K.U. Linderstrom-Lang and J.A. Schellman write, (quote: it is somewhat greater interest that the breaking of just a few of the many hundreds of peptide linkages in proteins often leads to extensive configurational changes) and (quote: if the secondary structure of a protein (or part of it) depends for its stability on the presence of closed loop of primary bonds, it will often make very little difference whether this loop is broken at a -S-S- linkage or at one of the peptide bonds). Liderstrom-Lang and Schellman also note, (quote: all the bonds need not be primary bonds; they must only be strong). This model is called as LS model. The model is in part supported by our previous studies of staphylococcal nuclease. Based on studies of others it is generally assumed that the uniquely high density of residue packing of proteins stabilizes the native structure. Nevertheless, according to LS model, hydrolytic cleavage of only one backbone peptide bond at a certain site of the native structure can lead to extensive configurational changes. Based on studies of others, the effect of cleavage of a back bone bond would unlikely propagate to the remote area of the structure if van der Waals force, hydrophobic interactions and hydrogen bonds were only forces for stabilization. Thus, LS model leads to the following hypothesis. There may be some structural domain (LS domain) in the native structure of proteins. This domain would be associated with some unique mechanism of spatially long-range residue-residue interactions (LS mechanism) to stabilize the native structure, resulting in cooperative folding of proteins. If such residue-residue interactions were of spatially long-range, the interactions could be electrostatic in nature. Based on this hypothesis we analyzed data of others as well as ours with fragment complexes and large fragments of pancreatic ribonuclease A and those of staphylococcal nuclease and ligand induced stabilization and data of others with 3D domain swapping, circular permutation analysis and evolution of cytochrome c fold. It is found that the hypothesis of LS mechanism can explain salient aspects of these diverse phenomena. Such consistency supports the model of LS mechanism. In previous years we have found that phosphate bound to cytochrome c interacts at pH 3.6 with presumably remote hydrophobic core residues of the protein presumably through electrostatic interactions. These results, combined with our data as well as others with cytochrome c and the model of Tsao, Evans and Wennerstrom of electrostatic interactions between two hydrophobic monolayers, have led to the following hypothesis. There may be a polarizable domain in the hydrophobic core of cytochrome c, which would be a subdomain of LS domain. The markedly greater stability of horse cytochrome c as compared with yeast iso-2- cytochrome c (shown by others) would be mainly due to a difference in electrostatic interactions of this polarizable domain with the surface charges through its polarizability (dipole in nature). This model of polarizable domain in the hydrophobic core is assumed to represent LS mechanism. This model predicts the following. (a) If an iso-2 mutant contained the hydrophobic core residues mutated to those corresponding to those of horse cytochrome c (the horse core-iso-2 surface chimera), it would not have an increased stability comparable to that of horse cytochrome c. (b) If an iso-2 mutant had all hydrophobic core and ionizable surface residues corresponding to those of horse cytochrome c (quasi-horse cytochrome c), it would have stability similar to horse cytochrome c. (c) If an iso-2 mutant had all surface ionizable side chains corresponding to those of horse cytochrome c (the horse surface-iso-2 core chimera), it would not have its stability significantly increased relative to wild type iso-2. If these three predictions were found to be true, there would be the surface charges-hydrophobic core electrostatic interactions which would be related to the stability difference between iso-2- and horse cytochromes c. This would support the idea that there may be a polarizable domain in the hydrophobic core, which would respond to the field of the surface charges through its polarizability. This would in turn provide a clue to the nature of LS mechanism. Previously, we have reported that prediction a is true. Predictions b and c are not tested. Thus, the current project is preparation of quasi horse cytochrome c to test whether this mutant shows the stability similar to that of horse cytochrome c. We have used pEMBLY 30-CYC7 phagemid that contains the structural gene for iso-2 (a gift from Dr. Barry T. Nall). As a starting point we used the previously prepared phagemid containing the gene of the horse core-iso-2 surface chimera. To this gene, 12 selected mutations have been introduced, resulting in the 25 mutations of iso-2 gene. These 12 new mutations represnt a change of distribution of ionizable surface residues from iso-2 to horse cytochrome c with respect to a majority of the ionizable surface residues of horse cytochrome c. The 12 new mutations are S2D, K4E, A7K, T8K, K9V, E22K, N26H, D60K, E88K, K89T, N92E, and T99K. Yeast strain GM-3C-2 (a gift from Dr. Nall) was transformed with the phagemid containing this iso-2 gene with the 25 mutations. The transformed yeast grew in media containing 2 per cent yeast extract, 2 per cent glycerol and 1 per cent ethanol (no peptone) although the degree of the growth was about a half of that expected for usual such transformants. The ability of transformed yeast cells to grow in such media containing non-fermentable carbon source indicates that the quasi horse cytochrome c (iso-2 containing the 25 mutations) is functional. Therefore, the structure of the 25-mutation mutant is expected to be similar to wild type iso-2. I believe that understanding of the LS mechanism is crucial for fuller understanding of the forces responsible for cooperative folding of proteins. Such understanding would help predict the structure-function of proteins based on the DNA sequences and design proteins for medical application. Furthermore, if the poralizable domain model were proved, the model would also contribute to fuller understanding of important protein functions such as the effect of phosphorylation.
| Taniuchi, Hiroshi; Schechter, Alan N; Shiloach, Joseph (2004) Linderstrom-Lang-Schellman's model for protein stabilization revisited. Curr Protein Pept Sci 5:275-86 |
| Taniuchi, H; Shi, Y; San Miguel, G I et al. (2001) A study of the influence of the hydrophobic core residues of yeast iso-2-cytochrome c on phosphate binding: a probe of the hydrophobic core-surface charge interactions. J Protein Chem 20:203-15 |
| Fisher, A; Shi, Y; Ritter, A et al. (2000) Functional correlation in amino acid residue mutations of yeast iso-2-cytochrome c that is consistent with the prediction of the concomitantly variable codon theory in cytochrome c evolution. Biochem Genet 38:181-200 |
