The understanding of the mechanisms involved in the interaction of proteins with inorganic surfaces is of major interest in both fundamental research and applications such as nanotechnology. However, despite intense research, the mechanisms and the structural determinants of protein/surface interactions are still unclear. We developed a strategy consisting in identifying, in a mixture of hundreds of soluble proteins, those proteins that are adsorbed on the surface and those that are not. If the two protein subsets are large enough, their statistical comparative analysis must reveal the physicochemical determinants relevant for adsorption versus non-adsorption. This methodology was tested with silica nanoparticles. We found that the adsorbed proteins contain a higher number of charged amino acids, particularly arginine, which is consistent with involvement of this basic amino acid in electrostatic interactions with silica. The analysis also identified a marked bias toward low aromatic amino acid content (phenylalanine, tryptophan, tyrosine and histidine) in adsorbed proteins. Structural analyses and molecular dynamics simulations of proteins from the two groups indicate that non-adsorbed proteins have twice as many π-π interactions and higher structural rigidity. The data are consistent with the notion that adsorption is correlated with the flexibility of the protein and with its ability to spread on the surface. Our findings led us to propose a refined model of protein adsorption.

Upon contact with biological fluids, nanoparticles (NPs) are readily coated by cellular compounds, particularly proteins, which are determining factors for the localization and toxicity of NPs in the organism. Here, we improved a methodological approach to identify proteins that adsorb on silica NPs with high affinity. Using large-scale proteomics and mixtures of soluble proteins prepared either from yeast cells or from alveolar human cells, we observed that proteins with large unstructured region(s) are more prone to bind on silica NPs. These disordered regions provide flexibility to proteins, a property that promotes their adsorption. The statistical analyses also pointed to a marked overrepresentation of RNA-binding proteins (RBPs) and of translation initiation factors among the adsorbed proteins. We propose that silica surfaces, which are mainly composed of Si–O⁻ and Si–OH groups, mimic ribose-phosphate molecules (rich in –O⁻ and –OH) and trap the proteins able to interact with ribose-phosphate containing molecules. Finally, using an in vitro assay, we showed that the sequestration of translation initiation factors by silica NPs results in an inhibition of the in vitro translational activity. This result demonstrates that characterizing the protein corona of various NPs would be a relevant approach to predict their potential toxicological effects.

The human bronchial epithelium is the first line of defense against atmospheric particles, pollutants, and respiratory pathogens such as the novel SARS-CoV-2. The epithelial cells form a tight barrier and secrete proteins that are major components of the mucosal immune response. Functional in vitro models of the human lung are essential for screening the epithelial response and assessing the toxicity and barrier crossing of drugs, inhaled particles, and pollutants. However, there is a lack of models to investigate the effect of chronic exposure without resorting to animal testing. Here, we developed a 3D model of the human bronchial epithelium using Calu-3 cell line and demonstrated its viability and functionality for 21 days without subculturing. We investigated the effect of reduced Fetal Bovine Serum supplementation in the basal medium and defined the minimal supplementation needed to maintain a functional epithelium, so that the amount of exogenous serum proteins could be reduced during drug testing. The long-term evolution of the epithelial cell secretome was fully characterized by quantitative mass spectrometry in two preclinical models using Calu-3 or primary NHBE cells. 408 common secreted proteins were identified while significant differences in protein abundance were observed with time, suggesting that 7–10 days are necessary to establish a mature secretome in the Calu-3 model. The associated Reactome pathways highlight the role of the secreted proteins in the immune response of the bronchial epithelium. We suggest this preclinical 3D model can be used to evaluate the long-term toxicity of drugs or particles on the human bronchial epithelium, and subsequently to investigate their effect on the epithelial cell secretions.