Identification and characterization 2007

This paper presents an effort to address two fundamental question in human prion biology, which existed in our minds at the time at which this work was done: (i) What is the critical region in the prion protein which is responsible for making it easier for prions to transform from alpha helix-based structures to beta sheet-based structures? and (ii) Is it possible to combinatorially explore this region through phage-display screening of a library of prion variants, for the ability to bind to immobilized prion molecules which had been deliberately converted into a beta sheet conformation by exposure to low pH (pH 4.0) under reducing conditions. It seemed from the existing literature of the time that a large proportion of naturally-occurring mutations found to be associated with spontaneous cases of new variant Crutzfeld-Jakob Syndrome (nvCJD; a prion-based disease) involved amino acid changes between residues 101 and 112 of the human prion protein. So, we resolved to genetically randomize this region and explore whether we could address the above questions. Instead of using the full-length human prion, we used a truncated form already shown to be capable of displaying all of the full-length molecule’s features, consisting of residues 90-231 (including the residues 101-112 which were of interest to us). We cloned the gene encoding the protein, and mounted it on the tips of filamentous bacteriophage. We also transformed the purified protein into a beta-sheet-based form through reduction and exposure to pH 4.0, and immobilized this prion form. We then used degenerate oligonucleotides to create a library of prions randomized in the region of residues 101-112 and mounted this library on phagemids with which we created a library of prion variants. The diversity of the library was determined to be of the order of twenty million variants; negligible, compared to the 12E20 theoretical variants possible, but sizeable nevertheless. We then spiked this library of phage-borne prion variants with close to 1:1 phage-born wild-type prion (already prepared; see above) and screened this library for the ability to bind to the immobilized beta-sheet-based form of the prion molecule, through several rounds of biopanning and amplification of phages. We discovered that two sequences climbed up in representation through all the rounds of biopanning and amplification. Along with wild-type prion which was also recovered (because the sheer numbers of phage bearing wild-type prion were many orders of magnitude higher than any variant), these two sequences appeared to consistently crop up, suggesting that they have a million-fold higher affinity for the beta-sheet based form of the prion than any other variant, or even the wild-type prion. We then cloned the gene and expressed one of these two variants of prion as a soluble protein, and found that it has a beta sheet-based structure at pH 10.0, but that under pH 8.0 this variant naturally aggregates and precipitates into amyloid fibers. It would appear that we had isolated what could be a super-prion, which was already amyloid-like, and capable of binding to beta-sheet-based prion entities immobilized on a surface. At this point, we were satisfied with the testing of our ideas and stopped working further because we did not have the motivation to do any further biosafety level II or level III work, particularly since it was only to satisfy curiosity. In summary, we proved that residues 101-112 which form the interface between the unstructured and structured parts of the prion-chain (like a hinge region) might play an important part in the protein’s structural transformation.