The structure of amyloid fibrils has been sought for a number of proteins involved in amyloid diseases. One of the main goals for determining the structure of amyloid has
been for drug design; once the structure of the fibrils is known, specific drugs can be designed to inhibit the formation of these toxic structures. However, if the structures that amyloidogenic proteins and peptides form are so dependent on their environment, they likely form many different structures in vivo, based on the different cellular compartments or extracellular space in which they form and/or deposit. Additionally, interactions between a drug and an amyloidogenic protein can change the energy landscape as well such that a different amyloid conformation could become more stable than that formed without the drug, which could be harmful or beneficial to the organism—it would be hard to predict this a priori. Therefore, it might be more beneficial to target therapeutics toward the complete inhibition of any type of aggregation by disease-related amyloid proteins, which is very challenging, or even something more upstream from fibril formation in disease progression.

Since amyloid fibrils are ordered self-assemblies, their use as bionanomaterials is of great interest. Amyloid could in principle be used as nanowires, gels, liquid crystals, and structural scaffolds. However, a better understanding of the energetics of amyloid will likely be required before any real applications result. As has been mentioned, the structure of many pathological amyloid fibrils are highly sensitive to the environment in which they are formed. Changing environmental conditions after the fibrils have formed can have a profound effect on their structure and stability, especially over a long period of time. For instance, β2-microglobulin fibrils formed at low pH dissolve when brought back to neutral conditions. Native amyloid may be better suited as a bionanomaterial because it has evolved to form amyloid, so it is likely that the structures formed are less frustrated and less susceptible to major changes in energy and structure upon an environmental change. Therefore, it is very important to carefully choose the amino acid sequences one uses to make amyloid-like bionanomaterials and to avoid using de novo designed sequences or disease-related amyloid sequences as a starting point if such sequences exhibit rough, highly concentration dependent amyloid landscapes.

As mentioned before, the stability of amyloid fibrils can be examined using fibril denaturation with chaotropes. To extract thermodynamic parameters from denaturation curves, approximations of two-state behavior are typically made. This type of analysis assumes there are no intermediates along the folding pathway, the folded state all adopts a single structure, and that the system is at equilibrium at the time of measurement. However, as discussed earlier, changing the environment of amyloid fibrils has a strong impact on the structure that it adopts. Adding chaotrope may change which structure is most stable, which would impact all of the assumptions made in doing this two-state approximation of amyloid denaturation. One must therefore be extremely careful when determining the stability of amyloid fibrils that these assumptions are in fact met (that the system is at equilibrium when measurements are taken and that the same structure is populated as chaotrope is added) before trying to obtain thermodynamic constants from amyloid denaturation curves.

It is important to understand how amyloid proteins transform from soluble monomeric units to insoluble, structured fibrils that cause disease so that viable treatment options for such maladies can be discovered. However, the polymerization mechanism most often used to describe amyloid formation, a nucleated polymerization, is likely too simplistic to describe the complexities of aggregation of pathological amyloid proteins. During a given in vitro aggregation reaction, species other than fibrils and monomers are observed, and the structures of aggregates formed depend strongly on the conditions in which the aggregation occurs and on the monomer concentration. This indicates that the folding energy landscapes of pathological amyloid proteins are not funneled, as would be the case for a typical protein; rather they are likely frustrated, containing many minima associated with different aggregate conformations that have similar energies. Therefore, slightly changing the conditions leads to a differently structured end product. For native amyloid, which has evolved its sequence to carry out its function from an amyloid structure, the folding energy landscape is likely funneled to avoid other structures that might not carry out its function properly. This has implications in the way we think about amyloid and study it biophysically, the use of amyloid fibrils for its material properties, and the design of therapeutics for treating amyloid diseases.



Sarah J. Siegel. Structure and Energetics of Amyloid