Daniel Chan and Dr. Allen Buskirk, Chemistry and Biochemistry

Proteins are of the utmost physiological importance in their roles as enzymes, structural elements, and antibodies. Correctly formed proteins catalyze biochemical reactions, initiate proper immune system response, and even promote the development of hair, bones, skin, muscles, and blood. Likewise, incorrectly formed proteins cause Alzheimer’s disease, cystic fibrosis, various forms of cancer, and Creutzfeld-Jacob (mad cow) disease1. One of the fundamental problems that scientists face today is understanding how proteins fold to yield correct or incorrect structures. Proteins consist of building blocks called amino acids, each of which is encoded by certain DNA sequences. The conventional wisdom in the field is that small DNA sequence changes make little difference in the folding pathway of a protein. However, recent studies by Subbian et al. suggest that these small differences may have powerful effects.

In our genetic analysis of protein folding we utilized two proteins, SbtE and ISP, that are both closely related by sequence and structures yet exhibit different folding pathways. The differences in the sequence of these two proteins map to the protein surfaces. The ISP subfamily includes a large number of acidic amino acid residues, which are charged at neutral pH, whereas the SbtE subfamily contains few charged residues. The proteins both share similar hydrophobic cores, contact order, and topology – properties known to strongly affect protein folding2.

We hypothesized that charged residues play a role in intermediates along the folding pathway, driving the folding reaction to completion. If these intermediates are long-lived, they can lead to aggregation and protein misfolding. SbtE misfolds through a protein folding pathway that becomes trapped in its kinetically stable product, forming its thermodynamically stable product at the rate of t½ = 1500 years. On the contrary, ISP folds correctly to form its thermodynamically stable product nearly one million times faster than SbtE. Thus despite the similar DNA sequences of SbtE and ISP, an important difference in folding pathways results in two different protein folding products.

Our interest was to identify which sequence differences of SbtE and ISP cause their different protein folding pathways. To do so we used an assay developed by Waldo et al. utilizing green fluorescent protein to quickly determine whether or not a protein folded correctly. When bacteria express GFP fused to a protein of interest, they fluoresce green under ultraviolet light, in direct proportion to the amount of the correctly folded protein3. For SbtE to fold correctly, it requires a pro-domain, which we have removed for our experimental setup. SbtE-GFP fusions as such, do not fold properly and interfere with GFP function. On the other hand, cloned bacteria expressing GFP-ISP fluoresce green because ISP folds properly to a thermodynamically stable conformer. This assay had already been recreated successfully prior to my involvement on this project.

I created an assortment of mutant SbtE fragments in hopes of identifying a particular SbtE sequence that folds correctly even without its pro-domain. This was done using error-prone polymerase chain reaction (PCR) which allowed me to create DNA sequence changes to SbtE by introducing mutations (roughly 1 in 1000 bases). In order to isolate the correctly folding, thermodynamically stable mutants of SbtE, I inserted the SbtE mutant genes upstream of the
GFP gene on our selection plasmid and created a library of bacteria with different forms of SbtE in the mid-106 range of number of mutants. I observed for mutant SbtE-GFP fusions that fluoresced green signified that SbtE-GFP mutants had changed folding pathways to now fold correctly.

To examine these results I plated colonies of the mutant SbtE-GFP bacteria on agar plates, and observed their degree of fluorescence under UV light. I also plated ISP-GFP bacteria as controls to determine a baseline level of green florescence and compared it to that of SbtE mutants looking for any that matched the intensity as that of ISP.

I picked certain SbtE mutants that did exhibit green florescence and grew them up in culture and plated them again to verify that they indeed fluoresced as brightly as the ISP-GFP controls. Multiple green fluorescing SbtE-GFP mutants were isolated (SbtE-Hits) and I selected the brightest three samples to have the DNA of the bacteria sequenced. The DNA sequences that we obtained of the SbtE-Hits showed similar mutations in certain sequences yet were inconclusive as to the exact reason why the SbtE-Hits were fluorescing green.

We suspected that the surface residues of the proteins influenced the folding and aggregation of intermediates affecting the folding pathway of the SbtE-Hits. Hydrophobic residues tend to aggregate and inhibit the folding pathway of proteins, while soluble residues tend to facilitate proper folding. To test a hypothesis that ISP and SbtE-Hits fold correctly due to increased soluble residues, I ran a solubility test on the SbtE-Hits against an ISP control using a protein gel. To our misfortune we discovered that our control, ISP, contained minimal soluble proteins suggesting that the folding pathway differences between SbtE and ISP were inherently not dependent on protein solubility. Our results from the protein gel also showed that the SbtE-Hits had similar insoluble characteristics as ISP. This further suggested that the green fluorescence of the SbtE-Hits was independent of the aggregation of pathway intermediates and of the folding pathway in general of both SbtE and ISP.

As it turns out, contrary to ideas expressed in literature of using GFP as an assay to monitor protein folding, I learned that GFP is not best suited as a determinant for the folding pathways of ISP and SbtE as it is too sensitive to change and expresses independently of the ISP and SbtE conformations. ISP and SbtE still remain model proteins to study the effects of DNA changes on protein conformation. Though our hypothesis remains that fine DNA sequence changes significantly affect protein-folding pathways, future research will require an alternative method from GFP to assay for these changes.

References

1. Cohen, F.E. (1999) Protein misfolding and prion diseases. Journal of Mol Biol293 (2), 313-320.
2. Subbian, E., Yukihiro, Y., and Ujwal, S. (2004) Positive selection dictates the choice between kinetic and
   thermodynamic folding and stability in subtilases, Biochemistry 43, 14348-14360.
3. waldo, G.S., Standish, B.M., Berendzen, J., and Terwillinger, T.C. (1999) Rapid protein-folding assay using
   green fluorescent protein. Nature Biotechnology, Vol17.