Evaluation of the Stx2 and Stx2-ribosomal P-stalk
It has been shown that the position of glutamate 259 (E259) likely blocks the interaction of Stx2 with SRL, rendering it catalytically inactive (Rudolph et al. 2020). We first studied the interaction between Y77 and E259 in the cryo-EM structure of Stx2A1 in complex with the native ribosomal P-stalk (PDB ID:7U6V). In this structure, Y77-E259 and Y114-E167 were located far from each other and did not exhibit polar interactions (Fig. 1A–1B). However, in Stx2 (PDB:7D6Q), these residues interacted via hydrogen bonding (Fig. 1C-1D). These results suggest that the interaction of Stx2 with the ribosome p-stalk blocks or inhibits Y77-E259 and Y114-E167 interactions.
Effect of shiga-toxin associated peptides on the Y77-E259 and Y114-E167 interactions
Docking analysis of the Stx2-peptides complexes revealed that out of the five known peptides, only one TVP (MAPPPRRRRA) blocked the interaction between the Y77 and E259 residues of Stx2A1 (Fig. 2 and Table 1). Therefore, it makes Stx2 catalytically active by exposing its active site (Y77), resulting in the inhibition of bacterial ribosomes. However, all peptides showed a close interaction with the Y77 and E259 residues of stx2A1 (data not shown), suggesting the role of interaction of peptides with these residues in STEC killing.
Table 1
Evaluation of the polar interaction of therapeutic peptides with Stx2
SN | Name | HPEPDOCK score | ClusPro score | Amino acids residues of Stx2 showing polar interaction with peptide | Whether blocking Y77-E259 or Y114-E167 interaction |
1 | WRWYCRR (Lino et al. 2011) | -229.232 | -964.8 | Y77, D94, Y114, E167, R170, F171, L200, P258, E259, Q261, I262 | No |
2 | TFNMWLPTFNQW (Li et al. 2016) | -206.687 | -877.7 | Y77, D94, W202, G203, R204, E259, N279 | No |
3 | MMARRRR (Watanabe-Takahashi et al. 2021) | -183.481 | -748.5 | Y77, D94, D111, S112, Y114, R170, L200, E259 | No |
4 | MAPPPRRRRA (Stearns-Kurosawa et al. 2011) | -182.587 | -576.5 | E72, Y77, V78, S112, G203, Q257, E259; Chain D-K7, D70 | Yes (Only block Y77-E259) |
5 | EWGRRMMGRGPGRRMMRWWR (Prada-Prada et al. 2020) | -213.327 | -900.1 | E72, Y77, Y114, T199, L200, G203, E259, Q261 | No |
The binding pocket of TVP was predicted by the CASTp 3.0 available online at http://cast.engr.uic.edu (Tian et al. 2018). The binding pocket is lined by chains A, D, and E of Stx2 (Fig. 3A-3B). The area and volume of the binding pocket was 381.6A°2 and 191.5A°3, respectively. Amino acids occupying the binding pocket are shown in Fig. 3C. TVP exhibited hydrogen bonding with E72, Y77, V78, S112, T115, Q257, and E259 in chain A and K7 and D70 in chain D., whereas hydrophobic interactions were formed by N74, N75, Y114, Q118, R125, V162, E167, R170, T199, L200, W202, L239, Q261, and I262 in chain A, and K7, I8, E9, K22, Q43, and G46 in chain D of Stx2 (Fig. 4A-4B).
The covariance matrix indicates that Y77 and E259 show anti-corelated motion (blue region), which indicates that Y77 and E259 are moving far from each other after interaction with the TVP (Fig. 5A). The Ramachandran plot showed that 92.9% and 7.1% of residues were in the favored and allowed regions, respectively. No residues were found in outlier regions (Fig. 5B). However, the overall G-factor value was slightly negative (-0.05), indicating residues in slightly unlikely conformations. The structure was also validated using another structure verification server called Errat, which showed an overall quality factor of 83.36 for Stx2-TVP. In ERRAT, an error value exceeding the 99% confidence level indicates a poorly modelled region. For both Y77 and E259, the error values were below the 50% confidence level, indicating that these residues are present in the optimal model region of the Stx2-TVP complex (Fig. 6).
Effect of TVP on the ribosome binding sites of Stx2
PyMOL was used to superimpose the structures of Stx2-TVP and Stx2-ribosomal P-stalk (Fig. 7A). The region near the ribosomal binding site (R172, R176, and R179) of Stx2 was also studied (Fig. 7B). Structural analysis of the Stx2-TVP complex demonstrated that two residues (R170 and F171) located near ribosomal binding sites (RBS) were present in the loop of the A subunit of Stx2 (Fig. 7C). Loops containing R170 and F171 are present between the two alpha helices (helix-loop-helix, HLH). HLH functions as a transcription factor that binds DNA and regulates gene expression (Massari and Murre 2000; Obata et al. 2022). However, in the case of Stx2-ribosomal P-stalk, these residues were located in the α-helix region of Stx2A1 (Fig. 7D).
Secondary Structure Analysis
The secondary structure of TVP was predicted using SOPMA server (Geourjon and Deleage 1995). It predicts the secondary structure based on the availability of amino acid sequences in the databases. The results showed that TVP was composed of 100% random coils (Fig. 8A). The Ramachandran plot showed that 60% and 40% of residues were in the favored and allowed regions, respectively. However, none of the residues were present in the outlier region. According to the PROCHECK analysis, the overall G-factor of TVP was − 0.55. The G-factor predicts the distance of each residue from the normal region of a plot. Negative G-factors indicate that the residues are not in the likely conformations.
Refinement of tertiary structure of TVP
The tertiary (3D) structure of the TVP was refined using the GalaxyRefine server. The accuracy of the model is defined by the GDA-HA, MolProbity, and clash scores. MolProbity score highlights steric clashes. Five models were generated using GalaxyRefine, of which Model 1 was selected because of its conformational properties. Additionally, for the refined model, GDT-HA was 0.8750, whereas MolProbity, clash score, and poor rotamers were 1.692, 15.5, and 0.0, respectively. The GDT-HA, MolProbity, clash score, and poor rotamers for the original model were 1, 1.742, 6.6, and 0, respectively. When compared with the other models, Model 1 was predicted to be the best refined model based on different scores. Moreover, the Ramachandran plot showed that 100% of the residues were in the favored region of the three-dimensional structure (Fig. 8B). The G-factor for this model was + 0.28, indicating residues in likely conformations. Superimposition of the refined model with the original TVP showed differences in structural configuration (Fig. 8C). No significant changes were observed for the RMSD; however, the Rg was less for the refined model of TVP, suggesting more compactness of the refined structure (Fig. 8D-8E).
Additionally, the QMEAN values (blue region) of the refined model are higher (0.67) than those of the original model (0.40), which indicates that the model scores are higher than those of the original model on average. In the ANOLEA profile, the initial model had many areas of high energy, which were greatly improved in the refined model, suggesting greater reliability (Fig. 9A–9B).
Physiochemical properties of tetravalent peptide
TVP has 174 atoms and its molecular weight was found to be 1.21 kDa. Further analysis showed that TVP had an isoelectric point (pI) value of 12.48, indicating its basic nature. The positive charge of TVP (+ 4) denotes its binding affinity for the negatively charged bacterial membranes. The amino acid composition showed the amphipathic nature of TVP, with four polar and six non-polar residues. Moreover, the membrane position of the TVP is globular, indicating its natural folding nature. These results suggest that TVP possesses characteristic properties that are essential for its effectiveness as an effective antimicrobial peptide. However, low aliphatic index, and high instability index suggests that TVP is thermally unstable. Moreover, the hydrophobicity index and grand average of hydropathicity (GRAVY) were negative, indicating its hydrophilic nature (Table 2).
Table 2
Physiochemical attributes of TVP
Property | Value |
Number of amino acids | 10 (Tiny = 2, Small = 5, Aliphatic = 2, Polar = 2, non-polar = 6, charged = 4, Basic = 4) |
Atomic composition | Carbon: 50 Hydrogen:90 Nitrogen:22 Oxygen:11 Sulfur:01 |
Total number of negatively charged residues (Asp + Glu) | 0 |
Total number of positively charged residues (Arg + Lys) | 4 |
Formula | C50H90N22O11S1 |
Extinction coefficients | 0 (will not be visible by UV spectrophotometry) |
Estimated half-life | > 30 hours (mammalian reticulocytes, in vitro and > 10 hours (Escherichia coli, in vivo) |
Instability index | 243.42 |
Aliphatic index | 20 |
Grand average of hydropathicity (GRAVY) value | -1.730 |
Allergenicity, Toxicity, and antigenicity analysis
Effective antimicrobial peptides do not have any allergenicity. No hits were found against the database of 2890 allergen representative peptides (ARPs), suggesting that all the five peptides were non-allergenic in nature. Toxin Pred analysis indicated the non-toxic nature of therapeutic peptides. Of the five peptides, only three had antigenic scores > 0.4, suggesting their protective antigenicity (Table 3) (Doytchinova and Flower 2007).
Table 3
Antigenicity, allergenicity, toxicity and physiological attributes of therapeutic peptides
SN | Name | Antigenicity score | Allergenicity | Toxicity | Boman Index | GRAVY score |
1 | WRWYCRR (Lino et al. 2011) | 1.62 | No | No | 5.7 | -2 |
2 | TFNMWLPTFNQW (Li et al. 2016) | 1.39 | No | No | 0.5 | -0.3 |
3 | MMARRRR (Watanabe-Takahashi et al. 2021) | -0.77 | No | No | 7.6 | -1.8 |
4 | MAPPPRRRRA (Stearns-Kurosawa et al. 2011) | -0.0045 | No | No | 5.4 | -1.7 |
5 | EWGRRMMGRGPGRRMMRWWR (Prada-Prada et al. 2020) | 0.47 | No | No | 4.6 | -1.7 |
Effect of antibiotics on the Y77-E259 and Y114-E167 interactions
Of the three antibiotics (active ingredients of FDA-approved peptides), only oritavancin diphosphate interacted with Y77, E259, and Y114 (Table 4). The same glycopeptide antibiotic exhibited the highest binding affinity. These analyses suggest a possible role of this antibiotic in the inhibition of STEC growth.
Table 4
Evaluation of the polar interaction of antibiotics with Stx2
SN | Name | CID number | Binding affinity | Amino acids residues of Stx2 showing polar interaction with antibiotics | Whether blocking Y77-E259 or Y114-E167 interaction |
1 | Oritavancin diphosphate (Chen and Lu 2020) | 53297457 | -14 | Y77, Y114, R125, L200, E259 | No |
2 | TELAVANCIN HYDROCHLORIDE (Chen and Lu 2020) | 9812715 | -14 | Y114, I271 | No |
3 | Daptomycin (Chen and Lu 2020) | 16134395 | -11 | D94, S112, R204, Q261, I262, | No |