In silico prospection of receptors associated with the biological activity of U1-SCTRX-lg1a: an antimicrobial peptide isolated from the venom of Loxosceles gaucho

Abstract

The emergence of antibiotic-resistant pathogens generates impairment to human health. U1-SCTRX-lg1a is a peptide isolated from a phospholipase D extracted from the spider venom of Loxosceles gaucho with antimicrobial activity against Gram-negative bacteria (between 1.15 and 4.6 μM). The aim of this study was to suggest potential receptors associated with the antimicrobial activity of U1-SCTRX-lg1a using in silico bioinformatics tools. The search for potential targets of U1-SCRTX-lg1a was performed using the PharmMapper server. Molecular docking between U1-SCRTX-lg1a and the receptor was performed using PatchDock software. The prediction of ligand sites for each receptor was conducted using the PDBSum server. Chimera 1.6 software was used to perform molecular dynamics simulations only for the best dock score receptor. In addition, U1-SCRTX-lg1a and native ligand interactions were compared using AutoDock Vina software. Finally, predicted interactions were compared with the ligand site previously described in the literature. The bioprospecting of U1-SCRTX-lg1a resulted in the identification of three hundred (300) diverse targets (Table S1), forty-nine (49) of which were intracellular proteins originating from Gram-negative microorganisms (Table S2). Docking results indicate Scores (10,702 to 6066), Areas (1498.70 to 728.40) and ACEs (417.90 to – 152.8) values. Among these, NAD + NH3-dependent synthetase (PDB ID: 1wxi) showed a dock score of 9742, area of 1223.6 and ACE of 38.38 in addition to presenting a Normalized Fit score of 8812 on PharmMapper server. Analysis of the interaction of ligands and receptors suggests that the peptide derived from brown spider venom can interact with residues SER48 and THR160. Furthermore, the C terminus (– 7.0 score) has greater affinity for the receptor than the N terminus (– 7.7 score). The molecular dynamics assay shown that free energy value for the protein complex of – 214,890.21 kJ/mol, whereas with rigid docking, this value was – 29.952.8 sugerindo that after the molecular dynamics simulation, the complex exhibits a more favorable energy value compared to the previous state. The in silico bioprospecting of receptors suggests that U1-SCRTX-lg1a may interfere with NAD + production in Escherichia coli, a Gram-negative bacterium, altering the homeostasis of the microorganism and impairing growth.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40203-024-00190-8.

Introduction

Loxosceles sp. (Araneae, Sicariidae) belongs to the subphylum Chelicerata, one of the three evolutionary lineages of arthropods. Systematically, within this subphylum, the class Arachnida comprises most of the Chelicerata, including forms such as spiders, scorpions, mites, and ticks (Ruppert and Barnes 1996).

Brown spiders are a group of spiders that produce venoms with human clinical manifestations. This venom has been studied for at least 60 years in different research groups worldwide. The extraction and characterization of this venom was motivated by several cases related to loxoscelism, injury caused by spider bites (Chaves-Moreira et al. 2017).

Three main toxin families are present in spider venom: phospholipases-D, astacin-like metalloproteases, and inhibitor cystine knot (ICK) peptides. Additionally, serine proteases, serpins, hyaluronidases, venom allergens, and translationally controlled tumor protein (TCTP) are also present. These toxins have essential biological properties that enable them to interact with a range of distinct molecular targets. Therefore, this toxin can be a source of bioactive molecules for use in the pharmaceutical industry (Chaves-Moreira et al. 2019).

The emergence of new multiresistant microorganisms is increasing each year, which can increase the risks of mortality and morbidity, consequently overloading public health systems and causing financial losses for countries stricken. Otherwise, the low efficiency of traditional medicines and treatments against these organisms requires research for the development of new bioactive molecules to control diseases caused by microorganisms (Ferri et al. 2017).

U1-UCRTX-lg1a (VGTDFSGNDDISDVQK) is an anionic peptide derived from phospholipase-D isolated from spider L. gaucho venom. This peptide showed antibacterial activity against Escherichia coli, Pseudomonas aeruginosa, Enterobacter cloacae presented minimal inhibitory concentrations between 2 and 5 µM. In this way, activity was only observed in gram-negative bacteria. Furthermore, in an experiment with human erythrocytes, hemolytic activity was not observed with U1-SCRTX-lg1a at within the inhibitory concentrations (Segura-Ramírez and Silva Júnior 2018). However, your active mechanism and molecular targets are not well elucidated, considering the breadth and diversity of antimicrobial peptide targets and their mechanisms of action (Li et al. 2022).

Virtual ligand screening and molecular docking are computational methods to identify protein targets and interactions. Among them, those methods can be used for the initial development of new drugs; likewise, simulations have been performed using many parameters at the same time since they are faster and more efficient than in vitro experiments (Glaab 2016). In this process, there are steps: protein and ligand preparation, creation of molecular models, molecular docking, analysis, and visualization of the results (Muniz Seif et al. 2023). In contrast, computational biology requires a high processing rate; however, this barrier has been superseded due to computational advancement and the emergence of online servers to perform analysis (Duhovny et al. 2002).

Therefore, the aim of this study was to use free bioinformatics tools to perform virtual screening and identify potential receptors associated with the antimicrobial activity of the U1-SCRTX-lg1a peptide and to further describe those ligand-receptor interactions.

Methods

Peptide characterization and minimum free energy

It was used the Heliquest server (https://heliquest.ipmc.cnrs.fr/) (Gautier et al. 2008), to determine physicochemical parameters (Net charge, hydrophobicity moment and molecular weight) of the U1-SCRTX-lg1petide (VGTDFSGNDDISDVQK). Theoretical pI, Instability index, grand average of hydropathicity (GRAVY) were obtained by the Expasy Server (https://web.expasy.org/cgi-bin/protparam/) (Gasteiger et al. 2005) and Pepcalc server (https://pepcalc.com/). To build the peptide tridimensional structure, free energy mini it was used by server I-TASSER (https://zhanggroup.org/I-TASSER/) (Zheng et al. 2021).

The state of minimum free energy has been obtained using UCSF Chimera software (https://www.cgl.ucsf.edu/chimera/) (Pettersen et al. 2004) the following settings were used: steepest descendent steps (100,000); steepest descendent steps size Å (0.02); conjugate gradient steps (10), conjugate gradient steps size Å (0.02), update interval (10), fixed atoms (none) after H, H-B, charges, and SR (amber force field ff14SB).

Search of potentials Gram-negative targets

To identify potential receptors for U1-SCRTX-lg1a, it was realized by PharmMapper server (http://www.lilab-ecust.cn/pharmmapper/) (Wang et al. 2017a, b). The peptide structure was submitted using the following parameters: generate conformers (yes); maximum generated conformations (300); full/complete pharmacophore mapping; all targets selected (v2010, 7302), and number of reserved matched targets (300). Those results were ranked by normalized fit score. All targets were classified based on origin and catalytic activity (https://www.uniprot.org) in the microorganism. It selected a total of forty-nine proteins derived from Gram-negative bacteria, without mutations in their models, those structures were provided by the Protein Data Bank (PDB) server (https://www.rcsb.org/).

Target and peptide molecular docking

Molecular docking between U1-SCRTX-lg1a and the selected targets it was performed by software PatchDock (https://bioinfo3d.cs.tau.ac.il/PatchDock/) (Schneidman-Duhovny et al. 2005). The PatchDock parameters were set to clustering RMSD (4.0) and complex type (default). The docking results were ranked based on the major score value to determine the optimal binding of U1-SCRTX-lg1a. It was collected also AREA, ACE (effective atomic contact energies), and transformation values. The docking results was classified by the major score value. In addition, the better target found in PatchDock was also performed by AutoDock vina (Trott and Olson 2010) to examined affinity of C-terminal, N-terminal segments and native ligand.

Analysis of ligands and receptors interaction

The U1-SCRTX-lg1a and receptors interaction was analyzed using UCSF Chimera software (https://www.cgl.ucsf.edu/chimera/) (Pettersen et al. 2004). The find bond tool was used with a relaxed constraint binding (2 Å and 20°) to examine the interaction. It was considered in this study only hydrogen bonds with distance less than 4 Å between heavy atoms. Furthermore, information on residue electron donors and acceptors was collected.

Comparison of predicted ligand sites

Receptor ligand sites were predicted using the PDBsum server (http://www.ebi.ac.uk/thornton-srv/databases/pdbsum/). For this, was used the Uniprot code (https://www.uniprot.org/) and the FASTA sequence of better scores found in molecular docking analysis. The results were compared to the ligand site obtained by molecular docking.

Molecular dynamics simulation

Molecular dynamics simulations were performed between PDBid:1wxi as a receptor and U1-SCRTX-lg1a as a ligand using Chimera 1.6 software. The simulations utilized the TIP3BOX solvent model, a 1:1 ratio, and 100 steps from steepest descent with a size of 0.02 Å to steepest descent and conjugate gradient steps with a size of 0.02 Å. Intermodal and intramodal hydrogen bonds were identified with constraints relaxed at 2 angstroms and 20 degrees.

Results and discussion

Peptide characterization and minimum free energy

The physicochemical analysis using Heliquest, Expasy, and Pepcal servers showed that the U1-SCTRX-lg1a peptide has a molecular weight of 1695.7 g/mol, a net charge of – 3, and a hydrophobic index/hydrophobic moment of 0.083/0.189. Energy minimization of the peptide structure resulted in a free energy of – 1502.51 kJ/mol (Fig. 1). The peptide contains four acidic, one aromatic, one alkaline, three aliphatic, and five polar amino acids. Aspartate has a negatively charged R group, phenylalanine has an aromatic side chain with relatively hydrophobic characteristics, lysine is positively charged and hydrophilic, and valine and isoleucine tend to group together inside proteins, stabilizing the protein structure through hydrophobic interactions. Serine, threonine, asparagine, and glutamine are more water-soluble amino acids that contain functional groups capable of forming hydrogen bonds (Nelson and Cox 2022).

Fig. 1.

Characterization and physicochemical properties of U1-SCRTx-lg1a. A Peptide helicoidal projection top view, yellow represents nonpolar residues, blue represents basic residues, pink represents asparagine and glutamine, gray represents alanine and glycine, red represents aspartic acid, lilac represents serine and tyrosine. The arrow indicates the hydrophobic portion of this projection. B Three-dimensional structure designed by UCSF software relative to the position of each amino acid residue. C The hydrophobicity is represented, red represents acidic, light green aromatic, blue basic, gray aliphatic and black polar. D Sequence peptide primary structure. The table at the end shows the amino acid sequence, molecular weight, NC (net charge), H/μH (hydrophobicity index/hydrophobic moment), Theoretical isoelectric point (pl), Instability index (estimate of the stability of peptide in a test tube) and GRAVY (value for a peptide in calculated as the sum of hydropathy values of all the amino acids, divide by the number of residues in the sequence). Heliquest server

The peptide presented a theoretical mass of approximately 1695.75 g/mol and is relatively small. The model obtained from the I-TASSER server showed a confidence score (C-score) of – 0.90, which was used to estimate the quality of predicted models. C-scores range from [– 5, 2], where a higher C-score indicates a more confident model (Yang and Zhang 2015). The best model was used for target search and molecular docking.

Physicochemical properties as charged, hydrophobic moment and approximately residues number are usually found in membranolytic antimicrobial peptides, there have membrane injury as the mainly activate mechanism due potential difference between membrane and peptide (Benfield and Henriques 2020).

In contrast, hemolytic activity was not observed with U1-SCRTX-lg1a at inhibitory concentrations. Therefore, this peptide uses non-membranolytic activity mechanisms to inhibit bacterial growth. Thus, the antimicrobial activity occurs for the interacting with intracellular targets, promoting homeostasis impairment and cell death. This mechanism particularly is found in non-charged or low-charged peptides, exampling: Doderlin (Silva et al. 2023), Rondonin (Riciluca et al. 2012) and Crotamine (Mas et al. 2019).

Potential Gram-negative target identification

The first step of bioprospecting receptors for U1-SCRTX-lg1a was determined by the PharmMapper Server. This is an online tool that uses pharmacophore mapping techniques to identify potential drug targets that can be used for virtual screening. The tool has a database of more than 7000 proteins, built based on protein information and pharmacophore models. When inserting a molecule, the tool generates a ranked list of proteins based on the similarity of their pharmacophores. This enables the identification of possible drug targets and the development of new compounds with therapeutic potential (Liu et al. 2010).

PharmMapper search resulted in 300 general targets (Table S1), among them 49 originated from Gram-negative targets (Table S2), however in this research only 10 better scores based on PatchDock results (Table 1). The PharmMapper results were normalized to fit the score between 8.216 and 9.998. The better receptors found using this method was peptidyl-dipeptidase dcp [PDB_id: 1y79], phosphoenolpyruvate carboxylase [PDB_id: 1jqn] and thermoresistant gluconokinase [PDB_id: 1ko8], whole was originated from Escherichia coli.

Table 1.

The 10 targets originated from Gram-negative microorganisms resulting from PharmMapper based on molecular docking

PM Rank	ID PDB	Normalized fit score	Target name	Catalytic activity	Origin
41	1y79	9.782	Peptidyl-dipeptidase dcp	Hydrolysis of unblocked	Escherichia coli
74	1jqn	9.592	Phosphoenolpyruvate carboxylase	Oxaloacetate + phosphate = hydrogencarbonate + phosphoenolpyruvate	Escherichia coli K12
81	1ko8	9.558	Thermoresistant gluconokinase	ATP + D-gluconate = 6-phospho-D-gluconate + ADP + H+	Escherichia coli K12
113	1m2x	9.433	Metallo-beta-lactamase type 2	A beta-lactam + H2O = a substituted beta-amino acid	Elizabethkingia meningoseptica
155	1hmu	9.188	Chondroitinase-AC	Eliminative degradation of polysaccharides containing 1,4-beta-D-hexosamine and 1,3-beta-D-glucuronosyl linkages to disaccharides containing 4-deoxy-beta-D-gluc-4-enuronosyl groups	Pedobacter heparinus
215	1wxi	8812	NH3-dependent NAD synthetase	ATP + deamido-NAD+  + NH4+  = AMP + diphosphate + H+  + NAD+	Escherichia coli K12
260	1k4m	8.513	Nicotinate-nucleotide adenylyltransferase	ATP + H+  + nicotinate beta-D-ribonucleotide = deamido-NAD+  + diphosphate	Escherichia coli K12
271	1geg	8.412	Diacetyl reductase [(S)-acetoin forming]	(S)-acetoin + NAD+  = diacetyl + H+  + NADH	Klebsiella pneumoniae
291	1kp2	829	Argininosuccinate synthase	ATP + L-aspartate + L-citrulline = 2-(N(omega)-L-arginino) succinate + AMP + diphosphate + H+	Escherichia coli K12

The rank of targets was ordered by major normalized fit score

ID PDB Identification code in protein data bank, PM Rank Rank in PharmMapper server, Normalized Fit Score value obtained by ratio of fit score and number of features, Catalytic activity obtained from uniprot.org, Origin Specie from the target was isolated, None catalytic activity not found

U1-SCTRX-lg1a peptide exhibited antimicrobial activity against Escherichia coli strains SBS636 and D31 with a minimum inhibitory concentration of 4.6 μM (Segura-Ramírez and Silva Júnior 2018).

Target and peptide molecular docking

The PatchDock is a software that utilizes a docking technique based on complementarity principles to model the molecular docking between two proteins. This allows for the simulation of protein docking in different contexts, molecular docking is a technique used to study the interaction between molecules, especially between a protein and a ligand, with the aim of predicting their binding affinity (Schneidman-Duhovny et al. 2005.

Molecular docking was performed by PatchDock for all 49 targets originated from Gram-negative organisms (Table S3), however, in this work only 10 better docking score targets (Table 2).

Table 2.

Shows the results of rigid docking obtained from the PatchDock

PM rank	ID PDB	Receptor name	Score	Area	ACE (kj/mol)
41	1y79	Peptidyl-dipeptidase dcp	10,702	1296.30	91.16
260	1k4m	Nicotinate-nucleotide adenylyltransferase	10,126	1256.60	417.90
74	1jqn	Phosphoenolpyruvate carboxylase	9974	1292.40	212.58
289	1h1l	Nitrogenase molybdenum-iron protein alpha chain	9924	1446.50	265.41
113	1m2x	Metallo-beta-lactamase type 2	9748	1285.10	357.24
215	1wxi	NH3-dependent NAD+ synthetase	9742	1223.6	38.39
81	1ko8	Thermoresistant gluconokinase	9556	1317.90	77.66
271	1geg	Acetoin(diacetyl) reductase	9534	1188.80	218.18
291	1kp2	Argininosuccinate synthase	9358	1488.70	367.27
155	1hmu	Chondroitinase-AC	9340	1185.90	214.19

U1-SCRTX-lg1a was used as the ligand, and the receptor target obtained by PharmMapper was used. The table is arranged according to the major scores

ID PDB Identification code in protein data bank, PM Rank Rank in PharmMapper server, Score Geometric shape complementarity score, Area Approximate interface area of the complex, ACE Atomic effective contact energy

The docking results indicate values score (10,702 to 6066), Area (1498.70 to 728.40), and ACE (417.90 to – 152.8). The ranking of receptors observed in PatchDock differs from that in the PharmMapper search. Among all proteins studied, the highest dock score was found for Peptidyl-dipeptidase dcp [PDB_id: 1y79], nicotinate-nucleotide adenylyltransferase [PDB_id: 1k4m], phosphoenolpyruvate carboxylase [PDB_id: 1jqn], nitrogenase molybdenum-iron protein alpha chain [PDB_id: 1h1l], metallo-beta-lactamase type 2 [PDB_id: 1m2x], and NH₃-dependent NAD⁺ synthetase [PDB_id: 1wxi] (Table 2).

Analysis of ligands and receptors interaction

Receptor-ligand interactions analysis plays a significant role in all biological processes, and computational tools are used to simulate these biological phenomena, such as three-dimensional structural modeling and molecular docking between ligands and receptors. Rigid docking is also a crucial tool in computational drug design, which enables efficient identification of connections between rigid molecules without any initial restrictions on position or orientation. The use of this approach enables a faster and more efficient exploration of optimal solutions, significantly reducing the computation time required to identify the correct binding configurations (Duhovny et al. 2002).

To better understand the docking analysis, we studied only stronger hydrogen bonds (≤ 4 Å) between U1-SCRTX-lg1a and the receptor. The H bonds of six highest docking scores are summarized in Table 3. These results showed that peptide as a donor or acceptor electrons in hydrogen bond interactions.

Table 3.

Hydrogen bonds and distance (≤ 4 Å) for the six highest scores receptors and U1-SCRTX-lg1a obtained by a docking result by PatchDock

PM rank	PDB ID	Receptor name	Energy	Donor	Acceptor	D-A distance (Â)
			Kg/mol
41	1y79	Peptidyl-dipeptidase dcp	– 75,558.81	ARG 141.A NH1	SER 6.LOG	2.008
				ARG 141.A NH2	ASP 9.L OD1	3.966
				ARG 268.A NH2	GLN 15.L O	3.814
				ASN 428.1A ND2	ASP 4.L OD1	2.014
				ASN 491.A ND2	ASP 4.L O	3.633
				ASN 8.L ND2	THR 69.A OG1	3.941
				ASP 9.L N	TYR 106.A OH	3.11
260	1k4m	Nicotinate-nucleotide adenylyltransferase	– 55,867.29	PHE 5.L N	ASP 66.C O	3.908
				LEU 69.A N	GLN 15.L OE1	3.932
				LYS 21.B NZ	ASP 10.L OD1	3.424
				LYS 67.B NZ	SER 6.L OG	3.824
74	1jqn	Phosphoenolpyruvate carboxylase	– 91,386.58	ASN 8.L N	GLN 532.A	2.754
				ASP 9.L N	GLN 532.A	2.303
				GLN 15.L NE2	ALA 470.A	2.047
				LYS 16.L NZ	GLU 574.A	3.506
				ARG 244.A	VAL 14.L	3.167
				ARG 244.A	VAL 14.L	2.523
				LYS 384.A	ASP 4.L	2.575
				ARG 392.A	ILE 11.L	3.406
289	1h1l	Nitrogenase molybdenum-iron protein alpha chain	– 201,573.34	VAL 1.L N	SER 9.D OG	3.673
				LYS 361.B NZ	VAL 1.L O	2.483
				LYS 6.D NZ	ASP 4.L O	3.789
				ASN 8.D ND2	ASP 10.L OD1	3.998
113	1m2x	Metallo-beta-lactamase type 2	– 66,967.77	LYS 16.L NZ	LYS 66.B O	3.493
				LYS 16.L NZ	ASP 227.B OD2	3.724
				LYS 43.B NZ	SER 6.L OG	2.29
				LYS 43.B NZ	ASP 10.L OD1	1.414
				LYS 66.B NZ	SER 12.L OG	3.311
				LYS 43.D NZ	ILE 11.L O	3.921
				LYS 66.D N	PHE 5.L O	2.528
				SER 225.D OG.A	VAL 1.L O	3.83
				TRP 277.D N	GLN 15.L OE1	3.674
215	1wxi	NH3-dependent NAD+ synthetase	– 29,952.81	SER 6.L N	ASP 223.A OD2	2.89
				GLY 7.L N	ASP 223.A OD2	3.796
				SER 48.A OG	ASP 9.L O	3.331
				GLN 51.A N	ASP 9.L OD1	3.961
				ASN 136.A ND2	VAL 1.L O	2.571
				ARG 142.A NH2	GLN 15.L OE1	0.646
				THR 160.A OG1	ASP 13.L OD1	2.679
				THR 172.A N	VAL 14.L O	3.869
				ASP 176.A N	VAL 14.L O	3.998
				ASP 223.A N	ASP 10.L OD1	1.815

Donor electron donor residue, Acceptor electron acceptor residue, D-A distance distance between heavy atoms of electron acceptor and donor, THR 3.L N Residue name/ Number/ Chain ID/ Shared electron atom, L Ligand ID

The rigid docking using PatchDock produced the following interactions: Peptidyl-dipeptidase Dcp from Escherichia coli with 7 bonds; phosphoenolpyruvate carboxylase from Escherichia coli K12 with 8 bonds; The NH₃-dependent NAD⁺ synthetase from Escherichia coli K12 showed the highest number of hydrogen bonds (10 bonds) (Fig. 2).

Fig. 2.

A NH(3)-dependent NAD(+) synthetase (dark gray) and U1-SCRTX-lg1a (red) formed a complex after molecular docking. B Ten hydrogen bonds (blue) were identified between the acceptor and ligand, with distances less than 4 angstroms. Chimera 1.6 tools

Ligand binding site analysis and molecular dynamic simulation

Comparison between predicted binding sites from the PDBsum and those observed using Chimera 1.6 software, revealed that U1-SCRTX-lg1a interaction may be colocalized or closely residues of ligand sites of each receptor found in this work.

Among the models with the six receptor highest dock scores, NH₃-dependent NAD⁺ synthetase showed the most satisfactory peptide interaction. It revealed 4 closely and two colocalized residues of active site. Although, Peptidyl-dipeptidase dcp and Phosphoenolpyruvate carboxylase both presented only one close binding site, the other 3 receptors did not present interactions with active sites (Table 4).

Table 4.

Study of receptor binding location prediction using PDBsum server compared to PatchDock docking analysis and molecular dynamics simulation

PM Rank	PDB ID	Sites with interactions involving ligands (PDBsum)	Ligand binding site residues (Patch Dock)	Molecular Dynamics Simulation
41	1y79	426(a), 469(a), 473(a), 498(a), 593(a), 594(a) 601(a), 607(a), 611(a), 614(a), 700(a), 702(a)	69(a), 106(a),141(a), 268(a), 428(a), 491(a)	NONE
260	1k4m	11(a), 12(a), 40(a), 45(a), 46(a) 85(a), 107(a), 109(a), 110(a), 118(a), 134(a), 177(a), 179(a), 181(a), 182(a), 185(a)	69(a), 21(b), 67(b), 66(c)	NONE
74	1jqn	396(a), 506(a), 543(a), 587(a), 699(a), 773(a), 832(a), 881(a), 901(a), 902(a)	244(a), 384(a), 392(a), 470(a), 532(a), 574(a)	NONE
289	1h1l	61(a), 87(a), 95(a), 153(a), 190(a), 194(a), 273(a), 355(a), 440(a), 1479(a), 1480(a), 68(b), 93(b), 106 (b), 107(b), 151(b), 349(d), 353(d)	361(b), 6(d), 8(d), 9(d)	NONE
113	1m2x	116(a), 118(a), 119(a), 120(a), 167(a), 196(a), 237(a), 221(a), 263(a), 285(a), 288(a), 811(a), 901(a), 902(a)	43(b), 66(b), 227(b), 43(d), 66(d), 225(d), 277(d)	NONE
215	1wxi	46(a), 48(a), 52(a), 53(a), 82(a), 88(a), 140(a), 173(a), 160(a), 165(a), 189(a), 400(a), 500(a), 600(a), 700(a),	223(a), 48(a), 51(a), 136(a), 142(a), 160(a), 172(a), 176(a)	48(a), 51(a), 90(a), 142(a), 146(a), 172(a), 223(a), 225(a), 226(a)

The bold residue is shared among binding sites, and the underlined residues are those near the binding amino acids

ID PDB Identification code in protein data bank, PM Rank Rank in PharmMapper server, 86(a) Residue number/Chain ID)

To better elucidate these interactions, molecular dynamics simulation was performed only between the peptide and NH3-dependent NAD + synthetase receptors. This analysis resulted in an energy value for the protein complex of – 214,890.21 kJ/mol, whereas with rigid docking, this value was – 29.952.81. This result suggests that after the molecular dynamics simulation, the complex exhibits a more favorable energy value compared to the previous state (Fig. 3).

Fig. 3.

A NH(3)-dependent NAD(+) synthetase and U1-SCRTX-lg1a (center) formed a complex after dynamic molecular simulation and solvent (withe): energy free -214,890.21 kJ/mol. B Focus on U1-SCRTX-lg1a. C Surface hydrophobicity to the interaction between the receptor and ligand. Chimera 1.6

NAD⁺ synthetase [1xwi] from E. coli, catalyzes ATP-dependent starch-NAD amidation to form NAD⁺. NAD plays roles in processes as diverse as calcium mobilization, DNA repair, and post-translational modification of proteins in eukaryotes (Jauch et al. 2005).

Molecular docking showed that U1-SCTRX-lg1a binds to SER₄₈ and THR₁₆₀ of the chain occupying an important active site for this protein. Therefore, this interaction suggests that U1-SCTRX-lg1a may compete by the ligand site of this receptor and can reducing its enzymatic activity, consequently modifying intracellular functions as nucleic acids metabolism, altering homeostasis resulting in Gram-negative bacteria death.

AutoDock vina analysis

The application of AutoDock Vina allowed a more detailed analysis of the protein–ligand interactions, providing valuable information about the binding affinity and the main amino acid residues involved in the interaction. A higher score represents a more favorable interaction between molecules (Trott and Olson 2010).

The models of the ligands indicated in the literature were obtained for comparison between their docking and the two lateral portions of the U1-SCRTX-lg1 toxin derived from Loxosceles gaucho venom.

The results suggest that the C-terminal of the U1-SCRTX-lg1a peptide exhibits a more favorable interaction with the 1wxi receptor compared to the N-terminus. Furthermore, the result showed that the binding between the NH₃-dependent NAD⁺ synthetase receptor and the diphosphate ligand (O7P24) which is maintained through stacking interactions involving ARG₁₄₂, ILE₄₇ and SER₄₈ with the adenine ring of the molecule (Jauch et al. 2005) was the one with the highest score. This connection was found in the literature, in docking with PatchDock, in molecular dynamics simulation and most favorably in Autodock vina (Table 5).

Table 5.

Interaction results between 1wxi and U1-SCRTX-lg1a using AutoDock vina sorted from lowest score to highest score

Score	Ligand	RMSD l.b./u.b
– 7.7	U1-SCRTX-lg1a—N term (VGTDFSGNDD)	0.0
– 7.4	Adenosine Monophosphate (C10H14N5O7P)	0.0
– 7.1	U1-SCRTX-lg1a—C term (GNDDISDVQK)	0.0
– 4.7	Diphosphate (O7P24)	0.0

Score measure used in the molecular docking process to assess the quality of the interaction between two molecules, Ligand Ligand name, RMSD l.b./u.b. Root Mean Square Deviation, range of RMSD values that are considered acceptable for assessing process accuracy. The l.b. (lower bound) refers to the lower bound of the range, while the u.b. (upper bound) refers to the upper bound

Furthermore, in crystalline by Jauch et al. (2005), the ribose portions of the adenosine monophosphate nucleotides that engage in hydrogen bonding interactions with the hydroxyl side chain of the THR₁₆₀ group did not show greater affinity compared to the C-terminal portion of the peptide derived from spider venom in Autodock vina. This connection was found in the PatchDock results and in the literature, but it was broken after the molecular dynamics simulation.

Conclusion

The adoption of different bioinformatic tools was successful in prospecting potential receptors associated with antimicrobial activity of U1-SCRTX-lg1a. It was used by PharmMapper to search receptors, PatchDock and AutoDock vina to mensurate interactions and UCSF chimera to molecular dynamics.

At the end of this study, was found 6 potential receptors originated from Gram-negative organisms. The NH₃-dependent NAD⁺ synthetase presented the best result, which was associated to NAD⁺ production, an important precursor in several cellular pathways from Escherichia coli K12. Therefore, U1-SCRTX-lg1a interaction may disrupt normal function of this enzyme generating intracellular alteration and growth impairment, corroborating bacterial in vitro experiments.

In silico assays show evidence that U1-SCRTX-lg1a acts on protein receptors in intracellular pathways that are important for the life of Gram-negative bacteria. Future studies could observe the interaction between the toxin produced by spider venom, nucleic acids or membrane proteins.

Finally, this study opens new ways to perform in vitro experiments to validate in silico results, as well as analogs designing to improve the biological activity of this peptide. In the search for the construction of an design of the U1-SCRTX-lg1a sequence, the results presented indicate a larger layer of the N-terminal portion of the peptide, that is, this side of the peptide could be the way to construct an antimicrobial peptide while maintaining a minimum pharmacological action and the lowest possible production cost.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Names in alphabetical order. Conceptualization and methodology: ASO, EJMS and P.I.S.Jr.; validation, formal analysis, investigation, resources, and data curation:ASO, EJMS and P.I.S.Jr; writing original draft: ASO and EJMS; review and editing, and visualization: ASO and EJMS and P.I.S.Jr.; supervision: P.I.S.Jr.; project administration and funding acquisition: P.I.S.Jr. All authors have read and agreed to the published version of the manuscript.

Funding

This research received financial support from the Research Support Foundation of the State of São Paulo (FAPESP/CeTICS), grant number 2013/07467–1, Brazilian National Council for Scientific and Technological Development (CNPq), grant numbers 472744/2012–7 and 161722/2021–0, and from Higher Education Personnel Improvement Coordination (CAPES) process number 88887.663437/2022–00.

Data and software availability

The physicochemical properties were determined via the Heliquest server https://heliquest.ipmc.cnrs.fr/. Potential receptors were screened through PharmMapper available in http://www.lilab-ecust.cn/pharmmapper/. Sequence and files of receptors were downloaded from website protein data bank https://www.rcsb.org/. A molecular docking method was used: PatchDock https://bioinfo3d.cs.tau.ac.il/PatchDock/ and https://vina.scripps.edu/. Ligand and receptor interaction, molecular presentation was built by free software UCSF chimera (version 1.16) https://www.cgl.ucsf.edu/chimera/. For prediction of ligand site, it was used by the PDBsum server http://www.ebi.ac.uk/thornton-srv/databases/pdbsum/. All files used in this study are available in https://github.com/loxoscelesgaucho/loxosceles-in-silico.git.

Declarations

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.