Bio-, chem-, and nanoinformatics approaches to study bionano interface
With the progress of biomedical and food technologies, the amount of new materials that come in contact with biological fluids and tissues is growing rapidly, and concerns are related but not limited to emerging risks for human health. The questions of biocompatibility of engineered materials arise naturally in respect to medical appliances such as stents, dental and cardiac implants, or prosthetic joints as they can cause immediate hazards upon introduction into the human body.
While microscopic surfaces can be scanned in a lab after exposure to biological fluids, the use of nanomaterials (NM) as implantable materials and components of medical devices poses more sophisticated regulatory challenges. NMs are being explored for a variety of biomedical tasks such as drug delivery, bioimaging, tissue engineering, and biosensors. Beside the medicine, in food industry, complex fluids like milk, protein and sugar mixtures and emulsions experience multiple contacts with engineered materials on the way to the final product and their state can be affected by interface-induced transformations, complexations, or fouling of the surfaces. Finally, common industrial NMs used in household chemistry and cosmetics, automotive and construction materials can cause chronic diseases upon exposure at the workplace.
In all these situations, adverse effects for human health can be triggered and controlled by molecular-level interactions at the biointerface. Understanding of these interactions and biointerface structure is therefore crucial for achieving better control over surface activity, for developing safety regulations, and reducing the associated health risks. Molecular simulation may provide critical tools to solve these problems efficiently.
The latter move faces a number of challenges. A key issue for the simulations is the enormous range of involved time and length scales. Nanoparticle-biointerface systems span lengthscales from the atomistic subnanometer distances (local interactions between amino acids and the surface) to hundreds of nanometers (endocytosis or protein corona). Moreover, protein adsorption and desorption is controlled by strong forces with the corresponding timescales of milliseconds, while the cell membrane rearrangement can take even longer. Finally, the number of biomolecules coming in contact with a NM in any common biological fluid is enormous. A protein corona of a single nanoparticle may contain hundreds of different proteins. For these reasons, the problem can only be addressed using coarse grained models and statistical methods.
Indeed, statistical methods based on nano-, chem- and bioinformatics have recently been successfully applied to toxicological studies. By analyzing the NM and biomolecule descriptors (e.g. surface charge, hydrophobicity, band gap), these methods have helped identify the NM properties of concern and to develop NM fingerprints that correlate well with their biological action. What these properties are in general and how they should be evaluated is the most critical question. A programme to develop nanoinformatics-based approaches is described in the U.S. Roadmap document (Nanoinformatics 2020) and a new EU-US nanoinformatics roadmap is currently being prepared by the community. Moreover, the draft H2020 NMBP programme for 2018-19 includes a call citing nanoinformatics as a crucial tool for addressing the NM safety.
The main goal of the workshop is to outline the progress and required essential steps needed to enable in silico screening of materials for biocompatibility and toxicity using bio-, chem-, and nanoinformatics. The key concept here is the material's descriptor that can quantify the material's biological potency, i.e. its ability to bind or unfold a certain type of molecules, bind cells, produce reactive species, or penetrate biological barriers. The descriptors can be used to construct materials' fingerprints and predictive models (e.g. QSARs) for the NM biological action, to improve NM application efficiency or safety using machine learning techniques.
We will address computational methodologies for systematic evaluation of descriptors, their experimental validation, and construction of databases for bio- and nanomaterials. In particular, the programme will cover the following topics:
• Systematic evaluation of physicochemical properties of common NMs and identification of intrinsic and extrinsic descriptors that determine the complexation of NMs with biomolecules.
• Calculation of advanced descriptors of biomolecules: 3D protein structure, adsorption affinity using molecular simulations
• Construction of predictive models (QSAR) for NM protein corona and for the events at the bionano interface based on NM and biomolecule descriptors using machine learning techniques
• Validation of the models using experiments and detailed simulations
• Development of simulation ontologies, databases of NM properties, read-across and data mining techniques
This event is associated with activities of the EU Nanosafety cluster, European Materials Modelling Council (EMMC) and European Materials Characterisation Council (EMCC).
Vladimir Lobaskin (University College Dublin) - Organiser & speaker
Tomasz Puzyn (University of Gdansk) - Organiser & speaker