Abstract
Approximate 70% of cobra venom is composed of cytotoxin (CTX), which is responsible for the dermonecrotic symptoms of cobra envenomation. However, CTX is generally low in immunogenicity, and the antivenom is ineffective in attenuating its in vivo toxicity. Furthermore, little is known about its epitope properties for empirical antivenom therapy. This study aimed to determine the epitope sequences of CTX using the immunoinformatic analyses and epitope-omics profiling. A conserved CTX was used in this study to determine its T-cell and B-cell epitope sequences using immunoinformatic tools and molecular docking simulation with different Human Leukocyte Antigens (HLAs). The potential T-cell and B-cell epitopes were 'KLVPLFY,' 'CPAGKNLCY,' 'MFMVSTPTK,' and 'DVCPKNSLL.' Molecular docking simulations disclosed that the HLA-B62 supertype exhibited the greatest binding affinity towards cobra venom cytotoxin. The namely L7, G18, K19, N20, M25, K33, V43, C44, K46, N47, and S48 of CTX exhibited prominent intermolecular interactions with HLA-B62. The multi-enzymatic-limited-digestion/liquid chromatography-mass spectrometry (MELD/LC–MS) also revealed three potential epitope sequences as 'LVPLFYK,' 'MFMVS,' and ‘TVPVKR’. From different epitope mapping approaches, we concluded four potential epitope sites of CTX as ‘KLVPLFYK’, ‘AGKNL’, ‘MFMVSTPKVPV’ and ‘DVCPKNSLL’. Site-directed mutagenesis of these epitopes confirmed their locations at the functional loops of CTX. These epitope sequences are crucial to CTX’s structural folding and cytotoxicity. The results concluded the epitopes that resided within the functional loops constituted potential targets to fabricate synthetic epitopes for CTX-targeted antivenom production.
Similar content being viewed by others
Introduction
The World Health Organization (WHO) has declared snakebite envenomation (SBE) as a category A Neglected Tropical Disease1. The cobra (Naja sp.) envenomation is unequivocally a feared snake species due to their cytotoxic venoms, leading causes of physical impairment that engenders lifetime handicaps in most survivors2. The cytotoxic envenomation is manifested as local dermonecrosis, which is a sequelae event due to cell death and subsequent loss-of-function injuries3,4,5,6. The venomics of Naja venom revealed that, the major venom toxins are phospholipase A2s (PLA2s), neurotoxins (NTXs) and cytotoxins (CTXs). PLA2s and CTXs constitute average 20% and 70% of the Naja venoms’ dry weight, respectively7,8,9,10. It is noteworthy that, synergisms of PLA2s and CTXs have been summarized to cause downstream cytotoxic effects upon envenomation11. Nevertheless, CTXs play major roles in venom-induced dermonecrosis due to their abundances in Naja venom12.
Cytotoxins are highly basic and amphipathic polypeptides13. They are 7 kDa with approximately 59–62 amino acid residues14. They appear in three-finger loops from five antiparallel β-sheets13. These β-sheets make up the hydrophobic core of the toxins, which are stabilized by four conserved disulfide bridges15. The hydrophobic core is also flanked by basic lysine and arginine residues responsible for cytotoxicity16. It allows the toxins to interact with the cell membranes and contributes to the formation of membrane pores, resulting in subsequent cell death cascade events17,18,19,20. Thus, it is said that the CTX’s loops are responsible for dermonecrosis in snakebite victims.
Although SBE poses a significant global health risk, immunotherapy with heterologous antivenoms is the only currently effective intervention. Antivenoms are usually produced by isolating polyclonal antibodies or antibody fragments from the serum or plasma of immunized animals with venom21. However, they present undesirable drawbacks, especially exposing patients to the risk of hypersensitivity reactions and anaphylaxis22,23. Besides, the antivenoms exhibit limited neutralizing ability on the CTX-induced necrosis12. The low efficacy of the antivenoms suggests the inadequate immunogenicity of the CTXs to produce CTX-specific antibodies. Several factors affecting the immunogenic capability of CTXs including the molecular structure, and sequence variations5,24. Small molecular size of the toxins may contribute to the sparse antigenicity25. Moreover, CTXs exhibit low immunogenicity due to their three-finger folded structures that limit its surface exposure for epitope recognition. It is conspicuous that the existing antivenoms are ineffective against dermonecrosis due to their poor pharmacokinetic profiles26. This sheds light in the recruitment of biotechnological tools to expedite the development of next generation antivenom, which allows the immunological characterization of venom toxins responsible for envenomation27,28.
By characterizing the epitopes of venom toxins, the possible immunogenic sites of toxins can be identified to aids in developing appropriate antivenoms specifically targeting the respective toxins. Computational analysis has been reported to examine the immunogenicity of toxins, such as three-finger toxins (3FTX), phospholipase A2s (PLA2s), cysteine-rich secretory proteins (CRISP), disintegrins (DIS), Kunitz peptides (KUN), L-amino acid oxidases (LAAO), natriuretic peptides (NP), snake C-type lectins (SNACLEC), snake venom metalloproteinases (SVMP) and snake venom serine proteases (SVSP) 25,29,30. Furthermore, high-throughput approaches such as peptide microarray, mass spectrometry (MS) epitope mapping, and phage display methods have also been developed to characterize the toxin-antivenom interaction29,30,31,32,33.
Epitopes are short molecular regions of an antigen recognized by the immune system34,35. The recognition of epitopes elicits immune responses to produce antibodies that bind to eradicate the antigen and inhibit its toxic effects36. This is the fundamental principle in current antivenom production, which involves the neutralization of venom toxins by the antibody37. Recent research demonstrates innovative venom-independent immunization strategies using synthetic epitopes for antibody production37. Such epitopes have shown more direct and potent protectivity than the immune response conferred by whole protein34,38. Notably, most ongoing research today focuses on the whole venom instead of toxin-specific antitoxins of interest39. The understanding of immunogenic properties of venom toxins shows the prospects of producing standardized synthetic epitopes for specific and effective antivenoms without causing toxicity to the host.
Despite the promising prospects of synthetic epitopes of toxins, limited knowledge is known on utilizing synthetic epitopes in CTXs-targeted therapeutics. Furthermore, the functional epitopes properties of CTXs remain unclear. Therefore, in this study, epitope characteristics of CTXs were determined to identify the potential epitopes of significant relevance as synthetic epitope leads for next generation antivenoms. This study used a conserved CTX sequence deduced from various CTX isoforms from different Naja species, aiming to overcome the predicaments of species diversity and geographical variation40. We applied immunoinformatic and multi-enzymatic-limited-digestion/liquid chromatography-mass spectrometry (MELD/LC–MS) to determine the immunogenic properties of the conserved CTX can be determined. Furthermore, we performed site-directed mutagenesis to validate the presence of epitope sites at the hydrophobic functional loops of CTX.
Results and discussions
Epitope prediction
T-cell epitope prediction
The list of predicted T-cell epitopes is summarized in Table 1. A total of 15 potential T-cell epitopes associated with 9 major histocompatibility complex (MHC) supertypes were identified. Some of these epitopes showed sequence and length similarities despite being classified into different MHC supertypes.
The parameters such as frequency of epitope among supertypes, affinity ranking, transporter associated with antigen processing (TAP) efficiency, and combined score were analyzed for ranking statistics (Table 1). The identical epitope sequences were then grouped. The frequency of T-cell epitopes present among 9 MHC supertypes was determined using equation (1).
Equation (1):
The frequency value determined which T-cell epitopes were predominant in the CTX. This can be deduced by ranking the T-cell epitopes according to their frequency of presence.
The sequences ‘NNKLVPLFY’ and ‘KMFMVSTPT’ exhibited the highest probability of 0.3333 across all MHC supertypes (Table 1). Both epitope sequences are made up of amino acid residues within the functional loops I and II of CTX, which are responsible for its cytotoxicity. The epitope sequences of ‘CPAGKNLCY’ and ‘MFMVSTPTK’ showed a probability of 0.2222, whereby ‘MFMVSTPTK’ shared 78% sequence homology to that of ‘KMFMVSTPT,’ which has a superior presence probability of 0.3333 with only discrepancy of an amino acid residue. The remaining predicted T-cell epitopes exhibited only probability values of 0.1111 across different MHC supertypes. CTX consists of three functional loops responsible for cytotoxicity13,14,16. Despite some predicted T-cell epitope sequences such as 'CPKNSVK' and 'DVCPKNSLL' resided entirely on the functional loops' amino acids, these sequences somehow demonstrated a lower presence probability value of 0.1111.
The affinity (supertype) scoring represents the predicted MHC binding affinity as -log50000(affinity score), whereby log50000 is the logarithm with base 50,000 for the affinity score. The higher the scoring, the higher the binding affinity. Based on the result, 'FMVSTPTKV' exhibited a higher affinity score (Table 1). This epitope sequence resided within the functional loop II of CTX. This was followed by 'KLVPLFYKT,' which resided on the functional loop I. Despite so, some epitopes on functional loops showed relatively lower affinity scores, for example, 'CPKNSLLVK,' 'DVCPKNSLL,' and 'NNKLVPLFY.'
The ranking analysis of the TAP efficiency of each predicted T-cell epitope sequence was also performed. The interactions of cytotoxic T-cells with other cells abide by a strict and specific recognition process dictated by T-cell epitopes presented on the surfaces of human leukocyte antigens (HLAs)41. However, before they are presented on the surfaces of HLAs, the transport associated with antigen processing (TAP) recognizes and transports these epitope sequences into the endoplasmic reticulum42. Therefore, the role of TAP must significantly influence the selection of T-cell epitopes. The TAP efficiency ranking showed that 'NNKLVPLFY' and 'CPAGKNLCY' exhibited TAP efficiency scores of 2.8000 and 2.6060, respectively. The third-ranked TAP efficiency sequence was 'KTCPAGKNL.' There was a drastic difference in TAP efficiency scores among the epitope sequences, presumably due to the strict specificity of TAP binding with peptide sequences. Notably, the amino acid residue 'KT' were found in the epitopes with lower TAP efficiency scores, indicating that these two amino acids significantly reduced the TAP efficiency. Structurally, these two amino acids comprise the β-sheet of loop I in the CTX. Their structural role may be accounting for their stringent non-interactive characteristics. Interestingly, none of the amino acid residues of 'CPAGKNLCY' was found to reside within the functional loops of CTX. This epitope sequence constituted the region that connected loops I and II of CTX, suggesting that this sequence may facilitate CTX-membrane interactions.
The next parameter used in predicting T-cell epitopes was combined score ranking. The combined score consists of an integrated peptide binding affinity, C-terminal cleavage affinity, and TAP efficiency scores, which are all significantly involved in the MHC Class I processing pathway43. To determine a compromise between sensitivity and specificity, a threshold of 0.75 for the combined score of predicted epitope sequences was used for selection. Thus, the predicted four epitope sequences 'FMVSTPTKV,' 'KLVPLFYKT,' 'MFMVSTPTK,' and 'VPVKRGCID' exhibited the highest chances as potential T-cell epitopes of CTX.
Based on the results, 'NNKLVPLFY' and 'KLVPLFYKT' were found within loop I, while 'FMVSTPTKV' and 'MFMVSTPTK' were situated in loop II. 'CPAGKNLCY' constituted the region between loops I and II of CTX. These sequences were the best sequences representing T-cell epitopes of CTX. Although 'CPAGKNLCY' was not situated in the functional loops of CTX, this epitope sequence exhibited a greater TAP efficiency score and presence probability among all MHC supertypes. Furthermore, this sequence was amphiphilic and highly basic, conserved in CTX for interaction with targets13. This epitope sequence also connected the two prominent β-sheets of loops I and II, which are vital to the overall functionality and stability of CTX. Since functional loops of CTX are important for cytotoxicity, and antivenoms have a binding affinity towards functional sites of toxins, it is feasible to deduce that the amino acid sequences that make up the functional loops were potential T-cell epitopes14,28,44.
B-cell epitope prediction
Protrusion index score
The protrusion index (PI) is a simple and effective concept for identifying protein regions that protrude from its surface. A high PI score is often related to sites of antigenicity45. The B-cell epitope prediction based on the PI score for CTX was summarized in Table 2. Four predicted linear B-cell epitopes were 'DVCPKNSLL,' 'MVSTPTKVP,' 'KLVPLFYKTCPAGKN,' and 'NTDRCN.' Of these B-cell epitopes, 'DVCPKNSLL,' 'MVSTPTKVP,' and 'KLVPLFYKTCPAGKN' were found on the functional loops of CTX. The residues within the functional loops were underlined in the predicted discontinuous B-cell epitopes.
Parker hydrophilicity prediction
The basis of utilizing hydrophilicity to predict epitopes is attributed to the rational nature of epitopes on the surface, and they are exposed to be potentially antigenic46. The Parker hydrophilicity prediction of CTX is illustrated in Fig. S1A (Supplementary File). It was found that the hydrophilicity peaks surpassed the threshold value around amino acid residues 4–6, 15–21, 30–44, and 55–59, which represented hydrophilic and potentially antigenic sequences of 'LKCNNKLVP,' 'YKTCPAGKNLCYK,' 'MVSTPTKVPVKRGCIDVCPKNSL' and 'KYCCNTDRCN,' respectively.
Kolaskar & Tongaonkar antigenicity prediction
Kolaskar & Tongaonkar antigenicity prediction analyzes the physiochemical properties and abundance of amino acids within CTX compared to B-cell epitope references. Higher antigenicity scores are associated with a higher potential to induce immune responses47. As illustrated in Fig. S1B, peaks were observed at positions 8–14, 37–38, 41–43, 46–47, and 50–57, which represented sequences 'NKLVPLFYKTCPA,' 'VPVKRGCI,' 'RGCIDVCPK,' 'VCPKNSLL' and 'NSLLVKYVCCNTDR' respectively.
Emini surface accessibility prediction
An essential criterion of a potent epitope is its accessibility to other molecules for interaction. Emini surface accessibility prediction was therefore used to analyze and evaluate the surface accessibility of the amino acid residues. The linear B-cell epitopes were predicted based on their surface accessibility characteristics. As shown in Fig. S1C, 'STPTKVPV' appeared to be the most prominent peak with high surface accessibility at positions 29–35.
Karplus & Schulz flexibility prediction
Flexibility in protein structures is a vital characteristic that influences a protein’s ability to perform binding, catalysis, and allosteric interactions48. Karplus & Schulz's flexibility prediction was used to analyze the structural flexibility contributed by individual amino acid residues of CTX. As shown in Fig. S1D, the flexible regions were observed at positions 4–7, 14–21, 30–39, 45–49, and 58, which represented the sequences 'LKCNNKLVPL,' 'FYKTCPAGKNLCYK,' 'MVSTPTKVPVKRGCID,' 'DVCPKNSLLVK' and 'CCNTDRC,' respectively.
Multiple sequence alignment of T-cell and B-cell epitopes
The predicted T-cell and B-cell epitopes were aligned using Clustal Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/)49,50,51. Based on the multiple sequence alignment, three potential T-cell epitope sequences of 'KLVPLFY,' 'CPAGKNLCY,' and 'MFMVSTPTK' were selected due to high consensus frequency (Fig. 1A). Of these sequences, 'KLVPLFY' and 'MFMVSTPTK' were found within the functional loops I and II, respectively. B-cell epitopes 'NKVLPLFY,' 'KTCPAGKNL,' 'MVSTPTKVP,' and 'DVCPKNSLL' were identified to have high consensus frequency (Fig. 1B). Interestingly, some of the identified T-cell and linear B-cell epitopes share significant sequence homology. For example, T-cell epitopes 'KLVPLFY,' 'CPAGKNLCY’ and ‘MFMVSTPTK’ shared similarities with B-cell epitopes ‘NKVLPLFY,' 'KTCPAGKNL,' 'MVSTPTKVP', respectively. Thus, four epitope sequences were concluded as potential epitopes for CTX. There were 'KLVPLFY,' 'CPAGKNLCY,' 'MFMVSTPTKVP,' and 'DVCPKNSLL.'
Multiple sequence alignment of predicted (A) T-cell epitopes, and (B) B-cell epitopes of cytotoxin. The red boxes represent the highly consensus residues.
Molecular docking and molecular dynamic simulations of cytotoxin with HLA supertypes
The HADDOCK 2.2 web server clustered 583 structures into 89 clusters, representing 36.44% of the water-refined models. The docking parameters are summarized in Table 3. The HADDOCK score represents the weighted sum of all energies (Van der Waals, electrostatic interactions, desolvation energy, buried surface area, and restraints violation energy) between the docking structures. This scoring parameter was used to rank the docking clusters52. A lower HADDOCK score indicates a more reliable docking structure with better energetic stability53,54. HLA-B62 possessed the lowest HADDOCK score of 257.6 ± 30.1, followed by HLA-B27 and HLA-A02 (Table 3). The Z-score is a scoring function that predicts the binding affinity of multiple ligands towards a receptor, whereby a more negative Z-score indicates significant binding affinity55. All HLA-CTX complexes had negative Z-scores, but HLA-B07 and HLA-B62 had Z-scores of -1.9 and -2.0, respectively. This result suggested that HLA-B07 and HLA-B62 had the best binding affinity towards the T-cell and B-cell epitopes CTX. In molecular docking, the root-mean-square deviation (RMSD) value is used to assess whether a docked simulation between a ligand and receptor in a specific binding orientation and conformation are reproducible and values of < 2.0 Å indicate successful and appropriate docking56. Only HLA-B07 and HLA-B62 exhibited RMSD values < 2.0 Å, thus implying that they are reliable and reproducible docking structures. Therefore, HLA-B62 exhibited the best binding affinity and stability towards CTX's epitopes.
Furthermore, we also performed Residue Interaction Network Generator (RING) analysis to determine the interacting residues between CTX and different HLA supertypes. RING analysis enables the identification of non-covalent interactions at the atomic level of a protein57. The RING analysis then revealed four highly interacting regions of 'NKLVPLFYK,' 'AGKNL,' 'MFMVSTPTLVPVK,' and 'DVCPKNSLLV' as the potential epitope sites, within the functional loops of CTX (Fig. 2A).
The molecular docking analysis of cytotoxin (CTX) with different Human Leukocyte Antigen (HLA) supertypes. (A) RING analysis for determination of interacting residues between CTX and eight HLA supertypes. Four regions indicated the highly frequent interacting sites (frequency ⩾3) suggesting these interacting sites were the potential epitopes. (B) Examples of interaction between CTX (red color) and the best docked HLA, HLA-B62 (green). The interacting residues were annotated, the dashed lines indicated hydrogen bonds.
The significant intermolecular interactions within HLA-CTX complexes were Van der Waals (VDW) forces, hydrogen bonds, and salt bridges (Table 3). In comparison, π-π stacking and cation-π interaction were only observed in the docking structures of HLA-A01, HLA-A24, and HLA-B27. We considered the composition of the bonds within the docking complex to select the best HLA-CTX. Furthermore, a balance between the percentage of interacting bonds at the T- and B-cell epitopes was also considered. The π-π stacking and cation-π interactions were excluded as there is still limited information that serves as a reference for the best complex selection in docking simulations. The number of salt bridges was used as the first reference point for this selection because they constitute the most substantial interaction among all non-covalent molecular bonds. Among all the HLA-CTX complexes, HLA-A01, HLA-B27, HLA-B58, and HLA-B62 contained higher numbers of salt bridges with CTX's epitopes (Table 3). HLA-A01 and HLA-B27 were excluded because they did not interact with B-cell epitopes. By comparing HLA-B58 and HLA-B62, HLA-B62 had a relatively higher percentage of interactions with T-cell and B-cell epitopes (Fig. 2B), suggesting that HLA-B62 was the best binding HLA supertype with CTX.
The stability of the HLA-B62-CTX complex was further examined using molecular dynamic (MD) simulations) in a water box for 150 ns at 300 K. The mean RMSD value (Supplementary File-Fig. S2 was determined for the docking complex as 8.73 ± 0.80 Å. The mean MMPBSA potential energy of the docking complex remained at -23.55 ± 9.71 kcal/mol, suggesting the complex was stable during the simulation period. The mean number of hydrogen bonds of 1.56 ± 0.99 indicated the formation of hydrogen bonds within the HLA-CTX complex during simulation.
Epitope-omic: MELD/LC–MS
The multi-enzymatic limited digestion (MELD) approach was first introduced by Morsa et al.58, which involves a cocktail of multiple diluted enzymes to generate peptide fragments. The synergistic effect of multi-enzymatic digestion results in enhanced sequence coverage and promising protein characterization58. This was due to the ability of the multi-enzymatic schemes to overcome the limitations of the mono-enzymatic digestion method, such as sequence-dependent digestion efficiencies and insufficient sequence coverage58. Therefore, the MELD/MS-based epitope mapping was employed to validate CTX’s epitopes. The epitope excision method was applied to exhibit no preference for either the “consecutive” or “assembled” epitopes59. This approach is based on the principle of specific binding between the antibody and antigen.
In this experiment, an anti-CTX monoclonal antibody with broad specificity was used to map the epitope sites on CTX. Upon binding to the anti-CTX, the CTX's epitopes were shielded from enzymatic proteolysis, and the exposed regions were digested. The reaction mixture was then subjected to LC–MS to identify the peptide(s) of the undigested region(s) known to be the epitopes. Prior to the MS-based epitope mapping, the immunoturbidimetric assay confirmed the formation of immunocomplexes between the anti-CTX and CTX. Based on the kinetic analysis (Fig. 3A), the formation of immunocomplexes peaked within the first minute and plateaued after that. The increased in absorbance reading reflects the turbidity formation. This is in corroborate with rapid formation of immunocomplexes as reported by O’Leary et al.60. The formation of immunocomplexes occurred from 3.125 to 50 µg/mL of anti-CTX, whereby the greatest turbidity was measured at 25 µg/mL of anti-CTX.
The alignment of epitope sequences determined by different epitope mapping approaches. (A) Immunocomplex kinetic analysis of CTX and anti-CTX, different concentrations of anti-CTX (3.125–50 µg/mL) were incubated with 12.5 µg/mL of CTX. Absorbance readings were measured at the wavelength (λ) of 340 nm (n = 3). (B) Multiple sequence alignment of MELD/LC–MS peptides with CTX revealed the presence of highly consensus sequences, annotated in red boxes. (C) Multiple sequence alignment and summary of epitope sequences determined from different mapping methods.
From the MELD/LC–MS results, 1:5000 diluted anti-CTX demonstrated a greater coverage of peptides mapping (Supplementary File-Table S1; Fig. S4). Thus, these identified peptide sequences were mapped to CTX (Fig. 3B). The multiple sequence alignments of LC–MS identified peptide sequences revealed three highly consensus sequences, ' LVPLFYK,' 'MFMVS,' and 'TVPVKR,' which were situated at the functional loops I and II of CTX (Fig. 3B).
In summary the epitope sequences of CTX determined from different mapping approaches (Fig. 3C) were finalized as ‘KLVPLFYK’, ‘AGKNL’, ‘MFMVSTPKVPV’ and ‘DVCPKNSLL’. Three of these sequences were located at the functional loops I, II, and III of CTX.
Site-directed mutagenesis and cytotoxicity assay
Given that the epitopes ‘KLVPLFYK’, ‘MFMVSTPKVPV’ and ‘DVCPKNSLL’ were situated at functional loop I–III of CTX- (Fig. 3C), we performed site-directed mutagenesis to further confirm the presence of epitopes at functional loops. The membrane interactions of CTXWT and CTXVAR were expected to vary, which subsequently affected their cytotoxicity.
A CTX variant (CTXVAR) contained four mutated sites within the potential epitope sequences on the functional loops. The gene sequences of CTXWT and CTXVAR are as follows, underlined residues indicate mutated regions while asterisk (*) represents identical residues:
These mutated residues (Fig. 4A) were introduced based on the fact that the changes in amino acids did not affect the three-fingered structure of CTX (Fig. 4B). The RMSD values of both models were < 2.0 Å, indicating that they were all of good docking quality and did not deviate from the ideal orientation61. Nevertheless, the DOPE energy score (Fig. 4C) of CTXVAR somehow showed variations at amino acids position 10–12, 17–20, 30–34, and 46–49 compared to CTXWT. These positions represented the four mutated sites that reside in the functional loops of CTX, whereby substantial variation was observed at functional loops I and II, which are vital towards the formation of the cytotoxin's hydrophobic core and cytotoxin-membrane binding motif13.
Site-directed mutagenesis analysis of cytotoxin (CTX) to determine its functionality. (A) Comparison of amino acid sequences between wildtype CTX (CTXWT) and variant (CTXVAR), red boxes indicate the mutagenesis sites. (B) Superimposed structural view of CTXWT (cyan color) and CTXVAR (purple color) showed similar three-finger folded structure. (C) DOPE per residue energy score against amino acid residue, CTXVAR demonstrated variations at amino acid residues 10–12, 17–20, 30–34, and 46–49 compared to CTXWT. (D) Cell viability of HaCat cell lines after treatment with CTXWT and CTXVAR for 24 h. PBS and Triton-X 100 were used as negative and positive controls, respectively. The cytotoxic effect of CTXVAR was slightly higher than the CTXWT; treatment groups. The data are shown as mean ± SEMs of three independent experiments (n = 3), the data was analyzed using one-way ANOVA whereby **** indicates p < 0.0001.
The alteration of amino acid residues in CTXVAR somewhat changed its pI value, whereby CTXVAR was more acidic (pI: 5.29) than CTX wildtype, CTXWT (pI: 9.37). The interaction between CTX and the cell membrane is known to be closely related to the types of CTX, namely S-type and P-type. The insertion of S-type CTX into the cell membrane was primarily relying on the structural adaptation of loop II62. On the other hand, loops II and III of the P-type CTX were found to be more critical in lipid membrane penetration63. Additionally, studies showed that loop I of both S- and P-type CTX was involved in the membrane interaction64,65. Altogether, no conclusive finding supports a more significant loop among the three functional loops of the CTX during membrane penetration. However, loop II has been reported to involve in the membrane insertion for both S- and P-type CTX, implying that it plays a prominent role in the interaction between the CTX and cellular membranes.
The cytotoxicity of CTXWT and CTXVAR was determined in the human skin keratinocytes (HaCat cell lines). Both CTXWT and CTXVAR significantly (p < 0.0001) reduced the viability of the HaCat cell lines (Fig. 4D). It was observed that the CTXVAR showed higher cytotoxicity than the CTXWT. The alteration of CTXWT at the functional loops produced a more acidic CTXVAR with enhanced cytotoxicity. Following the interaction of acidic CTX with cell membrane, it increased the acidity of cell membrane proteins to rigidify the membrane by expulsing water molecules, which ultimately shifts the cellular membrane to a gel phase66. Dehydration of the cellular membrane also influences dynamic membrane heterogeneity and facilitates membrane fusion, resulting in compromised membrane fluidity that leads to cell death67,68. The alteration of functional loops containing epitope sequences thus affected the cytotoxicity of CTX. The results confirmed the presence of epitope sequences at functional loops of CTX.
Materials and methods
Cytotoxin sequence
A conserved sequence of cytotoxin (CTX) was constructed in our laboratory40 based on consensus sequence obtained from multiple sequence alignment of the non-redundant and mature CTX sequences (NCBI Serpentes protein database, Taxid: 8570). These sequences constitute different CTX isoforms of different Naja species40. The final consensus CTX sequence is shown below:
This conserved CTX is highly matched to Naja atra cytotoxin (accession no: Q98959.1, Taiwan origin) with 95% sequence similarity40.
Epitope prediction
Prediction of T-cell epitope
T-cell epitope (CTL) prediction of the cobra venom cytotoxin was performed using the NetCTL 1.2 webserver (https://services.healthtech.dtu.dk/service.php?NetCTL-1.2) 43. The HLA supertypes selected for the prediction were A1, A2, A3, A24, A26, B7, B8, B27, B39, B44, B58, and B62; the threshold score for epitope identification was set at 0.50, while the weight on C terminal cleavage and TAP transport efficiency were 0.15 and 0.05, respectively. For the HLA-B62 supertype, a prototypical member of the supertype, HLAB*05:01, was used to represent it69. Statistical analyses were performed on each T-cell epitope candidate and ranked by four categories: frequency of epitope presence across supertypes, affinity score, transport associated with antigen processing (TAP) efficiency score, and combined score.
Prediction of B-cell epitope
The potential linear B-cell epitopes of CTX were identified using four different prediction algorithms, namely BepiPred-2.0 (https://services.healthtech.dtu.dk/services/BepiPred-2.0/) 70, IEDB analysis resource (http://tools.immuneepitope.org/bcell/) 71, antigenic peptide prediction method (http://imed.med.ucm.es/Tools/antigenic.pl) 72 and ABCPred (http://www.imtech.res.in/raghava/abcpred/) 73,74. Whereas the discontinuous B-cell epitopes were predicted using ElliPro webserver (http://tools.iedb.org/ellipro/) 75, provided by the Immune Epitope Database (IEDB) based on each amino acid residues' protrusion index score. Karplus & Schulz flexibility prediction (window size value of 7 and threshold value of 0.986), Kolaskar & Tongaonkar antigenicity (window size value of 7 and threshold value of 1.091), and Parker hydrophilicity prediction (window size value of 7 and threshold value of 0.895), was applied for the prediction. The predicted sequences from the outputs were aligned, whereby amino acid sequences with more than three overlapping frequencies were potential B-cell epitopes.
Human leukocyte antigen (HLA) distribution analysis
The T-cell epitopes (CTL) were also used for the HLA distribution analysis as the epitopes were also MHC ligands. The binding prediction of the epitopes to MHC Class I was evaluated using the peptide binding to MHC Class I molecules tool from the Immune Epitope Database (IEDB; http://tools.iedb.org/mhci/). NetMHCpan EL 4.1 was selected as the prediction method recommended by the IEDB for the binding prediction analysis across all alleles76. The peptide length was set as nonamer. The peptide-MHC I combinations with the highest scoring were filtered out from each peptide sequence input. A higher score indicates a higher affinity for the interaction between the peptide and MHC molecules. Prediction of the T-helper cell (HTL) epitopes for the CTX was performed using the peptide binding to MHC class II molecules tool from IEDB (http://tools.iedb.org/mhcii/). IEDB recommended NetMHCIIpan 4.1 EL was selected as the prediction method using specific MHC II alleles76. A 15-mer peptide length was selected. The peptide-MHC II interactions with the highest percentile rank in each set, indicating the strongest binding affinity, were filtered out for further analyses.
Molecular docking of cytotoxin with human leukocyte antigens (HLAs)
According to the predicted T-cell epitopes list, molecular docking analysis of T-cell epitopes with different HLA supertypes was performed on HADDOCK 2.2 to determine the interaction77,78. The PDB structures of different HLA supertypes were obtained from the RCSB PDB database. These HLA supertypes were HLA-A01 (3BO8), HLA-A02 (1AO7), HLA-A03 (3RL1), HLA-A24 (4F7P), HLA-B07 (5EO1), HLA-B27 (2BST), HLA-B58 (5IM7) and HLA-B62 (3C9N). CTX was set as the first molecule with functional loops' residues (residues 4–12, 27–36, 41–51) as active residues. The second molecule input was HLA supertypes, and all amino acid residues were set as active residues. The best docking clusters were determined by assessing each structure's HADDOCK score, Z-score, and RMSD value. Apart from that, the chosen docking cluster was accepted if the cluster maintained its native conformation, i.e., β-strands that are responsible for the stability of the structure. Later, the PDB of the best docking clusters was subjected to Residue Interaction Network Generator 2.0.1 (RING 2.0.1) analysis (http://old.protein.bio.unipd.it/ring/) 57 to determine the types and strength of both intra- and inter- non-covalent interactions of the docked structural clusters. The intermolecular non-covalent interactions between amino acids of CTX and HLA supertypes were mapped based on the aligned T-cell and B-cell epitopes.
Molecular dynamic (MD) simulations
The stability of CTX-HLA supertype complexes was further examined by molecular dynamic (MD) simulations using the GROMACS simulation package79,80. The MD simulations of all complexes were carried out for 150 ns in a solvated TIP3P water box using CHARMM 36 m force field, whereby the trajectory and energy files were written every 10 ps. Three chloride ions were added to the complex to neutralize the overall system. The protonation states were checked at pH 7 using the H + + web server for His, Lys, Arg, Asp, and Glu residues. Minimization was carried out for 5000 steps using Steepest Descent Method, and the convergence was achieved within the maximum force < 1000 (KJ mol-1 nm-1) to remove any steric clashes. All three systems were equilibrated at NVT and NPT ensembles for 100 ps (50,000 steps) and 1000 ps (1,000,000 steps), respectively, using time steps 0.2 and 0.1 fs, respectively, at 300 K to ensure a fully converged system for the production run. The production runs for simulation were carried out at a constant temperature of 300 K and a pressure of 1 atm or bar (NPT) using weak coupling velocity-rescaling (modified Berendsen thermostat) and Parrinello-Rahman algorithms, respectively. All bond lengths involving hydrogen atoms were kept rigid at ideal bond lengths using the Linear Constraint Solver (lincs) algorithm, allowing for a time step of 2 fs. The Verlet scheme was used for the calculation of non-bonded interactions. Periodic Boundary Conditions (PBC) were used in all x, y, and z directions. Interactions within a short-range cutoff of 1.2 nm were calculated in each time step. Particle Mesh Ewald (PME) was used to calculate the electrostatic interactions and forces to account for a homogeneous medium outside the long-range cutoff.
Epitope-omics: MELD/LC–MS
Multi-enzymatic-limited-digestion (MELD)
In this experiment, MELD was used to prepare the immunocomplexes for peptide mapping in epitope analysis. In brief, an anti-CTX monoclonal antibody (TPL–27_01_F7 scFv clone, Absolute Antibody, United Kingdom), was diluted according to the ratio of 1:5000, 1:10,000, and 1:30,000 in phosphate-buffered saline (PBS). The anti-CTX was mixed with pure CTX (see section “Expression of CTXWT and CTXVAR”) in 1:1 (v/v) for 2 h at room temperature to form immunocomplexes. Subsequently, the immunocomplexes were concentrated in PBS, and the unbound CTX was removed with a 10 kDa AmiconⓇ Ultra Centrifugal Filter Spin Column (Merck, United States). The immunocomplex was then reduced with dithiothreitol (DTT; Thermo Fisher Scientific, United States) and alkylated and iodoacetamide (IAM; Thermo Fisher Scientific, United States). This was followed by proteolytic digestion with multiple endoproteases. The endoproteases used in MELD were trypsin (Thermo Fisher Scientific, United States), chymotrypsin (Thermo Fisher Scientific, United States), and Lys-C (New England Biolabs, United States) in a 1:20 enzyme-to-protein ratio by mass. The reaction mixtures were incubated at 37 °C overnight before desalting using ZipTipⓇ pipette tips (Merck, United States). The digested peptides were then vacuum-dried before LC–MS analysis.
Examination of immunocomplex formation by immunoturbidimetric assay
The immunoturbidimetric assay protocol was modified from a reported study by O’ Leary and colleagues60. Briefly, a two-fold serial dilution of anti-CTX from 50 µg/mL in 50 µL PBS was added into a 96-well plate. Then, another 50 µL of CTX (25 µg/mL) was added into each well, resulting in a final concentration of 12.5 µg/mL. Absorbance readings were taken every minute at wavelength 340 nm using the Tecan microplate plate reader at 37 °C for 40 min. The kinetics of the immunocomplexes formation were then plotted and examined.
Liquid chromatography-mass spectrometry (LC–MS)
The identification of the MELD-peptides was conducted using the Agilent 1200 HPLC-Chip/MS Interface, coupled with Agilent 6550 iFunnel Q-TOF LC/MS instruments. The lyophilized samples were reconstituted in solvent A (0.1% formic acid in water). Subsequently, 1 µL of the sample was injected into the C18 enrichment and analytical column (Agilent Large Capacity Chip, 300 Å, C18, 160nL enrichment column & 75μm × 150mm analytical column) at a flow rate of 4 µL/min from Agilent 1200 Series Capillary pump and 0.5 µL/min from Agilent 1200 Series Nano Pump with solvent B (90% of acetonitrile in water with 0.1% formic acid). The peptide mixtures were separated for 50 min under an increasing gradient of solvent B from 5 to 70%, followed by an 8-min post-run analysis. The elution of peptides was configured as follows: positive ion mode, capillary voltage 1.9 kV, fragmentor voltage 360 V, gas temperature 325 °C, drying gas flow 5.0 L/min. Mass spectra were acquired in an MS/MS mode over a mass (m/z) range of 50–3000. Solvent A was used as a blank to eliminate peptide contamination from other sources. The data was then processed using the Peaks X + software, searching against the Serpentes protein database (Taxid: 8570).
Multiple sequence alignment
To finalize the potential epitope sequences of CTX, multiple sequence alignment of predicted epitopes was performed using the Clustal Omega49,50,51 and visualized using Jalview 2.11.1.381. Several parameters, such as overlapping sequences, degree of sequence conservation, and the number of amino acids residing within functional sites of CTX, were considered.
Functional epitope site mapping of cytotoxin by site-directed mutagenesis
To further confirm the presence of epitopes at the functional loops of CTX, functional epitope mapping of CTX was performed by site-directed mutagenesis.
Preparation of pET-22b ( +)-CTXWT and CTXVAR
Polymerase chain assembly (PCA) was used to synthesize wildtype CTX (CTXWT) and variant (CTXVAR) genes by Economy Gene Synthesis (GenScript, United States). Both genes were inserted into pET-22b ( +) plasmids at the NdeI/XhoI restriction sites (Supplementary File-Fig. S3), respectively, and transformed into Escherichia coli Dh10B cells for plasmid expansion.
Expression of CTXWT and CTXVAR
The pET-22b ( +)-CTXWT and CTXVAR plasmids were extracted from glycerol stock of DH10B cells using GF-1 Plasmid DNA Extraction kit (Vivantis, Malaysia) following the manufacturer’s protocol. The CTXWT and CTXVAR genes were amplified by touch-down PCR with their respective forward and reverse primers (Supplementary File 1-Table S2).
To express CTXWT and CTXVAR, pET-22b ( +) plasmids harboring the respective genes were first transcribed into RNA using the HiScribe™ T7 Quick High Yield RNA Synthesis Kit (New England Biolabs, United States). The transcribed RNAs were then expressed using NEBExpress® Cell-free E. coli Protein Synthesis System (New England Biolabs, United States) for in vitro protein translation, according to the manufacturer's protocol. After that, CTXWT and CTXVAR were purified using a HisTrap™ HP, 1 mL affinity column on ÄKTA Pure protein purification system (Cytiva, United States). The HisTrap column was equilibrated with buffer A (20 mM sodium phosphate, 0.5 M NaCl, pH 8.0) before the samples were injected into the column. CTXWT and CTXVAR were purified with a gradient elution from 4 to 100% of buffer B (20 mM sodium phosphate, 0.5 M NaCl, 500 mM imidazole, pH 8.0) at a flow rate of 500 uL/min. The presence of his-tagged expressed CTX and its purity were visualized by western blot.
Cytotoxicity of CTXWT and CTXVAR
The cytotoxicity of CTXWT and CTXVAR was determined using the cell-counting kit-8 (CCK-8) assay (Sigma Aldrich, United States). Briefly, 100,000 cells/mL of human skin keratinocytes, Hacat cell lines (Elabscience Biotechnology Inc, United States), were seeded in a 96-well plate at 37 °C, 5% CO2 overnight. The cells were treated with 100 µL of 8 µg/mL CTXWT and CTXVAR in serum-free cell culture media for 24 h. Ten microliters of CCK-8 reagent were added to each well and incubated for 1 h. The level of cell viability was quantified through the formation of orange-colored products upon reduction of the WST-8 salts present in the CCK-8 reagent by dehydrogenase in viable cells. The absorbance readings were obtained using the Tecan microplate reader at wavelength 450 nm. The percentage (%) of cell viability was calculated.
Homology modeling of CTXWT and CTXVAR
We performed homology modeling on MODELLER 9.2182 to determine if site-directed mutagenesis produced any structural changes to CTXVAR. The DOPE per residue profile of each amino acid residue within the CTXWT and CTXVAR was plotted on Gnumeric for comparison. CTXWT and CTXVAR models were subjected to further visual evaluation using BIOVIA Discovery Studio Visualizer 2021 by superimposing both models. The corresponding root-mean-squared deviation (RMSD) value was also determined for CTXWT and CTXVAR models.
Conclusion
Potential T-cell and B-cell epitopes of CTX were identified to be 'KLVPLFY,' 'CPAGKNLCY,' 'MFMVSTPTKVP,' and 'DVCPKNSLL’, whereby 'KLVPLFY,' ‘MFMVSTPTKVP,' and 'DVCPKNSLL’ were found to reside within the functional loops of CTX. 'CPAGKNLCY' constituted the region between loops I and II, which is vital for CTX’s structural folding and stability. HLA-B62 supertypes was identified to possess the highest binding towards CTX. The intermolecular bonds between the HLA-CTX complex were hydrogen bonds, Van der Waals interactions, and salt bridges. The crucial amino acid residues on CTX, namely L7, G18, K19, N20, M25, K33, V43, C44, K46, N47, and S48 were involved in intermolecular interactions with HLA-B62. The epitope-omic analysis revealed three highly potential epitope sequences ' LVPLFYK,' 'MFMVS,' and ‘TVPVKR,’ which were also located at the functional loops I and II. Different mapping approaches finalized the potential epitopes of CTX as ‘KLVPLFYK’, ‘AGKNL’, ‘MFMVSTPKVPV’ and ‘DVCPKNSLL.’ Furthermore, site-directed mutagenesis of these four epitope sites altered the physicochemical properties and cytotoxicity of CTX albeit the overall structure of CTX was retained. This concluded that, the identified epitope sequences were located at the functional loops of CTX. Since the functional loops of CTX form its hydrophobic three-finger folded core, our findings provide a justification of low immunogenicity of CTX, which results in inadequate efficacy of existing antivenoms targeting CTX. In conclusion, the CTX’s functional epitopes represented consensus epitope properties of different CTX isoforms. These functional epitopes can be fabricated as synthetic epitopes to produce toxin-targeted cross-reactive antivenom that antagonize the cytotoxic effects in dermonecrosis.
Data availability
The mass spectrometry proteomics datasets generated and/or analyzed during the current study are available in the ProteomeXchange Consortium (https://www.proteomexchange.org/) via the PRIDE partner repository with the dataset identifier PXD040420.
References
Chippaux, J. Snakebite envenomation turns again into a neglected tropical disease!. J. Venom. Anim. Toxins incl. Trop. Dis. 23, 1–2 (2017).
Panagides, N. et al. How the cobra got its flesh-eating venom: Cytotoxicity as a defensive innovation and its co-evolution with hooding, aposematic marking, and spitting. Toxins. 9, 1–22 (2017).
Gasanov, S. E. et al. Naja naja oxiana cobra venom cytotoxins CTI and CTII disrupt mitochondrial membrane integrity: Implications for basic three-fingered cytotoxins. PLoS ONE 10, 1–28 (2015).
Ismail, M., al-Bekairi, A. M., el-Bedaiwy, A. M. & Abd-el Salam, M. A. The ocular effects of spitting cobras: II. Evidence that cardiotoxins are responsible for the corneal opacification syndrome. J. Toxicol. Clin. Toxicol. 31, 45–62 (1993).
Liu, C. C. et al. Pathogenesis of local necrosis induced by Naja atra venom: Assessment of the neutralization ability of Taiwanese freeze-dried neurotoxic antivenom in animal models. PLoS Negl. Trop. Dis. 14, 1–20 (2020).
O’Brien, J., Lee, S. H., Gutiérrez, J. M. & Shea, K. J. Engineered nanoparticles bind elapid snake venom toxins and inhibit venom-induced dermonecrosis. PLoS Neg. Trop. Dis. 12, 1–20 (2018).
Utkin, Y. N. & Osipov, A. V. Non-lethal polypeptide components in cobra venom. Curr. Pharm. Des. 13, 2906–2915 (2007).
Yap, M. K. K., Fung, S. Y., Tan, K. Y. & Tan, N. H. Proteomic characterization of venom of the medically important Southeast Asian Naja sumatrana (Equatorial spitting cobra). Acta Trop. 133, 15–25 (2014).
Lauridsen, L. P., Laustsen, A. H., Lomonte, B. & Gutiérrez, J. M. Exploring the venom of the forest cobra snake: Toxicovenomics and antivenom profiling of Naja melanoleuca. J. Proteomics. 150, 98–108 (2017).
Offor, B. C., Muller, B. & Piater, L. A. A review of the proteomic profiling of African Viperidae and Elapidae snake venoms and their antivenom neutralisation. Toxins. 14, 723 (2022).
Hiu, J. J. & Yap, M. K. K. The myth of cobra venom cytotoxin: More than just direct cytolytic actions. Toxicon: X. 14, 100123 (2022).
Ho, C. H. et al. Analysis of the necrosis-inducing components of the venom of Naja atra and assessment of the neutralization ability of freeze-dried antivenom. Toxins. 13, 619 (2021).
Gasanov, S. E., Dagda, R. K. & Rael, E. D. Snake venom cytotoxins, phospholipase A2s, and Zn2+-dependent metalloproteinases: Mechanisms of action and pharmacological relevance. J. Clin. Toxicol. 4, 1–34 (2014).
Konshina, A. G., Dubovskii, P. & Efremov, R. G. Structure, and dynamics of cardiotoxins. Curr. Protein Pept. Sci. 13, 570–584 (2012).
Girish, V. M. et al. Identification and structural characterization of a new three-fingered toxin Hemachatoxin from Hemachatus haemachatus venom. PLoS ONE 7, e48112 (2012).
Dubovskii, P. V. & Utkin, Y. N. Cobra cytotoxins: Structural organization and antibacterial activity. Acta Naturae. 6, 11–18 (2014).
Forouhar, F. et al. Structural basis of membrane-induced cardiotoxin A3 oligomerization. J. Biol. Chem. 278, 21980–21988 (2003).
Dubovskii, P. V., Konshina, A. G. & Efremov, R. G. Cobra cardiotoxins: Membrane interactions and pharmacological potential. Curr. Med. Chem. 21, 270–287 (2014).
Feofanov, A. V. et al. Comparative study of structure and activity of cytotoxins from venom of the cobras Naja oxiana, Naja kaouthia, and Naja haje. Biochemistry 69, 1148–1147 (2004).
Wu, M. et al. The anticancer effect of cytotoxin 1 from Naja atra Cantor venom is mediated by a lysosomal cell death pathway involving lysosomal membrane permeabilization and cathepsin B release. Am. J. Chin. Med. 41, 643–663 (2013).
Laustsen, A. H., Ainsworth, S., Lomonte, B., Kini, R. M. & Chávez-Olórtegui, C. Editorial: Novel immunotherapies against envenomings by snakes and other venomous animals. Front. Immunol. 11, 1004 (2020).
Descotes, J. & Choquet-Kastylevsky, G. Gell and Coombs’s classification: Is it still valid? Toxicology 158, 43–49 (2001).
Leon, G. et al. Pathogenic mechanisms underlying adverse reactions induced by intravenous administration of snake antivenoms. Toxicon 76, 63–76 (2013).
Suzuki-Matsubara, M., Athauda, S. B., Suzuki, Y., Matsubara, K. & Moriyama, A. Comparison of the primary structures, cytotoxicities, and affinities to phospholipids of five kinds of cytotoxins from the venom of Indian cobra, Naja naja. Comp. Biochem. Physiol. C Toxicol. Pharmacol. 179, 158–164 (2016).
Chan, Y. W., Tan, C. H., Heh, C. H. & Tan, K. Y. An immunoinformatic approach to assessing the immunogenic capacity of alpha-neurotoxins in elapid snake venoms. Front. Pharmacol. 14, 1143437 (2023).
Laustsen, A. H., Johansen, K. H., Engmark, M. & Anderson, M. R. Recombinant snakebite antivenoms: A cost competitive solution to a neglected tropical disease? PLoS Negl. Trop. Dis. 11, e0005361 (2017).
Ferreira, J. et al. Assessment of the neutralizing potency of ovine antivenom in a swiss mice model of Bothrops jararaca envenoming. J. Venom. Anim. Toxins incl. Trop. Dis. 11, 336–349 (2005).
Liu, B. S. et al. Identification of immunoreactive peptides of toxins to simultaneously assess the neutralization potency of antivenoms against neurotoxicity and cytotoxicity of Naja atra venom. Toxins. 10, 1–17 (2018).
Ros-Lucas, A., Bigey, P., Chippaux, J. P., Gascón, J. & Alonso-Padilla, J. Computer-aided analysis of West Sub-Saharan Africa snakes venom towards the design of epitope-based poly-specific antivenoms. Toxins. 14, 418 (2022).
Engmark, M. et al. Cross-recognition of a pit viper (Crotalinae) polyspecific antivenom explored through high-density peptide microarray epitope mapping. PLoS Negl. Trop. Dis. 11, e0005768 (2017).
Krause, K. E. et al. An interactive database for the investigation of high-density peptide microarray guided interaction patterns and antivenom cross-reactivity. PLoS Negl. Trop. Dis. 14, e0008366 (2020).
Smith, G. P. Filamentous fusion phage: Novel expression vectors that display cloned antigens on the virion surface. Science 228, 1315–1317 (1985).
Camperi, S. A. et al. Synthetic peptides to produce antivenoms against the Cys-rich toxins of arachnids. Toxicon: X. 6, 100038 (2020).
Palatnik-de-Sousa, C. B., Soares, I. S. & Rosa, D. S. Editorial: Epitope discovery and synthetic vaccine design. Front. Immunol. 9, 826 (2018).
Patronov, A. & Doytchinova, I. T-cell epitope vaccine design by immunoinformatics. Open Biol. 3, 120139; https://doi.org/10.1098/rsob.120139 (2013).
Ledsgaard, L. et al. Antibody cross-reactivity in antivenom research. Toxins. 10, 393 (2018).
Bermúdez-Méndez, E. et al. Innovative immunization strategies for antivenom development. Toxins (Basel). 10, 1–37 (2018).
Kao, D. J. & Hodges, R. S. Advantages of a synthetic peptide immunogen over a protein immunogen in the development of an anti-pilus vaccine for Pseudomonas aeruginosa. Chem. Biol. Drug Des. 74, 33–42 (2009).
da Silva, W. D. & Tamourgi, D. V. The humoral immune response induced by snake venom toxins. Inflamm. Allergy Drug Targets. 10, 343–357 (2011).
Misuan, N., Mohamad, S., Tubiana, T. & Yap, M. K. K. Ensemble-based molecular docking and spectrofluorometric analysis of interaction between cytotoxin and tumor necrosis factor receptor 1. J. Biomol. Struct. Dyn. 16, 1–15 (2023).
Peters, B., Bulik, S., Tampe, R., Van Endert, P. M. & Holzhütter, H. G. Identifying MHC class I epitopes by predicting the TAP transport efficiency of epitope precursors. J. Immunol. 171, 1741–1749 (2003).
Zhang, G. L., Petrovsky, N., Kwoh, C. K., August, J. T. & Brusic, V. PREDTAP: a system for prediction of peptide binding to the human transporter associated with antigen processing. Immunome Res. 2, 1–12 (2006).
Larsen, M. V. et al. An integrative approach to CTL epitope prediction: A combined algorithm integrating MHC class I binding, TAP transport efficiency, and proteasomal cleavage predictions. Eur. J. Immunol. 35, 2295–2303 (2005).
Engmark, M. et al. High-throughput immunoprofiling of mamba (Dendroaspis) venom toxin epitopes using high-density peptide microarrays. Sci. Rep. 6, 1–14 (2016).
Mukonyora, M. A review of important discontinuous B-cell epitope prediction tools. J. Clin. Cell Immunol. 6, 1–5 (2015).
Sanchez-Trincado, J. L., Gomez-Perosanz, M. & Reche, P. A. Fundamentals and methods for T- and B-cell epitope prediction. J. Immunol. Res. 2017, 1–15 (2017).
Tahir ul Qamar. et al. Epitope-based peptide vaccine design and target site depiction against Middle East Respiratory Syndrome Coronavirus: An immune-informatics study. J. Transl. Med. 17,1–14 (2019).
Vihinen, M., Torkkila, E. & Riikonen, P. Accuracy of protein flexibility predictions. Proteins 19, 141–149 (1994).
Siever, F. et al. Fast, scalable generation of high-quality protein multiple sequence alignments using Clustal Omega. Mol. Syst. Biol. 7, 539 (2011).
Goujon, M. et al. A new bioinformatics analysis tools framework at EMBL–EBI. Nucleic Acids Res. 38, 695–699 (2010).
McWillian, H. et al. Analysis tool web services from the EMBL-EBI. Nucleic Acids Res. 41, 597–600 (2013).
Vangone, A. et al. Sense and simplicity in HADDOCK scoring: Lessons from CASP-CAPRI round 1. Proteins 85, 417–423 (2017).
Das, A. A. et al. Elevated level of circulatory sTLT1 induces inflammation through SYK/MEK/ERK signalling in coronary artery disease. Clin. Sci. (Lond) 133, 2283–2299 (2019).
Ramirez, L. M., Shekhtman, A. & Pande, J. Hydrophobic residues of melittin mediate its binding to αA−crystallin. Protein Sci. 29, 572–588 (2019).
Chaudhari, R., Tan, Z. & Zhang, S. Overview of drug polypharmacology and multitargeted molecular design in Drug Disc. Tech. 2, 258–275 (2017).
Hevener, K. E. et al. validation of molecular docking programs for virtual screening against dihydropteroate synthase. J. Chem. Inf. Model. 49, 444–460 (2009).
Clementel, D. et al. RING 3.0: Fast generation of probabilistic residue interaction networks from structural ensembles. Nucleic Acids Res. 50, 651–656 (2022).
Morsa, D. et al. Multi-enzymatic limited digestion: The next-generation sequencing for proteomics? J. Proteome Res. 18, 2501–2513 (2019).
Opuni, K. F. M. Mass spectrometric epitope mapping. Mass Spectrom. Rev. 37, 229–241 (2018).
O’Leary, M. A., Maduwage, K. & Isbister, G. K. Use of immunoturbidimetry to detect venom-antivenom binding using snake venoms. J. Pharmacol. Toxicol. Methods 67, 177–181 (2013).
Ramirez, D. & Caballero, J. Is it reliable to take the molecular docking top scoring position as the best solution without considering available structural data? Molecules 23, 1–17 (2018).
Dubovskii, P. V. et al. Variability in the spatial structure of the central loop in cobra cytotoxins revealed by X-ray analysis and molecular modelling. Toxins. 14, 149 (2022).
Su, Z. Y. & Wang, Y. T. Coarse-grained molecular dynamics simulations of cobra cytotoxin A3 interactions with a lipid bilayer: Penetration of loops into membranes. J. Phys. Chem. B. 115, 796–802 (2011).
Dubovskii, P. V. et al. Impact of membrane partitioning on the spatial structure of an S-type cobra cytotoxin. J. Biomol. Struct. Dyn. 36, 3463–3478 (2018).
Konshina, A. G., Dubovskii, P. V. & Efremov, R. G. Stepwise insertion of cobra cardiotoxin CT2 into a lipid bilayer occurs as an interplay of protein and membrane “Dynamic Molecular Portraits”. J. Chem. Inf. Model. 61, 385–399 (2021).
Kanwa, N., De, S. K., Maity, A. & Chakraborty, A. Interaction of aliphatic amino acids with zwitterionic and charged lipid membranes: Hydration and dehydration phenomena. Phys. Chem. Chem. Phys. 22, 3234–3244 (2020).
Malik, S. & Debnath, A. Dehydration induced dynamical heterogeneity and ordering mechanism of lipid bilayers. J. Chem. Phys. 154, 174904 (2021).
Cevc, G. & Richardsen, H. Lipid vesicles & membrane fusion. Adv. Drug Deliv. Rev. 38, 207–232 (1999).
Røder, G., Kristensen, O., Kastrup, J. S., Buus, S. & Gajhede, M. Structure of a SARS coronavirus-derived peptide bound to the human major histocompatibility complex class I molecule HLA-B*1501. Acta Crystallogr. Sect. F. Struct. Biol. Cryst. Commun. 64, 459–462 (2008).
Clifford, J. N. et al. BepiPred-3.0: Improved B-cell epitope prediction using protein language models. Protein Sci. 31(12), e4497 (2022).
Larsen, J. E., Lund, O. & Nielsen, M. Improved method for predicting linear B-cell epitopes. Immunome Res. 2, 2 (2006).
Kolaskar, A. S. & Tongaonkar, P. C. A semi-empirical method for prediction of antigenic determinants on protein antigens. FEBS Lett. 276, 172–174 (1990).
Saha, S. & Raghava, G. P. S. Prediction of continuous B-cell epitopes in an antigen using recurrent neural network. Proteins 65, 40–48 (2006).
Saha, S. & Raghava, G. P. S. Prediction methods for B-cell epitopes. Methods Mol. Biol. 409, 387–394 (2007).
Ponoarenko, J. et al. ElliPro: A new structure-based tool for the prediction of antibody epitopes. BMC Bioinf. 9, 1–8 (2008).
Reynisson B., Alvarez B., Paul S., Peters B. & Nielsen M. NetMHCpan-4.1 and NetMHCIIpan-4.0: Improved predictions of MHC antigen presentation by concurrent motif deconvolution and integration of MS MHC eluted ligand data. Nucleic Acids Res. 48, 449–454 (2020).
Honorato, R. V. et al. Structural biology in the clouds: The WeNMR-EOSC ecosystem. Front. Mol. Biosci. 8, 729513 (2021).
van Zundert, G. C. P. et al. The HADDOCK2.2 web server: User-friendly integrative modeling of biomolecular complexes. J. Mol. Biol. 428, 720–725 (2016).
Pronk, S. et al. GROMACS 4.5: A high-throughput and highly parallel open source molecular simulation toolkit. Bioinformatics. 29, 845–854 (2013).
Abraham, M. J. et al. GROMACS: High performance molecular simulations through multi-level parallelism from laptops to supercomputers. Software X. 1–2, 19–25 (2015).
Waterhouse, A. M., Procter, J. B., Martin, D. M. A., Clamp, M. & Barton, G. J. Jalview Version 2-A multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191 (2009).
Webb, B. & Sali, A. Comparative protein structure modeling using Modeller. Curr. Protoc. Bioinf. 54, 1–55 (2016).
Acknowledgements
This study was supported by Royal Society of Tropical Medicine and Hygiene (RSTMH) Small Grants Program 2021.
Author information
Authors and Affiliations
Contributions
Y.M.K.K. conceptualize the study and design of experiments, H.J.J. and F.J.K.Y. conducted the experiments, Y.M.K.K., T.H.S., H.J.J. and F.J.K.Y. analyzed the results, H.J.J. and F.J.K.Y. writing the original draft, Y.M.K.K. and T.H.S. review and edit the manuscript. All authors approve the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Hiu, J.J., Fung, J.K.Y., Tan, H.S. et al. Unveiling the functional epitopes of cobra venom cytotoxin by immunoinformatics and epitope-omic analyses. Sci Rep 13, 12271 (2023). https://doi.org/10.1038/s41598-023-39222-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-39222-2
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.
