My scientific work is focused on basic aspects of protein chemistry and using molecular biology and molecular genetics to understand structure-function relationships in proteins. The main areas of experience include disulfide bond formation and thiol-disulfide redox reactions, in particular in a chemical/biophysical setting with a strong focus on in vitro characterization.
All the projects described here are carried out in close collaboration with Martin Willemoes, Kresten Lindorff-Larsen. Kaare Teilum and Lars Ellgaard are collaborating on selected projects.
In nature, proteins are able to take on a huge range of functions both as structural entities and as enzymes, the latter being the most versatile and specific chemical catalysts known. Since the 1960s it has been understood that the protein structure is solely defined by its amino acid sequence. Following this realization, a huge effort has been made to predict protein structure from sequence, however, only recently are the daunting theoretical and computational challenges relating to this problem beginning to yield.
The emerging ability to predict structure from sequence opens the intriguing possibility of approaching the inverse problem of finding an amino acid sequence compatible with a predefined structure. Considering the diverse functional properties of proteins, the theoretical and practical implications of being able to build protein structure at will are tremendous.
Contemporary tools of chemistry and molecular biology allow protein sequences to be routinely generated from synthetic DNA and recent advances in the computational methods have made it possible in a few cases to design completely new proteins with specific and novel structures.
We are currently examining the virtues and limitations of the RosettaDesign software for design and re-design of protein structure. Re-design aims to follow in the footsteps of nature and design structures, from scratch and without involving prior knowledge of specific sequences to obtain a well folded protein of similar structure to that found in nature. On the other hand a long-term goal will be do design sequence that, at will, take up desired structures, not necessarily resembling any found in nature. For the time being, we are focusing on re-design.
In our current re-design project we are testing a large library of designed sequences to analyze the template-dependence for the design software. Some of these sequences fold into proteins with well-defined tertiary structure and we are currently investigating them using various biochemical and biophysical tools.
Development and application of strategies for selection and screening of functional proteins and enzymes
To be able to efficiently identify folded sequences we are developing a model system for genetic screening and selection for protein fold independent on enzymatic activity of the design target protein. In this system the folding of the target protein is linked to an enzymatic activity which is genetically selectable.
We hope that such genetic systems will enable us to select for improved designs based on the survival of E.coli that carry more efficiently folding and stable designed proteins. Such systems should also be applicable for the improvement of thermal stability of marginally stable natural proteins.
In addition to this, we are presently involved in several collaboration projects developing model systems that directly utilize the enzymatic activity of target proteins (HIV protease and the M2 proton channel from the influenza virus) for screening and selection purposes. Here the readout, in E. coli cells, is based on the actual enzymatic activity of the target enzymes. We are here looking both at ways to use FACS sorting to obtain information on crucial residues in target enzymes.
Problems relating to charge interactions in proteins
Evaluating charges in proteins is one of the major challenges in protein design. To gain more insight into how charged amino acid residues interact in proteins we have developed a simple model protein to study protein charges. This system is based on a protein with very few charges in which we can ask specific questions regarding charge dissipation in proteins both on the surface and through the interior in a very clean system.
Artificial disulfide bonds for sensing and for stabilization of proteins
Building on our experience with thiol-disulfide reactions and we are investigating the interplay between the intrinsic stability of the disulfide bond and the stability bestowed on the protein in which it is inserted. Although it is often stated that disulfide bonds stabilize proteins this is by no means a trivial issue. Thus in designing stabilizing disulfide bonds, not only does the intrinsic stability depend on the detailed geometric properties of the engineered disulfide bond, the potential gain global stability of the protein depends on the adverse effects of loss of dynamics as well as subtle strains on the structure.
We are studying these phenomena on a cytosolic dimeric protein from E. coli. Here we have generated and exceedingly stable disulfide bond that we hope also to apply for redox sensing in the cytosol of other cell types.
This work is supported by Danish Agency for Science, Technology and Innovation.