Sequence and immunologic conservation of Anaplasma marginale OmpA within strains from Ghana as compared to the predominant OmpA variant

A primary challenge in developing effective vaccines against obligate, intracellular, bacterial tick-borne pathogens that establish persistent infection is the identification of antigens that cross protect against multiple strains. In the case of Anaplasma marginale, the most prevalent tick-borne pathogen of cattle found worldwide, OmpA is an adhesin and thus a promising vaccine candidate. We sequenced ompA from cattle throughout Ghana naturally infected with A. marginale in order to determine the degree of variation in this gene in an area of suspected high genetic diversity. We compared the Ghanaian sequences with those available from N. America, Mexico, Australia and Puerto Rico. When considering only amino acid changes, three unique Ghanaian OmpA variants were identified. In comparison, strains from all other geographic regions, except one, shared a single OmpA variant, Variant 1, which differed from the Ghanaian variants. Next, using recombinant OmpA based on Variant 1, we determined that amino acid differences in OmpA in Ghanaian cattle as compared to OmpA Variant 1 did not alter the binding capacity of antibody directed against OmpA Variant 1, supporting the value of OmpA as a highly conserved vaccine candidate.


Introduction
The identification of protective antigens is a primary limitation of vaccine development against obligate, intracellular pathogens that use antigenic variation to establish persist infection within the mammalian host. These pathogens include, though are not limited to bacteria in the family Anaplasmataceae, such as Anaplasma marginale, A. phagocytophilum and Ehrlichia ruminantium. A. marginale causes bovine anaplasmosis which is a production limiting disease PLOS ONE | https://doi.org/10.1371/journal.pone.0217661 July 10, 2019 1 / 10 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 of cattle that occurs worldwide [1]. A single vaccine that protects against antigenically distinct variants is a high priority. This is particularly true for farmers in tropical regions of the world in which high infection prevalence and transmission pressure leads to pathogen diversification [2]. Whole outer membrane preparations induce protection against bovine anaplasmosis and in a proportion of vaccinates, protection against infection [3][4][5]. The immunodominant components of the outer membrane preparation include the major surface proteins Msp1, Msp2, Msp3, Msp4 and Msp5. Msp2 and Msp3 are best studied and are critical for immune escape and establishment and maintenance of persistent infection [6][7][8]. However, none of these major surface proteins have consistently produced protective immunity, though many trials have been conducted using both native and recombinant proteins [9][10][11].
Complete genome sequencing, proteomics and reverse vaccinology have identified a number of additional outer membrane proteins that could serve as vaccine candidates [12,13]. However, narrowing the possible candidates is challenging in the absence of knowledge about proteins that mediate essential functions. Adhesins and invasins are particularly relevant vaccine candidates because they are functionally essential and tend to have highly conserved regions. Additionally, blocking adhesion and invasion to host cells is lethal for obligate intracellular pathogens.
Recently, ompA, annotated as Am854 in the A. marginale genome, was identified as an adhesin playing an essential role in A. marginale entry into mammalian and tick cells [14]. Consequently, this protein serves as a highly relevant vaccine candidate. In West Africa, where prevalence of A. marginale is up to 75%, there is a near complete absence of information concerning the extent of genetic variation of A. marginale in general and ompA, specifically [15]. Additionally, genetic variation is generally higher in regions where pathogen prevalence and transmission pressure are high [2,16]. Thus, the genetic variation in ompA in such regions is a relevant question for vaccine development. In this paper we determine the variation in OmpA in A. marginale strains infecting cattle in Ghana and test whether this variation impacts antibody binding.

Animals and sampling
The study was carried out on properties owned by the University of Ghana, the Ghanaian Ministry of Food and Agriculture, and on private farms. Permission was received to handle and bleed animals from cattle owners and managers of private farms and research institutions. No endangered species were involved in this study. The cattle used in this study were treated with strict accordance of guidelines set by University of Ghana Institutional Animal Care and Use Committee and the Washington State University Animal Care and Use Committee. Endowment fund (GHP). The funding agencies did not play any role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

DNA extraction and amplification of ompA from whole blood
Genomic DNA was extracted from the peripheral blood samples collected in EDTA using a Gentra Puregene Blood kit (Qiagen, Germantown, MD) as described by the manufacturer's protocol. PCR to amplify full-length ompA was done using the following oligonucleotides: 5'-ATG CTG CAT CGT TGG TTA GC-3' and 5'-TTC AGG CGC GAC CAC TCC TG -3' , and PCR master mix (Roche, Basel, Switzerland). These oligonucleotides bind to the complementary bases of the start codon (TAC) and the last codon (GAA). Cycling conditions were as follows: initial denaturation at 94˚C for 5 min, annealing at 60˚C for 1 min, extension at 72˚C for 1 min and final extension at 72˚C for 7 min, for 35 cycles. The resulting PCR fragments were visualized by electrophoresis in 1% agarose gel stained with ethidium bromide. All sequences have been deposited in GenBank and have accession numbers MK882853-MK882884.

Cloning and sequencing of the recombinant OmpA
The amplicons were cloned into PCR 4 TOPO Vector plasmids and TOP 10 Esherichia coli cells were transformed using the manufacturer's protocol (Invitrogen, Carlsbad, CA). The resulting individual colonies were selected and grown overnight at 37˚C using standard procedures. Plasmid DNA was extracted from the cells using the Q1Aprep 1 Spin Miniprep kit (Qiagen, Germantown, MD). To confirm the presence of the desired inserts, plasmid DNA samples were digested with EcoRI and the resulting fragments were visualized on a 1% agarose gel stained with ethidium bromide.

OmpA expression and purification
Genomic DNA derived from the St. Maries strain of A. marginale was used as a template to amplify ompA as previously described [11]. The PCR product was cloned into pBAD202 Directional TOPO Expression vector (ThermoFisher Scientific, Waltham, MA) and inserted into TOP10 One Shot chemically competent cells. Positive transformation was confirmed by PCR and Sanger sequencing (Eurofins, Louisville, KY).
Protein expression, from a single randomly selected clone, was induced with 20% arabinose to a final concentration of 0.1% (v/v) once the OD600 reached 0.6 nm. The culture was grown for 4 hours, pelleted by centrifugation for 10 min at 10 000g and 4˚C. The pellets were flash frozen with liquid nitrogen and stored overnight at -80˚C.
Pellets were resuspended in PBS containing cOmplete ULTRA Tablets, Mini, EDTA-free, Protease Inhibitor Cocktail (Roche, Basel, Switzerland) and sonicated for 4 min using a high intensity cup at 90% intensity (Fisherbrand Model 705 Sonic Dismembrator) in intervals of 20 seconds ON/OFF. The lysate was then centrifuged at 30,000 x g for 30 min and 4˚C.
The first step of purification was performed using 5mL of HisPur Nickel NTA resin (Ther-moFisher Scientific, Waltham, MA) packed into a 10/10 GL Tricorn column (GE Healthcare, Chicago, IL) attached to an AKTA Avant HPLC apparatus (GE Healthcare, Chicago, IL). Sample loading was performed at 1mL/min and the remaining steps were performed at 2 mL/min. Chromatographic conditions were as follows: resin equilibration with 90% buffer A1 (50 mM Sodium phosphate 500 mM NaCl), 10% buffer B1 (50 mM Sodium phosphate 500 mM NaCl 150mM Imidazole pH 7.6) for 5 column volumes (CV); sample loading was performed using the sample pump; washout unbound for 5 CV in 10% Buffer B1, sample elution in 5 CV at 100% B1. Protein elution was detected measuring absorbance at 280nm and fractions were collected in 1ml aliquots at 12˚C. Pooled fractions correspondent to the entire peak that eluted at 100% B1 were buffer exchanged into Buffer A2 (20mM BisTris pH 6.5) using a Hi-trap G25 column attached to the AKTA and then chromatographed on a HPLC AKTA apparatus using a MonoQ GL 5/10 column (GE Healthcare, Chicago, IL). The anion exchange separation was performed at 1 mL/min using Buffers A2 and B2 (20mM BisTris 250mM NaCl pH 6.5) and protein elution was detected by measuring absorbance at 280 nm. Chromatographic conditions were: Isocratic (100% A2) for 5 CV, followed by sample loading using the sample pump, isocratic (100% A2) for 5 CV, 0-100% B2 for 25 CV, isocratic (100% B2) for 5 CV and isocratic (100% A2) for 10 CV. Protein peaks were collected using peak elution and protein purity and identity confirmed by SDS-PAGE and mass spectrometry (WSU central core facilities). Fractions containing pure OmpA were pooled and buffer exchanged by gel filtration using a G25 prepacked column attached to the AKTA into PBS pH 7.4. The data were collected, and the sample chromatograms were acquired using a Unicorn Chromatography Management System Version 7.0 (GE Healthcare, Chicago, IL). Sample concentration was determined by BCA (ThermoFisher Scientific, Waltham, MA). Successful protein purification was verified using standard SDS-PAGE (Fig 1).

Enzyme-Linked Immunosorbent Assay (ELISA)
ELISA's to detect anti-OmpA antibody were done as previously described [11]. Briefly, serial two-fold dilutions starting at 1:10 were used in triplicate to determine end-point titers. A positive titer was calculated as three standard deviations above the mean of the optical density of the pre-immune sera at the same dilution as the test sera [20]. The commercially available cELISA (VMRD, Pullman, WA) was used to measure anti-Msp5 antibody response using the manufacturer's instructions.

Statistical analyses
All data analyses were conducted using GraphPad Prism (version 8.1.0 for Windows, Graph-Pad Software, La Jolla CA). End point titers for OmpA were log 10 transformed. The Mann-Whitney two-tailed test was used to determine P values when comparing anti-OmpA endpoint titers and percent inhibition of anti-Msp5 antibodies in the St. Maries and Ghanaian strain-infected animals. A Mann-Whitney test was used to test null hypothesis that there is no greater diversity among Ghana strains than among the North America strains of ompA.

Conservation of ompA
Three new ompA variants were identified in A. marginale strains from cattle throughout Ghana (Table 1). The overall conservation among the three variants was high with >99% nucleotide and >98% amino acid identity. None of the three variants were identical to the There are three variants among all other previously identified ompA sequences. Variant 1 (V1) is the most common and is typified by the St. Maries reference strain. Variant 2 (V2) has one synonymous nucleotide substitution of a guanine to adenine at base pair 531, thus making the amino acid sequence identical to that of V1. V3, which represents only EMF, shares a single synonymous and a single non-synonymous SNP with all Ghanaian variants, as described above.

Antibody responses
In order to determine if these sequences differences affected antibody binding, OmpA V1, the most common variant, from the St. Maries strain of A. marginale was expressed as recombinant protein. The endpoint titers of the specific anti-OmpA response in animal infected with the St. Maries strain and the Ghanaian strains of A. marginale were compared. Sequence differences between OmpA V1 and the Ghanaian strains did not alter antibody binding (Fig 2A). In animals from Ghana, titers to OmpA V1 tended to be higher than in animals infected with the homologous St. Maries strain, though there was marked variation between animals in both groups. All animals in Ghana had titers to OmpA. The titers varied from 10 to 12,800 with a mean titer (+/-CI) of 2459 +/-3437. In the St. Maries-infected animals, three animals lacked titers to OmpA. The maximum titer was 640. The mean titer (+/-CI), including the negative animals was 166 +/-149.30. In order to verify that the higher titers in the Ghanaian animals did not reflect elevated anti-A. marginale responses in general, we measured the anti-Msp5 antibody response using a commercially available cELISA (Fig 2B). This cELISA detects antibody against the highly conserved Major Surface Protein 5. The percent inhibition in animals from Ghana was between 21-81% (mean +/-CI = 57% +/-10.96). The percent inhibition in animals infected with the St. Maries strain was between 34 and 73% (mean +/-CI = 61% +/-8.64).

Discussion
One of the major limitations of developing a vaccine against A. marginale is strain diversity and the inability to produce cross-protection against a variety of field strains with a single vaccine. Consequently, vaccine development and antigen selection have focused on the identification of outer membrane proteins that are broadly conserved across distant geographic areas. For example, Omp 5,9,12,13 and 14 are invariant during acute and persistent infection as well as tick transmission [21]. However, all of these Omps have variation between strains even when comparing strains from N. America including St. Maries, Florida, S. Idaho, Washington-Okanogan, and Oklahoma [22]. Interestingly, this study demonstrates that OmpA has less variation than these other outer membrane proteins, with a maximum of three amino acid differences between the most common OmpA variant, V1, and the Ghanaian variants. Additionally, these amino acid differences did not alter the capacity of antibody raised against the Ghanaian strains to bind the heterologous OmpA V1 supporting the value of OmpA as a highly conserved vaccine candidate. Importantly, the endpoint titers are widely variable between animals. For example, in animals Sequence and immunology conservation of A. marginale OmpA infected with Ghanaian strains, titers vary from 10 to 12,800, and in St. Maries infected animals titers vary from 0 to 640. This difference reflects the expected variability in the immune response in outbred animals, which must be addressed in vaccine development.
The reason for the overall higher titers directed against recombinant OmpA in the Ghanaian animals is unknown. However, to ensure that this difference does not reflect anti-A. marginale antibodies levels in general, the commercially available cELISA was used to evaluate the anti-msp5 antibody response. Msp5 is an immunodominant, highly conserved outer membrane protein, commonly used for diagnostic purposes [23]. The % inhibition between groups of animals was similar with mean of 61% and 57% for the St. Maries-strain infected and Ghanaian strain-infected animals, respectively.
OmpA serves as an adhesion and invasion for A. marginale and as such is an important vaccine candidate [14]. Here we demonstrate that OmpA is remarkably conserved in A. marginale from geographically diverse regions and the minimal difference between OmpA sequences does not affect antibody binding. Additionally, no amino acid diversity is present in the experimentally determined binding domain [14]. In one study, recombinant OmpA failed to induce protective immunity against tick challenge in cattle [11]. However, the binding domain is not a predicted B cell epitope and, in that study, the overall antibody response to a peptide containing the binding domain was low. Thus, appropriate adjuvants and delivery strategies need to be developed to overcome the marked variation in the immune response between animals and to maximize the potential protective capacity of the anti-OmpA immune response.