Surface-induced protein denaturation has important implications for the development of materials that are resistant and/or innocuous to biomolecules. Here, we studied the mechanism of lysozyme (T4L) unfolding on fused silica (FS) using single-molecule methods that provided direct insight into the cause of denaturation. Unfolding of T4L was monitored by Förster resonance energy transfer while simultaneously tracking the adsorption, diffusion, and desorption of individual molecules at the solid-solution interface. Results of high-throughput single-molecule analysis suggested that the unfolding of T4L on FS was mediated by surface diffusion and occurred on isolated nanoscale sites, which were relatively rare and distinct from the majority of the surface. These observations suggest that surface-mediated protein unfolding is a search process that is based on the exploration for denaturing sites by the protein. Ultimately, these findings have important implications for the design of protein-compatible surfaces.

The successful incorporation of enzymes into materials through multipoint covalent immobilization (MPCI) has served as the foundation for numerous advances in diverse fields, including biocatalysis, biosensing, and chemical weapons defense. Despite this success, a mechanistic understanding of the impact of this approach on enzyme stability has remained elusive, which is critical for realizing the full potential of MPCI. Here, we showed that the stabilization of lipase upon MPCI to polymer brush surfaces resulted from the rigidification of the enzyme with an increase in the number of enzyme-brush attachments. This was evident by a 10-fold decrease in the rates of enzyme unfolding and refolding as well as a reduction of the intrinsic fluctuations of the folded and unfolded states, which was measured by single-molecule (SM) Förster Resonance Energy Transfer imaging. Moreover, our results illuminate an important trade-off between stability and activity as a function of this decrease in structural dynamics of the immobilized lipase. Notably, as the thermal stability of lipase increased, as indicated by the temperature optimum for activity of the enzyme, the specific activity of lipase decreased. This decrease in activity was attributed to a reduction in the essential motions of the folded state that are required for catalytic turnover of substrate. These results provide direct evidence of this effect, which has long been a matter of speculation. Furthermore, our findings suggest that the retention of activity and stabilization of an enzyme may be balanced by tuning the extent of enzyme attachment.

Despite their widespread use in biocatalysis, the marginal stability of lipases can significantly limit their catalytic performance in industrial biotransformations. Here, we demonstrate that this limitation can be overcome by immobilization on poly­(sulfobetaine methacrylate) (PSBMA) polymer brushes. Specifically, the immobilization of Bacillus subtilis lipase A (lipA) on PSBMA brushes resulted in a 100-fold enhancement in turnover frequency relative to ambient conditions at the temperature optimum of the immobilized enzyme, which was also improved by immobilization. This significant enhancement in catalytic performance was due to the structural stabilization of lipA as well as changes in lipA conformational dynamics as measured using single-molecule Förster resonance energy transfer. Interestingly, the enhancement in catalytic performance of lipases depended strongly on the chemistry of the brush. These findings demonstrate that tuning the brush chemistry can lead to marked improvements in the catalytic efficiency of immobilized lipases, which may have major ramifications in industrial biocatalysis.

A long-standing goal in the field of biotechnology is to develop and understand design rules for the stabilization of enzymes upon immobilization to materials. While immobilization has sometimes been successful as a strategy to stabilize enzymes, the design of synthetic materials that stabilize enzymes remains largely empirical. We sought to overcome this challenge by investigating the mechanistic basis for the stabilization of immobilized lipases on random copolymer brush surfaces comprised of poly­(ethylene glycol) methacrylate (PEGMA) and sulfobetaine methacrylate (SBMA), which represent novel heterogeneous supports for immobilized enzymes. Using several related but structurally diverse lipases, including Bacillus subtilis lipase A (LipA), Rhizomucor miehei lipase, Candida rugosa lipase, and Candida antarctica lipase B (CALB), we showed that the stability of each lipase at elevated temperatures was strongly dependent on the fraction of PEGMA in the brush layer. This dependence was explained by developing and applying a new algorithm to quantify protein surface hydrophobicity, which involved using unsupervised cluster analysis to identify clusters of hydrophobic atoms. Characterization of the lipases showed that the optimal brush composition correlated with the free energy of solvation per enzyme surface area, which ranged from −17.1 kJ/mol·nm2 for LipA to −11.8 kJ/mol·nm2 for CALB. Additionally, using this algorithm, we found that hydrophobic patches consisting of aliphatic residues had a higher free energy than patches consisting of aromatic residues. By providing the basis for rationally tuning the interface between enzymes and materials, this understanding will transform the use of materials to reliably ruggedize enzymes under extreme conditions.

The surface diffusion of poly-L-lysine (PLL) in a planar nanoslit was studied using convex lens-induced confinement (CLiC) single-molecule tracking microscopy. Three surface chemistries were employed to understand the interplay of electrostatic and short-range interactions: an amine-functionalized silica surface, an oligo(ethylene oxide) (OEG)-modified surface, and a 1:1 mixture of the two ligands. Effective surface diffusion coefficients increased rapidly with slit height until saturating for slit heights <30 nm. While diffusion at a semi-infinite interface was significantly faster for OEG surfaces, the diffusion coefficient increased most rapidly with slit height for aminefunctionalized surfaces, resulting in surface diffusion within very thin slits being nearly independent of surface chemistry. Intermittent random walks were simulated within a planar slit geometry, using experimentally measured parameters obtained from diffusion at a single interface to account for the characteristic short-range interactions between PLL and each surface chemistry, and were in good agreement with experimental measurements.