S2: MODELLING BIONANO INTERACTIONS FOR PREDICTIVE TOXICOLOGY
Theme: Risk assessment & management
Over the last decade, in vitro and in vivo experiments have produced significant amount of veritable information that can be integrated into theoretical models with the aim of predicting possible health and environmental effects of engineered nanoparticles (NP). However, even the most systematic studies often leave the question of precise toxicity mechanisms associated with NPs wide open. One can speculate that the toxic effects can emerge either from membrane damage or from interaction of NPs, once they are inside the cell/tissue, with the internal cell machinery and signaling pathways. Therefore, NP ability to penetrate the cells, bind to key biomolecules, or perturb the normal pathways, either by physical (e.g. membrane disruption) or biochemical mechanisms (e.g. production of reactive oxygen species) can be important. Significant research efforts were invested recently into nanotoxicology in an attempt to identify nanomaterials' properties of concern, which can be responsible for these interactions, and model their relationships with adverse outcomes.
While statistical analysis allows one to relate the physical and chemical descriptors of the materials to certain toxicity endpoints, the mechanisms of action are not always known. The quantitative relations between the descriptors and the effect can only be deduced once we have a clear picture of all stages of interaction between the foreign agent and the biological tissue and of all stages of the systemic transport at the mechanistic level. The identification of molecular or nanomaterial properties responsible for the uptake or hazard can be facilitated by establishing the molecular initiating events or key events in each adverse outcome pathway and by mechanistic modelling of the underlying processes. A development of an intelligent, mechanism-aware testing strategy may require an identification of novel mechanistic endpoints. On the other hand, to build a successful predictive model, one should develop a new language suitable for description of bionano interactions and identify the relevant NP and biomolecule descriptors. In most cases, the standard physicochemical descriptors of NPs are not fit for this purpose as they do not allow to predict whether a specific NP would bind to the membrane, adsorb or deform the key molecule.
To create a basis for grouping the nanomaterials, read-across, and for development of quantitative models of toxicity, we propose to form a database of bionano interactions in addition to common physicochemical descriptors. We develop a computational scheme for a fast evaluation NP-biomolecule interactions. We use a bottom-up molecular simulation approach, which relates the advanced protein and lipid descriptors (sequence and structure descriptors) with basic molecular interactions at the interface. We address protein adsorption on metals, oxides, and carbon-based materials. We present: (1) a principal scheme of the model required for understanding of protein adsorption at various solid interfaces; (2) propose a set of advanced NP descriptors (e.g. hydration energy, ionisation potential, conduction band gap, Hamaker constant, etc.) and show how they can be calculated, (3) the scheme of model parameterisation using experiment or simulation, and (4) a bio/nanoinformatics-based scheme to predict free energy of adsorption of globular proteins at liquid/solid interfaces and (5) a way of ranking adsorption propensity of proteins. We also construct and validate a method of modelling the adsorption kinetics and competitive adsorption of proteins on NPs and formation of NP protein corona. We quantify the influence of NP surface curvature, charge and coating on the adsorption energies for important plasma proteins and rank the proteins by their binding affinity to the NPs. We show what properties of NPs can be most promising for predictive modelling of bionano interface and toxicity mechanisms.
This work is supported by EU H2020 project SmartNanoTox under grant agreement No. 686098.