognized by IgG from only calf 04B90. One protein (AM366) was recognized by IgG from calves 04B91 and 04B92 (Fig. 3). Immunoblotting was also performed using immune sera from outer membrane vaccinates to detect IgG2 (data not shown). The same pattern of spots was revealed, but in some cases the reactivity was weaker than with total IgG. In addition, total IgG from calf 04B90 detected four additional spots, but not enough LC-MS-MS data were obtained for a positive score when searching the annotated A. marginale genome. Otherwise, there were no differences in the numbers of spots detected when anti-IgG2 or anti-IgG was used (data not shown). Figure 4 summarizes the antibody recognition by all three animals.
To predict the cellular location of the newly identified proteins, topology prediction algorithms, including TMHMM and posttranslational modification prediction algorithm SignalP, were used (Table 1). Of the 21 putative antigenic proteins, 10 were predicted by SignalP (9) to contain a signal peptide. Thirteen of the remaining proteins that were not predicted to have a signal peptide were predicted by TMHMM (56) to contain transmembrane domains (Table 1). A single predicted internal membrane protein, AM1314 (VirB10), was identified. Proteins predicted to be within the outer membrane included the undefined proteins AM197, AM366, AM529, AM854, AM127, and AM956. Previously defined proteins also included AM075 (OMP14), AM1096 (OMA87), AM1142 (OpAG2), AM1223 (OMP10), AM1315 (VirB9), AM1220 (OMP7), and AM1164 (OMP4) (Table 1). Membrane-associated proteins included the undefined products AM072 and AM387, as well as AM097 (conjugal transfer protein), AM254 (elongation factor Tu [EF-Tu]), and the protein family of AM878, AM879, and AM880 (Anaplasma appendage-associated proteins) (Table 1).
Protein BLAST database searches (http://www.ncbi.nlm.nih.gov/BLAST/) were performed on undefined proteins to determine identity with protein sequences deposited in the NCBI database since the annotation of the A. marginale genome (Table 2). BLAST searches were also performed on annotated proteins other than MSP2, MSP3, and MSP5, in order to determine sequence identity in related organisms (Table 2). Three undefined proteins (AM197, AM854, and AM127) and nine defined proteins (AM075, AM097, AM254, AM956, AM1096, AM1142, AM1164, AM1314, and AM1315) had significant BLAST scores to putative proteins in A. ovis, A. phagocytophilum, Ehrlichia canis, E. chaffeensis, or E. ruminantium (Table 2). The remaining proteins (AM072, AM366, AM387, AM529, AM878, AM879, AM880, AM1220, and AM1223) did not have a significant BLAST score to any putative protein.
DISCUSSION
This study used a combined immunologic, proteomic, and genomic approach to discover new antigenic proteins in the protection-inducing outer membrane immunogen of A. marginale. Evidence for a protective role of humoral immunity against rickettsial pathogens has been established (8, 24, 29, 33), and more specifically there is evidence for a protective role of OMP-specific IgG2 antibodies. Complete protection against homologous strain challenge in A. marginale outer membrane-immunized cattle was associated with high IgG2 titers against OMP (16). Furthermore, passive transfer of OMP-specific immune sera, comprised mainly of IgG2a, IgG2b, and IgG3 antibodies, from immunocompetent mice provided SCID mice with significant protection from infection with E. chaffeensis (34, 35). Therefore, a type 1 immune response regulated by IFN- that influences isotype switching to A. marginale OMP-specific IgG2 may be important to prevent bacterial attachment and penetration of host cells and to enhance phagocytosis and complement fixation (21, 40).
There have been several recent attempts to identify novel proteins in A. marginale by using various antigen preparations (6, 53). In one study, cattle were immunized with membrane proteins fractionated by isoelectric point. Five proteins were identified in a fraction containing proteins with pIs ranging from pH 7.7 to 9.6, which induced partial protective immunity. These novel proteins were designated Ana43, Ana37, Ana32, Ana29, and Ana17 based on their molecular masses in kDa (53). Of these, sequence analysis of Ana29 predicted this to be an inner membrane protein (30), and partial amino acid sequencing of Ana32 identified this protein as OpAG3 (11, 53). In a separate study, proteins from A. marginale were separated by one- or two-dimensional electrophoresis and immunoblots were probed with sera from cattle that had been immunized with A. marginale and then challenged, in order to identify IgG2-reactive proteins (6). Using mouse antibodies specific for MSP1a, MSP2/MSP3, or MSP5, the investigators identified these proteins on one-dimensional gel immunoblots. Immune bovine serum detected proteins of similar sizes on both one- and two-dimensional gel immunoblots, including numerous proteins that comigrated with proteins of the same size as MSP2, as well as unknown proteins designated Am56, Am50, Am38, and Am28, based on their molecular masses in kDa (6). However, neither amino acid nor encoding gene sequences were identified for any of these proteins. Thus, in this study, the recognition of MSP1a, MSP2, MSP3, and MSP5 by postimmunization and challenge bovine serum was tentatively based on apparent molecular masses of immunoreactive proteins (6).
In the present study, a strategy was implemented to identify antigenic proteins in outer membrane preparations which have been shown to induce protection against homologous strain challenge in cattle (16, 58). The annotated A. marginale genome (11) provided a searchable database that was used to identify the gene sequences of 24 proteins based on LC-MS-MS analysis of immunoreactive spots. Among these were the well-characterized proteins MSP2, MSP3, and MSP5 (2, 3, 13, 14, 17, 46, 48, 60). Of the 75 spots selected, 16 were identified as MSP2, consistent with the high abundance of this protein in the outer membrane (43). The large number of MSP2 proteins reflects the generation during acute and persistent infection of antigenic variants of MSP2 that vary slightly in size (23). In a recent study, we identified nine different msp2 transcripts in acute-infection blood from several challenged cattle (1), which is consistent with the identification of multiple MSP2 protein spots by 2D electrophoresis and immunoblotting. Antigenic variation in MSP2 occurs by gene conversion of whole pseudogenes and small segments of pseudogenes into a single expression site (5, 12), resulting in the expression of proteins that have conserved amino- and carboxy-terminal regions and a central hypervariable region. Thus, the different variants would be expected to have similar molecular masses and different pIs, consistent with our data and those of Riding et al. (53). All of the 16 MSP2 spots yielded peptide sequences of VEVEVGYER (amino acids [aa] 130 to 138), AVEGAEVIE (aa 279 to 287), and/or DTGIASFNFAYFGGELGVR (aa 388 to 406), from the highly conserved amino or carboxy regions of MSP2, where relative amino acid sequence position is based on the msp2 11.2 genomic clone (46). MSP1a and MSP1b (43, 44) were not identified, perhaps because these proteins were not sufficiently abundant to be detected by LC-MS-MS or because the predicted pI of the A. marginale St. Maries strain MSP1a is 4.99 and those of MSP1b1 and MSP1b2 are 4.78 and 4.87, which would fall just outside of the range of the pH 5 to 8 gradient used for IEF. The presence of MSP1a in the outer membrane preparation used for 2D electrophoresis was verified by immunoblotting whole outer membranes (Fig. 2). Furthermore, CD4+ T cells from the vaccinated cattle responded specifically to native MSP1 protein and to selected peptides of MSP1a (unpublished observations). Additionally, while we were able to identify MSP4 by LC-MS-MS analysis of spots selected by molecular mass (data not shown), MSP4 was surprisingly not reactive with immune sera. However, it has been reported that MSP4 may not be very antigenic in cattle immunized with A. marginale (6).
Twenty-one of the proteins identified in our study had not previously been characterized immunologically, although two of these were reportedly expressed. The product of the OpAG2 gene of the msp2 operon was shown to be expressed on the surface of A. marginale during infection of bovine erythrocytes as well as within tick tissue (36). Also, an Anaplasma appendage-associated protein localizes with actin filaments on the surface of A. marginale during intracellular infection (57). However, the present study is the first report that either of these proteins induces an IgG response in outer membrane-immunized cattle. Of the remaining 19 newly identified proteins, 7 were undefined, in that they were annotated in the genome as putative proteins but had no known function. BLAST searches were performed on these undefined proteins to identify homologous sequences in other organisms that had been deposited in the GenBank database since the A. marginale genome was annotated. Two sequences had identity to hypothetical proteins in E. canis (AM197 and AM854) and one sequence had identity to a hypothetical protein in E. chaffeensis (AM127), while AM072, AM366, AM387, and AM529 did not have sequence similarity to any published proteins.
Potential functions of the remaining proteins were determined by comparison with known proteins in other bacteria. Proteins with designated functions include VirB9, VirB10, conjugal transfer protein, EF-Tu, and PepA cytosol amino peptidase. VirB9, VirB10, and conjugal transfer protein are considered part of the type IV secretion system (TFSS), which mediates secretion or cell-to-cell transfer of macromolecules, proteins, or DNA-protein complexes in gram-negative bacteria, such as Bordetella pertussis, Legionella pneumophila, Rickettsia prowazekii, Ehrlichia chaffeensis, and Anaplasma phagocytophilum (7, 18, 42). While the role of the TFSS in A. marginale has not been defined, the virB operon in Brucella suis was shown to be required for intracellular survival after phagocytosis (22, 41) and may therefore be critical to A. marginale survival in erythrocytes.
The identification of PepA cytosol amino peptidase and EF-Tu was somewhat unexpected because both are believed to be cytoplasmic proteins. PepA cytosol amino peptidase was so designated by genomic annotation (11) but by TMHMM was predicted to be anchored through the outer membrane and exposed to the outside environment. In gram-positive Staphylococcus epidermidis, this protein appears to be cell surface associated (52), but its location in gram-negative bacteria is unclear. EF-Tu was also predicted by TMHMM analysis to be surface associated, and the presence of EF-Tu in various compartments of bacteria has been previously documented. EF-Tu associates with the cell surface of Escherichia coli (32) and that of Lactobacillus johnsonii (28). Furthermore, in L. johnsonii, EF-Tu had properties of an adhesion factor, mediating attachment of bacteria to human intestinal cells (28). In Mycobacterium leprae and Mycobacterium pneumoniae, EF-Tu was identified as a cell wall-associated protein (19, 37). Thus, in some organisms EF-Tu appears to function in the bacterial cytoplasm for protein synthesis, whereas in other organisms EF-Tu appears to localize to the surface, where it can function in bacterial attachment to host cells.
Proteins with unknown functions include OMP4, OMP7, OMP10, OMP14, and OMA87. With the exception of OMA87, these proteins were designated members of the MSP2 superfamily because of sequence identity to a family of surface antigens known as pfam01617 (11). However, their functions in A. marginale are unclear.
The approach of using LC-MS-MS analysis of immunoreactive protein spots excised from 2D gels has several limitations. One unpredicted result from this study was that only 52% of the protein spots submitted for LC-MS-MS were positively identified as A. marginale proteins. Those spots that did not yield a positive identification in the A. marginale genome were also searched against the NCBI database, and spots 1 to 3, 51, 52, 62 to 65, and 67, comprising 13% of the immunoreactive spots, were determined to be of bovine origin (data not shown). It is probable that contamination with bovine protein in these spots quenched any identification of Anaplasma proteins, as it is unlikely that cattle would develop an antibody response to self-antigens following immunization. In fact, sera from calves immunized with outer membranes did not react on immunoblots to uninfected erythrocytes (data not shown). The failure to positively identify the additional protein spots by LC-MS-MS may have resulted from poor solubilization of very hydrophobic outer membrane proteins and therefore incomplete trypsinization.
There were several examples of differences between actual and predicted molecular masses or pIs. These discrepancies could result from posttranslational modification (as is known to occur with MSP1a, which is glycosylated [4, 27]) or partial protein degradation. Furthermore, algorithms to predict pIs are not completely accurate. There were also examples of identifying the same protein in more than one spot. Finding MSP2 and MSP3 in multiple spots was expected, because multiple variants of these proteins, which can differ slightly in molecular masses and pIs, are present during infection (1-3, 13-15, 17, 43, 44, 48). The peptides identified in spots 73 and 74 were conserved peptides derived from the closely related appendage-associated protein family members, so that the analysis was unable to assign an individual member to an individual spot. Additionally, the presence of conjugal transfer protein in spots 7 and 8 and of EF-Tu in spots 43 to 46 and 49 may have resulted from aberrant migration on the gels if these proteins were complexed with their substrates. EF-Tu interacts with tRNA during protein synthesis, and high-affinity binding to tRNA might have impeded the removal of tRNA during outer membrane preparation. Similarly, as a component of the TFSS, conjugal transfer protein may mediate secretion or cell-to-cell transfer of macromolecules and protein-DNA complexes. Thus, residual DNA and/or RNA bound to these two proteins may have affected their mobilities.
The identification of novel antigens in the complex outer membrane immunogen has expanded the number of targets recognized by immune serum IgG and thus provided a number of vaccine candidates for further study. For an effective vaccine, it is important that these proteins can be recognized by serum antibody and CD4+ T lymphocytes from a majority of cattle within a vaccinated population upon infection and that these proteins are conserved among different strains of A. marginale. Animals selected for this study expressed common major histocompatibility complex class II DRB3 alleles that are present in over 50% of Holstein-Friesian cattle (54). Furthermore, the majority of the novel antigenic proteins were recognized by IgG2 from all three animals (Fig. 4), indicating the stimulation of T-helper-lymphocyte responses that promote switching to IgG isotypes. If protective, these vaccine candidates may be useful at the population level.
The availability of the sequences encoding these proteins from other A. marginale strains will enable determination of protein sequence conservation among the different strains. Additionally, identification of these antigenic proteins may offer insight to immunoreactive proteins in related pathogens, such as A. phagocytophilum, E. chaffeensis, E. canis, and E. ruminantium. A. phagocytophilum contains proteins highly similar to A. marginale MSP2 (23, 31, 62), and BLAST searches of the A. marginale proteins identified in this study indicate that several have sequence identity to proteins in A. phagocytophilumand species within the genus Ehrlichia. Thus, the identification of these immunogenic A. marginale proteins may also provide direction towards specific antigens for vaccine development against related animal and human pathogens.
ACKNOWLEDGMENTS
We thank Kim Kegerreis, Shelly Whidbee, Beverly Hunter, Emma Karel, Ralph Horn, Maggie Tam, and Xioting Tang for excellent technical assistance and Don Knowles for providing PDquest 2D imaging software.
This research was supported by an American Society for Microbiology Watkins Minority Fellowship and by NIH NIAID grant R01-AI053692.
REFERENCES
1. Abbott, J. R., G. H. Palmer, K. A. Kegerreis, P. F. Hetrick, C. J. Howard, J. C. Hope, and W. C. Brown. 2005. Rapid and long-term disappearance of CD4+ T lymphocyte responses specific for Anaplasma marginale major surface protein-2 (MSP2) in MSP2 vaccinates following challenge with live A.marginale. J. Immunol. 174:6702-6715.
2. Alleman, A. R., and A. F. Barbet. 1996. Evaluation of Anaplasma marginale major surface protein 3 (MSP3) as a diagnostic test antigen. J. Clin. Microbiol. 34:270-276.
3. Alleman, A. R., G. H. Palmer, T. C. McGuire, T. F. McElwain, L. E. Perryman, and A. F. Barbet. 1997. Anaplasma marginale major surface protein 3 is encoded by a polymorphic, multigene family. Infect. Immun. 65:156-163.
4. Allred, D. R., T. C. McGuire, G. H. Palmer, S. R. Leib, T. M. Harkins, T. F. McElwain, and A. F. Barbet. 1990. Molecular basis for surface antigen size polymorphisms and conservation of a neutralization-sensitive epitope in Anaplasma marginale. Proc. Natl. Acad. Sci. USA 87:3220-3224.
5. Barbet, A. F., A. Lundgren, J. Yi, F. R. Rurangirwa, and G. H. Palmer. 2000. Antigenic variation of Anaplasma marginale by expression of MSP2 mosaics. Infect. Immun. 68:6133-6138.
6. Barigye, R., M. A. Garcia-Ortiz, E. E. Rojas Ramirez, and S. D. Rodriguez. 2004. Identification of IgG2-specific antigens in Mexican Anaplasma marginale strains. Ann. N. Y. Acad. Sci. 1026:84-94.
7. Baron, C., D. O'Callaghan, and E. Lanka. 2002. Bacterial secrets of secretion: EuroConference on the biology of type IV secretion processes. Mol. Microbiol. 43:1359-1365.
8. Beaman, L., and C. L. Wisseman, Jr. 1976. Mechanisms of immunity in typhus infections. VI. Differential opsonizing and neutralizing action of human typhus rickettsia-specific cytophilic antibodies in cultures of human macrophages. Infect. Immun. 14:1071-1076.
9. Bendtsen, J. D., H. Nielsen, G. von Heijne, and S. Brunak. 2004. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340:783-795.
10. Bitsaktsis, C., J. Huntington, and G. Winslow. 2004. Production of IFN- by CD4 T cells is essential for resolving ehrlichia infection. J. Immunol. 172:6894-6901.
10. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.
11. Brayton, K. A., L. S. Kappmeyer, D. R. Herndon, M. J. Dark, D. L. Tibbals, G. H. Palmer, T. C. McGuire, and D. P. Knowles, Jr. 2005. Complete genome sequencing of Anaplasma marginale reveals that the surface is skewed to two superfamilies of outer membrane proteins. Proc. Natl. Acad. Sci. USA 102:844-849.
12. Brayton, K. A., G. H. Palmer, A. Lundgren, J. Yi, and A. F. Barbet. 2002. Antigenic variation of Anaplasma marginale msp2 occurs by combinatorial gene conversion. Mol. Microbiol. 43:1151-1159.
13. Brown, W. C., T. C. McGuire, D. Zhu, H. A. Lewin, J. Sosnow, and G. H. Palmer. 2001. Highly conserved regions of the immunodominant major surface protein 2 of the genogroup II ehrlichial pathogen Anaplasma marginale are rich in naturally derived CD4+ T lymphocyte epitopes that elicit strong recall responses. J. Immunol. 166:1114-1124.
14. Brown, W. C., G. H. Palmer, K. A. Brayton, P. F. Meeus, A. F. Barbet, K. A. Kegerreis, and T. C. McGuire. 2004. CD4+ T lymphocytes from Anaplasma marginale major surface protein 2 (MSP2) vaccinees recognize naturally processed epitopes conserved in MSP3. Infect. Immun. 72:3688-3692.
15. Brown, W. C., G. H. Palmer, H. A. Lewin, and T. C. McGuire. 2001. CD4+ T lymphocytes from calves immunized with Anaplasma marginale major surface protein 1 (MSP1), a heteromeric complex of MSP1a and MSP1b, preferentially recognize the MSP1a carboxyl terminus that is conserved among strains. Infect. Immun. 69:6853-6862.
16. Brown, W. C., V. Shkap, D. Zhu, T. C. McGuire, W. Tuo, T. F. McElwain, and G. H. Palmer. 1998. CD4+ T-lymphocyte and immunoglobulin G2 responses in calves immunized with Anaplasma marginale outer membranes and protected against homologous challenge. Infect. Immun. 66:5406-5413.
17. Brown, W. C., D. Zhu, V. Shkap, T. C. McGuire, E. F. Blouin, K. M. Kocan, and G. H. Palmer. 1998. The repertoire of Anaplasma marginale antigens recognized by CD4+ T-lymphocyte clones from protectively immunized cattle is diverse and includes major surface protein 2 (MSP-2) and MSP-3. Infect. Immun. 66:5414-5422.
18. Christie, P. J., and J. P. Vogel. 2000. Bacterial type IV secretion: conjugation systems adapted to deliver effector molecules to host cells. Trends Microbiol. 8:354-360.
19. Dallo, S. F., T. R. Kannan, M. W. Blaylock, and J. B. Baseman. 2002. Elongation factor Tu and E1 beta subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol. Microbiol. 46:1041-1051.
20. Eriks, I. S., D. Stiller, W. L. Goff, M. Panton, S. M. Parish, T. F. McElwain, and G. H. Palmer. 1994. Molecular and biological characterization of a newly isolated Anaplasma marginale strain. J. Vet. Diagn. Investig. 6:435-441.
21. Estes, D. M., N. M. Closser, and G. K. Allen. 1994. IFN-gamma stimulates IgG2 production from bovine B cells costimulated with anti-mu and mitogen. Cell. Immunol. 154:287-295.
22. Foulongne, V., G. Bourg, C. Cazevieille, S. Michaux-Charachon, and D. O'Callaghan. 2000. Identification of Brucella suis genes affecting intracellular survival in an in vitro human macrophage infection model by signature-tagged transposon mutagenesis. Infect. Immun. 68:1297-1303.
23. French, D. M., W. C. Brown, and G. H. Palmer. 1999. Emergence of Anaplasma marginale antigenic variants during persistent rickettsemia. Infect. Immun. 67:5834-5840.
24. Gambrill, M. R., and C. L. Wisseman, Jr. 1973. Mechanisms of immunity in typhus infections. III. Influence of human immune serum and complement on the fate of Rickettsia mooseri within human macrophages. Infect. Immun. 8:631-640.
25. Ganta, R. R., C. Cheng, M. J. Wilkerson, and S. K. Chapes. 2004. Delayed clearance of Ehrlichia chaffeensis infection in CD4+ T-cell knockout mice. Infect. Immun. 72:159-167.
26. Ganta, R. R., M. J. Wilkerson, C. Cheng, A. M. Rokey, and S. K. Chapes. 2002. Persistent Ehrlichia chaffeensis infection occurs in the absence of functional major histocompatibility complex class II genes. Infect. Immun. 70:380-388.
27. Garcia-Garcia, J. C., J. de la Fuente, G. Bell-Eunice, E. F. Blouin, and K. M. Kocan. 2004. Glycosylation of Anaplasma marginale major surface protein 1a and its putative role in adhesion to tick cells. Infect. Immun. 72:3022-3030.
28. Granato, D., G. E. Bergonzelli, R. D. Pridmore, L. Marvin, M. Rouvet, and I. E. Corthesy-Theulaz. 2004. Cell surface-associated elongation factor Tu mediates the attachment of Lactobacillus johnsonii NCC533 (La1) to human intestinal cells and mucins. Infect. Immun. 72:2160-2169.
29. Hanson, B. A. 1983. Effect of immune serum on infectivity of Rickettsia tsutsugamushi. Infect. Immun. 42:341-349.
30. Hope, M., G. Riding, M. Menzies, and P. Willadsen. 2004. A novel antigen from Anaplasma marginale: characterization, expression and preliminary evaluation of the recombinant protein. Vaccine 22:407-415.
31. Ijdo, J. W., W. Sun, Y. Zhang, L. A. Magnarelli, and E. Fikrig. 1998. Cloning of the gene encoding the 44-kilodalton antigen of the agent of human granulocytic ehrlichiosis and characterization of the humoral response. Infect. Immun. 66:3264-3269.
32. Jacobson, G. R., and J. P. Rosenbusch. 1976. Abundance and membrane association of elongation factor Tu in E. coli. Nature 261:23-26.
33. Kaylor, P. S., T. B. Crawford, T. F. McElwain, and G. H. Palmer. 1991. Passive transfer of antibody to Ehrlichia risticii protects mice from ehrlichiosis. Infect. Immun. 59:2058-2062.
34. Li, J. S., F. Chu, A. Reilly, and G. M. Winslow. 2002. Antibodies highly effective in SCID mice during infection by the intracellular bacterium Ehrlichia chaffeensis are of picomolar affinity and exhibit preferential epitope and isotype utilization. J. Immunol. 169:1419-1425.
35. Li, J. S., E. Yager, M. Reilly, C. Freeman, G. R. Reddy, A. A. Reilly, F. K. Chu, and G. M. Winslow. 2001. Outer membrane protein-specific monoclonal antibodies protect SCID mice from fatal infection by the obligate intracellular bacterial pathogen Ehrlichia chaffeensis. J. Immunol. 166:1855-1862.
36. Lhr, C. V., K. A. Brayton, V. Shkap, T. Molad, A. F. Barbet, W. C. Brown, and G. H. Palmer. 2002. Expression of Anaplasma marginale major surface protein 2 operon-associated proteins during mammalian and arthropod infection. Infect. Immun. 70:6005-6012.
37. Marques, M. A., S. Chitale, P. J. Brennan, and M. C. Pessolani. 1998. Mapping and identification of the major cell wall-associated components of Mycobacterium leprae. Infect. Immun. 66:2625-2631.
38. Martin, M. E., K. Caspersen, and J. S. Dumler. 2001. Immunopathology and ehrlichial propagation are regulated by interferon-gamma and interleukin-10 in a murine model of human granulocytic ehrlichiosis. Am. J. Pathol. 158:1881-1888.
39. McGuire, T. C., W. C. Davis, A. L. Brassfield, T. F. McElwain, and G. H. Palmer. 1991. Identification of Anaplasma marginale long-term carrier cattle by detection of serum antibody to isolated MSP-3. J. Clin. Microbiol. 29:788-793.
40. McGuire, T. C., A. J. Musoke, and T. Kurtti. 1979. Functional properties of bovine IgG1 and IgG2: interaction with complement, macrophages, neutrophils and skin. Immunology 38:249-256.
41. O'Callaghan, D., C. Cazevieille, A. Allardet-Servent, M. L. Boschiroli, G. Bourg, V. Foulongne, P. Frutos, Y. Kulakov, and M. Ramuz. 1999. A homologue of the Agrobacterium tumefaciens VirB and Bordetella pertussis Ptl type IV secretion systems is essential for intracellular survival of Brucella suis. Mol. Microbiol. 33:1210-1220.
42. Ohashi, N., N. Zhi, Q. Lin, and Y. Rikihisa. 2002. Characterization and transcriptional analysis of gene clusters for a type IV secretion machinery in human granulocytic and monocytic ehrlichiosis agents. Infect. Immun. 70:2128-2138.
43. Palmer, G. H., A. F. Barbet, G. H. Cantor, and T. C. McGuire. 1989. Immunization of cattle with the MSP-1 surface protein complex induces protection against a structurally variant Anaplasma marginale isolate. Infect. Immun. 57:3666-3669.
44. Palmer, G. H., A. F. Barbet, W. C. Davis, and T. C. McGuire. 1986. Immunization with an isolate-common surface protein protects cattle against anaplasmosis. Science 231:1299-1302.
45. Palmer, G. H., A. F. Barbet, K. L. Kuttler, and T. C. McGuire. 1986. Detection of an Anaplasma marginale common surface protein present in all stages of infection. J. Clin. Microbiol. 23:1078-1083.
46. Palmer, G. H., G. Eid, A. F. Barbet, T. C. McGuire, and T. F. McElwain. 1994. The immunoprotective Anaplasma marginale major surface protein 2 is encoded by a polymorphic multigene family. Infect. Immun. 62:3808-3816.
47. Palmer, G. H., and T. C. McGuire. 1984. Immune serum against Anaplasma marginale initial bodies neutralizes infectivity for cattle. J. Immunol. 133:1010-1015.
48. Palmer, G. H., S. M. Oberle, A. F. Barbet, W. L. Goff, W. C. Davis, and T. C. McGuire. 1988. Immunization of cattle with a 36-kilodalton surface protein induces protection against homologous and heterologous Anaplasma marginale challenge. Infect. Immun. 56:1526-1531.
49. Palmer, G. H., F. R. Rurangirwa, K. M. Kocan, and W. C. Brown. 1999. Molecular basis for vaccine development against the ehrlichial pathogen Anaplasma marginale. Parasitol. Today 15:281-286.
50. Palmer, G. H., S. D. Waghela, A. F. Barbet, W. C. Davis, and T. C. McGuire. 1987. Characterization of a neutralization-sensitive epitope on the Am 105 surface protein of Anaplasma marginale. Int. J. Parasitol. 17:1279-1285.
51. Park, Y. H., Y. S. Joo, J. Y. Park, J. S. Moon, S. H. Kim, N. H. Kwon, J. S. Ahn, W. C. Davis, and C. J. Davies. 2004. Characterization of lymphocyte subpopulations and major histocompatibility complex haplotypes of mastitis-resistant and susceptible cows. J. Vet. Sci. 5:29-39.
52. Reis, M., M. Eschbach-Bludau, M. I. Iglesias-Wind, T. Kupke, and H.-G. Sahl. 1994. Producer immunity towards the lantibiotic Pep5: identification of the immunity gene pepI and localization and functional analysis of its gene product. Appl. Environ. Microbiol. 60:2876-2883.
53. Riding, G., M. Hope, D. Waltisbuhl, and P. Willadsen. 2003. Identification of novel protective antigens from Anaplasma marginale. Vaccine 21:1874-1883.
54. Sharif, S., B. A. Mallard, B. N. Wilkie, J. M. Sargeant, H. M. Scott, J. C. Dekkers, and K. E. Leslie. 1999. Associations of the bovine major histocompatibility complex DRB3 (BoLA-DRB3) with production traits in Canadian dairy cattle. Anim. Genet. 30:157-160.
55. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68:850-858.
56. Sonnhammer, E. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol. 6:175-182.
57. Stich, R. W., G. A. Olah, K. A. Brayton, W. C. Brown, M. Fechheimer, K. Green-Church, S. Jittapalapong, K. M. Kocan, T. C. McGuire, F. R. Rurangirwa, and G. H. Palmer. 2004. Identification of a novel Anaplasma marginale appendage-associated protein that localizes with actin filaments during intraerythrocytic infection. Infect. Immun. 72:7257-7264.
58. Tebele, N., T. C. McGuire, and G. H. Palmer. 1991. Induction of protective immunity by using Anaplasma marginale initial body membranes. Infect. Immun. 59:3199-3204.
59. van Eijk, M. J., J. A. Stewart-Haynes, and H. A. Lewin. 1992. Extensive polymorphism of the BoLA-DRB3 gene distinguished by PCR-RFLP. Anim. Genet. 23:483-496.
60. Visser, E. S., T. C. McGuire, G. H. Palmer, W. C. Davis, V. Shkap, E. Pipano, and D. P. Knowles, Jr. 1992. The Anaplasma marginale msp5 gene encodes a 19-kilodalton protein conserved in all recognized Anaplasma species. Infect. Immun. 60:5139-5144.
61. Winslow, G. M., E. Yager, K. Shilo, E. Volk, A. Reilly, and F. K. Chu. 2000. Antibody-mediated elimination of the obligate intracellular bacterial pathogen Ehrlichia chaffeensis during active infection. Infect. Immun. 68:2187-2195.
62. Zhi, N., N. Ohashi, and Y. Rikihisa. 1999. Multiple p44 genes encoding major outer membrane proteins are expressed in the human granulocytic ehrlichiosis agent. J. Biol. Chem. 274:17828-17836.(Job E. Lopez, William F. )
Department of Chemistry, Washington State University, Pullman, Washington 99164
ABSTRACT
Immunization with purified Anaplasma marginale outer membranes induces complete protection against infection that is associated with CD4+ T-lymphocyte-mediated gamma interferon secretion and immunoglobulin G2 (IgG2) antibody titers. However, knowledge of the composition of the outer membrane immunogen is limited. Recent sequencing and annotation of the A. marginale genome predicts at least 62 outer membrane proteins (OMP), enabling a proteomic and genomic approach for identification of novel OMP by use of IgG serum antibody from outer membrane vaccinates. Outer membrane proteins were separated by two-dimensional electrophoresis, and proteins recognized by total IgG and IgG2 in immune sera of outer membrane-vaccinated cattle were detected by immunoblotting. Immunoreactive protein spots were excised and subjected to liquid chromatography-tandem mass spectrometry. A database search of the A. marginale genome identified 24 antigenic proteins that were predicted to be outer membrane, inner membrane, or membrane-associated proteins. These included the previously characterized surface-exposed outer membrane proteins MSP2, operon associated gene 2 (OpAG2), MSP3, and MSP5 as well as recently identified appendage-associated proteins. Among the 21 newly described antigenic proteins, 14 are annotated in the A. marginale genome and include type IV secretion system proteins, elongation factor Tu, and members of the MSP2 superfamily. The identification of these novel antigenic proteins markedly expands current understanding of the composition of the protective immunogen and provides new candidates for vaccine development.
INTRODUCTION
Bacteria of the genera Anaplasma and Ehrlichia cause acute infection in immunologically nave hosts and are major causes of tick-borne disease in animals and humans. Studies with both bovine and murine models support the induction of a type 1 immune response leading to antigen-specific CD4+ T-lymphocyte proliferation, gamma interferon (IFN-) production, and immunoglobulin G2 (IgG2) antibody directed against the pathogen surface as being necessary for protection (10, 16, 25, 26, 34, 38, 61). However, the critical targets of this protective immune response have not been definitively identified for any of the pathogens in these two genera.
In several studies with Anaplasma marginale, immunization of cattle with purified outer membranes has been shown to induce protection against high levels of rickettsemia and clinical disease, and in one study, complete protection against infection was attained (16, 58). This protection is associated with high IgG2 titers against the immunodominant major surface protein 2 (MSP2) (16) and CD4+ T-lymphocyte proliferative responses to MSP1, MSP2, and MSP3 (16) as well as MSP4 (W. C. Brown, unpublished observations). However, none of these individual proteins has induced protection equivalent to that of the complex outer membrane immunogen (1, 16, 43, 44, 48, 49, 58). This discrepancy in ability to induce protection suggests at least two possible explanations. The first is that the critical outer membrane protein (OMP) antigens have not yet been identified. The limited repertoire identified to date has been biased by the methods used, primarily surface chemistry labeling and screening with panels of monoclonal antibodies (MAb) (45, 48, 50). However, the isolation of IFN--secreting CD4+ T-lymphocyte clones that responded specifically to outer membranes but did not recognize any of the previously identified MSPs (MSP1 to MSP5) indicates the presence of additional T-cell epitopes on as-yet-unidentified OMPs (17). The second possibility is that a combination of OMPs, alone or in a membrane context, is needed to induce complete protection. Addressing either possibility requires a more comprehensive knowledge of the composition of the protective outer membrane immunogen.
The recent availability of the complete genome sequence of the St. Maries strain of A. marginale allows new approaches to identify previously unknown OMPs in the complex outer membrane immunogen. In the present work, we report a combined immunologic, proteomic, and genomic approach for identification of novel antigenic OMP. The separation of the complex immunogen by using two-dimensional (2D) electrophoresis followed by detection of proteins bound by total IgG and IgG2 from outer membrane-immunized cattle restricted the identification to antigens relevant to the induction of an immune response. This approach also permitted identification of subdominant antigens that were weakly reactive with immune sera. Immunoreactive proteins were excised and subjected to liquid chromatography-tandem mass spectrometry (LC-MS-MS) for definitive identification by mapping to the annotated genome.
MATERIALS AND METHODS
Preparation of A. marginale outer membranes. Two splenectomized calves, C942b1 and C986b1, were inoculated intravenously with 1.8 ml cryopreserved blood stabilate mixed with 4.0 ml normal bovine serum. The stabilate was prepared from calf C533b1 infected with the St. Maries, Idaho, strain of A. marginale (20). On day 21 postinfection, when the levels of rickettsemia reached 23% and 24% for calves C942b1 and C986b1, respectively, fresh blood was obtained from each calf and cryopreserved in liquid nitrogen as previously described (20) for preparation of outer membranes to be used for immunization and fractionation. A. marginale organisms were isolated from thawed infected erythrocytes as previously described, with minor modifications (47). Briefly, infected erythrocytes were washed at least five times with phosphate-buffered saline (PBS) and centrifuged at 30,000 x g for 30 min after each wash. Infected erythrocytes were resuspended in PBS containing Complete Mini protease inhibitors by following the manufacturer's instructions (Roche, Penzberg, Germany). A. marginale organisms released from infected erythrocytes by sonication were centrifuged at 30,000 x g for 20 min and resuspended in 20% sucrose in 10 mM HEPES containing 50 μg per ml each of DNase I and RNase A. Outer membranes were prepared by sucrose density gradient centrifugation as previously described (58). Isolated A. marginale was disrupted by sonication, centrifuged at 1,500 x g for 15 min to pellet residual organisms, and layered onto the top of a sucrose step density gradient containing 2 ml of 32, 38, 44, 48, and 52% sucrose. Following centrifugation at 82,000 x g for 20 h at 4°C, three bands were visualized at 1.15 g per cm3, 1.17 g per cm3, and 1.22 g per cm3. The bands were collected and resuspended in ice-cold 10 mM HEPES buffer and centrifuged at 177,000 x g at 4°C for 1 h. This was repeated twice, and after a final wash, outer membrane pellets were either resuspended in PBS with 5 μg per ml gentamicin for use in immunization or processed through a ReadyPrep 2D cleanup kit (Bio-Rad, Hercules, Calif.) prior to 2D electrophoresis.
Outer membrane fractionation and immunoblotting. Infected erythrocytes, A.marginale organisms, and purified outer membrane fractions were solubilized and denatured by being boiled for 5 min in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer (0.025 mM Tris-HCl [pH 6.8], 2% SDS, 15% glycerol, 2.5% -mercaptoethanol, 0.02% bromophenol blue), and serial dilutions ranging from 0.008 to 1 μg protein were applied to a 4 to 20% gel. Following electrophoresis, proteins were transferred to nitrocellulose membranes by using a Criterion blotter (Bio-Rad). Membranes were then blocked overnight in I-Block blocking reagent (Applied Biosystems, Bedford, Mass.) containing 0.5% Tween 20 and probed for 1 h with either MSP5-specific MAb ANAF16C1 (60) or anti-erythrocyte MAb ANA8A, diluted to 2 μg per ml. Immunoreactivity was detected using goat anti-mouse secondary antibody from Western-Star chemiluminescent immunoblot detection systems (Applied Biosystems, Foster City, Calif.) according to the manufacturer's specifications.
Immunoblotting was performed similarly to detect MSP1a in the outer membrane fraction. Either 5 μg uninfected erythrocyte membranes or 1.25 and 2.5 μg outer membranes were electrophoresed, transferred to nitrocellulose, reacted with 2 μg per ml MAb ANA22B1 specific for MSP1a (4), and developed as described above.
Immunization. Three 7-month-old neutered male Holstein calves, designated 04B90, 04B91, and 04B92, with bovine lymphocyte antigen (BoLA) DRB3 haplotypes common within Holstein-Friesian breeds (54), were selected. DRB3 alleles were defined by PCR restriction fragment length polymorphism analysis of exon 2 (59). Sequencing of the BoLA DQA genes was performed as previously described (51). The nomenclature of bovine class II genes can be found at http://www.projects.roslin.ac.uk/bola and http://www.ebi.ac.uk/ipd/mhc/bola. BoLA DRB3 and DQA haplotypes for the calves in this study are as follows: for calf 04B90, DRB3 1101/1501, DQA 10011/2206, and DQB 10011/22021; for calf 04B91, DRB3 1201/2703, DQA 12011/2201, and DQB 0101/22031; and for calf 04B92, DRB3 0201/1201, DQA 0203, and DQB 12011/2201.
Calves were immunized four times with outer membranes from the 1.22-g-per-cm3 fraction of the sucrose density gradient described above. Briefly, each calf received a subcutaneous inoculation at weeks 0, 2, 4, and 8 of 60 μg total protein of membranes resuspended in 1.3 ml PBS containing 6 mg saponin (16, 58). The antibody titer following immunization was determined by immunoblotting using pre- and postimmunization sera as previously described, with minor modifications (16). Briefly, 150 μg of whole bacteria or uninfected erythrocytes (negative control) was solubilized in SDS-PAGE sample buffer, electrophoresed in a 10 to 20% Tris-HCl Criterion precast gel (Bio-Rad), and then transferred to a nitrocellulose membrane. After membranes were incubated overnight at 4°C in I-Block reagent, preimmune and immune sera from calves 04B90, 04B91, and 04B92 were diluted to 1:200, 1:1,000, and 1:5,000 in I-Block reagent and incubated with the membranes for 1 h, followed by a 1-h incubation with a 1:3,000 dilution of horseradish peroxidase (HRP)-conjugated goat anti-bovine IgG antibody (Kirkegaard and Perry, Gaithersburg, Md.). Immunoblots were washed with I-Block reagent with 0.5% Tween 20 for 1 h, and antigen-antibody binding was detected by chemiluminescence using an ECL Western blotting detection system (Amersham Biosciences, Piscataway, N.J.) according to the manufacturer's instructions.
2D electrophoresis and immunoblotting. Purified outer membranes in the 1.22-g-per-cm3 density band from sucrose density centrifugation as described above were processed with a ReadyPrep 2D cleanup kit (Bio-Rad) and solubilized in 5 M urea, 2 M thiourea, 4% CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}, 0.5% Bio-Lyte 3/10 ampholytes (Bio-Rad), 2 mM tributyl phosphine reducing agent, and 2% 3-(decyldimethyl-ammonio) propanesulfonate (SB3-10). Protein determinations were performed using a Quick Start Bradford (10a) protein assay (Bio-Rad) according to manufacturer specifications. Isoelectric focusing (IEF) was performed using a Protean IEF cell system (Bio-Rad) and 11-cm immobilized pH gradient (IPG) strips, pH 5 to 8, according to manufacturer specifications. Briefly, strips were rehydrated with 150 μg of solubilized outer membranes and focused for 30,000 V · h. The second dimension was performed on 10 to 20% SDS-PAGE gradient gels (Bio-Rad) using a Criterion cell system (Bio-Rad), per the manufacturer's instructions. Protein gels were stained with SYPRO Ruby (Bio-Rad) or transferred to a nitrocellulose membrane by using a Criterion blotter (Bio-Rad), blocked overnight in I-Block reagent with 0.5% Tween 20, and probed with immune serum from calf 04B90, 04B91, or 04B92 diluted at 1:200 in I-Block reagent. Antigen binding was detected with a 1:3,000 dilution of HRP-conjugated goat anti-bovine IgG antibody or HPR-conjugated sheep anti-bovine IgG2 antibody (Serotec, Raleigh, N.C.) followed by chemiluminescence as described above. The 2D gel and 2D immunoblot images were overlaid with aid from the 2D imaging software PDQuest (Bio-Rad), and corresponding spots were manually picked and prepared for LC-MS-MS.
In-gel trypsin digestion, LC-MS-MS, and A. marginale database search. For LC-MS-MS analysis, IgG-reactive protein spots were excised from SYPRO Ruby-stained gels and processed as previously described, with minor modifications (55). Individual protein spots were destained overnight in 10% methanol and 7% acetic acid at room temperature and then in 200 μl of high-pressure-liquid-chromatography-grade acetonitrile (ACN) (CH3CN) for two 5-min washes. ACN was completely evaporated by vacuum centrifugation, and gel pieces were reduced in 10 mM dithiothreitol for 30 min, alkylated in 50 mM iodoacetamide for 30 min, dehydrated with two 5-min incubations in ACN, rehydrated for 10 min in 100 mM ammonium bicarbonate, dehydrated again with two 5-min incubations in ACN, and completely dried by vacuum centrifugation. Gel pieces were rehydrated for 10 min in 20 ng per μl ice-cold sequencing-grade modified porcine trypsin (Promega, Madison, Wisc.) solution (solubilized in 50 mM ammonium bicarbonate) and overnight in 50 mM ammonium bicarbonate at 37°C with continuous shaking. Following centrifugation, the protein was extracted from individual gel spots for 10 min with 60 μl 5% formic acid in 50% ACN and pooled with the corresponding overnight supernatant. The extraction was repeated and pooled with the previous extract, and samples were evaporated by vacuum centrifugation until <25 μl remained. Samples were then subjected to LC-MS-MS analysis by use of an Esquire HCT electrospray ion trap (Bruker Daltonics, Billerica, Mass.) and an LC Packings (Dionex, Sunnyvale, Calif.) Ultimate nano-high-pressure-liquid-chromatography system with a Famos microautosampler and a Swichos microcolumn switching module (Dionex). The trap column was a C18 PepMap, 300 μm by 1 mm, 5 μm (LC Packings), and the nanoanalytical column was a C18 PepMap, 75 μm by 150 mm, 3 μm, 100 (LC Packings). For separation, 0.1% TFA with 90% ACN as solvent A was used. The column was rinsed with 5% solvent A for 3 min, and separation was performed at a flow rate of 250 nl per min by using a binary gradient starting at 20% solvent A and rising to 70% in 42 min. After elution, the column was rinsed with 90% solvent A for 10 min and subsequently equilibrated with 5% solvent A for 10 min. Nanospray tips were made in house by etching fused silica capillaries (20 μm by 360 μm) with hydrofluoric acid. Data were acquired in the Bruker Auto(MS2) mode, using a repeated cycle of a single MS survey scan followed by two MS-MS scans of parent ions determined from a dynamically updated list. Batches of MS-MS fragment ion lists were submitted to our local MASCOT server (www.matrixscience.com) for comparison to the protein sequences of the St. Maries strain of A. marginale. Protein identifications were based on one or more tryptic peptides yielding MASCOT ion scores greater than 16, corresponding to a probability of 95% or greater that the peptide match is not a random event.
In silico analysis of identified proteins. To predict the cellular locations of immunostimulatory proteins, posttranslation and topology prediction programs were used, including a signal peptide prediction algorithm, SignalP (http://www.cbs.dtu.dk/services/SignalP/) (9), and a transmembrane prediction algorithm, transmembrane hidden Markov model (TMHMM) (http://www.cbs.dtu.dk/services/TMHMM-2.0/) (56). Furthermore, protein similarity searches were performed for protein products by using Basic Local Alignment Search Tool (BLAST) (http://www.ncbi.nlm.nih.gov/BLAST/) and the public NCBI database. As a guideline, proteins with BLAST values of <1e-20 were considered to be similar to proteins in other organisms. Where applicable, Vector NTI suite 9 (Invitrogen)was used to predict pIs.
RESULTS
Evaluation of isolated A. marginale outer membranes. Immunoblot analysis of the different sucrose density gradient bands of sonicated A. marginale reacted with MAb specific for outer membrane protein MSP5 (60) revealed that, as described previously (58), the 1.22-g-per-cm3 band was the most highly enriched for outer membrane proteins (data not shown). Immunoblot analysis of infected erythrocytes, whole bacteria, and the outer membrane-enriched fraction from the 1.22-g-per-cm3 sucrose density gradient band, using MAb specific for A. marginale MSP5, revealed a 125-fold enrichment of MSP5 in outer membranes compared to infected erythrocytes and a 25-fold enrichment of MSP5 in outer membranes compared to A. marginale whole bacteria (Fig. 1A). Immunoblot analysis of the same antigens by use of a bovine erythrocyte-specific MAb that detects an undefined high-molecular-weight protein indicated there was a 25-fold decrease in erythrocyte contamination compared to infected erythrocytes and to whole A. marginale (Fig. 1B).
To further verify the presence of outer membrane proteins in the fraction to be used to immunize cattle and to perform 2D electrophoresis, immunoblotting was performed to detect MSP1a (Fig. 2). MSP1a is present in purified outer membranes but not in membranes from uninfected erythrocytes.
Seroconversion of outer membrane vaccinates. Three calves with common major histocompatibility complex class II haplotypes were selected for this study based on DRB3 allelic frequencies that are expressed in a majority of Holstein and Friesian cattle (54). To evaluate antibody responses after immunization with outer membranes, immunoblotting of A. marginale with serially diluted pre- and postimmunization sera from the three calves was performed. Preimmune sera were unreactive with A. marginale antigens at all dilutions. However, postimmunization sera diluted 1:5,000 from all three animals reacted with A. marginale antigens ranging from approximately 15 to 250 kDa (data not shown). As a negative control, immunoblotting was performed similarly, using uninfected bovine erythrocytes, and pre- and postimmunization sera from all three animals were negative at all dilutions.
Identification of novel antigenic OMP. To identify antigenic OMP, 2D electrophoresis and immunoblotting were performed with sera from the outer membrane vaccinates. Immunoblots were developed with anti-IgG (Fig. 3A to C) or anti-IgG2 (data not shown) antibody. Spots of interest were defined by overlay analysis of the stained 2D gels and immunoblots. A total of 75 spots (Fig. 3D), which coincided with the image overlay analyses, were excised from the gels, trypsinized, and processed by LC-MS-MS to obtain molecular masses of peptide fragments. Mass spectral data from the 75 spots were searched against the annotated A. marginale St. Maries strain genome database, yielding strong identification from 38 spots which contained a total of 24 proteins (Table 1). Protein scores in Table 1 are the sums of nonredundant individual ion scores of identified peptides, as reported by MASCOT. The table also reports, for each protein, the number of distinct peptides identified by the MS-MS analysis. Multiple proteins were identified in spots 17, 22, 49, 73, and 74. Of the 24 proteins identified by LC-MS-MS, MSP2, MSP3, and MSP5 have previously been described as immunogenic in cattle (2, 3, 13, 14, 16, 17, 39, 48, 49, 58). Of the 21 remaining proteins, 15 (AM075, AM097, AM127, AM197, AM254, AM387, AM854, AM956, AM1096, AM1142, AM1164, AM1220, AM1223, AM1314, and AM1315) were recognized by IgG antibodies from all three animals, and 5 (AM072, AM529, AM878, AM879, and AM880) of the 21 proteins were recognized by IgG from only calf 04B90. One protein (AM366) was recognized by IgG from calves 04B91 and 04B92 (Fig. 3). Immunoblotting was also performed using immune sera from outer membrane vaccinates to detect IgG2 (data not shown). The same pattern of spots was revealed, but in some cases the reactivity was weaker than with total IgG. In addition, total IgG from calf 04B90 detected four additional spots, but not enough LC-MS-MS data were obtained for a positive score when searching the annotated A. marginale genome. Otherwise, there were no differences in the numbers of spots detected when anti-IgG2 or anti-IgG was used (data not shown). Figure 4 summarizes the antibody recognition by all three animals.
To predict the cellular location of the newly identified proteins, topology prediction algorithms, including TMHMM and posttranslational modification prediction algorithm SignalP, were used (Table 1). Of the 21 putative antigenic proteins, 10 were predicted by SignalP (9) to contain a signal peptide. Thirteen of the remaining proteins that were not predicted to have a signal peptide were predicted by TMHMM (56) to contain transmembrane domains (Table 1). A single predicted internal membrane protein, AM1314 (VirB10), was identified. Proteins predicted to be within the outer membrane included the undefined proteins AM197, AM366, AM529, AM854, AM127, and AM956. Previously defined proteins also included AM075 (OMP14), AM1096 (OMA87), AM1142 (OpAG2), AM1223 (OMP10), AM1315 (VirB9), AM1220 (OMP7), and AM1164 (OMP4) (Table 1). Membrane-associated proteins included the undefined products AM072 and AM387, as well as AM097 (conjugal transfer protein), AM254 (elongation factor Tu [EF-Tu]), and the protein family of AM878, AM879, and AM880 (Anaplasma appendage-associated proteins) (Table 1).
Protein BLAST database searches (http://www.ncbi.nlm.nih.gov/BLAST/) were performed on undefined proteins to determine identity with protein sequences deposited in the NCBI database since the annotation of the A. marginale genome (Table 2). BLAST searches were also performed on annotated proteins other than MSP2, MSP3, and MSP5, in order to determine sequence identity in related organisms (Table 2). Three undefined proteins (AM197, AM854, and AM127) and nine defined proteins (AM075, AM097, AM254, AM956, AM1096, AM1142, AM1164, AM1314, and AM1315) had significant BLAST scores to putative proteins in A. ovis, A. phagocytophilum, Ehrlichia canis, E. chaffeensis, or E. ruminantium (Table 2). The remaining proteins (AM072, AM366, AM387, AM529, AM878, AM879, AM880, AM1220, and AM1223) did not have a significant BLAST score to any putative protein.
DISCUSSION
This study used a combined immunologic, proteomic, and genomic approach to discover new antigenic proteins in the protection-inducing outer membrane immunogen of A. marginale. Evidence for a protective role of humoral immunity against rickettsial pathogens has been established (8, 24, 29, 33), and more specifically there is evidence for a protective role of OMP-specific IgG2 antibodies. Complete protection against homologous strain challenge in A. marginale outer membrane-immunized cattle was associated with high IgG2 titers against OMP (16). Furthermore, passive transfer of OMP-specific immune sera, comprised mainly of IgG2a, IgG2b, and IgG3 antibodies, from immunocompetent mice provided SCID mice with significant protection from infection with E. chaffeensis (34, 35). Therefore, a type 1 immune response regulated by IFN- that influences isotype switching to A. marginale OMP-specific IgG2 may be important to prevent bacterial attachment and penetration of host cells and to enhance phagocytosis and complement fixation (21, 40).
There have been several recent attempts to identify novel proteins in A. marginale by using various antigen preparations (6, 53). In one study, cattle were immunized with membrane proteins fractionated by isoelectric point. Five proteins were identified in a fraction containing proteins with pIs ranging from pH 7.7 to 9.6, which induced partial protective immunity. These novel proteins were designated Ana43, Ana37, Ana32, Ana29, and Ana17 based on their molecular masses in kDa (53). Of these, sequence analysis of Ana29 predicted this to be an inner membrane protein (30), and partial amino acid sequencing of Ana32 identified this protein as OpAG3 (11, 53). In a separate study, proteins from A. marginale were separated by one- or two-dimensional electrophoresis and immunoblots were probed with sera from cattle that had been immunized with A. marginale and then challenged, in order to identify IgG2-reactive proteins (6). Using mouse antibodies specific for MSP1a, MSP2/MSP3, or MSP5, the investigators identified these proteins on one-dimensional gel immunoblots. Immune bovine serum detected proteins of similar sizes on both one- and two-dimensional gel immunoblots, including numerous proteins that comigrated with proteins of the same size as MSP2, as well as unknown proteins designated Am56, Am50, Am38, and Am28, based on their molecular masses in kDa (6). However, neither amino acid nor encoding gene sequences were identified for any of these proteins. Thus, in this study, the recognition of MSP1a, MSP2, MSP3, and MSP5 by postimmunization and challenge bovine serum was tentatively based on apparent molecular masses of immunoreactive proteins (6).
In the present study, a strategy was implemented to identify antigenic proteins in outer membrane preparations which have been shown to induce protection against homologous strain challenge in cattle (16, 58). The annotated A. marginale genome (11) provided a searchable database that was used to identify the gene sequences of 24 proteins based on LC-MS-MS analysis of immunoreactive spots. Among these were the well-characterized proteins MSP2, MSP3, and MSP5 (2, 3, 13, 14, 17, 46, 48, 60). Of the 75 spots selected, 16 were identified as MSP2, consistent with the high abundance of this protein in the outer membrane (43). The large number of MSP2 proteins reflects the generation during acute and persistent infection of antigenic variants of MSP2 that vary slightly in size (23). In a recent study, we identified nine different msp2 transcripts in acute-infection blood from several challenged cattle (1), which is consistent with the identification of multiple MSP2 protein spots by 2D electrophoresis and immunoblotting. Antigenic variation in MSP2 occurs by gene conversion of whole pseudogenes and small segments of pseudogenes into a single expression site (5, 12), resulting in the expression of proteins that have conserved amino- and carboxy-terminal regions and a central hypervariable region. Thus, the different variants would be expected to have similar molecular masses and different pIs, consistent with our data and those of Riding et al. (53). All of the 16 MSP2 spots yielded peptide sequences of VEVEVGYER (amino acids [aa] 130 to 138), AVEGAEVIE (aa 279 to 287), and/or DTGIASFNFAYFGGELGVR (aa 388 to 406), from the highly conserved amino or carboxy regions of MSP2, where relative amino acid sequence position is based on the msp2 11.2 genomic clone (46). MSP1a and MSP1b (43, 44) were not identified, perhaps because these proteins were not sufficiently abundant to be detected by LC-MS-MS or because the predicted pI of the A. marginale St. Maries strain MSP1a is 4.99 and those of MSP1b1 and MSP1b2 are 4.78 and 4.87, which would fall just outside of the range of the pH 5 to 8 gradient used for IEF. The presence of MSP1a in the outer membrane preparation used for 2D electrophoresis was verified by immunoblotting whole outer membranes (Fig. 2). Furthermore, CD4+ T cells from the vaccinated cattle responded specifically to native MSP1 protein and to selected peptides of MSP1a (unpublished observations). Additionally, while we were able to identify MSP4 by LC-MS-MS analysis of spots selected by molecular mass (data not shown), MSP4 was surprisingly not reactive with immune sera. However, it has been reported that MSP4 may not be very antigenic in cattle immunized with A. marginale (6).
Twenty-one of the proteins identified in our study had not previously been characterized immunologically, although two of these were reportedly expressed. The product of the OpAG2 gene of the msp2 operon was shown to be expressed on the surface of A. marginale during infection of bovine erythrocytes as well as within tick tissue (36). Also, an Anaplasma appendage-associated protein localizes with actin filaments on the surface of A. marginale during intracellular infection (57). However, the present study is the first report that either of these proteins induces an IgG response in outer membrane-immunized cattle. Of the remaining 19 newly identified proteins, 7 were undefined, in that they were annotated in the genome as putative proteins but had no known function. BLAST searches were performed on these undefined proteins to identify homologous sequences in other organisms that had been deposited in the GenBank database since the A. marginale genome was annotated. Two sequences had identity to hypothetical proteins in E. canis (AM197 and AM854) and one sequence had identity to a hypothetical protein in E. chaffeensis (AM127), while AM072, AM366, AM387, and AM529 did not have sequence similarity to any published proteins.
Potential functions of the remaining proteins were determined by comparison with known proteins in other bacteria. Proteins with designated functions include VirB9, VirB10, conjugal transfer protein, EF-Tu, and PepA cytosol amino peptidase. VirB9, VirB10, and conjugal transfer protein are considered part of the type IV secretion system (TFSS), which mediates secretion or cell-to-cell transfer of macromolecules, proteins, or DNA-protein complexes in gram-negative bacteria, such as Bordetella pertussis, Legionella pneumophila, Rickettsia prowazekii, Ehrlichia chaffeensis, and Anaplasma phagocytophilum (7, 18, 42). While the role of the TFSS in A. marginale has not been defined, the virB operon in Brucella suis was shown to be required for intracellular survival after phagocytosis (22, 41) and may therefore be critical to A. marginale survival in erythrocytes.
The identification of PepA cytosol amino peptidase and EF-Tu was somewhat unexpected because both are believed to be cytoplasmic proteins. PepA cytosol amino peptidase was so designated by genomic annotation (11) but by TMHMM was predicted to be anchored through the outer membrane and exposed to the outside environment. In gram-positive Staphylococcus epidermidis, this protein appears to be cell surface associated (52), but its location in gram-negative bacteria is unclear. EF-Tu was also predicted by TMHMM analysis to be surface associated, and the presence of EF-Tu in various compartments of bacteria has been previously documented. EF-Tu associates with the cell surface of Escherichia coli (32) and that of Lactobacillus johnsonii (28). Furthermore, in L. johnsonii, EF-Tu had properties of an adhesion factor, mediating attachment of bacteria to human intestinal cells (28). In Mycobacterium leprae and Mycobacterium pneumoniae, EF-Tu was identified as a cell wall-associated protein (19, 37). Thus, in some organisms EF-Tu appears to function in the bacterial cytoplasm for protein synthesis, whereas in other organisms EF-Tu appears to localize to the surface, where it can function in bacterial attachment to host cells.
Proteins with unknown functions include OMP4, OMP7, OMP10, OMP14, and OMA87. With the exception of OMA87, these proteins were designated members of the MSP2 superfamily because of sequence identity to a family of surface antigens known as pfam01617 (11). However, their functions in A. marginale are unclear.
The approach of using LC-MS-MS analysis of immunoreactive protein spots excised from 2D gels has several limitations. One unpredicted result from this study was that only 52% of the protein spots submitted for LC-MS-MS were positively identified as A. marginale proteins. Those spots that did not yield a positive identification in the A. marginale genome were also searched against the NCBI database, and spots 1 to 3, 51, 52, 62 to 65, and 67, comprising 13% of the immunoreactive spots, were determined to be of bovine origin (data not shown). It is probable that contamination with bovine protein in these spots quenched any identification of Anaplasma proteins, as it is unlikely that cattle would develop an antibody response to self-antigens following immunization. In fact, sera from calves immunized with outer membranes did not react on immunoblots to uninfected erythrocytes (data not shown). The failure to positively identify the additional protein spots by LC-MS-MS may have resulted from poor solubilization of very hydrophobic outer membrane proteins and therefore incomplete trypsinization.
There were several examples of differences between actual and predicted molecular masses or pIs. These discrepancies could result from posttranslational modification (as is known to occur with MSP1a, which is glycosylated [4, 27]) or partial protein degradation. Furthermore, algorithms to predict pIs are not completely accurate. There were also examples of identifying the same protein in more than one spot. Finding MSP2 and MSP3 in multiple spots was expected, because multiple variants of these proteins, which can differ slightly in molecular masses and pIs, are present during infection (1-3, 13-15, 17, 43, 44, 48). The peptides identified in spots 73 and 74 were conserved peptides derived from the closely related appendage-associated protein family members, so that the analysis was unable to assign an individual member to an individual spot. Additionally, the presence of conjugal transfer protein in spots 7 and 8 and of EF-Tu in spots 43 to 46 and 49 may have resulted from aberrant migration on the gels if these proteins were complexed with their substrates. EF-Tu interacts with tRNA during protein synthesis, and high-affinity binding to tRNA might have impeded the removal of tRNA during outer membrane preparation. Similarly, as a component of the TFSS, conjugal transfer protein may mediate secretion or cell-to-cell transfer of macromolecules and protein-DNA complexes. Thus, residual DNA and/or RNA bound to these two proteins may have affected their mobilities.
The identification of novel antigens in the complex outer membrane immunogen has expanded the number of targets recognized by immune serum IgG and thus provided a number of vaccine candidates for further study. For an effective vaccine, it is important that these proteins can be recognized by serum antibody and CD4+ T lymphocytes from a majority of cattle within a vaccinated population upon infection and that these proteins are conserved among different strains of A. marginale. Animals selected for this study expressed common major histocompatibility complex class II DRB3 alleles that are present in over 50% of Holstein-Friesian cattle (54). Furthermore, the majority of the novel antigenic proteins were recognized by IgG2 from all three animals (Fig. 4), indicating the stimulation of T-helper-lymphocyte responses that promote switching to IgG isotypes. If protective, these vaccine candidates may be useful at the population level.
The availability of the sequences encoding these proteins from other A. marginale strains will enable determination of protein sequence conservation among the different strains. Additionally, identification of these antigenic proteins may offer insight to immunoreactive proteins in related pathogens, such as A. phagocytophilum, E. chaffeensis, E. canis, and E. ruminantium. A. phagocytophilum contains proteins highly similar to A. marginale MSP2 (23, 31, 62), and BLAST searches of the A. marginale proteins identified in this study indicate that several have sequence identity to proteins in A. phagocytophilumand species within the genus Ehrlichia. Thus, the identification of these immunogenic A. marginale proteins may also provide direction towards specific antigens for vaccine development against related animal and human pathogens.
ACKNOWLEDGMENTS
We thank Kim Kegerreis, Shelly Whidbee, Beverly Hunter, Emma Karel, Ralph Horn, Maggie Tam, and Xioting Tang for excellent technical assistance and Don Knowles for providing PDquest 2D imaging software.
This research was supported by an American Society for Microbiology Watkins Minority Fellowship and by NIH NIAID grant R01-AI053692.
REFERENCES
1. Abbott, J. R., G. H. Palmer, K. A. Kegerreis, P. F. Hetrick, C. J. Howard, J. C. Hope, and W. C. Brown. 2005. Rapid and long-term disappearance of CD4+ T lymphocyte responses specific for Anaplasma marginale major surface protein-2 (MSP2) in MSP2 vaccinates following challenge with live A.marginale. J. Immunol. 174:6702-6715.
2. Alleman, A. R., and A. F. Barbet. 1996. Evaluation of Anaplasma marginale major surface protein 3 (MSP3) as a diagnostic test antigen. J. Clin. Microbiol. 34:270-276.
3. Alleman, A. R., G. H. Palmer, T. C. McGuire, T. F. McElwain, L. E. Perryman, and A. F. Barbet. 1997. Anaplasma marginale major surface protein 3 is encoded by a polymorphic, multigene family. Infect. Immun. 65:156-163.
4. Allred, D. R., T. C. McGuire, G. H. Palmer, S. R. Leib, T. M. Harkins, T. F. McElwain, and A. F. Barbet. 1990. Molecular basis for surface antigen size polymorphisms and conservation of a neutralization-sensitive epitope in Anaplasma marginale. Proc. Natl. Acad. Sci. USA 87:3220-3224.
5. Barbet, A. F., A. Lundgren, J. Yi, F. R. Rurangirwa, and G. H. Palmer. 2000. Antigenic variation of Anaplasma marginale by expression of MSP2 mosaics. Infect. Immun. 68:6133-6138.
6. Barigye, R., M. A. Garcia-Ortiz, E. E. Rojas Ramirez, and S. D. Rodriguez. 2004. Identification of IgG2-specific antigens in Mexican Anaplasma marginale strains. Ann. N. Y. Acad. Sci. 1026:84-94.
7. Baron, C., D. O'Callaghan, and E. Lanka. 2002. Bacterial secrets of secretion: EuroConference on the biology of type IV secretion processes. Mol. Microbiol. 43:1359-1365.
8. Beaman, L., and C. L. Wisseman, Jr. 1976. Mechanisms of immunity in typhus infections. VI. Differential opsonizing and neutralizing action of human typhus rickettsia-specific cytophilic antibodies in cultures of human macrophages. Infect. Immun. 14:1071-1076.
9. Bendtsen, J. D., H. Nielsen, G. von Heijne, and S. Brunak. 2004. Improved prediction of signal peptides: SignalP 3.0. J. Mol. Biol. 340:783-795.
10. Bitsaktsis, C., J. Huntington, and G. Winslow. 2004. Production of IFN- by CD4 T cells is essential for resolving ehrlichia infection. J. Immunol. 172:6894-6901.
10. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254.
11. Brayton, K. A., L. S. Kappmeyer, D. R. Herndon, M. J. Dark, D. L. Tibbals, G. H. Palmer, T. C. McGuire, and D. P. Knowles, Jr. 2005. Complete genome sequencing of Anaplasma marginale reveals that the surface is skewed to two superfamilies of outer membrane proteins. Proc. Natl. Acad. Sci. USA 102:844-849.
12. Brayton, K. A., G. H. Palmer, A. Lundgren, J. Yi, and A. F. Barbet. 2002. Antigenic variation of Anaplasma marginale msp2 occurs by combinatorial gene conversion. Mol. Microbiol. 43:1151-1159.
13. Brown, W. C., T. C. McGuire, D. Zhu, H. A. Lewin, J. Sosnow, and G. H. Palmer. 2001. Highly conserved regions of the immunodominant major surface protein 2 of the genogroup II ehrlichial pathogen Anaplasma marginale are rich in naturally derived CD4+ T lymphocyte epitopes that elicit strong recall responses. J. Immunol. 166:1114-1124.
14. Brown, W. C., G. H. Palmer, K. A. Brayton, P. F. Meeus, A. F. Barbet, K. A. Kegerreis, and T. C. McGuire. 2004. CD4+ T lymphocytes from Anaplasma marginale major surface protein 2 (MSP2) vaccinees recognize naturally processed epitopes conserved in MSP3. Infect. Immun. 72:3688-3692.
15. Brown, W. C., G. H. Palmer, H. A. Lewin, and T. C. McGuire. 2001. CD4+ T lymphocytes from calves immunized with Anaplasma marginale major surface protein 1 (MSP1), a heteromeric complex of MSP1a and MSP1b, preferentially recognize the MSP1a carboxyl terminus that is conserved among strains. Infect. Immun. 69:6853-6862.
16. Brown, W. C., V. Shkap, D. Zhu, T. C. McGuire, W. Tuo, T. F. McElwain, and G. H. Palmer. 1998. CD4+ T-lymphocyte and immunoglobulin G2 responses in calves immunized with Anaplasma marginale outer membranes and protected against homologous challenge. Infect. Immun. 66:5406-5413.
17. Brown, W. C., D. Zhu, V. Shkap, T. C. McGuire, E. F. Blouin, K. M. Kocan, and G. H. Palmer. 1998. The repertoire of Anaplasma marginale antigens recognized by CD4+ T-lymphocyte clones from protectively immunized cattle is diverse and includes major surface protein 2 (MSP-2) and MSP-3. Infect. Immun. 66:5414-5422.
18. Christie, P. J., and J. P. Vogel. 2000. Bacterial type IV secretion: conjugation systems adapted to deliver effector molecules to host cells. Trends Microbiol. 8:354-360.
19. Dallo, S. F., T. R. Kannan, M. W. Blaylock, and J. B. Baseman. 2002. Elongation factor Tu and E1 beta subunit of pyruvate dehydrogenase complex act as fibronectin binding proteins in Mycoplasma pneumoniae. Mol. Microbiol. 46:1041-1051.
20. Eriks, I. S., D. Stiller, W. L. Goff, M. Panton, S. M. Parish, T. F. McElwain, and G. H. Palmer. 1994. Molecular and biological characterization of a newly isolated Anaplasma marginale strain. J. Vet. Diagn. Investig. 6:435-441.
21. Estes, D. M., N. M. Closser, and G. K. Allen. 1994. IFN-gamma stimulates IgG2 production from bovine B cells costimulated with anti-mu and mitogen. Cell. Immunol. 154:287-295.
22. Foulongne, V., G. Bourg, C. Cazevieille, S. Michaux-Charachon, and D. O'Callaghan. 2000. Identification of Brucella suis genes affecting intracellular survival in an in vitro human macrophage infection model by signature-tagged transposon mutagenesis. Infect. Immun. 68:1297-1303.
23. French, D. M., W. C. Brown, and G. H. Palmer. 1999. Emergence of Anaplasma marginale antigenic variants during persistent rickettsemia. Infect. Immun. 67:5834-5840.
24. Gambrill, M. R., and C. L. Wisseman, Jr. 1973. Mechanisms of immunity in typhus infections. III. Influence of human immune serum and complement on the fate of Rickettsia mooseri within human macrophages. Infect. Immun. 8:631-640.
25. Ganta, R. R., C. Cheng, M. J. Wilkerson, and S. K. Chapes. 2004. Delayed clearance of Ehrlichia chaffeensis infection in CD4+ T-cell knockout mice. Infect. Immun. 72:159-167.
26. Ganta, R. R., M. J. Wilkerson, C. Cheng, A. M. Rokey, and S. K. Chapes. 2002. Persistent Ehrlichia chaffeensis infection occurs in the absence of functional major histocompatibility complex class II genes. Infect. Immun. 70:380-388.
27. Garcia-Garcia, J. C., J. de la Fuente, G. Bell-Eunice, E. F. Blouin, and K. M. Kocan. 2004. Glycosylation of Anaplasma marginale major surface protein 1a and its putative role in adhesion to tick cells. Infect. Immun. 72:3022-3030.
28. Granato, D., G. E. Bergonzelli, R. D. Pridmore, L. Marvin, M. Rouvet, and I. E. Corthesy-Theulaz. 2004. Cell surface-associated elongation factor Tu mediates the attachment of Lactobacillus johnsonii NCC533 (La1) to human intestinal cells and mucins. Infect. Immun. 72:2160-2169.
29. Hanson, B. A. 1983. Effect of immune serum on infectivity of Rickettsia tsutsugamushi. Infect. Immun. 42:341-349.
30. Hope, M., G. Riding, M. Menzies, and P. Willadsen. 2004. A novel antigen from Anaplasma marginale: characterization, expression and preliminary evaluation of the recombinant protein. Vaccine 22:407-415.
31. Ijdo, J. W., W. Sun, Y. Zhang, L. A. Magnarelli, and E. Fikrig. 1998. Cloning of the gene encoding the 44-kilodalton antigen of the agent of human granulocytic ehrlichiosis and characterization of the humoral response. Infect. Immun. 66:3264-3269.
32. Jacobson, G. R., and J. P. Rosenbusch. 1976. Abundance and membrane association of elongation factor Tu in E. coli. Nature 261:23-26.
33. Kaylor, P. S., T. B. Crawford, T. F. McElwain, and G. H. Palmer. 1991. Passive transfer of antibody to Ehrlichia risticii protects mice from ehrlichiosis. Infect. Immun. 59:2058-2062.
34. Li, J. S., F. Chu, A. Reilly, and G. M. Winslow. 2002. Antibodies highly effective in SCID mice during infection by the intracellular bacterium Ehrlichia chaffeensis are of picomolar affinity and exhibit preferential epitope and isotype utilization. J. Immunol. 169:1419-1425.
35. Li, J. S., E. Yager, M. Reilly, C. Freeman, G. R. Reddy, A. A. Reilly, F. K. Chu, and G. M. Winslow. 2001. Outer membrane protein-specific monoclonal antibodies protect SCID mice from fatal infection by the obligate intracellular bacterial pathogen Ehrlichia chaffeensis. J. Immunol. 166:1855-1862.
36. Lhr, C. V., K. A. Brayton, V. Shkap, T. Molad, A. F. Barbet, W. C. Brown, and G. H. Palmer. 2002. Expression of Anaplasma marginale major surface protein 2 operon-associated proteins during mammalian and arthropod infection. Infect. Immun. 70:6005-6012.
37. Marques, M. A., S. Chitale, P. J. Brennan, and M. C. Pessolani. 1998. Mapping and identification of the major cell wall-associated components of Mycobacterium leprae. Infect. Immun. 66:2625-2631.
38. Martin, M. E., K. Caspersen, and J. S. Dumler. 2001. Immunopathology and ehrlichial propagation are regulated by interferon-gamma and interleukin-10 in a murine model of human granulocytic ehrlichiosis. Am. J. Pathol. 158:1881-1888.
39. McGuire, T. C., W. C. Davis, A. L. Brassfield, T. F. McElwain, and G. H. Palmer. 1991. Identification of Anaplasma marginale long-term carrier cattle by detection of serum antibody to isolated MSP-3. J. Clin. Microbiol. 29:788-793.
40. McGuire, T. C., A. J. Musoke, and T. Kurtti. 1979. Functional properties of bovine IgG1 and IgG2: interaction with complement, macrophages, neutrophils and skin. Immunology 38:249-256.
41. O'Callaghan, D., C. Cazevieille, A. Allardet-Servent, M. L. Boschiroli, G. Bourg, V. Foulongne, P. Frutos, Y. Kulakov, and M. Ramuz. 1999. A homologue of the Agrobacterium tumefaciens VirB and Bordetella pertussis Ptl type IV secretion systems is essential for intracellular survival of Brucella suis. Mol. Microbiol. 33:1210-1220.
42. Ohashi, N., N. Zhi, Q. Lin, and Y. Rikihisa. 2002. Characterization and transcriptional analysis of gene clusters for a type IV secretion machinery in human granulocytic and monocytic ehrlichiosis agents. Infect. Immun. 70:2128-2138.
43. Palmer, G. H., A. F. Barbet, G. H. Cantor, and T. C. McGuire. 1989. Immunization of cattle with the MSP-1 surface protein complex induces protection against a structurally variant Anaplasma marginale isolate. Infect. Immun. 57:3666-3669.
44. Palmer, G. H., A. F. Barbet, W. C. Davis, and T. C. McGuire. 1986. Immunization with an isolate-common surface protein protects cattle against anaplasmosis. Science 231:1299-1302.
45. Palmer, G. H., A. F. Barbet, K. L. Kuttler, and T. C. McGuire. 1986. Detection of an Anaplasma marginale common surface protein present in all stages of infection. J. Clin. Microbiol. 23:1078-1083.
46. Palmer, G. H., G. Eid, A. F. Barbet, T. C. McGuire, and T. F. McElwain. 1994. The immunoprotective Anaplasma marginale major surface protein 2 is encoded by a polymorphic multigene family. Infect. Immun. 62:3808-3816.
47. Palmer, G. H., and T. C. McGuire. 1984. Immune serum against Anaplasma marginale initial bodies neutralizes infectivity for cattle. J. Immunol. 133:1010-1015.
48. Palmer, G. H., S. M. Oberle, A. F. Barbet, W. L. Goff, W. C. Davis, and T. C. McGuire. 1988. Immunization of cattle with a 36-kilodalton surface protein induces protection against homologous and heterologous Anaplasma marginale challenge. Infect. Immun. 56:1526-1531.
49. Palmer, G. H., F. R. Rurangirwa, K. M. Kocan, and W. C. Brown. 1999. Molecular basis for vaccine development against the ehrlichial pathogen Anaplasma marginale. Parasitol. Today 15:281-286.
50. Palmer, G. H., S. D. Waghela, A. F. Barbet, W. C. Davis, and T. C. McGuire. 1987. Characterization of a neutralization-sensitive epitope on the Am 105 surface protein of Anaplasma marginale. Int. J. Parasitol. 17:1279-1285.
51. Park, Y. H., Y. S. Joo, J. Y. Park, J. S. Moon, S. H. Kim, N. H. Kwon, J. S. Ahn, W. C. Davis, and C. J. Davies. 2004. Characterization of lymphocyte subpopulations and major histocompatibility complex haplotypes of mastitis-resistant and susceptible cows. J. Vet. Sci. 5:29-39.
52. Reis, M., M. Eschbach-Bludau, M. I. Iglesias-Wind, T. Kupke, and H.-G. Sahl. 1994. Producer immunity towards the lantibiotic Pep5: identification of the immunity gene pepI and localization and functional analysis of its gene product. Appl. Environ. Microbiol. 60:2876-2883.
53. Riding, G., M. Hope, D. Waltisbuhl, and P. Willadsen. 2003. Identification of novel protective antigens from Anaplasma marginale. Vaccine 21:1874-1883.
54. Sharif, S., B. A. Mallard, B. N. Wilkie, J. M. Sargeant, H. M. Scott, J. C. Dekkers, and K. E. Leslie. 1999. Associations of the bovine major histocompatibility complex DRB3 (BoLA-DRB3) with production traits in Canadian dairy cattle. Anim. Genet. 30:157-160.
55. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68:850-858.
56. Sonnhammer, E. L., G. von Heijne, and A. Krogh. 1998. A hidden Markov model for predicting transmembrane helices in protein sequences. Proc. Int. Conf. Intell. Syst. Mol. Biol. 6:175-182.
57. Stich, R. W., G. A. Olah, K. A. Brayton, W. C. Brown, M. Fechheimer, K. Green-Church, S. Jittapalapong, K. M. Kocan, T. C. McGuire, F. R. Rurangirwa, and G. H. Palmer. 2004. Identification of a novel Anaplasma marginale appendage-associated protein that localizes with actin filaments during intraerythrocytic infection. Infect. Immun. 72:7257-7264.
58. Tebele, N., T. C. McGuire, and G. H. Palmer. 1991. Induction of protective immunity by using Anaplasma marginale initial body membranes. Infect. Immun. 59:3199-3204.
59. van Eijk, M. J., J. A. Stewart-Haynes, and H. A. Lewin. 1992. Extensive polymorphism of the BoLA-DRB3 gene distinguished by PCR-RFLP. Anim. Genet. 23:483-496.
60. Visser, E. S., T. C. McGuire, G. H. Palmer, W. C. Davis, V. Shkap, E. Pipano, and D. P. Knowles, Jr. 1992. The Anaplasma marginale msp5 gene encodes a 19-kilodalton protein conserved in all recognized Anaplasma species. Infect. Immun. 60:5139-5144.
61. Winslow, G. M., E. Yager, K. Shilo, E. Volk, A. Reilly, and F. K. Chu. 2000. Antibody-mediated elimination of the obligate intracellular bacterial pathogen Ehrlichia chaffeensis during active infection. Infect. Immun. 68:2187-2195.
62. Zhi, N., N. Ohashi, and Y. Rikihisa. 1999. Multiple p44 genes encoding major outer membrane proteins are expressed in the human granulocytic ehrlichiosis agent. J. Biol. Chem. 274:17828-17836.(Job E. Lopez, William F. )