It is well known that ticks possess innate immunity that also affects their vector competence (Hajdusek et al., 2013). Although all of the involved mechanisms have not yet been fully clarified, microbe phagocytosis by tick hemocytes seems to be coupled to a primitive complement-like system, a variety of antimicrobial peptides and possibly reactive oxygen species (Kopacek et al., 2010). Studies in both hard and soft ticks have implicated several tSPIs in immune responses against different microbes, mostly identified in hemolymph (Figure 2, Table 2).

The most studied tick species in this context is R. (B.) microplus. Firstly, Ixodidin, a tSPI discovered in hemocytes, was reported to have strong inhibitory activity against Micrococcus luteus and, to a lesser extent against Escherichia coli (Fogaça et al., 2006). In addition to this bacterial clearance role, Ixodidin also possesses inhibitory activity against chymotrypsin and elastase serine proteases (Fogaça et al., 2006). However, it remains unclear whether the antimicrobial activity is due to protease inhibition or directly through peptide effects on bacterial membranes. Two additional inhibitors have been identified from R. (B.) microplus eggs: BmSI 6 and BmSI 7 (Boophilus microplus subtilisin inhibitors), that target Pr1 proteases from Metarhizium anisopliae, a fungus used as a biological insecticide (Sasaki et al., 2008). These Pr1 proteases induce host cuticle degradation, enabling hyphae penetration to obtain nutrition (Leger et al., 1987). Only BmSI 7—which shares several similarities with Ixodidin—has been well characterized. It harbors disulphide bonds and a trypsin inhibitor-like cysteine-rich domain (TIL), and as it is expressed in adult ovaries, midgut, SGs, hemocytes, body fat, and larvae, it could be involved in tick defense mechanisms, and/or in avoiding tick tissue degradation (Sasaki et al., 2008). Random sequencing of a tick body fat cDNA library enabled the discovery of, BmCI (Boophilus microplus chymotrypsin inhibitor), a chymotrypsin inhibitor that belongs to the Kunitz/BPTI inhibitor family (Lima et al., 2010). BmCI strongly and specifically inhibits chymotrypsin, and also hNE with reduced specificity. BmCI gene expression analysis demonstrated higher expression in hemocytes and ovaries than in SGs and body fat. Following infection with M. anisopliae, increased BmCI expression was only observed in tick hemocytes clearly suggesting a role in the tick defense system (Lima et al., 2010). Finally, among BmTIs (Boophilus microplus trypsin inhibitors) identified in both larvae and eggs (Tanaka et al., 1999), a double Kunitz-containing inhibitor, BmTI-A, had increased transcript levels in ovaries following Babesia bovis infection (normally transmitted transovarially in ticks), also suggesting a probable role in the tick immune system (Rachinsky et al., 2007).

IrAM (I. ricinus alpha macroglobuline), was identified in the hemolymph of I. ricinus as a α2M composed of two non-covalently linked subunits (Buresova et al., 2009). Alternative splicing occurring during IrAM synthesis generates seven bait variants, increasing the spectrum of targeted proteases. IrAM transcripts are detected throughout all tick developmental stages. Gene expression was higher in SGs from partially engorged females than in hemocytes or ovaries, even though IrAM protein is only detected in tick hemocytes. RNAi experiments revealed that IrAM is not involved in tick fitness, mortality, or fecundity (Buresova et al., 2009). However, IrAM enhanced phagocytosis and elimination of the bacteria Chryseobacterium indologenes due to its active thioester bonds, which was not observed with other bacteria such as B. burgdorferi (Buresová et al., 2006; Buresova et al., 2009). IrAM likely interacts with the major C. indologenes virulence factor, a metalloprotease, suggesting a role in the tick immune system during phagocytosis of metalloprotease-producing bacteria (Buresova et al., 2009).

Following the discovery of over-expressed genes in Dermacentor variabilis tick body fat and midgut in response to Rickettsia montanensis infection (Ceraul et al., 2007), several Kunitz/BPTI inhibitors were identified including DvKPI (Dermacentor variabilis Kunitz protease inhibitor with five Kunitz domains) (Ceraul et al., 2008). DvKPI was up-regulated six-fold in infected ticks that had fed for three days. These results suggested that DvKPI is involved in tick responses to Rickettsia, avoiding massive colonization which would be detrimental to ticks.

Several tSPIs have also been identified in Ornithodoros moubata soft tick hemolymph, including TAM, (tick α-macroglobulin), the second most abundant protein after vitellogenin (Kopacek et al., 2000). TAM is a tetrameric glycosylated protein that displays similar structural features present on IrAM, and exerts inhibitory activity against both trypsin and thermolysin (Kopacek et al., 2000). Comparing the conserved cysteine motifs between human and limulus α2Ms enables the prediction of disulfide bridge patterns which explain the atypical molecular arrangement of the four TAM bait region variants, likely arising from alternate splicing (Saravanan et al., 2003). While TAM was initially detected in tick hemocytes, significant up-regulation has also been reported in SGs and gut after 1 day of feeding (Saravanan et al., 2003). TAM is believed to be involved in tick defense systems, but an additional role as an anti-coagulant has also been postulated (Keller et al., 1993; van de Locht et al., 1996).

As for vertebrates, effective clotting is critical in ticks to limit hemolymph loss and to inhibit pathogens from entering into the tick through the wound. Little is known about the proteins involved in tick hemolymph clotting but several tSPIs have been implicated in this defense system due to homology with known proteins (Figure 2).

Four Rhipicephalus appendiculatus tick serpins have been identified as potential clotting enzymes involved in the hemolymph coagulation cascade: R. appendiculatus serpin-1, -2, -3, and -4 (RAS-1, -2, -3, and -4) (Mulenga et al., 2003b). All exhibit similarities ranging from 25 to 30% with limulus intracellular coagulation inhibitor type 1 (LICI-1) from the Japanese horseshoe crab Tachypleus tridendatus. RAS-3 harbors also amino acids similarities with LICI-2 and is comparably expressed in all tick organs, while RAS-1, -2 and -4 expression was stronger in SGs than in other tick organs (Mulenga et al., 2003b).

HLS2, a serpin from Haemaphysalis longicornis, contains an RCL with high sequence similarities to both vertebrate and invertebrate serpins, and may interact with both chymotrypsin and thrombin (Imamura et al., 2005). HLS2 also demonstrates similar molecular features to RAS-3 and LICI, and its mRNA has only been detected in the hemolymph of partially or fully engorged nymphs and females, suggesting a possible role in endogenous hemolymph circulation. In the same tick species, the serpin HLSG-1 harbors high similarity (32–44%) to Japanese and mangrove horseshoe crab clotting factor C precursors, mouse manan-binding lectin serine protease 1, and rat/mouse hepsin proteins (Mulenga et al., 2001). In partially fed ticks HLSG-1 transcripts were weakly expressed in the midgut and strongly detected in the SGs.

As strict haematophagous acari, ticks require blood meals to complete their development and reproduction. These arthropods are pool-feeders and create haemorrhagic pools from which they collect nutritive fluids while biting, and interestingly, some female hard ticks can imbibe enough blood to increase in size by as much as 100 times (Sonenshine and Anderson, 2014). This unique feeding method implies the existence of very effective blood uptake and digestion mechanisms, in which several tSPIs have now been implicated (Figure 2).

In 2001, Mulenga et al. identified HLSG-2 in H. longicornis ticks, a serpin with chymotrypsin-like protease selectivity (Mulenga et al., 2001). HLSG-2 transcripts were only detected in partially fed ticks and expression increased in parallel with feeding duration. HLSG-2 expression was strong in midgut and weak in salivary glands, suggesting probable links with blood uptake and digestion processes (Mulenga et al., 2001). In the same tick species, another serpin-1 (HLS-1) harboring similarities to the I. ricinus SG serpins RAS-2 and RAS-1 (Sugino et al., 2003) was identified. Taking into account additional sequence homologies with other anticoagulation factors, its specific expression in midgut of partially fed ticks, as well as the fact that clotting time can be delayed by HLS-1 in a dose-dependent manner, all indicate that HLS-1 has a probable role in tick blood meal uptake as well as maintaining blood fluidity in the midgut (Sugino et al., 2003). Then in 2010, Miyoshi et al. identified a Kunitz-type tSPI exclusively expressed in the midgut of adult ticks named HlMKI (Haemaphysalis longicornis midgut Kunitz-type inhibitor) (Miyoshi et al., 2010). Immunofluorescent analysis demonstrated that HlMKI likely interacts with HlSP, a hemolytic serine protease expressed in the tick midgut, and that both proteins harbored similar expression patterns with a 72 h peak during feeding (Miyoshi et al., 2007). HlMKI displayed inhibitory activity against the HlSP protein, and against chymotrypsin and elastase to a lesser extent. Altogether these results suggested that HlMKI can regulate blood digestion in tick midgut via HlSP modulation (Miyoshi et al., 2010). Lastly, another Kunitz-type tSPI was identified in H. longicornis hemocytes, and was named HlChI because of its chymotrypsin inhibitory profile (Alim et al., 2012). HlChI has strong chymotrypsin inhibitory activity but low trypsin inhibitory activity. The HlChI gene is expressed in larvae, nymphs, and adults during all feeding phases, and transient up-regulation has been clearly detected from feeding initiation to repletion. During feeding, HlChI is mainly localized in hemocytes, though low expression levels were also detected in midgut, salivary glands, ovaries, and the epidermis (Alim et al., 2012). Several HlChI-RNAi-injected ticks died 48 h after feeding, while others ingested significantly smaller and slower blood meals, laid fewer eggs, and demonstrated lower larvae conversion (Alim et al., 2012). During feeding, HlChI expression in hemocytes peaks at 96 h, coinciding with low proteolytic activity and a low homeostasis level maintained by a few principal inhibitors (Franta et al., 2010; Alim et al., 2012). Thus HlChI is likely an indirect but essential actor in both vital blood feeding and tick reproduction processes.

In R. appendiculatus ticks, three serine protease inhibitors of chymotrypsin or trypsin and named RAMSP 1-3 (R. appendiculatus midgut serine proteinases 1-3) were identified, and are likely involved in feeding processes or blood digestion (Mulenga et al., 2003a). RAMSP-1 and -2 transcripts have only been detected in partially fed ticks, while RAMSP-3 mRNA was detected in both unfed and partially fed ticks, with stronger signals in the latter. RAMSPs expression is not restricted to the midgut because RAMSP-1 is equivalently expressed in all tick organs, while RAMSP-2 is weakly expressed in both SGs and midgut, and RAMSP-3 is more weakly expressed in SGs than midgut (Mulenga et al., 2003a). Additionally, all three RAMSPs are expressed in tick carcasses (whole tick without SGs and midgut), suggesting widespread distribution of these inhibitors in other tick tissues (Mulenga et al., 2003a). In terms of RAS serpins, RAS-1 and RAS-2 mRNAs can be detected at all life stages, as well as in both sexes, with positive detection in 4-day partially fed and fully engorged ticks (Mulenga et al., 2003b). This suggests gene expression both during and after feeding, although they were not expressed in saliva, likely because they do not contain signal peptide sequences (Mulenga et al., 2003b). Indeed, RAS-1 and RAS-2 might be associated with feeding in both SGs and the midgut by modulating blood uptake and digestion. Confirmation occurred when significantly fewer fully engorged nymphs were counted when ticks were fed on vaccinated compared to non-vaccinated cattle (Imamura et al., 2006).

Then in 2014, NGS technologies enabled the identification of 22 RMS serpins (RMS-1 to RMS-22) from R. (B.) microplus (Tirloni et al., 2014a,b), among which 18 full-length coding sequences were identified (Rodriguez-Valle et al., 2015). While serpin consensus patterns were conserved, these 18 members of the RMS family showed high amino acid sequence variability suggesting a broad spectrum of targeted serine proteases. Transcription levels of RMS-13, -15, -16 in SGs, or RMS-6, -7, -9, -17 in both SGs and the midgut, suggested that they play a role in the blood meal process, either in uptake or during digestion (Rodriguez-Valle et al., 2015). Functionally, RMS-3 strongly inhibits chymotrypsin and elastase, but only weakly trypsin and thrombin. RMS-15 is a strong thrombin inhibitor and RMS-6 a chymotrypsin inhibitor. Finally while RMS-21 and -22 are not secreted, they were detected in the midgut and SGs suggesting a probable role in proteolysis activity during blood digestion (Rodriguez-Valle et al., 2015). Among the previously mentioned R. (B.) microplus BMTIs, the Kunitz inhibitors BmTi A and BmTI D were also believed to play an important role in feeding by inhibiting human prekallikrein (HuPK) implicated in the coagulation cascade (Sasaki et al., 2004), thus facilitating blood fluid uptake.

The serpin AamS6 was identified in the Amblyomma americanum tick (Mulenga et al., 2013). Both AamS6 mRNA and protein are strongly expressed in the SGs and midgut in unfed ticks, as well as during the first 72 h of feeding, before fading at 96 h, suggesting injection into the bite site and a role in tick anchorage to the host (Chalaire et al., 2011). As expected according to its serpin-like sequences, AamS6 interacted clearly and selectively with trypsin, chymotrypsin, elastase, and chymase, but also surprisingly with papain-like cysteine proteases, indicating that it is a cross-class protease inhibitor (Mulenga et al., 2013). In addition, AamS6 seems to transiently interact with fibrin-lysing plasmin. Despite only in vitro data providing the proof for plasmin-AamS6 interaction, it was hypothesized that this inhibitor could sustain blood flow to the tick feeding site and prevent clot formation (Mulenga et al., 2013). In this series of experiments, AamS6 only delayed recalcification time (RCT, the time to plasma clotting once calcium ions and blood-clotting co-factor(s) are reintroduced to citrate plasma), and did not inhibit any of the three coagulation pathways (extrinsic, intrinsic, and common, Figure 3). Platelet aggregation inhibition was also reported, that, when combined with previous results, bears out AamS6's involvement in inhibiting blood coagulation both in the midgut and via saliva secretion (Mulenga et al., 2013). Later, an A. americanum transcriptomic study identified a novel serpin, AAS19, which has the most fully conserved RCL among the ixodid ticks (Kim et al., 2015; Porter et al., 2015). The presence of the tripeptide Arg-Gly-Asp which constitutes a “RGD” motif, in the AAS19 sequence suggests a potential relationship between integrin GPIIb-IIIa and AAS19 during platelet aggregation (Nurden, 2014). Additionally, as AAS19 is abundantly and mostly expressed in the midgut at 96 h during feeding, it likely ensures that host blood doesn't clot during feeding (Porter et al., 2015). AAS19 demonstrates inhibitory activity against plasmin, FXa, and FXIa, and at a lower efficacy rate against FXIIa, FIIa (activated thrombin), FIXa, and tryptase, demonstrating its broad activity spectrum. In addition, the three coagulation pathways have delayed clotting in the presence of AAS19, likely due to weak thrombin inhibition. Altogether these results support AAS19 involvement in blood digestion (Kim et al., 2015).

Finally, Ixophilin was identified as a Kunitz thrombin inhibitor secreted in the midgut of I. scapularis and which shares homology with Hemalin and Boophilin (see below) (Narasimhan et al., 2013). Ixophilin was preferentially expressed in adult and nymphal midgut and was induced upon feeding, consistent with a potential role in preventing blood clotting in the midgut. In addition, Ixophilin mice immunization experiments demonstrated that ixophilin was necessary for efficient engorgement (Narasimhan et al., 2013).

Tick oviposition and egg laying are essential aspects of the tick life cycle determining tick population expansion. These processes are regulated by many proteins, including tSPIs (Figure 2). Inhibitors may be endogenous to certain organs and, for ovaries, it appears that some proteins—including tSPIs—can be captured from the midgut by receptor-mediated endocytosis followed by incorporation into the eggs (Tufail and Takeda, 2009).

Several tSPIs named BmTIs, with a similar target spectrum (trypsin, plasmin, and HuPK), have been discovered in the eggs and larvae of R. (B.) microplus (Tanaka et al., 1999). However, differing concentrations and inhibitor specificity changes have been reported between egg and larval stages (Andreotti et al., 2001; Sasaki et al., 2004). Andreotti et al. showed a 69.7 and 71.3% reduction in both engorged tick number and egg weight respectively when female ticks were fed on BmTIs-immunized cattle, confirming a crucial role for BmTIs in egg production and development (Andreotti et al., 2002). BmTI-6, a Kunitz tSPI identified in ovaries, was also expressed in tick body fat and demonstrated specific inhibitory activity against trypsin and trypsin-like proteases, such as plasmin (Sasaki and Tanaka, 2008). As for other BmTIs, BmTI-6 is suspected to play a role in controlling endogenous proteases in the egg and larval stages (Andreotti et al., 2001; Sasaki et al., 2004).

The RMS-3 serpin is expressed in the SGs of semi-engorged females, but lower expression levels were also observed in the midgut and ovaries of R. (B.) microplus (Rodriguez-Valle et al., 2012). In vitro feeding assays showed that both egg weight and larval transformation rates were reduced in female ticks pre-fed on anti-RMS-3 sheep serum, thus implicating RMS-3 in reproduction and egg development (Rodriguez-Valle et al., 2012). An additional Kunitz serine protease inhibitor with trypsin and kallikrein inhibitory activities named RmKK was recently discovered in R. (B.) microplus eggs (Abreu et al., 2014). Although the native protein was obtained from eggs, the RmKK transcript was only detected in the midgut, suggesting possible midgut expression and subsequent transport to the ovaries and egg incorporation. Interestingly, kallikrein inhibitors were thought to protect against undesired egg proteolysis (Willadsen and Riding, 1980). Lastly, four other serpins were also recently implicated in tick embryogenesis regulation or vitellogenesis: RMS-19 and RMS-20, which are expressed in all tissues and at all stages, RMS-6 which is only detected in ovaries, and RMS-21 and -22 which are only detected in eggs (Rodriguez-Valle et al., 2015).

Studies on the dog tick Rhipicephalus sanguineus identified a group of proteins belonging to the Kunitz/BPTI tSPI family, called RsTIs in larval stages (Sant'Anna Azzolini et al., 2003). Three of which—RsTIQ2, RsTIQ7 and RsTIS5—inhibit trypsin, neutrophil elastase, and human plasmin. RsTIQ2, and to a lesser extent RsTIQ7, also inhibit HuPK (Sant'Anna Azzolini et al., 2003). Because RsTIs are similar in structure and inhibitory activity to previously described tSPIs from R. (B.) microplus, a similar role in egg production has been hypothesized.

A human follistatin-related-protein (FRP) homolog was also identified in H. longicornis ticks, and is implicated in tick oviposition (Zhou et al., 2006). This protein harbors three distinct domains, a follistatin-like domain, a Kazal domain and two EFh calcium–binding motifs. Polyclonal antibodies revealed FRP presence in tick salivary glands, midgut, body fat, hemocytes, and a strong expression in ovaries (Zhou et al., 2006). RNAi experiments silencing FRP in adult rabbit-fed ticks as well as the use of anti-FRP antibodies showed significant negative effects on tick oviposition while no differences were observed in feeding duration, engorgement weight, and survival (Zhou et al., 2006).

Finally, when the aforementioned A. americanum AAS19 gene was RNAi silenced, ticks imbibed much less blood and presented curious body deformities compared to controls, likely due to deficiencies in hemostasis regulation (Kim et al., 2016). In addition, ticks that fed on rAAS19-immunized rabbits took smaller blood meals and detached prematurely. Following a second round of infestation on these rabbits, ticks also failed to lay eggs, suggesting an important role for AAS19 in both tick homeostasis and reproduction (Kim et al., 2016).

Hemostasis in vertebrates is a tightly regulated process to avoid blood leakage following injury (Aird, 2003). Many biochemical mechanisms are involved in blood clot formation with a central enzymatic cascade that can be activated by three different pathways (Figure 3). When the vessel epithelium is damaged, tissue factors (TF) expressed by epithelium cells initiate clot cascades, leading to activation of extrinsic blood coagulation pathways. Displaying TF triggers factor VII activation thus forming TF/VIIa complexes which then activate central factor X. The second intrinsic pathway is activated by contact factors—XII, prekallikrein (PK) and high molecular weight kininogen (HMWK)—that cause successive activation of intermediate factors (factors XI, IX), and IXa/VIIIa complex formation, which also activates factor X. Thus intrinsic and extrinsic coagulation pathways merge in order to activate Factor X to Xa. Then in the “common pathway,” activated factor Xa interacts with its co-factor Va and forms the pro-thrombinase complex. This complex then processes pro-thrombin (factor II) into thrombin (factor IIa) that converts fibrinogen into fibrin which then polymerizes to constitute a clot. Ticks prevent clot formation in the host, in the micro hematoma at the bite site, but also, in tick mouthparts and midgut. They inject numerous proteins via their saliva into the blood bowl, mostly tSPIs that affect serine proteases or their zymogens (Factors II, VII, IX, X, XI, XII, and prekallikrein) implicated in clot formation (Tatchell, 1969; Ribeiro et al., 1985). As anticoagulant factors, these inhibitors are essential in allowing ticks to take their blood meal, and concomitantly ingest and transmit pathogens. In addition, during the long feeding period, the blood in the tick midgut is maintained in a fluid state until repletion. To retain blood fluidity, anticoagulants are secreted from the tick midgut epithelium to the midgut lumen. Coagulation inhibitors secreted into the midgut mostly inhibit thrombin whereas anticoagulants secreted from SGs mainly inhibit FXa.

Because ticks are blood-feeding arthropods requiring hours to weeks to complete their blood meal, they have developed several mechanisms to evade host rejection during this long feeding period (Ribeiro, 1995; Francischetti et al., 2009; Kazimirova and Stibraniova, 2013; Kotal et al., 2015). Their saliva injected into the wound site carries many components able to manipulate vertebrate host responses, including anti-inflammatory compounds as well as immunomodulators acting on both innate and acquired immunity, and among which are included several tSPIs.

In A. americanum, the aforementioned anticoagulants AamS6 and AAS19 can also target plasmin, known for its role in pro-inflammatory cytokine release, monocyte and dendritic cell chemotaxis, neutrophil attraction, tissue remodeling, and wound healing, suggesting involvement in the regulation of monocyte, macrophage, and dendritic cell functions, and in inflammatory responses (Syrovets et al., 2012). In addition, 17 different serpins called lopsins (L1-L17) were found to be expressed in various organs such as SGs, midguts, ovaries, and the carcasses of partially fed A. americanum ticks (Mulenga et al., 2007). Sequence analysis revealed the presence of glycosaminoglycan binding sites on all lopsins, similar to several other proteins involved in the modulation of blood coagulation, inflammatory responses, or immune cell migration (Munoz and Linhardt, 2004). However, further investigations are required to identify their inhibitory targets and to decrypt mechanisms governing lopsins' likely immunomodulatory activities.

Iris from I. ricinus is also a powerful immunosuppressive molecule (Leboulle et al., 2002a). It is secreted at the tick-bite site and strongly inhibits elastase-like proteases (leukocyte elastase and pancreatic elastase) with rapid kinetics, thus repressing host inflammation (Prevot, 2006). Iris regulates innate immune mechanisms by suppressing T lymphocyte proliferation and inducing Th2-type immune responses with increased IL-4 and by inhibiting typical Th1 molecule production (IL2, IFN-γ) (Leboulle et al., 2002a). IRS-2, another serpin exhibiting specific anti-chymotrypsin activity, was identified from I. ricinus SGs (Chmelar et al., 2011). In SGs, IRS-2 is highly expressed at 2 days with maximum expression 6 days after attachment, suggesting a role in the early stage of feeding. IRS-2 inhibits both tissue swelling and neutrophil migration into inflamed tissue, modulates T cell differentiation, and decreases IL-6 production at both protein and mRNA levels in spleen dendritic cells activated by B. burgdorferi (Chmelar et al., 2011). In addition, by decreasing STAT-3 signaling molecule phosphorylation, IRS-2 impairs Th17 cell development (Páleníková et al., 2015). IRS-2 is also the only known serpin that targets both cathepsin G and chymase. Both of these proteases are secreted following neutrophil (cathepsin G) and mast cell (chymase) activation, and are involved in a huge range of physiological processes associated with the acute inflammatory response, particularly in crosstalk between neutrophils and platelets (Zarbock et al., 2007). IRS-2 also affects thrombin-induced platelet aggregation, thus likely playing multiple roles in inflammation and hemostasis particularly through the modulation of PAR activation (Chmelar et al., 2011).

The serpin Ipis-1, which shares 95.5% sequence identity with Iris, was isolated from SGs of fed Ixodes persulcatus and may be associated with immunomodulatory effects on both innate and acquired immune responses (Toyomane et al., 2016). Ipis-1 may directly interact with and inhibit T cells and CD14⁺ cells (mainly macrophages, neutrophils, and dendritic cells), with an as yet unidentified inhibitory mechanism. Similarly, Ipis-1 affects cytokine and chemokine activity via currently unknown mechanisms (Toyomane et al., 2016).

Based on previous I. scapularis sialome exploration (Ribeiro et al., 2006), a salivary Kunitz inhibitor with unusual structure was characterized and named tryptogalinin (Dai et al., 2012; Valdés et al., 2013). Tryptogalinin has a broad spectrum (potentially explained by the presence of three intrinsic disordered regions) and a high affinity for serine proteases playing an important role in inflammation and host immune responses (Heutinck et al., 2010). In fact, tryptogalinin inhibits trypsin, α-chymotrypsin, plasmin, matriptase, elastase, and especially human skin β tryptase found in mast cells (Valdés et al., 2013). Mast cell β-tryptase has a key role in host-inflammatory responses by stimulating chemoattractant release, such as IL-8, and by inducing IL-1β mRNA expression (Caughey, 2007). Tryptogalinin causes excitation of sensory neurons and has anti-inflammatory activity by repressing tryptase-PAR type 2 activation which normally mobilizes intracellular calcium stores to increase intracellular Ca²⁺ concentrations (Payne and Kam, 2004).

The glycosylated Kunitz inhibitor, TdPI (tick-derived protease inhibitor), was isolated from SGs of R. appendiculatus adult females (Paesen et al., 2007). TdPI is only expressed during the four first hours of feeding and manipulates host-immune defenses during the tick feeding process. TdPI potently inhibits trypsin and moderately affects human plasmin and human tryptase activity (Paesen et al., 2007). Mast cells and eosinophils initiate and/or amplify inflammation at the tick feeding site, mostly via the production and degranulation of several pro-inflammatory molecules such as histamine and tryptase. Accordingly, TdPI transcription coincides with that of RaHBPs (R. appendiculatus Histamine-Binding Proteins) known to sequester histamine in ticks (Paesen et al., 1999), thus highlighting a probable role for TdPI as a human tryptase inhibitor via complex formation (Paesen et al., 2007). In addition, TdPI may bind to tryptase inside mast cells, and may suppress its autocatalytic activation step (Sakai et al., 1996).

In R. (B.) microplus, BmSI 7 inhibitor regulates bovine neutrophil elastase pro-inflammation activity, due to its strong elastase inhibitory activity. Thus by reducing inflammation and/or avoiding tick tissue degradation, it enables the tick to thwart the host immune system (Sasaki et al., 2008). Similarly, RMS-3 is likely secreted into tick saliva during feeding, and carries a B-Lymphocyte epitope, highlighting a probable role in host-immune response modulation (Rodriguez-Valle et al., 2012). Finally, 12 further proteins have been extracted from R. (B.) microplus larvae, initially BmTI A-F, then BmTI 1-7, among which only BmTI 2 and BmTI 3 were further characterized as follows (Sasaki et al., 2004). Both BmTIs were classified into the BPTI-Kunitz family as they are suspected to have two Kunitz domains. Both inhibit trypsin and hNE, BmTI 2 demonstrating greater trypsin inhibition, and BmTI 3 stronger hNE inhibition, suggesting a role inhibiting the host's inflammatory response (Sasaki et al., 2004).

While ticks are biting, they must control and limit wound neovascularization and cell proliferation responses to enable blood meal uptake. Thus, they inject salivary proteins to favor apoptosis and slow down neovascularization as reported for I. scapularis (Francischetti et al., 2005a), and interestingly, tSPIs have also been implicated in this role.

Significant similarities are shared between BmCI from R.(B.) microplus (Lima et al., 2010) and dendrotoxins. Dendrotoxins are non-inhibitory Kunitz proteins able to block different ion channels (Na⁺, K⁺, Ca⁺) involved in both cell proliferation (Lang et al., 2005) and apoptosis (Nutt et al., 2002). BmCI appears to be highly cytotoxic and causes fibroblast cell death via pro-apoptotic activity, without affecting cell cycle integrity, likely by Ca⁺ channel activity regulation (Lima et al., 2010).

Haemangin was identified as a salivary Kunitz inhibitor from H. lonigicornis carrying one single Kunitz domain, which is up-regulated during blood feeding (Islam et al., 2009). It strongly inhibits trypsin and chymotrypsin, poorly inhibits elastase, and is able to efficiently stimulate degradation of both heavy and light plasmin chains during fibrinolysis, therefore strongly supporting plasmin-dependent fibrinolysis inhibition (Islam et al., 2009). Haemangin also inhibits chick aortic explant angiogenesis; human umbilical vein endothelial cell (HUVEC) differentiation, proliferation, and tube formation; and chick ChorioAlantoic Membrane (CAMs) neovascularization, demonstrating that it can impede with both pre-existing vessel angiogenesis and neovascularization (Islam et al., 2009). Haemangin also significantly induces apoptosis in HUVECs (Nagata, 2000), and affects wound healing in an artificially wounded HUVEC monolayer (Islam et al., 2009). Anti-haemangin RNAi experiments showed that ticks completely failed to make blood pools by 72 h and achieved significantly diminished engorgement by 144 h, while control ticks become engorged by 120 h, with the simultaneous increase of neovascularization in knockdown ticks (Islam et al., 2009). High-throughput studies also indicated that Haemangin can utilize multiple intracellular signaling pathways to negatively regulate angiogenesis, and angiogenesis-dependent wound healing (Islam et al., 2009). Interestingly, it was noticed that soft ticks—which are fast blood-feeders compared to hard tick—lack Haemangin homologs, thereby highlighting the importance of these molecules during the long blood feeding processes of hard ticks (Francischetti et al., 2005a).

BmTI-A from R. (B.) microplus also strongly inhibits neovascularization, and prevents new vessel formation in vitro by inhibiting plasma kallikrein, plasmin, and elastase (Soares et al., 2016). In addition, BmTI-A inhibits endothelial cell viability and proliferation through kallikrein inhibition (Lang et al., 2005), leading to an absence of bradykinin release, which normally stimulates cell growth and survival (Andoh et al., 2010). The inhibition of plasma kallikrein and plasmin by BmTI-A prevents angiogenic growth factor release during tick infestation, thereby averting cell adhesion mechanisms (Soares et al., 2016). Neutrophil elastase degrades a broad spectrum of extracellular matrix and cell surface proteins, and the release of growth factors such as TGF-β, PDGF, and VEGF, induces cell proliferation and migration. By inhibiting neutrophil elastase, BmTI-A could also inhibit angiogenesis (Wada et al., 2007). BmTI-A is also believed to have an inhibitory action on both vessel formation and wound healing, thus enabling R. (B.) microplus to continue feeding (Soares et al., 2012).

Finally, the anticoagulant Amblyomin-X from A. cajennense can also act as a proteasome inhibitor (Chudzinski-Tavassi et al., 2010). Amblyomin-X demonstrates cytotoxicity in different tumor cells lines, but not in healthy human fibroblast cells (Chudzinski-Tavassi et al., 2010; Morais et al., 2016). In human tumor cells, Amblyomin-X alters the expression of several genes related to the cell cycle (related to the G0/G1 phase, as supported by observed G0/G1 alterations), causes endoplasmic reticulum stress marker accumulation, slightly modulates intracellular calcium concentration [Ca2⁺], causes mitochondrial dysfunction, cytochrome C release, poly (ADP-ribose) polymerase (PARP) cleavage, and activates the caspase cascade (Chudzinski-Tavassi et al., 2010; Morais et al., 2016). Additional in vitro assays on endothelial cells showed that Amblyomin-X delays the cell cycle, inhibits cell proliferation and adhesion, tube formation, and membrane expression of the platelet-endothelial cell adhesion molecule-1 (PECAM-1) (Drewes et al., 2012). In vivo, Amblyomin-X reduces tumoral mass in a murine melanoma model, decreases the number of metastatic events (Chudzinski-Tavassi et al., 2010), and also inhibits VEGF-A-induced angiogenesis in the dorsal subcutaneous tissue in mice (Drewes et al., 2012). Similar effects were observed in chicken chorioallantoic membrane (CAM). These investigations confirm the anti-tumorigenic and anti-angiogenic properties of Amblyomin-X.

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