2.1. Protein cloning, expression and purification

A full-length cDNA sequence of the gene that encodes Iripin-4 was obtained from the salivary glands of female I. ricinus ticks fed on guinea pigs by reverse transcription using the Transcriptor First Strand cDNA Synthesis Kit (Roche Applied Science, Penzberg, Germany) during a salivary-gland transcriptome project (Schwarz et al., 2013). Iripin-4 was amplified using the primers shown in Supplementary Table S1, cloned into a Champion pET-SUMO expression vector (Life Technologies, Carlsbad, California, USA) using NEBuilder HiFi DNA Assembly Master Mix (New England Biolabs, Ipswich, Massachusetts, USA) and transformed into Escherichia coli Rosetta 2 (DE3) pLysS competent cells (Novagen, Merck Life Science, Darmstadt, Germany). An overnight culture (10 ml l⁻¹) was inoculated into autoinduction TB medium (100 µg ml⁻¹ kanamycin, 34 µg ml⁻¹ chloramphenicol) and incubated for 24 h at 25°C. Bacterial cells were then harvested and disrupted with a cell disruptor. Soluble SUMO-tagged Iripin-4 was purified using a HisTrap FF column (GE Healthcare, Chicago, Illinois, USA) and eluted with 200 mM imidazole. After the first purification, the His and SUMO tags were cleaved using SUMO protease [1:100(w/w)] overnight at laboratory temperature. Samples were then reloaded onto a HisTrap column to separate the tags from native serpin. This step was followed by ion-exchange chromatography using a HiTrap Q HP column (GE Healthcare), which ensured sufficient protein purity. The final protein concentration was 1.17 mg ml⁻¹ in 20 mM Tris pH 7.4, 20 mM NaCl and the protein was stored at −80°C.

2.2. Analysis of complex formation of Iripin-4 with proteases

The formation of covalent complexes of Iripin-4 was performed as reported previously (Chmelar et al., 2011). Selected serine proteases (trypsin, thrombin, fXa, APC, fXIa, plasmin and elastase) were incubated with equimolar amounts of Iripin-4 for 1 h at laboratory temperature (final concentration of 1 µM). Serpin–protease complex formation was analysed on reducing SDS–PAGE using 12% gels followed by Coomassie staining. The second-order inhibition rate constants for native Iripin-4 against recombinant human granzyme B were then defined as described in Chmelar et al. (2011). In order to detect substrate hydrolysis while avoiding saturation of the reaction, the concentration used was chosen based on the biochemical properties of granzyme B. Thus, one unit of granzyme B hydrolyzes 1 pmol min⁻¹ Ac-IEPD-pNA substrate (200 µM) at 37°C. The assay was performed in triplicate at 37°C and stopped at each time point by addition of the aforementioned substrate. Prior to substrate addition, Iripin-4 was pre-incubated with granzyme B for 0–14 min at a protease concentration of 1 U µl⁻¹ (0.2 U µl⁻¹ final concentration). The assay was measured at five different concentrations of Iripin-4 and plotted against the serpin concentration. The slope of the line of best fit provided an estimate of the second-order rate constant. An appropriate buffer consisting of 20 mM Tris pH 7.4, 150 mM NaCl, 0.01% Triton X-100 was used. The rate of substrate hydrolysis was measured using a plate reader (BioTek Synergy H1 Plate Reader, USA; absorbance at 405 nm) in a 96-well plate.

2.3. Protein crystallization, X-ray data collection and processing

A suitable protein concentration for crystallization screening was determined using the Pre-Crystallization Test (PCT; Hampton Research, Aliso Viejo, California, USA). Crystallization experiments were performed using the sitting-drop vapour-diffusion technique in 96-well Swissci MRC 2-drop or 3-drop crystallization plates (Molecular Dimensions, Newmarket, Suffolk, United Kingdom). Initial screening of crystallization conditions and further crystallization experiments were carried out using an OryxNano crystallization robot (Douglas Instruments, Hungerford, United Kingdom). Commercially available crystallization kits (Crystal Screen, PEGRx, PEG/Ion and Index from Hampton Research and JBScreen JCSG++ from Jena Bioscience) were used to screen for crystallization conditions. Protein solution:well solution ratios of 1:0.5 and 1:1 (0.4:0.2 µl and 0.4:0.4 µl) were used with protein concentrations of 1.17 and 2.3 mg ml⁻¹ and the drop solution was equilibrated against 50 µl reservoir solution at 4°C. Suitable crystallization conditions were identified as JBScreen JCSG++ conditions H8 [25%(w/v) PEG 3350, 0.1 M bis-Tris pH 5.5, 0.2 M sodium chloride] and H10 [25%(w/v) PEG 3350, 0.1 M bis-Tris pH 5.5, 0.2 M ammonium acetate]. A simple 2D gradient was used for these conditions to optimize the crystal quality in a larger volume (0.5:0.5 µl).

For data collection, both native and cleaved Iripin-4 crystals were mounted directly from the crystallization drop and flash-cooled in liquid nitrogen without further cryoprotection prior to measurements. Diffraction data were collected on BL14.1 at the BESSY II electron-storage ring operated by Helmholtz-Zentrum Berlin (Mueller et al., 2015). Data for all crystals were processed using XDS (Kabsch, 2010) with the XDSAPP graphical user interface (Sparta et al., 2016). The solvent content of the crystal was analysed with MATTHEWS_COEF from the CCP4 suite (Winn et al., 2011) Data-collection statistics for both data sets are summarized in Table 1.

Table 1 X-ray data-collection and refinement statistics Values in parentheses are for the highest resolution shell.   Native Iripin-4 Cleaved Iripin-4 Data collection  X-ray diffraction source BL14.2, BESSY II, Germany BL14.2, BESSY II, Germany  Wavelength (Å) 0.9184 0.9184  Temperature (K) 100 100  Detector PILATUS 6M PILATUS 6M  Crystal-to-detector distance (mm) 175.66 373.61  Rotation range per image (°) 0.10 0.10  Total rotation range (°) 270.00 360.00  Exposure time per image (s) 0.10 0.10  Resolution range (Å) 44.61–2.30 (2.44–2.30) 46.42–2.00 (2.12–2.00)  Space group P3121 [No. 152] P21 [No. 4]  Molecules in asymmetric unit 1 4  a, b, c (Å) 78.93, 78.93, 117.78 65.70, 138.41, 80.22  α, β, γ (°) 90.00, 90.00, 120.00 90.00, 107.70, 90.00  Mosaicity (°) 0.105 0.128  Total No. of reflections 289932 (43011) 634544 (98469)  No. of unique reflections 19373 (3085) 91690 (14692)  Multiplicity 14.97 6.92  Average I/σ(I) 22.47 (4.30) 9.10 (2.32)  Completeness (%) 99.70 (99.70) 99.40 (98.60)  CC1/2 99.90 (79.70) 99.70 (77.30)  Rmeas† (%) 9.60 (62.40) 16.70 (85.30)  Overall B factor from Wilson plot (Å2) 39.20 23.00 Refinement  Resolution range (Å) 44.62–2.30 46.42–2.00  No. of reflections in working set 18404 89590  Final R‡/Rfree§ (%) 20.30/25.40 19.30/25.30  Mean B value (Å2) 39.80 28.00  No. of atoms in the asymmetric unit   Protein 2962 11869   Ligand: chloride ion   4   Ligand: nickel ion 1     Water 196 1276   Total 3168 13196  R.m.s. deviations   Bond lengths (Å) 0.007 0.007   Angles (°) 1.444 1.455  Average B factors, overall (Å2) 42.0 28.0  Ramachandran plot   Most favoured (%) 95.99 98.17   Allowed (%) 4.01 1.83  MolProbity score 1.46 [99th percentile¶] 1.27 [99th percentile¶]  PDB code 7zbf 7zas †Rmeas = , where the average intensity 〈I(hkl)〉 is taken over all symmetry-equivalent measurements and I(hkl) is the measured intensity for any given reflection. ‡R = , where Fobs and Fcalc are the observed and calculated structure factors, respectively. §Rfree is equivalent to R but is calculated for 1.87% of the reflections chosen at random and omitted from the refinement process. ¶The 100th percentile is the best among structures of comparable resolution; the 0th percentile is the worst.

2.4. Structure determination and refinement

The structures of Iripin-4 were solved using the CCP4 suite (Winn et al., 2011). The MOLREP (Vagin & Teplyakov, 2010) molecular-replacement method using MrBUMP, an automated pipeline for molecular replacement (Krissinel et al., 2018), was used for the cleaved conformation, and the structure of Iripin-2 (PDB entry 3nda; Chmelar et al., 2011), with 56.5% sequence identity, generated by an automated model search using the BALBES molecular-replacement pipeline (Long et al., 2008) was used for the native conformation. The resolution cutoff for native Iripin-4 was 2.3 Å due to the presence of two ice rings at 2.25 and 1.95 Å resolution. The structures were refined using REFMAC5 (Murshudov et al., 2011) from the CCP4 package (Winn et al., 2011) and were manually rebuilt in Coot (Emsley et al., 2010) based on evaluation of the electron-density peaks. Improvement during refinement was monitored by structure validation throughout the refinement process. Water molecules were added to the model using the REFMAC5 interface. The accepted solvent molecules had hydrogen-bond contacts with permissible geometry in the range 2.5–3.5 Å to protein atoms or to existing solvent. At this point, residues and water molecules with two possible conformations were included, and their alternative conformations were added for further refinement. In the final refinement steps, the Ni²⁺ ion (native structure) and four chloride ions (cleaved structure) were incorporated into the respective (2F_o − F_c) and (F_o − F_c) electron-density maps using coordinates from the ligand data bank in Coot (Emsley et al., 2010). The MolProbity server (Chen et al., 2010) and the wwPDB OneDep validation system (Young et al., 2017) were used for final qualitative model validation. All figures were created by PyMOL (Schrödinger) or CCP4mg (Mc­Nicholas et al., 2011). To visualize the cleaved conformation, chain C was used. The refinement statistics are summarized in Table 1.

3.1. Recombinant Iripin-4 production and purification

The Iripin-4 gene transcript amplified from tick cDNA showed three amino-acid substitutions, H78Q, G155E and G307D, compared with the deposited Iripin-4 sequence (GADI01002650; Supplementary Fig. S1) used for primer design. This is not an artefact but is most likely to be the result of high inter-tick variability in the genes expressed in salivary glands, as reported in Kotál et al. (2021). The resulting protein has no substitutions in the RCL or the hinge region, and therefore should retain the same inhibitory function and specificity. The final yield of Iripin-4 was 17 mg (∼2 mg per litre of medium), which is a sufficient concentration for further experiments (Supplementary Fig. S2). Using the NetNGlyc 1.0 and NetOGlyc 4.0 servers, the amino-acid sequence of Iripin-4 was estimated to contain two potential N-glycosylation sites [N-X-(S/T)] and four potential O-glycosylation sites (Supplementary Table S2).

3.2. Complex formation of Iripin-4 with proteases

Iripin-4 is not a suitable inhibitor of the selected serine proteases. No complex formation was observed between Iripin-4 and any of the tested serine proteases. In addition, Iripin-4 acted as a substrate for porcine elastase, as can be observed as a double band on a NuPAGE gel corresponding to native and cleaved serpin (Fig. 1). For this reason the second-order rate constant for Iripin-4 inhibition of granzyme B was determined (Fig. 2). Granzyme B is a component of cytolytic granules produced by natural killer and cytotoxic T lymphocytes. Granzyme B serves as a mediator of inflammation, controls the coagulation cascade (the ability to trigger inflammatory cytokines) and has proteolytic activity that is crucial for programmed cell death (Garzón-Tituaña et al., 2020). Iripin-4 protects cells by inhibiting granzyme B, and is also partly responsible for prolonging the attachment of the tick to the host. Mutants of Iripin-4 show similar but insignificant activity towards native Iripin-4 (data not shown).

Figure 1 Analysis of heat-stable complex formation between Iripin-4 and tested proteases. No complex formation was observed. There is visible cleavage of Iripin-4 by porcine elastase, as shown by a black arrow. Proteins were resolved on 12% NuPAGE Bis-Tris gels and visualized by Coomassie staining.

Figure 2 The relationship between Iripin-4 concentration and kobs, the observed second-order rate constant for the binding of human granzyme B to Iripin-4. kobs was determined using a para-nitroanilide substrate (Ac-IEPD-pNA).

3.3. Crystallization of Iripin-4 conformations

Based on initial screening, several successful crystallization conditions were found at both protein concentrations. Preliminary data collection indicated that optimization of crystallization was required, and therefore a protein concentration of 1.17 mg ml⁻¹ with a larger drop volume (2 µl) was used for further optimization. Crystals of native Iripin-4 grew under conditions consisting of 25%(w/v) PEG 3350, 0.1 M bis-Tris pH 5.5, 0.2 M ammonium acetate (Supplementary Fig. S3). The difference in the crystallization conditions for the cleaved conformation was the presence of 0.2 M sodium chloride instead of 0.2 M ammonium acetate as used in the crystallization of the native serpin (Supplementary Fig. S4). One of the most likely reasons for obtaining crystals of both conformations was the storage time of the protein sample. In the case of native Iripin-4 the protein sample was freshly obtained, whereas the crystals of cleaved Iripin-4 were formed from a protein sample that had been stored at −20°C for six months before optimization was initiated.

3.4. The partially relaxed state of the native conformation

The structure of native Iripin-4 was solved at 2.3 Å resolution and the crystal belonged to space group P3₁21 (No. 152), with unit-cell parameters a = 77.78, b = 77.78, c = 109.49 Å, α = 90.0, β = 90.0, γ = 120.0°. The crystal contained one molecule in the asymmetric unit with 51.33% solvent content and a Matthews coefficient of 2.53 ų Da⁻¹. Detailed refinement statistics are given in Table 1.

The overall structure of Iripin-4 consists of all 376 residues with a typical fold of native serpin secondary structure with the RCL exposed on top of β-sheet C as a protease `bait (substrate)' (Fig. 3; Huntington, 2011). The structure consists of nine α-helices and three β-sheets (A, B and C) arranged sequentially in the order αA–β1–αB–αC–αD–β2–αE–β3–αF–β4–β5–β6–β7–β8–αG–αH–β9–β10–αI–β11–β12–β13–β14 (Fig. 4b). β-Sheet A consists of five β-strands, β-sheet B of five β-strands and β-sheet C of four β-strands. The crystal structure has a tentatively assigned nickel cation (Ni²⁺) located between two molecules of Iripin-4 with crystallographic symmetry that interacts with His253 and Asp257 of both molecules (Supplementary Fig. S5). The presence of the Ni²⁺ ion, which was the most logical choice, is probably due to release of this ion during affinity purification.

Figure 3 The RCL of native Iripin-4. (a) Front view of the RCL (blue) showing the inserted residues into β-sheet A (cyan) and highly flexible residues of the hinge region of the RCL (purple). (b) View rotated by 50°.

Figure 4 Crystal structures of Iripin-4. (a) Native Iripin-4; the position of the RCL partially inserted into the breach region of β-sheet A is highlighted. (b) Side view of the native conformation. (c) Cleaved Iripin-4 displaying an additional β-strand in β-sheet A as a result of RCL insertion. (d) Side view of the cleaved conformation. Both crystal structures are displayed as cartoons; β-sheet A is in cyan and the RCL is in blue. The position of the P1 cleavage site is marked by a blue asterisk.

The crystal structure of Iripin-4 shows the typical partially relaxed state of native serpins, in which the RCL, which is located between β-sheet A and β-sheet C, is partially inserted into the breach region (Fig. 4a; the top part of β-sheet A; Whisstock & Bottomley). In contrast, the remaining part of the RCL is exposed (Fig. 4b). This state is likely to be the result of a crystallization process that trapped the serpin in one of two possible states, given the high flexibility of the RCL and the dynamic equilibrium between the partially inserted and the fully exposed native state. The six amino acids Ala326, Gly327, Thr328, Glu329, Ala330 and Ala331 are inserted between the third β-strand (s3A) and the fifth β-strand (s5A) of β-sheet A (Fig. 3). Three amino acids in the exposed part of the RCL, namely Thr334 and Gly335, which are part of the hinge region, and Leu343, were modelled into the structure despite the lower electron density observed here, indicating high flexibility of these two regions of the RCL (Fig. 3).

3.5. Structural transition to the cleaved conformation

The crystal of cleaved Iripin-4 diffracted to 2.00 Å resolution and belonged to space group P21 (No. 4), with unit-cell parameters a = 65.70, b = 138.41, c = 80.22 Å, α = 90.0, β = 107.7, γ = 90.0°. The crystal contains four molecules in the asymmetric unit, with a solvent content of 40.65% and a Matthews coefficient of 2.07 ų Da⁻¹. Detailed refinement statistics are given in Table 1.

The structure of the cleaved conformation consists of a typical mixed α–β secondary serpin structure with an N-terminal helical region and a C-terminal β-sheet fold. The overall structure consists of nine α-helices (A–I) and three β-sheets arranged sequentially in the order αA–β1–αB–αC–αD–β2–αE–β3–αF–β4–β5–β6–β7–β8–αG–αH–β9–β10–αI–β11–β12–β13–β14–β15 (Fig. 4c). The additional β12 strand results from the insertion of an RCL between the two β-strands of β-sheet A to form an additional strand called s4A; more precisely, it is located between the β11 strand (s5A) and the β4 strand (s3A) (Fig. 4d). A chloride ion (Cl⁻) is present in the model at the same position in all chains and interacts with Leu215 in the gate region. The gate region (Fig. 5) is responsible for stabilizing the RCL and preventing its premature insertion into β-sheet A (Irving et al., 2015).

Figure 5 Structural superposition of Iripin-4 conformations. The native Iripin-4 (cyan) with an exposed RCL (blue) is aligned with cleaved Iripin-4 (grey) containing a new additional β-strand (marine) labelled s4A. The three regions responsible for insertion of the RCL into β-sheet A are marked in circles (Khan et al., 2011). The hinge region is marked with an oval.

The four molecules in the final model differ in the number of modelled amino acids near the P1 site. Chains A and D contain 373 residues, with amino acids Ser342, Leu343 and Val344 missing, compared with chain B, which lacks Ser342 and Leu343 and consists of 374 residues, and chain C, in which only Ser342 is missing and which consists of 375 residues (out of a total of 376 residues). The missing residues were observed as an absence of electron density after the P1 residue Glu341 located at the end of the inserted RCL, now called s4A (Fig. 5). Alignment of the chains using PyMOL (Schrödinger) showed minor differences, which were confirmed by r.m.s.d. calculations between chain C (3014 atoms) and the other three chains: chain A, r.m.s.d. of 0.181 Å (2351 atoms); chain B, r.m.s.d. of 0.270 Å (2424 atoms); chain D, r.m.s.d. of 0.271 Å (2367 atoms).

3.6. Structural differences in native and cleaved Iripin-4

Structural comparison of the native and cleaved conformations of Iripin-4 shows that the β-strands in β-sheet A, together with α-helix D, α-helix E, α-helix F and α-helix I, shift to make space for the inserted RCL (Fig. 5). Since the cleaved conformation of serpin is structurally most related to the structure of the covalent complex between the serpin and the protease, it can be concluded that these structural changes will be the same or at least similar to the changes required for formation of the covalent complex. These observations also show how a significant conformational change is required for transition from the native to the cleaved state and also for protease inhibition. In addition to the structural changes in the shutter and hinge regions, a shift of the gate region, the loop part of β-sheet C between the s3C and s4C strands, is evident (Fig. 5).

The aforementioned breach, shutter, hinge and gate regions are involved in the incorporation of the RCL, specifically the hinge region of the RCL, into β-sheet A through coordinated movements in this complex process (Khan et al., 2011; Pearce et al., 2007). This process is referred to as the S-to-R transition, which involves a transition from a native stressed state to a more relaxed cleaved state with greater thermal stability (Whisstock et al., 2000). Comparison of the native and cleaved structures of Iripin-4 (Fig. 5) shows significant movement in the breach and shutter regions. The breach region, which is the starting point in the inhibitory mechanism, was slightly open compared with other solved structures of native serpins from I. ricinus (Fig. 6) and thus is more flexible than the shutter region, which is responsible for controlling or preventing undesired conformational change. The shutter region has more than 85% conserved residues and must open to incorporate the RCL, which helps to stabilize this transition (Irving et al., 2000). The polar contacts between s3A and s5A, more precisely Glu316–Asn167, Lys318–Asn167 and Thr319–Ile169, were disrupted by the insertion of the RCL compared with the cleaved structure with the additional β-s4A strand. For proper inhibition of the RCL by the protease, the transition must occur outside the gate region (Khan et al., 2011) and result in cleavage of the serpin instead of the formation of a covalent complex with the protease.

Figure 6 Superposition and alignment of I. ricinus native serpin conformations. (a) Superposition of the Iripin-4 (PDB entry 7zbf), Iripin-1 (PDB entry 7qtz; Chlastáková et al., 2023) and Iripin-8 (PDB entry 7pmu) crystal structures displayed as cartoons. The structure models are differentiated as follows: Iripin-4 is in blue, Iripin-1 is in red and Iripin-8 is in green. The structure of Iripin-1 was modified because of the absence of residues in its RCL, which are coloured salmon for emphasis. (b) A closer view of the RCL residues displayed as sticks; the colours are as described previously. (c) Structure-based sequence alignment of I. ricinus native serpins. The secondary-structure elements are depicted as arrows for β-strands and spirals for α-helices. The RCL is highlighted in a black box and the P1 and P1′ residues are marked according to the nomenclature of Schechter & Berger (1967).