Whenever any synthetic material is introduced into a biological system, either through design or through environmental uptake, issues of biocompatibility and potential deleterious health effects are of paramount importance. A few studies of the interaction of carbon nanoparticles with proteins and nucleotides have been reported.18,19 While there is still controversy as to whether pristine C60 and its derivatives are health and environmentally dangerous20 there is strong evidence that fullerenes can be cytotoxic21-23 and that the exposure of living cells to C60 fullerene derivatives and that their aggregates can result in disruption of the cell membrane due to lipid peroxidation.24 An improved fundamental understanding of interactions between fullerenes and cell membranes is central to understanding potential deleterious effects of fullerenes and fullerene-based molecules on living cells as well as to help in the design of effective nanoparticle-based biomaterials. Recently, we have shown through molecular dynamics (MD) simulation studies that because of their size (approximately 1 nm in diameter) and strong van der Waals (dispersion) interactions (due to a high density of surface atoms), fullerenes behave neither like classical hydrophobic molecules nor like hydrophobic surfaces.25-27 We found that association of a pair of C60 fullerenes is dominated by the strong direct fullerene−fullerene attractive interactions. Surprisingly, water-induced interactions between the fullerenes were found to be repulsive. Specifically, it was found that the high atomic density on the fullerene surface gives rise to highly favorable water−fullerene interactions due to strong carbon−water dispersion interactions. The energetic origin of this water-induced repulsive interaction between the carbon nanoparticles is in stark contrast to the entropy-driven association observed for conventional hydrophobic nonpolar organic solutes.
To investigate the implications of the unique properties of fullerenes on their interaction with and passive transport into lipid membranes, we have conducted all-atom MD simulations of a C60 fullerene in a fully hydrated di-myristoyl-phoshatidylcholine (DMPC) lipid membrane. To date computational studies of molecule transport in lipid membranes have been primarily limited to small molecules.28-44 A few atomistic simulations studies45,46 have looked at the interaction of carbon-based nanoparticles with model cell membranes but have not addressed the permeability of these nanoparticles through cell membrane. Only two simulation works address issues of permeability of fullerenes through model cell membranes. The first employed a coarse-grained model to investigate the transport of model nanoparticles (fullerene-like) through a model lipid membrane.47 While this work showed some qualitative consistency with atomistic simulations discussed below the results of these simulations should be taken with caution due to relatively crude representation of water and lipid molecules. The second study utilized atomistic simulations to compare transport of C60 and C60(OH)20 from water into a di-palmitoyl-phosphatidylcholine (DPPC) bilayer48 and found that a pristine fullerene can readily “jump” into the bilayer and translocate through membrane, while C60(OH)20 was not able to do so over the entire length of the simulation. The authors explained the observed difference in the ability of the pristine and OH-modified fullerenes to penentrate into the bilayer by analyzing the potential of mean force (PMF) acting on the fullerene as a function of its position relative to the bilayer center of mass. However, as we discuss below, the PMF obtained from that work is inconsistent with our data as well as experimental data on solubility of fullerenes in different environments.