Global Protein Stabilization Does Not Suffice to Prevent Amyloid Fibril

Jul 2, 2018 - Global Protein Stabilization Does Not Suffice to Prevent Amyloid Fibril ... Vall d'Hebron Institute of Oncology (VHIO), Vall d'Hebron Un...
0 downloads 0 Views 4MB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Global protein stabilization does not suffice to prevent amyloid fibril formation Patrizia Marinelli, Susanna Navarro, Manuel Bano Polo, Bertrand Morel, Ricardo Graña-Montes, Anna Sabé, Francesc Canals, Maria Rosario Fernandez, Francisco Conejero-Lara, and Salvador Ventura ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.8b00607 • Publication Date (Web): 02 Jul 2018 Downloaded from http://pubs.acs.org on July 3, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Graphical Abstract 39x30mm (600 x 600 DPI)

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

Global protein stabilization does not suffice to prevent amyloid fibril formation. Patrizia Marinelli1+, Susanna Navarro1+, Manuel Baño-Polo1+, Bertrand Morel2, Ricardo Graña-Montes1, Anna Sabe3, Francesc Canals3, Maria Rosario Fernandez1, Francisco Conejero-Lara2 and Salvador Ventura1** 1

Institut de Biotecnologia i de Biomedicina and Departament de Bioquímica i Biologia Molecular. Universitat Autònoma de Barcelona. E-08193 Bellaterra (Spain). 2 Departamento de Química Física e Instituto de Biotecnología, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain 3

Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron University Hospital, 08135 Barcelona, Spain

* *To whom correspondence should be addressed: Salvador Ventura, Institut de Biotecnologia i de Biomedicina, Campus Universitari de Bellaterra, 08193 Cerdanyola, Barcelona, Spain Tel.: (+34) 93 586 8956 ; E-mail: [email protected] +These authors contributed equally.

1 ACS Paragon Plus Environment

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

ABSTRACT Mutations or cellular conditions that destabilize the native protein conformation promote the population of partially unfolded conformations, that, in many cases, assemble into insoluble amyloid fibrils, a process associated with multiple human pathologies. Therefore, stabilization of protein structures is seen as an efficient way to prevent misfolding and subsequent aggregation. This has been suggested to be the underlying reason why proteins living in harsh environments, such as the extracellular space, have evolved disulfide bonds. The effect of protein disulfides on the thermodynamic and kinetics of folding has been extensively studied, but much less is known on its effect on aggregation reactions. Here, we designed a single point mutation that introduces a disulfide bond in the all-α FF domain, a protein that, despite being devoid of preformed β-sheets, forms β-sheet-rich amyloid fibrils. The novel and unique covalent bond in the FF domain dramatically increases its thermodynamic stability and folding speed. Nevertheless, these optimized properties cannot counteract the inherent aggregation propensity of the protein; thus indicating that a high global protein stabilization does not suffice to prevent amyloid formation, unless it contributes to hide from exposure the specific regions that nucleate the aggregation reaction.

2

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

INTRODUCTION Protein misfolding and aggregation is associated with an increasing number of human diseases, ranging from dementia to diabetes1. Despite the proteins behind these pathologies are not related in sequential or structural terms, in many cases, their aggregation leads to the formation of amyloid fibrils displaying a common cross-β motif2. The adoption of such macromolecular architecture is not restricted to disease-linked polypeptides, but rather constitutes a generic feature of proteins chains3,4. Thus, in addition to the native fold, an alternative, stable, aggregated state is accessible to proteins. For any given protein, both states compete in the cell5. Thus, in globular proteins, the adoption of a stable native state constitutes a major protective strategy against aggregation6. The onset of aggregation reactions usually requires partial unfolding7 that expose to solvent aggregation-prone regions that were previously shielded in the native structure of globular proteins8. This implies that aggregation is connected to protein thermodynamic and/or kinetic stability9 and explains why mutations, or cellular conditions, that destabilize the native structure and/or increase unfolding rates are associated to pathological phenotypes in several diseases10. Accordingly, it has been suggested that rational overstabilization of protein structures might be a convenient way to overcome protein aggregation11. On the other hand, in some cases aggregation rates have no correlation with changes in global stability exerted by mutations but rather with the changes in stability of specific conformational states that nucleate aggregation12,13. Therefore, avoidance of aggregation by protein stabilization may need to account for the effects on the conformational landscape. Natural disulfide bonds can stabilize proteins to a large extent and strongly reduce conformational fluctuations14, providing a defense against unfolding in harsh environments, such as the extracellular space. By crosslinking sequentially distant regions of the polypeptide chain, disulfide bonds decrease the entropy of the unfolded ensemble, making it less favorable compared with the folded conformation. In some cases, the stabilizing effect is so high that the protein readily unfolds when its disulfides are reduced15. Thus, rational design of new disulfide bonds is a suitable way to increase the stability of globular proteins16 and potentially decrease their aggregation propensity17. Here, we take advantage of the single Cys residue in position 57 of the yeast URN1 FF

3 ACS Paragon Plus Environment

Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

domain18,19 to build up a novel disulfide bond connecting two sequentially distant regions of the protein by introducing a single point mutation. FF domains are small protein-protein interaction modules of 50–70 residues and characterized by the presence of two conserved Phe residues at their N- and C-termini20. This fold consists of three α-helices arranged as an orthogonal bundle with a 310 helix in the loop connecting the second and the third helices18,21 (Figure 1A). FF domains are involved in RNA splicing, signal transduction and transcription processes22,23, they are present in a variety of eukaryotic nuclear transcription and splicing factors, and their sequences are well conserved from yeast to humans20. The folding and aggregation landscapes of several of these domains, including the URN1 FF domain, have been characterized in detail24,25,26,27. This domain is a perfect protein module to address the connection between protein stability and aggregation, since destabilization of the native structure, either by the solution conditions o by postraductional modifications leads to the formation of an intermediate that despite keeping most of the α-helical structure selfassembles into highly ordered amyloid fibrils28,24,29. We have shown previously that altering polypeptide chain connectivity by introducing novel disulfide bridges have a high impact in the folding and amyloid formation of an all-β-protein: the SH3 domain, for which aggregation is initiated by the self-assembly of the preformed β-sheets upon domain destabilization19. However, to the best of our knowledge, this issue has not been previously addressed for an all-α -protein, in which preformed β-sheets elements do not exist. We show here that the designed disulfide bond dramatically increases the global stability of the URN1 FF domain and its folding efficiency. However, in contrast to what we observed for the SH3 domain19, these optimized properties were unable to preclude the aggregation of the domain into amyloid fibrils; thus indicating that global protein stabilization does not suffice to prevent per se aggregation, unless it prevents the access of aggregation-prone structural elements to the solvent. RESULTS AND DISCUSSION Mutagenesis of yeast URN1 FF domain and disulfide bond formation The FoldX algorithm30 was used to design an URN1 FF domain variant in which the wild type cysteine residue (Cys57) is linked to a selected mutated residue by a disulfide bridge. Lys23 was chosen as an optimal candidate for mutation into a Cys 4

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 30

residue, considering their proximity (distance between the β-carbons = 3.2Å), face-toface orientation, and proper dihedral angles (Figure 1A and 1B). The FF mutant (FFK23C-SS) was produced by site-directed mutagenesis and expressed in E. coli. The designed covalent bond was spontaneously formed in the cytosol as confirmed by the absence of detectable carbamidomethylation with iodoacetamide (IAA) in the recombinant purified protein. The wild type domain (FF-WT) displays a MW of 7118.58 Da (expected 7118.89 Da), which upon derivatization with IAA increases to 7175.72 Da (expected 7175.87 Da). FF-K23C-SS exhibits a MW of 7091.17 Da (expected 7091.78 Da). In the presence of IAA, FF-K23C-SS kept a MW of 7091.77 (expected for an IAAform 7206.90 Da), indicative of the absence of free thiol groups (Table S1 and Figures S1 and S2).

Conformational properties of FF domains under physiological conditions The far-UV CD spectrum of FF-K23C-SS under native conditions corresponds, as in the case of FF-WT, to that of a canonical α-helical protein displaying, however, a slightly more intense peak at 210 nm. The reduced form of the mutant (FF-K23C-SH) also exhibits the typical α-helical pattern but displays lower ellipticity (Figure S3A). The analysis of the intrinsic fluorescence contributed by the two tryptophan (Trp) residues at positions 27 and 56 of the URN1 FF domains shows that the mutant FF-K23C-SS emission spectrum displays lesser intensity, with a maximum slightly blue-shifted compared with that of the FF-WT, suggesting deeper burial of Trp residues into the globular structure. Conversely, the reduced mutant FF-K23C-SH spectrum shows an increase of fluorescence intensity with a red-shift respect to the wild type suggesting a higher exposure of Trp residues and a certain opening of the globular structure (Figure S3B). This behaviour is confirmed by the bis-ANS binding ability of FF-K23C-SH, which shows a higher intensity in the fluorescence emission with a red-shift of its maximum compared with that in the presence of the wild type domain indicating an increased exposure of hydrophobic clusters. Bis-ANS emission spectrum in the presence of FF-K23C-SS indicates a binding similar to that of FF-WT, with a low exposure of hydrophobic regions (Figure S3C). The one-dimensional NMR (1H-NMR) spectrum of the three domains under native conditions (298K and pH 5.7) displays a wide signal 5 ACS Paragon Plus Environment

Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

dispersion of resonances with good peak sharpness, characteristic of folded molecules (Figure S3D). Thermal denaturation of URN1 FF domains Thermal stability of FF domains was analyzed at pH 5.7 by circular dichroism (CD), intrinsic fluorescence and differential scanning calorimetry (DSC) (Figure 1). The thermal denaturation curves were followed by far-UV CD at 222 nm (Figure 1C), and by intrinsic fluorescence emission at 360 nm (Figure 1D) showing a single cooperative transition for both FF-WT and FF-K23C-SH. FF-K23C-SS turned to be highly stable, not being completely denatured even at 368K. DSC scans of the three FF domains exhibited a single cooperative transition corresponding to a two-state unfolding process without a significant population of intermediate partially unfolded states, according to the ratio between the van’t Hoff and calorimetric enthalpies (Figure 1E). The thermal stability of the FF domain is drastically affected by the Lys23 mutation as revealed by the extraordinary increase of the FF-K23CSS melting temperature (Tm), ≈ 20K higher than the wild type domain, and the significant decrease of FF-K23C-SH Tm, ≈15K lower than the wild type (Table S2). When they could be measured, the transition unfolding temperatures obtained by the three independent methods were very similar (Table S2) indicating that the reduced FF mutant is less stable compared to the FF-WT, and that FF-K23C-SS is significantly more stable towards thermal denaturation than the wild-type domain. The DSC analysis shows that the difference in DH between FF-WT and FFK23C-SS is only ≈ 1 kcal/mol, and therefore that a large majority of the stabilizing effect can be attributed to entropic factors. This view is supported by the large differences in the heat capacity change of unfolding (DCp) (Table S2). The DCp between the folded and unfolded states of proteins is thought to arise mainly from the exposure of buried sidechains to solvent; thus the low DCp of FF-K23C-SS suggests that it posseses a compact unfolded state. The temperature-induced unfolding of FF domains was also monitored by 1HNMR (Figure S4). FF-WT displays native signal dispersion and peak sharpness until 328 K. The intensity of the signals gradually decays until 348K when the spectrum collapses and the resonances at low fields are hardly detectable, indicating the absence of a preferential folded conformation. In the case of FF-K23C-SS, the spectra remain

6

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

essentially unchanged until 343K and even at 348K, the spectrum presents good signal dispersion, although with decreased peak intensity. FF-K23C-SH spectrum displays good signal dispersion until 320K and a rapid loss of the signal is observed above this temperature. The analysis of the spectra at 338K demonstrates clearly the different thermal stability of the three FF domains in the present study (Figure S4D). Equilibrium unfolding of FF domains The resistance against chemical denaturation with guanidinium chloride (GuHCl) was also used to analyze differences in the stability of the three different FF domains. The chemical denaturation curves at equilibrium were followed at 298K and pH 5.7 monitoring the changes in molar ellipticity at 222 nm (Figure 1F) and the intrinsic fluorescence at 360 nm (Figure 1G) under increasing GuHCl concentrations. FF-WT and FF-K23C-SH displayed a single visible transition indicative of a cooperative unfolding process. The reaction was reversible. The main thermodynamic parameters of the unfolding reaction were calculated from the equilibrium curves assuming a two-state model (R > 0.99) (Table 1). The stability of FF-WT calculated from fluorescence and CD measurements was similar with ΔG(H20)U-F ≈ 5.5 kcal/mol and [GuHCl]50% of 2.8 M. In agreement with the thermal denaturation data, FF-K23C-SH resulted to be less stable than the wild type domain, exhibiting a ΔG(H20)U-F of ≈ 3.6 kcal/mol and [GuHCl]50% of 2.3 M. The value of [GuHCl]50% for FF-K23C-SS was too high (˃5 M) and a final baseline for the fully unfolded state could not be accurately measured, which precluded fitting and calculation of thermodynamic parameters. Therefore, the FF-K23C-SS, as well as FFWT, were studied analyzing the equilibrium denaturation with the stronger chaotropic agent guanidinium isothiocyanate (GITC) (Figure 1H). FF-K23C-SS displayed a ΔG(H20)U-F 3.9 kcal/mol higher than the FF-WT, with an increase in [GITC]50% of 1.58 M (Table 1), which confirms that the intramolecular disulfide notably stabilizes FF K23C-SS in front of chemical denaturation (Figure 1H and Table 1). Interestingly, the dithiol form of the protein showed a marked destabilization relative to the FF-WT suggesting that the mutated residue is not sterically or electrostatically acceptable in the reduced state. Folding and unfolding kinetics of FF domains

7 ACS Paragon Plus Environment

Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

The kinetics of folding and unfolding of the FF domains were determined by stopped-flow at pH 5.7 and 298 K under a wide range of urea concentrations. In all cases, the folding and unfolding traces by fluorescence fit well into single exponential functions. The chevron plots appear to be linear in the complete range of denaturant concentrations studied, indicating the lack of detectable intermediates, according to a two-state model (Figure 2). The rate constants for folding (kf) and unfolding (ku) and their dependence on the denaturant concentration (mf and mu) are shown in Table 2. The kinetic data of FF domains are in good agreement with the equilibrium stabilities reported above and with the respective urea unfolding curves (Figure S5 and Table 1). The FF K23C-SH folding rate appears similar to that of FF-WT, while its unfolded rate results highly affected, suggesting that the mutated Lys plays an important role in maintaining the native conformation even though is not likely involved in the folding nucleus formation as indicated by the calculated φF-U value of 0.15 (Figure 2A). Due to its high resistance to urea denaturation, FF-K23C-SS kinetic measurements were performed in the presence of guanidinium chloride (GuHCl) (Figure 2B). We could not mesure the refolding limb of the chevron plot for FF-WT in GuHCl, which precluded calculation of its kinetic constants. However, FF-K23C-SS shows a folding rate much faster than the kf calculated for the wild type using urea as denaturant. This acceleration of the folding reaction can be univocally attributed to the presence of the disulfide bridge, since the dithiol form exhibits a folding rate close to that of FF-WT, as discussed above. The disulfide bond acts also upon the unfolding rate, which is significantly slower than that of the wild type domain, indicating the disulfide bond contribution to the maintenance of the native conformation. FF aggregation under native conditions The FF domains were incubated at pH 5.7 and 310 K during one week to test their ability to self-assemble into macromolecular structures. The presence of amyloid aggregates was preliminary analyzed using the amyloid specific dye Th-T binding (Figure 3A). The analysis of FF-K23C-SH solution in presence of Th-T revealed a significant increase in the fluorescence emission, while both wild type and FF-K23C-SS solutions did not bind the amyloid diagnostic dye. In agreement with these results, Congo red (CR) binding was observed only for the FF-K23C-SH aggregates, with the characteristic red-

8

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

shift at ≈ 540 nm and increase in the absorbance maximum of the dye (Figure 3B). The changes in secondary structure were monitored by far-UV CD (Figure 3C). Both FF-WT and FF-K23C-SS proteins retained an α-helical native spectrum with the characteristic minima at 210 and 222 nm. In contrast the FF-K23C-SH domain, exhibited a β-sheet enriched conformation, with a single minimum at 217 nm. Then, the bis-ANS dye was used to monitor the exposure of hydrophobic clusters to the solvent in the putatively aggregated states, with a detectable increase of bis-ANS emission being observed only for the FF-K23C-SH domain (Figure 3D). The morphological analysis of the FF aggregates by transmission electron microscopy (TEM) confirmed the presence of amyloid-like fibrillar structure only for the FF-K23C-SH, with the FF-WT and FF-K23CSS solutions lacking aggregates (Figure 3E). Thus, as previously decribed24, FF-WT does not form amyloid-like aggregates under physiological conditions. The mutation of the Lys23 to a Cys promotes amyloid formation when it is in its reduced form, while its oxidation into a disulfide bond precludes this pro-aggregative effect. Conformational and stability analysis of FF domain under low pH Both FF-WT and FF-K23C-SS remain essentially folded in the 3-10 pH range, according to CD and intrinsic fluorescence measurements (Figure S6). We have shown that at pH 2.5 the URN1 FF domain forms a Molten Globule (MG) state that retains most of the α-helical secondary structure content characteristic of the protein native state24. The FF-K23C-SS was analyzed in this condition to study the effect of the disulfide introduction upon the MG state. The far-UV CD spectrum of FF-K23C-SS at pH 2.5 corresponds to that of an α-helical conformation, as in the native state but with slightly reduced ellipticity; a similar behavior can be observed in the case of the wild type domain (Figure 4A). The intrinsic fluorescence spectra of both acid induced FF domain states show an increase of signal emission and a slight red-shift compared with that of their native states according to a higher opening of their globular structures at low pH (Figure 4B). The FFWT spectrum displays, however, higher fluorescence emission than the FF-K23C-SS domain showing a stronger exposure of Trp residues in the wild type MG than in the conformation attained by the FF mutant. These data suggest that the covalent bond formation might regulate the compactness of the globular structure around Trp residues 9 ACS Paragon Plus Environment

Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

at low pH as well as at physiological conditions. As expected, both FF domains promote an increase in the bis-ANS fluorescence emission compared with the native proteins confirming the exposure of hydrophobic regions under acidic conditions (Figure 4C). Interestingly, the FF-K23C-SS spectrum shows a higher bis-ANS intensity emission than the wild type domain, which contrasts with the data obtained by intrinsic fluorescence, indicating that a different conformational rearrangement occurs in presence of the disulfide bond. As reported previously by our group, the URN1 FF thermodynamic stability is drastically affected at lower pHs24. To test the pH effect upon the FF mutant stability, the thermal unfolding of FF domains was analyzed at pH 2.5 by intrinsic fluorescence and by differential scanning calorimetry (DSC) (Figure 5). The thermal curves of the FF-WT and FF-K23C-SS were followed monitoring the emission changes at 360 nm. In both cases, a single cooperative transition was observed and the data could be fitted accurately to a two-state temperature-induced unfolding model (R> 0.99). The two-state unfolding process without intermediate partially unfolded states was confirmed by the DSC analysis. The FF-K23C-SS domain is significantly more stable toward thermal denaturation than the wild-type under low pH as reflected by its melting temperature (Tm), ≈ 20 K higher than the FF-WT Tm (Table S3). Indeed, the Tm of FF-K23C-SS at pH 2.5 resembles that of the wild-type at pH 5.7. The similarity in the enthalpies exhibited by the FF-K23C-SS and the wild type proteins suggest that the changes in thermal stability at this pH are due to differences in the entropic values of the respective unfolding reactions. Amyloid fibril formation by FF domains under acidic conditions In order to study the pH-effect upon FF domains amyloid formation, the aggregation kinetics of FF proteins were analyzed at pH 2.5 and 310 K monitoring the changes over time in the fluorescence of the amyloid staining dye Th-T (Figure 6A). We first confirmed that the disulfide bond in FF-K23C-SS remains intact in this condition using MS. Samples incubated for 24 h, 48 h and 72 h were analyzed with and without incubation with IAA, all displaying the same MW (~ 7091 Da), correspondent to a disulfide bonded form (Table S1 and Figure S1 and S2). The presence of the disulfide link in FF-K23C-SS upon 72 h of incubation was confirmed after its reduction with dithiothreitol, which rendered a species of 7093.35 Da (expected 7093.78 Da) and the

10

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

subsequent derivatization of its two free SH groups with IAA, which increased the size to 7206.23 Da (expected 7206.90 Da) (Table S1 and Figures S1 and S2). The FF-K23C-SS aggregation kinetics were faster (k = 0.445 ± 0.005 h-1) than that of the wild type (k = 0.212 ± 0.003 h-1), which might owe to the higher exposure of hydrophobic regions observed in the FF-K23C-SS soluble monomer under acidic conditions (Figure 4C). The FF aggregation reactions were also followed at different time points by transmission electron microscopy (TEM) (Figure 6B). Abundant early FFK23C-SS aggregates were visible after 60 min of incubation, whereas few wild type aggregates were observed, in agreement with the Th-T intensity in the respective kinetics. The early FF-K23C-SS aggregates seem to correspond to oligomeric assemblies with amyloid-like features. At 240 min, the wild type domain formed ordered fibrils while only short assemblies were observable for the FF-K23C-SS domain. Different morphologies were observed after 1000 min of incubation indicating that the disulfide introduction may interfere in the fibril structuration. Differences in fibrillar structure were also evident after 48 h of incubation at acidic pH (Figure S7). FF mature fibrils incubated for one week at pH 2.5 and 310K were analyzed by Th-T, CR and bis-ANS binding (Figure 7A,B,C). Both FF fibrils resulted positive to amyloid dyes displaying, however, different binding strength, indicating a probable distinct conformational rearrangement of the fibrils, as also suggested by the dissimilar fluorescence emission intensity shown in the presence of bis-ANS. The observation of FF mature fibrils by transmission electronic microscopy (TEM) confirmed the presence of two distinct morphological species (Figure 7D). The visualization of the FF-WT mesh showed long, unbranched and twisted fibrils with a diameter of ≈14 nm as previously reported for the FF wild type domain under acidic conditions24. On the other hand, unbranched straight linear fibrils with diameters of 22-30 nm were observed for the FF-K23C-SS sample. The evaluation of the FF fibrils secondary structure by Attenuated Total Reflectance Fourier Transform Infrared spectroscopy (ATR-FTIR) revealed similar amyloid structural components for both fibrils (Figure 7E). As shown in Table S4, deconvolution of the absorbance spectra in the amide I region shows an intense signal at ≈1615 cm−1 and ≈1618 cm−1 for wild type and mutant fibrils, respectively, indicating the presence of predominant intermolecular amyloid β-sheets31,32,33. Common bands around 1641 cm−1 associated to β-sheets are also observed for both fibrils. While, a small signal at ≈1658 cm−1 corresponding to α-helical structure is only detected for the FF-K23C-SS fibrils. These data suggest that despite the 11 ACS Paragon Plus Environment

Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

presence of a common β-sheet signature, the residual presence of α-helical components might modulate the resulting fibrils morphology. CONCLUSION Protein stability is considerably increased by naturally occurring disulfide crosslinks34,35,36,14. Remarkably, the presence of disulfide bonds in numerous precursor proteins (55%) involved in amyloidosis suggests that they might play an important role also in aggregation reactions17. In the present work, a novel disulfide bridge in the URN1 FF domain was designed in an attempt to study simultaneously the impact of chain crosslinking upon the stability and aggregation of an all α helical protein model. The Lys23 to Cys mutation has a differential impact on the FF domain stability in the reduced and oxidized states. The dithiol form resulted to be significantly less stable than the wild-type domain. The lower value of the unfolding ΔH in FF-K23C-SH, as calculated by DSC experiments, suggests that this destabilization may be caused by the loss or weakening of intramolecular and protein-solvent interactions in the native state, consistent with the observed opening of the globular structure and exposure of hydrophobic regions. In contrast, the disulfide in the FF mutant stabilized the domain against both thermal and chemical denaturation, with an increase in Tm of ≈ 34 K and a ΔΔGU-F at 298 K of ≈ 5 kcal/mol relative to the dithiol form. The slight decrease of the ΔH value compared with that of the wild type protein suggests that the large stabilization conferred by the covalent bond is due essentially to the entropy reduction of the unfolded state. Additionally, a 45% decrease in the ΔCp of FF-K23C-SS was observed, likely resulting from a more compact unfolded state with restricted hydration of side chains, as observed for other disulfide mutants37, and consistent with thermodynamic and kinetic data in this work. It is worth to mention that, to the best of our knowledge, the overstabilization attained in K23C-SS is the highest obtained in a small domain with a single point mutation38. The conformational entropy reduction in the unfolded state due to disulfide bond linkage is generally associated with an acceleration in the folding rate. Decrease in unfolding rates depends on whether the regions involved in the crosslink are or not nativelike structured in the transition state. In this work, an intramolecular disulfide in the FFK23C-SS leads to an extraordinary increase of the folding rate, more 45-fold faster than that of the wild type, approaching the folding speed limit39. The decrease in the unfolding

12

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

rate of FF-K23C-SS suggests that the opening of the connected regions may represent a rate-limiting step in the unfolding process of the FF fold, as confirmed by the elevated unfolding rate observed for the dithiol form. Disulfide bonds are present in 15 % of the human proteome, in 65 % of secreted proteins, and in more than 50 % of those implied in amyloidosis40,17. The effect of disulfide bonds upon amyloid fibril formation is still far from being completely understood. However, studies in proteins like antibody light chains converge to indicate that these covalent connections impact significantly the aggregation-propensity of disease-linked polypeptides41. Our data show that the high stability and compactness of the FF-K23C-SS structure at pH 5.7 prevents the protein from aggregation. Indeed, in these conditions the stability of the natural domain suffices to protect it from side aggregation reactions. By contrast, the lower stability of the dithiol form leads to a very fast amyloid fibrillation, in agreement with the increased exposure of hydrophobic regions exhibited under physiological conditions. The observation that a destabilizing point mutation allows amyloid formation under, otherwise, protective native conditions, is consistent with the pro-aggregational effect reported for many destabilizing diseaselinked mutations1. In order to understand the impact of disulfide formation on the FF domain amyloid propensity in a destabilizing environment, aggregation was analyzed at pH 2.5. Surprisingly, despite its Tm at low pH was equivalent to that of the wild type domain under physiological conditions, the FF-K23C-SS variant self-assembled into amyloid fibrils faster than the FF-WT domain at pH 2.5, although this latter domain is much less stable in these conditions. Indeed, in spite of its higher stability, the conformational characteristics of FF-K23C-SS at this pH are also compatible with a molten globule-like state42. The disulfide bond imposes a topological restraint in the helix 3 and in the loop between helix1-2 leaving probably the helix 1 sterically free to initiate the amyloid formation. This region of the protein exhibits the highest aggregation propensity at pH 2.5 and indeed the dissected helix suffices to form amyloid fibrils24. Therefore, this work illustrates how the global stabilization of a protein does not suffice to prevent the amyloid formation, unless it contributes to hide the regions that nucleate the aggregation process from exposure and intermolecular interaction. Our observations are consistent with a recent study on the disease-linked b-2 microglobulin protein, which reports that the exposure of aggregation-prone residues is a major determinant of the aggregation propensity, even in the context of highly stable proteins43. Our data provide strong support 13 ACS Paragon Plus Environment

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

to the so-called amyloid-stretch hypothesis, which postulates that the aggregation propensity of a protein maps to specific short sequence stretches, rather than being uniformly distributed along the sequence. The results also illustrate how the structural/topological characteristics of the aggregation-prone state determine the kinetics, structure and morphology of the final amyloid assembly and how its stabilization might favour the accumulation of potentially toxic oligomers. Overall, the present work shows how, by itself, a thermodynamically overstabilized conformation does not prevent protein aggregation and adds to recent evidences indicating that aggregation might occur from native states in the absence of destabilization44, that structural perturbations of the folded state might account for aggregation propensities45 or that the unbinding of a ligand suffices to promote protein deposition46. Thus, the landscape of protein conformations that might access aggregationcompetent states turns to be more complex than it is usually assumed. METHODS Design of disulfide bridge The yeast URN1 FF-WT protein consists of 59 amino acidic residues and presents only a cysteine residue on position 57. To design the disulfide bond, FoldX algorithm (version 3b6) was used to screen the best position for the mutation of the second cysteine30. The disulfide bond formation was modeled using the Disulfide by Design 2.0 (DbD2) program, a web-based tool that predicts potential disulfide links analyzing the Bfactor (temperature factor) of protein regions involved in predicted disulfide bonds47. The relative difference in stability between the wild-type and mutant structure was calculated employing the FoldX BuildModel command by rotating the same residues in the wildtype and mutant structure according to the FoldX rotamer database30. The URN1 FF-WT Protein Data Bank (PDB) 2JUC was used as input. Mass Spectrometry Analysis MALDI-TOF analysis were performed on an Autoflex Speed mass spectrometer (Bruker, Bremen), using sinapinic acid as matrix. A mixture of protein standards of known Mw was used for internal calibration. To determine the number of free cysteine residues, samples before and after carbamidomethylation with iodoacetamide were analyzed. Alkylation was accomplished by addition of 75 mM iodoacetamide in 50 mM

14

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

ammonium bicarbonate pH=8 to 5µL of a 90µM solution of the protein in sodium acetate, followed by incubation at room temperature for 90 min. The samples were then purified using ZipTip C4 micro reverse-phase columns (Millipore) before MS analysis. A sample of K23C 72 was also analyzed by MALDI-TOF MS after reduction of the disulfide bridge with DTT, and then after further alkylation with iodoacetamide. Reduction was performed in 6M Urea, 50 mM ammonium bicarbonate, 50 mM DTT, for 1h at r.t. Sample was then purified by ZipTip C4, and carbamidomethylated by addition of 75 mM iodoacetamide as above. Methods for cloning, mutagenesis and expression; sample preparation; thermal denaturation; NMR spectroscopy; chemical and pH denaturation; folding and unfolding kinetics; transmission electron microscopy, Th-T and congo red binding and Circular Dichroism (CD) and bis-ANS binding, are described in detail in the Supporting Information. ASSOCIATED CONTENT Supporting information available: Mass spectrometry analysis (Table S1). Thermodynamic parameters obtained from thermal denaturation at pH 5.7 (Table S2); thermodynamic parameters obtained from thermal unfolding at pH 2.5 (Table S3); secondary structure bands in the deconvoluted absorbance FTIR spectra of FF fibrils formed at pH 2.5 (Table S4). Mass spectrometry analysis (Figures S1 and S2). Conformational analysis of FF domains (Figure S3); 1H-NMR spectra at different temperatures (Figure S4). Equilibrium unfolding curves with urea (Figure S5). Depencence of the Far-UV CD and intrinsic fluorescence signals on the solution pH (Figure S6). Morphological properties of FF-WT and FF-K23C-SS aggregates after incubation for 48 h at pH 2.5 (Figure S7). Supplementary Methods and Supplementary References. AUTHOR INFORMATION Corresponding Author: E-mail: [email protected] Salvador Ventura: orcid.org/0000-0002-9652-6351 The authors declare that they have no competing financial interests. ACKNOWLEDGMENTS We thank E. Gil for preliminary studies. 15 ACS Paragon Plus Environment

Page 17 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

FUNDING SOURCES This work was funded by the Spanish Ministry of Economy and Competitiveness BIO2016-783-78310-R to S.V and by ICREA, ICREA-Academia 2015 to S.V.

16

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 30

REFERENCES (1) Chiti, F., and Dobson, C. M. (2017) Protein Misfolding, Amyloid Formation, and Human Disease: A Summary of Progress Over the Last Decade. Annu. Rev. Biochem. 86, 27–68. (2) Fernàndez-Busquets, X., de Groot, N. S., Fernandez, D., and Ventura, S. (2008) Recent structural and computational insights into conformational diseases. Curr. Med. Chem. 15, 1336–1349. (3) Fändrich, M., Fletcher, M. A., and Dobson, C. M. (2001) Amyloid fibrils from muscle myoglobin. Nature 410, 165–166. (4) Guijarro, J. I., Sunde, M., Jones, J. A., Campbell, I. D., and Dobson, C. M. (1998) Amyloid fibril formation by an SH3 domain. Proc. Natl. Acad. Sci. U. S. A. 95, 4224– 4228. (5) Jahn, T. R., and Radford, S. E. (2008) Folding versus aggregation: Polypeptide conformations on competing pathways. Arch. Biochem. Biophys. 469, 100–117. (6) Monsellier, E., and Chiti, F. (2007) Prevention of amyloid-like aggregation as a driving force of protein evolution. EMBO Rep. 8, 737–742. (7) Chiti, F., and Dobson, C. M. (2009) Amyloid formation by globular proteins under native conditions. Nat. Chem. Biol. 5, 15–22. (8) Rousseau, F., Serrano, L., and Schymkowitz, J. W. H. (2006) How Evolutionary Pressure Against Protein Aggregation Shaped Chaperone Specificity. J. Mol. Biol. 355, 1037–1047. (9) Castillo, V., Espargaró, A., Gordo, V., Vendrell, J., and Ventura, S. (2010) Deciphering the role of the thermodynamic and kinetic stabilities of SH3 domains on their aggregation inside bacteria. Proteomics 10, 4172–4185. (10) Booth, D. R., Sunde, M., Bellotti, V., Robinson, C. V., Hutchinson, W. L., Fraser, P. E., Hawkins, P. N., Dobson, C. M., Radford, S. E., Blake, C. C. F., and Pepys, M. B. (1997) Instability, unfolding and aggregation of human lysozyme variants underlying amyloid fibrillogenesis. Nature 385, 787–793. (11) Espargaró, A., Castillo, V., de Groot, N. S., and Ventura, S. (2008) The in Vivo and in Vitro Aggregation Properties of Globular Proteins Correlate With Their Conformational Stability: The SH3 Case. J. Mol. Biol. 378, 1116–1131. (12) Varela, L., Morel, B., Azuaga, A. I., and Conejero-Lara, F. (2009) A single mutation in an SH3 domain increases amyloid aggregation by accelerating nucleation, but not by destabilizing thermodynamically the native state. FEBS Lett. 583, 801–806. (13) Ruzafa, D., Varela, L., Azuaga, A. I., Conejero-Lara, F., and Morel, B. (2014) Mapping the structure of amyloid nucleation precursors by protein engineering kinetic analysis. Phys. Chem. Chem. Phys. 16, 2989–3000. (14) Arolas, J. L., D’Silva, L., Popowicz, G. M., Aviles, F. X., Holak, T. A., Ventura, S., Akasaka, K., Tachibana, H., Segawa, S., Kobayashi, Y., and al., et. (2005) NMR structural characterization and computational predictions of the major intermediate in oxidative folding of leech carboxypeptidase inhibitor. Structure 13, 1193–1202. (15) Creighton, T. E., and Goldenberg, D. P. (1984) Kinetic role of a meta-stable nativelike two-disulphide species in the folding transition of bovine pancreatic trypsin 17 ACS Paragon Plus Environment

Page 19 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

inhibitor. J. Mol. Biol. 179, 497–526. (16) Takagi, H., Takahashi, T., Momose, H., Inouye, M., Maeda, Y., Matsuzawa, H., and Ohta, T. (1990) Enhancement of the thermostability of subtilisin E by introduction of a disulfide bond engineered on the basis of structural comparison with a thermophilic serine protease. J. Biol. Chem. 265, 6874–6878. (17) Mossuto, M. F., Bolognesi, B., Guixer, B., Dhulesia, A., Agostini, F., Kumita, J. R., Tartaglia, G. G., Dumoulin, M., Dobson, C. M., and Salvatella, X. (2011) Disulfide Bonds Reduce the Toxicity of the Amyloid Fibrils Formed by an Extracellular Protein. Angew. Chemie Int. Ed. 50, 7048–7051. (18) Bonet, R., Ramirez-Espain, X., and Macias, M. J. (2008) Solution structure of the yeast URN1 splicing factor FF domain: Comparative analysis of charge distributions in FF domain structures-FFs and SURPs, two domains with a similar fold. Proteins 73, 1001–1009. (19) Graña-Montes, R., de Groot, N. S., Castillo, V., Sancho, J., Velazquez-Campoy, A., and Ventura, S. (2012) Contribution of Disulfide Bonds to Stability, Folding, and Amyloid Fibril Formation: The PI3-SH3 Domain Case. Antioxid. Redox Signal. 16, 1– 15. (20) Bedford, M. T., and Leder, P. (1999) The FF domain: a novel motif that often accompanies WW domains. Trends Biochem. Sci. 24, 264–265. (21) Allen, M., Friedler, A., Schon, O., and Bycroft, M. (2002) The structure of an FF domain from human HYPA/FBP11. J. Mol. Biol. 323, 411–416. (22) Jiang, W., Sordella, R., Chen, G.-C., Hakre, S., Roy, A. L., and Settleman, J. (2005) An FF Domain-Dependent Protein Interaction Mediates a Signaling Pathway for Growth Factor-Induced Gene Expression. Mol. Cell 17, 23–35. (23) Smith, M. J., Kulkarni, S., and Pawson, T. (2004) FF Domains of CA150 Bind Transcription and Splicing Factors through Multiple Weak Interactions. Mol. Cell. Biol. 24, 9274–9285. (24) Castillo, V., Chiti, F., and Ventura, S. (2013) The N-terminal helix controls the transition between the soluble and amyloid states of an FF domain. PLoS One 8, e58297. (25) Jemth, P., Day, R., Gianni, S., Khan, F., Allen, M., Daggett, V., and Fersht, A. R. (2005) The Structure of the Major Transition State for Folding of an FF Domain from Experiment and Simulation. J. Mol. Biol. 350, 363–378. (26) Korzhnev, D. M., Religa, T. L., and Kay, L. E. (2012) Transiently populated intermediate functions as a branching point of the FF domain folding pathway. Proc. Natl. Acad. Sci. 109, 17777–17782. (27) Korzhnev, D. M., Religa, T. L., Banachewicz, W., Fersht, A. R., and Kay, L. E. (2010) A Transient and Low-Populated Protein-Folding Intermediate at Atomic Resolution. Science (80-. ). 329, 1312–1316. (28) Marinelli, P., Castillo, V., and Ventura, S. (2013) Trifluoroethanol modulates amyloid formation by the all α-helical URN1 FF domain. Int. J. Mol. Sci. 14, 17830– 17844. (29) Marinelli, P., Navarro, S., Graña-Montes, R., Bañó-Polo, M., Fernández, M. R., Papaleo, E., and Ventura, S. (2018) A single cysteine post-translational oxidation

18

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 30

suffices to compromise globular proteins kinetic stability and promote amyloid formation. Redox Biol. 14, 566–575. (30) Schymkowitz, J., Borg, J., Stricher, F., Nys, R., Rousseau, F., and Serrano, L. (2005) The FoldX web server: an online force field. Nucleic Acids Res. 33, W382– W388. (31) Sarroukh, R., Goormaghtigh, E., Ruysschaert, J.-M., and Raussens, V. (2013) ATR-FTIR: A “rejuvenated” tool to investigate amyloid proteins. Biochim. Biophys. Acta - Biomembr. 1828, 2328–2338. (32) Zurdo, J., Guijarro, J. I., and Dobson, C. M. (2001) Preparation and characterization of purified amyloid fibrils. J. Am. Chem. Soc. 123, 8141–8142. (33) Villegas, V., Zurdo, J., Filimonov, V. V, Avilés, F. X., Dobson, C. M., and Serrano, L. (2000) Protein engineering as a strategy to avoid formation of amyloid fibrils. Protein Sci. 9, 1700–1708. (34) Matsumura, M., and Matthews, B. W. (1991) Stabilization of functional proteins by introduction of multiple disulfide bonds. Methods Enzymol. 202, 336–356. (35) Betz, S. F. (1993) Disulfide bonds and the stability of globular proteins. Protein Sci. 2, 1551–1558. (36) Darby, N., and Creighton, T. E. (1995) Disulfide Bonds in Protein Folding and Stability, in Protein Stability and Folding, pp 219–252. (37) Betz, S. F., and Pielak, G. J. (1992) Introduction of a disulfide bond into cytochrome c stabilizes a compact denatured state. Biochemistry 31, 12337–12344. (38) Dehouck, Y., Grosfils, A., Folch, B., Gilis, D., Bogaerts, P., and Rooman, M. (2009) Fast and accurate predictions of protein stability changes upon mutations using statistical potentials and neural networks: PoPMuSiC-2.0. Bioinformatics 25, 2537– 2543. (39) Eaton, W. A., Muñoz, V., Hagen, S. J., Jas, G. S., Lapidus, L. J., Henry, E. R., and Hofrichter, J. (2000) Fast Kinetics and Mechanisms in Protein Folding. Annu. Rev. Biophys. Biomol. Struct. 29, 327–359. (40) Smith, D. P., and Radford, S. E. (2001) Role of the single disulphide bond of beta(2)-microglobulin in amyloidosis in vitro. Protein Sci. 10, 1775–1784. (41) Morgan, G. J., Usher, G. A., and Kelly, J. W. (2017) Incomplete Refolding of Antibody Light Chains to Non-Native, Protease-Sensitive Conformations Leads to Aggregation: A Mechanism of Amyloidogenesis in Patients? Biochemistry 56, 6597– 6614. (42) Baldwin, R. L., and Rose, G. D. (2013) Molten globules, entropy-driven conformational change and protein folding. Curr. Opin. Struct. Biol. 23, 4–10. (43) Camilloni, C., Sala, B. M., Sormanni, P., Porcari, R., Corazza, A., De Rosa, M., Zanini, S., Barbiroli, A., Esposito, G., Bolognesi, M., Bellotti, V., Vendruscolo, M., and Ricagno, S. (2016) Rational design of mutations that change the aggregation rate of a protein while maintaining its native structure and stability. Sci. Rep. 6, 25559–25570. (44) Soldi, G., Bemporad, F., Torrassa, S., Relini, A., Ramazzotti, M., Taddei, N., and Chiti, F. (2005) Amyloid Formation of a Protein in the Absence of Initial Unfolding and Destabilization of the Native State. Biophys. J. 89, 4234–4244. 19 ACS Paragon Plus Environment

Page 21 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

(45) Del Poggetto, E., Bemporad, F., Tatini, F., and Chiti, F. (2015) Mutations of Profilin-1 Associated with Amyotrophic Lateral Sclerosis Promote Aggregation Due to Structural Changes of Its Native State. ACS Chem. Biol. 10, 2553–2563. (46) Masino, L., Nicastro, G., Calder, L., Vendruscolo, M., and Pastore, A. (2011) Functional interactions as a survival strategy against abnormal aggregation. FASEB J. 25, 45–54. (47) Craig, D. B., and Dombkowski, A. A. (2013) Disulfide by Design 2.0: a web-based tool for disulfide engineering in proteins. BMC Bioinformatics 14, 346.

20

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 30

TABLES Table 1. Thermodynamic parameters of FF domains equilibrium unfolding b

DG(H2O)U-F (Kcal/mol)

CD FF-WTd FF-K23C-SHd FF-K23C-SSd FF-WTe FF-K23C-SSe FF-WTf FF-K23C-SHf

5.80 ± 0.42 3.84 ±0.40 -

c

-RTm F-U (Kcal/mol·M)

a

Intrinsic Fluorescence 5.21 ± 0.21 3.37 ± 0.49 4.56 ± 1.08 8.46 ± 0.32 3.78 ± 0.49

CD 1.19 ± 0.11 0.93 ± 0.09 -

Intrinsic Fluorescence 1.10 ± 0.02 1.47 ± 0.19 2.97 ± 0.55 1.94 ± 0.29 0.57 ± 0.04

[Denaturant]50% (M)

CD 2.88 ± 0.47 2.43 ± 0.48 >5 -

Intrinsic Fluorescence 2.80 ± 0.22 2.29 ± 0.63 >5 0.91 ± 0.30 2.49 ± 0.50 >6 5.40 ± 0.71

a

Gibbs energy of unfolding at [denaturing agent]=0 m value, dependence of free energy of unfolding with denaturating agent c The denaturating agent concentration required to unfold 50% of the protein molecules d Parameters obtained by the equilibrium unfolding with Guanidinium hydrochloride e Parameters obtained by the equilibrium unfolding with Guanidinium thiocyanate f Parameters obtained by the equilibrium unfolding with Urea b

Table 2. Folding kinetic parameters for FF domains kf (s-1)

ku (s-1)

FF-WT (Urea) FF-K23C-SH (Urea)

2789 ± 189

0.32 ± 0.12

1734 ± 105

5.89 ± 0.17

FF-K23C-SS (GuHCl)

131940 ± 3782 (19315e)

0.06 ± 0.03 (0.08f)

mf (Kcal/mol·M) 0.94 ± 0.02 0.97 ± 0.02 1.62 ± 0.01

mu (Kcal/mol·M) 0.26 ± 0.04

a -RTm F-U (Kcal/mol·M) 0.71 ± 0.06

0.11 ± 0.03 1.11 ± 0.09

b

[ ]50% (M) 7.56

c ΔG U-F (Kcal/mol) 5.34

d ΔΔG U-F (Kcal/mol) ̶

g ΔG U-F (Kcal/mol) 5.21 ± 0.21h

0.64 ± 0.05

5.26

3.37

-1.97

3.78 ± 0.49i

1.62 ± 0.10

5.35

8.65

3.31

8.46 ± 0.32j

a

Dependence of the Gibbs energy of unfolding with urea/GuHCl

b

The urea/ GuHCl concentration required to unfold 50% of the protein molecules

c

Gibbs energy of unfolding with urea/GuHCl determined from the kinetic parameters

d

The difference in free energy between the mutant protein and WT protein

e

Calculated at 4M of GuHCl

f

Calculated at 6.5M of GuHCl

g

Gibbs energy of unfolding with hGuHCl, iUrea or jGITC at equilibrium

21 ACS Paragon Plus Environment

Page 23 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

FIGURE LEGENDS Figure 1. Structure and stability of FF-WT and FF-K23C-SS domains. (A) Ribbon representation of wild type FF domain (PDB code: 2JUC). Lys23 (magenta) and Cys57 (orange) β-carbons distance are shown. (B) Structural model of FF-K23C-SS mutant showing disulfide bond formed by mutated Cys23 (red) and wild type Cys57 (orange). The figure was prepared with PyMOL (www.pymol.org). Thermal stabilities were analyzed by (C) far-UV CD signal change at 222 nm, (D) intrinsic fluorescence emission at 360 nm and (E) differential scanning calorimetry. Equilibrium unfolding curves with guanidinium hydrochloride (GuHCl) were analyzed at 298 K by (F) far-UV CD at 222 nm and by (G) tryptophan intrinsic fluorescence at 360 nm. (H) FF-WT and FF-K23C-SS equilibrium unfolding curve with guanidinium isothiocyanate (GITC) monitored by intrinsic fluorescence. In (F), (G) and (H), empty and solid symbols correspond to unfolding and refolding experiments, respectively.

Figure 2. Folding and unfolding kinetics of FF domains. (A) Dependence of the folding and unfolding rate constants (k) on urea concentration of FF-WT (green) and FF-K23C-SH (red). (B) Dependence of the folding and unfolding rate constants (k) on guanidinium chloride concentration of FF-WT and FF-K23C-SS. Figure 3. Morphological, structural and tinctorial properties of FF aggregates at pH 5.7. FF aggregates were analyzed by (A) Th-T fluorescence, (B) Congo red binding (C) far-UV CD spectroscopy, (D) binding to Bis ANS dye, and (E) TEM microscopy. The scale bar represents 200 nm. Figure 4. Conformational analysis of FF domains at pH 2.5. FF protein solutions analyzed at pH 2.5 and at pH 5.7 were monitored at 298 K by (A) far-UV CD signal, (B) tryptophan intrinsic fluorescence and (C) bis-ANS fluorescence. Figure 5. Thermal unfolding of FF domains at pH 2.5. Thermal stabilities under acidic conditions were analyzed by (A) intrinsic fluorescence emission at 360 nm and (B) differential scanning calorimetry.

Figure 6. Aggregation kinetics of FF domains under acidic conditions. Protein solutions were incubated with agitation at pH 2.5 and 310 K. Aggregation kinetics were followed by (A) Th-T fluorescence. (B) Transmission electron microscopy images of negatively stained aggregates at different time points are shown. Bar scale represents 200 nm. Figure 7. Tinctorial, morphological and structural properties of mature FF fibrils formed at pH 2.5. FF protein solutions were prepared at 100 µM at pH 2.5 and incubated for one week

at 310 K under agitation. The fibrils were analyzed by (A) Th-T fluorescence, (B) Congo red binding and (C) binding to Bis-ANS. (D) Negatively stained URN1 FF fibrils were visualized by TEM microscopy at different magnification. Scale bar represents 500 (upper panels) and 200 nm (lower panels). (E) URN1 FF fibrils were characterized by ATR-FTIR showing the absorbance spectra of the amide I region (colored thick line) and the component bands (continuos thin lines). The sum of individual spectral components after Fourier self-deconvolution closely matches the experimental data.

22

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 30

Figure 1

23 ACS Paragon Plus Environment

Page 25 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Figure 2

24

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 30

Figure 3

25 ACS Paragon Plus Environment

Page 27 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Figure 4

26

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 30

Figure 5

27 ACS Paragon Plus Environment

Page 29 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Chemical Biology

Figure 6

28

ACS Paragon Plus Environment

ACS Chemical Biology 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 30

Figure 7

29 ACS Paragon Plus Environment