Amyloidogenic Properties of Short α-l-Glutamic Acid Oligomers

Publication Date (Web): September 11, 2015. Copyright © 2015 American .... Shuo Wang , Youguo Zhang , Qiang Li , Rongqin Sun , Lin Ma , Liangchun Li...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/Langmuir

Amyloidogenic Properties of Short α‑L‑Glutamic Acid Oligomers Agnieszka Hernik, Wojciech Puławski, Bartłomiej Fedorczyk, Dagmara Tymecka, Aleksandra Misicka, Sławomir Filipek, and Wojciech Dzwolak* Department of Chemistry, Biological and Chemical Research Centre, University of Warsaw, 00-927 Warsaw, Poland S Supporting Information *

ABSTRACT: Poly-L-glutamic acid (PLGA) forms amyloid-like β2fibrils with the main spectral component of vibrational amide I′ band unusually shifted below 1600 cm−1. This distinct infrared feature has been attributed to the presence of bifurcated hydrogen bonds coupling CO and N−D (N−H) groups of the main chains to glutamate side chains. Here, we investigate how decreasing the chain length of PLGA affects its capacity to form β2-fibrils. A series of acidified aqueous solutions of synthetic (L-Glu)n peptides (n ≈ 200, 10, 6, 5, 4, and 3) were incubated at high temperature. We observed that n = 4 is the critical chain length for which formation of aggregates with the β2-like infrared features is still observed under such conditions. Interestingly, according to atomic force microscopy (AFM), the self-assembly of (L-Glu)n chains varying vastly in length produces fibrils with rather uniform diameters of approximately 4−6 nm. Kinetic experiments on (L-Glu)5 and (L-Glu)200 peptides indicate that the fibrillation is significantly accelerated not only in the presence of homologous seeds but also upon cross-seeding, suggesting thereby a common selfassembly theme for (L-Glu)n chains of various lengths. Our results are discussed in the context of mechanisms of amyloidogenic fibrillation of homopolypeptides.



pH induces a rapid coil-to-helix transition.13−16 With increasing temperature, helical PLGA tends to convert into β-sheet-rich aggregates. Since the α-helical fold is plausible for chains above certain critical length, only disordered and β-sheet conformations are accessible to short oligomers of α-L-glutamic acid.17 In fact, for a homopolypeptide, the chain length is a key variable to be taken into account while studying its propensities to adopt various secondary structures (e.g., ref 18). Chain-length dependence of structural dynamics of homopolypeptides is not only important from a theoretical point of view,19,20 but is also closely linked to the molecular basis of profoundly pathogenic amyloidogenesis in the case of Huntington’s disease.21 Itoh and colleagues were the first to describe β-sheet aggregates of PLGA (termed β2) with an unusual infrared feature: the amide I′ band characteristically red-shifted below 1600 cm−1.22 We proposed in our earlier works that this exotic optical trait can be explained by the presence of networks of three-centered hydrogen bonds coupling side-chain carboxyl and main chain −NH (−ND) groups as hydrogen donors to main-chain > CO groups as bifurcating acceptors.23,24 As such a structural arrangement requires perfect spatial packing of Glu side chains within the amyloid β2-fibril, it can be easily perturbed or prevented by latent defects in PLGA (either of

INTRODUCTION Self-assembly of amyloid fibrils, the insoluble linear β-sheet-rich aggregates of misfolded proteins, has been linked to dozens of degenerative maladies including Alzheimer’s and Parkinson’s diseases.1,2 Many proteins that are not amyloidogenic in vivo can be converted into amyloid-like fibrils in vitro under denaturating conditions. Fändrich and Dobson demonstrated that various poly-α-amino acids including poly-L-lysine and poly-L-glutamic acid (PLGA) form such fibrils, as well.3 Recent studies have proven that fibrils with amyloid-like characteristics may be obtained from copolymer poly-α-amino acids with randomized amino acid sequences, for example of Glu and Ala,4 or of Glu, Lys, Ala, and Tyr,5 or even from non-α-poly amino acids, e.g. ε-poly-L-lysine.6 These findings strongly support the idea that the ability to form amyloid fibrils is not restricted to a handful of disease-associated proteins but is rather a generic polymeric property of polypeptides primarily driven by main chain interactions.3 This also explains the main rationale of employing homopolypeptides such as PLGA as model amyloidogenic precursors. Conformational transitions in PLGA have been studied for decadeslong before the emergence of the amyloid context.7−11 Depending on pH and temperature, PLGA adopts different secondary structures in aqueous solutions. Because of repulsive culombic interactions between charged Glu side chains, only disordered conformations of PLGA are observed at pH above 5 and in the absence of multivalent cations.10−12 For sufficiently long PLGA chains, protonation of side chains at low © 2015 American Chemical Society

Received: August 5, 2015 Revised: September 4, 2015 Published: September 11, 2015 10500

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir

Static and Time-Lapse FT-IR Spectra. For acqusition of static FTIR spectra (Figure 2), 256 interferograms of nominal resolution of 2 cm−1 were coadded. The spectra were acquired at 25 °C using a CaF2 transmission cell and 0.05 mm Teflon spacer on Nicolet iS50 FT-IR spectrometer (Thermo, USA) equipped with a DTGS detector. During the measurements, the spectrometer’s sample chamber was continuously purged with dry air. Time-lapse FT-IR spectra (Figures 1 and 5) were collected in a similar way; however, the number of interferograms coadded for a single spectrum was reduced to 16. During measurements, the temperature in the cell (40 °C) was controlled through a dedicated Peltier system. From each sample’s spectrum the corresponding buffer and water vapor spectra were subtracted. Baseline correction was performed with GRAMS software (Thermo). All further experimental details were the same as specified earlier.26 AFM. Collected samples of aggregates were initially diluted 100 times with acidified water (pH 4). A small droplet (8 μL) of fibril suspension was swiftly deposited onto freshly cleaved mica and left to dry for 24 h. AFM tapping-mode measurements were carried out using a Nanoscope III atomic force microscope (Veeco, USA) and TAP300Al sensors, res. frequency 300 kHz (BudgetSensors, Bulgaria).

covalent nature, or as entrapped bodies) leading to thermodynamically frustrated fibrils with the conventional infrared amide I′ band features of antiparallel β-sheet (socalled β1-fibrils25,26). There are several fascinating aspects of the molecular architecture of β2-fibrils (e.g., their chiral superstructures with the handedness determined by molecular chirality of the main chain24), but despite efforts of several groups (including those of Keiderling, Kubelka, and Bouř23−28) the self-assembly process and the β2-structure itself are not presently understood in sufficient detail. So far, it has not been determined how the capacity of (L-Glu)n peptides to adopt β2structure changes with shortening the main chain length. This uncertainty precludes design of realistic molecular dynamics of the β2-self-assembly and hampers the search for adequate models for high-resolution structural studies. The aim of this work is to address these problems by investigating the chainlength dependence of PLGA’s β2-type amyloidogenic selfassembly, and in particular: establishing the critical size of aggregation-prone (L-Glu)n fragment.





RESULTS AND DISCUSSION β-Fibrillar aggregates of PLGA are known to enhance, in an amyloid-like manner, fluorescence of Thioflavin T,23,29 which may be conveniently used to monitor the progress of formation of fibrils by (Glu)n peptides.4 Such an approach has two obvious drawbacks: it does not discriminate between β1 and β2 fibrils and may lead to misinterpretation of the final fluorescence plateau due to free fluorophore depletion as a false end of the fibril elongation phase. Thus, we chose timelapse FT-IR spectroscopy to monitor the α → β2 process instead. Figure 1A presents stacked FT-IR spectra (amide I′ vibrational band) of (Glu)200 undergoing gradual α → β2 transition at pD 4.1 and 40 °C over the period of 24 h. The spectra show a gradual decrease of the 1641 cm−1 component assigned to α-helical structures coinciding with an increase of the sharp peak at 1596 cm−1 assigned to the β2 structure. These changes are accompanied by the 1705 cm−1 band, assigned to side-chain −COOD stretches, blue-shifting and splitting into two sharper peaks at 1730 and 1720 cm−1. We have attributed this characteristic spectral evolution accompanying the transition of soluble α-helical PLGA into insoluble fibrils to the formation of networks of bifurcated hydrogen bonds coupling CO and N−D groups of the main chains to glutamate side chains.23,24 The self-assembly of β2 fibrils is expected to depend strongly on the initial concentration of soluble peptide. This has been confirmed in the following experiment on concentration-dependent kinetics of the α → β2 transition shown in Figure 1B (each trajectory being calculated as the time-dependent intensity of the β2 infrared component at 1596 cm−1). The lag-phase corresponding to nucleation of fibrils is already very short for the 1 wt % sample and practically unmeasurable for the two higher concentrations. For the slightly diluted (Glu)200 samples (0.5 and 0.3 wt %), the lagphase becomes much longer. The kinetic data in Figure 1B illustrates the obvious intermolecular aspect of the amyloidogenic self-assembly: the concentration of soluble peptide affects both the probability of nucleation events and the rate of diffusion-limited elongation. However, concentration-independent intramolecular interactions also contribute to the peptide’s innate tendencies (or lack thereof) to aggregate. Determination of chain-length dependence of the amyloidogenic propensity is a starting point to evaluate relative importance of these contributions. In order to

MATERIALS AND METHODS

Samples. PLGA ((Glu)200). We will use term “(Glu)200” to refer to the particular commercial fraction of PLGA used in this study. (Glu)200 (as sodium salt, cat. No. P4761, Lot # 096 K5103 V, MW 15−50 kDa) was from Sigma, USA. D2O and DCl were purchased from ARMAR Chemicals, Switzerland. Branching of homopolypeptide main chains, which may occur in commercial preparations of PLGA, is undesirable, as it prevents proper folding and packing of polypeptide structures. The (Glu)200 lot used in this work is the same as in our previous studies, and it has been extensively characterized in terms of linearity and MW.24,26 Short α-L-Glutamic Acid Oligomers. (Glu)10, (Glu)6, (Glu)5, (Glu)4, and (Glu)3 were obtained by means of solid phase peptide synthesis (SPPS) with the use of preloaded with Fmoc-Glu(OtBu)Wang resin (capacity of 0.55 mmol/g). The Fmoc deprotection was performed using 20% piperidine in DMF for 5 min and after that for 20 min with a fresh portion of a deprotection solution. Each glutamic acid coupling was carried out with an excess of Fmoc-Glu(OtBu)−OH (3 equiv), coupling reagent TBTU (3 equiv), and DIPEA (5 equiv) dissolved in DMF for over 1 h. The Kaiser test was performed to check the completeness of each coupling step. After washing of peptide-resin with DMF, DCM and Et2O polyglutamic acid peptides were cleaved from solid support with a mixture of TFA:phenol:H2O:TIS (88:5:5:2, v/v/v/v) by constant shaking for 4 h. After filtration of resin, TFA was evaporated, and the residue was precipitated from Et2O. The crude products were analyzed on RP-HPLC, purified, and lyophilized. The purity of peptides was determined by RP-HPLC and characterized with an LC-IT-TOF mass spectrometer. All further experimental details and chromatographic as well as ESI-HR-MS-based characterization of synthesized peptide samples are placed in the Supporting Information. Preparation of β2 Aggregates from (Glu)n Peptides. For the experiments reported in Figures 2, 3, and 4, the following protocol of preparation of β2-aggregates was employed. Peptides in the form of sodium salts were dissolved in D2O at 1 wt % concentration. Clear solutions were gradually acidified at room temperature with diluted DCl to pD 4.1 (uncorrected pH-meter readout) and subjected to 72-hlong quiescent incubation at 60 °C. Under these conditions, β2 aggregates formed and precipitated over time (the process appears to be fastest in the case of (Glu)200). Insoluble aggregates were subsequently subjected to FT-IR/AFM measurements. For kinetic experiments, freshly acidified peptide samples were swiftly transferred to CaF2 cell. Depending on the case, seeds (i.e., fibrils preformed at 60 °C, as described above, and sonicated afterward) were added to the peptide solution at a 1:33 seed-soluble peptide mass ratio (for seeding of 0.3 wt % samples of soluble peptides). 10501

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir

on the spectra of β2-(Glu)n; data not shown). Until a more detailed study of this problem is completed, one could argue that the proximity of an ionized N-terminal −NH3+ (−ND3+) group to a carboxyl group buried within the tightly packed solvent-sequestered β2-aggregate could keep the latter one in the ionized state, as well.31 The monotonous intensity gain of the 1585 cm−1 band for n decreasing from 200 to 4 could be rationalized as a gradually increasing ratio of carboxyl groups interacting with N-termini versus all remaining carboxyl groups. This scenario can be envisioned for charged N-terminal amine groups selectively forming salt bridges with C-terminal carboxyl groups (in a framework of head-to-tail assembled (Glu)n chains) with Glu side-chain carboxyl groups remaining neutral and involved in bifurcated hydrogen bonds. The relatively high intensity of the 1585 cm−1 band, especially for β2-(Glu)4, where it coincides with reduced splitting between 1596 and 1639 cm−1 peaks as they shift to 1604 and 1633 cm−1, respectively, suggests that coupling of transition dipole moments and “intensity borrowing” must be considered in any adequate theoretical explanation of the β2-(Glu)n infrared signatures. The spectra of aggregated peptides (from n = 200 to n = 4) are clearly different from that of (Glu)3, where no visible aggregation took place within 72 h of incubation at 60 °C. The 1675 cm−1 peak is likely to arise from traces of TFE. We conclude that, according to FT-IR spectroscopy, the critical length of (Glu)n backbone permissive to the formation of amyloid fibrils (at least under the conditions used in this study) corresponds to four Glu residues. Subsequently, we have used AFM imaging to verify the presence of fibrillar structures in β2-(Glu)n samples. Representative tapping mode AFM images corresponding to samples previously studied using infrared absorption (Figure 2) are

Figure 1. (A) Time-lapse spectra of PLGA ((Glu)200) undergoing conformational transition from the α-helical to β2-amyloid structure collected over the period of 24 h. Experimental conditions: 1 wt % (Glu)200 in D2O, pD 4.1, 40 °C. (B) Concentration-dependent kinetics of the α → β2 transition in (Glu)200 plotted according to normalized intensities of the main amide I′ band component assigned to β2-fibrils (at ca. 1596 cm−1) during the first 5 h of aggregation.

address this problem, acidified D2O-based solutions of (Glu)n peptides of various length were subjected to prolonged incubation at 60 °C known to strongly favor conversion of αhelical PLGA into fibrillar β-aggregates.22−26 We have noted that insoluble aggregates formed first in (Glu)n samples of high MWs with only the (Glu)3 solution remaining clear after 72 h of incubation. Figure 2 presents FT-IR data collected afterward. While the spectra of β2-aggregates formed by (Glu)200 and (Glu)10 are strikingly similar, with the number of Glu residues further decreasing, minor spectral changes begin to be apparent. From the spectra of (Glu)6 onward, the amide I′ band’s main component shifts from 1596 cm−1 to higher wavenumbers with a new broad band at ca. 1585 cm−1 gradually emerging. The −COOD region encompassing the split peaks at 1730 and 1720 cm−1 in the case of β2-(Glu)200 gradually reshapes, with the 1730 cm−1 peak decreasing in intensity and a new band being formed at ca. 1705 cm−1, the frequency coinciding with the −COOD band’s position in α-helical (Glu)200 (Figure 1). This trend continues until n reaches 4, for which the 1705 cm−1 peak becomes dominant in the −COOD region, while the 1585 cm−1 band tops the low frequency range. In fact, the steady intensity gain of this band observed for β2-(Glu)n with decreasing n is quite puzzling given its most likely assignment to antisymmetric stretching vibrations of ionized −COO¯ groups,30 and the fact that at pD 4.1, the majority of carboxyl groups should not be ionized (further gradual acidification of fibril suspensions to pD 3.6, 3.1, and 2.6 has virtually no impact

Figure 2. Chain-length dependence of FT-IR spectra of (Glu)n peptides subjected to aggregation-promoting incubation. Experimental conditions: 1 wt % (Glu)n in D2O, pD 4.1, subjected to 72 h-long incubation at 60 °C; spectra were collected afterward at 25 °C.

presented in Figure 3. All analyzed samples except for (Glu)3 reveal long straight unbranched aggregates that appear to gain crystalline character with n approaching 4. Individual fibrils tend to align laterally and form superstructural assemblies, which become twisted in the case of β2-(Glu)200. Similar observation has been reported in our previous work based on electron microscopy.24 According to the infrared and AFM data shown in Figures 2 and 3, there is a clear correlation between the β2-like spectral fingerprint features and fibrillar morphology 10502

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir

Figure 3. AFM tapping-mode images of various (Glu)n aggregates obtained under the same conditions reported in Figure 2. Insets show cross sections of selected fibrillary specimen.

for aggregates self-assembled from oligo α-L-glutamic acids at least four residues long. Even the largest superstructural β2(Glu)n aggregates are built of thin individual fibrils. Using height-mode AFM, we have carried out an extensive statistical characterization of the thickness of such entities and of the larger structures they tend to form. Figure 4 presents averaged values of diameter of elementary fibrils and larger assemblies of β2-(Glu)n aggregates obtained from multiple cross-section measurements of accessible specimen based on height-mode AFM. Intriguingly, the measured diameters of individual fibrils of (Glu)n peptides are in the same narrow range of approximately 4−6 nm. This holds true even for fibrils assembled from units dramatically different in terms of number of residues (e.g., (Glu)4 and (Glu)200). This unexpected finding provides important insights into the spatial packing of polyglutamate main chains within the β2-fibril structure: for example, this result seems to be quite incompatible with “layered” amyloid models proposed for Aβ fibrils, where amyloid cross-section increases proportionally with main chain length (e.g., ref 32). We note that, under the conditions of this study, well-defined twisted superstructures are formed only by (Glu)200 chains (Figures 3 and 4), implying that the length threshold of the main chain may depend on the level of hierarchical self-assembly of fibrils. The so-far presented results seem to suggest that protonated (Glu)n chains of various lengths (except for n ≤ 3) selfassemble into amyloid-like fibrils according to a common structural theme involving antiparallel β-sheets, bifurcated hydrogen bonds, and similar 3D-packing patterns. Such a universal chain-length-independent structural motif should be conducive to the capacity to cross-seed, i.e., induce and catalyze growth of fibrils built of chains of different length. Proper docking interactions between structurally compatible amyloid seeds and monomers would lead to elongation of the seeds and shortening of the lag-phase, which is the rationale for conducting the following kinetic experiments on selected pairs of (Glu)5 and (Glu)200 peptides, which were accessible

Figure 4. Histograms showing averaged diameters of single fibrils (top panel) and fibrillar superstructures (bottom panel) of various (Glu)n aggregates. The data were extracted from multiple cross section measurements of fibrils using AFM height mode.

to us as high-purity chemicals in sufficient quantities. Figure 5 shows time-lapse FT-IR spectra of appropriately acidified 0.3 wt % solutions of (Glu)5 and (Glu)200 undergoing fibrillation in the absence and presence of homologous and heterologous 10503

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir

Figure 5. Time-lapse FT-IR spectra of (Glu)5 (left column) and (Glu)200 (right column) undergoing the α → β2 transition in the absence of seeds (A), in the presence of homologous seeds (B), and upon cross-seeding with the opposite type of seeds (C). Kinetics of the α → β2 transition in (Glu)5 and (Glu)200 for different fibrillation scenarios plotted according to normalized intensities of the 1596 cm−1 peak. (D) Experimental conditions: in each case, concentration of soluble peptide in D2O, pD 4.1, was 0.3 wt %, the mass ratio of soluble peptide:seed was 33:1, and the spectra were collected at 40 °C over the period of 24 h.

amyloid fibrilsthe infrared features of daughter fibrils induced upon cross-seeding are typical for amyloid formed de novo and independent of the type of seed used.33 According to our results, in an appropriately acidified environment, even short (Glu)n peptides remain strongly amyloidogenic, sharing similar morphological and conformational characteristics, as well as the ability to cross-seed one another. The observation that (Glu)6, (Glu)5, and (Glu)4 form β-pleated aggregates contradicts previous circular dichroismbased studies by Rinaudo and Domard claiming that, for oligoL-glutamic acids, the minimum n for β-sheet-formation is close to 8−10.17,34 This discrepancy appears to have two likely sources. First, the degree of polymerization and homogeneity of (Glu)n samples used in this study are firmly controlled. The samples were synthesized using the SPPS method, then purified by RP-HPLC and analyzed afterward using mass spectrometry, whereas in the works by Rinaudo and Domard, fractions of polydisperse PLGA were separated on ion-exchange chromatographic columns and analyzed in terms of degree of

seeds. Under the low concentration regime, the lag phase of (Glu)200 fibrillation was sufficiently long (Figure 1B) to enable detection of seeding activity of added fibrils. In the absence of seeds, the (Glu)5 sample remains in the nucleation phase throughout a 24-h-long incubation period with the corresponding FT-IR spectra remaining intact, whereas (Glu)200 enters an elongation phase after an initial delay of approximately 3 h (Figure 5A,D). Addition of homologous seeds results in pronounced acceleration of fibrillation of both peptides with the lag phases disappearing altogether (Figure 5B). The key result of this experiment is shown in panel C, where the two monomeric peptides are cross-seeded with opposite types of β2template. The observed fast spectral changes are compelling evidence that β2-fibrils built of (Glu)n chains of different lengths act as equally effective catalysts of amyloid growth as homologous seeds. This, in turn, further supports the idea that, regardless of degree of polymerization, (Glu)n chains form amyloid-like fibrils according to a generic structural theme. We report no evidence of self-propagating polymorphism of (Glu)n 10504

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir polymerization using viscosimetric measurements.17,34 We argue that the approach employed in our work is more adequate and reliable in providing homogeneous (Glu)n samples of defined n. Second, unlike in our study, samples used in the previous work had both N- and C-termini covalently modified which, especially for short peptides, may affect their ability to form aggregated β-sheets.26 It should be stressed, however, that double Glu → Val substitutions in derivatives of (Glu)10 peptide studied by Keiderling et al. were shown to adapt well to the β2-architecture and maintain both the infrared traits and fibrillar morphology.27 The most puzzling, yet fascinating finding of this study is the virtual independence of β2-fibrils’ diameter of (Glu)n chains’ length. This observation adds a new and strict constrain to hypothetical structural models of the fibril, which should also accommodate the antiparallel arrangement of β-strands, and the presence of bifurcated hydrogen bonds27. The coexistence of these three conditions contradicts certain otherwise attractive structural models, for example, those based on β-helix, which would entail a parallel arrangement of strands.35 In Figure 6, we are laying out our basic hypothesis on possible spatial packing of (Glu)n chains within β2-fibril. Panel A shows the structure of an aggregate consisting of layered antiparallel β-sheets with individual strands corresponding to whole extended short chains (e.g., (Glu)5). These strands are packed head-to-tail to enable formation of periodic salt bridges between N- and C- termini, as postulated above. Although shortening of the intersheet distance (compressing the aggregate along vertical Z axis) could satisfy conditions for bifurcated hydrogen bonds involving Glu side chains, such structure, upon unrestricted growth, would intuitively expand in all three directions ,which hardly resonates with the fibrillar nature of (Glu)n aggregates. In panel B, a similar arrangement of layered antiparallel β-sheets built of long polypeptide chains (e.g., (Glu)200 ) is shown. Each infinite sheet can be conceptualized as a continuous assembly of β-sheet-turn motives with a stretch of main chain in β-turn conformation connecting two antiparallel β-strands. Such motives have been implicated as building blocks in polyglutamine-like (e.g., (Lys)2(Gln)41(Lys)2,36 or (Lys)2(Gln)24(Lys)2Trp37) amyloid fibrils associated with Huntington disease. As the turn-rich side of the aggregate is expected to have a diminished capacity to recruit soluble monomers, this would result in a more directional growth. In fact, the growth would become unidirectional should the stacked β-sheets be arranged circularly in a rosette-like pattern, as shown in Figure 6C with turns protruding to the outside and inside. For a single βhairpin-turn stretch involving 13 Glu residues (i.e., two βstrands each 5-residues-long connected by a 3-residue-long turn region), the fibril’s external diameter would be roughly 6.5 nm, and the internal channel approximately 1 nm thick. An internal channel of ca. 1.2 nm in diameter running through an infinite βhelix was proposed by Perutz et al. as a structural model for polyglutamine fibrils.35 Although the sketch of a structural concept of (Glu)n amyloid fibrils shown in Figure 6C is hypothetical at best, it does accommodate the main structural constrains deciphered so far from the experimental studies on β2-(Glu)n, and on derivatives of the closest homopolypeptide analogue: polyglutamine.36,37 The rosette-like packing shares similarities with the structural model proposed earlier for a particular type of polyamide fibrils (notably Kevlar) by Dobb et al.38 Certainly, plenty of further work is needed to untangle the intricacies of the amyloidogenic self-assembly of (Glu)n

Figure 6. Examples of possible spatial arrangements of individual chains of (Glu)5 (A) and (Glu)200 (B) consistent with aggregated antiparallel β-sheet architecture and the presence of bifurcated hydrogen bonds characteristic for the β2-fibrils (only two such bonds are indicated for the sake of clarity). A single β-hairpin turn in B consists of 13 Glu residues. (C) Hypothetical model of 3D-packing of individual antiparallel β-sheets into an amyloid-like fibril. Visualizations were prepared using the VMD program.39

peptides and to build a plausible comprehensive and accurate model of β2-fibrils. The goals are well worth the efforts, however, given the biomedical importance of homopolypeptide aggregation and emerging applications of PLGA and its derivatives in biodegradable drug delivery (e.g., refs 40,41), tumor targeting gene carriers (e.g., ref 42), or surface modifications for enhanced cell adhesion (e.g., ref 43). In conclusion, among oligo-α-L-glutamic acids, only peptides shorter than tetramers appear to lack the capacity to form amyloid-like fibrils. Fibrils self-assembled from chains of quite different lengths feature remarkably similar diameters, infrared spectral characteristics, and the ability to cross-seed one another. Our study suggests that (Glu)n peptides form fibrils according to a single unifying structural theme regardless of their length. 10505

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir



(12) Song, S.; Asher, S. A. UV Resonance Raman Studies of Peptide Conformation in Poly (L-lysine), Poly (L-glutamic acid), and Model Complexes: The Basis for Protein Secondary Structure Determinations. J. Am. Chem. Soc. 1989, 111, 4295−4305. (13) Krejtschi, C.; Hauser, K. Stability and Folding Dynamics of Polyglutamic Acid. Eur. Biophys. J. 2011, 40, 673−685. (14) Gooding, E. A.; Sharma, S.; Petty, S. A.; Fouts, E. A.; Palmer, C. J.; Nolan, B. E.; Volk, M. pH-dependent Helix Folding Dynamics of Poly-glutamic Acid. Chem. Phys. 2013, 422, 115−123. (15) Donten, M. L.; Hamm, P. pH-jump Induced α-helix Folding of Poly-L-glutamic Acid. Chem. Phys. 2013, 422, 124−130. (16) Gregory, M. J.; Anderson, M.; Causgrove, T. P. Measurement of Energy Barriers to Conformational Change in Poly-L-glutamic Acid by Temperature-Derivative Spectroscopy. Chem. Phys. 2013, 420, 1−6. (17) Rinaudo, M.; Domard, A. Circular Dichroism Studies on alphaL-glutamic Acid Oligomers in Solution. J. Am. Chem. Soc. 1976, 98, 6360−6364. (18) Dzwolak, W.; Muraki, T.; Kato, M.; Taniguchi, Y. Chain-length Dependence of α-helix to β-sheet Transition in Polylysine: Model of Protein Aggregation Studied by Temperature-Tuned FTIR Spectroscopy. Biopolymers 2004, 73 (4), 463−469. (19) Ricchiuto, P.; Brukhno, A. V.; Paci, E.; Auer, S. Communication: Conformation State Diagram of Polypeptides: A Chain Length Induced α-β Transition. J. Chem. Phys. 2011, 135 (6), 061101. (20) Brukhno, A. V.; Ricchiuto, P.; Auer, S. Tracking Polypeptide Folds on the Free Energy Surface: Effects of the Chain Length and Sequence. J. Phys. Chem. B 2012, 116 (29), 8703−8713. (21) Wetzel, R. Physical Chemistry of Polyglutamine: Intriguing Tales of a Monotonous Sequence. J. Mol. Biol. 2012, 421 (4), 466− 490. (22) Itoh, K.; Foxman, B. M.; Fasman, G. D. The Two β Forms of Poly (L-glutamic acid). Biopolymers 1976, 15, 419−455. (23) Fulara, A.; Dzwolak, W. Bifurcated Hydrogen Bonds Stabilize Fibrils of Poly (L-glutamic) Acid. J. Phys. Chem. B 2010, 114, 8278− 8283. (24) Fulara, A.; Lakhani, A.; Wójcik, S.; Nieznańska, H.; Keiderling, T. A.; Dzwolak, W. Spiral Superstructures of Amyloid-like Fibrils of Polyglutamic Acid: An Infrared Absorption and Vibrational Circular Dichroism Study. J. Phys. Chem. B 2011, 115, 11010−11016. (25) Yamaoki, Y.; Imamura, H.; Fulara, A.; Wójcik, S.; Bożycki, Ł.; Kato, M.; Keiderling, T. A.; Dzwolak, W. An FT-IR Study on Packing Defects in Mixed β-Aggregates of Poly (L-glutamic acid) and Poly (Dglutamic acid): A High-Pressure Rescue from a Kinetic Trap. J. Phys. Chem. B 2012, 116, 5172−5178. (26) Fulara, A.; Hernik, A.; Nieznańska, H.; Dzwolak, W. Covalent Defects Restrict Supramolecular Self-Assembly of Homopolypeptides: Case Study of β2-Fibrils of Poly-L-Glutamic Acid. PLoS One 2014, 9, e105660. (27) Chi, H.; Welch, W. R.; Kubelka, J.; Keiderling, T. A. Insight into the Packing Pattern of β2 Fibrils: A Model Study of Glutamic Acid Rich Oligomers with 13C Isotopic Edited Vibrational Spectroscopy. Biomacromolecules 2013, 14, 3880−3891. (28) Kessler, J.; Keiderling, T. A.; Bour, P. Arrangement of Fibril Side Chains Studied by Molecular Dynamics and Simulated Infrared and Vibrational Circular Dichroism Spectra. J. Phys. Chem. B 2014, 118, 6937−6945. (29) Babenko, V.; Dzwolak, W. Thioflavin T Forms a NonFluorescent Complex with α-helical Poly-l-glutamic Acid. Chem. Commun. 2011, 47 (38), 10686−10688. (30) Barth, A. The Infrared Absorption of Amino Acid Side Chains. Prog. Biophys. Mol. Biol. 2000, 74 (3), 141−173. (31) Anderson, D. E.; Becktel, W. J.; Dahlquist, F. W. pH-induced Denaturation of Proteins: A Single Salt Bridge Contributes 3−5 kcal/ mol to the Free Energy of Folding of T4 Lysozyme. Biochemistry 1990, 29 (9), 2403−2408. (32) Petkova, A. T.; Ishii, Y.; Balbach, J. J.; Antzutkin, O. N.; Leapman, R. D.; Delaglio, F.; Tycko, R. A Structural Model for Alzheimer’s β-amyloid Fibrils Based on Experimental Constraints from

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b02915. Full synthetic procedure and characterization of (Glu)3, (Glu)4, (Glu)5, (Glu)6, (Glu)10 (PDF)



AUTHOR INFORMATION

Corresponding Author

*Phone: +48 22 552 6567; Fax: +48 22 822 5996; E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was supported by the National Science Centre of Poland, grant no. DEC-2011/03/B/ST4/03063. The study was carried out at the Biological and Chemical Research Centre, University of Warsaw, established within the project cofinanced by EU from the European Regional Development Fund under the Operational Programme Innovative Economy, 2007-2013, and with the use of CePT infrastructure financed by the same EU programme.



ABBREVIATIONS: AFM, atomic force microscopy; CD, circular dichroism; FT-IR, Fourier transform infrared; MW, molecular weight; PLGA, poly-L-glutamic acid; RP-HPLC, reversed-phase high-performance liquid chromatography; SPPS, solid phase peptide synthesis; TFE, trifluoroethanol



REFERENCES

(1) Chiti, F.; Dobson, C. M. Protein Misfolding, Functional Amyloid, and Human Disease. Annu. Rev. Biochem. 2006, 75, 333−366. (2) Uversky, V. N.; Fink, A. L. Conformational Constraints for Amyloid Fibrillation: The Importance of Being Unfolded. Biochim. Biophys. Acta, Proteins Proteomics 2004, 1698, 131−153. (3) Fändrich, M.; Dobson, C. M. The Behaviour of Polyamino Acids Reveals an Inverse Side Chain Effect in Amyloid Structure Formation. EMBO J. 2002, 21, 5682−5690. (4) Colaco, M.; Park, J.; Blanch, H. The Kinetics of Aggregation of Poly-glutamic Acid Based Polypeptides. Biophys. Chem. 2008, 136 (2), 74−86. (5) Lai, J.; Fu, W.; Zhu, L.; Guo, R.; Liang, D.; Li, Z.; Huang, Y. Fibril Aggregates Formed by a Glatiramer-Mimicking Random Copolymer of Amino Acids. Langmuir 2014, 30 (24), 7221−7226. (6) Lai, J.; Zheng, C.; Liang, D.; Huang, Y. Amyloid-like Fibrils Formed by ε-Poly-L-lysine. Biomacromolecules 2013, 14 (12), 4515− 4519. (7) Blout, E. R.; Schmier, I.; Simmons, N. S. New Cotton Effects in Polypeptides and Proteins. J. Am. Chem. Soc. 1962, 84, 3193−3194. (8) Fasman, G. D. Poly-α-amino Acids. Protein Models for Conformational Studies; Marcel Dekker, Inc.: New York, 1967. (9) Doty, P.; Wada, A.; Yang, J. T.; Blout, E. R. Polypeptides. VIII. Molecular Configurations of Poly-L-glutamic Acid in Water-Dioxane Solution. J. Polym. Sci. 1957, 23, 851−861. (10) Yan, J. F.; Vanderkooi, G.; Scheraga, H. A. Conformational Analysis of Macromolecules. V. Helical Structures of Poly-L-aspartic Acid and Poly-L-glutamic Acid, and Related Compounds. J. Chem. Phys. 1968, 49, 2713−2726. (11) Appel, P.; Yang, J. T. Helix-Coil Transition of Poly-L-glutamic Acid and Poly-L-lysine in D2O. Biochemistry 1965, 4, 1244−1249. 10506

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507

Article

Langmuir Solid State NMR. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (26), 16742− 16747. (33) Surmacz-Chwedoruk, W.; Nieznańska, H.; Wójcik, S.; Dzwolak, W. Cross-seeding of Fibrils from Two Types of Insulin Induces New Amyloid Strains. Biochemistry 2012, 51 (47), 9460−9469. (34) Rinaudo, M.; Domard, A. β-structure Formation and Its Stability in Aqueous Solutions of α-L-glutamic Acid Oligomers. Macromolecules 1977, 10 (3), 720−721. (35) Perutz, M. F.; Finch, J. T.; Berriman, J.; Lesk, A. Amyloid Fibers Are Water-Filled Nanotubes. Proc. Natl. Acad. Sci. U. S. A. 2002, 99 (8), 5591−5595. (36) Kar, K.; Hoop, C. L.; Drombosky, K. W.; Baker, M. A.; Kodali, R.; Arduini, I.; van der Wel, P. C. A.; Horne, W. S.; Wetzel, R. βhairpin-mediated Nucleation of Polyglutamine Amyloid Formation. J. Mol. Biol. 2013, 425 (7), 1183−1197. (37) Buchanan, L. E.; Carr, J. K.; Fluitt, A. M.; Hoganson, A. J.; Moran, S. D.; de Pablo, J. J.; Skinner, J. L.; Zanni, M. T. Structural Motif of Polyglutamine Amyloid Fibrils Discerned with Mixed-Isotope Infrared Spectroscopy. Proc. Natl. Acad. Sci. U. S. A. 2014, 111 (16), 5796−5801. (38) Dobb, M. G.; Johnson, D. J.; Saville, B. P. Supramolecular Structure of a High-Modulus Polyaromatic Fiber (Kevlar 49). J. Polym. Sci., Polym. Phys. Ed. 1977, 15 (12), 2201−2211. (39) Humphrey, W.; Dalke, A.; Schulten, K. VMD - Visual Molecular Dynamics. J. Mol. Graphics 1996, 14, 33−38. (40) Tansey, W.; Ke, S.; Cao, X. Y.; Pasuelo, M. J.; Wallace, S.; Li, C. Synthesis and Characterization of Branched Poly (L-glutamic acid) As a Biodegradable Drug Carrier. J. Controlled Release 2004, 94, 39−51. (41) Li, C. Poly (L-glutamic acid)−Anticancer Drug Conjugates. Adv. Drug Delivery Rev. 2002, 54, 695−713. (42) Tian, H.; Guo, Z.; Lin, L.; Jiao, Z.; Chen, J.; Gao, S.; Zhu, X.; Chen, X. pH-responsive Zwitterionic Copolypeptides As Charge Conversional Shielding System for Gene Carriers. J. Controlled Release 2014, 174, 117−125. (43) Richert, L.; Arntz, Y.; Schaaf, P.; Voegel, J. C.; Picart, C. pH Dependent Growth of Poly (L-lysine)/Poly (L-glutamic) Acid Multilayer Films and Their Cell Adhesion Properties. Surf. Sci. 2004, 570, 13−29.

10507

DOI: 10.1021/acs.langmuir.5b02915 Langmuir 2015, 31, 10500−10507