Crystal structure of Echinococcus multilocularis cystatin B reveals a novel feature in classical stefins

Abstract

Alveolar echinococcosis (AE), caused by the metacestode larval of Echinococcus multilocularis, is one of the most lethal helminthic diseases in humans. Current treatment options, such as albendazole, are limited in their efficacy, highlighting the need for a deeper understanding of the parasite-host interaction to identify new therapeutic targets. One promising area of research involves helminth-derived cystatins, which are known to modulate host immune responses to facilitate parasite survival. A cystatin homologue from E. multilocularis (EmCystatin-B) was identified and analyzed. EmCystatin-B was cloned, expressed and purified. Its expression patterns were evaluated by western blot, qPCR and Immunohistochemistry The EmCystatin-B structure was solved by X-ray crystallography. EmCystatin-B was expressed in the mature protoscoleces as well as in the cytosol and nucleus of the metacestode vesicles. Moverover, EmCystatin-B adopts a conserved typical cystatin fold, but also exhibits unique structural features. Notably, a novel feature characterized by two intermolecular disulfide bridges between Cys4 in a EmCystatin-B molecular and Cys76 in adjacent molecule was discovered. Further investigation demonstrated this distinctive feature appears to be involved in the oligomerization of EmCystatin-B, facilitating a monomer-dimer-tetramer assembly pathway. The crystal structure of EmCystatin-B reveals a novel feature in classical stefins, and provides species-specific insights into the sequence, structure, and functional characteristics of EmCystatin-B.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-35060-0.

Introduction

Alveolar echinococcosis (AE) is a neglected parasitic zoonosis caused by infection with the metacestode larval of the tapeworm Echinococcus multilocularis¹. It severely threatens the health of human and livestock mainly spreading in regions of the Northern Hemisphere, such as Central Asia and Tibetan plateau of Western China, Central and Eastern Europe, and North America^2,3. It is estimated that the annual incidence of AE ranges from 0.03 to 1.2 cases/100 000 population, while the mortality rate in untreated or insufficiently treated patients is > 90% within 10–15 years after diagnosed, indicating that AE remains a serious cosmopolitan public health issue^4–6.

Humans get infected by accidentally ingesting food and water contaminated with infective eggs, which develop into cyst-like metacestode vesicles in the liver⁷. The tumor-like infiltrative growth and metastasis of the vesicles lead to massive lesions mainly in the liver and other host organs adjacent to the liver (for example, the spleen) or distant locations (such as the lungs, or the brain)^8,9. Currently, early diagnosis and radical surgery followed by prolonged anti-infective prophylaxis with mebendazole or albendazole is the most effective strategy for treatment and control of AE^10,11. However, several critical problems including lower early diagnostic efficiency, poor cure rates and certain side effects of these two anthelminthic drugs remain to be solved^11,12. Therefore, better understanding of the parasite-host interaction contribute to developing alternative effective approaches for reducing this disease burden.

The cystatin superfamily (MEROPS database ID: I25) is a group of functionally and evolutionarily related proteins, which are usually non-selective, reversible, competitive, and tight-binding inhibitors of cysteine proteinases including the C1A papain-like and C13 legumain-like families¹³. According to their sequence, spatial structure and physiological roles, it could be classified into three different subfamilies: stefins (type I cystatins or I25A), cystatins (type II cystatins or I25B), and kininogens (type III cystatins or I25C)¹³. Stefins including cystatins A and B are generally intracellular small single-domain proteins (~ 11 kDa) devoid of carbohydrate content and disulfide bonds^13,14. Cystatins including cystatins C, D, E/M, F, G, S, SA, and SN are extracellular single-chain proteins (13–14 kDa) containing a signal peptide and two disulfide bridges but lacking any carbohydrate moiety^13,14. Kininogens are usually extracellular secreted multi-domain glycoproteins (60–120 kDa) containing disulfide linkages and I25B repeats^13,14. All the members have the typical cystatin fold characterized by a four or five stranded anti-parallel β-sheet wrapped around one central N-terminal α-helix^15,16. Moreover, three regions mediated the interaction between cystatin and its target cysteine proteinase are conserved, which includes the N-terminal segment (G-X motif), two β-hairpin loops L1 (Q-X-V-X-G motif) and L2 (aromatic Trp or His residue)^13–16.

Cystatins have been reported to be involved in a variety of physiological and pathological processes, but their main biological function is to protect organism from being damaged by their endogenous cysteine cathepsins and other papain-like enzymes^13–15, as well as to function as an effective defense mechanism against cysteine proteases secreted from the invading pathogens^17,18. Currently, many helminth cystatins have been identified and widely demonstrated to have the ability to modulate host’s immune responses via several different mechanisms¹⁹, including inhibiting antigen processing and presentation in antigen-presenting cells (APC)^20,21, up-regulating nitric oxide (NO) production²², suppressing T-cell proliferation²³, inducing anti-inflammatory cytokine responses^24,25, and interfering with pattern recognition receptors processing²⁵. However, to date, no E. multilocularis derived member of the cystatin superfamily has yet been identified and structurally characterized.

In this study, a cystatin homologue has been identified and characterized from the cestode E. multilocularis. The sequence analysis revealed that it belongs to the stefin subfamily, and its sequence identity and similarity to Homo sapiens Cystatin-B (HsCystatin-B) are higher than HsCystatin-A, hence designated here as EmCystatin-B. The mRNA level of EmCystatin-B was much higher in the mature protoscoleces than in the metacestode vesicles, and EmCystatin-B was localized in the cytosol and nucleus of the metacestode vesicles as well as in the proliferating germinative cells of E. multilocularis. Next, its crystal structure was also determined, and the results showed that EmCystatin-B has the conserved typical cystatin fold with a three-stranded anti-parallel β-sheet (β1-β3) wrapped around a central five-turn central N-terminal α-helix (α1). Moreover, it has several significant structural differences with HsCystatin-B, especially in the flexible loop regions. Interestingly, a distinctive feature characterized by two intermolecular disulfide bridges between Cys4 in a EmCystatin-B molecular and Cys76 in adjacent molecule was discovered, and it has never been described in other cystatins. Further analysis indicated that the EmCystatin-B dimer forms possibly via domain swapping or disulfide bridges. Meanwhile, our data showed EmCystatin-B tetramer was maintained by four intermolecular disulfide bridges, which is a distinctive mechanism from the common hand shaking mechanism reported in previous studies. This work sheds a species-specific light on sequence, structural and functional characteristics of EmCystatin-B, enriches the knowledge on the cystatin superfamily and help better understanding the E. multilocularis-host interaction.

Methods

Sequence analysis and identification of EmCystatin-B

The amino acid sequence of EmCystatin-B was obtained from the NCBI database (GenBank accession no. CDI97789.1). Sequences were analyzed using the BLAST searches (http://www.ncbi.nlm.nih.gov/BLAST). Physical and chemical parameters of EmCystatin-B were predicted and analyzed by the ExPASy ProtParam tool (http://web.expasy.org/protparam). The SignalP 6.0 server (https://services.healthtech.dtu.dk/services/SignalP-6.0/) was used to check for the potential presence or absence of signal peptides. Sequences were aligned in MAFFT version 7.490 implemented in Geneious Prime version 2023.1.2 using the L-INS-i method, with settings for BLOSUM62 scoring matrix, gap opening penalty of 1.53 and offset value of 0.14. Evolutionary analysis was conducted by maximum likelihood method in MEGA11.0 program, the evolutionary history was inferred by using the maximum likelihood method and JTT matrix-based model. The selected species, protein and accession numbers of Cystatin-B used in the phylogenetic tree analysis are listed in Supplementary Table S1.

Parasite in vitro cultivation

In vitro cultivation of metacestode vesicles and protoscoleces was performed using host cell conditioned medium according to a previously established method^26–28. Briefly, the metacestode vesicles were collected from an infected Kunming mouse and subsequently cultivated in DMEM medium, supplemented with 10% FBS, in a culture flask covered with host feeder Hela cells. The cultivated hydatid tissue was collected and the protoscoleces (PSCs) were isolated by filtering the homogenised hydatid tissue through a 70 μm cell strainer, then washed three times in phosphate-buffered saline (PBS). The PSCs obtained by isolation and cultivated E. multilocularis metacestode vesicles were subsequently employed in the experimental analyses.

mRNA expression analysis of EmCystatin-B

The total RNA were extracted from cultivated E. multilocularis metacestode vesicles and isolated PSCs using the RNeasy Mini Kit (Qiagen). The total cDNA was reverse transcribed using the Evo M-MLV Reverse Transcription Kit with gDNA Clean (Accurate Biology). To determine the mRNA transcription level of EmCystatin-B in E. multilocularis metacestode vesicles and PSCs, real-time quantitative PCR (qPCR) was conducted using Hieff TM qPCR SYBR^® Green. The Master Mix (Yeasen Biotechnology) utilised EmCystatin-B specific primers (Supplementary Table S2). The housekeeping gene Emelp (E. multilocularis ERM-like protein, GenBank accession number: AJ012663) was employed as an internal control, with the specific primers (Supplementary Table S2)²⁹.

Molecular cloning of EmCystatin-B

Total RNA was extracted from in vitro cultivated E. multilocularis metacestode vesicles³⁰ by using the RNEasy kit (Qiagen, Germany) according to the manufacturer’s instructions. The purified total RNA was used for cDNA synthesis by using PrimeScript II 1st Strand cDNA Synthesis Kit (TAKARA, Shiga, Japan) according to the manufacturer’s instructions. The complete coding sequence of EmCystatin-B was amplified by high-fidelity polymerase chain reaction (PCR) with specific primers (Supplementary Table S2) containing NdeI and XhoI restriction sites respectively. The PCR products were digested and purified, then cloned into the bacterial expression vector pET-22b (Novagen, Madison, USA) (EmCystatin-B construct). The mutants including C4S, P75S, C76S, C4/76S were generated using the QuikChange method, among them, two PCRs were conducted on C4/76S. The primer sequences for site-directed mutagenesis PCR are listed in Supplementary Table S2.

Expression and purification of recombinant EmCystatin-B and size-exclusion chromatography assays

EmCystatin-B and mutant constructs were transformed into Escherichia coli BL21(DE3) competent cells (Novagen) for protein expression. The bacterial cultures were grown to an OD₆₀₀ of 0.6, after which 0.4 mM isopropyl-β-d-thiogalactoside (IPTG) was added for induction. Then, the cultures were incubated at 16 °C for 12–16 h. Bacteria were subsequently harvested. For the purification of these proteins, the bacteria were lysed by ultrasonication in lysis buffer (20 mM Tris-HCl, 300 mM NaCl, pH 8.0). The lysate was centrifuged at 16,000 g for 30 min to remove cell debris. The supernatant was loaded onto a Ni²⁺ NTA-agarose column (GE Healthcare) and eluted with lysis buffer containing 250 mM imidazole. A Superdex 200 gel filtration column (GE Healthcare) was subsequently preequilibrated with a buffer containing 20 mM Tris-HCl, 300 mM NaCl (pH 8.0), and then the protein was loaded onto the column and the fractions containing proteins were collected. Size-exclusion chromatography assays (SEC) were performed using a Superdex 200 Increase 10/300 gel filtration column (GE Healthcare) at a flow rate of 0.8 ml/min with absorbance monitored at 280 nm. The purified proteins were separated and identified by 15% reducing (+β-Me) SDS-PAGE gels staining with coomassie blue.

Non-reduced SDS-PAGE assay

Non-reduced samples were lysed and collected using 2 × SDS loading buffer (0.1 M Tris, 4% SDS, 20% glycerol and 0.02% bromophenol blue) without β-mercaptoethanol. Next, electrophoresis was conducted at a constant voltage of 150 V until completion. The gel was stained with Coomassie Brilliant Blue R250. The non-reduced SDS-PAGE marker was purchased from Real-Times (Beijing) Biotechnology Co.,Ltd with the product number RTD6136.

Enzyme activity assay

The inhibitory effect of EmCystatin-B on human Cathepsin B (CTSB) was evaluated using an enzyme activity assay. Human CTSB was obtained from Sigma-Aldrich, and the experimental procedure was adapted from the CTSB activity assay protocol provided by Merck. Briefly, enzymatic activity was measured fluorometrically by monitoring the release of 7-amino-4-methylcoumarin (AMC) from the substrate Z-Arg-Arg-7-amido-4-methylcoumarin (Sigma-Aldrich, C5429). The assay buffer consisted of 352 mM potassium phosphate, 48 mM sodium phosphate, and 4.0 mM EDTA, adjusted to pH 6.0 at 40 °C. Human CTSB (50 nM) was incubated with varying concentrations (0-100 nM) of EmCystatin-B or the control inhibitor E-64 (Sigma). The increase in fluorescence intensity was recorded over 5 min at excitation and emission wavelengths of 348 nm and 440 nm, respectively. Data were analyzed using GraphPad Prism (Version 9.5.0; GraphPad Software Inc., San Diego, USA).

Preparation of polyclonal antibody

Recombinant EmCystatin-B protein (40 µg) was emulsified in 100 µL of MnJ(β) colloidal manganese adjuvant (Mnstarterbio Co., Ltd., China) per mouse. Female Kunming mice were subcutaneously immunized with this formulation according to the following schedule: a total of four immunizations were administered, with the first two doses given 7–14 days apart, followed by subsequent doses at 7-day intervals. Blood samples were collected one week after the final immunization, and antisera were separated. Antigen-specific antibodies were subsequently purified from the antisera by affinity chromatography using Protein A/G resin (Thermo Fisher Scientific).

Western blot analysis

Protoscoleces extracted from Kunming mice infected with E. multilocularis were processed using ultrasonic extraction in RIPA strong lysis buffer (Servicebio, China) containing protease and phosphatase inhibitors (Selleck, China). The protein concentration was then quantified using the BCA method (Beyotime Biotechnology, China). The extracted protein was boiled at 100 °C for 10 min. A total of 40 µg of protein was loaded onto a 15% acrylamide gel (Vazyme, China) and transferred to a PVDF membrane. The membrane was blocked with 5% bovine serum albumin (BSA) (Biofroxx, Germany) at room temperature for 2 h. Samples were incubated overnight at 4 °C with EmCystatin-B antibody (1:2000 dilution). The membrane was washed three times with TBST for 5 min each. An enzyme-linked immunoassay was performed using anti-mouse IgG-HRP (Invitrogen, 1:10,000) at room temperature for 1 h, followed by three additional TBST washes (5 min each). Detection was achieved using the ECL method (Millipore) according to the manufacturer’s instructions. The membrane was stripped using an antibody stripping solution (Thermo Fisher Scientific) following the manufacturer’s protocol and reanalyzed using a Bio-Rad ChemiDoc Touch instrument. The same procedure was applied for GAPDH protein detection, with the exception that the GAPDH antibody (Proteintech) was diluted at a ratio of 1:5,000.

Immunofluoresence of EmCystatin-B in the E. multilocularis metacestode vesicles

In vitro cultivated metacestode vesicles were exposed to 50 µM EdU in HM medium for 4 h^30,31. The EdU-containing medium was then removed, and the vesicles were fixed with 4% paraformaldehyde (PFA) at room temperature for 20 min. This was followed by three 5-minute washes with PBS. The vesicles were permeabilized with 1% Triton X-100 in PBS for 10 min and then washed three times with PBS (5 min each). Blocking was performed using a 5% BSA solution at room temperature for 1 h. EmCystatin-B antibody (1:1000) was added and incubated at 4 °C for 8 h. The vesicles were washed three times with PBS (5 min each). Detection of EmCystatin-B was achieved using Alexa 488-conjugated anti-mouse IgG (Cell Signaling Technology, 1:1500) at room temperature for 1 h, followed by three 5-minute PBS washes. EdU detection was performed using Click-iT EdU Alexa Fluor 555 (Life Technologies, 1:1500) according to the manufacturer’s instructions. DNA was counterstained with 4’,6-diamidino-2-phenylindole (DAPI) after EdU detection. Fluorescence images were captured using a Nikon 80i fluorescence microscope.

Crystallization of rEmCystatin-B

Purified rEmCystatin-B was collected and concentrated to ~ 10 mg/ml with a Centricon centrifugal filter unit (molecular weight cutoff [MWCO], 10,000) for crystallization. The crystallization of rEmCystatin-B was carried out by sitting-drop vapor diffusion at 16 °C. The reservoir solution consisted of 0.2 M magnesium acetate tetrahydrate, 20% PEG3350, pH 7.9. About 0.6 ml of the protein solution was mixed with an equal volume of the reservoir solution. The rhombus-shaped rEmCystatin-B crystals were observed after 24–48 h (Fig. S1).

Data collection, structure determination and refinement

The diffraction data were collected at 100 K under the synchrotron radiation at beamline BL02U1 of the Shanghai Synchrotron Radiation Facility (SSRF). The data sets were integrated and scaled with the XDS package. The structure of the EmCystatin-B was determined by molecular replacement with the structure of the Clonorchis sinensis Stefin-1 (Protein Data Bank accession code 5ZC1)³² as the initial search model with the program Phaser. The programs Refmac5 and Coot9 in the CCP4 suite were used for the refinement and model building. Ramachandran plots were generated with Coot9. The statistics for data processing and structure refinement are shown in Table 1.

Table 1.

Crystallographic data collection and refinement statistics.

	Structure of EmCystatin-B
Data collection
Wavelength (Å)	0.97915
Space group	P21
Cell dimensions
a, b, c (Å)	72.72, 66.11, 92.60
α, β, γ (°)	90.0, 103.42, 90.0
Resolution (Å)	50-1.99
Rsym or Rmerge	0.104
I/σI	2.5 (at 1.99Å)
Completeness(%)	98.9(89.1)
Redundancy	5.9(3.4)
Refinement
Resolution (Å)	50-1.99
No. reflections	57,418
Rwork/Rfree	0.211/ 0.241
No. atoms
Protein	6286
Water	361
B-factors
Protein	45.03
Water	43.36
R.m.s. deviations
Bond lengths (Å)	0.0073
Bond angles (°)	1.7513
Ramachandran analysis
Favored region (%)	98.20
Allowed region (%)	0.64
Outlier (%)	1.16

Values in parentheses are for highest-resolution shell.

Molecular modeling for EmCystatin-B binding to human cathepsin B

The crystal structures of both human cathepsin B (accession number 3K9M)³³ and EmCystatin-B (accession number 9UOZ ) were obtained from the Protein Data Bank The starting structures used for local docking are from the protein docking prediction server ZDOCK (https://zdock.umassmed.edu/). To prepare the structures for docking, the unbound structures were superimposed over the bound complex and the resulting superposed structure was used as the starting structure for local docking in Rosetta. The receptor and ligand proteins were docked with the default settings. The presentations were made with the PyMOL Molecular Graphics System, version 2.2.3 (Schrödinger LLC).

Results

Sequence analysis reveals EmCystatin-B belongs to the stefin subfamily

The full-length cDNA of EmCystatin-B contains an open reading frame (ORF) composed of 297 bp coding for a 98-amino-acid protein (GenBank accession no. CDI97789.1) with a putative molecular mass of 10.9 kDa and 6.57 theoretical isoelectric point (pI) using the ExPASy program. Next, the multiple sequence alignments (Fig. 1) revealed that EmCystatin-B contains the three conserved regions important for the interaction with its cysteine proteases targets^16,33,34, such as the N-terminal G-G motif (N), the Q-I-V-N-G motif within the first β-hairpin loop (L1), and the L-P motif in stefins within the second β-hairpin loop (L2), in contrast to the P-W motif in cystatins. Additionally, EmCystatin-B also carries a relatively conserved C-terminal D-X-L-X-Y-F motif, which is typical for stefins^14,35. On the other hand, EmCystatin-B does not possess the large sequence insertion in the central part of the molecule, which is always found between strands β2 and β3 in cystatins, and it also has no signal peptide sequence, which is a hallmark of secreted cystatins^13,14,37. Phylogenetic analysis showed that EmCystatin-B is closely related to E. granulosus Cystatin-B (Fig. S2). Therefore, these results indicated that EmCystatin-B belongs to the stefin subfamily. Additionally, the results of BLAST analyses from the NCBI database showed that its sequence identity and similarity to HsCystatin-B are higher than HsCystatin-A (Supplementary Table S3). Hence, it was designated here as EmCystatin-B.

Fig. 1.

Multiple amino acid sequence alignment of EmCystatin-B with representative members of the cystatin superfamily. EmCystatin-B is compared with representative members including HsCystatin-A and HsCystatin-B (type I family), HsCystatin-D and HsCystatin-F (type II family). The secondary structure elements of EmCystatin-B are depicted in cyan for α-helix and magenta for β-strands. Residues identical with those of EmCystatin-B are shaded gray. The three conserved regions including G-X motif (N), two β-hairpin loops Q-X-V-X-G motif (L1) and L-P or P-W motif (L2) forming the reactive center are boxed in red, the critical consensus positions are highlighted in red for EmCystatin-B. The conserved C-terminal D-X-L-X-Y-F motif in classical stefins is boxed in green. The disulfide bridges are labeled by the connecting black lines and indicated with D1 and D2 for cystatins D/F. Signal peptide is indicated by SigPep. GenBank accessions: EmCystatin-B, CDI97789.1; HsCystatin-A, AAH10379; HsCystatin-B, AAH10532; HsCystatin-D, AAH62678; HsCystatin-F, CAG46658.

Expression of EmCystatin-B in E. multilocularis larvae

Cystatin-B has been shown to be an endogenous cysteine cathepsin inhibitor abundant in the cytosol, mitochondria and nucleus of various cell types³⁶. We then analyzed EmCystatin-B expression in the mature protoscoleces and in the metacestode vesicles which have a large population of proliferating germinative cells. As shown in Fig. 2A, we found that the mRNA level of EmCystatin-B was much higher in the mature protoscoleces than in the metacestode vesicles. Indirect immunofluorescence microscopy on the metacestode vesicles using polyclonal antibodies raised against recombinant EmCystatin-B (Fig. S3) demonstrated that EmCystatin-B was localized in the cytosol and nucleus of the metacestode vesicles, and also expressed in the proliferating germinative cells of E. multilocularis (Fig. 2B).

Fig. 2.

Expression of EmCystatin-B in E. multilocularis larvae. (A) Relative mRNA levels of EmCystatin-B in the mature protoscoleces (mp) and metacestode vesicles (mv). Data are shown as means ± the standard deviations (SD). ****, P < 0.001. (B) Immunofluoresence analysis of EmCystatin-B in the E. multilocularis metacestode vesicles. Immunofluorescence was performed using anti-EmCystatin-B antibody (green) and DAPI for nuclei staining (blue). The proliferating germinative cells labeled with EdU (red). Arrow indicates the nuclei labeled by merged fluorescence. Scale bar, 10 μm.

The overall structure of EmCystatin-B

The evidences from size-exclusion chromatography (SEC) and SDS-PAGE showed that purified rEmCystatin-B was successfully obtained, while containing the higher-order aggregates in the solution (Fig. 3A,B). The crystal structure of rEmCystatin-B was determined by molecular replacement and refined to 1.99 Å resolution with an R_free of 0.241 (Table 1). The crystallographic data indicated that rEmCystatin-B was crystallized in a monoclinic space group P21, and there are eight protomers in the asymmetric unit (Fig. S4) (Table 1). Almost all the residues of rEmCystatin-B could be modeled into a well-defined electron density map except for the first two N-terminal residues (Met1-Pro2), thereby the final model contained 102 residues (Met3-His104). In the crystal lattice, the structure of EmCystatin-B reveals an oligomeric assembly stabilized by both three-dimensional domain swapping and intermolecular disulfide bridges. The canonical cystatin fold, characterized by a five-turn N-terminal α-helix (α1) packed against a four-stranded antiparallel β-sheet, is formed across two polypeptide chains. In this domain-swapped dimer, one protomer contributes its α1-β1-L1 segment to complete the fold of its adjacent partner, and vice versa (Fig. 3C). Consequently, a self-contained, monomeric cystatin fold is not observed in the crystal structure. Each monomeric cystatin-like conformations adopted a typical cystatin fold similar to that of human homologues stefins A and B^16,33,34, comprising a three-stranded anti-parallel β-sheet (β1-β3) wrapped around a central five-turn central N-terminal α-helix (α1) (Fig. 3D). The structural elements that typically constitute the tripartite wedge-shaped reactive center in canonical monomeric cystatins—the N-terminal segment and the L1 and L2 loops—are present in the EmCystatin-B sequence but are reconfigured in the crystal structure due to domain swapping (Fig. 3D). In the swapped dimer, these elements are contributed by different protomers (Fig. 3D), and their spatial arrangement is not compatible with the formation of a continuous inhibitory wedge against cysteine proteases as seen in monomeric inhibitors^16,33,34. This suggests that the oligomeric state may regulate its inhibitory function. Additionally, although the two cysteine residues (Cys4 and Cys76) in EmCystatin-B molecule were close in space, no intramolecular disulfide bonds or carbohydrate side-chains were discovered in monomeric EmCystatin-B (Fig. 3D). These results further confirmed that EmCystatin-B is a type I cystatin.

Fig. 3.

Expression, purification, and crystal structure of rEmCystatin-B. (A) Purification of rEmCystatin-B by size-exclusion chromatography (SEC). The peaks (peak1 and peak2) are labeled. (B) SDS-PAGE analysis of the purified rEmCystatin-B. M: protein size marker; Lanes 1–2: purified rEmCystatin-B protein for peak 1; Lanes 3–4: purified rEmCystatin-B protein for peak 2. (C) The overall structure of the EmCystatin-B tetramer. The two cysteine residues (C4 and C76) are labeled in rainbow, and the C termini and α1 helix are indicated. (D) The ribbon diagram of rEmCystatin-B is shown in colored by secondary structure elements (α1 helix in cyan, β1–β3 strands in magenta). The N and C termini (labeled N and C), and two β-hairpin loops (L1 and L2) are indicated. The predicted critical residues are highlighted in red, and the two cysteine residues (C4 and C76) are labeled in yellow. The hairpin loops L1 and L2 and the N terminus are involved in the binding to cysteine cathepsins.

Structural differences between EmCystatin-B and HsCystatin-B

The full lengths of both EmCystatin-B and HsCystatin-B encode 98 amino acids, and they all bear upon a similar typical cystatin fold characterized by an anti-parallel β-sheet wrapped around a central N-terminal α-helix (Fig. 4A). However, in contrast to five β strands in one HsCystatin-B molecule, EmCystatin-B only contains three β strands in one molecule (Fig. 4A). Next, when these two solved structures were superimposed, the mean square root mean square deviation (RMSD) indicated a considerably large difference of 3.039 Å (Fig. 4A). The results showed that the main structural differences between the two were focused on the α1 (region I) and loop regions (II-IV) among α1-β1, β1-β2, and β2-β3 (L2) of EmCystatin-B (Fig. 4). In region I, the differences among residues 20 L/18H, V24/Q22, K26/R24 and Q31/E29 make the α1 a little deflection (Fig. 4B). As the most flexible regions of the protein, loops are well-reported to be crucial for binding with their ligands and can always adopt multiple conformations³⁸. Here, there are significant structural differences in the three loop regions, such as residues L33/E31, R35/KE33, P37/F35, and E39/V37 in region II, residues H61/G60, G63/E62, V88/T87, V91/A90, and P94/E94 in region III, residues P75/P74, C76/H75 in region IV (L2) (Fig. 4B). Especially, the P75 and C76 might play an important role in the oligomerization of EmCystatin-B. And the results obtained from molecular docking of EmCystatin-B and human cathepsin B showed that the binding mode of EmCystatin-B is totally different from that of HsCystatin-A with cathepsin B (Fig. S5). It was suggested that the mechanisms involved in the binding of EmCystatin-B and HsCystatin-B with their cysteine proteases targets are likely to be different.

Fig. 4.

Analysis of structural differences between EmCystatin-B and HsCystatin-B. (A) Superimposition of ribbon diagrams of EmCystatin-B (magenta) and HsCystatin-B (green). The main structural differences between the two were focused on the α1 (region I) and loop regions (II-IV) among α1-β1, β1-β2, and β2-β3 (L2) of EmCystatin-B. The regions I-IV are boxed in red. (B) The boxed regions (I-IV) are enlarged. The large conformational changes are drawn as stick representations. In region I, Q19, L20, V24, K26 and Q31 of EmCystatin-B are indicated. In region II, L33, R35, P37 and E39 are labeled. In region III, H61, D62, G63, V88, V91 and P94 are indicated. In region IV, P75 and C76 are labeled.

The EmCystatin-B dimer forms via domain swapping or disulfide bridges

The three-dimensional domain swapping mechanism has been reported to be a common feature of homodimerization by members from the cystatin superfamily^39,40. Here, in this solved structure of EmCystatin-B, we observed that two adjacent molecules could arrange in a domain-swapped dimer, consisting of two α-helixes and two four-stranded anti-parallel β-sheets, in which three β-strands come from one subunit and the fourth β-strand comes from the other monomer (Fig. 5A,B). In contrast to other reported cystatin homodimer, the N-terminal region α1-β1-L1 of one EmCystatin-B molecule swapped out and integrated at the equivalent position of the adjacent molecule, then formed a four-stranded anti-parallel β-sheet, and vice versa (Fig. 5B). The two monomeric β1-L1 and β1’-L1’ aligned in an anti-parallel fashion to form a hinge region, which could bridge the two monomers (Fig. 5B,C). In the hinge region, several hydrogen bonds formed by L1 and L1’ contributed to maintaining a more stable conformation of this domain-swapped dimer (Fig. 5C).

Fig. 5.

The formation of EmCystatin-B dimer could be mediated by domain swapping or disulfide bridges. (A) The two monomers of EmCystatin-B are shown in gray and green. (B) The domain-swapped dimer of EmCystatin-B. The N-terminal region α1-β1-L1 of one EmCystatin-B molecule swapped out and integrated at the equivalent position of the adjacent molecule. The two monomeric β1-L1 and β1’-L1’ aligned in an anti-parallel fashion to form a flexible hinge region (boxed in red), which could bridge the two monomers. (C) Top view of the flexible hinge region formed by the L1 loops. Several hydrogen bonds formed by L1 and L1’ are indicated. (D) The two monomers of EmCystatin-B are shown in cyan and green. The two cysteine residues (C4 and C76) are labeled in yellow. (E) The disulfide-bridged dimer of EmCystatin-B. There were two intermolecular disulfide bridges (boxed in red) formed by the C4 in the N-terminal of a EmCystatin-B molecule and the C76 in L2 of the adjacent molecule. (F) The electron density of two intermolecular disulfide bridges in EmCystatin-B dimer. The electron density of four cysteine residues (C4, C4’, C76 and C76’) are shown.

As a typical feature of type II cystatins, two conserved intramolecular disulfide bridges have been identified in nearly all basal metazoans^14,37. However, it has been well-demonstrated that classical stefins lacked intramolecular disulfide linkages due to no cysteine residues or unpaired cysteines with a free thiol¹⁵. Here, the EmCystatin-B amino acid sequence contains three cysteine residues including Cys4, Cys65 and Cys76, but no intramolecular disulfide bonds were found in the 3D structure of EmCystatin-B (Figs. 3C and 5D). However, further structural analysis showed that there were two intermolecular disulfide bridges formed by the Cys4 in the N-terminal of a EmCystatin-B molecule and the Cys76 in L2 of the adjacent molecule (Fig. 5E,F), which is a novel distinct feature of EmCystatin-B and has never been described in other stefins. Therefore, the adjacent two EmCystatin-B monomers were bridged by these two disulfide bonds, which could provide enough binding energy for the formation of EmCystatin-B dimer, and also make its conformation more stable.

In general, the formation of EmCystatin-B dimer is likely to be mediated by two different mechanisms including domain swapping and disulfide bridges. However, our results could not provide enough evidences to confirm which one is dominated, and it also needs more further investigations in the future study.

The EmCystatin-B tetramer formation adopts a distinctive mechanism

The formation of tetramers of human cystatin-B has been reported to be closely related with extensive intermolecular contacts, termed hand shaking, through which three-dimensional domain-swapped dimers become entwined^13,41. And this process is mainly mediated by concurrent trans to cis isomerization of proline 74^41,42, which is widely conserved throughout the stefins and cystatins. However, our results showed that EmCystatin-B oligomerized to a tetramer as a consequence of the formation of four intermolecular disulfide bridges between Cys4 in one molecule and Cys76 in adjacent molecule (Fig. 6A). Although the crucial proline residue is conserved in EmCystatin-B, we did not observe its isomerization and thereby the hand shaking mechanism (Fig. 6B,C). Therefore, the formation and stability of EmCystatin-B tetramer was maintained mainly by intermolecular disulfide bridges, which is a distinctive mechanism from previous studies.

Fig. 6.

The formation of EmCystatin-B tetramer adopts a distinctive mechanism. (A) EmCystatin-B tetramer was maintained by intermolecular disulfide bridges. The electron density of four intermolecular disulfide bridges formed by C4 and C76 in EmCystatin-B tetramer are shown. (B) Ribbon representation of the tetramer of EmCystatin-B. The intermolecular disulfide bridges (colored in yellow) are boxed in red. (C) The HsCystatin-B tetramer forms via hand shaking mechanism. The trans to cis isomerization of proline 74 leads to HsCystatin-B tetramer formation. (D) SDS-PAGE analysis of the purified rEmCystatin-B and its mutants including P75S, C4S, C76S and C4/76S. M: protein size marker. SDS-PAGE analysis of purified proteins under reducing (+β-Me) conditions. (E) Non-reducing SDS-PAGE analysis of the purified rEmCystatin-B and its mutants including P75S, C4S, C76S and C4/76S. M: protein size marker. SDS-PAGE analysis of purified proteins under non-reducing (-β-Me) conditions. (F) Size-exclusion chromatography analysis of the purified rEmCystatin-B and its mutants including P75S, C4S, C76S and C4/76S. The calibrated chromatograms with marked molecular weights were shown. WT: wild-type rEmCystatin-B protein; P75S: Pro75 was mutated to Ser75; C4S: Cys4 was mutated to Ser4; C76S: Cys76 was mutated to Ser76; C4/76S: both Cys4 and Cys76 were mutated to serine. (G) The inhibitory activity of EmCystatin-B on human cathepsin B. E-64: the inhibitor of cathepsin B.

To further verify the mechanism of EmCystatin-B oligomerization, we designed four point mutations (P75S, C4S, C76S and C4/76S) based on the solved crystal structure of EmCystatin-B. The results of Native-PAGE showed that the P75S mutant almost did not change the aggregation state of the EmCystatin-B, while either the single mutation (C4S and C76S) or double mutation (C4/76S) of the cysteine in EmCystatin-B significantly reduced the aggregation degree of the protein (Fig. 6D,E), which was also confirmed by the size-exclusion chromatography assays (Fig. 6F). These results indicated that intermolecular disulfide bridges between Cys4 in one molecule and Cys76 in adjacent molecule play an indispensable role in the formation of EmCystatin-B protein oligomers, rather than the common hand shaking mechanism in human cystatin-B. Moreover, inhibition assays results showed that EmCystatin-B can significantly inhibit the activity of human cathepsin B (CTSB) in a concentration-dependent manner. The P75S mutant did not affect EmCystatin-B’s inhibitory activity on CTSB, while the disruption of disulfide bonds such as C4S, C76S and C4/76S mutant leads to a marked decrease in its inhibitory activity, indicating that the oligomerization of the protein is required for inhibiting the activity of CTSB.

Discussion

Parasite cystatins have been reported to function as an immunomodulatory molecule aiding in the establishment or persistence of successful parasitism in the hosts by suppressing host immune responses¹⁹. Recent investigations showed that they have the therapeutic potential for several immune-mediated disorders including inflammatory bowel disease (IBD)⁴³, allergy⁴⁴, rheumatoid arthritis⁴⁵, sepsis⁴⁶, and type-1 diabetes⁴⁷. As an evolutionarily conserved protein, orthologues of cystatins have been identified and functionally characterized across a wide range of living organisms, from protozoa to mammals, especially in several helminths^14,19. Herein, we identified a cystatin B homologue from the tapeworm E. multilocularis and determined its structural features and characteristics. This work not only provides insights into the E. multilocularis-host interaction and enriches the knowledge on the cystatin superfamily, but also identifies a novel helminth-derived immunoregulatory molecule, laying the foundation for its future functional evaluation and therapeutic development.

The sequence and structural features revealed that EmCystatin-B is a member of the type I subfamily of stefins (Figs. 1 and 3). EmCystatin-B does not present a typical signal peptide for secretion, indicating it is likely to be an intracellular cytoplasmic protein. Like all the members of cystatin family, the EmCystatin-B molecule bears a tripartite wedge-shaped reactive center formed by three highly conserved regions including the N-terminal G-G motif, the G-I-V-N-G motif within the loop L1, and the L-P motif within the loop L2, which would be functionally competent for cysteine protease binding^16,32,33. What’s more, the monomeric cystatin-like conformation molecule adopted a typical cystatin fold characterized by a three-stranded anti-parallel β-sheet (β1-β3) wrapped around a central five-turn central N-terminal α-helix (α1) (Fig. 3C). In contrast to HsCystatin-B, there are several significant structural differences on the α1 and flexible loop regions among α1-β1, β1-β2, and β2-β3 (L2) of EmCystatin-B, especially the conserved P75 and the key residue C76 (Fig. 4), which might play an important role in the maintenance of its stable conformation, indicating its species-specific structural characteristics. However, whether it has the similar biological activities and functions with most well-characterized members of the cystatin superfamily or not remains unclear and needs more further investigations in the future study.

Stefins are usually devoid of disulfide bridges, which are a typical feature of type II cystatins⁴⁸. Here, the crystallographic analysis showed that EmCystatin-B indeed contains no intramolecular disulfide bonds, but a distinct feature characterized by intermolecular disulfide bridges formed by the Cys4 in a EmCystatin-B molecule and the Cys76 in adjacent molecule was discovered in oligomerized EmCystatin-B including dimer and tetramer (Figs. 5 and 6), which differs from that of typical intracellular stefins. Due to their ability of stabilizing proteins in the oxidizing extracellular environment, disulfides are generally found in most of secreted proteins while rather rare in cytosolic proteins⁴⁹. Therefore, it reminds us that whether EmCystatin-B is a secreted or cytosolic protein still required to be confirmed in the further study. On the other hand, the potential physiological functions of disulfide bridges in EmCystatin-B also need to be clarified.

Similar to stefin A, HsCystatin-B has been reported to be able to dimerize, oligomerize and form amyloid fibrils under in vitro conditions⁵⁰. It has been reported that the cystatin dimer formation is mainly induced by the domain swapping mechanism^51,52. However, in this work, we discovered another novel mechanism mediated by intermolecular disulfide bridges, which could also be responsible for the dimerization of EmCystatin-B (Fig. 5D,F). That is to say, the EmCystatin-B dimer could have two different conformations. Furthermore, our results showed that the EmCystatin-B tetramer could be maintained by four intermolecular disulfide bridges, which is a distinct feature from the common hand shaking mechanism (Fig. 6). Enzyme activity inhibition experiments showed that the oligomeric EmCystatin-B also have an inhibitory effect on CTSB. There were significant differences in enzyme activity inhibition among different oligomers. This indicates that our structure is not caused by the special conditions of crystallization, but represents a physiologically relevant complex conformation (Fig. 6G). Nevertheless, we still have no idea that either the domain-swapped dimer or disulfide-bridged dimer oligomerizes to the EmCystatin-B tetramer, and also needs more further investigations to make it clear. To sum up, as shown in Fig. 7, EmCystatin-B could oligomerize through monomer-dimer-tetramer pathway, which is mainly mediated by the intermolecular disulfide bridges between Cys4 in one molecule and Cys76 in adjacent molecule. Meanwhile, this structural characteristic might be the reason why there were some higher-order aggregates of rEmCystatin-B in the elution buffer as shown in Fig. 3.

Fig. 7.

A proposed model for the oligomerization of EmCystatin-B. EmCystatin-B is likely to oligomerize through monomer-dimer-tetramer pathway. The monomeric EmCystatin-B oligomerizes to dimer via domain swapping or disulfide bridges. The EmCystatin-B tetramer could be maintained by four intermolecular disulfide bridges, which is a distinct feature from the common hand shaking mechanism in human cystatins.

Taken together, we have identified a novel stefin homologue in E. multilocularis and determined its crystal structure. These results indicated that EmCystatin-B was expressed in the mature protoscoleces as well as in the cytosol and nucleus of the metacestode vesicles. What’s more, EmCystatin-B not only has the highly conserved characteristics of the cystatin superfamily, but also presents several distinctive features from the reported cystatins, especially the formation of intermolecular disulfide bridges and their potential involvement in the oligomerization of EmCystatin-B. This work provides enough evidences for enriching the knowledge on cystatin superfamily and laying the structural foundation for the functions of EmCystatin-B in the E. multilocularis-host interaction.

Conclusions

The crystal structure of EmCystatin-B reveals a novel feature in classical stefins, and provides species-specific insights into the sequence, structure, and functional characteristics of EmCystatin-B. This work not only advances our understanding of the cystatin superfamily but also lays the structural foundation for the development of novel therapeutic strategies targeting cystatin-mediated immune modulation in AE.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

W.H.: Methodology, Investigation, Conceptualization, Writing-Original Draft; Z.C.: Investigation, Resources, Validation, Funding acquisition; Z.X.: Investigation, Data Curation, Visualization; S.Z.: Data Curation, Formal analysis; X.L.: Visualization, Formal analysis; N.D.: Software, Formal analysis; Z.D.: Formal analysis, Validation; Y.L.: Formal analysis, Investigation; W.L.: Validation, Formal analysis; H.J.: Resources, Formal analysis; X.S.: Software, Writing-Review & Editing, Funding acquisition; S.H.: Writing-Review & Editing, Project administration, Conceptualization, Funding acquisition. All authors reviewed the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Nos. 82102428 to SH, 32200450 to XS and 82272365 to ZC), and the Natural Science Foundation of Hunan Province (2022JJ40663 to SH).

Data availability

The atomic co-ordinates and structure factors of EmCystatin-B obtained by X-ray crystallography have been deposited in the Protein Data Bank (PDB) database with the accession code 9UOZ. A full wwPDB X-ray structure validation report for 9UOZ has been provided in Supplementary Material 6. The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

All animal experiments were conducted in strict accordance with China regulations on the protection of experimental animals and specifically approved by the Ethics Committee of Xiangya School of Basic Medical Science, Central South University (Permit Number: 2021-KT25). This study is reported in accordance with the ARRIVE guidelines (https://arriveguidelines.org).