Prion domains

D. Prion Domains Form the Filament Backbone Other Domains Are... 

Prion domain Middle domain Translation termination domain... 

II. Prion Domains and Functional Domains A. Prion Domains... 

All fungal prion proteins have a so-called prion domain and a functional domain. The prion domain is a region of the polypeptide chain that is necessary and sufficient for prion formation and maintenance (Fig. 1 Wickner et al., 2002). For Ure2p and Sup35p, the functional domain is responsible for the cellular activity of the normal form of the protein. 

In Ure2p and Sup35p, the prion domain is at the N-terminus and the functional domain is at the C-terminus. In Rnqlp and HET-s, these positions are reversed. 

The prion domains of Ure2p, Sup35p, and Rnqlp have unusual amino acid compositions, with abnormally high contents of the polar uncharged residues, Asn and Gin, and relatively low contents of charged and hydrophobic amino acids. In contrast, the amino acid composition of the HET-s prion domain is more typical of a normal globular protein. 

The sizes of the prion domains vary from 70 amino acid residues (HET-s) to 250 amino acid residues (Rnqlp) (Fig. 1), but their borders are not precisely defined by functional criteria since these domains may be subjected to sizeable truncations while retaining their prionogenic properties (Edskes and Wickner, 2002 Ross et al., 2005). Nor are they precisely defined by the extent of N/Q-rich tracts since not all prion domains have them and the preponderance of these residues can vary within a given prion domain. Since the functional domains tend to have globular folds (Section II.B), an operational definition of a prion domain may be that it terminates with the last (or first) residue that is not part of the functional domain, although this definition is applicable only when that fold is known. 

Even then, however, there remains an ambiguity because evidence suggests that at least some prion domains, thus defined, actually consist... 

As with Ure2p, expression of a Sup35p variant lacking the prion domain restores the translation termination activity of the full-length protein in... 

As for HET-s, the structure of its N-terminal domain has yet to be determined however, NMR data suggest that residues 1-227 form a well-folded domain (Balguerie et al, 2003). HET-s is a special case because the prion form is the active form, so that the prion domain (residues 218-289) is needed for function. Although the prion domain alone is sufficient for prion maintenance, it is not competent in heterokaryon incompatibility. For this activity, at least HET-s157-289 must be expressed (Balguerie et al., 2004). 

Filaments of full-length Ure2p are wider than prion domain filaments and they are not smooth-sided (Fig. 5) rather they have a backbone that closely resembles prion domain filaments in width, surrounded by globular domains—presumably, the G-terminal functional domains. This interpretation is supported by the results of protease digestion experiments that trim filaments of full-length Ure2p down to 4-nm core fibrils that closely resemble prion domain filaments assembled de novo (Fig. 5 Baxa et al, 2003). 

Chimeras in which the prion domains of Ure2p and Sup35p are fused with other unrelated proteins also form filaments (Baxa et al., 2002 Diaz-Avalos et al., 2005 King and Diaz-Avalos, 2004 Schlumpberger et al, 2000)... 

Taken together, these observations establish that it is the prion domain alone that is responsible for filament formation. 

Prion domains alone or prion domain fusions of Sup35p (King and Diaz-Avalos, 2004), Ure2p (Brachmann et al., 2005), and HET-s (Nazabal... 

IV. Filament Formation and Prion Conversion Are Based on Amyloidosis of the Prion Domains... 

There is now strong cumulative evidence that in the Ure2p, Sup35p, and HET-s systems, filamentation is based on polymerization of the prion domains into amyloid filaments. In this process, the prion domains undergo a conformational change from a natively unfolded state in which they are sensitive to proteolysis to a compact folded state rich in -structure in which... 

Ure2p filaments are not smooth-sided and their width, at 20 nm, exceeds the range normally associated with amyloid filaments. However, these apparent discrepancies are reconciled by the consideration that it is only the filament backbone of polymerized prion domains that is amyloid (Section III.C and III.D). Its width of 4 nm or so is in the typical amyloid range and these filaments are smooth-sided (Fig. 5). The greater width of Ure2p filaments and the fact that they are not smooth-sided is explained by the amyloid backbone being decorated by still folded globular domains. 

In consideration of thermal stability, a calorimetric investigation of Ure2p prion domain-containing filaments detected no evidence of these domains denaturing up to 105°C (Baxa et al., 2004). In comparison, most proteins denature at temperatures of 50—70°C and rarely exceed 80—90°C, except for proteins of extreme thermophiles. 

Cross-/] structure has been demonstrated for Sup35pNM filaments. Serio et al. (2000) observed a 0.47-nm reflection by X-ray diffraction, and subsequently this reflection was shown to be meridional both by X-ray fiber diffraction (Kishimoto et al., 2004) and electron diffraction (King and Diaz-Avalos, 2004). In the Ure2p system, cross-/ structure has been established by electron diffraction from prion domain filaments preserved in vitreous ice (Fig. 7 Baxa et al, 2005). In addition, a 0.47-nm reflection was detected by both X-ray diffraction and electron diffraction from filament preparations of full-length Ure2p and the Ure2p1 65-GFP fusion, indicating that they contain the same structure (Fig. 7 Baxa et al, 2005). 

As noted above, many natively unfolded proteins become folded in the presence of an appropriate interaction partner. Fungal prion domains subscribe to this paradigm whereby the partner is itself (each other) and the interaction represents homotypic polymerization into amyloid. 

---