1. Introduction
The family Anaplasmataceae includes bacterial pathogens that cause disease in humans and animals. These pathogens include Anaplasma marginale and A. phagocytophilum, which cause bovine and granulocytic anaplasmosis, respectively. The Ehrlichial species Ehrlichia ruminantium, E. chaffeensis, and E. canis cause ‘heartwater’ in ruminants and monocytic ehrlichiosis in humans and dogs, respectively. Overall, effective measures to prevent these diseases are lacking. The fact that these pathogens are all tick transmitted presents opportunities to develop novel, broadly applicable interventions to prevent disease. However, major fundamental knowledge gaps about the molecular interactions between the tick and the pathogen that allow for successful transmission of the pathogen limit our ability to develop such interventions. These interventions could include antibodies or other immune effectors produced by vaccination of the host and delivered during tick feeding or novel chemical inhibitors.
With few exceptions, iron is an essential and limiting nutrient for bacterial pathogens and their hosts. It is a critical component of eukaryotic and prokaryotic cellular functions and is indispensable for nucleic acid and lipid synthesis, protein translation, and energy metabolism and generation [
1,
2]. Because iron is highly reactive and essential, it is tightly regulated in eukaryotic systems, and bacterial pathogens have developed varied molecules and mechanisms for obtaining iron from the eukaryotic host. Thus, methods that block the ability of the pathogen to acquire iron may serve as the foundation of intervention strategies.
However, we know very little about the requirement for, or the molecules and mechanisms that mediate iron uptake in
Anaplasma and
Ehrlichia spp. during tick colonization. On the basis of the presence of a heme biosynthesis pathway,
Anaplasma and
Ehrlichia spp. are expected to require iron. Deferoxamine, an iron chelator, nearly completely abrogates
E. chaffeensis and partially reduces
A. phagocytophilum infection in monocytes [
3]. A robust understanding of iron uptake and metabolism in macrophages has served as the foundation for understanding how
E. chaffeensis and other pathogens acquire iron from host cells [
4,
5].
Relative to macrophages and other mammalian cells, iron uptake and metabolism in ticks is poorly understood. Because ticks lack heme oxygenase and thus do not acquire iron from heme, it has been proposed that ticks may acquire iron from mammalian transferrin in the blood meal [
6]. However, ticks apparently lack transferrin receptors (TfR), and the mechanisms and molecules used to acquire iron from the blood meal are unknown. Additionally, the mechanisms of iron regulation are poorly understood. For example, in vertebrates, the iron-responsive protein (IRP) and iron-responsive elements (IREs) interact to form a regulatory system that controls translation of many genes involved in iron metabolism, including ferritins that store cytosolic iron, TfR1, and divalent metal transporter 1 (DMT1), among others. In contrast, ferritin 1 is the only gene in ticks with an IRE regulating the translation of ferritin 1, which stores cytosolic iron in the midgut [
6]. These differences in iron uptake and metabolism in tick cells as compared to mammalian cells suggest the requirement and strategies used by the pathogen may vary depending on the species of the host and type of host cell.
Based on bioinformatics, there are few identifiable genes or iron-dependent regulators directly involved in iron uptake in
Anaplasma and
Ehrlichia spp. [
7,
8]. These pathogens apparently lack iron-dependent repressors, iron response regulators, heme and transferrin receptors, heme oxygenase required to liberate iron from heme, and siderophores. However, three genes are predicted to be involved in iron uptake and are conserved among
Ehrlichia and
Anaplasma spp. In
A. marginale, these genes are
Am069,
Am240, and
Am392 (locus identifiers based on the St. Maries strain, RefSeq GCF_000011945.1). The proteins encoded by these genes are orthologs of the siderophore-independent iron uptake systems first defined in
Neisseria spp. (nFbpABC) and
Haemophilus influenzae (hFbpABC) and found in many other Gram-negative pathogens [
9,
10,
11]. Typically, these genes are encoded on an operon and are under the negative regulatory control of the ferric uptake regulator (
fur) [
10,
12,
13]. However, the
Anaplasmataceae lack a
fur regulator, and these genes are dispersed throughout the genome rather than being encoded in an operon. There are no data regarding a possible role of these genes in iron transport.
In this study, we used A. marginale, which is transmitted by Dermacentor spp. ticks, to address these knowledge gaps by testing two hypotheses: (1) iron is required for A. marginale replication in D. andersoni cells; (2) the FbpA, FbpB, and FbpC orthologs in A. marginale are differentially expressed in response to iron deprivation during tick cell colonization. The findings are discussed and reviewed in the context of knowledge gaps in our understanding of iron acquisition by obligate intracellular, tick-borne bacterial pathogens.
3. Discussion
Together, our results support the hypothesis that iron is required for A. marginale replication in tick cells. This conclusion is based on the following: (1) live A. marginale as detected through transcript was readily detected, even at later time points of iron reduction; (2) A. marginale levels increased with the restoration of iron, indicating replication; (3) A. marginale colony size, according to fluorescent microscopy, remained static in the face of iron reduction, but increased with the restoration of iron, directly indicating replication. However, iron reduction also likely leads to death of some A. marginale. For example, between 24 h and 48 h of iron reduction, there was 1.5-fold reduction in A. marginale levels. The loss was more marked between 24 h and 72 h with a 2.6-fold decrease in A. marginale levels, although between 72 and 96 h, the levels remained stable.
Additionally, despite the lack of a
fur regulator,
A. marginale has some limited ability to respond to altered host cell iron levels. An important caveat is the inability to isolate the effects of iron deprivation on the host cells from the direct effects of iron deprivation on the intracellular bacteria. The approach used to deplete iron in DAE100 cells resulted in measurably decreased iron levels according to ICP/MS and a decrease in ferritin 1 protein levels in the absence of cell death [
14]. Additionally, the overall low cytotoxicity of the treatment provides some assurance of the health of the cells during the experiments. However, negative effects of Bpdl and EtOH on the tick cells were measured using exclusion of trypan blue stain. In essence, this is a measure of membrane integrity. It is possible that the EtOH or Bpdl had cytotoxic effects that were not yet reflected in alteration in membrane integrity, and consequently the cytotoxic effects of EtOH and Bpdl were underestimated. For example, the decrease in
Am069 and increase in
Am392 transcripts at 96 h of iron deprivation was the opposite of what one would expect on the basis of earlier time points, suggesting some level of metabolic derangement in the cells at the later time points.
Bacterial pathogens influence iron uptake and availability in host cells. In the case of the
Anaplasmatceae, many knowledge gaps remain in our understanding of iron acquisition (
Figure 8). Additionally, the iron uptake and transport mechanisms in the
Anaplasma spp. differ from those of the
Ehrlichia spp. In
E. chaffeensis, for example, transferrin receptors accumulate on the pathogen-containing vacuole early in infection of mammalian monocytic cell lines, and the
E. chaffeensis vacuole interacts with the TfR recycling endosome [
3,
29]. Decreased transcription of TfR markedly inhibits
E. chaffeensis infection, and this effect can be reversed with the addition of holotransferrin [
3,
29]. Additionally, the
E. chaffeensis vacuole is mildly acidic, thus facilitating the dissociation of iron from transferrin [
29]. Together these data indicate that holotransferrin is one source of iron for
E. chaffeensis in monocytes. In a related or additional mechanism for iron acquisition, an
E. chaffeensis effector protein, Etf-3, is exported into the macrophage cytoplasm via the type IV secretion system (T4SS) and induces breakdown of ferritin and release of Fe
2+ into the cytosol, thus increasing availability of iron for
E. chaffeensis [
5].
In contrast, transferrin receptors do not co-localize with
A. phagocytophilum-containing vacuoles in monocytes [
3]. Erythrocytes, the primary host cell for
A. marginale in mammals, do not express transferrin receptors, while ticks lack identifiable transferrin receptors [
14]. Additionally, the
A. marginale-containing vacuole in
D. andersoni cells has a neutral pH [
30]. Thus,
Anaplasma spp. likely use iron from sources other than transferrin. Additionally,
A. marginale and
A. phagocytophilum lack an ortholog of Eft-3, although enhancing ferritinophagy via other effectors or mechanisms is a possible means of iron acquisition in tick cells.
The final steps in iron uptake by Gram-negative bacteria are transport across the bacterial outer membrane, periplasm, and inner membranes.
Neisseria spp. have a transferrin-binding protein, TbpA, which binds holo-transferrin and then passes the iron to nFbpA for transport across the periplasm [
15,
31]. While an outer membrane-exposed iron-binding protein has not been identified in the
Anaplasmataceae, they all have orthologs of the siderophore-independent iron transporter, FbpA, FbpB, and FbpC (
Figure 8). Unlike other Gram-negative bacteria that have this system, the
Anaplasmataceae, including
A. marginale, lack an identifiable
fur regulator, and these three genes do not form an operon. However, transcript levels of these genes in
A. marginale were modestly responsive to the manipulation of iron levels in tick cells, although the responses were not homogeneous among genes.
Am069, the predicted periplasmic, iron transport protein, was upregulated in response to iron reduction, similar to the orthologs nFbpA and hFbpA in
Neisseria spp. and
Haemophilus spp., respectively, as well as FutA1 of
Synechosystis spp. [
9,
13,
32]. This response was modest, which may reflect a reduced ability to respond to environmental changes compared to facultative bacteria, likely due to their residence in a stable intracellular niche. Similarly, the related pathogen,
Rickettsia rickettsii, lacks a
fur regulator, but has a limited transcriptional response to iron depletion in host cells [
33].
In contrast, Am240 and Am392 were both downregulated in the first 24–72 h of iron reduction, although expression levels recovered once iron levels were restored, indicating some ability to adapt to altered cellular environments. The difference in transcriptional response between the periplasmic iron transporter (Am069), the permease (Am240), and the ATPase (Am392) may reflect their different roles in iron transport. In the face of iron reduction, the increase in Am069 transcript may be a direct response to increased demand for iron, while the decrease in Am240 and Am392 may reflect a more general response to iron reduction and a resulting decrease in the ability of the bacteria to generate energy.
Interestingly, two of the three proteins with the greatest structural similarity to Am069 are FutA proteins from Cyanobacteria (
Synechocyctis sp. and
T. erythraeum), while the third is cFbpA from
C. jejuni (
Table 1). Unlike the canonical nFbpA and hFpbA,
Synechocystis sp. FutA1 and
C. jejuni cFbpA preferentially bind Fe
2+, while
T. erythraeum is thought to bind Fe
3+, although there is less experimental data for this organism [
20,
21,
22]. Cyanobacteria, as one of the oldest lifeforms on earth, evolved in low-oxygen conditions and are thought to have maintained the ability to transport Fe
2+, which is more abundant in anaerobic conditions, but acquired ferrireductases in order to adapt to an aerobic environment [
20].
C. jejuni has an intracellular life stage that is likely microaerophilic, which may account for the binding preference for Fe
2+ of cFbpA.
A. marginale, as an obligate intracellular bacterium, also likely resides in a microaerophilic environment [
34]. Thus, it is possible that Am069 has an affinity for Fe
2+, although this must be experimentally demonstrated.
Synechocystis Fut1A,
C. jejuni cFbpA, and
T. erythraeum FutA coordinate iron in a different fashion than the canonical nFbpA and hFbpA, which use a histidine, glutamate, two adjacent tyrosine residues, and a phosphate counter ion (
Table 1,
Figure S1). In contrast, on the basis of X-ray crystallography,
C. jejuni cFbpA and
Synechocystis sp. FutA1 use one histidine and four tyrosine residues for iron coordination (
Table 1,
Figure S1) [
20,
22].
T. erythraeum FutA uses three tyrosine residues, and no counter ion, although this protein maintains a conserved histidine and tyrosine residues that are involved in iron coordination in FutA1 of
Synechocystis sp. and cFbpA of
C. jejuni (
Table 1,
Figure S1) [
21]. Iron coordination in
A. marginale is predicted to be most similar to
T. erythraeum FutA using only three tyrosine residues (
Table 1,
Figure S1). Unlike,
T. erythraeum FutA, Am069 does not have the conserved histidine and tyrosine residues in the N terminus of the proteins that contribute to iron coordination in
Synechocystis FutA1 and
C. jejuni cFbpA (
Figure S1).
In summary, we demonstrated that
A. marginale requires iron for replication in tick cells. Based on bioinformatics and protein modeling, we found that
Am069,
Am240, and
Am392, although not organized in an operon, are orthologs of the siderophore-independent iron transport system, FbpA, found in many Gram-negative bacteria (
Figure 8). Although
A. marginale lacks a
fur regulator, these genes are differentially expressed in response to iron starvation and recovery. Additionally, FutA1 from the Cyanobacterium
Synechocystis sp. has the greatest structural similarity to Am069. Protein localization experiments are required to determine if any portion of Am069 is surface-exposed, and thus a potential vaccine target. More broadly, X-ray crystallography, heterologous expression, and iron binding experiments must be performed to determine if Am069 is required for iron transport, the binding preference for Fe
2+ as compared to Fe
3+, and to determine how iron coordination is achieved. Together, this information will help build our understanding of the unique physiology of the
Anaplasmataceae.
4. Methods
4.1. DAE100 Cell Culture Conditions and A. marginale Strain
DAE100 cells (source: Dr. Uli Munderloh in the department of Entomology at the University of Minnesota) isolated from
D. andersoni embryonated eggs were used in all experiments [
35]. Cells were maintained at 34 °C and grown in L15B complete medium as previously described [
36]. The St. Maries strain of
A. marginale, originally isolated at Washington State University and maintained by the USDA-ARS Animal Disease Research Unit at Washington State University, was used in all experiments [
37]. L15B buffered with HEPES (Millipore-Sigma, Burlington, VT, USA), NaHCO
3 (Millipore-Sigma, Burlington, VT, USA) and NaOH (Millipore-Sigma, Burlington, VT, USA) (
Anaplasma medium) was used to maintain
A. marginale infected DAE100 cells [
38]. To ensure iron stores were adequately depleted, DAE100 cells used in iron depletion experiments were grown in L15B medium without ferrous sulfate (FeSO
4) for two weeks prior to each experiment [
14].
4.2. Production, Storage, and Quantitation of Host Cell-Free A. marginale
To isolate A. marginale free of host cells, flasks of heavily infected DAE100 cells were pelleted via centrifugation at 1800× g for 15 min. The cell pellet was resuspended in sucrose-phosphate-glutamate buffer (SPG) made of 230mM sucrose (Avantor, Allentown, PA, USA), 4.41 mM KH2PO4 (Avantor, Allentown, PA, USA), 6.98 Mm K2HPO4 (Millipore-Sigma, Burlington, VT, USA), and 4.92 mM C5H8KNO4·H2O (Millipore-Sigma, Burlington, VT, USA) dissolved in water (ThermoFisher Scientific, Fitchburg, WI, USA), vortexed, then sonicated in a cup horn sonicator (Fisherbrand Model 705 Sonic Dismembrator) at 30% amplitude until ≈90% of cells had lysed, as visualized on a wet mount slide. Cellular debris was pelleted by centrifugation at 200 rcf for 5 min. The A. marginale was aliquoted into cryovials and preserved at −80 °C.
Anaplasma marginale DNA was extracted from cryopreserved SPG stocks using a DNeasy Blood and Tissue kit (Qiagen, Valencia, CA, USA). The quantity of
A. marginale per microliter of SPG stock was calculated using quantitative PCR and a standard curve with
msp5-specific primers and PerfeCTa SYBR Green FastMix (Quantabio, Beverly, MA, USA) [
39]. As
msp5 is a single copy gene, the number of msp5 copies reflects the number of bacteria (
Table 2).
4.3. Experimental Design Overview
Cells were infected with
A. marginale 24 h prior to the initiation of treatments, except for the experiments to determine the cytotoxic effects of Bpdl and EtOH on DAE100 cells. To induce iron depletion in all experiments, DAE100 cells were treated every 24 h up to 96 h with Bpdl (
Figure 2A). To determine if the effects of iron depletion were reversible,
A. marginale-infected DAE100 cells were treated with Bpdl for 24 h only. Following this treatment, cells then received iron-replete
Anaplasma media for the reminder of the experiment (
Figure 2B). To confirm that the reduction in
A. marginale levels was due to iron reduction in the host cell rather than unknown effects of the chelator,
A. marginale-infected cells were treated every 24 h with Bpdl plus various concentrations of FeSO
4 (
Figure 2A). Cells treated with the carrier only, EtOH, served as the control.
Cells were harvested starting 24 h following the first treatment and each 24 h thereafter. Cells treated with Bpdl or the carrier were harvested up to 96 h post-treatment, while cells treated with Bpdl plus FeSO4, were harvested up to 72 h post-treatment. Following harvest, cells were stained with trypan blue to determine viability, or total RNA was extracted to enumerate A. marginale levels relative to DAE100 cells or Am069, Am240, and Am392 transcript levels relative to the housekeeping gene rpoH, using RT-qPCR and calculation of the 2−∆∆Ct.
4.4. Iron Depletion and Infection of DAE100 Cells
For cytotoxicity, iron deprivation and reversal of iron deprivation studies, cellStar T25 flasks (Greiner Bio-One, Monroe, NC, USA) were seeded with 5 mL of medium containing 5 × 105 cells/mL. DAE100 cells were infected at an A. marginale multiplicity of infection (MOI) of 350:1. For the rescue of A. marginale with the addition of FeSO4 in the presence of Bpdl, 24-well CellStar cell culture plates (Greiner Bio-One, Monroe, NC, USA) were seeded with 5 × 105 of DAE100 cells in 500 µL of medium. Because of the smaller area of the wells as compared to T25 flasks, an MOI of 100:1 was used.
To infect DAE-100 cells, vials of cell-free A. marginale cryopreserved in SPG buffer were thawed and pelleted by centrifugation at 13,000× g for 5 min. The A. marginale pellets were resuspended in either buffered L15B complete medium or in L15B complete medium without FeSO4. The infected flasks or plates were spun at 200× g for 5 min to force contact between the cells and bacteria, then incubated at 34 °C for 2 h to allow pathogen entry. After 2 h, the medium was replaced with Anaplasma medium using gentle pipetting, and the cells were incubated for 24 h. Plates were incubated in a sealed BD Campy Container with a GasPak (Scientific Equipment Company, Aston, PA USA).
At 24 h post-infection, cells were fed with regular
Anaplasma medium,
Anaplasma medium containing 100 µM of Bpdl (Sigma Aldrich, St. Louis, MO, USA), or
Anaplasma medium containing 2 µL/mL of 100% of EtOH (Koptec, King of Prussia, PA, USA), as the carrier control for Bpdl. Treatments continued every 24 h for 96 h. At each time point, 24 h, 48 h, 72 h, and 96 h post-treatment, cells were pelleted and frozen in RNAlater (Ambion, Austin, TX, USA) at −80 °C for RNA extraction (
Figure 1A). These experiments were performed four independent times.
4.5. Cytotoxicity of Iron Chelator on DAE100 Cells
Trypan blue, which only stains non-viable cells, was used to measure the cytotoxicity of 100 µM of Bpdl and EtOH on uninfected and A. marginale-infected DAE100 cells. Cells were harvested every 24 h starting 24 h after treatment with Bpdl or EtOH for 96 h. For staining, cells were mixed 1:1 with 0.4% trypan blue (Millipore-Sigma, Burlington, VT, USA), and viable and non-viable cells were counted using a hemocytometer and light microscopy. Three technical replicates were counted in each experiment. The experiment was performed three independent times.
4.6. Reversal A. marginale Growth Inhibition
To determine if the effect of iron depletion on
A. marginale replication was reversible, infected DAE100 cells were treated with 100 µM Bpdl for 24 h. Cells then received
Anaplasma medium for the remainder of the experiment. Cells were treated and harvested every 24 h up to 96 h and stored for RNA extraction as described above (
Figure 1B). These experiments were performed three independent times.
4.7. Rescue of A. marginale Replication by Addition of FeSO4
In order to restore iron availability in the presence of 100 µM Bpdl, A. marginale-infected DAE100 cells were treated daily for 24 h, 48 h, and 72 h with 20 mM, 50 mM, or 100 mM FeSO4 (Millipore-Sigma, Burlington, VT, USA). Controls groups were treated with either 100% EtOH at 2 µL/mL or 100 µM Bpdl on the same schedule. These experiments were performed four independent times.
4.8. RT-qPCR
Specific primers were designed to amplify
rpoH,
Am069,
Am240, and
Am392 of
A. marginale and
β-actin of
D. andersoni using Integrated DNA Technologies (IDT) and were tested for amplification efficiency (
Table 2). RNA was extracted from DAE100 cells using an RNeasy Mini Kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions. Extracted RNA was treated with DNAse twice using a Turbo DNase Kit (Ambion, Austin, TX, USA) and cleaned using an RNA Clean and Concentrator-25 kit (Zymoresearch, Irvine, CA, USA). RNA concentrations were measured using a Nanodrop ND-1000 spectrophotometer (ThermoFisher Scientific, Fitchburg, WI, USA). RNA was converted to cDNA using the Superscript III Reverse Transcriptase kit (Invitrogen, Carlsbad, CA, USA) as described in the manufacturer’s protocol.
Next, a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) was used to perform RT-qPCR. Each RT-qPCR reaction contained cDNA, SsoAdvanced™ Universal SYBR Green Supermix (Bio-Rad, Hercules, CA, USA), RNase/DNase free water, and 20 nM forward and reverse primers (Integrated DNA Technologies, Coralville, IA, USA). Thermocycling conditions for all reactions were as follows: denaturation at 95 °C for 30 s, then 40 cycles of 95 °C for 5 s, 60 °C for 15 s, followed by a melt curve analysis from 65–95 °C in 0.5° increments for 5 s/step.
2−∆∆CT was used to quantify A. marginale relative to DAE100 cells. Specifically, the ratio of the abundance of A. marginale rpoH transcript in treated cells (Bpdl, Bpdl-24 h only, and Bpdl + FeSO4) to EtOH-only treated cells (at the same time post infection) was normalized to D. andersoni β-actin under the same treatment conditions. Using D. andersoni β-actin compensates for any differences in the number of DAE100 cells. The same approach was used to measure the relative transcript levels of Am069, Am240, and Am392. In this case, the ratio of the abundance of the target gene (Am069, Am240, or Am392) transcript in treated cells (Bpdl or Bpdl-24 h only) to the target gene in EtOH-only treated cells (at the same amount of time post A. marginale infection) was normalized to expression of the housekeeping gene, rpoH, under the same conditions. The use of rpoH compensates for differences in A. marginale levels between samples.
4.9. Immunofluorescent Microscopy and Image Analysis
To visually evaluate the effects of iron reduction on
A. marginale growth, 1 × 10
6 DAE100 cells were seeded in a Nunc™ Lab-Tek™ II Chamber Slide™ System (ThermoFisher Scientific, Fitchburg, WI, USA). Cells were infected with
A. marginale using an MOI of 350:1, spun at 200×
g for 5 min, then incubated for 2 h before replacing the media via gentle pipetting. Cells were treated daily with Bpdl or EtOH, as described above (
Figure 1). At each time point, slides were fixed with 4% PFA (Invitrogen, Carlsbad, CA, USA) for 10 min, then rinsed well with phosphate buffered saline PBS (Quality Biological, Gaithersburg, MD, USA). All antibodies were diluted to 0.5 µg/mL in PBS with 0.05%, Tween 20 (Thermo Scientific Pierce, Fitchburg, WI, USA) and 3.0% bovine serum albumin (Millipore-Sigma, Burlington, VT, USA).
Anaplasma marginale outer membranes were labeled for 1 h with AnaR49A1 anti-Msp2 primary antibody [
30] (source: Guy H. Palmer, in the Paul Allen School for Global Health at Washington State University), rinsed, then labeled with Goat anti-Mouse IgG (H+L) Alexa Fluor 594 secondary antibody (Invitrogen, Carlsbad, CA, USA) and Alexa Fluor 488 Phalloidin (Invitrogen, Carlsbad, CA, USA) for 1 h. Slides were rinsed and mounted with Prolong Diamond Antifade with DAPI (Invitrogen, Carlsbad, CA, USA) and coverslips.
Deconvolution microscopy was performed using an Axio Imager.M1 microscope (Carl Zeiss Microimaging, Thornwood, NY, USA) equipped with an X-Cite 120 Fl Illuminating system (EXFO Photonic Solutions, Mississauga, ON, Canada) and an AxioCam MR3 digital camera. Image acquisition was performed under two settings. To estimate bacterial load (Alexa Fluor 594) and cellularity (Alexa Fluor 488) for each treatment at each time point, six non-adjacent regions were blindly selected and imaged as 512 × 512 pixel two-channel image stacks using a Plan-Apochromat 20×/0.8 M7 objective (pixel scaling: 0.324 nm × 0.324 nm) at z-axis steps of 250 nm. Examples of bacterial colony morphology (Alexa Fluor 594) were then acquired as 256 × 256 pixel single-channel image stacks using a Plan-Apochromat 63×/1.4 oil M7 objective (pixel scaling: 0.103 × 0.103 nm) and 200 nm z-axis steps. Treatment with Bpdl reduced DAE100 cell adherence to slides in a time-dependent manner. Thus, only four of six image stacks were suitable for measurement of the ‘Bpdl 24 h only’ treatment group at 72 h, and none of the image stacks for either Bpdl treatment group were suitable for measurement at 96 h. All image processing and measurements were done using the Fiji distribution of ImageJ software (
http://fiji.sc/ accessed on 23 March 2022) [
40].
Deconvolution of original image stacks was accomplished using the DeconvolutionLab2 plugin [
41]. For image stacks acquired using the 20× objective, the Richardson–Lucy Total Variation algorithm (lambda 0.001) was used with z-axis Tukey apodization and a residual less than 0.01 as a final iteration constraint. For 63× objective acquisitions, the Richardson–Lucy algorithm was used with no apodization and a residual less than 0.001 as a final iteration constraint. Theoretical point spread functions were generated using the Richards and Wolf 3D optical model as implemented in the PSF Generator plugin [
42].
The ‘auto local threshold’ function was used to objectively segment fluorescent structures, using the image stack histogram and default settings of the Bernsen algorithm for Alexa Fluor 488, and the Phansalkar algorithm for Alexa Fluor 594. The ‘analyze particles’ function was used to measure the area (sq microns) of each discrete threshold structure present in each image stack slice. This experiment was performed once.
4.10. In Silico Identification of A. marginale Genes Involved in Iron Transport
To identify A. marginale genes potentially involved in iron transport, we first used the KEGG to search for orthologs of ABC transporters that bind and transport iron. The three resulting genes, Am069, Am240, and Am392, were translated in silico, and the CDD from NCBI was used to search for conserved domains.
On the basis of KEGG orthology, Am069 is predicted to bind iron, but this was not supported by the presence of conserved domains. To help address this discordance, we performed structural modeling using I-TASSER [
16,
17,
18]. In contrast, Am240 and Am392 both have conserved domains that give a strong indication of function, and thus structural modeling for these two proteins was unnecessary.
For structural modeling of Am069, the amino acid sequence was uploaded into I-TASSER. The resulting best structural model among the five provided by I-TASSER is reported in the results. Additionally, the three proteins in the PDB with the closest structural similarity to Am069, on the basis of TM-align scores, as provided by I-TASSER, are also reported. Finally, the overlay between the predicted structure of Am069 and best PDB match are presented (
Figure 6).
For
Figures S1–S3, multiple sequence alignments were performed with SnapGene (GSL Biotech LLC, San Diego, CA, USA) using the Clustal Omega alignment algorithm and the Clustal X coloring scheme. TMHMM was used to predict the transmembrane domains for Am 240 and its orthologs, as presented in
Figure S2 [
43].
4.11. Statistical Analysis
GraphPad Prism V9 Software (San Diego, CA, USA) was used for most of the statistical analyses. The data from the cytotoxicity assays and transcriptional analysis of Am069, Am240, and Am392 were analyzed by two-way ANOVA followed by a Dunnett’s test for multiple comparisons. Raw numbers from the cytotoxicity assays were used in the statistical analysis. The data from RT-qPCR experiments quantitating A. marginale relative to tick cell numbers were log2 transformed, and significant differences between groups were determined using two-way ANOVA followed by a Tukey’s test for multiple comparisons.
Intracellular growth of bacterial colonies was measured as the total area of threshold anti-Msp2 fluorescence divided by the total area of threshold phalloidin. The effects of treatment, incubation period, and the interaction term were analyzed using a generalized linear model (PROC GLIMMIX; SAS version 9.4, SAS Institute Inc., Cary, NC, USA). The data were well fit by the beta distribution, and comparisons of interest were made by including an LSMEANS statement for the interaction term sliced by the incubation periods and significance values adjusted by the step-down Bonferroni (Holm) procedure. The effects of treatments and time on colony sizes were compared by visual inspection of empirical cumulative distribution functions (PROC NPAR1WAY, EDF; SAS).