Evaluation of Ochratoxin Recognition by Peptides Using Explicit Solvent Molecular Dynamics

Abstract

Biosensing platforms based on peptide recognition provide a cost-effective and stable alternative to antibody-based capture and discrimination of ochratoxin-A (OTA) vs. ochratoxin-B (OTB) in monitoring bioassays. Attempts to engineer peptides with improved recognition efficacy require thorough structural and thermodynamic characterization of the binding-competent conformations. Classical molecular dynamics (MD) approaches alone do not provide a thorough assessment of a peptide’s recognition efficacy. In this study, in-solution binding properties of four different peptides, a hexamer (SNLHPK), an octamer (CSIVEDGK), NFO4 (VYMNRKYYKCCK), and a 13-mer (GPAGIDGPAGIRC), which were previously generated for OTA-specific recognition, were evaluated using an advanced MD simulation approach involving accelerated configurational search and predictive modeling. Peptide configurations relevant to ochratoxin binding were initially generated using biased exchange metadynamics and the dynamic properties associated with the in-solution peptide–ochratoxin binding were derived from Markov State Models. Among the various peptides, NFO4 shows superior in-solution OTA sensing and also shows superior selectivity for OTA vs. OTB due to the lower penalty associated with solvating its bound complex. Advanced MD approaches provide structural and energetic insights critical to the hapten-specific recognition to aid the engineering of peptides with better sensing efficacies.

Molecular Modeling and Simulation Tools in the Development of Peptide-Based Biosensors for Mycotoxin Detection: Example of Ochratoxin

“…The foregoing requires that a clear energetic distinction be established between bound and unbound states, for which energetically accessible receptor and mycotoxin conformations should be calculated. For such type of molecular estimations and visualization, all-atom MD simulations are appropriate and quite frequently used [ 28 , 29 , 112 , 113 , 114 , 115 ].…”

“…Our group has focused on the use of a conformational sampling approach involving time dependent bias-potential [ aka biased exchange metadynamics (BEMD)] for the purpose of modeling peptide bioreceptor design and synthesis [ 28 , 29 ]. BEMD involves conformational sampling at the same temperature but explores the conformational landscapes of the biomolecule by introducing a time-dependent bias potential as a function of the chosen collective variables (CV) [ 122 , 123 , 124 , 125 ].…”

“…In other words, while the sampling approaches determine the ensemble average of states, the trajectories do not necessarily represent the actual dynamic time sequence of events and therefore limits any direct kinetic estimates of the biomolecular events from such simulations. In such a scenario, predictive analytics such as Markov State Model (MSM) can greatly aid the prediction of dynamic properties involved in peptide–mycotoxin binding, e.g., binding free energy, dissociation constant, and ‘ on ’ and ‘ off ’ rates [ 28 , 29 ]. MSM approaches involve the linear transformation of the configuration space into collective coordinates that are then sorted by ‘ slowness ’ into discrete states, such that the transitions within each discrete state are fast but the transitions across different states are slow.…”

“…Over the past few decades, the Bioelectronics, Biosensors and Biochips (C3B ® ) group has been engineering a wide range bioanalytical microsystems in the service of human health and medicine [ 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 ]. Recently, our group has also been involved in the development of peptide-based optical and electrochemical biosensors for continual OTA monitoring [ 28 , 29 , 30 , 31 , 32 ]. Peptides can be chemically synthesized to virtually any mycotoxin target.…”

“…In addition to improving the mycotoxin binding properties of the peptide receptors, the grand challenge in the fabrication of any biosensor for mycotoxin monitoring lies in the ability to predict and control the mycotoxin-binding properties of the molecular recognition receptor of the biotransducer. While it is often assumed that the mycotoxin binding behavior of a peptide in-solution is replicated even when the peptide is chemically modified for immobilization, chemically conjugated for labeling and/or physicochemically immobilized on a solid support, this assumption may not necessarily be true [ 28 , 29 , 30 , 31 , 32 ]. Many studies have in fact shown that a variety of factors in the synthesis of the molecular recognition receptor, in the fabrication of the biotransducer using that receptor, and in the operational processes associated with the use of that biotransducer can promote, inhibit or have no effect on the in-solution affinities and selectivity of the biological recognition molecule [ 29 , 31 , 45 , 48 ].…”