Research in Computational Biophysics

About Proteins and Protein Folding

Proteins are the nano-machines of life. Virtually every molecular task in a living organism involves highly specific actions of these biomolecules. Oxygen transport, muscle activation, light sensation in the retina of our eyes, smelling, immune reaction, tissue regeneration, hormones and enzymagic reactions, photosynthesis, bioluminence - everything is governed and regulated by proteins.

Each protein has a well defined and complex three-dimensional structure, which it requires to fulfill its specific function. Although we understand how the genetic information in the DNA encodes the chemical structure of its corresponding protein, it is still not really understood how the nascent chain of amino acids arranges itself into the often incredibly complex three-dimensional structure. This process, which is known as protein folding, is crucial for generating correctly functioning proteins. If this process goes wrong - due to defects in the genetic information, unfavorable conditions in the environment of the nascent protein chain, or simply by stochasticity - then the protein does not acquire its correct structure, and can not properly fulfill its physiological function. Many diseases, including Alzheimers, Creutzfeld-Jakobs, other neuro-degenerative illnesses, as well as various forms of cancer, have their molecular origin in incorrectly folded proteins.

It is therefore of utmost importance that we understand the determinants of protein folding and stability in their molecular details. Not only is this a fascinating intellectual puzzle for basic research, but it also constitutes the very basis to successfully treat or prevent illnesses like the ones mentioned above.

My research in this field aimed to advance our understanding of these processes by means of powerful compuational physics-based methods, Molecular Dynamics Simulations, in conjunction with statistical methods capable of identifying and extracting the relevant information from the high-dimensional data obtained by this method.

Protein Denaturation

At the Max-Planck-Institute for Biophysical Chemistry, I worked in the Department for Theoretical and Computational Biophysics, and investigated protein denaturation and stability.

The solvent environment of a protein governs its stability, and cosolvents like urea can induce protein denaturation and unfolding. Other cosolvents, like TMAO, can actually increase protein stability and serve in some organisms as protective agents against cellular stress. The molecular mechanism by which these denaturing or stabilizing cosolvents affect protein stability is not yet well understood, and was the focus of my research and topic of my PhD thesis.

The main results of my work can be summarized as follows:

• Urea acts through direct interactions with the protein, rather than through an indirect mechanism via alteration of the water structure.
• Less polar amino acids, as well as the protein backbone, exhibit particularly favorable interactions with urea, whereas more polar amino acids have a higher solvation coefficient with water.
• Urea displaces water from the solvation shell of the protein, and releases it into the bulk solvent. This is entropically favorable, in particular because one urea molecule substitutes for approximately three water molecules. As a net effect, urea acts as interface between bulk water and protein surfae, weakening the hydrophobic effect.
• The enthalpics of hydrogen bonding for all different donor:acceptor combinations does not support a dominant role of urea:protein hydrogen bonds upon unfolding.
• Formation of protein-urea hydrogen bonds is not the cause for preferential solvation by urea, but rather a consequence of it.
Urea (green) binding to exposed protein surface.
• In its dynamics, urea-induced protein denaturation is not necessarily an actively-driven process. Urea molecules can bind to temporarily exposed less polar protein surface and - like a ratchet - render reversible fluctuations in the natural dynamics as irreversible, leading to a sequence of partial unfolding events. The resulting shift in the equilibrium is observed as unfolding.
• The denaturation mechanism of Guanidinium, another common denaturant, seems to be different than that of urea, despite their geometric similarity. Due to its ionic nature, a combination of contributions from polar effects and urea-like effects might be conceivable.
• The stabilizing osmolyte TMAO exhibits unfavorable interactions with the backbone, which would be exposed upon unfolding. The compact folded conformation is therefore energetically favored. This mechanism of stabilization is very analoguous to the destabilizing effect of urea, just with inversed sign. This might suggest that the positive or negative effect of cosolvents on protein stability is generally determined by the energetics of the interactions.

Aside from main project, I also worked on a side project investigating the affinity of ligand binding to the transport protein Snurportin 1, and the structural response of the receptor. On of the main results of this work is that the binding affinity for different ligands is largely determined by the entropic contribution of the solvation water upon ligand unbinding. Further, binding of the ligand induces an expected qualitative change in the dynamics of the receptor binding site.

Protein Folding - Functional Mutant Design

At Stanford University, I worked in the NIH Simbios Center, focussing on research in protein folding and functional mutant design.

Chaperonins are nature's invention to facilitate correct protein folding, and prevent misfolding. These molecules are huge protein complexes themselves, and assist substrate proteins in their folding process. In that regard, they can be regarded as molecular machines that assemble new machines. How chaperonins work on the molecular level is not yet exactly understood, although promising theories have emerged in recent years. With their ability to prevent or undo misfolding of proteins, a better understanding of chaperonins and the design of chaperonins with improved functionality is the first step in a nanomedicine-approach to combat misfolding related diseases.

My research aimed at verifying existing hypotheses for chaperonin action, and moreover based on that to assess and improve methods for computational design of chaperonin mutants with improved functionality.

In an international collaboration, we investigated the relation between the water density inside the chaperonin cavity and the ability of the chaperonin to promote substrate folding. Our worked provided strong support for the hypothesis that chaperonins faciliate protein folding mainly by an enhanced hydrophobic effect inside the chaperonin cavity, the place where the substrate protein is kept during its folding process.

Moreover, the significant correlation between the local water density inside the cavity of different mutants of the prokaryotic chaperonin GroEL and their respective folding activities provided a direct relation between a structural property and a functional property of the chaperonin. As such, it provided a starting point for the design of functional mutants.

For computational mutant screening, however, an accurate as well as efficient sampling method is critical. We evaluated the possibilities of a joint approach of Molecular Dynamics Simulations and the 3D Reference Interaction Site Model. Both have their respective strengths and limitations, and we showed that these two methods can be used complimentary to each other for efficient and accurate screening of mutants. This approach will allow to tailor proteins for improved or modified functionality.

As a side project, I developed a computational tool, SimSAXS for the efficient computation of Small Angle X-ray Scattering (SAXS) structures from atomic coordinates. This software can be used to bridge the gap between theory and experiment, which is an important, yet often hardly accessible, link in structural biology.