The study was ethically approved by the local Animal Health and Welfare Authority (Diputación Foral de Alava) with reference No. 1820, 12th May 2015, following Spanish ethical guidelines and animal welfare regulations (Real Decreto 53/2013).

The His-tag recombinant A. phagocytophilum human NY18 isolate (Asanovich et al., 1997) proteins MSP4 (AFD54597) and HSP70 (KX891324) were produced in Escherichia coli BL21 cells (Champion pET101 Directional TOPO Expression kit, Carlsbad, CA, USA), after induction with IPTG and purified using the Ni-NTA affinity column chromatography system (Qiagen Inc., Valencia, CA, USA) as previously described (Villar et al., 2015b). Recombinant purified proteins showed purity higher than 85% of total proteins and were used to immunize rabbits to purify IgGs from preimmune and immunized animals (Montage Antibody Purification Kit and Spin Columns with PROSEP-A Media, Millipore, Billerica, MA, USA) for Western blot and antibody inhibition analyses as previously described (Villar et al., 2015b).

The A. phagocytophilum human NY18 isolate was propagated in cultured HL60 human promyelocytic cells as previously described (de la Fuente et al., 2005). The A phagocytophilum-infected cells (~1 × 10⁷ cells) were collected when 70–80% of the cells were infected as determined by detection of intracellular morulae in stained cytospin cell smears. Host cell-free bacteria were isolated from cell lysates after five passages through a 27-gauge syringe, followed by differential centrifugation in Percoll gradients as previously described for A. marginale to separate bacteria from host cell debris (Lis et al., 2014). The pellet of purified A. phagocytophilum was resuspended in 200 μl of SPG buffer (0.25 mM sucrose, 10 mM sodium phosphate, 5 mM L-glutamic acid, pH 7.2), and 5 μl of sequencing-grade trypsin (Promega, Madison, WI, USA) was added to half of the cell reaction mixture. Bacteria were incubated at 37°C for 30 min and then centrifuged at 10,000 × g for 15 min and resuspended in Laemmli protein loading buffer, boiled for 5 min and loaded onto a 12% SDS-PAGE and analyzed by Western blot using rabbit antibodies against recombinant MSP4 and HSP70 proteins as previously described (Villar et al., 2015b).

The active A. phagocytophilum HSP70 and MSP4 proteins were modeled using I-TASSER (Zhang, 2008), and the unbound (apo)-HSP70 protein was modeled using Robetta (Kim et al., 2004). All tertiary models were optimized with the Schrödinger's Maestro Protein Preparation Wizard (Li et al., 2007). All steric clashes were resolved via minimization with the default settings in the Schrodinger's Maestro package. For the tertiary models, the Protein Preparation Wizard clusters at the highest degree of hydrogen bonding in equilibrium were used. Monte Carlo orientations were performed (100,000) for each cluster. The optimized structure is based on electrostatic and geometric scoring functions. The membrane positioning for MSP4 was calculated by the OPM database (Lomize et al., 2006) and generated using the Desmond systems builder (Bowers et al., 2006) as part of the Schrodinger's Maestro package. The protein-protein docking was assessed using the SwarmDock server (Torchala et al., 2013) that incorporates flexible docking by exploring in proximity to the Cartesian center of mass of the target protein. Minimization steps are included for the whole system. The poses are calculated based on the most energy favorable poses, minimized once again and sent to the user. We chose to analyze the top 10 poses since these were highly energy favorable (−41 to −54 kcal/mol). The server also produces the residue contacts made between both proteins. All structures were visualized and analyzed using the Visual Molecular Dynamics (Humphrey et al., 1996).

The adhesion of recombinant E. coli strains to HL60 human cells was characterized as previously reported (de la Fuente et al., 2001; Villar et al., 2015b). Briefly, E. coli strains producing A. phagocytophilum MSP4, HSP70, and mutant HSP70 with truncated peptide-binding domains that are involved in protein-protein interactions (Villar et al., 2015b) recombinant proteins were grown and induced as described before. The E. coli cells transformed with expression vector alone were used as negative control. Cell densities were determined and adjusted to 10⁸ cells per ml in Luria Broth (LB). One hundred microliters (10⁷ bacteria) culture were added to 900 μl of 10⁶ cells per ml suspensions of HL60 human cells in LB. Human cells and bacteria were incubated for 30 min at 37°C with occasional agitation. Cells were then collected by centrifugation, washed two times in PBS and resuspended in 100 ml of PBS. Elimination of unbound bacteria from human cells with bound bacteria was performed by Percoll (Sigma, St. Louis, MI, USA) gradient separation (de la Fuente et al., 2001). The band containing human cells was removed with a pipette and washed in PBS. The final cell pellet was lysed in 1 ml of sterile water and 5 μl plated onto LB agar plates containing 100 μg of ampicillin per ml. Two replicates were done for each experiment. Adhesive bacteria were quantitated as the number of colony forming units (CFU) recovered from each test and compared to the control values by Chi2 test (P = 0.001; N = 2).

The HL60 human cells incubated as described above with E. coli strains producing A. phagocytophilum MSP4 and HSP70 recombinant proteins or transformed with expression vector alone as controls were pelleted and fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer for 1 h at room temperature. Fixed cells were washed three times in 0.1 M phosphate buffer with 4% glucose and embedded in 2% agar at 60°C. Samples were post-fixed using 2% OsO₄ for 2 h at room temperature, three times washed and dehydrated in a graded series of acetone (30–100%) solutions for 15 min at each step. Samples were infiltrated with 25, 50, and 75% solutions of Spi Pon Epoxy resin (Structure Probe, Inc. Supplies, West Chester, PA, USA) diluted in anhydrous acetone for 1 h at each step. Samples were left in 100% resin overnight, transferred to embedding molds and polymerized at 60°C for 48 h. Ultrathin sections were contrasted in ethanolic uranyl acetate and lead citrate, carbon coated and observed in a JEOL 1010 TEM (JEOL Ltd., Akishima, Tokio, Japan) at an accelerating voltage of 80 kV. Images were captured using a Mega View III camera (Olympus Soft Imaging Solutions GmbH, Münster, Germany).

Anaplasma phagocytophilum MSP4 (AFD54597) and HSP70 (AAC31306) amino acid sequences were aligned using MAFFT version 7, applying a gap opening penalty of 3 (range 1–3, default 1.53; Katoh and Standley, 2013). Sequence homology was calculated using Clustal Omega (Sievers et al., 2011). The linear B-cell epitopes on A. phagocytophilum MSP4 and HSP70 proteins were predicted using the Bepipred Linear Epitope Prediction tool (http://tools.immuneepitope.org/bcell/; Haste Andersen et al., 2006; Larsen et al., 2006; Ponomarenko and Bourne, 2007). Subsequently, to search for epitope homology, each predicted epitope within each protein was aligned with the full sequence of the other protein.

The inhibitory effect of rabbit IgG antibodies against MSP4 and HSP70 recombinant proteins on A. phagocytophilum human NY18 (Asanovich et al., 1997) and sheep (Alberdi et al., 2015) isolates infection of HL60 human cells was conducted as described previously for tick cells (Villar et al., 2015b). The inhibitory effect of IgG antibodies purified from MSP4 and MSP4-HSP70 immunized and control sheep at 0 and 94 days post-infection (dpi) on A. phagocytophilum human NY18 isolate infection of HL60 human cells was conducted using the same experimental approach as for rabbit IgG. The IgGs were purified from sheep sera using the NAb Protein G spin kit (Thermo Fisher Scientific, Waltham, MA, USA) following manufacturer's recommendations. HL60 cells were pooled and used to seed 24-well plates for each assay. Each well received 1 × 10⁶ cells in RPMI 1,640 medium (Gibco, Thermo Fisher, Madrid, Spain) 48 h prior to inoculation with A. phagocytophilum. Infected cultures for inoculum were harvested when infection reached 80% and host cells were mechanically disrupted with a syringe and 26-gauge needle. Rabbit or sheep purified IgGs (2.2–2.4 mg/ml) were mixed with inoculum (1:1) for 60 min before being placed on the cell monolayers. Each monolayer then received 100 μl of the inoculum plus IgG mix and plates were incubated at 37°C for 30 min. The inoculum was removed from the wells and cell monolayers washed three times with PBS. Complete medium (1 ml) was added to each well and the plates were incubated at 37°C. The control for each trial included inoculum incubated with rabbit pre-immune IgG or control sheep IgG. Four replicates were done for each treatment. After 7 days, cells from all wells were harvested, resuspended in 1 ml PBS and frozen at −70°C. Samples were thawed and solubilized with 1% Triton-X100 and processed for A. phagocytophilum detection by PCR after DNA extraction using TriReagent (Sigma) according to the manufacturer's recommendations. Anaplasma phagocytophilum infection levels were determined by msp4 real-time PCR normalizing against human β-actin as described previously (de la Fuente et al., 2005) but using oligonucleotide primers MSP4-L (5′-CCTTGGCTGCAGCACCACCTG-3′) and MSP4-R (5′-TGCTGTGGGTCGTGACGCG3′), with PCR conditions of 5 min at 95°C and 35 cycles of 10 s at 95°C, 30 s at 55°C and 30 s at 60°C. Results were compared between treatments by the Student's t-test with unequal variance (P = 0.05; N = 4).

HL60 cells were incubated with 4 μM MSP4 and HSP70 recombinant proteins or their combination in culture media for 1 h at 37°C and 5% CO2 in a humidified atmosphere. For antigen combination, equal molar ratios of each protein, equivalent to one part of HSP70 and two parts of MSP4, were incubated at 4°C on a rotator overnight. HL-60 cells (4 × 10⁵ cells/well) were fixed in 4% paraphormaldehyde (PFA) in PBS for 1 h at room temperature (RT), and then incubated with a 6x-His epitope tag monoclonal antibody (3 μg/ml mouse IgG1, Thermo Fisher 4A12E4) for 1 h at RT. After washing, cells were incubated with FITC-conjugated goat anti-mouse IgG (1/100, Sigma F2012) for 1 h at RT. Protein binding was assessed by flow cytometry using a FACScalibur® Flow Cytometer, equipped with the CellQuest Pro® software (BD-Biosciences, San Jose, CA, USA) as previously described (Seidman et al., 2015; Hebert et al., 2017). Incubation with PBS was used as negative control. The viable cell population was gated according to forward scatter and side scatter parameters. To determine the effect of MSP4 and HSP70 recombinant proteins on A. phagocytophilum infection, HL60 cells were incubated with 4 μM MSP4 and HSP70 recombinant proteins or their combination for 1 h, after which A. phagocytophilum human NY18 isolate bacteria purified as described above were added and incubated with the host cells in the continued presence of recombinant protein for 2 h. Unbound bacteria and proteins were removed and the infection was allowed to proceed for 48 h. Then, cells were harvested and infection levels determined by PCR and statistically analyzed as described above. The Rhipicephalus microplus Subolesin recombinant protein (SUB; Merino et al., 2011) and PBS were included as controls. Four replicates were done for each treatment.

The recombinant MSP4 was formulated alone or combined with HSP70. Antigen combination was done as described above in Section Protein Inhibition Assay. Recombinant antigens or saline control were adjuvated in Montanide ISA 50 V2 (Seppic, Paris, France; Merino et al., 2013). Nine 3-month old lambs of the Latxa breed (Basque Country, Spain) were selected from the experimental sheep flock maintained at NEIKER and were kept indoor during the experiment. This flock has no known history of ticks or tick-borne diseases. However, blood of lambs and their dams were analyzed prior to the start of the study to check their status for hemoparasites Theileria, Babesia, and Anaplasma spp. as previously described (Hurtado et al., 2015). All animals were negative for these hemoparasites. Three groups of 3 lambs each with similar live weight were formed. Lambs from each group were injected subcutaneously in the loose skin of the axilla (armpit) using a sterile syringe with removable needle 20 G × 1″ (9.0 × 25 mm), and taking aseptic precautions. Lambs were immunized three times on days 55, 30, and 10 before experimental infection with 1 ml doses of MSP4 (100 μg/dose), MSP4-HSP70 (100 μg of equal molar ratios of each protein/dose) or adjuvant/saline as control. The strain of A. phagocytophilum used for experimental infection originated from an infected lamb in Norway, which suffered TBF but was negative to other tick-borne pathogens (Alberdi et al., 2015; Stuen et al., 2015). The inoculum consisted of A. phagocytophilum infected heparinised blood that had been stored at −70°C with 10% dimethyl sulfoxide (DMSO). Once unfrozen, 1 ml of infected blood containing 1.8 × 10⁶ infected neutrophils per ml was intravenously inoculated into each experimental lamb through the jugular vein using sterile winged infusion sets with needle 21 G × 3/4″ (0.8 × 19 mm).

Whole blood and serum samples were collected from the jugular vein of each lamb previous to each immunization, daily starting on infection day during 10 days, and at weekly intervals until the end of the experiment at 94 dpi (Supplementary Table 1). Rectal temperatures were taken daily until 10 dpi and then weekly until 94 dpi (Supplementary Table 1). Lambs were also weighed periodically (Supplementary Table 1). Hematological analyses including leukocyte and erythrocyte cell counts, leukocyte cell differentiation (percent neutrophils, lymphocytes, monocytes and eosinophils), hemoglobin levels, hematocrit, mean cell volume (MCV), and mean corpuscular hemoglobin (MCH) were performed with an electronic counter (Hemavet 950, Drew, USA) in blood samples collected in EDTA-containing tubes (Supplementary Table 1). Blood smears stained with Giemsa stain were examined to investigate the presence of A. phagocytophilum in blood cells (Supplementary Table 1). At least 100 neutrophils were counted and examined to calculate the number of infected neutrophils per milliliter blood of each lamb throughout the experiment. The differential percent of A. phagocytophilum-infected neutrophils was calculated as the difference between values at different dpi and values at 3 dpi when infected neutrophils were first detected (Supplementary Table 1).

An indirect ELISA test was performed to detect IgG antibodies against MSP4 and HSP70 proteins in immunized and control lambs using serum samples collected before each immunization and at 0, 7, 10, and 94 dpi. High absorption capacity polystyrene microtiter plates were coated with 100 μl (0.01 μg/μl solution of purified recombinant MSP4 or HSP70 protein) per well in carbonate-bicarbonate buffer (Sigma). After an overnight incubation at 4°C, coated plates were blocked with 100 μl/well of blocking solution (5% skim milk in PBS). Serum samples or PBS as negative control were diluted (1:100, v/v) in blocking solution and 100 μl/well were added into duplicate wells of the antigen-coated plates. After an overnight incubation at 4°C, the plates were washed three times with a washing solution (PBS containing 0.05% Tween 20). A donkey anti-sheep IgG-peroxidase conjugate (Sigma) was added (diluted 1:1000 in blocking solution) and incubated at room temperature for 1 h. After three washes with washing solution, 100 μl/well of substrate solution (Fast OPD; Sigma) was added. Finally, the reaction was stopped with 50 μl/well of 3N H₂SO₄ and the optical density (OD) was measured in a spectrophotometer at 450 nm. Antibody titers were expressed as OD_450nm (OD_lambsera OD_PBScontrol). The antigen-specific antibody response in immunized lambs was corroborated by ELISA using pooled sera collected at 0 dpi, but incubating sera with A. phagocytophilum purified from infected HL60 human cells as described above in Section Antibody Inhibition Assay. Results from rectal temperature and hematological analyses, differential percent of A. phagocytophilum-infected neutrophils and antibody titers were compared between immunized and control groups by two-way ANOVA test (P = 0.05; N = 3).

For the analysis of A. phagocytophilum infection in lambs during the trial, DNA was extracted from 200 μl of blood using the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany), including negative extraction controls every 9 samples. DNA was stored at −20°C until subsequent analysis. The presence of Anaplasma spp. was firstly determined using a real-time PCR assay for the screening of piroplasms and Anaplasma spp. (RTi-PCR1) that targets the 16S rRNA gene of Anaplasma spp. and the 18S rRNA gene of piroplasms of the genera Theileria and Babesia (Hurtado et al., 2015). The assay also includes an internal amplification control (IAC) to monitor for possible inhibition of the PCR reaction. All samples positive to Anaplasma spp. in the RTi-PCR1 were analyzed with a multiplex PCR assay that specifically amplifies the major surface protein 2 (msp2) gene of A. phagocytophilum, and the msp4 gene of Anaplasma ovis (RTi-PCR2). Sequences of primers and probes, as well as details on cycling conditions were as reported previously (Hurtado et al., 2015). Analyses were performed in 20 μl volume reactions using an ABI PRISM 7500 Fast Sequence Detection System (Applied Biosystems, Foster City, CA, USA). For the quantitative analysis of A. phagocytophilum infection levels, DNA was extracted from 200 μl blood samples using Nucleospin 96 Blood (Machery-Nagel, Düren Germany). A quatitative real-time PCR was conducted on DNA samples using a Quantitect SYBR Green RT-PCR Kit and a Rotor Gene Q thermocycler (Qiagen, Inc. Valencia, CA, USA) following manufacturer's recommendations. A dissociation curve was run at the end of the reaction to ensure that only one amplicon was formed and that the amplicon denatured consistently in the same temperature range for every sample (Ririe et al., 1997). The DNA levels were normalized against sheep aldolase B using primers Ovi-ALDOB-F: CCCATCTTGCTATCCAGGAA and Ovi-ALDOB-R: TACAGCAGCCAGGACCTTCT, and the genNorm method (ddCT method as implemented by Bio-Rad iQ5 Standard Edition, Version 2.0; Livak and Schmittgen, 2001). Normalized Ct values were compared between immunized and control groups by Student's t-test with unequal variance (P = 0.05; N = 3).

The subcellular localization of MSP4 and HSP70 proteins was characterized in A. phagocytophilum purified from infected HL60 human promyelocytic leukemia cells, mock treated or surface digested with trypsin and loaded onto polyacrylamide gels for Western blot analysis using rabbit antibodies specific against recombinant proteins. The results showed that as a transmembrane protein, MSP4 was partially resistant to trypsin digestion in A. phagocytophilum from HL60 cells, while HSP70 was extracellular and exposed to protease digestion (Figure 1A).

The apo and bound A. phagocytophilum HSP70 models showed that the major structural difference between the two HSP70 tertiary structures is at the C-terminus (residues 400–533), with a 40–90 Å α-carbon backbone root mean square deviation (RMSD; Supplementary Figure 1A). The A. phagocytophilum MSP4 showed a 25° ± 1° tilt from the membrane with its N/C-terminus oriented toward the cytosol, and the β-sheets buried within the membrane with remaining β-hairpin loops exposed extracellularly (Supplementary Figure 1B). The models suggested a limited number of possible MSP4-HSP70 binding positions due to the membrane orientation of MSP4 (Supplementary Figure 1C). The energy score of the optimum MSP4-HSP70 bound state (Supplementary Figure 1C) was calculated at −46 kcal/mol. Furthermore, although the residue map showed that the majority of protein-protein contacts are formed between the β-sheets of MSP4 buried within the membrane and the C-terminus of HSP70, several contacts between the β-hairpin loops of MSP4 and the N-terminus of HSP70 are exposed extracellularly, and therefore these residues may act as markers for mutational studies and antibody targeting (Supplementary Figure 1C). These models supported the interaction between A. phagocytophilum MSP4 and HSP70 proteins when localized on the bacterial membrane.

The binding of HSP70 and MSP4 to HL60 human cells was characterized using recombinant proteins and E. coli producing surface-exposed A. phagocytophilum proteins. The results demonstrated that MSP4 and HSP70 are involved in binding to human promyelocytic leukemia cells (Figures 1B,C). Furthermore, E. coli producing the mutant HSP70 with truncated peptide-binding domains that are involved in protein-protein interactions did not bind to human HL60 cells, thus supporting the role of this protein in interactions with host cells. The interaction of recombinant E. coli producing A. phagocytophilum MSP4 (Figure 1D) and HSP70 (Figure 1E) proteins with HL60 human cells was also characterized by electron microscopy in comparison with control E. coli cells to provide additional evidence for the role of these proteins in the interaction with vertebrate host cells.

To provide additional support for the role of A. phagocytophilum MSP4 and HSP70 proteins in the interaction with and infection of vertebrate host cells, recombinant proteins and antibodies against these proteins were used to evaluate their effect on pathogen infection of HL60 human cells. Anti-MSP4 and anti-HSP70 or recombinant MSP4 and HSP70 proteins were incubated with HL60 cells prior to infection with A. phagocytophilum. The results showed an inhibitory effect of anti-MSP4 and anti-HSP70 antibodies on infection of human cells with A. phagocytophilum human NY18 (Figure 2A) and sheep (Figure 2B) isolates when compared to cells treated with pre-immune serum. Furthermore, incubation with HSP70 and MSP4-HSP70 but not MSP4 recombinant proteins inhibited infection of human cells with A. phagocytophilum human NY18 isolate when compared to cells incubated with PBS or SUB controls (Figure 2C).

These results evidenced a role for MSP4 and HSP70 proteins in A. phagocytophilum adhesion to vertebrate host cells, and suggested a role for HSP70 during pathogen infection. These results also suggested that these proteins might constitute candidate protective antigens to prevent or control pathogen infection.

To gain additional information on the role of A. phagocytophilum MSP4 and HSP70 proteins in host-pathogen interactions, sheep that are natural hosts for this pathogen were selected for immunization with recombinant proteins followed by experimental infection with A. phagocytophilum. Groups of three lambs each were immunized with recombinant MSP4, MSP4-HSP70 combination or adjuvant/saline control and infected with a sheep isolate of A. phagocytophilum. Then, several parameters including rectal temperature, animal weight, hemoglobin content, and hematological variables were evaluated in immunized and control A. phagocytophilum-infected lambs to correlate with TBF main clinical signs (Supplementary Table 1).

The results showed signs of TBF in lambs infected with A. phagocytophilum. Evidence of A. phagocytophilum in neutrophils was obtained for all animals (Supplementary Table 1). Fever was evident in animals from all groups, primarily between 3 and 9 dpi (Figure 3A and Supplementary Table 1). Although immunized lambs tend to gain more weight, differences with controls were not significant (Supplementary Table 1). Control sheep showed evidence of anemia at 4 dpi, and between 8 and 10 dpi with all animals being anemic at 9 dpi, a result that correlated with low erythrocyte counts at 8–10 dpi (Figure 3B and Supplementary Table 1). The percent neutrophils in the leukocyte population increased in all animals after A. phagocytophilum infection between 4 and 9 dpi (Figure 3C and Supplementary Table 1). A severe neutropenia was evident in all animals after 59 dpi and lasted until the end of the experiment at 94 dpi (Figure 3C and Supplementary Table 1). Although monocyte levels were within normal values throughout the experiment, an increase was observed in all animals after infection between 2–10, 38–45, and 59–94 dpi (Figure 3D and Supplementary Table 1).

The results showed that anti-MSP4 (Figure 4A) and anti-HSP70 (Figure 4B) IgG antibody titers were higher in MSP4-immunized than in control animals from second immunization until 94 dpi. In MSP4-HSP70 immunized lambs, anti-MSP4 IgG antibody titers were significantly higher until 10 dpi (Figure 4A), while anti-HSP70 IgG antibody titers remained higher then in control animals until 94 dpi (Figure 4B). The increase in the IgG antibody response to A. phagocytophilum MSP4 and HSP70 proteins after experimental infection was higher in immunized than in control animals (Figures 4A,B), suggesting an anamnestic response that may be protective against pathogen infection.

The IgG antibody response in MSP4 and MSP4-HSP70 immunized lambs was specific for MSP4 and HSP70 proteins as supported by the dilution effect observed after incubating sera collected before infection (0 dpi) with A. phagocytophilum purified from infected HL60 human cells (Figure 4C). However, the antibody response in MSP4-immunized lambs raised a question regarding the anti-HSP70 response in these animals. Possible explanations to this question are the production of polyreactive antibodies, and the existence of common B-cell epitopes between A. phagocytophilum MSP4 and HSP70 proteins. The MSP4 and HSP70 proteins were aligned and the linear B-cell epitopes were predicted and aligned to both protein sequences (Supplementary Figure 2A). A total of 14 and 32 linear B-cell epitopes were predicted for MSP4 and HSP70, respectively. Only epitopes longer than 8 amino acids were included in further analysis, resulting in 5 and 12 linear B-cell epitopes identified in MSP4 and HSP70, respectively. Only one epitope of MSP4 (DGATGYAI) aligned without gaps to a region of HSP70 (DGQTAVTI) with 50% identity (Supplementary Figure 2A). Three epitopes from HSP70 (FNDAQRQATKDAGTI, AGIKDNSKV and SNCSTDTLQQ) aligned without gaps to regions of MSP4 (FVAVGRDATLTPDNF, AGIPASNRV and AVCACSLLIS), with 26, 44, and 20% identity, respectively. These results suggested that antibodies against MSP4 epitopes (i.e., DGATGYAI) could cross-react with a region of HSP70 (DGQTAVTI), thus explaining the anti-HSP70 response in MSP4-immunized lambs. Furthermore, these epitopes were highly conserved because A. phagocytophilum MSP4 and HSP70 protein sequences show a high homology between different strains (Supplementary Figures 2B–D). In 56 of the MSP4 sequences available containing this region, the B-cell epitope was conserved (Supplementary Figure 2D). This region was conserved in all HSP70 sequences available in GenBank (Supplementary Figure 2D).

To address the role of A. phagocytophilum MSP4 and HSP70 proteins as potential targets for the development of vaccines for the control of pathogen infection in vertebrate hosts, their potential protective capacity was characterized in immunized lambs.

The IgG antibody levels to MSP4 and HSP70 antigens remained higher after infection in immunized animals when compared to controls (Figures 4A,B). Although all animals showed signs of fever after infection with similar fever relapses, rectal temperature decreased faster in lambs immunized with MSP4-HSP70 (Figure 3A). The anemia typical of TBF was evident in control sheep at 8–10 dpi, while in immunized animals it did not occur (MSP4 group) or was observed at 9 dpi only (MSP4-HSP70 group; Figure 3B). Erythrocyte counts were not affected in immunized animals (Supplementary Table 1). The analysis of leukocytes, lymphocytes and eosinophils showed lower values at various dpi in immunized animals when compared to controls (Supplementary Table 1). In contrast, neutrophil and monocyte levels were higher in immunized animals when compared to controls at different dpi (Figures 3C,D; Supplementary Table 1). These results showed that while immunized animals presented evidence of TBF such as fever and neutropenia, the response to immunization resulted in less severe anemia in response to infection.

Although the percent of infected neutrophils was apparently higher in immunized than in control animals at some dpi, the results suggested differences in the initial infection rate despite the injection of the same amount of unfrozen infected blood (Supplementary Table 1). These differences could be explained by variations in cell viability between different inoculums, resulting in animal-to-animal variations in the initial infection rate. Therefore, the differential percent of infected neutrophils with respect to the initial value at 3 dpi was used to characterize the effect of vaccination to normalize for these differences. The results showed a significant decrease in A. phagocytophilum-infected neutrophils in animals immunized with the MSP4-HSP70, but not MSP4 antigen at 8–10 dpi when compared to controls (Figure 5A). Furthermore, the A. phagocytophilum normalized DNA levels were significantly lower in lambs immunized with MSP4 and MSP4-HSP70 antigens at 17 dpi (Figure 5B). Taken together, these results suggested that the number of infected neutrophils decreased at 8–10 dpi in response to immunization with MSP4-HSP70, while pathogen levels per cell were lower in immunized lambs when compared to controls at 17 dpi.

An antibody inhibition assay using IgG from immunized sheep at 0 and 94 dpi was conducted to further characterize the antibody response in immunized lambs in relation with the protective capacity of MSP4 and MSP4-HSP70 antigens (Figure 6). While rabbit IgG antibodies against A. phagocytophilum MSP4 and HSP70 recombinant proteins inhibited pathogen infection of HL60 human cells (Figures 2A, 7), sheep IgG collected from immunized animals before infection (0 dpi) and after infection at the end of the experiment (94 dpi) did not affect pathogen infection (Figure 6). These results evidenced differences in the IgG response between immunized rabbits and lambs, and provided support for the limited protection against A. phagocytophilum infection observed in sheep immunized with MSP4 and MSP4-HSP70.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.