Read with caution!
This post was written during early stages of trying to understand a complex scientific problem, and we didn't get everything right. The original author no longer endorses the content of this post. It is being left online for historical reasons, but read at your own risk.
A helpful review from 2003 by Detlev Risener about the amino acid sequence, domain structure, and post-translatational modifications (two glycosylation sites, GPI-anchor, disulfide bond) of the PrP protein, and characterization of the structural and energetic states of the two forms of the protein.
Below is a nice schematic of the protein linear sequence and modification sites, taken from this review.
Here’s another review that looks good, on a similar topic, from 2011:
Surewicz WK and Apostol MI (2011), Topp Curr Chem 305:135-68.
Notes from the Surewicz and Apostol 2011 review (more to come):
PrPc: monomeric, alpha-helical
PrPsc: oligomeric, B-sheet-rich
(same protein, differentially folded)
253 aa protein, once C-terminal and N-terminal signal sequences (target to membrane) are removed, 209 aa functional protein. tethered to cell membrane by GPI anchor.
Most structural studies conducted with recombinant (r) PrPc, because it is hard to get good yeild from brain-derived material. This rPrPc will not have glycosylations or GPI anchor, but at least one study suggest that it is structurally similar to PrPC in vivo (using what techniques?)
C-terminal domain (resi 121-231): “structured”/”folded” and best studied
N-terminal domain: very fleixble, and at least partially disordered. (includes octapeptide repeats – proline and glycine rich). Histidines and Tryptophans are essential for copper binding (Cu2+). There is an x-ray crystal structure of a small peptide (HGGGW) bound to copper and in a B-turn structure. All four His-containing octarepeats may bind to Cu2+ this way. This region also has been found to bind other small molecule ligands such as glycosaminoglycans, hemin, and oligonucleotides/nucleic acids. NMR experiments suggest N-temrinal domain may exist in equilobrium between unfolded and partially-folded B-turn state. pH may shift the equilibrium (neutral = folded).
Important: octapeptide repeat does not appear to be ESSENTIAL for prion infectivity, but may likely faciilitate it. It has been thought to faciliate self-interaction between PrP molecules, but it does not form the B-sheet-rich core of PrPsc. Insertion of additional octapeptide repats lead to familial prion diseases in humans and this has also held up in controlled mice experiments.
C-terminal domain: more than 12 X-ray crytal and NMR structures of this region of human PrP! Consensus structure: three alpha-helices and 2 short-B strands in antiparallel sheet. Helix 2 and 3 have the disulfide bond between them (Cys179Cys214). PrPc has an *unusually small hydrophobic core* and surfaceis very hyodrophilic and charged.
- Flexibility and amino-acid identity in one particular loop (aa165-175) seem to corrlate with cross-species transmission barriers (stiff infects stiff, loose infects loose), and mutations in this loop in mice can lead to spoontanteous prion diseases.
- Most genetic mutations in inherited human prion diseaseas map to helix 2 and 3 and the loop between them. Many mutations influence charge – some neaautralizing charg (D178N, D202N, E211Q), some introducing new charge (T188K, T188R, H187R, Q217R) or reverse from negative to positive charge (E196K, E200K). There are NMR structures solved of E200K, Q212P and H187R. There are also crystal structures solved of *D178N* and F198S. All of these structures of mutants showed only minor structural differences, and provided little insight into pathogenicity.
- Some of the crystal structures have revealed homo-dimers, which may be relevant for how two molecules might interact before converting to the pathogenic state. Interactions are between the two-stranded B-sheet with its symmetry mate in the crystal to form a 4-stranded B-sheet. ANd this has been seem in multiple strucutres of human and sheep PrPc. This interaction centers around resi 129, the position of a Met/Val polymorpism which affects susceptability to prion diseases. This position could be a nucleation site and influence PrPc –> PrPsc conversion.
The other intermolecular interaction observed in xtal structures is domain swapping of helices 2 and 3. This requires breaking and reforming the dislfide bond. No evidence to date of domain swapped PrP in vivo.
More on copper binding:
- It has been hypothesized that PrpSc could be a transporter of or sink for copper. Big discrepancies in literature about affinity for PrPC to Cu2+ (femtomolar to micromolar).
- there may be two more Cu2+-binding sites in PrP beyond the octapeptide repeats (one in the C-terminal domain). In animal studies, those given extra copper and those given chelator to reduce copper levels BOTH had delayed onset of symptoms, so relationship between copper and prion pathogenicity seems to be complex.
Folding / Energy Landscapes:
- PrP folding is extremely fast, making conventional kinetics studies (even with stopped-flow devices) challenging.
- There seems to be residual structure present in PrP (native-like) even under strong denaturing conditions, as determined by NMR. This may faciliate rapid folding.
- Some studies have found evidence for a folding intermediate, which would be rate-limiting, as it folds from urea to the native state, but another study showed no evidence for a folding intermediate. When found, the intermediate state is more stable (and more populated) at mildly acidic conditions. Such conditions could be found as PrPc on the membrane is endocytosed – such vesicles are acidified.
- Mutant studies: PrP folding pathway has been found to be very sensitive to changes at the surface of interaction between helix 2/helix 3 (where the majority of genetic mutations map, such as D178N). Most familial mutations lead to a drastic increase in the stability of the fodling intermediate, and there was less of a correlation with how they influenced the global thermodynamic stability of the native state of PrP. Also, the fully unfolded state was always at least 10-fold higher than this folding intermediate, suggesting that the mutants may lead to enahnced misfolding/pathogenicity by populating the folding intermediate, rather than populating the fully unfolded PrP. (This would allow genetic mutants to begin to aggregate even in the absence of exogenous seeds, say from Mad Cows.) (K wonders whether it is neccesary that an intermedaite on the folding pathway would also be an intermediate on the UNFOLDING pathway, which may be the more relevant situation in vivo. She thinks it is not… and should be explcitly tested.)
NMR studies suggest that there may be at least three folded states of PrP: “pertrubed native state” where the C-termini of helix 2 and helix 3 have rearrangedd; a partially denatured state with a hyper-stable core surrounding helix 2 and 3 and the disulfide bond; and a ~molten globule-like state. Note clear to K that the thermodynamic folding intermediate has been associated with any of these structural states.