We are interested in following how different proteins can form well-ordered fibrils. We study the following proteins:

α-synuclein and Parkinson's Disease: α-synuclein (αSN) is the major protein involved in the formation of insoluble deposits known as Lewy Bodies in the brains of Parkinson's patients. αSN is natively unfolded in solution, but assumes amyloid structures spontaneously. We conduct basic research on how α-synuclein aggregates (1, 2) and forms non-amyloid structures in lipids (3). We recently solved a low-resolution structure of what is thought to be the oligomeric intermediate in the aggregation process (2) (see picture below) and are continuing to study the properties of this oligomer. In addition, we are involved in a large project with Pfizer to develop compounds that inhibit the aggregation of αSN with the aim of developing drugs against the disease. We have developed assays (4, 5) that have allowed us with Pfizer to complete a high-throughput screen of compounds that inhibit αSN aggregation and are now analyzing the best of these compounds for their molecular properties.
We have also shown that αSN aggregates in a fundamentally different manner in the presence of SDS, forming what looks like beads on a string (1) - as shown in the figure below. Here there is a (light gray) shell of protein surrounding the (dark gray) hydrocarbon core of the SDS micelles, and each bead is presumably linked by amyloid-forming bridging segments.
Fas4 and corneal dystrophies: Several corneal dystrophies (CD) arise from the deposition of protein as aggregates in the otherwise transparent corneal tissue, ultimately leading to blindness. Transforming growth factor Beta Induced protein (TGFBIp) is the major component in many of these diseases, and we are studying the mechanism by which this aggregation occurs in order to develop ways to prevent it. TGFBIp is a 683-residue protein consisting of four domains. We have shown that the C-terminal domain (Fas4) is a good model for the behavior of the entire protein in terms of how CD-inducing mutants affect stability of the protein (6). Interestingly, different mutations in Fas4 lead to different aggregates, some of which are amyloid-like and others of which are more amorphous. We are continuing these studies with Fas4 to understand the molecular mechanisms behind these unwanted processes.

Bacterial amyloid: Functional or useful amyloid illustrates how Nature makes good use of something which is potentially dangerous. When amyloid does occur in Nature as a beneficial phenomenon (e.g. in melanosomes, insect cocoons and bacterial biofilm), it is because the amyloid-producing organisms have developed strategies to control amyloid formation in time and space. We have shown that the main amyloid protein in E. coli called CsgA, which produces curli, fibrillates under a wide variety of conditions - i.e. it is really "hardwired" to fibrillate (7). We have developed assays to identify the occurrence of amyloid structures in different bacterial communities, and have shown that this very wide-spread, with up to 50% of all bacterial species producing amyloid (8, 9), including several pathogenic Gram-positive species (10). We recently identified a new amyloid-producing operon in the widespread Pseudomonas family (11).
This operon is organized in a different way to E. coli's curli operon and we are currently analyzing the different components involved in this. Our work has been summarized in several recent reviews (12, 13).

Glucagon: This 29-residue peptide hormone does not fibrillate under physiological conditions, but has a high tendency to do so during production in the pharmaceutical industry. We have carried out numerous studies of how different types of fibrils are formed under different circumstances, illustrating the principle of fibrillar polymorphism (14-25). This has been summarized in several reviews (26, 27).
S6 (ribosomal protein from T. thermophilus): With the study of fibrillation of S6 at low pH, we were the first group to publish a detailed protein engineering study of a protein that can be induced to fibrillate with a pronounced lag time from a quasi-native state, demonstrating that fibrillation requires partially unfolded regions and that "minimization" of the backbone by truncation to alanine favours this process (28).