Proteomic Analysis of Cattle Tick Rhipicephalus (Boophilus) microplus Saliva: A Comparison between Partially and Fully Engorged Females

The cattle tick Rhipicephalus (Boophilus) microplus is one of the most harmful parasites affecting bovines. Similarly to other hematophagous ectoparasites, R. microplus saliva contains a collection of bioactive compounds that inhibit host defenses against tick feeding activity. Thus, the study of tick salivary components offers opportunities for the development of immunological based tick control methods and medicinal applications. So far, only a few proteins have been identified in cattle tick saliva. The aim of this work was to identify proteins present in R. microplus female tick saliva at different feeding stages. Proteomic analysis of R. microplus saliva allowed identifying peptides corresponding to 187 and 68 tick and bovine proteins, respectively. Our data confirm that (i) R. microplus saliva is complex, and (ii) that there are remarkable differences in saliva composition between partially engorged and fully engorged female ticks. R. microplus saliva is rich mainly in (i) hemelipoproteins and other transporter proteins, (ii) secreted cross-tick species conserved proteins, (iii) lipocalins, (iv) peptidase inhibitors, (v) antimicrobial peptides, (vii) glycine-rich proteins, (viii) housekeeping proteins and (ix) host proteins. This investigation represents the first proteomic study about R. microplus saliva, and reports the most comprehensive Ixodidae tick saliva proteome published to date. Our results improve the understanding of tick salivary modulators of host defense to tick feeding, and provide novel information on the tick-host relationship.


Introduction
The cattle tick Rhipicephalus (Boophilus) microplus is a one-host tick that feeds on bovines. It is considered one of the most harmful cattle parasites in sub-tropical areas of the world due to its economic importance [1]. The economic losses associated with R. microplus parasitism are (i) direct, i.e., blood loss and lesions that predispose animals to myiasis and anaemia, reducing weight gain and milk production, and (ii) indirect, via the transmission of tickborne pathogens such as Babesia spp. and Anaplasma marginale [2,3].
There is evidence that tick salivary protein profiles change during tick feeding [21][22][23]. However, it is unclear whether the compounds secreted through R. microplus saliva vary throughout tick lifecycle. The identification of tick bioactive salivary components may be a potentially useful tool to more fully understand tick modulation of host physiological system. Moreover, this information may become valuable in the potential identification of novel target antigens for the development of anti-R. microplus vaccines and of potential lead compounds for pharmacological applications [24,25]. The aim of this work was to identify proteins secreted in saliva of R. microplus female ticks at two different feeding stages, and to gain insight into the putative role(s) these proteins play in regulating the tick-host relationship. For this purpose, we performed a proteomic characterization of saliva from partially engorged and fully engorged R. microplus tick females.

Ethics statement
All animals used in these experiments were housed in Faculdade de Veterinária, Universidade Federal do Rio Grande do Sul (UFRGS). This study was conducted considering ethic and methodological aspects in agreement with the International and National Directives and Norms by the Animal Experimentation Ethics Committee of the Universidade Federal do Rio Grande do Sul (UFRGS). The protocol was approved by the Comissão de É tica no Uso de Animais (CEUA) -UFRGS.

Ticks
R. microplus ticks, Porto Alegre strain, free of pathogens such as Babesia spp. and Anaplasma spp. were obtained from a laboratory colony maintained as previously described [26]. Ticks used in this study were exclusively fed on Hereford calves (Bos taurus taurus) acquired from a tick-free area. The calves were infested with 10day-old R. microplus larvae.

Saliva collection
Fully engorged female (FEF) ticks were obtained after the spontaneous detachment from the calves. Partially engorged female (PEF) ticks were carefully detached from the calves' skin by hand, between the 17 th and 20 th days post-infestation. Mean length of PEF and FEF ticks was 4.5 mm (ranging from 4 to 5 mm) and 11 mm (ranging from 9 to 12.5 mm), respectively. Before saliva collection, any host contaminating tissue in tick mouthparts was removed using a scalpel blade and surgical forceps. PEF and FEF ticks were rinsed with sterile distilled water and induced to salivate by dorsal injection of 2 or 5 mL pilocarpine (2% in PBS), respectively [27,28]. The saliva accumulated in the mouthparts was periodically collected using a pipette tip from ticks maintained at 37uC in a humid chamber for approximately 3 h. The saliva was stored at 280uC upon use. Saliva protein concentration was determined according to the bicinchoninic acid method (BCA Protein Assay, Pierce, Rockford, USA), as previously described [29].

In solution digestion, liquid chromatography and tandem mass spectrometry (LC-MS/MS) analysis
Three micrograms of protein from PEF and FEF tick saliva were reduced (10 mM DTT), alkylated (50 mM iodoacetamide) and digested with 1 mg modified trypsin (Promega Co., Madison, WI, USA) overnight at room temperature. LC-MS/MS was performed using a Thermo Electron LTQFT hybrid linear ion trap-FTICR mass spectrometer. Samples were loaded into a capillary C18 column (75 mm67.5 cm) and injected into the mass spectrometer at approximately 500 nL/min. The gradient elution was 0-90% acetonitrile/0.1 M acetic acid over 2 h. Data was collected in a top 10 mode, meaning that one FT scan (100 K resolution) taken was followed by 10 MS/MS fragmentation spectra of the top intensity ions collected in the linear ion trap. After MS/MS fragmentation was performed on a particular parent ion, m/z was placed on an exclusion list to enable greater dynamic range and prevent repeated analysis of the same peptide.
Electrospray voltage was set to 2.5 kV, and capillary temperature was 210uC.

1D gel electrophoresis and LC-MS/MS (1D-LC-MS/MS)
Saliva samples (25 mg) of both PEF and FEF were electrophoresed in 12% SDS-PAGE and stained with Coomassie brilliant blue. Subsequently, stained gel band slices (42 to PEF and 15 to FEF) were excised and individually subjected to trypsin digestion, as previously described [34]. The resulting peptides were analyzed using an electrospray ionization (ESI) quadrupole time-of-flight (Q-TOF) MicroTM mass spectrometer (Waters, Milford, MA, USA) coupled to a capillary liquid chromatography system nanoACQUITY UPLC (Waters, Milford, MA, USA). The peptides were eluted from a reverse-phase C18 column toward the mass spectrometer. Charged peptide ions (+2 and +3) were automatically mass selected and dissociated in MS/MS experiments. MS/MS spectra were searched against the database described above (item 2.3) using the MASCOT software version 2.2 (Matrix Science, London, UK) with the following parameters: tryptic specificity, one missed cleavage and a mass measurement tolerance of 0.2 Da in the MS mode and 0.2 Da for MS/MS ions. The carbamidomethylation of cysteine was set as a fixed modification, and methionine oxidation was set as variable modifications. The Scaffold software version 4.0.5 (Proteome Software Inc., Portland, OR) was used to validate MS/MS based peptide and protein identifications. Peptide identifications were accepted if they exceeded specific database search engine thresholds. Mascot identifications required ion scores higher than the associated identity scores of 20 and 35 for doubly and triply charged peptides, respectively. Protein identifications were accepted if they contained at least 2 identified peptides. To be included in this analysis, all peptide sequences had to have 100% identity with assigned proteins.

Functional annotation and classification of proteins
For functional annotation of the proteins, BLAST tools were used to compare the protein sequences to the NCBI (http://www. ncbi.nlm.nih.gov/) and GeneOntology protein database [35]. The ScanProsite and Pfam servers were used to search for conserved protein domains [36,37]. Functional annotation of identified tick proteins was based on previously published tick sialomes with some modifications (immunoglobulin-binding proteins were added to this classification) [4].

Results and Discussion
Blood is the only form of nutrition taken by ticks, and large blood meals are required for their development and survival. Ticks are pool feeders that accomplish feeding by lacerating small blood vessels and sucking up the blood that flows to the wound, the socalled feeding site [4][5][6][7]. Within minutes of inserting the hypostome into host skin, ticks secrete an amorphous adhesive substance (cement) that anchors them onto host skin and secures attachment throughout the feeding period [38]. When completely attached to the wound site, most ticks slowly feed off the pooled blood at the feeding site for several days [39]. The tick feeding cycle includes (i) the preparatory feeding phase, when the tick attaches onto host skin and creates the feeding lesion; (ii) the slow feeding phase, when the tick swallows moderate amounts of blood, begins to transmit pathogens, and grows new tissue to prepare itself for (iii) the rapid feeding phase, when it feeds to repletion [38,39]. The tick feeding style triggers tissue repair and other defense responses, like hemostasis, inflammatory reactions, pain or itching, and immune rejection [4][5][6][7]. Like other blood-sucking parasites, R. microplus ticks have developed a complex and sophisticated collection of pharmacological bioactive proteins and lipids produced by salivary glands that counteract host defenses and allow successful parasitism [4,5]. During blood meal acquisition, salivary glands undergo remarkable growth and differentiation accompanied by significant increase in protein synthesis [21][22][23]. Ticks concentrate the blood meal by secreting excess water and ions back into the host through salivary secretion [40]. After detachment from the host, a signal triggers tick salivary gland degeneration [41,42]. R. microplus ticks attach to its host as unfed larvae, and then proceed to feed and molt through nymphal and immature adult stages in a period that stretches to 12 days. After mating, adult pre-engorged females (PEF) increase blood meal ingestion rapidly, and by the 21 st or 22 nd day these fully engorged females (FEF) complete feeding and detach [43,44]. Adult ticks used in this study were collected between days 17 and 22 after experimental infestation. Thus, data presented here represent part of the slow feeding phase and of the final rapid feeding phase. Consistent with reports that other tick species change salivary expression profiles during feeding [21][22][23], data in this study reveals remarkable, quantitative and qualitative differences in saliva content of R. microplus at different feeding stages, suggesting modulation of protein expression during these stages. The saliva collection procedure yielded approximately 0.1 mL per PEF tick, and on average 0.8 mL of saliva per FEF tick. Despite the low amount of saliva secreted by PEF ticks using the pilocarpine-induced method, their salivary secretion had a higher protein concentration (3.22 mg/mL), compared with those obtained from FEF ticks (1.75 mg/mL). This is in accordance with an increased expression of saliva proteins that are important in hematophagy, during slow feeding phase (PEF). Most of these proteins may have been turned off in FEF. This could also be explained by fast degeneration of salivary glands in FEF ticks immediately after detaching from the host [41,42]. In the same way, as the salivary gland is responsible for hydrodynamic equilibrium in ticks [45] it is supposed that it excretes more water in the rapid feeding phase (FEF) than in the slow feeding phase (PEF), so the volume of saliva is higher in FEF, however protein concentration is lower. The proteomic analysis of R. microplus saliva allowed identifying 187 and 68 proteins from tick and cattle, respectively. Sequences from tick identified proteins were deposited as Transcriptome Shotgun Assembly project at DDBJ/ EMBL/GenBank under the accessions GBBO00000000 and GBBR00000000. The versions described in this paper are the first version, GBBO01000000 and GBBR01000000, respectively Based on SDS-PAGE analysis summarized in Figure 1, PEF saliva has a wider variety of proteins than FEF, as revealed by the number of identified proteins (147 to PEF and 112 to FEF) as well as in number of spectral counts, which can represent a semiquantitative approach (Table 1, 2 and 3). These data represent an apparent difference between PEF and FEF saliva. Interestingly, we observed high amounts of host proteins, which are presented predominantly in FEF saliva ( Table 4). The tick proteins identified in this study were classified as (i) putative secreted proteins and (ii) putative housekeeping proteins, and were then divided into groups according to their molecular function (Tables 1, 2, 3 and Figure 2) consistent with previous published tick sialomes [4].

Hemelipoprotein and other transporter proteins
Hemelipoproteins are the most abundant proteins in PEF and FEF saliva, based on protein band intensity ( Figure 1) and spectral count ( Table 3). In SDS-PAGE, these proteins appeared as two predominant bands between 95 and 130 kDa ( Figure 1) consistent with a previous study that reported that the major hemelipoprotein present in R. microplus hemolymph (HeLp) consists of two subunits (92 and 103 kDa) [46,47]. Although HeLp has no fullsequence deposited in any protein database, peptides corresponding to N-terminal sequence of HeLp subunits match the sequences for hemelipoproteins identified in tick saliva here, corresponding to HeLp-A and HeLp-B subunits [46]. HeLp has the ability to bind eight heme molecules, the prosthetic group released from hemoglobin digestion, and deliver them to tick tissues [46]. As a predominant protein in hemolymph, the presence of HeLp in R. microplus saliva could be explained by the phenomenon of hemolymph components incorporation by salivary glands, leading to secretion in saliva [48]. However, in other tick species, the transcriptional profile and protein localization of these hemelipoproteins in salivary glands of adult and unfed ticks suggest that they could act in different pathways during blood-feeding [18,49,50]. Previous studies have described these proteins in saliva from other ticks, which indicates that they are a conserved feature among different tick species [17,18,20], suggesting that HeLp may play vital role(s) in tick feeding and survival.
Since this protein could transport other compounds such as cholesterol, phospholipids and free fatty acids, in addition to heme [47], it is possible that they are secreted in the feeding site carrying small pharmacologic active molecules. It may also be postulated that hemelipoproteins perform non-classical yet unknown functions at the tick-feeding site. Recently, the main hemelipoprotein form in Dermacentor marginatum was shown to be a carbohydratebinding protein with galactose-and mannose-biding specificity able to agglutinate red blood cells [51]. In addition, as ticks use the pool-feeding strategy to feed [39], hemolysis at the feeding site is plausible due to the presence of digestive peptidases in saliva (Table 1 and 2). It is known that both heme and the heme-binding protein hemopexin have pro-inflammatory and anti-inflammatory properties, respectively [52][53][54]. Thus, the presence of hemelipoproteins could lower free heme concentration at the feeding site, preventing inflammation.
It may be speculated that HeLp is also essential to heme storage and/or detoxification in ticks. An important adaptation that coevolved with blood feeding is heme sequestration by heme-binding proteins and heme excretion, both of which prevent oxidative stress and tissue damage [55]. Interestingly, R. microplus ticks are unable to synthesize heme de novo [56], so hemelipoproteins could be critical components of a mechanism for sequestration, storage and utilization of host heme [46,49]. Due to their high concentration in tick saliva, it is possible that relatively high concentrations of hemelipoproteins are present at the feeding site. This may allow re-ingestion of these proteins along with blood. In this scenario, hemelipoproteins may act as heme transporter when hemoglobin digestion begins in the midgut, since the high content of heme in the cytosol of midgut cells suggests a heme transport pathway from the digestive vesicles through the cytosol to reach the midgut basal surface, where heme is transferred to hemolymph to be delivered to the ovary [57,58]. These molecules may be internalized in midgut cells by endocytosis, mediated by specific receptors, as described in mammal cells (e.g. heme-carrier protein hemopexin) [59]. This hypothesis is supported by the results of midgut proteome analysis of Dermacentor variabilis, where a hemelipoprotein was identified by LC-MS/MS, but not in the midgut cDNA library [60], suggesting that this protein is delivered from other tissue/secretion. Furthermore, D. marginatus major hemolymphatic hemelipoprotein was immuno-localized inside the midgut cells [51]. In the same way, hemelipoproteins may act in an excretory system to remove heme excess, obtained from blood ingestion, binding heme and re-injecting it into the host. This hypothesis of heme-binding agrees with the fact we detected a high amount of hemelipoproteins in PEF than in FEF saliva, and this reduction of hemelipoproteins in FEF saliva was accompanied by an increase in the host heme-binding proteins ( Figure 1, Figure 2, Table 3 and Table 4). These findings are compatible with a mechanism in which, towards the end of feeding, the tick replaces hemelipoprotein as heme-carrier by host derived heme-carrier proteins, including serum albumin, hemopexin, apolipoprotein and peroxiredoxin ( Figure 2 and Table 4). This may be possible at this stage because, after completing feeding, hemelipoproteins are necessary for vitellogenesis [61]. However, the presence of heme in tick saliva is yet to be demonstrated and needs further investigation. Similarly, ferritin is present only in PEF saliva (Table 1). Ferritin is an important iron reservoir, working as a protective mechanism against free iron overload. It is considered to be crucial for Ixodes ricinus development and reproduction [62,63]. Apparently, the absence of ferritin in FEF saliva is functionally compensated by serotransferrin, an iron-carrier protein from the host (Table 4). These observations strongly suggest the existence of a cooperative system between tick and host carrier-proteins, especially those involved in heme and/or iron regulation during blood-feeding. The role of these proteins in tickhost needs further investigation

Lipocalins
Lipocalins are single modular proteins of around 200 amino acids that fold tightly in a b-barrel with potential for binding small hydrophobic molecules in a central pocket. The tertiary structures of lipocalin are greatly conserved, even when amino acid sequence similarities are low [64,65]. In most organisms lipocalins are characterized by the consensus structural conserved regions (SCRs) that are characteristic of kernel lipocalins [66], while tick proteins assigned to the lipocalin family lack the typical SCR [67]. Annotation of the most recently identified tick lipocalins is based on homology with annotated histamine-binding proteins from other tick species, based on the presence of the characteristic tick histamine-binding domain (PF02098) as described in the Pfam database [37,[67][68][69]. PEF and FEF R. microplus secrete 50 different lipocalins in saliva (Table 1, Table S1 and Table S2. doi:10.1371/journal.pone.0094831.g001 lipocalins spectral counts are higher in FEF than in PEF (Table 1, 2 and 3). The presence of high amounts of lipocalins in cattle tick saliva is comparable with data from the O. moubata saliva proteome, showing that lipocalins are the most abundant salivary protein in this species [17]. Some of these R. microplus identified lipocalins have similarities with some described tick lipocalins, which have antihemostatic and immunomodulatory activities [68,69,[71][72][73][74][75][76][77][78][79][80], such as amine-binding molecules. The high content of lipocalins in tick saliva is compatible with their antihemostatic and immunomodulatory roles during tick parasitism [4][5][6][7]. Since histamine and serotonin secreted by the host at the feeding site induce cutaneous inflammation, ticks have to overcome their activities in order to complete feeding [4][5][6][7]. Sequestering these host molecules may be a mechanism used by R. microplus against these defensive reactions that affect thick attachment to hosts [81,82]. The high content of lipocalins in R. microplus saliva also could be related to level necessary to block the near micromolar concentration of biogenic amines and prostaglandins that accumulate at the feeding site [4]. The importance of this mechanism for tick feeding is underlined by the fact that R. microplus-resistant cattle have its status reverted to susceptible when treated with anti-histamines (H1 antagonists) [83]. Besides, a recent study that demonstrated that tick-resistant cattle sera have a higher IgG titer against lipocalins, compared to susceptible animals, stresses the importance of this class of proteins for blood-feeders [70]. The presence of a high concentration of lipocalins in FEF (Table 1, 2, 3 and Figure 2) is intriguing, because at this stage blood sucking is completed, and the tick does not need to modulate host defense mechanisms. It is possible that lipocalins found in FEF saliva signal the role(s) of these molecules during the    last stages of the rapid feeding phase, when the tick takes huge amounts of blood or prepares to detach from host skin.

Secreted conserved proteins
Transcriptomical analyses of salivary gland of hard and soft ticks have provided reliable data on blood-feeding behavior [4,5,[7][8][9][10][11][12][13][14][15][16]. The repertoire of tick salivary gland transcripts found is much broader and complex than anticipated, with many proteins without similarities to proteins in the NCBI database. Most of these new proteins were identified just as hypothetical secreted conserved proteins [4]. Proteins included in this group are the most abundant proteins in R. microplus saliva, and PEF saliva is richer in these proteins than FEF saliva (Table 1, 2, 3). The presence of these proteins in R. microplus saliva, as observed in the present study, confirms that some previously described hypothetical secreted conserved proteins are actually secreted proteins. Members of this type of proteins in R. microplus are 70-460 amino acid proteins (predicted molecular weight varying from 6.6 to 51.9 kDa) and some of them migrate as 34-60 kDa proteins when separated in SDS-PAGE ( Figure 1 and Table S1), suggesting that they have post-translational modifications. Given the higher number of these proteins present in tick saliva, it is reasonable to conclude that they have a role in tick feeding. The A. americanum AV422 protein (AamAV422) is a member of the secreted conserved protein group that is differentially up-regulated in response to contact with host and/or exposure to feeding stimuli [84,85]. This protein is secreted and injected in the host within the first 24 h of tick attachment onto the host. Apparently, AamAV422 is involved in the mediation of tick anti-hemostasis and anti-complement functions, since rAamAV422 delays plasma clotting time in a dose responsive manner, prevents platelet aggregation and reduces the formation of terminal complement complexes [84,85]. R. microplus secreted protein 20 is 99% identical to AamAV422, and is secreted in PEF and FEF saliva (Table 3). Like AamAV422, it may act as an anti-hemostatic and anti-complement protein [85]. Further studies are necessary to better characterize this group of salivary proteins, and may represent an opportunity to discover new targets for parasite control.

Peptidase inhibitors
The tick feeding style of lacerating host tissue and sucking host blood from the pool formed at the bite site is expected to strongly trigger host defense responses as hemostasis, inflammation, and complement systems [4,5,86]. These responses are dependent on the action of several peptidases, such as procoagulant (thrombin, factor Xa and other coagulation factors), pro-inflammatory (neutrophil elastase, proteinase-3, chymase, tryptase, kallikrein, cathepsin L, cathepsin B, cathespin S, cathepsin C and cathepsin G) and complement enzymes (factors B, C, D and component 2) [4,5,86,87]. These host defenses are highly regulated by specific endogenous inhibitors, maintaining homeostasis. From this perspective, it has been suggested that ticks secrete peptidase inhibitors to disrupt host defenses, facilitating feeding [88].
Serpins. proteins that belong to the serpin (serine protease inhibitor) superfamily are expressed in all branches of life [89]. They have a role in the control of several endopeptidase cascades in many organisms [90]. In mammalians, most serpins play crucial roles, controlling endopeptidases involved in blood coagulation, fibrinolysis, inflammation, and complement activation [89,91]. It is assumed that tick secreted serpins disrupt host homeostatic balance in order to facilitate parasitism [88]. Recently, 18 fulllength serpin encoding sequences were described in R. microplus [92], three of which (RmS-3, RmS-6 and RmS-17) were identified in PEF and FEF saliva (Table 3). Notably, PEF saliva has a high number of spectral counts of this protein family ( Table 3), suggesting that inhibition of serine endopeptidases involved in host defense system is important earlier in blood tick feeding. It was shown that tick-resistant cattle sera have high titers of antibodies against RmS-3, compared to tick-susceptible animals, suggesting its importance in the tick-host relationship [93]. Furthermore, the administration of an antibody against RmS-3 linear epitope by artificial feeding decreases the reproductive capacity of R. microplus females by 81% [93]. However, the precise role of these inhibitors in R. microplus saliva remains unclear. The presence of these serpins in R. microplus saliva could be responsible, at least partially, for the anti-thrombin [94] and anti-thrombotic [95] properties of its saliva, including their local and systemic alterations [26]. Moreover, some other pharmacological activities of R. microplus saliva may be associated to serpins, such as immunomodulatory activity [96][97][98][99]. The potential effect of these proteins on host systems are supported by several studies showing serpins from hematophagous parasites act as anti-coagulant and anti-inflammatory agents, being essential for a successful blood meal [96][97][98][99][100][101][102]. Clearly, data showing that the use of serpins as vaccinal antigens impairs tick development reinforces the importance of these proteins in regulating tick physiology [103][104][105][106][107].
a2-macroglobulin (a2M). these are large glycoproteins and are present in the body fluids of both invertebrates and vertebrates, being secreted as glycosylated polypeptides with a molecular mass of about 180 kDa [108]. Three a2M were identified in PEF and FEF saliva (Table 3), and based on spectral counts all three seem to be most abundant in PEF, relatively to FEF. In vertebrates, a2M proteins have been found to regulate host cell apoptosis [109], inhibit several serum peptidases like thrombin [110], factor Xa [111] and kallikreins [112], mediate Tcell proliferation [113] and induce proliferation and activation of Accession numbers for tick identified proteins were deposited as Transcriptome Shotgun Assembly project at DDBJ/EMBL/GenBank under the accessions GBBO00000000 and GBBR00000000. The versions described in this paper are the first version, GBBO01000000 and GBBR01000000, respectively. b Accession numbers of best matches identities obtained using BLASTP against the non-redundant protein database in GenBank. doi:10.1371/journal.pone.0094831.t002 macrophages [114]. Tick saliva a2M may be linked to interference in inflammation and immunomodulation, and it may be an additional salivary anti-coagulant. It is still unclear whether these a2M act as immunomodulators or as anticoagulants, this role needs to be elucidated. However, the fact that such inhibitors (as a2M proteins and serpins) are secreted mostly in PEF saliva (Table 3) reinforces the idea that inhibition of host-defenses endopeptidases is important as early as in the beginning of the blood meal. TIL domain-containing proteins. proteins belonging to the TIL (trypsin inhibitor-like) domain-containing group have been reported in blood-feeding mosquitoes and tick sialomes [5].  Ixodidin, an example of this group of inhibitors, was isolated from R. microplus hemolymph. In addition to antimicrobial activity, ixodidin has anti-trypsin and anti-elastase activities [115]. Only FEF saliva has peptides matching this group of proteins, including ixodidin (Table 2). These proteins may act similarly to host endopeptidases inhibitors, increasing the inhibition of the target endopeptidases. Additionally, presence of these proteins at the final phase of blood meal acquisition suggests that they have a possible role as an antimicrobial protein to prevent (or control) infection in ticks after blood-meal acquisition. Their interfering role in tick-vectoring ability, regulating the quantity or even the specificity of pathogens ticks transmit remains to be addressed. Thyropin. thyropin (thyroglobulin type-1 domain protease inhibitors) is a family of proteins characterized by the presence of thyroglobulin type-1 domain repeats [116,117]. The well characterized type-1 domain-containing protein was described in the sea anemone Actinia equina and has been shown to inhibit either cysteine or cation-dependent peptidases [118], including cathepsin L, cathepsin S, papain and cruzipain [117,119]. PEF and FEF saliva contains three thyropins (Table 1 and 2). It is possible that these proteins inhibit some host cysteine endopeptidases, contributing to the immunomodulatory effects of tick saliva. This hypothesis has yet to be proved, since thyropins have not been functionally characterized in ticks to date. Proteins containing these domains are present in several tick sialomes [4], and their presence was previously also detected in O. moubata and R. sanguineus saliva [17,19].
Cystatin. cystatins comprise a large family of reversible and tight-binding inhibitors of papain-like enzymes and legumains [120], which are involved in biological processes like antigen processing and presentation, phagocytosis, neutrophil chemotaxis during inflammation and apoptosis [121][122][123][124]. Two proteins of the cystatin family were identified in PEF and FEF, with higher spectral counts in PEF saliva (Table 1 and 3). The most abundant (RmCys2b -AGW80658.1) is a member of type 2 cystatin [125] and is present predominantly in PEF saliva (Table 3). It is able to inhibit cathepsin B, cathepsin L and cathepsin C (L. F. Parizi, personal communication). As these enzymes are important in some immunologic processes, these cystatins in R. microplus saliva could act as immunomodulators during the slow feeding phase of cattle tick parasitism, as previously shown for other tick cystatins, facilitating blood feeding and pathogen transmission [126][127][128][129][130]. The importance of these inhibitors in blood feeding was underscored in studies that showed that neutralization of cystatins (through gene silencing in ticks or vaccines) significantly reduces tick feeding ability [128,131,132].
Kunitz-type inhibitors. members of the Kunitz-type family are particularly well characterized as inhibitors of a large number of serine endopeptidases [133]. One protein containing Kunitz domains was found only in FEF saliva (Table 2). Interestingly, this protein contains nine in tandem Kunitz domains, a remarkable difference among well characterized inhibitors of this class in other ticks, which range between one and five domains [25,134]. These inhibitors have been characterized as acting upon thrombin, factor Xa, factor XIIa, trypsin and elastase [25]. This raises the suggestion they contribute to R. microplus saliva anticoagulant activity [26,94,95].

Glycine-rich proteins
This group of proteins is described in several tick sialomes and has distinct subdivisions [4]. In ticks, proteins containing glycinerich (Gly-rich) and proline-rich (Pro-rich) repeat motifs are associated with tick-cement functions [135,136]. Ten proteins of this superfamily were found exclusively secreted in PEF saliva (Table 1). These proteins have been identified also in O. moubata and R. sanguineus saliva [17,19]. The presence of these proteins at this stage lends strength to the hypothesis that they are important in the formation of a cement cone that affords tick attachment to the host during initial feeding phase. Three of these proteins contain the motif [LPAE]-P-G, that are known as targets of proline hydroxylase (data not shown) [137,138], a post-translational modification which allows cross-linking between proteins, a characteristic present in cement proteins [139]. The identification of these proteins at this developmental stage is in accordance with a previous study on A. americanum, where genes codifying for this superfamily of proteins are up regulated at the early stages of parasitism [84].

Enzymes
Peptidases. parasite secreted enzymes may play a wide array of roles in host tissues. Analysis of PEF tick saliva allowed the identification of two metallopeptidases (Table 1). In this sense, metallopeptidases, frequently associated with vascular damage, tissue remodeling and degradation of serum compounds [140] may have a role modulating host responses against ticks. As shown in other ticks, this salivary metallopeptidases may be linked to fibrin(ogen)lysis [141], bradykinin degradation [142], and angiogenesis inhibition [143]. In PEF saliva, a trypsin-like enzyme similar to factor-D from D. variabilis was identified (Table 1). This enzyme may interfere with host inflammation and blood clotting, acting as plasminogen activator or protein C activator, similarly to what has been reported for I. scapularis saliva [144]. The secretion of metallopeptidases and trypsin-like enzymes in tick saliva is stage-dependent, since the analysis performed here indicates that FEF saliva does not have significant amounts of these enzymes. The presence of these proteins in PEF saliva could also be explained by the fact that host defense modulation is crucial for blood feeding at this time.
In FEF tick saliva, only one endopeptidase was identified, the tick heme-binding aspartic peptidase (THAP) ( Table 2). Here, we report, for the first time, the presence of THAP in cattle tick saliva. THAP is able to hydrolyze hemoglobin and vittelin, and thus is supposed to have a role in R. microplus digestion and embryogenesis [145,146]. It may be hypothesized that THAP acts as a digestive enzyme secreted in the host during the fast engorgement phase. During blood meal acquisition, THAP may start the digestion process of blood components in the hemorrhagic pool at the tick attachment site. Similarly, this activity could explain the presence of a cathepsin-B in PEF saliva (Table 1), as this type of enzymes has been described to hydrolyze hemoglobin in other tick species [147,148]. In the same way, saliva of both PEF and FEF secretes a serine-carboxipeptidase (Table 3). Since a serine-carboxipeptidase from midgut was able to hydrolyze bovine hemoglobin in Haemaphysalis longicornis, it suggests that it also may be involved in digestion of the blood meal at feeding site [149]. In this way, the presence of these digestive enzymes in saliva may be associated with the presence of heme-binding proteins, since the free-heme delivered by hemoglobin digestion at the feeding site has to be sequestered, because heme has pro-inflammatory properties [52] and impairs blood meal acquisition.
Phospholipase A2. phospholipases A2 (PLA2) are secreted enzymes that have been implicated in several biological processes, such as modification of eicosanoid generation, inflammation and host defense [150,151]. Two PLA2 proteins were found in PEF saliva (Table 1). Secretory PLA2 are common and important components of bee and snake venoms, and have hemolytic, antiplatelet aggregation, and anticoagulant effects through their ability to interact with cells or by the degradation of phospholipid, thus generating free arachidonic acid [152]. Likewise, in A. americanum these proteins are suggested to act in the hemolytic activity of saliva [153,154]. The presence of PLA2 in PEF is in accordance with those digestive enzymes described above, which also may play a role in host blood cells lyses, facilitating the tick digestive process at feeding site. Additionally, these enzymes may act as antiplatelet and anticoagulant agents [152], facilitating blood feeding and reinforcing the notion that defense modulation in PEF is crucial for blood feeding.

Immunity-related proteins
Antimicrobial peptides. antimicrobial peptides (AMPs) are widely distributed in nature and are essential components of the first defense line against infections [155]. In invertebrates, which have only innate immunity, AMPs are extremely effective and work as powerful weapons against bacteria and fungi [156]. Microplusin is an AMP from R. microplus that belongs to the group of cysteine-rich AMPs with histidine-rich regions at N-and Ctermini, which have been implicated in sequestration of zinc, a microbial growth factor [157,158]. Proteins of the microplusin-like and histidine-rich families are present in the saliva of both PEF and FEF (Table 2 and Table 3). The role(s) of these proteins in tick saliva may be associated with the prevention of microbial proliferation at the tick-feeding site. Moreover, since a lot of saliva is ingested together with the diet, especially in pool feeders, it could be assumed that the AMP may also act in the midgut of ticks.

Putative housekeeping proteins
In R. microplus, we identified putative housekeeping proteins, predominantly in PEF saliva (Table 1 and 3). Putative housekeeping proteins in tick saliva have been identified in O. moubata and R. sanguineus [17,19]. The presence of this kind of protein in tick saliva is supported by observations showing apocrine and merocrine secretion in tick salivary glands [159]. Moreover, these housekeeping proteins can be secreted in non-classical pathways to the extracellular environment [160,161]. Presence of these proteins in tick saliva is underlined by the fact that hosts infested with A. americanum develop antibodies against housekeeping proteins during different tick feeding stages (A. Mulenga, personal communication).
The presence of housekeeping proteins in tick saliva may have further biological importance, since these proteins may play different roles in the tick-host interface. For example, since HSP70 is present in PEF saliva, it may be involved in tick-host relationship ( Table 1). In an experimental model of disease, HSP70 administration prevents inflammatory damage and promotes the production of anti-inflammatory cytokines [162]. Similarly, a study showed that HSP70 from Mycobacterium turbeculosis has antiinflammatory properties, inhibiting pro-inflammatory cytokine production by IL-10 driven down-regulation of transcriptional factor in dendritic cells [163]. Other examples of housekeeping protein involve enzymes linked to detoxification (Table 1). Glutathione S-transferase (GST) is a protein that catalyzes the conjugation of glutathione with several xenobiotic and endogenous substances [164]. In this sense, GST seems to be closely associated with detoxification and acaricide resistance [165]. Additionally, it has been proposed that GST secreted by parasite salivary glands has immunomodulatory activity due to the alteration of cytokine gene expression profile, modulation of immune cell proliferation and decrease in oxidative ability of phagocytes [166]. Further studies are necessary to elucidate the role of this class of proteins in tick saliva, since this appears to be a conserved feature among different tick species [17,19].

Host proteins
A large number of bovine proteins were identified in the saliva of both PEF and FEF, being present predominantly in FEF saliva, relatively to PEF saliva ( Table 4). The presence of host proteins in tick saliva has been reported in other ticks species [17][18][19][20]. These proteins are the majority secreted proteins in R. sanguineus saliva [19]. It was demonstrated that ticks transport intact proteins across the digestive system to the hemolymph [167]. Furthermore, some of the host proteins described in R. microplus proteome have been found in salivary glands of other tick species [12,18,20,48], suggesting that the presence of host proteins in tick saliva may be a real and common recycling system present in ticks, not a result of contamination during saliva collection. Furthermore, the presence of different classes of host proteins in the saliva of the two tick developmental stages suggested the existence of this selective uptake process (Table 4 and Figure 2). For example, in PEF saliva we observed a predominance of housekeeping proteins (actin, nuclear proteins like histone and HSP90) and hemoglobin subunits peptides (Table 4 and Figure 2). In FEF saliva this pattern switches dramatically due to: (i) transporter and/or proteins associated with metabolism of heme and iron, like serum albumin, peroxiredoxin, serotransferrin, apolipoprotein and hemopexin; (ii) immunity, like immunoglobulins chains and C3 complement protein; (iii) peptidase inhibitors of the serpin superfamily; and (iv) other proteins (Table 4 and Figure 2). Similarly, rabbit proteins involved in heme and iron metabolism (as serum albumin, serotransferrin and hemopexin); immunity (C3 complement protein); and serpins were identified in R. sanguineus saliva [19]. However, as in R. sanguineus saliva was collected from 5-7 days partially fed adults ticks [19], it is not possible to compare these differences among different developmental stages, as found in R. microplus.
We are mindful of the possibility that tick saliva proteins in FEF may not represent exactly what occurs at the end of the blood feeding. However, it is remarkable that the majority of host proteins in FEF saliva have heme-binding and endopeptidase inhibitory functions similar to some of the tick proteins in PEF saliva ( Figure 2). A quite interesting question is: if these proteins are returned intact, can they exert their biological function in the host? For instance, mammalian serpins were detected in FEF saliva (Table 4), so the question is: do these host serpins inhibit host serine endopeptidases of defense pathways as the tick prepares to detach? Whether these proteins are returned to the host as intact proteins or products of partial hydrolysis remains to be clarified. However, as in R. sanguineus saliva [19], it seems that host serum albumin is secreted intact into the host, since SDS-PAGE analysis reveals a ,60 kDa protein (Figure 1), which is intact [168]. Taken together with previous results that show the existence of a separate pathway for uptake and digestion of albumin in relation to hemoglobin incorporation into midgut cells [57], these results may be evidence of the existence of a system to recycle serum albumin. However, if serum albumin secreted into host is carrying some molecule along needs to be further clarified. In addition, it is important to note that several of these mammalian proteins, when undergoing limited proteolysis, generate peptides, some of which are bioactive, presenting antimicrobial action [169,170], as well as vasoactive peptides [171] which may enhance parasitism.
The presence of immunoglobulin chains in tick saliva could be explained as a part of the tick self-defense system, since immunoglobulin remains as an active protein in tick hemolymph [172]. In addition, the existence of immunoglobulin-binding proteins in both the tick salivary gland and hemolymph indicates that hemolymph and salivary gland cooperate to remove foreign proteins that could be deleterious for tick development during feeding [48]. An observation that support this hypothesis is that, in R. microplus, immunoglobulin-binding proteins from tick were found in the same developmental stage at which host immunoglobulin was found, in FEF saliva ( Table 2 and 4). Differently from R. microplus, saliva immunoglobulin was not identified in R. sanguineus [19]. In spite of that, as these proteins were identified only in FEF in R. microplus, the presence in FEF saliva of R. sanguineus cannot be ruled out.
Despite reports of the presence of host proteins in tick saliva, this remains a neglected issue in the study of tick biology. It is interesting to note that while long-term blood feeders like R. microplus and R. sanguineus saliva contains considerable amounts of host proteins, the saliva of the short-term blood feeder, such as O. moubata, contains only a few host proteins [17,19]. The demonstration of these proteins in tick saliva raises several questions to be further explored, and may reveal novel insights into tick-host relationship.

Conclusion
The advancements in transcriptomic and proteomic analyses in recent years have opened unprecedented opportunities to identify putative targets for tick control into the variety of tick salivary transcripts and proteins. Saliva of ticks are far more complex than anticipated, having hundreds of different tick proteins as well as a high content of host proteins, which could have a role in several pathways associated with tick survival. A complete identification of tick salivary compounds and their identification and characterization remains a major research challenge that will help understand how host modulation by ticks occurs. The proteomic approach allows a comprehensive analysis of saliva composition and provides novel information to guide further studies about molecular, biochemical, immune biological, pharmacological as well as physiological characterization of these proteins. In R. microplus it is technically challenging to study defined feeding time points, and this is the reason why all previous studies have utilized saliva of fully engorged ticks. It is conceivable that after detaching from the host (or most probably just before detaching) ticks stop secreting proteins, indeed, salivary gland degeneration starts at this point. So, all studies conducted with saliva or salivary glands from FEF ticks must be carefully interpreted. This study, comparing saliva from PEF and FEF ticks, helps identify tick proteins that are important in the tick feeding process. These data could contribute to the understanding of tick salivary gland physiology and the tickhost relationship as well clues to approach new immunologically based tick control.
To date, only a few reports have explored R. microplus saliva. Compared to other hematophagous parasites, there is relatively little information on the molecular composition of R. microplus saliva. This is the first comprehensive proteomic study on R. microplus saliva. It is important to note that ticks produce minute amounts of saliva, which makes it difficult to work with as biological material, and as such it is less well characterized than salivary glands. Although some proteins reported here have already been cloned from cDNA libraries of tick tissues, they were never purified from or identified in R. microplus saliva.
Despite the success of tick transcriptomic studies, which provide a global view of gene expression profiles in tick salivary glands, proteomic analysis of saliva provides unique information regarding proteins that are actually secreted. In conclusion, considering the great importance of this parasite, this study improves knowledge on the tick salivary arsenal composition and gives novel insights to clarify the mechanisms associated with the tick-host relationship.

Supporting Information
Table S1 Tick and host proteins identified in partially engorged female saliva by 1D-LC-MS/MS. (DOCX)