Prions of Yeast and Anti-Prion Systems

National Institute Of Diabetes And Digestive And Kidney Diseases

Investigators

Linked publications, trials & patents

Abstract

In 1994 we discovered prions (infectious proteins) infecting yeast (1) analogous to the transmissible spongiform encephalopathies of mammals. URE3 is a prion of Ure2p, and PSI+ is a prion of Sup35p (1,2), each an amyloid of the respective protein (reviewed in ref. 3). Our discovery showed that proteins can be genes. We defined the first prion domain for any protein, a segment necessary and sufficient for prion propagation (2). Unexpectedly, shuffling the prion domain amino acid sequences of Ure2p or Sup35p did not alter the ability of these domains to support prion formation (4, 5), suggesting that the amyloid beta-sheet structure is parallel in-register (6). We showed by solid-state NMR (with Rob Tycko of NIDDK) that the infectious amyloids of Ure2p, Sup35p and Rnq1p are indeed folded, in-register, parallel beta sheets (7, 8). Subsequently, infectious human prion protein amyloid (PrP-Sc) was found to have the same amyloid architecture as our yeast prion amyloids. The folded parallel in-register beta sheet amyloid filament architecture is characterized by each amino acid residue forming a line along the long axis of the filament with side chains of this line of identical residues facing the same direction (perpendicular to both the long axis of the filament and the direction of the peptide bonds). The structure is strongly stabilized by positive interactions among the identical residues: a line of hydrophobic interactions between the side chains of adjacent identical hydrophobic residues OR a line of H-bonds between the side chains of identical hydrophilic residues (Q, N, S, T). Charged residues are very few in the prion/amyloid-forming domains, as they destabilize the structure. A new monomer joining the ends of the filaments must fold at the same places as the molecules already in the filament in order to have this stabilizing structure. Thus, this led us to explain how a given protein sequence can template its conformation, and thus how a protein can act as a gene (9). This is the first and only explanation that has been offered for the templating of protein conformation that is central to the prion phenomenon and amyloid diseases. Prion-forming ability of Ure2p and Sup35p are not conserved among yeast and fungal species (10, 11), and the prion (amyloid)-forming parts of Ure2p and Sup35p have normal non-prion functions. PSI+ and URE3 are rare in wild strains (12, 13) and most variants are toxic or even lethal, showing that these are diseases of yeast (14). Overproducing Btn2p or Cur1p, acting with Hsp42, cures all variants of the URE3 prion. Normal levels of each cure nearly all URE3 prions except those with the highest seed number (15, 16). Btn2p cures by collecting Ure2p aggregates at a single locus (sequestration), increasing the likelihood that one of the progeny cells will not get any prion seeds and so be cured (15). These are antiprion systems. We find that overactive proteasomes prevent prion curing by Btn2 or Cur1 (17) and partially inactive proteasomes result in URE3 curing as a result of dramatically elevated levels of Btn2 and Cur1 (18). We propose that proteasomes overwhelmed by denatured proteins in stress conditions automatically turn on the Btn2 and Cur1 systems to help in the clean-up. The disaggregase Hsp104 is necessary for the propagation of all amyloid-based yeast prions, but cures PSI+ if overexpressed. We find that mutant Hsp104 specifically lacking the prion-curing activity generatates PSI+ 15x more often, and that most PSI+ variants isolated in the Hsp104 mutant are cured by restoration of normal levels of w.t. Hsp104, showing that this is an anti-prion activity (19). Inositol polyphosphates are signalling molecules and facilitate the correct folding of some enzymes. We found Siw14p, a pyrophosphatase specific for 5-diphosphoinositol pentakisphosphate (5PP-IP5) is anti-prion for PSI+ (20). We showed that most PSI+ prions require 5PP-IP5 or related inositol polyphosphates for their propagation, and 1PP-IP5 has a prion-inhibiting action in the absence of the inositol-5 pyrophosphates (20). However, the mechanisms of these effects are still not clear. We found that nonsense-mediated mRNA decay pathway components Upf1, Upf2 and Upf3, at normal expression levels, cure most PSI+ prions arising in their absence (21). The Upf proteins normally complex with the translation termination component, Sup35p, and thereby block most PSI+ prion formation and cure most of those PSI+ prions arising in their absence (21). Upf1p blocks amyloid formation by Sup35p in vitro, and co-localizes with Sup35p aggregates in vivo in PSI+ cells. We infer that normal protein-protein interactions prevent the abnormal protein-protein interactions that produce prions. This is a potentially generalizable approach to prion diseases, in that proteins normally interacting with a prion/amyloid protein may be induced to interact more strongly than normal and reverse the amyloidosis/prion disease. We also find that ribosome-associated chaperones Ssb, Ssz1 and Zuo1, that insure proper folding of nascent proteins, cure most of the PSI+ prions arising in their absence (22). Mutation of any of these genes results in 15-fold elevated prion generation frequency. Triple mutant ssz1 upf1 hsp104T160M cells produce PSI+ up to 5000-fold more often than wild type, but most of these prions arising are cured by replacing any one of the defective genes (23). Thus these antiprion systems act independently, and the real frequency of prions arising in normal cells is much higher than had been appreciated, as most variants arising are cured by these systems before they can be detected in the usual type of test. We used transposon mutagenesis and next-generation sequencing to find proteins that prevent growth defects that would otherwise be produced by the URE3 prion (24). We found that Lug1p/Ylr352wp prevents lethality on non-fermentable carbon sources (e.g. glycerol) that is produced by the URE3 prion in the absence of Lug1p (24). This effect is suppressed by overproduction of Hap4p, a transcription factor promoting expression of mitochondrial-bound proteins. A defect in Gln1p (glutamine synthase) also suppress the growth defect of lug1 URE3 strains. Ure2p is known to be a transcription factor mediating nitrogen catabolite repression. A poor nitrogen source inactivates this Ure2p function, but growth of normal strains on poor nitrogen sources does not require Lug1p, and so identifies a new function for Ure2p. Lug1p is an F-box protein, a substrate-directing subunit of an E3 ubiquitin ligase but its targets are not yet known. We also found that mutation of any of a wide array of chaperones results in a selective disadvantage for URE3 - carrying cells. Thus, cells act to limit the pathology produced on prion infection. We recently identified 19 human proteins whose expression in yeast cures URE3 or PSI+ or both (25). We find that human BAG proteins act by impeding interaction of J-proteins (Hsp40s) with Hsp70s in the process of Hsp104-catalyzed cleavage of prion amyloid filaments, a process essential for prion propagation. BAG proteins are known to interact with human Hsp70s suggesting that they may have similar effects on human prions as they do on yeast prions. Multiple systems prevent prion formation, cure most of the prions that do manage to arise, and limit the pathology from the few prions that evade the other systems (26). Just as we utilize humoral, cellular and innate immune systems to treat or cure or limit the damage from viral and bacterial infections, we suggest that these anti-prion systems will prove to be useful in treatment or prevention of prion/amyloid diseases of humans. 1. Wickner RB (1994) Science 264: 566 - 569. 2. Masison DC & Wickner RB (1995) Science 270: 93 - 95. 3. Wickner RB, Edskes HK, Ross ED, Pierce MM, Baxa U, Brachmann A & Shewmaker F (2004) Ann. Rev. Genetics 38: 681-707. 4. Ross ED, Baxa U, Terry MJ & Wickner RB (2004) Mol. Cell. Biol. 24: 7206-7213. 5. Ross ED, Edskes HK, Terry MJ & Wickner RB (2005). Proc Natl Acad Sci U S A 102: 12825 - 12830. 6. Ross ED, Minton AP & Wickner RB (2005) Nature Cell Biol. 7: 1039-1044. 7. Shewmaker F, Wickner RB & Tycko R (2006) PNAS 103: 19754 - 19759. 8. Gorkovskiy A, Thurber KR, Tycko R, Wickner RB (2014) PNAS 111:E4615-22. 9. Wickner RB, Edskes HK, Shewmaker F, Nakayashiki T 2007 Nat. Rev. Microbiol. 5: 611-618. 10. Edskes HK, Engel A, McCann LM, Brachmann A, Tsai H-F, Wickner RB (2011) Genetics 188:81 90. 11. Edskes HE, Khamar HJ, Winchester C-L, Greenler AJ, Zhou A, McGlinchey RP, Gorkovskiy A, Wickner RB (2014) Genetics, 198: 605-616. 12. Nakayashiki T, Kurtzman CP, Edskes HK, Wickner RB (2005) PNAS 102:10575-80. 13. Kelly AC, Shewmaker FP, Kryndushkin D, Wickner RB (2012) PNAS 109: E2683 - E2690. 14. McGlinchey R, Kryndushkin D, Wickner RB (2011) PNAS 108:5337 - 41. 15. Kryndushkin D, Shewmaker FP, Wickner RB (2008) EMBO J. 27: 2725 - 2735. 16. Wickner RB, Bezsonov E Bateman DA (2014) PNAS 111: E2711-20. 17. Bezsonov EE, Edskes HK, Wickner RB (2021) Genetics 217: doi: 10.1093/genetics/iyab013. 18. Edskes HK, Stroobant EE, DeWilde M, Bezsonov EE, Wickner RB (2021) Genetics 218: doi:10.1093/genetics/iyab037 19. Gorkovskiy A, Reidy M, Masison DC, Wickner RB (2017) PNAS 114: E4193-E4202. 20. Wickner RB, Kelly AC, Bezsonov EE, Edskes HE (2017) PNAS 114: E8402-E8410. 21. Son M, Wickner RB (2018) PNAS 115: E1184-E1193. 22. Son M, Wickner RB (2020) PNAS 117: 26298-26306. 23. Son M, Wickner RB (2022) PNAS 119: e2205500119. 24. Edskes HK, Mukhamedova M, Edskes BK, Wickner RB (2018) Genetics 209:789-800. 25. Wu S, Edskes HE, Wickner RB (2023) Proc Natl Acad Sci U S A. 120:e2314781120. 26. Wickner RB, Hayashi Y, Edskes HK (2025) J. Neurochem. 169:e70045.