From the time when Chris Anfinsen unfolded protein and then allowed them to refold by removing the denaturant, it has been assumed that proteins fold b moving from an amorphous configuration to the folded structure along a pathway which minimizes the global energy. Unfolding and refolding, however, is not the process which occurs most often in nature. Proteins ar synthesized sequentially by the ribosome and extruded either into the cytosol, or through a membrane into another compartment. There are no physical means for measuring the conformation of the peptide as it is synthesized. During this year we chose to adopt the hypothesis that proteins are synthesized into a known conformational state--either helix, beta strand, or turn. Both turns and isolated beta strands are rather amorphous. Helices, by contrast, are well-defined objects with local structure. Using the Maruzen plastic models, it was possible to explore th structure and stability of a number of different helices. While the alpha helix is most common in folded protein structures, the pi helix is just one member of a whole family of related helices. In working with the structure of these variant helices, a defect in the hydrogen bonding structure of the pi helix was accidentally created. In trying to fix this defect, the defec moved along the helix. After practicing with creating and manipulating the defect motion for a while, it became clear that solutions could be created in proteins. Solitary activity was first observed as a non-dissipating wav in a canal in England in the 1850s. In the summer of 1981, David Rawn and I worked with solutions in triangulated networks, but it took another 15 years until we could map the solitary activity into protein structure. The soliton theory was applied to the domains of wheat germ agglutinin (WGA). Like many CYS containing proteins which form disulfides, each domai of WGA has a pair of adjacent CYS residues. After the solitons had decomposed the initial pi helix into several fragments, the remaining structure quite naturally selected CYS-18, rather than CYS-17, to bond to CYS-3. Once the first correct disulfide was formed, the remaining three disulfides quite naturally formed. Three of the four domains of WGA worked this way, as well as proInsulin and Basic Pancreatic Trypsin Inhibitor (BPTI). Ribonuclease-A, however, was not resolvable using the soliton theory. Richard Judson programmed his molecular dynamics program, CCEMD, to accept a series of soliton-transformed protein structures. The energy barrier between each solitary transformation state is very small. We have produced simulations and movies of the solitary transformation and folding of proteins. Our experiments show that nature may indeed fold proteins by moving from a known structure (i.e. the feedstock pi helix) to another know structure (i.e. the X-ray or NMR determined folded structure) by a series o solitary transformations. If this theory turns out to be true, then protei folding may be a very rapid and deterministic process.
