TYPE Original Research PUBLISHED 16 January 2023 DOI 10.3389/fpls.2022.1060410 Exploring cryptic amyloidogenic OPEN ACCESS regions in prion-like proteins EDITED BY Paul Martin Harrison, McGill University, Canada from plants REVIEWED BY Sean Cascarina, Carlos Pintado-Grima 1, Jaime Santos 1†, Valentı́n Iglesias 1,2,3†, Colorado State University, United States Zoe Manglano-Artuñedo 1, Irantzu Pallarès 1 Prashant Jethva, Washington University in St. Louis, and Salvador Ventura 1* United States 1 Departament de Bioquı´mica i Biologia Molecular, Institut de Biotecnologia i Biomedicina, *CORRESPONDENCE Universitat Autònoma de Barcelona, Barcelona, Spain, 2 Barcelona Institute for Global Health, Salvador Ventura Barcelona Centre for International Health Research (ISGlobal, Hospital Clı´nic-Universitat de Barcelona), Barcelona, Spain, 3 Nanomalaria Group, Institute for Bioengineering of Catalonia (IBEC), salvador.ventura@uab.es The Barcelona Institute of Science and Technology, Barcelona, Spain † These authors have contributed equally to this work SPECIALTY SECTION This article was submitted to Plant Proteomics and Protein Structural Biology, Prion-like domains (PrLDs) are intrinsically disordered regions (IDRs) of low a section of the journal sequence complexity with a similar composition to yeast prion domains. Frontiers in Plant Science PrLDs-containing proteins have been involved in different organisms’ RECEIVED 03 October 2022 regulatory processes. Regions of moderate amyloid propensity within IDRs ACCEPTED 19 December 2022 have been shown to assemble autonomously into amyloid fibrils. These PUBLISHED 16 January 2023 sequences tend to be rich in polar amino acids and often escape from the CITATION detection of classical bioinformatics screenings that look for highly Pintado-Grima C, Santos J, Iglesias V, Manglano-Artuñedo Z, Pallarès I and aggregation-prone hydrophobic sequence stretches. We defined them as Ventura S (2023) Exploring cryptic cryptic amyloidogenic regions (CARs) and recently developed an integrated amyloidogenic regions in prion-like proteins from plants. database that collects thousands of predicted CARs in IDRs. CARs seem to be Front. Plant Sci. 13:1060410. evolutionary conserved among disordered regions because of their potential to doi: 10.3389/fpls.2022.1060410 stablish functional contacts with other biomolecules. Here we have focused on COPYRIGHT identifying and characterizing CARs in prion-like proteins (pCARs) from plants, © 2023 Pintado-Grima, Santos, Iglesias, Manglano-Artuñedo, Pallarès and a lineage that has been poorly studied in comparison with other prionomes. We Ventura. This is an open-access article confirmed the intrinsic amyloid potential for a selected pCAR from Arabidopsis distributed under the terms of the thaliana and explored functional enrichments and compositional bias of pCARs Creative Commons Attribution License (CC BY). The use, distribution or in plant prion-like proteins. reproduction in other forums is permitted, provided the original KEYWORDS author(s) and the copyright owner(s) are credited and that the original cryptic amyloidogenic regions, prion-like domains, plants, functional publication in this journal is cited, in interactions, bioinformatics accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Abbreviations: CAR, Cryptic amyloidogenic region; pCAR, Prionic cryptic amyloidogenic region; CARs- DB, Cryptic amyloidogenic regions database; PrLDs, Prion-like domains; LCD, Low complexity domain; IDPs, Intrinsically disordered proteins; IDRs, Intrinsically disordered regions; PPIs, Protein-protein interactions; LIPs, Linear interacting motifs; LLPS, Liquid-liquid phase separation; MLOs, Membraneless organelles; MED9, Mediator of RNA polymerase II transcription subunit 9. Frontiers in Plant Science 01 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 1 Introduction arbitrary residue exclusion or window-size length assumption. This prediction scheme allowed us to identify and characterize Since the description of the first self-perpetuating cryptic amyloidogenic regions (CARs) in experimentally proteinaceous infectious pathogens in the late 20th century validated intrinsically disordered regions (IDRs), which were (Prusiner, 1982) and the subsequent discovery of yeast prions previously considered to be depleted of aggregation-prone (Wickner, 1994), the prion concept has progressively evolved regions (Santos et al., 2021). CARs are sequences of mild with our understanding of the phenomenon. Protein domains in amyloidogenic potential, yet able to assemble into fibrils when a variety of polypeptides were found to share similar sequential in isolation. These stretches have likely been conserved within composition to the domains required for yeast prion conversion, IDPs because their potential to act as functional interacting although they do not necessarily show conformational motifs with other biomolecules overpasses a low risk for transitions or mechanisms of self-propagation and were pathogenic aggregation. In fact, CARs consistently overlap termed prion-like domains (PrLDs). Despite initially thought with interacting motifs in IDPs. We collected these predictions to be linked to disease, further characterization of prion-like in a comprehensive database named CARs-DB (Pintado-Grima polypeptides revealed their role in regulating key biological et al., 2022a). functions in a wide variety of organisms (True and Lindquist, Given the generic disordered nature of PrLDs, in this work, 2000; Chakrabortee et al., 2016; Yuan and Hochschild, 2017; we conducted a computational screening to identify CARs in Pallares et al., 2018; Tetz and Tetz, 2018). In this context, under plant PrLDs (pCARs) from different plant species to understand physiological conditions, most PrLDs-containing proteins would their role in this kingdom. We discovered that CARs are not transition towards aggregated states, and instead, they tend widespread in model plants’ PrLDs, with gene ontology to be involved in functional interactions with other proteins and enrichment analysis suggesting they are connected to critical nucleic acids (Alberti et al., 2009; King et al., 2012; Iglesias biological processes. In addition, we validated amyloid et al., 2015). formation for a selected pCAR from Arabidopsis thaliana (A. PrLDs are often enriched in glutamine, asparagine, serine, thaliana), demonstrating that this class of plant protein glycine, and tyrosine (Michelitsch and Weissman, 2000; Alberti sequences bears the potential to spontaneously self-assemble, et al., 2009; Toombs et al., 2010; Diaz-Caballero et al., 2018) and at least in vitro. depleted of hydrophobic amino acids, thus having an overall polar character. The characteristic compositional bias of PrLDs is at the core of diverse predictive bioinformatics tools that aim 2 Materials and methods at identifying such signatures in protein sequences (Toombs et al., 2012; Lancaster et al., 2014; Afsar Minhas et al., 2017). 2.1 Data collection and dataset Most of the current protein aggregation predictors were generation designed to predict highly hydrophobic sequence stretches in disease-associated proteins, a feature that does not match the The reference proteomes of five different plant model polar nature of PrLDs (Fernandez-Escamilla et al., 2004; organisms, including Arabidopsis thaliana (mouse-ear cress; Conchillo-Sole et al., 2007; Garbuzynskiy et al., 2010; Maurer- UP000006548), Zea mays (maize; UP000007305), Oriza sativa Stroh et al., 2010; Sormanni et al., 2015). Only a few of them, like (rice; UP000059680), Solanum lycopersicum (tomato; pWaltz (Sabate et al., 2015b), and its implementation in PrionW UP000004994) and Nicotiana tabacum (tobacco plant; (Zambrano et al., 2015), use adapted aggregation scales aimed to UP000084051) were extracted from UniProt (UniProt, 2021) identify sequences of milder amyloid propensity, denominated (Release 2022_03). PrLDs were screened with PLAAC soft amyloid cores (SACs). Albeit pWaltz has been successfully (Lancaster et al., 2014) using a core length of 60 amino acids used to identify SACs within PrLDs whose correspondent and relative weighting of background probabilities obtained peptides have been experimentally validated to form amyloids from input sequences. PLAAC searches protein sequences to (Sant’anna et al., 2016; Batlle et al., 2017a), this algorithm suffers identify probable prion subsequences exploiting a hidden- from technical limitations, including the exclusion of proline Markov model (HMM) algorithm using the selected residues on its predictions, the identification of a single hit per background probabilities. pCARs longer than six residues were PrLD or the rigidity of its sequence scoring system that obtained from PrLDs by applying the Waltz algorithm at unavoidably employs a sliding window of 21 residues. thresholds 73.5, 80, and 85, as optimized in our original study In recent work, we implemented a pWaltz-related approach of CARs (Santos et al., 2021) and CARs-DB (Pintado-Grima to identify protein regions with amyloid potential in hydrophilic et al., 2022a) to account for polar amyloidogenic regions. The sequential contexts by decreasing the actual detection threshold Waltz algorithm (Maurer-Stroh et al., 2010) explores the of the Walz algorithm (Maurer-Stroh et al., 2010), -a well- sequence determinants of amyloid structure using position- characterized amyloidogenic predictor- without any further specific scoring matrices derived from analyzing a database of Frontiers in Plant Science 02 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 short amyloid-forming and non-forming sequences. Modified of the aggregated sample was incubated for 20 minutes with 900 Waltz thresholds optimized in our original CAR work were µl of CR at a final concentration of 5 µM before recording the selected based on previous experimental evidence (Sabate spectra. CR absorbance spectra were recorded from 400 to 700 et al., 2015b). nm. A control sample with 100 µl of the aggregation buffer was prepared (CR-Free). 2.2 Gene Ontology (GO) enrichment analysis 2.5 Far circular dichroism spectroscopy Functional enrichments were assessed with the database for Far-UV circular dichroism (CD) spectra of the peptide annotation, visualization, and integrated discovery DAVID before and after aggregation were recorded on a Jasco J- (Sherman et al., 2022) using each proteome as background. 815CD spectrometer (Halifax, Canada) at 25°C. The spectra GO terms of biological process (BP_DIRECT), molecular were acquired from 260 nm to 200 nm at 0.2 nm intervals, 2 nm function (MF_DIRECT), and cellular component bandwidth, 2 s of response time, and a 200 nm/min scan speed (CC_DIRECT) were selected for the study with a modified on a 0.1 cm quartz cell. To prevent unwanted self-association Fisher cutoff p-value of 0.1. The Bejamini false discovery rate through the experiment and have a representative spectrum of method was used to control increased error rates for multiple the initial time-point, the initial aliquot was resuspended in 25 tests (Benjamini and Hochberg, 1995). mM Sodium acetate 100 mM NaCl pH 5; 10 accumulations were recorded and averaged for each measurement. 2.3 Peptide preparation and aggregation 2.6 Attenuated total reflectance Fourier The peptide encoding the mediator of RNA polymerase II transform infrared spectroscopy transcription subunit 9 (MED9) pCAR region Ac- QYQQFQQQQHFIQQQQFQ-NH 2 was purchased from Attenuated total reflectance Fourier transform infrared SynPeptide (Shanghai, China) with a purity >95%. To mimic (ATR-FTIR) spectroscopy experiments were performed in a the protein environment, terminal charges were neutralized by Bruker Tensor 27 FTIR spectrometer (Bruker Optics Inc) with N-terminal acetylation and C-terminal amidation. Peptide a Golden Gate MKII ATR accessory. Samples were dried under powder was dissolved in hexafluoroisopropanol to a an N2 (g) stream and measured at a spectral resolution of 2 cm-1 concentration of 1 mg/mL, aliquoted, and vacuum dried with within the 1800–1500 cm-1 range (16 accumulations). Dara a SpeedVac (Thermo Fisher Scientific, Waltham, USA). For the recording and normalization were performed using the OPUS aggregation reactions, aliquots were solubilized with 20 µL MIR Tensor 27 software and deconvoluted with the Peak Fit 4.12 hexafluoroisopropanol and diluted to a concentration of 50 program (Systat Software Inc., San Jose, CA, USA). mM in 20 mM Tris and 100 mM NaCl pH 8. 150 mL of peptide solutions were incubated in a 96 wells plate (non- treated) (Sarstedt, Germany) for two days at 37°C with 2.7 Transmission electron microscopy continuous agitation at 100 RPM. A detailed description of the aggregation conditions is described in the supplementary 5 µL of the aggregated sample, previously diluted to a MIRRAGGE spreadsheet (Martins et al., 2020). concentration of 5 µM, was placed onto glow-discharged carbon-coated copper grids for 1 min. Sample excess was blotted with ashless filter paper, and grids were washed in 2.4 Binding to amyloid dyes distillate water drops. Negative staining was performed with 2% (w/v) uranyl acetate for 1 min. A TEM JEM-1400 (JEOL, End-point samples of the MED9 pCAR were incubated with Peabody, USA) microscope was used, operating at an 40 µM thioflavin-T (Th-T) for 20 minutes, and their accelerating voltage of 120 kV. fluorescence emission spectra were recorded using a Spark plate reader (Tecan, Männedorf, Switzerland). Fluorescence intensity was measured by exciting at 440 nm, bandwidth of 5 3 Results nm, and collecting the emission from 460 to 600 nm with a 1 nm interval. 3.1 Prevalence of pCARs in plant PrLDs Congo-red (CR) binding to end-point samples recovered from the plate was tested using a Cary 100 UV/Vis Since the original characterization of prion domains in yeast Spectrophotometer (Varian, Palo Alto, United States). 100 µl (Chernoff et al., 1995; Serio et al., 1999), PrLDs with similar Frontiers in Plant Science 03 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 features have been identified and studied in other organisms such as viruses (Tetz and Tetz, 2018), bacteria (Iglesias et al., 2015; Yuan and Hochschild, 2017; Choi et al., 2022), protozoan (Pallares et al., 2018) or humans (An and Harrison, 2016; Iglesias et al., 2019). Plants have received less attention, with only a few studies focusing on PrLDs’ coverage and function (Chakrabortee et al., 2016; Dorone et al., 2021; Garai et al., 2021). To identify the potential presence of CARs in PrLDs (pCARs) from plants, the proteomes of five different model higher plants were first screened with the PLAAC algorithm. Around 1% of the total proteome of these species was predicted as PrLDs, spanning from 0.62% (O. sativa) to 1.57% (A. thaliana). In comparison with other studied model organisms (Gil-Garcia et al., 2021), this average is similar to the proportion of human, bacteria, and viruses PrLDs and lower than S. cerevisiae (~3%) (Figure 1; Table 1). For predicted plant PrLDs, pCARs were identified with Waltz at the three designated thresholds 73.5, 80, and 85. As FIGURE 1 expected, the percentage of PrLDs CAR positive (containing at Proportion of predicted PrLDs for five models of plants in least one CAR) increased in all species as the threshold became comparison with other described organisms. Around 1% of the total proteome is predicted as prion-like, similar to the values less stringent, hence allowing the detection of an increasing obtained for humans, bacteria, and viruses. S. cerevisiae represents number of pCARs of polar nature (Figure 2A). Remarkably, at an exceptional case of a proteome highly enriched in PrLDs. threshold 73.5, more than 90% of the predicted PrLDs contained TABLE 1 Identification of pCARs from PrLDs for the five different plant model organisms. PrLDs Waltz PrLDs PrLDs CAR CARs/PrLD Average CAR Organism Proteins PrLDs (%) threshold CARs CAR+ + (%) CAR+ length Arabidopsis thaliana 27474 432 1.57 85 679 303 70.14 2.24 11.6 80 983 356 82.41 2.76 12.66 73.5 1484 402 93.06 3.69 14.69 Zea mays (maize) 56926 404 0.71 85 667 302 74.75 2.21 12.58 80 918 356 88.12 2.58 13.78 73.5 1358 380 94.06 3.57 15.99 Oryza sativa (rice) 43672 270 0.62 85 429 190 70.37 2.26 11.48 80 595 220 81.48 2.7 13.19 73.5 864 253 93.7 3.42 15.52 Solanum lycopersicum (tomato) 34655 405 1.17 85 689 307 75.8 2.24 13 80 957 354 87.41 2.7 14.2 73.5 1366 377 93.09 3.62 15.88 Nicotiana tabacum (tobacco plant) 61673 883 1.43 85 1471 683 77.35 2.15 12.38 80 1989 763 86.41 2.61 13.98 73.5 2913 847 95.92 3.44 15.96 PrLDs were screened with PLAAC, and their corresponding CARs were predicted with Waltz at the three designated thresholds. Frontiers in Plant Science 04 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 A B C FIGURE 2 General statistics for pCARs from A thaliana (blue), Z. mays (green), O. sativa (red), S. lycopersicum (ochre), and N. tabacum (purple), in comparison with general DisProt CARs (black). Most PrLDs are CAR-positive (A), often with more than one predicted pCARs (B). Lower Waltz thresholds contain more pCARs with longer average sequences (C). such regions. According to the analysis, for the 5 tested plant functional interactions. Our Aggrescan algorithm (Conchillo- species, PrLDs tend to display several pCARs, with and average Sole et al., 2007) failed to identify 71.5% (703/983) of A. length of >10 residues (Figures 2B, C). Comparing these metrics thalianas’ pCARs detected at the 80.0 threshold, indicating with generic intrinsically disordered regions (IDRs) extracted that, in the context of PrLDs, they do not confer to the from DisProt (Santos et al., 2021), the positive rate of CARs in containing protein a high risk for aggregation. PrLDs is higher than in IDRs, as well as the number of CARs/ LCD (low complexity domain, either PrLD or IDR) and their size. This indicates that CARs are more prevalent in PrLDs than 3.2 Amino acid composition of pCARs in in IDRs, likely because of the particular amino acid bias of prion A. thaliana LCD sequences. As in IDRs (Santos et al., 2021), the widespread Once the presence of pCARs in all five plant species was representation of CARs in plants is suggestive of a link with confirmed and provided that the proportion of PrLDs and protein functionality. We compared the coincidence of pCARs pCARs were similar, we selected the best-known plant model from A. thaliana with linear interaction motifs (LIPs) as organism A. thaliana for subsequent analyses. described in the MobiDB database (Piovesan et al., 2021), and We compared the amino acid composition of PrLDs from A. observed a 63.5%, 65.6%, and 67.6% of pCAR residues thaliana with that of the complete proteome (Figure 3A). As overlapping with LIPs for the 73.5, 80 and 85 thresholds, expected from the encoded residue preferences in the employed respectively. This suggests that these regions might be PLAAC algorithm, in the PrLDs from A. thaliana are depleted of conserved because of their contribution to establishing hydrophobic residues and enriched in b-breakers (GP), two A B FIGURE 3 pCARs composition by threshold in A thaliana compared with their PrLDs and its proteome. Compared with the proteome, PrLDs already have an intrinsic bias toward polar and b-breakers amino acids. In contrast, pCARs are enriched in hydrophobics and depleted of b-breakers (A). Most polar amino acids are QNs, and aromatic residues seem to modulate the amyloidogenic load of pCARs (B). Barplot draws error bars in the plot with 95% confidence interval. Frontiers in Plant Science 05 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 features that play against the uncontrolled and generic of interpCARs is longer than that of pCARs; 20.4 and 12.7 at aggregation of these disordered regions (Rousseau et al., 2006; threshold 80.0, respectively. Theillet et al., 2013). Indeed, glycines and prolines (G+P) alone In a way, the distribution and biased composition of pCARs equal the total content of hydrophobic residues, around 25%, within PrLDs is reminiscent of the proposed stickers and spacers whereas, in the total proteome, non-polar residues are threefold model of LLPS (Wang et al., 2018; Martin et al., 2020; Holehouse more abundant than G+P. Next, we compared pCARs et al., 2021; Bremer et al., 2022); with the Y+F+W+N enriched composition against PrLDs. pCARs exhibit a higher pCARs acting as stickers that facilitate protein interactions and percentage of hydrophobic residues and a lower content in b- the P+G+H+R enriched interpCARs as spacers, connecting breakers than PrLDs. As expected, the proportion of G+P in stickers and providing them the adequate solubility and pCARs decreases as the Waltz threshold becomes more disorder context. stringent. The percentage of hydrophilic residues in pCARs resembles that in PrLDs, being insensitive to the applied threshold, suggesting that they might provide background 3.3 Functional enrichments of pCARs in control to the equilibrium between disorder and assembly in A. thaliana these sequences. To better dissect the relevant physicochemical features To explore the potential contribution of pCARs to the hidden in these amino acid groups, we split them into seven biological roles of A. thaliana prion-like proteins, we categories: Q+N, charged, rest of polars, tyrosine, rest of performed a functional enrichment analysis using the DAVID aromatics, aliphatics, and b-breakers. As observed in database (Sherman et al., 2022). Figure 3B, only two residues (Q+N) account for almost 25% of Regarding the molecular functions of pCARs, significant the amino acids in PrLDs, a situation that contrasts with these enrichments were found for regulative processes, especially for residues’ low contribution to this plant proteome composition nucleic acids, including DNA, RNA and mRNA binding, but (7.9%). Asp and Gln are known to play a critical role in prion also protein binding (Figure 5; Supplementary Table 1). This is conformation (Perutz et al., 2002; Derkatch et al., 2004; Jiang interesting, as CARs have already been described to mediate key et al., 2004; Sabate et al., 2015b; Zambrano et al., 2015; Zhang PPIs in IDPs (Santos et al., 2021; Pintado-Grima et al., 2022a). It et al., 2016). Interestingly enough, Q+N are more frequent in is not surprising that, provided the previously proposed pCARs than in complete PrLDs, their number increasing with regulatory function of CARs, pCARs are found enriched in the stringency of the prediction, indicative that they would be biological activities of gene expression, transcription, and important contributors to the homo- or heterotypic interactions translation. Related enrichments have been described in these stretches might establish. The higher proportion of amyloidogenic (Antonets and Nizhnikov, 2017) and prion-like hydrophobic amino acids in pCARs, relative to its PrLDs containers, can be traced to increases in both aromatic and aliphatic residues. However, it seems that the presence of aromatic residues is a distinctive feature of pCARs, their proportion increasing with the detection threshold. Specifically, tyrosines (Y) account in all the cases for more than 50% of the total aromatics, likely due to its dual polar/ apolar character and its unique hydrogen bonding capability. This bias toward Y is reminiscent of what has been observed in liquid-liquid phase separation (LLPS), where phenylalanine and tyrosine are not interchangeable (Wang et al., 2018), likely because Y best balances solubility and assembly potential (Diaz-Caballero et al., 2018). Not surprisingly, regions between pCARs (interpCARs) also differ in residues composition from pCARs (Figure 4). Compared with pCARs, interpCARs are highly enriched in G +P and depleted in aromatic residues. Despite Q+N are also FIGURE 4 Residue composition comparison between A. thaliana pCARs abundant in interpCARs there is a strong preference for Q over and interpCARs (regions between two pCARs in any PrLD N in these regions, indicating that their properties are not sequence) at threshold 80. pCARs are enriched in aromatic entirely exchangeable. Another interesting observation is that residues and depleted of b-breakers (GP). An opposite tendency is observed for interpCARs, which seem to compensate the the cationic arginine (R) and histidine (H) residues are preferred sequential amyloidogenic load by reducing the aromatics and at interpCARs, but this bias does not apply to other charged increasing b-breakers. residues like lysine, glutamate, and aspartate. The average length Frontiers in Plant Science 06 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 FIGURE 5 GO enrichments of pCARs from A. thaliana by threshold. Significant enrichments were found in cellular components (CC), biological processes (BP), and molecular functions (MF) at the three designated thresholds: 73.5 (green), 80 (yellow), and 85 (blue). pCARs are mainly associated with nuclear components and MLOs and are found in several regulatory processes and plant-specific functions. (Garai et al., 2021) proteins from plants. The fact that a RNA binding proteins (Mann and Donnelly, 2021), with many significant proportion of PrLDs is pCAR-positive makes it of them bearing PrLDs (Harrison and Shorter, 2017; Gotor challenging to disentangle the contributions of the complete et al., 2020). PrLD sequence and pCARs only in a given protein to the detected enrichments. However, even when the A. thaliana prionome is used as background for the GO analysis, 3.4 MED9 pCAR form amyloid fibrils significant enrichments in regulatory processes such as regulation of transcription (BP), transcription factor activity The concept of CARs has been gaining strength by means of (MF), or nucleus (CC) were still observed (Supplementary the continuous experimental validation of predicted segments of Figure 1). A. thaliana’s pCARs are also enriched in plant- hydrophilic nature that form amyloid fibrils in vitro (Santos specific biological processes such as leaf and flower et al., 2021; Pintado-Grima et al., 2022a). Our goal is to increase development or root meristem growth. In fact, there is already the body of evidence on the intrinsic amyloidogenic potential of experimental evidence for plant PrLDs being involved in the these sequences, when they are disconnected from their autonomous flowering pathway, specifically for the interpCARs regions. Accordingly, we examined a new pCAR luminidependens PrLD from A. thaliana (Chakrabortee from A. thaliana’s MED9 protein (Q8RWA2) in this work. et al., 2016). MED9, the mediator of RNA polymerase II transcription pCARs are found to be associated with nuclear components, subunit 9, is a nuclear protein that is part of the mediator ribonuclear complexes, and nucleus, the latter with the lowest p- complex. It acts as a coactivator involved in the transcriptional values observed (p<10-61) among all GO terms analyzed. Certain regulation of almost all RNA polymerase II-related genes. The membraneless organelles (MLOs), such as cytoplasmatic stress N-terminal domain of MED9 (1-149) is predicted as a PrLD by granules or P-bodies were also enriched in pCARs. PrLDs are PLAAC. Within this PrLD, an 18-residue pCAR comprising recruited to MLOs in response to different stimuli (Boncella residues 77 to 94 (77-QYQQFQQQQHFIQQQQFQ-94) et al., 2020; Iglesias et al., 2021), and they need to have some emerges as an interesting pCAR that we selected by its overall condensate driving element that facilitates the formation of polar composition (67%), significant aromatic content (22%), liquid droplets. This would be especially relevant when the low Waltz score (77.87), and for being undetectable by well- protein phase separates by itself, in the absence of additional established aggregation and amyloid predicting algorithms like co-factors (Pintado-Grima et al., 2022b); as described, pCARs Aggrescan (Conchillo-Sole et al., 2007; De Groot et al., 2012), might act as sticker regions facilitating PrLDs’ coalescence and Tango (Fernandez-Escamilla et al., 2004) or ZipperDB the stabilization of these cellular compartments (Tsoi et al., (Thompson et al., 2006). 2021). In general, RNA and mRNA binding and processing are MobiDB (Piovesan et al., 2021) consistently predicts MED9 associated with pCARs. This might reflect a situation in which PrLD to belong to a long-disordered region (1-150) (Figure 6A). many of the protein constituents of MLOs are originally nuclear AlphaFold (Jumper et al., 2021; Varadi et al., 2022) predict that Frontiers in Plant Science 07 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 A B C D E F FIGURE 6 Experimental characterization of the predicted pCAR derived peptide in MED9 from A thaliana. (A) MED9 diagram showing the location of the predicted disordered regions (yellow), PrLD (violet), secondary structures (blue) and the selected pCAR (red). The sequence of pCAR is shown in the red box. (B) Far-UV circular dichroism spectra of the peptide, evolution from the initial disordered soluble state into mature amyloid fibrils at 50 mM and 48 hours of incubation at 37 °C with continuous agitation at 100 RPM. (C) Thioflavin-T fluorescence emission spectrum when excited at 440nm in the absence (dashed line) and presence (solid line) of the aggregated sample; note the characteristic fluorescence enhancement at 480 nm when the dye is bound to amyloid-like aggregates. (D) CR spectral changes in the absence (dashed line) and in the presence (solid line) of the incubated peptide. (E) FTIR spectrum in the amide I region of the incubated sample. The red line corresponds to the absorbance spectra and the blue area indicates the inter-molecular b-sheet contribution to the total area upon Gaussian deconvolution. (F) Representative TEM micrographs illustrating the fibrils of the incubated peptide. Frontiers in Plant Science 08 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 residues 32-64 and 70-97 might form two alpha-helices, which, than the hydrophobic stretches often identified in the core of in the absence of tertiary contacts, are probably only attained pathogenic amyloid fibrils; to keep the protein soluble but allow upon binding to a molecular partner (i.e., coiled-coil formation). specific inter-molecular interactions. In this scenario, the The MED9 pCAR maps within the second predicted helix concept of CARs, whose existence has already been validated (Figure 6A), suggesting a potential implication in PPIs in generic IDPs (Santos et al., 2021), is translated here to PrLDs, through a folding upon binding mechanism, as reported in a since these domains are, by definition, intrinsically disordered. significant proportion of previously characterized CARs (Santos In the PrLD context, pCARs would define the boundaries of the et al., 2021). regions that would concentrate this mild amyloid potential, To assess the amyloidogenic potential of MED9 pCAR, we flanked by longer and highly soluble interpCARs segments, analyzed whether a peptide corresponding to its sequence was acting analogously to the way stickers and spacers do in LLPS able to self-assemble into amyloid fibrils in vitro. We incubated (Martin et al., 2020). the peptide at 50 mM in 20 mM Tris and 100 mM NaCl pH 8 for pCAR composition suggests that these regions within PrLDs 48 hours at 37 °C and 100RPM. After this time, the initial CD arise from a reduction of b-breakers and an increase in signal corresponding to the disordered soluble peptide was hydrophobic residues, especially of an aromatic character, significantly reduced and shifted to a faint b-sheet signature, relative to the rest of the domain. In addition, Tyr seems to indicating protein aggregation (Figure 6B). The end-point play a unique role in pCARs, since it units aromaticity and sample was positive for amyloid dye binding (Th-T and CR), polarity, allowing interactions with water in the soluble state and suggesting the formation of amyloid-like structures (Figures 6C, pi-pi interactions in the assembled state. Notably, the content in D). To assess the secondary structure content of the insoluble polar residues of pCARs is equal to or higher than that of the aggregates not accessible by CD, we used FTIR and recorded the complete PrLDs or even the whole proteome, which ensures that amine I region of the infrared spectra (1700-1600 cm-1). We their amyloid propensity is framed in an overall soluble observed a dominant contribution of the 1630 cm-1 band (70% background. In this context, the polar Gln and specially Asn of the area) associated with the formation of amyloid-like residues, which are highly enriched in PrLDs relative to the intermolecular b-sheets (Figure 6E). A morphological analysis proteome but also in pCARs relative to these domains, would of the aggregates by TEM confirmed the fibrillar nature of the also contribute to control the amyloid load, since their side aggregates (Figure 6F). chains have a high tendency to form hydrogen bonds with the Together, this data confirms that the MED9 pCAR has the solvent and thus enable solubility, but at the same time, they can inherent potential to aggregate into amyloid fibrils. form very tight interactions within the dehydrated interfaces of Furthermore, the coincidence of predicted coiled-coil and b-strands (Balbirnie et al., 2001; Peccati et al., 2020). The amyloid propensities in the same sequence has been previously preference for Asn relative to Gln in pCARs seems to make observed in a large number of human PrLDs and yeast prions sense since Asn richness promotes the formation of benign (Fiumara et al., 2010; Batlle et al., 2021). Indeed, a human assemblies, whereas Gln richness promotes the formation of member of the mediator complex, MED15, is a bona fide toxic aggregates (Halfmann et al., 2011). prion-like protein whose PrLD transits between disordered, The present analysis uncovered a previously unexplored coiled-coil, and amyloid states (Batlle et al., 2021). As putative amyloid sequence space in plants PrLDs, relative to mentioned above, the MED9 pCAR is just one example of a previous studies, including ours; where amyloid-prone regions conjunct of amyloid-competent sequences not detectable by identification was limited by rationally implemented but classical aggregation predictors like our own Aggrescan arbitrary requirements, including a fixed size, the absence of (Conchillo-Sole et al., 2007). Pro or the belief that a single region would suffice to nucleate aggregation (Sabate et al., 2015a; Batlle et al., 2017b; Fernandez et al., 2017), essentially because the bias toward studying disease- 4 Discussion associated amyloids made us wrongly thought that highly hydrophilic sequences, like pCARs, could not form amyloid The prion phenomenon continues to attract high interest. In structures autonomously. Importantly, the presence of these recent times, significant effort has been devoted to cryptic amyloidogenic regions seems to relate to RNA- and understanding the role PrLDs play in biological pathways mRNA-related processes occurring in the nucleus, highlighting across species, beyond those initially described in yeast prions. the significant role that PrLDs containing CARs might play in Although PrLDs seem to share a moderate amyloid propensity regulating plant-specific biochemical pathways. (Sabate et al., 2015b), the containing proteins do not always Although the pCAR concept indeed requires extensive behave as bona fide prions, but rather exploit these LCDs to experimental validation before their presence can be established promote the functional interactions needed for their activity. as a general principle modulating the activity of prion-like Amyloid regions in PrLDs need to be milder and more polar proteins, it is also certain that for most of the experimentally Frontiers in Plant Science 09 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 validated CARs, structural analysis indicated that these initially version of the manuscript. All authors have read and agreed to disordered regions might fold upon interaction with a partner, or, the published version of the manuscript. alternatively, that they mediate multimerization. This suggests that its higher amyloidogenic load, relative to the rest of the PrLD sequence, cannot be purged out by natural selection because CARs Funding are needed for functional interactions. Surprisingly, in many instances, the folded structures adopted by CARs in functional This work was funded by European Union’s Horizon 2020 complexes correspond to a-helices, often involved in coiled-coil research and innovation programme under GA 952334 dimerization (Batlle et al., 2021). Chameleon sequences adopt an (PhasAGE). SV was supported by the Spanish Ministry of a-helix in the folded state of globular proteins but display a Science and Innovation (PID 2019-105017RB-I00), by ICREA hidden propensity to form b-sheets when unfolded, and a-helix/ and ICREA Academia 2020. CP-G was supported by the b-strand-discordant stretches were proposed to be associated with Secretariat of Universities and Research of the Catalan amyloid fibril formation already in 2001 (Kallberg et al., 2001). Government and the European Social Fund (2021 FI_B What makes special CARs, such as those of A. thaliana MED9 or 00087). JS was supported by the Spanish Ministry of Science human MED15, is that they might populate three different states, and Innovation via a doctoral Grant (FPU17/01157). VI was disordered in isolation, a-helix in functional PPIs, and b-sheets if supported by the Spanish Ministry of Universities and the they eventually transition to the amyloid state; with the three European Union-Next Generation EU (ruling 02/07/2021, propensities encoded in the same sequence but controlled by the Universitat Autònoma de Barcelona). protein microenvironment and the sequential context. We believe the approach described in the present study may translate to the detection of pCARs in other interesting Conflict of interest prionomes, to build a comprehensive repository of polar amyloidogenic regions for further experimental validation and The authors declare that the research was conducted in the benchmarking future predictive algorithms. absence of any commercial or financial relationships that could be construed as a potential conflict of interest. Data availability statement Publisher’s note The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession All claims expressed in this article are solely those of the number(s) can be found in the article/Supplementary Material. authors and do not necessarily represent those of their affiliated The datasets are also available online at: http://carsdb.ppmclab. organizations, or those of the publisher, the editors and the com/prionplants. reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher. Author contributions CP-G, JS, and VI generated the dataset. CP-G and VI Supplementary material performed the analyses. ZM-A conducted the experimental validation. CP-G, IP, and SV conceptualized the project. CP- The Supplementary Material for this article can be found G, JS, and SV wrote the manuscript with the contributions of all online at: https://www.frontiersin.org/articles/10.3389/ authors. SV and IP acquired funding and reviewed the final fpls.2022.1060410/full#supplementary-material References Afsar Minhas, F. U. A., Ross, E. D., and Ben-Hur, A. (2017). Amino acid An, L., and Harrison, P. M. (2016). The evolutionary scope and neurological composition predicts prion activity. PloS Comput. Biol. 13, e1005465. doi: 10.1371/ disease linkage of yeast-prion-like proteins in humans. Biol. Direct. 11, 32. doi: journal.pcbi.1005465 10.1186/s13062-016-0134-5 Alberti, S., Halfmann, R., King, O., Kapila, A., and Lindquist, S. (2009). A Antonets, K. S., and Nizhnikov, A. A. (2017). Predicting amyloidogenic systematic survey identifies prions and illuminates sequence features of proteins in the proteomes of plants. Int. J. Mol. Sci. 18. doi: 10.3390/ prionogenic proteins. Cell 137, 146–158. doi: 10.1016/j.cell.2009.02.044 ijms18102155 Frontiers in Plant Science 10 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 Balbirnie, M., Grothe, R., and Eisenberg, D. S. (2001). An amyloid-forming Halfmann, R., Alberti, S., Krishnan, R., Lyle, N., O’donnell, C. W., King, O. D., peptide from the yeast prion Sup35 reveals a dehydrated beta-sheet structure for et al. (2011). Opposing effects of glutamine and asparagine govern prion formation amyloid. Proc. Natl. Acad. Sci. U.S.A. 98, 2375–2380. doi: 10.1073/pnas.041617698 by intrinsically disordered proteins. Mol. Cell 43, 72–84. doi: 10.1016/ Batlle, C., Calvo, I., Iglesias, V., Lynch, C. ,. J., Gil-Garcia, M., Serrano, M., et al. j.molcel.2011.05.013 (2021). MED15 prion-like domain forms a coiled-coil responsible for its amyloid Harrison, A. F., and Shorter, J. (2017). RNA-Binding proteins with prion-like conversion and propagation. Commun. Biol. 4, 414. doi: 10.1038/s42003-021- domains in health and disease. Biochem. J. 474, 1417–1438. doi: 10.1042/BCJ20160499 01930-8 Holehouse, A. S., Ginell, G. M., Griffith, D., and Boke, E. (2021). Clustering of Batlle, C., De Groot, N. S., Iglesias, V., Navarro, S., and Ventura, S. (2017a). aromatic residues in prion-like domains can tune the formation, state, and Characterization of soft amyloid cores in human prion-like proteins. Sci. Rep. 7, organization of biomolecular condensates. Biochemistry 60, 3566–3581. doi: 12134. doi: 10.1038/s41598-017-09714-z 10.1021/acs.biochem.1c00465 Batlle, C., Iglesias, V., Navarro, S., and Ventura, S. (2017b). Prion-like proteins Iglesias, V., De Groot, N. S. , and Ventura, S. (2015). Computational analysis of and their computational identification in proteomes. Expert Rev. Proteomics 14, candidate prion-like proteins in bacteria and their role. Front. Microbiol. 6, 1123. 335–350. doi: 10.1080/14789450.2017.1304214 doi: 10.3389/fmicb.2015.01123 Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: A Iglesias, V., Paladin, L., Juan-Blanco, T., Pallares, I., Aloy, P., Tosatto, S. C. E., practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B et al. (2019). In silico characterization of human prion-like proteins: Beyond (Methodol.) 57, 289–300. doi: 10.1111/j.2517-6161.1995.tb02031.x neurological diseases. Front. Physiol. 10, 314. doi: 10.3389/fphys.2019.00314 Boncella, A. E., Shattuck, J. E., Cascarina, S. M., Paul, K. R., Baer, M. H., Iglesias, V., Santos, J., Santos-Suarez, J., Pintado-Grima, C., and Ventura, S. Fomicheva, A., et al. (2020). Composition-based prediction and rational (2021). SGnn: A web server for the prediction of prion-like domains recruitment to manipulation of prion-like domain recruitment to stress granules. Proc. Natl. stress granules upon heat stress. Front. Mol. Biosci. 8, 718301. doi: 10.3389/ Acad. Sci. U.S.A. 117, 5826–5835. doi: 10.1073/pnas.1912723117 fmolb.2021.718301 Bremer, A., Farag, M., Borcherds, W. M., Peran, I., Martin, E. W., Pappu, R. V., Jiang, Y., Li, H., Zhu, L., Zhou, J. M., and Perrett, S. (2004). Amyloid nucleation et al. (2022). Deciphering how naturally occurring sequence features impact the and hierarchical assembly of Ure2p fibrils. role of asparagine/glutamine repeat and phase behaviours of disordered prion-like domains. Nat. Chem. 14, 196–207. doi: nonrepeat regions of the prion domains. J. Biol. Chem. 279, 3361–3369. doi: 10.1038/s41557-021-00840-w 10.1074/jbc.M310494200 Chakrabortee, S., Kayatekin, C., Newby, G. A., Mendillo, M. L., Lancaster, A., Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., et al. and Lindquist, S. (2016). Luminidependens (LD) is an arabidopsis protein with (2021). Highly accurate protein structure prediction with AlphaFold. Nature 596, prion behavior. Proc. Natl. Acad. Sci. U.S.A. 113, 6065–6070. doi: 10.1073/ 583–589. doi: 10.1038/s41586-021-03819-2 pnas.1604478113 Kallberg, Y., Gustafsson, M., Persson, B., Thyberg, J., and Johansson, J. (2001). Chernoff, Y. O., Lindquist, S. L., Ono, B., Inge-Vechtomov, S. G., and Liebman, Prediction of amyloid fibril-forming proteins. J. Biol. Chem. 276, 12945–12950. doi: S. W. (1995). Role of the chaperone protein Hsp104 in propagation of the yeast 10.1074/jbc.M010402200 prion-like factor [psi+]. Science 268, 880–884. doi: 10.1126/science.7754373 King, O. D., Gitler, A. D., and Shorter, J. (2012). The tip of the iceberg: RNA- Choi, K. J., Quan, M. D., Qi, C., Lee, J. H., Tsoi, P. S., Zahabiyon, M., et al. (2022). binding proteins with prion-like domains in neurodegenerative disease. Brain Res. NANOG prion-like assembly mediates DNA bridging to facilitate chromatin 1462, 61–80. doi: 10.1016/j.brainres.2012.01.016 reorganization and activation of pluripotency. Nat. Cell Biol. 24, 737–747. doi: Lancaster, A. K., Nutter-Upham, A., Lindquist, S., and King, O. D. (2014). 10.1038/s41556-022-00896-x PLAAC: a web and command-line application to identify proteins with prion-like Conchillo-Sole, O., De Groot, N. S., Aviles, F. X., Vendrell, J., Daura, X., and amino acid composition. Bioinformatics 30, 2501–2502. doi: 10.1093/ Ventura, S. (2007). AGGRESCAN: a server for the prediction and evaluation of bioinformatics/btu310 “hot spots” of aggregation in polypeptides. BMC Bioinf. 8, 65. doi: 10.1186/1471- Mann, J. R., and Donnelly, C. J. (2021). RNA Modulates physiological and 2105-8-65 neuropathological protein phase transitions. Neuron 109, 2663–2681. doi: 10.1016/ De Groot, N. S., Castillo, V., Grana-Montes, R., and Ventura, S. (2012). j.neuron.2021.06.023 AGGRESCAN: method, application, and perspectives for drug design. Methods Martin, E. W., Holehouse, A. S., Peran, I., Farag, M., Incicco, J. J., Bremer, A., Mol. Biol. 819, 199–220. doi: 10.1007/978-1-61779-465-0_14 et al. (2020). Valence and patterning of aromatic residues determine the phase Derkatch, I. L., Uptain, S. M., Outeiro, T. F., Krishnan, R., Lindquist, S. L., and behavior of prion-like domains. Science 367, 694–699. doi: 10.1126/ Liebman, S. W. (2004). Effects of Q/N-rich, polyQ, and non-polyQ amyloids on the science.aaw8653 de novo formation of the [PSI+] prion in yeast and aggregation of Sup35 in vitro. Martins, P. M., Navarro, S., Silva, A., Pinto, M. F., Sarkany, Z., Figueiredo, F., Proc. Natl. Acad. Sci. U.S.A. 101, 12934–12939. doi: 10.1073/pnas.0404968101 et al. (2020). MIRRAGGE - minimum information required for reproducible Diaz-Caballero, M., Navarro, S., Fuentes, I., Teixidor, F., and Ventura, S. (2018). AGGregation experiments. Front. Mol. Neurosci. 13, 582488. doi: 10.3389/ Minimalist prion-inspired polar self-assembling peptides. ACS Nano 12, 5394– fnmol.2020.582488 5407. doi: 10.1021/acsnano.8b00417 Maurer-Stroh, S., Debulpaep, M., Kuemmerer, N., Lopez de la Paz, M., Martins, Dorone, Y., Boeynaems, S., Flores, E., Jin, B., Hateley, S., Bossi, F., et al. (2021). A I. C., Reumers, J., et al. (2010). Exploring the sequence determinants of amyloid prion-like protein regulator of seed germination undergoes hydration-dependent structure using position-specific scoring matrices. Nat. Methods 7, 237–242. doi: phase separation. Cell 184, 4284–4298 e4227. doi: 10.1016/j.cell.2021.06.009 10.1038/nmeth.1432 Fernandez, M. R., Batlle, C., Gil-Garcia, M., and Ventura, S. (2017). Amyloid Michelitsch, M. D., and Weissman, J. S. (2000). A census of glutamine/ cores in prion domains: Key regulators for prion conformational conversion. Prion asparagine-rich regions: implications for their conserved function and the 11, 31–39. doi: 10.1080/19336896.2017.1282020 prediction of novel prions. Proc. Natl. Acad. Sci. U.S.A. 97, 11910–11915. doi: Fernandez-Escamilla, A. M., Rousseau, F., Schymkowitz, J., and Serrano, L. 10.1073/pnas.97.22.11910 (2004). Prediction of sequence-dependent and mutational effects on the Pallares, I., De Groot, N. S., Iglesias, V., Sant’anna, R., Biosca, A., Fernandez- aggregation of peptides and proteins. Nat. Biotechnol. 22, 1302–1306. doi: Busquets, X., et al. (2018). Discovering putative prion-like proteins in plasmodium 10.1038/nbt1012 falciparum: A computational and experimental analysis. Front. Microbiol. 9, 1737. doi: 10.3389/fmicb.2018.01737 Fiumara, F., Fioriti, L., Kandel, E. R., and Hendrickson, W. A. (2010). Essential role of coiled coils for aggregation and activity of Q/N-rich prions and PolyQ Peccati, F., Diaz-Caballero, M., Navarro, S., Rodriguez-Santiago, L., Ventura, S., proteins. Cell 143, 1121–1135. doi: 10.1016/j.cell.2010.11.042 and Sodupe, M. (2020). Atomistic fibrillar architectures of polar prion-inspired heptapeptides. Chem. Sci. 11, 13143–13151. doi: 10.1039/D0SC05638C Garai, S., Citu,, Singla-Pareek, S. L., Sopory, S. K., Kaur, C., and Yadav, G. (2021). Complex networks of prion-like proteins reveal cross talk between stress and memory Perutz, M. F., Pope, B. J., Owen, D., Wanker, E. E., and Scherzinger, E. (2002). pathways in plants. Front. Plant Sci. 12, 707286. doi: 10.3389/fpls.2021.707286 Aggregation of proteins with expanded glutamine and alanine repeats of the glutamine-rich and asparagine-rich domains of Sup35 and of the amyloid beta- Garbuzynskiy, S. O., Lobanov, M. Y., and Galzitskaya, O. V. (2010). peptide of amyloid plaques. Proc. Natl. Acad. Sci. U.S.A. 99, 5596–5600. doi: FoldAmyloid: a method of prediction of amyloidogenic regions from protein 10.1073/pnas.042681599 sequence. Bioinformatics 26, 326–332. doi: 10.1093/bioinformatics/btp691 Pintado-Grima, C., Barcenas, O., Manglano-Artunedo, Z., Vilaca, R., Macedo- Gil-Garcia, M., Iglesias, V., Pallares, I., and Ventura, S. (2021). Prion-like Ribeiro, S., Pallares, I., et al. (2022a). CARs-DB: A database of cryptic proteins: from computational approaches to proteome-wide analysis. FEBS Open amyloidogenic regions in intrinsically disordered proteins. Front. Mol. Biosci. 9, Bio 11, 2400–2417. doi: 10.1002/2211-5463.13213 882160. doi: 10.3389/fmolb.2022.882160 Gotor, N. L., Armaos, A., Calloni, G., Torrent Burgas, M., Vabulas, R. M., De Pintado-Grima, C., Barcenas, O., and Ventura, S. (2022b). In-silico analysis of Groot, N. S., et al. (2020). RNA-Binding and prion domains: the yin and yang of pH-dependent liquid-liquid phase separation in intrinsically disordered proteins. phase separation. Nucleic Acids Res. 48, 9491–9504. doi: 10.1093/nar/gkaa681 Biomolecules 12. doi: 10.3390/biom12070974 Frontiers in Plant Science 11 frontiersin.org Pintado-Grima et al. 10.3389/fpls.2022.1060410 Piovesan, D., Necci, M., Escobedo, N., Monzon, A. M., Hatos, A., Micetic, I., segments of proteins. Proc. Natl. Acad. Sci. U.S.A. 103, 4074–4078. doi: 10.1073/ et al. (2021). MobiDB: intrinsically disordered proteins in 2021. Nucleic Acids Res. pnas.0511295103 49, D361–D367. doi: 10.1093/nar/gkaa1058 Toombs, J. A., Mccarty, B. R., and Ross, E. D. (2010). Compositional Prusiner, S. B. (1982). Novel proteinaceous infectious particles cause scrapie. determinants of prion formation in yeast. Mol. Cell Biol. 30, 319–332. doi: Science 216, 136–144. doi: 10.1126/science.6801762 10.1128/MCB.01140-09 Rousseau, F., Serrano, L., and Schymkowitz, J. W. (2006). How evolutionary Toombs, J. A., Petri, M., Paul, K. R., Kan, G. Y., Ben-Hur, A., and Ross, E. D. pressure against protein aggregation shaped chaperone specificity. J. Mol. Biol. 355, (2012). De novo design of synthetic prion domains. Proc. Natl. Acad. Sci. U.S.A. 1037–1047. doi: 10.1016/j.jmb.2005.11.035 109, 6519–6524. doi: 10.1073/pnas.1119366109 Sabate, R., Rousseau, F., Schymkowitz, J., Batlle, C., and Ventura, S. (2015a). True, H. L., and Lindquist, S. L. (2000). A yeast prion provides a mechanism for Amyloids or prions? that is the question. Prion 9, 200–206. doi: 10.1080/ genetic variation and phenotypic diversity. Nature 407, 477–483. doi: 10.1038/ 19336896.2015.1053685 35035005 Sabate, R., Rousseau, F., Schymkowitz, J., and Ventura, S. (2015b). What makes a Tsoi, P. S., Quan, M. D., Choi, K. J., Dao, K. M., Ferreon, J. C., and Ferreon, A. C. protein sequence a prion? PloS Comput. Biol. 11, e1004013. doi: 10.1371/ M. (2021). Electrostatic modulation of hnRNPA1 low-complexity domain liquid- journal.pcbi.1004013 liquid phase separation and aggregation. Protein Sci. 30, 1408–1417. doi: 10.1002/ Sant’anna, R., Fernandez, M. R., Batlle, C., Navarro, S., De Groot, N. S., Serpell, pro.4108 L., et al. (2016). Characterization of amyloid cores in prion domains. Sci. Rep. 6, Uniprot, C. (2021). UniProt: the universal protein knowledgebase in 2021. 34274. doi: 10.1038/srep34274 Nucleic Acids Res. 49, D480–D489. doi: 10.1093/nar/gkaa1100 Santos, J., Pallares, I., Iglesias, V., and Ventura, S. (2021). Cryptic amyloidogenic Varadi, M., Anyango, S., Deshpande, M., Nair, S., Natassia, C., Yordanova, G., regions in intrinsically disordered proteins: Function and disease association. et al. (2022). AlphaFold protein structure database: massively expanding the Comput. Struct. Biotechnol. J. 19, 4192–4206. doi: 10.1016/j.csbj.2021.07.019 structural coverage of protein-sequence space with high-accuracy models. Serio, T. R., Cashikar, A. G., Moslehi, J. J., Kowal, A. S., and Lindquist, S. L. Nucleic Acids Res. 50, D439–D444. doi: 10.1093/nar/gkab1061 (1999). Yeast prion [psi +] and its determinant, Sup35p. Methods Enzymol. 309, Wang, J., Choi, J. M., Holehouse, A. S., Lee, H. O., Zhang, X., Jahnel, M., et al. 649–673. doi: 10.1016/S0076-6879(99)09043-6 (2018). A molecular grammar governing the driving forces for phase separation of Sherman, B. T., Hao, M., Qiu, J., Jiao, X., Baseler, M. W., Lane, H. C., et al. (2022). prion-like RNA binding proteins. Cell 174, 688–699 e616. doi: 10.1016/ DAVID: a web server for functional enrichment analysis and functional annotation of gene j.cell.2018.06.006 lists, (2021 update). Nucleic Acids Res. 50 (W1), W216–W221. doi: 10.1093/nar/gkac194 Wickner, R. B. (1994). [URE3] as an altered URE2 protein: evidence for a prion Sormanni, P., Aprile, F. A., and Vendruscolo, M. (2015). The CamSol method of analog in saccharomyces cerevisiae. Science 264, 566–569. doi: 10.1126/ rational design of protein mutants with enhanced solubility. J. Mol. Biol. 427, 478– science.7909170 490. doi: 10.1016/j.jmb.2014.09.026 Yuan, A. H., and Hochschild, A. (2017). A bacterial global regulator forms a Tetz, G., and Tetz, V. (2018). Prion-like domains in eukaryotic viruses. Sci. Rep. prion. Science 355, 198–201. doi: 10.1126/science.aai7776 8, 8931. doi: 10.1038/s41598-018-27256-w Zambrano, R., Conchillo-Sole, O., Iglesias, V., Illa, R., Rousseau, F., Theillet, F. X., Kalmar, L., Tompa, P., Han, K. H., Selenko, P., Dunker, A. K., Schymkowitz, J., et al. (2015). PrionW: a server to identify proteins containing et al. (2013). The alphabet of intrinsic disorder: I. act like a pro: On the abundance glutamine/asparagine rich prion-like domains and their amyloid cores. Nucleic and roles of proline residues in intrinsically disordered proteins. Intrinsically Acids Res. 43, W331–W337. doi: 10.1093/nar/gkv490 Disord. Proteins 1, e24360. doi: 10.4161/idp.24360 Zhang, Y., Man, V. H., Roland, C., and Sagui, C. (2016). Amyloid properties of Thompson, M. J., Sievers, S. A., Karanicolas, J., Ivanova, M. I., Baker, D., and asparagine and glutamine in prion-like proteins. ACS Chem. Neurosci. 7, 576–587. Eisenberg, D. (2006). The 3D profile method for identifying fibril-forming doi: 10.1021/acschemneuro.5b00337 Frontiers in Plant Science 12 frontiersin.org