Amphiphilic peptide Mastoparan-B induces conformational changes within the AdeB efflux pump, down-regulates adeB gene expression, and restores antibiotic susceptibility in an MDR strain of Acinetobacter baumannii

A series of computer-based experiments was done to find the mechanism antiefflux activity of MP-B against the MDR strain of A. baumannii (AB). We generated 20 poses of apo-MP-B using the InterPep tool. Analysis revealed that apo MP-B formed H-bonds to the backbone of five amino acids in Helix-5.
Published in Protocols & Methods
Like

Share this post

Choose a social network to share with, or copy the URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

 Published in Proteins. 2023;91:1205–1221

Methods

  Prediction of the protein-peptide binding site

 Identifying ligand-binding sites on the protein surface is crucial in the structure-based drug design. Due to the flexibility of peptides and the transient nature of protein-peptide interactions, they are difficult to study experimentally. Notably, small-molecule ligands can bind deeply buried sites in protein despite possible conformational changes during association.43,44 Therefore, we predicted the most likely protein–peptide site and the druggability of the unknown target for MP-B by Site Map tool version 2.4 (Schrödinger, LLC, New York, NY, 2009), which is one of the highly used tools for finding binding sites of the ligand (apo-MP-B) with a high degree of confidence. Similarly, the InterPep (an arbitrary forest, machine learning) pipeline (http://wallnerlab.org/InterPep/) was used to establish the highest plausible peptide-binding site with an accuracy of 80% and a recall of 20%. It is an excellent starting point for docking protocols or experiments investigating peptide interactions.

 Efflux pump inhibitor and molecular docking

 Mastoparan-B with MF: C78H138N20O16 was used as an RND efflux pump inhibitor. Here, the 3D structure of this peptidylamide was retrieved from the PubChem database (Compound CID:86289587) and used as a ligand. For molecular docking, we generated 20 ligand poses using the LigPrep module in Maestro v11 (https://www.schrodinger.com/products/ligprep). The best binding pose was selected from the PyMOL graphic program (https://pymol.org). At the same time, we used LigPrep software to generate an accurate, energy-minimized 3D molecular structure AdeB docking site. To verify the results, the SiteMap tool was used to select the target for ligands docking with Glide and to evaluate docking sites by showing how well the poses show the correct complementarity to the receptor. Free energy was minimized using the universal force field (UFF) and converted to pdbqt format for PyRx0.8 virtual verification, the energy minimization was performed at neutral pH 7.0 ± 2 using the OPLS 2005 least squares force field database. The force field parameter includes the bonded (intramolecular) and nonbonded (intermolecular) values, bond length, stereoisomerism, and dihedral angles. The simulation was repeated three times with different initial conditions using artificial intelligence software Gromacs 2018 (http://manual.gromacs.org/documentation/current/installguide/index.html) to avoid any dependency of the results on the initial conditions. Finally, molecular docking was performed via AutoDock/vina in the Schrödinger platform to assess the ligand and receptor H-bond order. Intermolecular hydrogen bonding between the apo-MP-B and the helix-5 backbone was detected.  Potential hydrogen bond site was energy minimized using the default settings including size, druggability score, amino acid exposure, enclosure, hydrophobicity, hydrophilicity, donor/acceptor ratio, and used for ligand-receptor docking.

 RESULTS 

  Antimicrobial susceptibility and PCR sequencing

 In a previous investigation, we isolated 65 strains of A. baumannii from two hospitals in Kerman, Iran.39 One isolate that exhibited the highest MIC profiles against multiple antibiotics was selected for this study. The isolate showed high MICs against gentamicin, kanamycin, ciprofloxacin, levofloxacin, ceftazidime, ceftriaxone, imipenem, colistin, piperacillin/tazobactam, tetracycline, and azithromycin. The combination of antibiotics with sub-MIC of MP-B had a synergism effect and reduced the MIC against gentamicin and kanamycin to 8 μg/mL, ciprofloxacin, and levofloxacin to 4 μg/mL, ceftriaxone to 2 μg/mL, carbapenems to 4 μg/mL, and tetracycline to 2 μg/mL, respectively. However, no changes in the MICs against colistin, piperacillin-tazobactam, azithromycin, trimethoprim-sulfamethoxazole, and chloramphenicol were observed. Furthermore, we examined the antibacterial activity of MP-B against this MDR isolate. Based on the results, MP-B significantly prevented the growth of A. baumannii MDR strain with MIC = 1 μg/mL, where 99.9% of the cells were unable to grow in the TSB medium (p ≤ 0.05). A growth curve test was performed to confirm that the MIC reduction wasn't due to bacterial death. We found regular growth of the organism at 0.5 μg/mL MP-B after 8 h of incubation, while no bacterial growth was observed at higher concentrations of MP-B. To confirm that restoration of antibiotic vulnerability was caused by MP-B repression of the efflux pump, we performed PCR amplification of the adeB gene, a main component of the RND efflux pump using a primer pair designed to detect the complete nucleotide sequence of the adeB gene. The gene was further purified with a purification kit and was sequenced in both directions. The nucleotide sequence was compared in the NCBI database by blasting with a set of analogous sequences. The amino acid composition of the AdeB protein was then retrieved from this nucleotide sequence. Furthermore, the expression of the adeB gene in the presence of sub-MIC of MP-B (0.5 μg/mL) by relative qRT-PCR showed a 20-fold reduction of the adeB expression compared to the control.

 The stereochemical trajectory of predicted AdeB protein

 The stereochemical trajectory of predicted AdeB protein of multidrug efflux pump superfamily, including E. coli AcrB, P. aeruginosa MexB, Neisseria gonorrhoeae MtrD, and Campylobacter jejuni CmeB were uncovered using x-ray crystallography. Among them, the AcrAB-TolC efflux system of E. coli and the MexAB-OprM efflux system of P. aeruginosa are the paradigm models and the most well-studied. However, there is a paucity of information in the case of the AdeB structure and function in A. baumannii MDR strains. The 3D trajectory of the RND efflux pump in the present study displayed a triplex system containing three subunits; (i) a U-shape channel positioned at the top of the AdeC external membrane protein composed of both α-helices and β-sheets (Figure 5A), (ii) a large periplasmic membrane cocktail AdeA protein with the coil–coil conformation in which the β-barrel folds along with α-helices creating pockets connected by large loops (the β-barrel are extended in the fringe, while α-helices indulged at the interior side) that stabilized central cavity. Moreover, a periplasmic cleft is formed between subdomains PC1 and PC2 of HAE-RND proteins. This cleft marks the gateway of substrates to the periplasm. The multidrug binding locations are responsible for creating a passage for substrate export are located deep inside this cleft. Hence, the periplasmic cleft is critical for the proper function of the pump, (iii) an inner membrane AdeB transporter in the form of helix-turn-helix conformation with a narrow pore (Figure 5A). The large periplasmic domain molecules were created by two extracellular loops bonding together TM1 with TM2 and TM7 with TM8. These periplasmic domains were further divided into the inner membrane the proximal porter structure and the distal channel domain. The porter domain is made up of the four subdomains PN1, and PN2, while, the docking area consists of the DN and DC subdomains. These features fit with the variety of antibiotics recognized by AdeB protein. The N-terminal TM1 leads to the PN1 subdomain that connects the PN2, similarly, TM7 leads to the PC1 subdomain connected to the PC2. Subdomains PN1, PN2, PC1, and PC2 form the portal domain, whereas subdomains DN and DC contribute to form the docking domain of the pump. The distal end view of the AdeB protein loose conformation (L*OO) is occluded, containing three chains, namely chains A, B, and C, shown in blue, red, and green colors. A composite figure showing the locations of the predicted bound MP-B at the AdeB periplasmic cleft entrance. Yet, a narrow pore containing a high amount of Phe and Ala acts as a substrate-trapping center for trapping different types of antibiotics. The combination of deep neural network with I-TASSER significantly improved the threading template quality boosted the accuracy of the final model through optimization fragment assembly with C-score = 1.41, TM-score = 0.99, and RMSD (Å) = 4.4 ± 2.9. The results suggested the RND efflux pumps emerged from a common ancestor. P-score = 1, respectively. Furthermore, the AdeB quality showed 96.2% of the amino acids' backbones remained in the Ramachandran favored region with torsional angles at low clash point (0.87) and MolProbity 97.92. In addition, the AdeB in apo-form showed amino acids backbone right-handed while inbound state they were left-handed. In addition, the obtained results of the improved local quality model estimation are based on a new version of QMEAN using the Swiss Model server for three chains A, B, and C. In addition, the secondary structure of the AdeB protein was delineated by the PROSETTA-fold software and showed the Z-score = 2.3 which is within the acceptable range of 10. Molecular docking analysis in combination with other computational techniques was used to find the mechanism underlying MP-B inhibition of the AdeB protein. The molecular docking approach can be used to model the interaction between a small molecule and a protein at the atomic position, for docking apo-MP-B to applicable binding sites on the AdeB transporter. The AutoDock Vina was used to predict the binding modes of the AdeB inner membrane. We observed that helix-5 was a prime candidate for the inhibitor binding point. The comparison of molecular docking simulations before and after MP-B attachment showed that all apo-MP-B poses exclusively anchor to helix-5 via amino acids Val 499. Phe binds to 454, Thr 474, Ser 461, Gly 465, and Tyr 468, respectively. Nevertheless, the other RND efflux components such as AdeA or AdeC did not contribute to the docking process. Upon docking of the ligand with the receptor, there was a rotameric shift in the orientation of the dihedral angles (Φ/ψ) of the involved amino acids by distances of 9.0 Ǻ, 9.3 Ǻ, and 9.6 Ǻ, respectively resulting in changes the bond interaction such as free energy, bond stretching (VB), angle-bending (VA), dihedral angles (Φ/Ψ), and conformational isomerism. These rotameric shifts in the involved amino acids change the interior conformation of the AdeB protein upon binding to MP-B. This remodel was recognized by AlphFold2 at the subatomic level. The attachment of MP-B with involved amino acids helix-5. Overall prediction of the ligand conformation, its position and exposure within pose sites, and assessment of the binding affinity H-bond conformation between MP-B (illustrated as an unheroic circle) with relative amino acids in helix-5.

 

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Abstract
Mastoparan B (MP-B) is an amphiphilic peptide with a potent antimicrobial activity against most Gram-negative bacteria. However, there is little information
available on the inhibition of the Acinetobacter baumannii resistance-nodulationcell-division (RND) efflux pump using this antimicrobial peptide. Here, we carried out a series of in-silico experiments to find the mechanisms underlying the
anti-efflux activity of MP-B using a multi-drug resistant (MDR) strain of
A. baumannii (AB). According to our findings, MP-B demonstrated a potent antibacterial activity against an MDR-AB (minimum inhibitory concentration
[MIC] = 1 μg/mL) followed by a 20-fold reduction in the adeB gene expression
in the presence of sub-MIC of this peptide. Using Groningen Machine for Chemicals Simulation (GROMACS) via PyMOL Graphical User Interface (GUI), (we
observed that, the AdeB transporter had conserved helix-turn-helix regions and
a tight pore rich in Phe and Ala residues. To understand how inhibition of the
AdeB is achieved, we generated 20 apo-MP-B poses using the InterPep and SiteMap tools. The high-quality model was created by homology modeling and used
for docking via AutoDock/Vina to identify the MP-B binding sites. We established that the most apo-MP-B formed H-bonds to the backbone of five amino
acids in the Helix-5. As a result, the dihedral angles of the involved amino acids
shift by 9.0–9.6 Ǻ, causing a change in the conformation of the AdeB protein.
This led to helix conformation stereoisomerization and block the AdeB activity.
MP-B presumably has dual mechanisms. (1) It blocks the AdeB transporter by
changing its conformation. (2) MP-B influences the adeB gene expression by
binding to G-protein which laterally controls efflux regulators like MarA, RamA,
SoxS, and Rob proteins.
KEYWORDS
AlphaFold2, efflux pumps, Mastoparan-B, molecular docking, protein