A tick saliva serpin, IxsS17 inhibits host innate immune system proteases and enhances host colonization by Lyme disease agent

Lyme disease (LD) caused by Borrelia burgdorferi is among the most important human vector borne diseases for which there is no effective prevention method. Identification of tick saliva transmission factors of the LD agent is needed before the highly advocated tick antigen-based vaccine could be developed. We previously reported the highly conserved Ixodes scapularis (Ixs) tick saliva serpin (S) 17 (IxsS17) was highly secreted by B. burgdorferi infected nymphs. Here, we show that IxsS17 promote tick feeding and enhances B. burgdorferi colonization of the host. We show that IxsS17 is not part of a redundant system, and its functional domain reactive center loop (RCL) is 100% conserved in all tick species. Yeast expressed recombinant (r) IxsS17 inhibits effector proteases of inflammation, blood clotting, and complement innate immune systems. Interestingly, differential precipitation analysis revealed novel functional insights that IxsS17 interacts with both effector proteases and regulatory protease inhibitors. For instance, rIxsS17 interacted with blood clotting proteases, fXII, fX, fXII, plasmin, and plasma kallikrein alongside blood clotting regulatory serpins (antithrombin III and heparin cofactor II). Similarly, rIxsS17 interacted with both complement system serine proteases, C1s, C2, and factor I and the regulatory serpin, plasma protease C1 inhibitor. Consistently, we validated that rIxsS17 dose dependently blocked deposition of the complement membrane attack complex via the lectin complement pathway and protected complement sensitive B. burgdorferi from complement-mediated killing. Likewise, co-inoculating C3H/HeN mice with rIxsS17 and B. burgdorferi significantly enhanced colonization of mouse heart and skin organs in a reverse dose dependent manner. Taken together, our data suggests an important role for IxsS17 in tick feeding and B. burgdorferi colonization of the host.


Introduction
Ticks and tick-borne diseases (TBD) impact public and veterinary health globally.Among those, Lyme disease (LD) caused by Borrelia species is one of the most important human TBD that has the most world-wide public health impact.The spirochete, Borrelia burgdorferi that is transmitted by Ixodes spp.ticks is responsible for LD in United States (US) and Europe, while B. afzelii and B. garinii are responsible for LD in Eurasia [1][2][3].Recently, a second LD pathogen B. mayonii was described in the US [4].Like other TBD agents, except a vaccine against tick-borne encephalitis approved by FDA in United States in 2021, there is currently no effective human vaccine against the LD agent.
In the absence of effective vaccines against the LD agent, avoidance of infectious tick bites is the only prevention method against LD currently.Despite a plethora of methods aimed at reducing infectious tick bites [5][6][7], LD cases have continued to increase.Confirmed and probable LD cases reported to the US Centers for Disease Control and Prevention have steadily risen from just under 20,000 in 1996 to more than 40,000 annual cases since 2008 (www.cdc.gov).According to insurance database, between 2010-2018, 476,000 LD cases were diagnosed and treated each year, with economic losses estimated at ~$786 million annually [8,9].
Given the ongoing rise in LD cases and search for better preventative measures, tick-antigen based vaccines have emerged among the most promising LD prevention approaches.This is based on evidence that repeatedly infested model animals that acquire immunity against tick feeding are protected against transmission of TBD agents including B. burgdorferi [10][11][12][13].
To date, there is no evidence of transovarial transmission (or passed from female ticks to larval ticks) of LD agents.Larval ticks acquire the spirochetes during feeding on infected reservoir hosts and then transtadially transmit to nymphs, which in turn transtadially transmit to

PLOS PATHOGENS
Tick saliva serpin inhibits innate immune responses and enhances colonization of Lyme disease agent adult ticks [28].Major transmission events of B. burgdorferi occur after the tick has fed for more than 48 h [29][30][31].The small size of the nymph tick and pain suppressants in its saliva that mask its presence on human skin allows the tick to go unnoticed and feed long enough for more than 36-48 h to transmit the LD agent [32].For that reason, although both nymph and adult ticks are capable of transmitting LD agents to the human host, most reported LD cases were associated with infectious nymph bites [3,33].On this basis, we recently identified tick saliva proteins of B. burgdorferi infected I. scapularis nymphs that were secreted every 12 h throughout feeding [25].
This study was initiated to understand functional roles of I. scapularis tick saliva serine protease inhibitor (Serpin; GenBank accession# EEC18973.1 or XP_002415308.5) in tick feeding and B. burgdorferi colonization of the host.We later found that IxsS17 was among homologs (orthologs) to Amblyomma americanum serpin (AAS) 19 that were characterized by the functional domain reactive center loop (RCL) being 100% conserved in all tick species according to currently available data [34,35].We also reported that IxsS17 and its homologs are among the proteins being injected into animals by adult I. scapularis [23], A. americanum [35], and Rhipicephalus microplus [27] ticks.In our recent study, we found that B. burgdorferi infected I. scapularis nymphs predominantly secreted IxsS17 at 48h feeding time point when major B. burgdorferi transmission events are expected [25].Additionally, we showed that RNAi silencing of IxsS17 [36] and its A. americanum homolog, AAS19 [23] caused mortality and reduced tick feeding efficiency.This evidence suggested that functions of IxsS17 and its homologs are related to tick feeding and transmission of tick-borne pathogens including B. burgdorferi.Consistent with functional analyses of IxsS17 homologs in A. americanum (AAS19; [35]), R. microplus (RmS-15; [37,38]), and recently in R. haemaphysaloides (RHS8; [39]) and I. ricinus (Iripin-8;), we provide new information that IxsS17 is an anticoagulant that is potentiated by binding heparin.Significantly, we further show that IxsS17 promotes B. burgdorferi colonization of the host by inhibiting host inflammation, blood clotting, and complement system effector proteases.

I. scapularis serpin (IxsS) 17 is not redundant and is conserved across Ixodidae tick species
BLASTP search of IxsS17 (EEC18973.1 or XP_002415308.5)amino acid sequence against entries in GenBank retrieved one sequence match of more than 77% amino acid identity per tick species except for Dermacentor silvarum, which has two matches (S1 Fig) .The next highest matches to IxsS17 in I. scapularis and other tick species showed amino acid identity levels of less than 50%.This indicates that IxsS17 and its homologs, in other tick species, are not redundant except for D. silvarum, which has two matches that differ by an 11 amino acid deletion to IxsS17: KAH7955208.[40].In this manuscript, we show that the I. scapularis genome (RefSeq GCF-016920785.2) encode for 62 serpins including IxsS17 as revealed by unique serpin RCLs (S1 Table ).Pairwise and global alignment of the 62 RCLs with coverage set to between 80-100% confirmed that IxsS17 RCL was not redundant as all other RCLs are 29-52% identical to IxsS17.
When compared to its homologs in other tick species both EEC18973.1 [41,42] and XP_002415308.5 [43] have a 53 amino acid sequence extension at the amino terminus end.SignalP 6.0 software did not identify the signal peptide in EEC18973.1 or XP_002415308.5

Yeast expressed rIxsS17 inhibits trypsin-like proteases and its inhibitory function is affected by hexa-histidine tag location
Canonical mode of serpin inhibitory activity is mechanical disruption of the target protease which starts with the C-terminal reaction center loop (RCL) irreversibly trapping the target protease [44].Determined to investigate the effect of the hexa-histidine (His) fusion tag on inhibitory functions of rIxsS17, we successfully expressed, and affinity purified three rIxsS17 constructs: (1) the hexa-histidine tag located at the N-terminal or ( 2 [49,50].In serpins, glycosylation is important for proper protein secretion, stability, and their half-life extension [46,51]. We initially used the C-terminal hexa-His-tagged rIxsS17 in substrate hydrolysis assays of 17 serine proteases related to host responses against tick feeding (Fig 3A and S2 Table).This screen showed that rIxsS17 (1 μM) inhibited pancreatic trypsin (1.5 nM) by 96-100%, rat skin trypsin IV (2.0 nM; in house expressed) by 90-99%, blood clotting factor (f) Xa (2.3 nM) by 79-80% followed by inhibition of blood clotting fXIa (3.7 nM), plasmin (33.7 nM), and
Next, we tested if proximity of the hexa-His-fusion tag to the RCL or its absence affected inhibitory activity of rIxsS17 against selected proteases (Fig 3B).As shown, rIxsS17 with N-terminal hexa-His-fusion tag had an 8.6% decrease in the inhibitory activity against trypsin.This suggests that beside the C-terminus domain that contains RCL, extension of the N-terminus region of the serpin might affect its inhibitory activity against trypsin.For both factor Xa and thrombin, inhibitory activity of the three constructs were similar.Since C-terminal histidine and non-tagged rIxsS17 have equal inhibitory activity, either of them was used in our downstream assays.
To determine the efficiency and rate at which rIxsS17 inhibits pancreatic trypsin, trypsin IV, and factor Xa, stoichiometry of inhibition (SI) and association rate of constant (k a ) were calculated (Fig 4).As shown, the SI (amount of rIxsS17 needed to inhibit one molecule of protease) for C-terminal His-tagged rIsS17 against trypsin, trypsin IV, and factor Xa was estimated at 12.9, 10.5, and 68 respectively (Fig 4A -4C).The rate of rIxsS17 (k a ) inhibition of trypsin, trypsin IV, and factor Xa was 2.7 ± 0.003 x10 3 M -1 s -1 , 3.9± 0.0001 x10 3 M -1 s -1 , and 5.4 ± 1.1 x 10 2 M -1 s -1 , respectively (Fig 4D -4F).(1 μM) against 17 serine proteases (with indicated concentrations) in the substrate hydrolysis assays.Substrate hydrolysis was monitored at A 405nm every 11s for 30 min at 30˚C.Inhibition rate was calculated using the formula: 100-V i /V 0 x 100, where V i = activity in the present of, and V 0 in the absence of rIxsS17.(B) Inhibition activity of C-terminal, N-terminal and non-Histidine tagged rIxsS17 against trypsin, factor Xa and human thrombin was determined using the substrate hydrolysis assay.Data represents mean ± SEM calculated from 3 biological replicates.The difference was analyzed using ANOVA in GraphPad Prism 9 and is statistically significant when P value � 0.05.https://doi.org/10.1371/journal.ppat.1012032.g003 The concentration of rIxsS17 (1 μM) used in the inhibitory assays may not reflect the physiological levels of this protein in tick saliva, however, is at optimal concentration of a single tick salivary recombinant protein to be biologically active in-vitro or ex-vivo (1-6 μM) according to Chmelař et al., 2016 [52].The reason is in the complex salivary mixture, this high concentration could be achieved by combination with numerous redundant proteins.

rIxsS17 interacts with both innate immune system effector proteases and regulatory protease inhibitors as revealed by protein-to-protein interaction analysis
Fig 5A -5C, and Table 1 and S1 File summarize the differential precipitation protein-protein interaction analysis of rIxsS17 and human plasma proteins.Since low amounts of rIxsS17 were detected in fractions 1-6, we pooled fractions 1-3, and 4-6 while fractions 7-10 where individually analyzed in LC-MS/MS analysis (S1 File).Next, we used PSOPIA (prediction server of protein-to-protein interactions; https://mizuguchilab.org/PSOPIA/) to analyze the interaction if NSAF (normalized spectral abundance factor) value was higher in rIxsS17 and human plasma mixture compared to plasma only controls.The interactions that were predicted with more than 75% likelihood by PSOPIA were considered true (Table 1).This analysis confirmed substrate hydrolysis results and revealed novel insights that rIxsS17 likely interacts with both effector proteases and regulatory protease inhibitors of the innate immune system (Table 1).Various molar ratios of rIxsS17 to proteases (0, 2.5, 5, 10, 20, 25, 50) were incubated for 15 min at 37˚C with constant concentration of bovine trypsin (1.5 nM) or rat trypsin IV (2.0 nM) or factor Xa (13.9 nM).In the presence of appropriate substrates, residual enzymatic activity was measured and plotted against rIxsS17: protease molar ratio.The SI was determined by extrapolating to the rIxsS17: protease ratio where protease activity is zero (Y axis = 0).The inhibition rate (k a ) of rIxsS17 was determined against bovine trypsin (D), rat trypsin IV (E) and factor Xa (F).Different concentrations of rIxsS17 (50,100,200, 400, 600 and 1000 nM) were incubated with constant amounts of bovine trypsin (14.6 nM), trypsin IV (12.5 nM) or factor Xa (13.9 nM) for different periods of time (0, 1, 2, 4, 6, 8, 10 and 15 min) at 37˚C.The residual protease activity was measured and plotted against time to determine the pseudo-first order constant, k obs .Consequently, the second-rate constant (k a ) was determined by the best fit line slope of the k obs values that were plotted against rIxsS17 concentration.https://doi.org/10.1371/journal.ppat.1012032.g004 Differential precipitation and PSOPIA analysis revealed that rIxsS17 interacted with blood clotting system factors (f) II (prothrombin), fX, fXII, plasma kallikrein, and plasminogen alongside blood clotting regulatory protease inhibitors; antithrombin III (serpin), heparin Protein-to-protein interaction using differential precipitation of proteins (DiffPOP) analysis reveals novel IxsS17 functional insights.25 μg of affinity purified rIxsS17 (A-C) or rIxsS4 (D-F) was incubated with human plasma in reaction buffer (20mM Tris-HCL and 150mM NaCl pH 7.4) overnight at 37˚C.The reaction was stabilized using Phosphoprotein Kit-Buffer A and subjected to repeated precipitation (X10) using methanol and acetic solution (90% methanol to 1% acetic acid).Appropriately washed precipitates of each fraction were resolved on 10% SDS-PAGE and transferred onto PVDF membrane for western blot analysis using monospecific antibodies to rIxsS17.A and D = human plasma only, B and E = human plasma mixed with rIxsS17 or rIxsS4, and C and F = rIxsS17 or rIxsS4 alone.Ladder (L), Number (1-9) represents each fraction from differential precipitation.Please note that, fraction 10 for rIxsS17 is not shown in this figure; however, its LC-MS/MS data analysis is available in S1 File.

rIxsS17 binds glycosaminoglycans (GAGs)
Homology modeling predicted that IxsS17 secondary structure, which was scored at Coulombic electrostatic values of -10 to 10 on ChimeraX server has a single basic positive patch located near the RCL (Fig 6A and 6B).Basic positive patch could potentially bind negatively charged ligands such as GAGs [53,54].Consistently, docking analysis conducted by AutoDock Vina and ADT v1.5.4 demonstrated that the IxsS17 secondary structure is likely to bind with heparin (Fig 6B).For the docking, nine poses were predicted and the result with the binding affinity of -12.6 Kcal/mol and the lower bound and upper bound RMSD as 0 were selected to be the best docked conformation.Further, the result generated was visualized by PyMOL, which shows that 5 amino acids of IxsS17 (Asp185, Lys188, Lys210, Ser211, Thr212) were interacting with heparin; and out of 5 binding sites, 2 heparin binding sites (Lys 188 and 210) lies within the positively charged or basic patch which was predicted by Chimera X.Finally, we confirmed that rIxsS17 has high binding affinity for heparin followed by chondroitin sulphate and dermatan sulphate (Fig 6C).However, it was interesting to observe that rIxsS17 bound to heparin, but not to heparan sulphate (Fig 6C).It might be because of structural variations between heparin and heparan sulphate, such as the chain of heparan sulphate is generally longer, with higher molecular weight (30kDa) than heparin (15kDa).Furthermore, l-iduronic acid predominates in heparin while d-glucuronic acid represents most of the uronic acid found in heparan sulfate.This changes the structure configuration resulting into alteration in binding affinity.Most importantly, heparin is the complete modified version of heparan sulphate and contains highest negative charge density of any known biological macromolecule which will increase its binding affinity to the positive patch of rIxsS17 [55].The relative binding affinity to heparin was further determined showing that rIxsS17 bound on the heparin column and was eluted at 0.25-1M of NaCl (Fig 6D).For positive control, IxsS17 A. americanum homolog, rAAS19 which is known for high binding affinity to heparin and having 4 basic patches [35] was eluted at higher concentrations of NaCl (1-3M).The negative control r1E1 (KF990169) does not bind heparin and does not have a basic patch [45], therefore, came out in the flowthrough.

Binding of heparin significantly enhances anti-blood clotting effects of rIxsS17
In preliminary studies, we empirically determined that 2 μM of rIxsS17 delayed plasma clotting by more than 60 seconds compared to buffer control.Next, we tested if the combination of rIxsS17 and 17 kDa heparin had synergistic anti-plasma clotting effect.As heparin is an approved blood clotting disorder therapeutic [56][57][58], it is interesting to note that pre-incubating plasma with the rIxsS17 and heparin mixture significantly delayed plasma clotting up to 532.9 seconds compared to clotting time for buffer control (64.5 seconds), rIxsS17 only (184 seconds), and heparin only (407 seconds) ( Fig 7).It is also notable that plasma clotting was also delayed to 474 seconds when a reaction was assembled from plasma that was incubated separately with rIxsS17 and heparin.

rIxsS17 inhibits complement activation via the mannose-binding lectin pathway and rescues B. burgdorferi from complement-mediated killing
Consistent with protein-to-protein interaction (in silico and ex vivo) showing that rIxsS17 interacted with complement system serine proteases (C2, C1s and factor I), our data shows that IxsS17 is an inhibitor of the complement system (Table 1 and Fig 8).We successfully used the WIESLAB complement system kit to independently assess three complement activation pathways.The results demonstrated that rIxsS17 significantly inhibited deposition of the complement membrane attack complex (MAC) via the mannose-binding lectin (MBL) complement activation pathway and moderately via the classical and alternative complement activation pathways (S4 Table ).In the initial screen, rIxsS17 molar excess (4 μM) reduced MAC deposition by ~40, 62, and 99% via the classical, alternative and MBL pathway, respectively (S4 Table ).Moreover, we found that rIxsS17 dose dependently reduced by more than 55% MAC deposition through 31 nM of rIxsS17 (Fig 8).In this study, dose response analysis was not done for the classical and alternative complement activation pathway.B. burgdorferi can activate three complement pathways resulting in several host defense mechanisms that include: opsonization, phagocyte recruitment, priming of the adaptive immune system, and bacteriolysis [59,60].Serum-mediated bacteriolysis has been used to test the sensitivity of LD spirochetes to normal human serum [61,62].Next, we tested if rIxsS17 was able to rescue B. burgdorferi from complement-mediated killing in vitro (Fig 9A -9D).Consistent with its inhibitory effect against complement activation (Fig 8), rIxsS17 dose dependently rescued the complement sensitive B. burgdorferi strain B314/pBBE22luc from complement killing.At 1 h post incubation, B. burgdorferi survival rates ranged from 73-100% and were not different among the tested groups (Fig 9A).At 2 and 2.5 h, only the positive control (complement resistant B. burgdorferi strain B314/pCD100) and the 1μM rIxsS17 groups had higher survival rates than negative control, heat-inactivated rIxsS17 and PBS (protein buffer control) (Fig 9B and 9C).At 3 h of incubation, 0-14% of the negative control survived while survival increased to 19-21% (22 ± 2.7%), 25-31% (28 ± 1.8%), 35-47% (42 ± 3.7%) and 55-70% (64 ± 7.9%) in the presence of 0.25, 0.5, 0.75 and 1 μM rIxsS17, respectively (Fig 9D).The positive control had survival rates of 43-67% (56 ± 12.1%).In heat-inactivated normal human serum, survival rates of all the tested groups ranged from 97-100%.

Co-injecting rIxsS17 enhances B. burgdorferi colonization of C3H/HeN heart and skin (ear)
Next, we tested if rIxsS17 supported B. burgdorferi colonization of the C3H/HeN Lyme disease mouse model.We initially inoculated six groups of mice (4 mice per group) with BSK-II Anti-blood clotting effects of heparin binding on rIxsS17 was determined in the recalcification time assay.Universal coagulation reference human plasma was preincubated with (1) reaction buffer control, or (2) heparin only, or (3) rIxsS17 only, or (4) rIxsS17 and heparin preincubated together, or (5) rIxsS17 and heparin pre-incubated separately.CaCl 2 was added to trigger blood clotting and the reaction was monitored at A 650nm every 20s for 20 min.The A 650nm data were then fitted in the Sigmoidal doseresponse lines: blue (buffer control), red (heparin only), green (rIxsS17 only), black (rIxsS17 and heparin preincubated together), and brown (rIxsS17 and heparin pre-incubated separately).Clotting time was interpolated from the sigmoid line when A 650nm increases by 10% with 95% confident interval.Drop vertical lines A, B, C, D, and E = clotting time for buffer only (circle), rIxsS17 only (triangle), heparin only (square), rIxsS17 and heparin preincubated with plasma separately (upside-down triangle), rIxsS17 and heparin pre-incubated with plasma together (diamond).
https://doi.org/10.1371/journal.ppat.1012032.g007medium (negative control), B. burgdorferi in BSK-II (positive control) and B. burgdorferi mixed with various amounts of rIxsS17 (1, 2, 5 and 10 μM).B. burgdorferi infection was confirmed in all treated groups by in-vitro cultivation and conventional PCR (S5 and S6 Tables).Spirochete burden in the ear, heart, joint and bladder tissues did not show significant differences between the groups (S4 Fig).However, we observed a pattern of B. burgdorferi load decreasing with increasing concentration of rIxsS17 in the heart and bladder (S4 Fig) .Moreover, IgG antibody titer against B. burgdorferi was statistically lower in the higher rIxsS17 concentration groups (S5 Fig).
We decided to repeat the experiment with reduced concentrations of rIxsS17: 0.06, 0.125, 0.25 and 0.5 μM (Fig 10).We show that the spirochete load in the heart tissue of rIxsS17 injected mice with 0.06 and 0.125 μM of rIxsS17 was 5.7 and 4.3 folds significantly higher than B. burgdorferi only group (Fig 10A).Likewise, there is an apparent trend (P < 0.1) that the spirochete load in 0.06 and 0.125 μM rIxsS17 injected mice was 1.8 and 2.3 folds higher than the B. burgdorferi only group in ear tissues (Fig 10C).Similarly, there is an apparent high B. burgdorferi load in joints of mice that were co-injected with 0.125 and 0.25 μM rIxsS17 than control (Fig 10B ) and there is no apparent difference in the bladder (Fig 10D).To determine whether high concentrations of rIxsS17 affected the survival of B. burgdorferi in the inoculum, we incubated B. burgdorferi with 0, 0.06, 0.125, 0.25, 0.5, 1, 5, and 10 μM of rIxsS17 in vitro.This analysis revealed rIxsS17 did not have a negative effect on B. burgdorferi in culture as spirochete survival ranged from 95-98% up to 24h of observation (S7 Table ).
It is also notable that IgG titers to B. burgdorferi lysate antigen detected in ELISA of the B. burgdorferi control and rIxsS17-treated groups were not statistically different (S6 Fig) .However, IgM titers of the 0.06 and 0.125 μM rIxsS17 co-injected groups were significantly higher than 0.25 and 0.50 μM of rIxsS17 co-injected groups.Interestingly the IgM antibody of mice that were co-injected with 0.25 and 0.50 μM of rIxsS17 did not show any significant difference with B. burgdorferi control mice (Fig 11A and 11B).Furthermore, immune sera of 0.06 and 0.125 μM rIxsS17 co-inoculated mice bound multiple bands on western blots of lab cultured B. burgdorferi lysate (Fig 11C).

Discussion
This study provides data showing that I. scapularis serpin (IxsS) 17 regulates key functions that are important to tick feeding and B. burgdorferi colonization of the host.It builds on previous studies done by our lab that characterized the A. americanum tick serpin 19 as the only tick serpin that has its functional domain RCL perfectly conserved (100%) in all tick species as per available data [34,35].Like most parasites, ticks have the propensity for encoding redundant molecular systems.For tick serpins, it is common for multiple paralogs or isoforms showing more than 70% amino acid identity being transcribed by the same tick species [34,40,63] suggesting redundancy.Thus, the finding that the IxsS17 amino acid sequence does not show any matches to other I. scapularis serpins with more than 50% amino acid identity suggests that this protein represents a non-redundant tick serpin.It is notable that except for D. silvarum which encodes for two homologs to IxsS17 with more than 77% amino acid identity, 12 other tick species encoded single serpin homologs to IxsS17, with the RCL being 100% conserved.It is important to note that the two IxsS17 homologs in D. silvarum have the same RCL and the difference is limited an 11 amino acid deletion.We would like to note while amino acid sequence of IxsS17 suggested no redundancy, we are unable to know in this study if IxsS17 is also not functionally redundant.
The RCL plays an important role in the inhibitory functions of serpins [44], and thus it is not surprising that the functional roles of IxsS17 is similar to its homologs in A. americanum, AAS19 [35], R. microplus, RmS15 [37,38], and recently I. ricinus, Iripin-8 [64,65].Collectively, substrate hydrolysis and protein-to-protein interaction data in this study indicate that IxsS17 is an inhibitor of innate immunity effector proteases associated with inflammation, nociception (pain sensing), hemostasis, and complement innate immune defenses all of which must be blocked by ticks to feed and transmit tick-borne pathogens.In addition to food digestion [66], pancreatic trypsin which was highly inhibited by rIxsS17 is also found in blood circulation, accelerates blood clotting in the presence of calcium ions, pro-thromboplastic lipid, factor V, VII, and X [67], and it is also the major activator of protease-activated receptor 2 (PAR2) that initiates inflammation signaling [68].Likewise, trypsin IV, also highly inhibited by rIxS17 is associated with signaling of cutaneous local inflammation and nociception by activation of PAR2 on cutaneous neurons [69].Cathepsin G regulates the inflammatory responses by stimulating production and maturation of cytokines and chemokines and controls the functional state of immune cells [70].It is interesting to note that like I. ricinus Iripin-8 and A. americanum AAS19 [35,65], rIxsS17 inhibited plasmin.At a glance, IxsS17 inhibition of plasmin is counterintuitive because plasmin is known for degradation of fibrin clots and preventing platelet aggregation by cleaving PAR1 [68,71], which will benefit tick feeding.However, plasmin has additional functions on inflammation and wound healing by directly interacting with various cell types including leukocytes (monocytes, macrophages, and dendritic cells) and cells of the vasculature (endothelial cells, smooth muscle cells) as well as soluble factors of the immune system and components of the extracellular matrix [72,73].In these interactions, plasmin contributes to inflammation and thus, its inhibition by IxsS17 will help tick feeding through prevention of inflammatory processes.While substrate hydrolysis reveals that rIxsS17 weakly inhibited human thrombin and did not inhibit bovine thrombin, our protein-to-protein interaction data show that this protein interacted with thrombin (or blood clotting factor II) similar to its homologs; AAS19, Iripin-8, and Rm-15 [38,65,74].
Our protein-to-protein interaction data also revealed functional insights of rIxsS17.The finding that rIxsS17 interacted with both effector proteases and regulatory protease inhibitors of the innate immune system is intriguing because serpins are known for their role in inhibiting functions of effector serine and cysteine proteases [44].Thus, it is surprising to note that rIxsS17 interacted with both effector proteases and serpin regulators of the blood clotting pathway, antithrombin III (AT), heparin cofactor II (HCII), kininogen-1 and protease C1 inhibitor (C1 inhibitor) of the complement system.Glycosaminoglycans (GAGs) such as heparin [75] serve as cofactors to enhance the activity of some mammalian serpins such as AT [76,77] and HCII [78][79][80].Comparative secondary structure modeling predicted basic patches in IxsS17, which were confirmed as functional heparin binding sites using GAG plate binding and heparin column assays.As IxsS17 and host serpins (AT and HCII) bind heparin, it is potentially possible that the observed interaction between IxsS17 and host serpins was because of heparin serving as a bridge between IxsS17 and host serpins.Likewise, as C1 inhibitor is heavily glycosylated [81,82], we hypothesize that IxsS17 may interact with the C1 inhibitor by binding onto GAGs linked to this protein.It is important to also note that in silico analysis predicted direct interactions between rIxsS17 and protease inhibitors with more than 95% chance.These observations warrant further investigations.
The function of serpins is effected by mechanical disruption of the target protease that starts with the target protease being trapped at P1 residue in the RCL [44].Our results demonstrated that positioning the histidine fusion tag at the amino-terminus end and not the C-terminus end reduced the mechanical efficiency of rIxsS17 which was restored by cleaving off the Histidine tag.The observed stoichiometry inhibition of rIxsS17 against trypsin, rat trypsin IV, and factor Xa were high and not close to the ideal 1:1 serpin-to-target protease ratio.These findings could be explained by evidence some of the serpins such as blood clotting regulatory serpins antithrombin III and heparin II that require binding of glycosaminoglycans (GAGs) to enhance their inhibitory potency [58].Similarly, we have shown that heparin binding potentiated functions of AAS19, the IxsS17 homolog in A. americanum [74].Likewise in this study, we determined that the putative basic patch in IxsS17 comparative tertiary structure was functional and bound GAGs including heparin, an approved therapeutic against blood clotting disorders [56,71].Consistent with AAS19 [74], heparin binding by rIxsS17 significantly increased its anti-coagulation activity.The finding that heparin binding potentiated the function of rIxsS17 suggests that when tick injects this protein into the host, native IxsS17 binds GAGs to enhance its anticoagulant functions.
Our protein-to-protein interaction findings showing that rIxsS17 interacted with complement system proteases.rIxsS17 likely blocks complement activation by potentially interfering with complement component C2 (associated with classical and MBL pathway), C1s (classical pathway), and factors I (alternative pathway).The evidence led us to investigate the effect of this protein on complement activation.Complement system activation can be initiated via binding of specific antibodies (Classical pathway), mannose binding lectins (MBL pathway) or small-scale activation of complement component C3 (Alternative pathway) [59,60].B. burgdorferi can activate all 3 complement pathways and results in direct complement-mediated killing of the spirochete [83].Moreover, Schuijt et al., [84] showed that the MBL pathway is of paramount importance in the eradication of B. burgdorferi.Here, we showed that rIxsS17 facilitated B. burgdorferi survival and promoted localization via inhibition of MBL pathway.By inhibiting the deposition of MAC in MBL pathway, rIxsS17 rescued B. burgdorferi from complement-mediated killing in vitro.
Prompted by evidence that IxsS17 is highly secreted by B. burgdorferi infected nymphs [25] within 24-48h of tick feeding, an open-window for transmission of LD agent, suggested that this protein is important to the transmission of B. burgdorferi.Kota ´l et al., [65] reported that Iripin-8 significantly influenced nymphal I. ricinus feeding but did not promote B. afzelii transmission as revealed by RNAi silencing.Here, we took a different approach and assessed the effect of co-injecting C3H with rIxsS17 and B. burgdorferi and show that this protein promoted colonization of C3H mice.rIxsS17 co-inoculation increased B. burgdorferi loads in heart and ear tissue, but not distal tissues like joint and bladder of C3H/HeN mice at 21 days post inoculation.This finding is relatively in agreement with in vivo experiment of Coumo et al., [85] that MBL deficiency mice had higher antibody titers and harbored significantly higher B. burgdorferi in skin tissue than deeper tissue (heart, joint and bladder) at 14 days post inoculation.Interestingly, our results showed that rIxsS17 effects on B. burgdorferi localization of mouse tissue is inverse dose dependent as revealed by B. burgdorferi load and IgM titers being higher in mice that were injected with low dose (0.06 and 0.125 μM) than high dose (0.25, 0.5, 1, 2, 5 and 10 μM) of rIxsS17.According to a study of Fikrig et al., [86], IgM to B. burgdorferi whole cell in infected mice peaks at day 14 post inoculation while IgG increases continuously even at day 180 after infection.Since IgG response to B. burgdorferi is much later during infection, it would explain why IgG titer in our experiment which we detected at day 21 (3 weeks) post inoculation did not show significant difference.
Although actual amount of IxsS17 that the tick injects into the mammalian host during feeding is unknown, it is estimated to be in picomolar range.Thus, it makes logical sense that low concentrations of rIxsS17 supported B. burgdorferi colonization and vice versa for the high dose.We hypothesize that the effects of low dose rIxsS17 used in this study better resemble the dose of the native protein in tick saliva.It will also be interesting to further investigate if the low spirochete load in mice that received the high dose of rIxsS17 was due to direct toxicity of this protein to B. burgdorferi or that the high dose over stimulated the innate immune system leading to clearance of the spirochetes.
LD control and prevention is challenging.The rise and spread of LD, and the fact that individuals can get LD more than once when bitten by an infected tick requires the development of novel effective vaccine against this vector-borne disease [87].The search for vaccine target antigens is shifting from the pathogen toward tick molecules, with the purpose of reducing tick density and B. burgdorferi infection among tick population and blocking the transmission of LD agents [17,87,88].In line with this, our data show that IxsS17 is an important protein in both tick feeding and B. burgdorferi colonization of the host, and it represents a possible target antigen in vaccines to prevent transmission of tick-borne disease agents.The mechanism by which IxsS17 protected B. burgdorferi from host innate immune response warrants future investigations.

Ethics statement
The use of animals strictly followed the animal use protocol approved by Texas A&M University Animal Care and Use Committee under the number IACUC 2020-0089.

Phylogeny and comparative sequence analysis
Amino acid sequence of IxsS17 was obtained from NCBI protein database (Accession # EEC18973.1)and blasted using the protein-protein BLAST (blastP) tool (non-redundant protein sequences (nr) and transcriptome shotgun assembly proteins (tsa_nr) database), resulting in 12 IxsS17 homolog sequences from different tick species.A homo sapiens antithrombin (Accession # CAA48690.1)was used as the out group in phylogenetic analysis of the sequences using MEGA X software [89].Pairwise sequence alignment analysis between IxsS17 and its homologs was used to determine protein identities.Based on the multiple sequence alignment analysis, the best protein model prediction and overall mean distance calculation, a phylogenetic tree was generated using maximum likelihood statistical method, bootstrap 1000 replicates, Le_Gascuel_2008 model and Gamma distributed with invariant sites (G+I) [90].In previous studies, we showed that IxsS17 sequence was likely not redundant because its functional domain RCL did not show amino acid identity of more than 55% to other I. scapularis serpins [34,35].To confirm these analyses, we compared the amino acid sequence of IxsS17 RCL to other I. scapularis serpin RCLs that were annotated in the recently updated I. scapularis genome [43].Briefly, the extracted RCL regions from previously published serpins [40] were used as a query database to search each of the 34,235 protein sequences from the current I. scapularis reference genome (ASM1692078v2) using DIAMOND blastp v2.0.15 [91].Each of the proteins with a positive hit to the RCL query was manually curated to confirm that it is a serpin with a complete RCL region.Finally, RCL regions from each of the confirmed I. scapularis serpins was compared to IxsS17 RCL using Needleman-Wunsch Global Alignment tool from NCBI.

Expression of recombinant (r) IxsS17
Expression of mature rIxsS17 (without signal peptide) was done in Pichia pastoris using the pPICZαA yeast expression plasmid as described [35,92].The coding domain for mature IxsS17 nucleic acid sequence was retrieved from GenBank (Accession # EEC18973.1)and optimized for expression in P. pastoris.Two rIxsS17 expression constructs with His-fusion tag placed at amino-or carboxyl-terminus were custom synthesized and cloned into pPICZαA (Biomatik, Wilmington, Delaware).To facilitate cleaving off the His-fusion-tag, the Tobacco etch virus (TEV) protease cutting site (ENLYFQG) was inserted between His-fusion tag and the IxsS17 coding sequence.The TEV protease was produced in house (see below).
Routinely, rIxsS17 expression plasmid was transformed into P. pastoris X33 (Thermo Fisher Scientific, Hanover Park, IL, USA), cultured at 28˚C in a bio-shaker (MaxQ 400, Thermo Fisher Scientific), and expression of rIxsS17 induced by feeding cultures daily with 5% methanol as a carbon source.For large scale expression (1 L batches), cultures were incubated for 5 days.The pPICZαA yeast expression plasmid secretes the recombinant protein into culture media.Thus, spent media was collected and rIxsS17 was precipitated out by ammonium sulfate saturation (525 g/L) at 4˚C overnight with stirring.Subsequently, precipitated rIxsS17 was dialyzed against His-tag affinity column binding buffer (100 mM Tris, 500 mM NaCl, 5mM imidazole, pH 7.4) and processed for affinity purification using the Hi-Trap Chelating HP column (Cytiva, Marlborough, MA, USA) under native conditions.Affinity purification was then confirmed by standard sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), silver staining using the Pierce Silver Stain Kit staining kit (ThermoScientific, USA), and western blotting analysis using the mouse monoclonal anti-Histidine antibody (GenScript, Piscataway, NJ) to determine purity.The concentration of rIxsS17 was determined using a bicinchoninic acid (BCA) kit (ThermoScientific, USA).The protein was stored at -80˚C until use.

Cleaving off the histidine tag from affinity purified rIxsS17
The TEV protease was produced in house using expression vector MBP-TEVcs (ENLYFQ/G)-His6-TEVΔ (220-242)-R5, a gift from Alice Ting (Addgene plasmid # 135456; http://n2t.net/addgene:135456; RRID: Addgene_135456) and Escherichia coli BL21 (DE3) (ThermoScientific, USA).In brief, the expression vectors were transformed into the E. coli BL21 strain.The transformed E. coli cells were grown in SOB medium (RPI Research Product International, Mount Prospect, IL) at 37˚C until the OD 600 reached 0.4-0.6.The TEV expression was then induced by cultivation with 1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) at 30˚C for 5 h.The cells were collected following by sonication in binding buffer (100 mM Tris, 500 mM NaCl, 5mM Imidazole, 10% glycerol, pH 7.4) and centrifugation at 10,000 × g, 4˚C for 30 min.After that, TEV was purified from the supernatants using the Hi-Trap Chelating HP column (Cytiva, Marlborough, MA, USA) under native conditions as described previously.The purity of the protein was assessed by Coomassie blue staining following SDS-PAGE gel electrophoresis.The concentration of TEV was determined using a NanoDrop 8000 spectrophotometer (Thermo-Scientific, USA).Finally, TEV was stored at -80˚C until use.
To cleave off the His fusion tag, affinity purified rIxsS17 (3 μg) and TEV (1 μg) were mixed and incubated at 4˚C overnight in cleaving buffer (50 mM Tris-HCl, 0.5 mM EDTA, 1 mM DTT, pH 8.0).The rIxsS17 and TEV ratio was empirically determined in preliminary studies.Subsequently, the reaction mix was dialyzed into Hi-Trap chelating column binding buffer and then purified as described above.Fractions of non-tagged-rIxsS17 were expected to elute into flow through and column wash fractions while both His-tagged TEV and non-cleaved rIxsS17 bound onto the affinity column matrix and were recovered into elution fractions.SDS-PAGE with silver staining and western blotting analysis using the His-tag antibody were used to confirm purification of His-tag free rIxsS17.Quantification was done as described above.
rIxsS17 deglycosylation.In a 50 μl reaction, 10 μg of rIxsS17 was deglycosylated using 2 μl of protein deglycosylation mix II (NEB, MA, USA) under denaturing condition following the instructions from the manufacturer.The reaction was incubated at 37˚C for 16 h.Deglycosylated rIxsS17 was analyzed by SDS-PAGE and silver staining.

Profiling inhibitor functions against innate immune system proteases
Inhibitory activity of rIxsS17 against 17 serine proteases related to host responses against tick feeding was determined using substrate hydrolysis assays as described [35,48,93].The serine proteases and their substrates used in this study are listed in S2 Table .1μM of rIxsS17 (Histagged at N-or C-terminal or His-tag removed) was incubated with empirically verified serine protease amount at 37˚C for 15 min in reaction buffer (20 mM Tris-HCl, 150 mM NaCl, 0.1% BSA, pH 7.4).Subsequently, 200 nM of the appropriate peptide substrate was added to the final reaction volume of 100 μL and hydrolysis was monitored at 405 nm wavelength every 11s for 30 min at 30˚C using microplate reader (Biotek Synergy H1, Winooski, VT, USA).The assay was performed in duplicates of 3 biological replications.Data were subjected to onephase decay analysis PRISM 9 to determine plateau values as proxies for initial velocity of substrate hydrolysis or residual enzyme activity.

Stoichiometry of inhibition (SI).
Preliminary substrate analysis revealed that molar excess of rIxsS17 inhibited pancreatic trypsin, trypsin IV, and blood clotting factor Xa by more than 70%-nearly 100%.To estimate the molar ratio of rIxsS17 to the target protease (pancreatic trypsin, trypsin IV, and factor Xa) required for 100% inhibition of enzyme activity of the target protease, stoichiometry of inhibition (SI) analysis was done as described [35,48,74].Various molar ratios of His-tagged rIxsS17 to proteases (0, 2.5, 5, 10, 20, 25, 50) were incubated for 15 min at 37˚C with constant concentration of bovine trypsin (1.5 nM) or rat trypsin IV (2.0 nM) or factor Xa (13.9 nM).The colorimetric substrate was added; and residual protease activity was determined as described above.SI or the molar ratio of rIxsS17 to protease was determined by plotting the percentage residual protease activity against serpin to protease ratios, fitting data onto the linear regression in PRISM 9, and extrapolation to the ratio which resulted in total loss of protease activity [94].
Affinity constant (k a ) calculation.The rate of rIxsS17 inhibiting bovine trypsin, trypsin IV and fXa was determined using the discontinuous method [93,94].Different concentrations of rIxsS17 (50,100,200,400, 600 and 1000 nM) were incubated with constant amounts of bovine trypsin (14.6 nM), trypsin IV (12.5 nM) or fXa (13.9 nM) for different periods of time (0, 1, 2, 4, 6, 8, 10 and 15 min) at 37˚C.The colorimetric substrate was added; and residual protease activity was assayed as described above.The pseudo-first order constant, k obs , was determined from the slope of a semi-log plot of the residual protease activity against time.The second-rate constant (k a ) was determined by the best fit line slope of the k obs values that were plotted against rIxsS17 concentration [94].
Homology modeling and prediction of basic patches and docking.Secondary structure modeling of IxsS17 was done on ChimeraX molecular modeling server [95].The mature IxsS17 protein amino acid sequence was pasted into Alphafold comparative modeling software on ChimeraX server, and the best IxsS17 comparative secondary structure model was reported.Putative basic patches were predicted using the electrostatic potential calculator in ChimeraX.Molecular docking study using the ligand molecule heparin (PubChem ID 22833565) with IxsS17 protein was conducted using Autodock Vina and Auto Dock Tools (ADT) v 1.5.4 from the Scripps Research Institute [96,97].The ligand was oriented suppositionally to allow flexible rotation and thus explore the most probable binding positions, while the receptor was kept rigid.The grid maps were calculated by Autogrid which represents the center of active site pocket for the ligand.The generated results were visualized by using PyMOL viewer (https:// pymol.org/2/).
Glycosaminoglycan (GAG) binding and effect on rIxsS17 function.Secondary structure modeling predicted at least one basic patch in IxsS17 comparative modeling structure.To test if the rIxsS17 basic patch is functional, rIxsS17 binding affinity of glycosaminoglycans (GAGs): heparin (Sigma-Aldrich, MO, USA), chondroitin sulphate A (Sigma-Aldrich), heparan sulphate (Galen Laboratory Supplies, North Haven, CT) and dermatan sulphate (Galen Laboratory Supplies) was done as previously described [48].A GAG-binding microplate (Galen Laboratory Supplies) was coated with 200 μL of GAG at the concentration of 25 μg/mL in binding buffer (100 mM NaCl, 50 mM Na-acetate, 0.2% Tween, pH 7.2) and incubated overnight at room temperature.After washing with binding buffer, the plate was blocked with 250 μL of 1% bovine serum albumin in PBS for 1 h at 37˚C.Thereafter, different concentrations of rIxsS17 (0, 1, 2, 5, 10 and 20 μg/mL) in 200μL of blocking buffer was added and incubated for 2 h at 37˚C.After the wash, 200 μL of HRP-conjugated anti-histidine antibody (1:5,000 dilution) was added.Following by addition of 200 μL of 1-step Ultra TMB ELISA substrate (Thermo Scientific), 100 μL of hydrochloric acid (1N) was used to stop the reaction; and the OD 450 nm was determined using a microplate reader (Biotek Synergy H1).
Heparin binding assay.Approximately 300 μg of rIxsS17 or rAAS19 (positive control) or r1E1 (negative control) was applied to the Hi-Trap heparin column (Cytiva).After washing with 10 mM phosphate buffer pH 7.4, the proteins were eluted using a gradient concentration (0.25-0.5-1.0-.2.0-3.0M) of NaCl.Samples included the protein before binding, flowthrough, wash, and elution were collected, resolved on 10% gel of SDS-PAGE, and subjected to silver staining for analysis.
Recalcification time assay.Prompted by preliminary findings that rIxsS17 was an inhibitor of blood clotting factors and it bound GAGs including heparin, we assayed the effect of heparin and rIxsS17 mixture on plasma clotting in a recalcification time assay, which evaluates the blood clotting system holistically [98].Five groups: (1) Buffer control (20 mM Tris-HCl, 150 mM NaCl pH 7.4), (2) 17 kDa heparin sodium salt (0.5 μg/mL) (Sigma-Aldrich, USA), (3) rIxsS17 (2 μM: empirically determined to delay plasma clotting by more than 60 seconds), (4) rIxsS17 and heparin mixture incubated separately, and (5) rIxsS17 and heparin incubated together were incubated in 40 μL of 20 mM Tris-HCl, 150 mM NaCl, pH 7.4 buffer for 5 min at 37˚C.Subsequently, 50 μL of pre-warmed (37˚C) universal coagulation reference human plasma (UCRP) (ThermoScientific, USA) was added to each group and incubated at 37˚C for an additional 5 min.After adding 10 μL CaCl 2 (final concentration of 150 mM) to trigger plasma clotting, optical density (A) was monitored at 650 nm wavelength every 20s for 20 min using the microplate reader (Biotek Synergy H1).Data from the recalcification time assay were plotted onto sigmoid line in PRISM 9. Initiation of plasma clotting (or clotting time) was interpolated from the sigmoid line when A 650nm increased by 10%, with 95% confident interval as published [74].
Differential Precipitation of Proteins (DiffPOP) and in silico protein to protein interactions.Differential precipitation of protein-to-protein interaction between rIxsS17 and human plasma was done as described [99].In a 1.5 mL vial, a reaction mix of 150 μL reaction containing 25 μg rIxsS17 and 20 μL of human plasma in reaction buffer (20 mM Tris-HCl, 150 mM NaCl, pH 7.4) were incubated at 37˚C overnight.Human plasma only and rIxsS17 only were also incubated in reaction buffer as negative controls.After the incubation, 100 μL of Phosphoprotein Kit-Buffer A (Clontech Laboratories, New York, NY) was added to stabilize the reaction prior to fractionation.To fractionate, precipitation solution (90% methanol/ 1% acetic acid) was added to the stabilized reaction mix, vortexed and incubated at room temperature for 5 min.Precipitates were collected by centrifugation at 14,000 rpm (or max speed) at 4˚C.The supernatant was transferred into a new 1.5 mL tube and process repeated until desired fractions are obtained (10 fractions in total).The pellet was washed in 400 μL of icecold acetone, air-dried, re-suspended in 100 μL reaction buffer and stored in -80˚C until use.
The expectation for this approach is that rIxsS17 will co-precipitate with its interactors.To determine fractions that co-precipitated with rIxsS17, each fraction was resolved on 10% SDS-PAGE gels and subjected to standard western blotting analysis using the monospecific antibody to rIxsS17.The monospecific antibody to rIxsS17 was purified from immune serum of rabbits that were repeatedly fed on by I. scapularis as previously described [24,48].The positive signal was developed using chemiluminescent substrates (ThermoScientific, USA).
Fractions that contain potential complexes with rIxsS17 were processed for LC-MS/MS analysis using the method published previously [25].To identify proteins, extracted tandem mass spectra was searched against the database of non-redundant human proteins from Gen-Bank using the Prolucid program in the Integrated Proteomics Pipeline (IP2) as published [100].The parameters used to identify potentially true rIxsS17 interactors included detecting at least two peptides in two of three independent LC-MS/MS runs and normalized spectral abundance factors (index for relative protein abundance in exceeded plasma only control values).Subsequently, these interactions were validated using in silico methods on the PSOPIA (prediction server of protein-to-protein interactions; https://mizuguchilab.org/PSOPIA/) server [101] and readouts with more than 75% likelihood to interact were considered as true.
Complement activity assay.Following up on protein-to-protein interaction results that rIxsS17 also interacted with complement system factors, we investigated its effect on complement pathway activations using the WIESLAB Complement Classical, Alternative and Mannose-binding Lectin (MBL) pathway Kits (Malmo ¨, Sweden).The kits allowed for independent assessment of rIxsS17 on the three complement activation pathways as measured by C5b-C9 or Membrane Attack Complex (MAC) deposition.For initial screening of rIxsS17 possible effect on the complement pathways, we started with high dose of rIxsS17, at 4 μM (20 μg) in 100 μL reactions.Subsequently, inhibition activity assessment of a serially dilute rIxsS17 (0.0078-4 μM) was done.First, rIxsS17 was pre-incubated with positive control (human serum provided with the kit) at 37˚C for 30 min.Then, the samples were added to the wells (provided in the kits) along with a blank (diluent only), negative control and positive control, and incubated at 37˚C for 60 min.The assay was performed in duplicates.After the washing step, conjugate and substrate were subsequently added following the instructions of the manufacturer.Finally, absorbance was read at 405 nm on a microplate reader (Biotek Synergy H1, Winooski, VT, USA).The effect of rIxsS17 on MAC deposition was calculated as follow: (Sample-NC)/ (PC-NC) x100 where NC is negative control and PC is positive control.
Borrelia burgdorferi complement sensitivity assay.Prompted by preliminary findings that rIxsS17 dose dependently reduced deposition of the MAC, we assessed its effect on rescuing complement sensitive spirochete as previously described [61,62,102].The complement sensitive B. burgdorferi (B314/pBBE22luc) and complement resistant (B314/pCD100) strains were kindly gifted by the Skare lab (TAMU Health Science Center).Both strains were propagated in BSK-II media at 32˚C, 1% CO 2 .For the assay, 15 μL of normal human serum (NHS) (Complement technology, TX, USA) was pre-incubated with serial dilutions of either C-terminal-His/non-tagged rIxsS17 (0.25, 0.5, 0.75, and 1μM), heat-inactivated rIxsS17 (1μM) or protein buffer (PBS; Phosphate buffered saline) at 37˚C for 30 min prior to addition of 85 μl of B. burgdorferi B314/pBBE22luc at the concentration of 10 6 cells/mL and inoculated in a bioshaker at 32˚C, 100 rpm.NHS with B. burgdorferi B314/pBBE22luc or B314/pPCD100 were included as negative and positive controls.Survival of spirochetes was assessed at 1.5, 2, 2.5 and 3 h post incubation.Spirochetes were counted from randomly chosen fields (10-15 fields) under dark-field microscope.Spirochete viability was judged based on cell mobility, membrane integrity, and cell lysis as described [102].Spirochete survival rates were calculated from 3 biological replicates.Heat-inactivated NHS (hiNHS) was used as the no complement-activity control and for normalization.
Effect of IxsS17 on B. burgdorferi colonization of C3H/HeN Lyme disease mouse model.Routinely, B. burgdorferi strain B31 (MSK5; kindly gifted by the Skare lab) were cultured in BSK-II medium and virulence plasmid Ip25 and Ip28-1 were verified using PCR primers (S3 Table ) as described [103].Groups of C3H/HeN mice (Charles River Laboratories, Wilmington, MA) (4 mice/group) were intradermally inoculated with 10 4 B. burgdorferi spirochetes or 10 4 B. burgdorferi spirochetes with various amounts of rIxsS17 (0.06, 0.125, 0.25, 0.5, 1, 2, 5 and 10 μM).Another group of 4 mice were inoculated with BSK-II + PBS as the negative control.At 21 days post inoculation, blood and tissue samples were collected from all mice.Serum was extracted from blood; and genomic DNA was extracted from tissues using DNeasy Blood and Tissue kit (Qiagen, MD, USA).B. burgdorferi infection was assessed by ELISA, western blot, in-vitro cultivation, PCR, and real-time qPCR methods.
Western blotting.Two μg of Bb lysate was resolved by SDS-PAGE and transferred to PDVF membrane.The membrane was blocked in 5% skim milk and then incubated with mouse antisera (diluted to 1:200) overnight at 4˚C.The anti-mouse IgM-HRP conjugates (ThermoScientific) at 5,000-fold dilution was used to detect primary antibodies.Finally, signal was detected using SuperSignal West Femto Maximum Sensitivity Substrate (ThermoScientific) under Biorad ChemiDoc MP Imaging system.
In-vitro cultivation of B. burgdorferi.In-vitro cultivation of B. burgdorferi was used to confirm colonization in mice organs.Within 1 h of completing necropsy, mice organs were submerged in 3-5 mL of BSK-II with appropriate antibiotics and antifungals and incubated at 32˚C, 1% CO 2 .The culture was examined bi-weekly for the presence of B. burgdorferi under dark-field microscope.
Real-time quantitative PCR.Real-time qPCR was used to quantify B. burgdorferi (Bb) in mouse organs targeting Bb flab; and Murine β-Actin was used as an internal control and for normalization (Primer sequences are listed in S3 Table) [103,104].The qPCR assays were performed in 10 μL reactions with 5 μL iTaq Universal SYBR Green Supermix (Bio-rad, Hercules, CA), 300 nM each primer and 10-50 ng mouse organ gDNA on a Bio-rad CFX96 real time system (Bio-rad).Data was analyzed by the comparative (C t ) method with the equation: Fold change = 2 -ΔΔCt = [(C t Flab-C t β-Actin) Bb-rIxsS17 co-injected group-(C t Flab-C t β-Actin) Bb only group] [105].The spirochete burden in mouse organs was expressed as the fold change of B. burgdorferi load in mice that were co-injected with rIxsS17 compared with mice injected with B. burgdorferi only.

Statistical analysis
Data was analyzed using GraphPad Prism 9 software and represented as mean ± SEM with statistical significance (P < 0.05) detected using the Student's t-test and two-tailed ANOVA.

Fig 1 .
Fig 1.Amino acid sequence analysis of IxsS17.(A) Phylogenetic tree was constructed using the MEGA-X software, Maximum Likelihood method and Le_Gascuel_2008 model with Bootstrap set to 1,000 replications.Group A, B, C and D represent amino acid identity levels of IxsS17 to its homologs in percentage.(B) Multiple sequence alignment of IxsS17 reactive center loop (EEGSEAAAVTGFVIQLRTAAF) and its homologs as well as the antithrombin III outlier was done in MacVector using T-Coffee specifications.Amino acids in the grey box are identical.https://doi.org/10.1371/journal.ppat.1012032.g001 ) C-terminal ends or (3) cleaved off using the inhouse produced Tobacco etch virus (TEV) protease (Fig 2).The rIxsS17 are glycosylated like other tick (IxsS-1E1 [AID54718.1],AAS19 [JAI08902.1],AAS27 [GAGD01011247.1], and AAS41 [JAI08957.1])and human serpins (antithrombin III and vaspin) [24,35,45-48].After deglycosylation treatment, protein sizes reduced ~2.5-5.0 kDa (S3 Fig).Glycosylation is the most common post-translational modification of proteins when the carbohydrate units are attached to the protein backbone either by N-or O-glycosidic bonds or both

Fig 2 .
Fig 2. Expression and affinity purification of recombinant (r) IxsS17.(A) Graphical illustration of three different rIxsS17 expression constructs that were custom synthesized: (1) C-terminal hexa-histidine tag, (2) N-terminal hexahistidine tag and Tobacco Etch Virus (TEV) cutting site (ENLYFQG) included, and (3) hexa-histidine tag is cleaved off at the TEV cutting site in the non-tagged rIxsS17.Please note all three recombinant constructs contain full-length sequence of rIxsS17.(B) Western blotting analysis of daily expression of rIxsS17 in Pichia pastoris culture.Culture (1 mL) were precipitated by ammonium sulfate saturation and resolved on 10% SDS-PAGE.rIxsS17 were detected in western blot using the HRP-conjugated monoclonal antibody to the hexa-histidine tag.(C) Silver staining and (D) Western blotting analysis of affinity purified rIxsS17.The hexa-histidine tag was detected in the C-terminal (Lane 1) and N-terminal-His-rIxsS17 (Lane 2) but not in the non-tagged rIxsS17 (Lane 3).https://doi.org/10.1371/journal.ppat.1012032.g002

Fig 3 .
Fig 3. Inhibition profiling of rIxsS17 against 17 serine proteases related to host responses during tick feeding.(A) Inhibition rates of C-terminal-Histidine tagged rIxsS17(1 μM) against 17 serine proteases (with indicated concentrations) in the substrate hydrolysis assays.Substrate hydrolysis was monitored at A 405nm every 11s for 30 min at 30˚C.Inhibition rate was calculated using the formula: 100-V i /V 0 x 100, where V i = activity in the present of, and V 0 in the absence of rIxsS17.(B) Inhibition activity of C-terminal, N-terminal and non-Histidine tagged rIxsS17 against trypsin, factor Xa and human thrombin was determined using the substrate hydrolysis assay.Data represents mean ± SEM calculated from 3 biological replicates.The difference was analyzed using ANOVA in GraphPad Prism 9 and is statistically significant when P value � 0.05.

Fig 4 .
Fig 4. rIxsS17 is a moderate inhibitor of trypsin, rat trypsin IV and factor Xa. Stoichiometry inhibition (SI) analysis calculated the amount of rIxsS17 needed to inhibit one molecule of bovine trypsin (A), rat trypsin IV (B), and factor Xa (C).Various molar ratios of rIxsS17 to proteases (0, 2.5, 5, 10, 20, 25, 50) were incubated for 15 min at 37˚C with constant concentration of bovine trypsin (1.5 nM) or rat trypsin IV (2.0 nM) or factor Xa (13.9 nM).In the presence of appropriate substrates, residual enzymatic activity was measured and plotted against rIxsS17: protease molar ratio.The SI was determined by extrapolating to the rIxsS17: protease ratio where protease activity is zero (Y axis = 0).The inhibition rate (k a ) of rIxsS17 was determined against bovine trypsin (D), rat trypsin IV (E) and factor Xa (F).Different concentrations of rIxsS17(50,100, 200, 400, 600 and 1000 nM) were incubated with constant amounts of bovine trypsin (14.6 nM), trypsin IV (12.5 nM) or factor Xa (13.9 nM) for different periods of time (0, 1, 2, 4, 6, 8, 10 and 15 min) at 37˚C.The residual protease activity was measured and plotted against time to determine the pseudo-first order constant, k obs .Consequently, the second-rate constant (k a ) was determined by the best fit line slope of the k obs values that were plotted against rIxsS17 concentration.

Fig 5 .
Fig 5.Protein-to-protein interaction using differential precipitation of proteins (DiffPOP) analysis reveals novel IxsS17 functional insights.25 μg of affinity purified rIxsS17 (A-C) or rIxsS4 (D-F) was incubated with human plasma in reaction buffer (20mM Tris-HCL and 150mM NaCl pH 7.4) overnight at 37˚C.The reaction was stabilized using Phosphoprotein Kit-Buffer A and subjected to repeated precipitation (X10) using methanol and acetic solution (90% methanol to 1% acetic acid).Appropriately washed precipitates of each fraction were resolved on 10% SDS-PAGE and transferred onto PVDF membrane for western blot analysis using monospecific antibodies to rIxsS17.A and D = human plasma only, B and E = human plasma mixed with rIxsS17 or rIxsS4, and C and F = rIxsS17 or rIxsS4 alone.Ladder (L), Number(1)(2)(3)(4)(5)(6)(7)(8)(9) represents each fraction from differential precipitation.Please note that, fraction 10 for rIxsS17 is not shown in this figure; however, its LC-MS/MS data analysis is available in S1 File.

Fig 6 .
Fig 6.IxsS17 binds heparin and the putative binding sites are located on the positive basic patch.(A)Comparative modeling of IxsS17 secondary structure was predicted on Chimera X server and heparin binding predicted using Autodock Vina and Auto Dock Tools.Heparin ligand (in blue) was arranged accordingly to be flexible to rotate and to explore the most probable binding positions (in red) while the receptor was kept rigid.RCL = reactive center loop.(B) Two heparin binding sites at Lysine 188 and 210 are located on the positive basic patch (Electrostatic potential is color coded: positive is blue; neutral is white and negative is red).(C) Binding affinity of rIxsS17 to 4 different GAGs: heparin (black circle), heparan sulphate (red square), dermatan sulphate (green triangle) and chondroitin sulphate (yellow upside-down triangle).The rIxsS17 was added into 96-well microplates previously coated with different GAG at the concentration of 0, 1, 2, 5, 10 and 20 μg/mL.Binding was detected using HRP-conjugated antibody to the histidine tag and documented as A 450nm .The data represent mean ± SEM from 3 biological replicates.(D) Silver staining of rIxsS17, rAAS19 (positive control), and r1E1 (negative control) eluted from heparin column.Recombinant proteins (~300 μg) were applied to the heparin column.After washing, the proteins were eluted using a gradient concentration of NaCl (0.25-0.5-1.0-2.0-3.0M).Serpins = the proteins before applying to the column, FT = Flow through.https://doi.org/10.1371/journal.ppat.1012032.g006

Fig 7 .
Fig 7.Heparin binding enhances rIxsS17 anti-coagulant activity.Anti-blood clotting effects of heparin binding on rIxsS17 was determined in the recalcification time assay.Universal coagulation reference human plasma was preincubated with (1) reaction buffer control, or (2) heparin only, or (3) rIxsS17 only, or (4) rIxsS17 and heparin preincubated together, or (5) rIxsS17 and heparin pre-incubated separately.CaCl 2 was added to trigger blood clotting and the reaction was monitored at A 650nm every 20s for 20 min.The A 650nm data were then fitted in the Sigmoidal doseresponse lines: blue (buffer control), red (heparin only), green (rIxsS17 only), black (rIxsS17 and heparin preincubated together), and brown (rIxsS17 and heparin pre-incubated separately).Clotting time was interpolated from the sigmoid line when A 650nm increases by 10% with 95% confident interval.Drop vertical lines A, B, C, D, and E = clotting time for buffer only (circle), rIxsS17 only (triangle), heparin only (square), rIxsS17 and heparin preincubated with plasma separately (upside-down triangle), rIxsS17 and heparin pre-incubated with plasma together (diamond).

Fig 8 .
Fig 8. rIxsS17 dose dependently inhibited membrane attack complex deposition of the Mannose-Binding Lectin pathway.A Blank, Negative control, Positive control or Positive control incubated with 2-fold-serial dilution of rIxsS17starting from 4 μM were added to the Mannan binding lectin (MBL) pathway Kit and incubated at 37˚C for 60 min.After the washing step, conjugate and substrate were subsequently added following the instructions of the manufacturer.Mac deposition rate was calculated using the formular: (Sample-NC)/(PC-NC) x100%, where NC is negative control and PC (or 0.0 μM of rIxsS17) is positive control.Data is presented as percent inhibition of MAC deposition mean ± SEM calculated from 3 biological replicates.https://doi.org/10.1371/journal.ppat.1012032.g008

Fig 11 .
Fig 11.Mouse groups co-inoculated with low dose of rIxsS17 have higher IgM titers compared to high dose injected mice.ELISA plates previously coated with B. burgdorferi crude antigen (200 ng/well) were tested with mouse sera that diluted at 1:250 (A) and 1:500 (B).IgM titers were determined using anti-mouse IgM monospecific antibody conjugated with HRP and absorbance were read at 450 nm.The data were presented as mean ± SEM; each dot is individual mouse.NC = negative control; Bb = B. burgdorferi.Statistical significance was evaluated using t test in GraphPad Prism 9 (***:P value � 0.001, ****: P value � 0.0001).(C).Western blot analysis of Bb lysate (2 μg) incubated with antisera from mice (diluted to 1:200) and anti-mouse IgM antibody-HRP conjugate (diluted to 1:5000).Images were taken at 18 seconds of exposure.Asterisks indicate extra and intense bands detected in mice challenged with Bb plus 0.06 and 0.125 μM rIxsS17.https://doi.org/10.1371/journal.ppat.1012032.g011

Table 1 . Fast fractionation and LC-MS/MS analyses identification of human plasma proteins that interacted with rIxsS17 and validation using in silico protein to protein interaction prediction PSOPIA software.
(Murakami and Mizuguchi, 2014)present likely protein to protein interactions(Murakami and Mizuguchi, 2014)https://doi.org/10.1371/journal.ppat.1012032.t001