Potential physiological roles of prion conversion in yeast
Prion disease is one of the neurodegenerative disorders, and includes Creutzfeldt–Jacob disease and Gerstmann–Straussler–Scheinker syndrome in humans and bovine spongiform encephalopathy and scrapie in animals 1. These disorders are caused by propagation of prions, which are self-replicating infectious β-sheet-rich fibrillar aggregates (amyloid) of the mammalian prion protein (PrP). The “protein-based” inheritance of prions is also observed in yeast and filamentous fungi and fungal prions share common characteristics with mammalian prions 2–4. All yeast prion proteins identified to date contain glutamine/asparagine (Gln/Asn)-rich domains that play critical roles in the formation of self-propagating amyloids 4–7. Remarkably, Alberti et al. 7 have recently showed that approximately 100 Gln/Asn-rich yeast proteins may have the potential to behave as yeast prions.
The potential generality of yeast prions encoded by the yeast genome raises the possibility that yeast might actively utilize reversible prion conversion to regulate certain cellular functions as a molecular switch. This intriguing hypothesis is consistent with the observations made in various yeast prion states. [PSI+], the prion state resulting from aggregation of the translation termination factor Sup35, provides a means to uncover hidden genetic variation and produce new heritable phenotypes 8. Furthermore, [PSI+] yeast show resistance to a range of stressors 9. As the prion conversion of the transcriptional repressor Cyc8 and the chromatin remodeling factor Swi1 is involved in transcriptional regulation of a large fraction of the yeast genome, it can potentially facilitate substantial, complex phenotypic changes 5, 6. In Podospora, [Het-s], the prion state induced by aggregation of a non-Asn/Gln protein HET-s plays a role in heterokaryon incompatibility, which appears to be advantageous to cells 2. Despite these observations, it has yet to be clarified whether eukaryotic cells actively utilize prion conversion to conduct specific cellular functions or whether these observations are simply the consequences of artificially formed prion states that do not occur in nature. How could we distinguish between these two scenarios?
If prion conversion plays an active physiological role in cells, it would be anticipated that such prion states must also be present in wild strains 10. Based on this speculation, the existence of yeast prion states, [PSI+] and [URE3] (the prion state induced by aggregation of the transcriptional regulator Ure2), was investigated in 70 wild strains but none were identified, suggesting that the prion states have detrimental effects on cells. Moreover, a significant fraction of [PSI+] and [URE3] variants was shown to be sick or lethal 10. Although [PIN+], the prion state resulting from aggregation of Rnq1, was found in wild strains 10–12, [PIN+] wild strains were more likely to be heterozygous at the RNQ1 locus than non-prion [pin−] wild strains. This suggests that the polymorphism of RNQ1 was selected to protect cells from the harmful effects of the [PIN+] prion 13, 14. In light of these findings, [PSI+], [URE3], and [PIN+] may represent disease states 10, 13. In contrast, Halfmann et al. 12 investigated the existence of [PSI+] and [MOT+] (the prion state triggered by aggregation of a transcriptional repressor/activator Mot3) states in approximately 700 wild strains and found that 10 out of 690 and 6 out of 96 strains harbored [PSI+] and [MOT+] states, respectively. These prion states were demonstrated to confer a range of diverse phenotypes that enhanced survival compared to non-prion yeast under certain selective conditions.
While the aforementioned studies demonstrated that prion states do indeed exist in nature, it remains to be determined whether organisms under selective pressures of environmental stress regulate cellular functions by active prion state conversion and whether the active induction of prion states is truly linked to physiological cellular adaptation. While over half of the [PSI+] variants were reported to be sick or lethal 13, induction of [PSI+] prion states has been shown to confer a survival advantage under some stress conditions 15. Thus, prion conversion might help an organism adapt to environmental stress. However, our understanding of whether and how prion conversion in response to environmental stress enhances cell survival is limited. Furthermore, the specific mechanisms associated with active prion conversion leading to cellular fitness adaptation remain elusive.
Identification of Mod5 as a novel non-Gln/Asn yeast prion protein
In order to address these issues, we designed and performed a genome-wide screen in yeast to identify proteins whose overexpression and consequential aggregation induces the prion state conversion of a chromosomally expressed Sup35-Q62 chimera protein 16, 17. A novel yeast prion, Mod5, was identified through this screen 18. Mod5, a yeast tRNA isopentenyltransferase, forms self-perpetuating amyloid fibers in vitro, which can induce the formation of the dominantly and cytoplasmically heritable prion states [MOD+], when introduced into yeast in vivo. Structurally, Mod5 does not contain any Gln/Asn-rich domains, and is therefore a prototypical member of a new class of yeast prion proteins.
Acquired antifungal resistance of [MOD+] prion state
Following its identification, we investigated the physiological consequences of the prion conversion from non-prion [mod−] to [MOD+]. Although Mod5 catalyzes the transfer of an isopentenyl group to A37 in the anticodon loop, a previous study suggested that the loss of Mod5 leads to increased ergosterol synthesis by Erg20 because the two enzymes compete for a common substrate, dimethylallyl pyrophosphate 19. [MOD+] yeast, in which Mod5 is aggregated and thus possesses a reduced pool of functional enzyme, indeed contained higher levels of ergosterol. This suggests that [MOD+] cells may have increased antifungal resistance since most common antifungal drugs act on ergosterol directly or indirectly by targeting its synthetic pathway. As expected, [MOD+] yeast are resistant to antifungal drugs such as fluconazole and ketoconazole (Fig. 1). Significantly, culture of [mod−] non-prion yeast in the presence of antifungal drugs increases the de novo formation of [MOD+] colonies. Thus, when yeast that require antifungal resistance for survival are exposed to antifungal drugs, prion conversion from [mod−] to [MOD+] is favored in order to allow the organism to endure the environmental stress.
Prion conversion of Mod5 for cellular adaptation to environmental stress
As prion conversion can affect cellular phenotypes without changing genomic information, it is reasonable that prion conversion acts as a fast on-demand “molecular switch” in response to environmental changes (Fig. 2). Such rapid phenotypic changes would be required particularly if the organism is faced with environmental stressors that can induce cell death. In contrast, a fast phenotypic switch would not be required for adaptation to new environments that are not deleterious. For this type of cellular adaptation, organisms may make use of random genomic mutations to enhance a fitness advantage such that the selection of these mutations over time may lead to evolution.
Prion conversion of Mod5 provides yeast with a flexible, fast, and on-demand strategy to respond to environmental stresses and acquire a fitness advantage. However, another crucial question arises as to whether the stress-resistant prion state can be cured to a non-prion state when the prion-state yeast is returned to non-stressful culture conditions. To this end, [MOD+] shows a slight growth disadvantage compared to [mod−] yeast in nutrient-rich media in the absence of antifungal drugs. Therefore, when a mixture of [MOD+] and [mod−] cells are cultured together without antifungal drugs, the fraction of [MOD+] cells is decreased over time, resulting in the disappearance of [MOD+] phenotypes. Thus, the phenotypic curing of [MOD+] was achieved via the growth disadvantage of [MOD+] yeast under non-stress conditions (Fig. 2). This suggests that one may not be able to easily identify prion states unless the yeast is cultured under stress environments in which prion states may potentially offer beneficial traits. Instead, our results suggest that stressful culture conditions might be more effective in inducing and enriching prion states such that identification of novel prion states becomes more efficient.
Another possible mechanism that causes phenotypic reversion is the loss of infectious prions in a cell either by impairment or enhancement of the prion disaggregation process 20. Such disaggregation activity is mediated by the activities of molecular chaperones such as Hsp104, Hsp70, and Hsp40, which synergistically act to sever prion aggregates. Specifically, by altering the activity of Hsp104, the population of prion state yeast would gradually diminish and the phenotypes associated with the particular prion states would be cured 4, 21. However, of the various yeast prion states, only [PSI+] is known to be returned to a non-prion state by overexpression of Hsp104. Thus, it remains unclear whether an increased expression level of Hsp104 could cure unknown prion states. Furthermore, the rate of prion disaggregation by the molecular chaperones would be influenced by prion (amyloid) conformations with different physical properties 20. Thus, the prion conformation would also be a critical factor for phenotypic reversion by the chaperone machinery.
Stress response by altered tRNA modification
In our study, we have established that yeast can alter cell metabolism and antifungal resistance by converting soluble Mod5 to an amyloid form, which results in adaptation to the environmental stressor and acquisition of a survival advantage. While increased ergosterol levels may confer antifungal resistance, the phenotypic changes stemming from decreased tRNA modification levels caused by Mod5 aggregation may also contribute to the enhanced survival of [MOD+] yeast. How then may altered tRNA modification affect the stress response? Recently, it has been shown that tRNA modifications such as 2′-O-methylcytidine (Cm) and 5-methylcytosine (m5C) increase following hydrogen peroxide exposure and that the loss of enzymes catalyzing tRNA modifications sensitizes the affected yeast to hydrogen peroxide. This indicates that tRNA modifications are critical features of the cellular stress response 22. More generally, the dynamic regulation of tRNA modification may be closely linked to translational control in cellular stress response pathways 23. For instance, in yeast exposed to hydrogen peroxide, the proportion of tRNA (Leu(CAA)) containing m5C at the wobble position is increased and this results in selective translation of mRNA from genes enriched in the TTG codon 24. Interestingly, oxidative stress was demonstrated to increase protein expression from the TTG-enriched ribosomal protein gene Rpl22a, but not its non-enriched paralog. These results reveal the link between tRNA modifications, protein synthesis, and the oxidation stress response for cell survival 24. The modification in the vicinity of the anticodon loop of tRNA by Mod5 could also modulate translation of mRNA, but it remains to be addressed how aggregation of Mod5 in [MOD+] yeast impacts cellular survival by altering protein synthesis.
Prion conversion for environmental adaptation in higher organisms
The cytoplasmic polyadenylation element binding protein (CPEB) from Aplysia, ApCPEB, which stabilizes activity-dependent changes in synaptic efficacy was demonstrated to be assembled into prion-like amyloidogenic self-propagating multimers 25, 26. Prion conversion of ApCPEB increases the capacity to stimulate translation of CPEB-regulated mRNA, which might be involved in the maintenance of long-term synaptic morphology and function related to memory storage. As transient expression of CPEB induces the assembly of stress granules 27, the stimulation of protein synthesis by the prion conversion of CPEB might also be involved in cellular stress responses.Recent findings indicate that endogenous prion conversion in response to environmental selection may be responsible for a broader spectrum of cellular adaptation in living systems than previously appreciated 12, 18. The implications are not limited to biological sciences, but may have substantial economical and human health repercussions. For instance, drug resistance of fungi and insects against plant agricultural chemicals is important in agriculture. In addition, the acquisition of multiple drug resistance by fungi has been a historical problem in medicine. Thus, understanding the mechanism of acquisition of drug resistance via prion conversion may provide a novel strategy to cope with such problems. Furthermore, our study provides new insights for the future examination of endogenous prion conversion in environmental adaptation by higher organisms such as plants and animals.
Conclusions and perspectives
Our results raise the hypothesis that amyloid formation in some forms of neurodegeneration, such as Alzheimer's and Huntington's disease, may represent a positive adaptation for cell survival in response to selective pressure from the environment. A hallmark of many neurodegenerative disorders associated with the onset of middle- or old-age is amyloid formation of the causal proteins in affected brains. Aggregation-prone proteins can form oligomers and misfold into diverse amyloid conformations under different conditions 28–31. Although insoluble amyloids with specific conformations may possess cytotoxic properties, a growing body of evidence has suggested that conformationally aberrant soluble forms of monomeric or oligomeric species may be more toxic to neurons and are responsible for neurodegeneration. Consistent with this notion, amyloid formation observed in these diseases may represent a positive environmental adaptation of aging cells, as it is also conceivable that neurons respond to various stress stimuli such as oxidative stress in the aging cellular environment by converting more toxic soluble monomeric or oligomeric species to lesser or non-toxic amyloid forms in order to gain survival advantages. As the manner by which neurons cope with harmful environmental stresses is critical for cell survival, it will be of significant interest to examine all possible mechanisms that may provide neuroprotection in such conditions. It is therefore tempting to speculate that eukaryotic cells have evolved to struggle with deleterious environments associated with aging by utilizing prion conversion, thereby avoiding the dependence on genomic changes which may potentially be accompanied by the risk of harmful genetic mutation. In light of this speculation, identification of functional prions in response to environmental stress in mammalian systems would be a challenge in the future.
In summary, prion conversion can provide a flexible mechanism for cellular adaptation to gain survival advantages. We propose the hypothesis that phenotypic changes by active prion conversion under selective pressure of environmental stress may be responsible for a broader spectrum of cellular adaptation in living systems.
We thank Kelvin Hui for his comments on the manuscript. Funding was provided by JST PRESTO, Grants from MEXT (Priority Area on Protein Society), the Next Program, the Sumitomo Foundation and the Novartis Foundation (Japan) for the Promotion of Science (M.T.).