First molecular and functional characterisation of ferritin 2 proteins from Ornithodoros argasid ticks

Ferritins are iron-binding proteins that play critical functions in iron metabolism. Tick ferritins are essential in blood feeding, reproduction, iron transport, and protection of ticks from the iron-mediated oxidative stress during blood feeding and digestion. In ixodids, ferritin 2 (Fer2) is responsible for iron transport into peripheral tissues, it is critically involved in tick reproduction and has been identified as a good candidate antigen to be included in anti-tick vaccines. In argasids, information on the molecular and functional characteristics of ferritins is almost nonexistent. Given the potential of ixodid Fer2 as a vaccine target, the aim of the current study was to characterise the Fer2 orthologues in Ornithodoros erraticus (OEFer2) and O. moubata (OMFer2), including functional analyses by RNAi gene knockdown and the assessment of the protective efficacy of recombinant Fer2 protein in an animal vaccination trials. Characterisation and analysis of the OMFer2 and OEFer2 amino acid sequences showed high similarity to each other, and high similarity to the Fer2 sequences of ixodid species as well, confirming that Fer2 is highly conserved between both tick families and suggesting a similar function in the physiology of both argasid and ixodid ticks. Fer2 gene knockdown in O. moubata reduced egg hatchability rate and the subsequent number of emerging nymphs-1 up to 71%. Conversely, Fer2 gene knockdown in O. erraticus did not affect the treated ticks even though the Fer2 mRNA expression level was reduced by 90%. The recombinant form of O. moubata Fer2 (tOMFer2) was highly immunogenic and induced strong humoral responses when administered to rabbits formulated with Montanide adjuvant. The protective effect of the anti-tOMFer2 response was limited. While in O. erraticus, we did not observe any protective effect, in O. moubata it induced a significant reduction in oviposition without affecting the other parameters analysed. Accordingly, Fer2 seems to be involved in O. moubata embryogenesis. This study provides the first data on the molecular and functional characterisation of Fer2 in soft tick species and paves the way for further studies aimed at unveiling the functional aspects of Fer2 in soft ticks and confirming its potential as a vaccine candidate antigen.


Introduction
Ticks are haematophagous arthropods that belong to two major families, Ixodidae (hard ticks) and Argasidae (soft ticks). Usually, ixodids are exophilic ticks that persist on the soil and vegetation and actively seek hosts when the season is suitable. After attachment to an appropriate host, they feed for several days, ingest huge amounts of blood and, once engorged, drop off the host returning to the soil, where they moult or, in the case of females, oviposit and die. By contrast, argasid ticks are endophilic/nidicolous parasites. In nature, they live inside the nests and burrows of their hosts. In synanthropic environments, they colonise animal facilities and human dwellings, hiding in holes, cracks, and fissures of the floor, walls, and ceiling where they remain protected from adverse climatic conditions and have regular access to host blood. Most argasids are fast feeders, taking less than one hour to complete engorgement. After that, they drop off the host and hide inside their shelters to moult or reproduce. Adult specimens can feed and reproduce several times and are resistant to starvation, surviving for years without feeding (Sonenshine and Roe, 2014).
Ticks transmit a wide variety of infectious diseases worldwide; in fact, ticks are among the most important vectors of pathogens affecting livestock, humans, pets, and wildlife (Jongejan and Uilenberg, 2004). Among argasid ticks, several species of Ornithodoros are vectors of microbial agents causing severe diseases. O. erraticus and O. moubata are the main vectors of the African swine fever (ASF) virus and the human relapsing fever spirochetes Borrelia hispanica and B. crocidurae in the Mediterranean and B. duttoni in mainland Africa (Cutler, 2010;Boinas et al., 2014;Sánchez-Vizcaíno et al., 2015).
Traditional measures for controlling ticks and tick-borne pathogens are primarily based on using chemical acaricides. However, the application of these products has drawbacks, such as the rapid development and selection of resistant tick strains, pollution of the environment, and contamination of products of animal origin (Willadsen, 2006). Some decades ago, these problems already raised the need to develop alternative tick control methods. Immunological control with anti-tick vaccines was the method of choice because, among other advantages, its application could avoid the aforementioned drawbacks. The first two anti-tick vaccines were commercialised in the 1990 s: TickARD in Australia and GAVAC in Latin America. Both vaccines were developed to fight the ixodid tick Rhipicephalus microplus and were based on the intestinal antigen Bm86, a glycoprotein expressed in the luminal membrane of the enterocytes of the tick midgut (de la Fuente et al., 1999;Rodríguez-Valle et al., 2012). Although with variable efficacy, the application of these vaccines has demonstrated that vaccination of hosts is a sustainable method for the control of tick populations (de la Fuente et al., 2007). Since then, important research efforts have been made for developing new and more efficacious tick vaccines Ribeiro et al., 2021;Tabor, 2021). Success in tick vaccine development largely depends on identifying new and highly protective antigens. This task may be accomplished by selecting candidate protective antigens with important biological functions for ticks and that share conserved structural and sequence motifs, which would facilitate the simultaneous control of several tick species (de la Fuente and Contreras, 2015;de la Fuente et al., 2016).
Ticks are known for their strict host-blood-dependent lifestyle, relying absolutely on their host's blood as the unique source of nutrients needed for their survival and reproduction (Mans and Neitz, 2004). To obtain their blood meal from the host, ticks have evolved a range of redundant mechanisms to circumvent the host's haemostatic, inflammatory and immune responses. Additionally, ticks have evolved specific strategies to neutralise the oxidative damage caused by the potentially toxic molecules released upon the digestion of the host blood, such as the excess of iron and haem group (Hajdusek et al., 2016;Whiten et al., 2018). Specifically, after a blood meal, part of haemoglobin-derived haem is directly absorbed for vitellogenesis, and the excess haem is concentrated in midgut digestive cell organelles that specialise in haem sequestration, termed hemosomes (Lara et al., 2003). Additionally, ticks ingest huge amounts of iron, which is essential for tick survival. However, iron excess is toxic, and its homeostasis must be tightly regulated (Galay et al., 2013).
Ferritins (Fer) are iron-binding proteins involved in iron homeostasis and present in almost all organisms (Arosio et al., 2009). Fer have been identified in ticks, where they play critical functions in iron metabolism and are essential in blood feeding, reproduction, iron transport, and protection of the tick from iron-mediated oxidative stress during blood feeding and digestion (Hajdusek et al., 2010;Galay et al., 2014aGalay et al., , 2014b. Two Fer have been described in ixodid ticks: ferritin 1 and 2. Ferritin 1 (Fer1) is an intracellular protein, it is involved in iron storage and homeostasis and is closely related to the mammalian heavy-chain ferritins. Ferritin 2 (Fer2) is a secreted protein without a known functional orthologue in vertebrates (Hajdusek et al., 2009). Fer2 is mainly synthesised in the tick midgut during the tick developmental stages and is secreted to the tick haemolymph. Fer2 functions as an iron transporter from the gut to the peripheral tissues to meet iron requirements (Hajdusek et al., 2009). RNA interference (RNAi) gene knockdown of tick Fer2 negatively impacted tick feeding performance, oviposition, and egg hatching, suggesting that Fer2 would be an interesting target antigen for anti-tick vaccines (Hajdusek et al., 2009;Galay et al., 2013). This was later confirmed in several vaccine trials using recombinant Fer2 from different ixodid species, including Ixodes ricinus, I. persulcatus, R. microplus, R. annulatus, Hyalomma antolicum, and Haemaphysalis longicornis. In these trials, a reduction in fed body weight, oviposition, and hatching rate was observed in the ticks fed on vaccinated hosts (Hajdusek et al., 2010;Galay et al., 2014a;Manjunathachar et al., 2019;Githaka et al., 2020).
All the functional studies performed hitherto on tick molecules involved in iron transport and metabolism have been carried out in ixodid species (Galay et al., 2015;Hajdusek et al., 2016), while equivalent studies on iron homeostasis in argasid ticks are nonexistent. Regarding argasid ferritins, to our knowledge, the only information available is the nucleotide sequences of Fer1 from O. moubata, O. coriaceus, and O. parkeri (Kopácek et al., 2003;Francischetti et al., 2008aFrancischetti et al., , 2008b, the identification of Fer1 in the midgut proteomes of O. moubata and O. erraticus (Oleaga et al., 2015(Oleaga et al., , 2017, and the recent finding of several transcripts homologous to the ixodid Fer2 in the midgut transcriptomes of O. erraticus and O. moubata (Oleaga et al., 2017b(Oleaga et al., , 2018.
In the light of the above, and given the potential of Fer2 as a vaccine target, in the present study, we aimed to characterise the Fer2 orthologues of O. erraticus and O. moubata, including functional analyses by RNAi gene knockdown and the assessment of their protective efficacy as vaccine antigens against both argasid species. This study provides the first data on the molecular and functional characteristics of Fer2 in soft tick species.

Ticks and tick material
The O. moubata and O. erraticus tick specimens used in the current study were obtained from two laboratory colonies kept in the IRNASA (CSIC) insectary. The O. moubata colony was initiated from specimens kindly donated by Dr Philip Wilkinson (Institute for Animal Health, Pirbright, United Kingdom), originally from Malawi (13 • 59 ′ 00 ′′ S 33 • 47 ′ 00 ′′ E). The O. erraticus colony was established from specimens captured in nature in the province of Salamanca (Spain) (40 • 58 ′ 00 ′′ N 5 • 39 ′ 50 ′′ O). Both colonies were maintained at 28 • C, 85% relative humidity, and are regularly fed on rabbits.
Tick midguts were obtained from O. moubata and O. erraticus females. The ticks were dissected in sterile phosphate-buffered saline (PBS) pH 7.4 at 4 • C, and the midguts were removed and collected in batches of 20 and 40 specimens for O. moubata or O. erraticus, respectively, and immediately preserved in RNAlater until RNA extraction. Additional batches of midguts of both species, obtained from unfed females and females at 48 h after feeding, were collected in PBS to prepare extracts of soluble and membrane proteins as previously described (Oleaga et al., 2017a). Protein concentrations in midgut extracts were measured using the DC Protein Assay kit I (Bio-Rad), and samples were stored at − 20 • C.

Reverse transcription PCR amplification, cloning, and sequencing of ferritin 2 orthologues
Total RNA was extracted from midgut tissues and purified with the RNeasy Mini Kit (Qiagen) following the manufacturer's instructions. Reverse transcription (RT) was performed from total RNA using the 1st Strand cDNA Synthesis kit (Roche) and the oligo(dT) 15 primer according to the manufacturer instruction.
PCR amplification of Fer2 cDNA was performed using specific primers designed with the Primer3 Plus software (Untergasser et al., 2012) from the nucleotide sequences of O. erraticus Fer2 (OEFer2) (GenBank accession number GFWV01008002) and O. moubata Fer2 (OMFer2) (GenBank accession number MZ332506), which were obtained in previous studies (BioProjects PRJNA401392 and PRJNA377416) (Oleaga et al., 2017b(Oleaga et al., , 2018. The primers were designed to amplify, clone, and express the mature forms of the proteins, i.e. without signal peptide (tOEFer2 and tOMFer2). To facilitate cDNA subcloning in the pQE-30 expression vector, the primers included a site for the KpnI restriction enzyme. The primer sequences and the PCR conditions are shown in Supplementary Table 1. The PCR products were electrophoresed in 1% agarose gels stained with GelRed (Biotium), then cut from the gels and purified using the StrataPrep DNA Gel Extraction Kit (Agilent Technologies). The concentration of the purified amplicons was estimated by spectrophotometry at 260 nm (NanoDrop 2000, Thermo Scientific).
The Fer2 cDNAs were then cloned into the sequencing vector pSC-A using the StrataClone PCR Cloning kit (Stratagene) following the manufacturer's instructions. The resulting recombinant plasmids (tOEFer2-pSCA and tOMFer2-pSCA) were transformed into Escherichia coli Solo-Pack cells, and the cells were seeded on agar plates containing 100 µg/ ml ampicillin, 25 μg/ml kanamycin, and 80 µl/plate of 2% X-gal, and incubated overnight at 37 • C. Several recombinant colonies were selected from each transformation, inoculated into 5 ml cultures of Luria-Bertani (LB) medium with ampicillin and kanamycin at the same concentrations as above, and incubated overnight at 37 • C. After that, the cells were lysed, and plasmid DNA was purified using the Qiaprep Spin Miniprep (Qiagen) kit. The corresponding inserts in the plasmids were verified by digestion with EcoRI (Promega) for 2 h at 37 • C and electrophoresis of digestion products in 1% agarose gel. After that, inserts were sequenced in both strands using primers T3 and T7 and the sequences were compared to the published O. erraticus and O. moubata Fer2 sequences (GFWV01008002, MZ332506) using the Chromas 2.6.2 and Multalin (http://multalin.toulouse.inra.fr/multalin/) (Corpet, 1988) tools. At least three clones of each recombinant were sequenced to verify the correctness of the sequences.
Next, BLASTp searches for Fer2 homologous sequences in the NCBInr database were performed. The retrieved sequences showing E value lower than 10 − 70 were selected, aligned to the O. erraticus and O. moubata Fer2 proteins using ClustalW 2.1 (http://www.ebi.ac.uk/ Tools/msa/clustalw2/), and their phylogenetic relation analysed using the Mega 5.05 package (Tamura et al., 2011).

O. erraticus and O. moubata ferritin 2 gene knockdown by RNAi and phenotypic analysis
Gene knockdown was induced by injecting the ticks with doublestranded RNA (dsRNA) fragments specific to the Fer2 gene of each Ornithodoros species. The degree of gene silencing was assessed by measuring the level of Fer2 mRNA by real-time RT-PCR, and the phenotypic effect was assessed by feeding the dsRNA-treated O. erraticus and O. moubata ticks on rabbits.

Synthesis of dsRNAs
For dsRNA synthesis, primers T7Fer2_OE_Fw and T7Fer2_OE_Rv, amplifying a 272 nucleotide cDNA fragment of the OEFer2 gene, and T7Fer2_OM_Fw and T7Fer2_OM_Rv, amplifying a 258 nucleotide cDNA fragment of the OMFer2 gene, were designed. Additionally, primers T7Luc_Fw and T7Luc_Rv, which amplified a 323 nucleotide cDNA fragment of the bacterial luciferase gene (Luc), were also designed. Luc-dsRNA was used as a negative control (Boldbaatar et al., 2010). These primers contained the T7 promoter sequence to facilitate subsequent dsRNA synthesis (Supplementary Table 1). Amplification of the Fer2 target fragment of each species was performed from the Fer2 cDNAs cloned in pSC-A (tOEFer2-pSCA and tOMFer2-pSCA), and amplification of the luciferase target fragment was performed using a commercial plasmid as the template (pGEM-Luc Vector, Promega). PCR conditions are shown in Supplementary Table 1. PCR products were purified from 1% agarose gels using the QIAquick gel extraction kit (Qiagen) followed by precipitation in sodium acetate and, sequencing of both strands to verify they had been correctly amplified. dsRNA synthesis was carried out using the Megascript RNAi kit (Promega) following the manufacturer's protocol. The dsRNAs were quantified by spectrophotometry at 260 nm, their integrity was checked in 1% agarose gels ( Supplementary Fig. 1), and preserved at − 80 • C until use. dsRNA concentrations were as follows: dsRNA_OMFer2 = 1.3 × 10 12 molecules/µl; dsRNA_OEFer2 = 9.1 × 10 11 molecules/µl; dsRNA_Luc = 5.9 × 10 12 molecules/µl.

dsRNA injection into ticks
Before injection, tick specimens were sequentially washed by immersion and shaking in the following series of solutions as indicated by Kocan et al. (2011): tap water, 3% hydrogen peroxide, two washes in distilled water, 70% ethanol, and two more washes in distilled water. Ticks were then dried on paper towels and immobilised with adhesive tape on a glass plate. Each tick was injected with 1 µl of the corresponding dsRNA in the lower right quadrant of the ventral side using a 33-gauge Hamilton syringe.
Each dsRNA (dsRNA_OMFer2, dsRNA_OEFer2, dsRNA_Luc) was injected into 35 newly moulted 4-month-old females of each species. In parallel, as an injection control, another group of 35 females were injected with 1 µl of TRIS-HCl buffer. After injection, ticks were kept at 28 • C and 85% relative humidity.

Real-time RT-PCR analysis to confirm gene silencing
The degree of gene silencing was individually assessed in five ticks from each group, five days after injection. Each tick was dissected, and the total RNA was purified from the whole internal organs using the RNeasy Mini Kit (Qiagen). Purified RNA was then treated with Turbo DNA-free Kit (Ambion) and used as a template for cDNA synthesis with the 1st Strand cDNA Synthesis kit (Roche) and the oligo(dT)15 primer.
In these cDNA samples, the OEFer2 and OMFer2 mRNA levels were quantified and normalised against the mRNA actin level, which was used as a reference gene. With this aim, specific primers were designed from OEFer2 (qOEFer2_Fw, qOEFer2_Rv), OMFer2 (qOMFer2_Fw, qOMFer2_Rv), O. erraticus actin (GFWV01010289) (qOEActin_Fw, qOEActin_Rv), and O. moubata actin (AB208021) (qOMActin_Fw and qOMActin_Rv) using the Primer3Plus software (Untergasser et al., 2012) (Supplementary Table 1). Real-time PCR was performed using the SYBR Premix Ex Taq kit (Takara) and the 7900HT Fast Real-Time System (Applied Biosystems). The reaction mixture was recommended by the kit manufacturer, and the PCR conditions are shown in Supplementary  Table 1.
The degree of silencing was calculated using the Ct method (Cikos et al., 2007) by comparing the Fer2 gene expression level in OEFer2 and OMFer2 dsRNA-treated ticks to the Fer2 gene expression level in the luciferase dsRNA-treated samples.

Analysis of the Fer2 gene silencing phenotype
The effect of gene knockdown on feeding, reproduction, and survival of O. erraticus and O. moubata females was assessed in 20 ticks from each group. Ticks were fed on rabbits five days after dsRNA injection. The amount of blood ingested, survival rate, oviposition rate, and the number of fertile eggs laid by each female were measured. The mean and standard deviation of each parameter were calculated for each group. The differences between ticks treated with dsRNA_OMFer2 or dsRNA_OEFer2 and ticks treated with dsRNA_Luc and TRIS-HCl buffer were assessed using one-way ANOVA followed by Dunnet's t-test. Values of p ≤ 0.05 were considered significant.

Recombinant expression and purification of ferritin 2 of O. moubata
As will be described and analysed in the results section, the silencing of the O. moubata Fer2 gene significantly affected egg viability in this species. Accordingly, we tested the protective potential of the recombinant mature form of O. moubata Fer2 (tOMFer2) in a preliminary rabbit vaccination trial.
The recombinant protein was produced as follows. The truncated cDNA sequence, tOMFer2, was subcloned into the expression vector pQE-30 (Qiagen) and expressed in Escherichia coli M15 cells (Qiagen). For this, the recombinant plasmid tOMFer2-pSCA was digested with KpnI restriction endonuclease (Promega). After digestion, the insert was purified from an agarose gel and ligated to the KpnI predigested pQE-30 vector using the enzyme T4 DNA ligase (Thermo Scientific). This construct was used to transform E. coli M15 cells. Single recombinant clones were selected, and plasmid DNA was extracted and sequenced to confirm the presence and the correct orientation of the tOMFer2 cDNA insert. The selected clone was inoculated into LB medium with 100 μg/ ml of ampicillin and 25 μg/ml kanamycin and incubated at 37 • C. Protein expression was induced by adding 1 mM Isopropyl β-D-1-thiogalactopyranoside (IPTG) (Fisher Scientific).
The induced cells were collected and lysed by sonication in lysis buffer containing 50 mM NaH 2 PO 4 , 300 mM NaCl, 10 mM imidazole, pH 7.9. The protein was expressed in a 100% insoluble form and, consequently, it was solubilised with 8 M urea, purified by nickel affinity chromatography in denaturing conditions, and dialysed against PBS pH 7.4, for 24 h at 4 • C according to the procedure described by Díaz-Martín et al. (2011).
The concentration of the purified protein was assessed by densitometry in Coomassie blue-stained SDS-PAGE gels, followed by interpolation into a bovine serum albumin (BSA) standard curve. Purified protein was stored at − 20 • C.

Animal vaccination assay with tOMFer2
The immunogenicity and protective efficacy of tOMFer2 were evaluated in rabbits as follows.

Rabbit immunisation
Two groups of three rabbits per group were used, one group as control and the other immunised with recombinant tOmFer2. In the vaccinated group, each animal received three subcutaneous doses of tOMFer2 administered fortnightly. Each dose consisted of 100 µg of tOMFer2 in 1 ml of PBS emulsified with an equal volume of adjuvant (Montanide ISA 61 VG). Control rabbits were administered an emulsion of PBS and adjuvant.
Rabbits were bled immediately before administration of the first antigen dose (preimmune sera), 14 days post-immunisation (14 d.p.i.), immediately before tick infestation, and 28 days post-immunisation (28 d.p.i.), at 14 days after the infestation. Blood samples were allowed to clot, and sera were removed and stored at − 80 • C.

Measurement of serum antibody levels
The antibody titre of the immune sera to the tOMFer2 was assessed by serial dilution ELISA following standard procedures (Pérez-Sánchez et al., 2019). Briefly, polystyrene plates (Sigma) were coated with 100 ng of tOMFer2 recombinant per well, the sera were diluted in PBS containing 0.05% Tween 20 in a two-fold dilution series starting at 1/100, and peroxidase-conjugated anti-rabbit IgG (Sigma) was used diluted 1/10,000. Ortho-phenylene-diamine (OPD) was used as a chromogen, and the reactions were stopped with 3 N sulphuric acid. The serum titre was defined as the highest dilution giving more than twice the reactivity of the corresponding preimmune serum at the same dilution.
After titration, the reactivity of the immune sera to the four midgut protein extracts (soluble and membrane proteins from fed and unfed females) from each species (O. moubata and O. erraticus) was tested by ELISA following the protocol described above (sera 1/100 diluted; antirabbit IgG 1/10,000 diluted), and by western blotting following standard procedures and using a chemiluminescent substrate (de la Torre-Escudero et al., 2013).

Tick infestation and evaluation of vaccination efficacy
Two weeks after the third tOMFer2 dose, batches of 15 females, 30 males, and 50 nymphs-3 of O. moubata and similar batches of O. erraticus were allowed to feed on every rabbit. After ticks fed to repletion, they were weighed and then monitored for survival rate, egg laying, and subsequent hatching to larvae for O. erraticus or nymph-1 for O. moubata and nymph-3 moulting rate.
The vaccine efficacy (E) of tOMFer2 was calculated according to the formula established by Contreras and de la Fuente (2016), which is based on the reduction in the studied developmental processes in ticks fed on vaccinated animals, compared to ticks fed on controls. Here, vaccine efficacy was calculated as E = 100 x [1-(S x F x N)], where S and F represent the reduction in the survival and fertility (newly hatched larvae/nymphs-1 per female) of female ticks, respectively, and N represents the reduction in survival of nymphs-3 (Pérez-Sánchez et al., 2019).

Statistical analysis
The values obtained for the parameters analysed in the vaccine trial and the RNAi gene knockdown experiment are summarised as the mean ± standard deviation per group. The overall differences between the vaccinated and control group in the vaccine trial and between the groups treated with dsRNA and the controls in the RNAi gene knockdown experiment were compared by a one-way ANOVA to assess the level of significance of these differences. In the RNAi experiment, a posthoc test (Dunnett's T-test) was also applied. Differences were considered significant for a probability of error ≥ 95% (p ≤ 0.05). All statistical analysis was performed using SPSSv27 (IBM).

Sequences of ferritin 2 from O. erraticus and O. moubata
Previous studies on midgut transcriptomes of O. moubata and O. erraticus (BioProjects: PRJNA377416, PRJNA401392) (Oleaga et al., 2017b(Oleaga et al., , 2018 provided us with several transcripts whose sequences were annotated as Fer2. This gave us knowledge of their cDNAs sequences to design primers and procedures for their cloning and expression, and to perform in silico analysis of their predicted amino acid sequences. The nucleotide sequences of OMFer2 and OEFer2 were deposited into Genbank under the accession numbers MZ332506 and GFWV01008002, respectively. The open reading frames (ORF) of OMFer2 and OEFer2 comprise 594 bp, which encoded two proteins 194 amino-acid-long. Signal peptides were present in both of them with the predicted cleavage sites between positions 19-20 and 17-18 for OMFer2 and OEFer2, respectively. Both sequences lacked transmembrane domains and GPI-anchors and possessed two O-glycosylation sites at residues 101 and 108. Additionally, OEFer2 possesses an N-glycosylation site at residue 19 (Fig. 1).
Analysis of putative conserved domains showed that both ferritins contain a ferritin-like diiron domain profile between residues 29 and 179 and seven conserved amino acid residues in the ferroxidase centre loop in positions 46, 53, 80, 81, 84, 126, and 161, which are important for metal binding (Galay et al., 2013). The calculated molecular mass and pI for Fer2 without the signal peptide are 20.6 kDa and 5.25, respectively.
The amino acid sequences of OMFer2 and OEFer2 were aligned with the Fer2 orthologues from nine ixodid species, and their phylogenetic relation was analysed by the neighbour-joining method. The 11 aligned tick Fer2 amino acid sequences showed high levels of residue conservation and primary structure similarity (Fig. 2). The OMFer2 and OEFer2 amino acid sequences showed high homology to each other, reaching up to 85.3% of sequence identity, while their sequence identity  to the nine ixodid Fer2 ranged between 57.1% and 67.8%. OMFer2 showed the highest identity percentages (63.1-66.3%) to the Fer2 of the three Ixodes sp. included, I. ricinus (ACJ70653), I. persulcatus (AGX01000), and I. scapularis (EEC19111). Similarly, OEFer2 showed the highest identity percentages (65.1-67.8%) to the Fer2 of the same Ixodes sp. and to the Fer2 of Dermacentor andersoni (QLM00666) (65.7%) and Hyalomma anatolicum (ALJ92580) (65.4%). As shown in Fig. 2, the amino acid residues of the ferroxidase centre loop (highlighted in red) are conserved in all the eleven aligned sequences.
The phylogenetic relationship of the 11 Fer2 amino acid sequences is shown in Fig. 3, where three well-defined clades can be observed. One includes the Fer2 of the three Ixodes species (prostriata ticks). The second includes the Fer2 of the six species, and four genera of metastriata ticks (Rhipicephalus, Dermacentor, Hyalomma, and Haemaphysalis), and the third one comprises the Fer2 of the two argasid ticks.
The secondary and tertiary structures of ferritins were highly conserved. The Phyre analysis for OMFer2 and OEFer2 retrieved practically the same model, mainly made up of 4-helical bundles, based on the same template, the 'crystal structure analysis of e173a variant of the soybean2 ferritin sfer4 ′ (PDB code c3a9qR). For the OMFer2 sequence, 182 residues (92% of the sequence) were modelled with 100% confidence; for the OEFer2 sequence, 178 residues (90% of the sequence) were modelled with 100% confidence. Fig. 4 shows the 3-D structures of OMFer2 and OEFer2 and highlights the position of the seven conserved domains in the ferroxidase centre loop.

OMFer2 and OEFer2 gene knockdown and effect on feeding, survival, and reproduction of O. erraticus and O. moubata
To evaluate the importance of Fer2 on blood feeding and reproduction of ticks, gene silencing through RNAi was performed. Females of O. moubata and O. erraticus were individually injected with either 1.3 × 10 12 molecules of dsRNA_OMFer2 or 9.1 × 10 11 molecules of dsRNA_OEFer2 or with 5.9 × 10 12 molecules of dsRNA_Luc or with 1 µl of TRIS-HCl buffer in the control groups. The degree of silencing was determined by measuring the Fer2 mRNA levels by real-time RT-PCR and the effect of gene knockdown was assessed five days after injection by feeding the treated females on rabbits. significantly reduced the expression level of the OMFer2 gene (p < 0.001), reaching an average reduction of 97.1%. In this same species, treatment with dsRNA_OEFer2 had a less intense and variable effect, reaching an average reduction of 24.3% in OMFer2 gene expression level and individual reductions ranging from 0% in two ticks to 43.1-49.2% in three ticks ( Supplementary Fig. 2B). In O. erraticus females, both dsRNAs (dsRNA_OMFer2 and dsRNA_OEFer2) significantly (p < 0.001) reduced the expression level of the OEFer2 gene, reaching average reduction percentages of 65.1% and 90.2%, respectively (Fig. 5).
The effect of gene knockdown was assessed at five days postinjection in 20 specimens from each group feeding upon rabbits.
In O. moubata females, the OMFer2 gene knockdown did not reduce survival (not shown), the amount of ingested blood, or fecundity (number of laid eggs) ( Table 1). By contrast, it significantly affected egg viability. In females injected with dsRNA_OMFer2, the number of fertile eggs was significantly lower (p < 0.001) than in the control groups (injected with dsRNA_Luc or buffer), showing reductions in the number of newly emerged nymphs-1 of 70.5% compared to those injected with dsRNA_Luc, and 71.7% compared to the buffer-treated group. In females injected with dsRNA_OEFer2, a similar significant (p < 0.05), although less intense effect was observed, which consisted in a reduction in the number of live nymphs-1 of 29.3% and 32.4% compared to the groups treated with dsRNA_Luc and buffer, respectively (Table 1). In both groups of O. moubata females, numerous unhatched eggs were observed, some of them containing the formed larva inside, which, however, did not emerge as nymph-1 (Fig. 6).
The O. erraticus females injected with dsRNA_OEFer2 or dsRNA_OMFer2 did not show significant differences in mortality, amount of blood ingested, fecundity, or egg viability, compared to the O. erraticus females treated with dsRNA_Luc or injection buffer alone (Table 1).

Recombinant protein production
Recombinant tOMFer was expressed as an almost entirely insoluble protein. Once solubilised in urea buffer and purified on nickel affinity columns, it migrated in SDS-PAGE gels as a single band showing the expected molecular weight of around 23 kDa (including His-tag) (see Fig. 7A). The yield of tOM_Fer2 expression and purification was 11 mg of protein per litre of culture.

Vaccination assay
RNAi gene knockdown of OMFer2 had a significant negative effect on the viability of eggs laid by O. moubata females. This result prompted us to test the capacity of recombinant tOMFer2 to induce protective immune responses by conducting a preliminary vaccine trial using the recombinant protein as an antigen.
Rabbit immunisation with recombinant tOMFer2 induced strong humoral responses, reaching anti-tOMFer2 IgG antibody titres higher than 1/12,000 at 14 d.p.i. (not shown). Fig. 8 shows that reactivity of sera from the vaccinated rabbit peaked at 14 d.p.i. and remained high at 28 d.p.i. The sera from control rabbits did not react to the recombinant tOMFer2 (not shown). Fig. 8 also shows that the anti-tOMFer2 antibodies did not react in ELISA with any protein extracts from O. moubata and O. erraticus midguts, regardless of whether they were obtained from fed (48 h after feeding) or unfed females.
The reactivity of anti-tOMFer2 antibodies to the native Fer2 present in the midgut extracts of the two species was also analysed by western blotting. Fig. 7B shows that the anti-tOMFer2 serum recognised two very faint bands with a molecular weight compatible with Fer2 on the insoluble fractions (P-0) of midgut extracts from unfed females of the two species. However, in the midguts of fed ticks, this reactivity was somewhat more intense, with native Fer2 detected in the two fractions of O. moubata midgut (S-1, P-1) and the insoluble fraction of O. erraticus midgut (P-1).
The vaccine's protective effect was assessed by feeding on each rabbit, immunised and controls, 30 males, 15 females, and 50 nymphs-3 of O. moubata and equivalent batches of O. erraticus. The amount of blood ingested, oviposition, number of viable eggs laid by females, moulting rates of immature stages, and mortality rates of all individuals were assessed and compared between ticks fed in vaccinated and control rabbits. The protective effect of the immune response induced by the tOMFer2 is summarised in Table 2. In O. moubata, no significant differences were observed between the vaccinated and control groups, at any developmental stage, in the amount of blood ingested, mortality rate, and moulting rate of the nymphs. However, the number of eggs laid by females fed on vaccinated rabbits was 23.6% lower (p ≤ 0.05) than those laid by females fed on control rabbits. This reduction in oviposition was not accompanied by lower egg viability, as observed in the RNAi assay (Table 2).
In O. erraticus, the immune response induced by tOMFer2 had no  protective effect. No differences could be observed between ticks fed on control or vaccinated rabbits, in any developmental stages and in any of the analysed parameters (Table 2).

Discussion
During the digestion of their blood meal, ticks are exposed to huge amounts of free iron, which has beneficial and harmful effects. For this reason, iron homeostasis is tightly regulated by a set of proteins that manage iron absorption, use, transport, and storage (Galay et al., 2015).  In ixodid ticks ferritins Fer1 and Fer2, play a pivotal role in iron-storage and transport (Hajdusek et al., 2009;Galay et al., 2013Galay et al., , 2014aHajdusek et al., 2016). Ixodid Fer2 is a secreted protein responsible for iron transport into peripheral tissues, with an essential role in tick feeding and reproduction, which has been identified as a good candidate antigen to be included in anti-tick vaccines (Galay et al., 2014a;Tabor, 2021).
As mentioned in the introduction section, several transcripts annotated as Fer2 were found in the midgut transcriptomes of O. erraticus and O. moubata (BioProjects: PRJNA377416, PRJNA401392). Moreover, the expression level of these transcripts significantly increased at 48 h postfeeding in response to blood meal, suggesting that this protein could play essential functions in early blood digestion phases in Ornithodoros ticks, which highlights its potential as an antigen target for tick vaccine development (Oleaga et al., 2017b(Oleaga et al., , 2018. Accordingly, we decided to clone and characterise the Fer2 orthologues of O. erraticus and O. moubata, to perform functional studies by RNAi gene knockdown of Fer2, and to test the protective efficacy of recombinant Fer2 protein in an animal vaccination trial. Characterisation and analysis of the OMFer2 and OEFer2 amino acid sequences showed very similar primary and secondary structures to each other and high similarity to the Fer2 sequences of ixodid species, confirming that Fer2 is highly conserved between both tick families (Githaka et al., 2020). Similarly, the 3-D structure of Fer2 is also highly conserved in ticks and hematophagous insects (Pham and Winzerling, 2010). This high amino acid conservation in tick sequences suggest that Fer2 could be included in a broad-spectrum vaccine against multiple tick species (Parizi et al., 2012).
RNAi is a valuable tool for studying gene function in ticks and a recognised method for screening tick protective antigens (de la Fuente and Merino, 2013). The crucial role of Fer2 in successful blood feeding and reproduction of hard ticks was demonstrated by gene silencing through RNAi in I. ricinus (Hajdusek et al., 2009), H. longicornis (Galay et al., 2013), and H. antolicum (Manjunathachar et al., 2019). Fer2 gene knockdown increased mortality rates in these species and influenced fed body weight, reproduction, egg morphology, and egg hatchability rate.
In the current study, gene knockdown of Fer2 in O. moubata reduced the egg hatchability rate -and subsequently the number of emerging nymphs-1-up to 71%, without affecting the rest of the parameters studied, namely the mortality rate, the amount of blood ingested and the fertility. In contrast, gene knockdown of Fer2 in O. erraticus did not affect the treated ticks even though the expression mRNA level of OEFer2 was reduced by 90%. This lack of effect was unexpected, particularly when compared to that observed for O. moubata. The reason behind this different effect is unknown, but perhaps it could be related to particular anatomo-physiological differences between O. moubata and O. erraticus, as for example the lack of anatomical connection between the midgut and hindgut in O. moubata, which avoids the passage of haematin and other toxic digestion waste to the hindgut to be excreted (Sonenshine and Roe, 2014). It can also be hypothesized that OEFer2 and OMFer2 be subjected to different regulatory mechanism and protein Fig. 8. ELISA. IgG antibody response in rabbits vaccinated with recombinant antigen tOMFer2. Reactivity of antirecombinant tOMFer2 rabbit sera to soluble (S) and insoluble (P) protein extract of midgut. Values are the mean OD ± SD at 492 nm from vaccinated rabbit group. Sera were taken before immunization (preinmune), 14 days postimmunization (14 d.p.i.) and 28 days post-immunization (28 d.p.i.), and were used at 1/100 dilution. S-0 and P-0, supernatant and pellet from midgut homogenates of unfed ticks. S-1 and P-1, supernatant and pellet from midgut homogenates of fed ticks.  Hajdusek et al. (2016), it is possible that OEFer2 could be subjected to a slower turnover than OMFer2, or that some leftover mRNA in the cell after knockdown may maintain a basic functional level of the protein in tick. Currently, we lack data that support these hypotheses, so that additional studies will be necessary to confirm or rule out them. We do not believe this lack of effect to be due to the particular time post dsRNA treatment at which the analysis was made. There are several reasons for this: (i) the degree of silencing achieved at five days posttreatment when ticks were fed was over 90% in both species, and (ii) previous studies on O. erraticus and O. moubata subolesin showed that the highest degree of silencing was achieved five days after dsRNA injection and remained high for at least 12 days post-treatment (Manzano-Román et al., 2012). Similar studies for Fer2 would elucidate the degree of Fer2 gene knockdown at different times post-injection and the corresponding phenotypic effects.
According to our current results, in O. moubata, Fer2 would be involved in embryogenesis but not in other biological processes known to be affected by Fer2 gene knockdown in ixodid ticks (Hajdusek et al., 2010;Galay et al., 2013).
The significant reduction in the number of viable eggs caused by OMFer2 gene silencing prompted us to test the ability of recombinant tOMFer2 to induce protective immune responses. The vaccine efficacy of Fer2 has already been demonstrated in several ixodid species, albeit with variable effects, causing higher mortality rates, reduced feeding and lower oviposition and hatchability in ticks fed on vaccinated animals (Hajdusek et al., 2010;Galay et al., 2014b;Manjunathachar et al., 2019;Githaka et al., 2020). The effects elicited by the protective immune response against Fer2 in ixodids were, in most assays, similar to those induced by the RNAi silencing of the Fer2 gene (Galay et al., 2014a;Hajdusek et al., 2016;Manjunathachar et al., 2019).
After the RNAi assay, we evaluated the usefulness of OMFer2 as a vaccine target. We first demonstrated that tOMFer2 is highly immunogenic and induced strong humoral responses when administered to rabbits formulated with the Montanide adjuvant, in parallel to that observed for the recombinant Fer2 proteins from ixodids (Galay et al., 2014a;Manjunathachar et al., 2019).
The anti-tOMFer2 immune sera poorly reacted to soluble and membrane midgut proteins of both Ornithodoros species according to the results of ELISA and western blotting, in which most of the reactivity observed can be attributed to the non-specific recognition of host IgG and other proteins, which were also recognised by the preimmune sera. Despite this, the western blots demonstrated that the immune sera recognised several bands compatible with the OMFer2 and OEFer2 native proteins on the midgut extracts of both species. This reactivity was more intense for the midgut extracts obtained at 48 h post-feeding, in agreement with the upregulation of OMFer2 and OEFer2 mRNAs observed in the midgut transcriptomes of both species in response to blood feeding (Oleaga et al., 2017b(Oleaga et al., , 2018. The low reactivity of immune sera with the midgut extracts observed in the ELISA and western blots could be explained by the assumption that the amount of OMFer2 present in the midgut would be minimal because part of this protein is secreted into the haemolymph, where it is commonly detected (Hajdusek et al., 2009;Galay et al., 2013).
On the other hand, given that Fer2 is a soluble protein, it would be expected that its presence in the midgut would be limited to the soluble fraction (S-0, S-1) of the protein extracts. It is thus remarkable that in the western blots, Fer2 is also detected in the insoluble fractions (P-0, P-1). It may be possible that OMFer2 and OEFer2 are located inside granules within the digestive cells, as has been described in H. longicornis, and that, during the process of preparation of midgut protein extracts, centrifugation at 10 5 x g may have concentrated these granules in the insoluble fraction (Galay et al., 2013).
The protective effect of the anti-tOMFer2 response on the O. moubata and O. erraticus ticks was quite limited. In O. erraticus, we did not observe any protective effect. This result was unexpected because of the high amino acid sequence identity between OMFer2 and OEFer2. Some cross-species protection would be expected, as has been observed among species of ixodids (Galay et al., 2014a;Githaka et al., 2020), and that the anti-OMFer2 antibodies would also be useful to protect against O. erraticus infestations. As this was not the case, our current results suggest that the slight differences in the amino acid sequence between OMFer2 and OEFer2 could be the cause of the difference in protection. Similar observations have been made with tick vaccines based on the Bm86 antigen, in which slight differences in the sequence of the Bm86 antigen can cause significant differences in the vaccine efficacy among different strains of the same R. microplus species .
In O. moubata, the protective effect mainly resulted in a significant reduction of oviposition without affecting the other analysed parameters. This effect was not observed in the RNAi assay, in which the main effect was the reduction of egg viability. This discrepancy is not surprising because the mechanism by which RNAi and host antibodies exert their blockade effect against a particular molecule is different (Galay et al., 2014a). In ixodids, vaccination with recombinant Fer2 also reduced the oviposition rates. This reduced egg production was related to decreased tick engorgement weight and impaired ovaries' development (Hajdusek et al., 2010;Githaka et al., 2020). It is known that vitellogenesis is triggered when the tick reaches a critical engorgement weight during a blood meal (Weiss and Reuben Kaufman, 2001). However, in O. moubata, the reduced oviposition produced by the vaccine was not related to reduced blood intake. In H. longicornis, the antibodies induced against HIFer2 were detected in ovaries and eggs, suggesting that antibodies interfered with the HIFer2 function in this location and disrupted tick reproduction (Galay et al., 2014a(Galay et al., , 2018. It cannot be ruled out that a similar phenomenon may occur in O. moubata. However, similar to those carried out in ixodids, further studies will be needed to elucidate these aspects in argasids. The low effects of our vaccination assay could be attributed to antibody activity. It is assumed that a high level of host antibodies against Fer2 was necessary for the blockade of Fer2 function (Galay et al., 2014a) and that the IgG in haemolymph disappears quickly after engorgement, possibly by degradation and/or absorption (adhesion to tissues) (Chinzei and Minoura, 1987). This would prevent a percentage of host antibodies from reaching its target tissue, reducing the vaccine's protective effect by blocking OMFer2 function (Galay et al., 2014a). Accordingly, future studies should aim to induce more potent and longer-lasting humoral responses.

Conclusions
Fer2 is a protein that plays an important role in iron homeostasis. It is essential for tick survival and reproduction in ixodids, making it an interesting antigen candidate for tick vaccine development. The current study is the first one carried out on Fer2 from argasid ticks, for which previous molecular and functional information was almost nonexistent. Here we confirm the nucleotide and amino acid sequences of Fer2 for two argasid species, O. moubata and O. erraticus. These two argasid Fer2 showed high sequence identity to each other and high similarity to the Fer2 from ixodids. The high sequence identity of Fer2 between tick families suggests a similar function in the physiology of both argasid and ixodid ticks. The results of the functional assay and vaccination trial carried out in this study supported this idea because both reduced female reproduction, just as observed in ixodids. While RNAi gene silencing of OMFer2 strongly inhibited egg viability, vaccination with the recombinant tOMFer2 reduced oviposition without affecting egg viability. Unexpectedly, gene silencing and vaccination affected both Ornithodoros species differentially, suggesting anatomo-physiological differences between them that require further investigation. The current work paves the way for further studies aimed at unveiling the functional aspects of Fer2 in soft ticks and confirming its vaccine value through assays that enhance protective immune responses.

Animal welfare statement
The procedures for tick feeding and experimentation with rabbits were approved for the Ethical and Animal Welfare Committee of the Institute of Natural Resources and Agrobiology (IRNASA) and the Ethical Committee of the Spanish National Research Council (CSIC, Spain) (Permit Number 742/2017).

Funding
This research was funded by project "RTI2018-098297-B-I00 ′′ (MCIU/AEI/FEDER, UE), granted by the Spanish Ministry of Science, Innovation and Universities, the State Research Agency (AEI) and the European Regional Development Fund (ERDF), and project "CLU-2019-05-IRNASA/CSIC Unit of Excellence", granted by the Junta de Castilla y León and co-financed by the European Union (ERDF "Europe drives our growth").

CRediT authorship contribution statement
AO and RPS conceived and designed the study, interpreted the data and drafted the manuscript; AO and RPS performed the RNAi and vaccination assays; SGP performed the real-time RT-PCR analysis and provided critical review and revisions. All authors read and approved the final manuscript.

Declaration of Competing Interest
The authors report no declarations of interest.