Characterizing aggregate growth and morphology of alanine-rich polypeptides as a function of sequence chemistry and solution temperature from scattering, spectroscopy, and microscopy

Abstract

The aggregation behavior and stability of a series of alanine-rich peptides, which are included as components of peptide-polymer conjugates, were characterized using a combination of biophysical techniques. Light scattering techniques were used to monitor changes in peptide morphology and size distributions as a function of time and temperature. The results show large particles immediately upon dissolution in buffer. At room temperature, these particles relaxed to reach a mostly monomeric peptide state, while at higher temperatures, they grew to form aggregates. Circular dichroism spectroscopy (CD) was used to monitor temperature- and time- dependent conformational changes as a function of peptide sequence and incubation time. CD measurements reveal that all of the sequences are helical at low temperatures with transitions to non-helical conformation with increased temperature. Samples incubated at room temperature were able to recover their original helicity. At increased temperature, the shorter and longer peptide sequences showed notable changes in conformation, and were not able to recover their original helicity after 72 hours. After incubation for up to one week, β-sheet conformations were observed in these two cases, while only α-helical conformation loss was observed for the peptide of intermediate molecular weight. Transmission electron microscopy measurements reveal the formation of fibrils after 72 hours of incubation at 60 °C for all samples, in agreement with the scattering measurements. Additional quenching experiments show that peptide aggregation can be stalled when solutions are cooled to room temperature.

Introduction

Unfolding, self-assembly and irreversible polypeptide aggregation can be used as a controlled and tunable process to create nanostructures with industrial and biomedical applications.[1–5] Molecular-scale understanding of these processes is needed to better design and predict the behavior and relative stability of aggregate species as a function of peptide sequence and solution environment.[6,7] For instance, polypeptide and peptide–polymer conjugate materials have been tuned to produce biomaterials upon self-assembly or aggregation, as these combine advantageous biochemical properties not usually found in inorganic and other organic component materials.

Previously reported polypeptide-based hybrid materials have shown biological functionality, with controlled self-assembly capabilities.[2,5] Of interest among these conjugates are materials with assembly that can be controlled via temperature,[8–10] pH,[11,12] ionic environment,[13,14] driven by hydrophobic interactions,[15,16] or the triggering of the formation of specific secondary structure in the peptide domains.[2,5,17] Triggered assembly by controlling solution conditions permits significant flexibility in assembly triggers and final structures the conjugates. With the aim of producing polypeptide-polymer conjugates that may capture the properties of select structural, multiblock proteins and show triggered assembly, we have previously reported the unfolding and solution behavior of a series of alanine-rich peptides as a function of solution temperature.[18] Alanine-rich peptides were selected because they have been demonstrated historically to be model helical peptides at low solution temperatures with a transition to random coil structures at higher temperatures.[19–21]

Additionally, alanine-rich peptides have been shown to associate into various structures. Blondelle and coworkers found that polyalanine-based peptides were able to undergo conformational conversion into β-pleated-sheet complexes.[22] Shinchuck et al, also investigated a series of polyalanine peptides varying in length from 7–20 alanine residues which underwent fibril formation under various temperature and pH conditions.[23] Giri et al examined the pH effects on a series of polyalanine peptides varying in length. The two longest sequences within that study, 11 and 17 alanine-residues, both showed enhancement of β-sheet content and formation of fibrils.[24] Additionally, Bernacki et al characterized a series polyalanine length variants undergoing association into small oligomers, fibrils and large spherical agglomerates dependent on peptide length and temperature. Although they observed fibrilization to occur in the 19-repeat peptide, they found that the longest species studied (25 repeats) did not actually undergo fibrilization at elevated temperature, but formed large spherical agglomerates, demonstrating that tuning the length of the alanine-rich peptides does indeed have an effect on structure of the association.[25] Furthermore, Castelletto et. al. characterized an alanine-rich surfactant-like peptide (A₆K) which was able to form single-wall nanotubes. Using X-ray diffraction, they proposed that the structure of the assembly was due to electrostatic charge from the lysine residues, coupled with the efficient packing of the peptide.[26] Additionally, Castelletto et. al. has also studied alanine-rich amphiphilic peptides (KA₆K and KA₆E) with homotelechelic or heterotelechelic charged termini. Specifically, they found that the KA₆E sequence was able to self-assemble into tape-like fibrils at low concentration, while hydrogelation was observed at higher concentration. This assembly was attributed to the favorable interactions between the termini.[27] Castelletto et. al. also investigated the self-assembly of an alanine-rich peptide functionalized with RGD (A₆RGD). The assembly of this surfactant-like peptide was found to be concentration dependent, with oligomers being formed at lower concentrations (0.1 wt%), and peptide fibrils (1.0wt%) and long peptide fibers (2.5 wt%) formed at higher concentrations. At concentrations between 2–15 wt% vesicles were found to co-exist with fibers.[28]

Polypeptide aggregation and self-assembly is often quantified by performing ex-situ characterization of polypeptide and protein solutions - e.g., by performing isothermal incubations for set amounts of time, quenching and then performing solution characterization.[29] However, this approach limits the understanding of the aggregation kinetics to comparisons of initial and final conformational and associate behavior. In this regard, simultaneous multiple sample light scattering (SMSLS) was developed to allow for in-situ monitoring of the aggregation events of polypeptides and protein solutions, providing detailed understanding of assembly processes and reducing the required solution volumes to perform aggregation studies.[30] Additionally, SMSLS allows one to monitor several solutions in parallel, reducing the total time needed to obtain kinetic data to select time points for further isothermal incubations and/or solution characterization. This technique could also be combined with ex-situ techniques to further test and validate the aggregation behavior of polypeptide and protein solutions.

This report focuses on the thermally induced aggregation behavior of a series of alanine-rich peptides with similar chemistry and various chain lengths. These sequences were selected from previously reported polypeptides that had been demonstrated to form β-sheet structures under select solution conditions;[31–33] the use of peptides derived from these longer sequences is of interest in developing associative domains in multiblock peptide-polymer conjugates. For better monitoring and understanding of the kinetics and dynamics of these aggregation events, four characterization techniques were used. Circular dichroism (CD) was used to characterize the secondary structure of the solution pre- and post-incubation combination. Dynamic light scattering (DLS) and transmission electron microscopy (TEM) were used to characterize the morphology of polypeptide aggregates as a function of incubation time and solution temperature. Finally, SMSLS was used to evaluate the qualitative in-situ aggregation kinetics of the studied sequences. Peptide aggregates with semi-uniform morphology, and the presence of β-sheet conformation were observed at elevated temperatures, with good agreement across the four characterization techniques. The combination of these four techniques allowed a more comprehensive understanding of the aggregation behavior of the studied sequences.

Materials and methods

Peptide synthesis and purification

All materials were purchased from Fischer Scientific (Pittsburgh, PA) except where otherwise indicated. Peptides were synthesized on a Rink amide resin (ChemPep, Wellington, FL). Specifically, the AQK18 sequence listed in Table 1 was synthesized with a PS3 peptide synthesizer (Protein Technologies, Tucson, AZ). Sequences AQK27 and AQK35 were synthesized with a Focus XC peptide synthesizer (AAPTec Inc, Louisville, KY). Fmoc-alanine, Fmoc-lysine(boc), Fmoc-glutamic acid (t-butyl), Fmoc-glutamine(trt), and Fmoc-phenylalanine were all purchased from ChemPep. The N-terminus of each peptide was acetylated, and peptides were cleaved in 95% trifluoracetic acid (TFA), 2.5% H2O, and 2.5% triisopropylsilane (Sigma-Aldrich, St. Louis, MO). TFA was mostly evaporated, and peptides were then precipitated twice into cold ethyl ether. Samples were redissolved in water, frozen in liquid nitrogen, and lyophilized. Dried samples were then reconstituted in water and purified by preparative-scale reverse-phase high-performance liquid chromatography (RP-HPLC) using a Waters Xbridge BEH130 Prep C-18 column. The mobile-phase comprised gradients of degassed, deionized water with 0.1% TFA and acetonitrile with 0.1% TFA, at a flow rate of 21 ml/min. Peptide was detected by UV absorbance at 214 nm, and fractions were collected and lyophilized. Molecular weights of the purified peptides were verified by electrospray ionization mass spectroscopy (ESI-MS).

Table 1.

Alanine-rich peptide sequences

Name	Sequence	Mass
AQK18	K(AAAQ)4K	1680.9
AQK27	K(AAAQ)3K(AAAQ)3K	2491.7
AQK35	K(AAAQ)4K(AAAQ)4K	3174.5

Sample preparation

Potassium phosphate buffer stock solutions were prepared by dissolving potassium phosphate dibasic (Fisher Scientific) in deionized water (MilliQ, Millipore) to reach 10 mM phosphate concentration and titrated to pH 7.2 ± 0.05 (termed pH 7 below) using 5 N aqueous hydrochloric acid solution (Fisher Scientific) and filtered with a 0.22 μm syringe filter (Millipore). Lyophilized peptide powder was dissolved into the potassium phosphate buffer at a concentration of 2, 4 and 10 g/L. Samples were spun at 15000 RCF for 10 minutes. After incubation at the selected temperatures, aliquots from solution were subject to gentle handling and were taken by pipetting for immediate characterization. Final peptide concentrations were validated using UV-Vis spectrophotometry (Agilent 8453, Santa Clara, CA).

Circular dichroism spectroscopy (CD)

Experimental characterization of the average secondary structure of the various peptides was conducted via circular dichroism spectroscopy on a Jasco 810 circular dichroism spectropolarimeter (Jasco Inc, Easton, MD, USA). CD spectra were recorded using a quartz cell with 0.1 cm optical path length. Samples for full wavelength scans incubated at various temperatures were cooled for three minutes at 0 °C prior to the start of the experiment. For each wavelength scan, the scanning rate was 50 nm/min, with a response time of 4 s. Wavelengths from 195 nm to 250 nm were recorded at increments of 0.5 nm. (MRE) values at 222 nm ([Θ]_MRE,222) were used to calculate fraction of α-helicity was calculated by using a previously established method that is based on idealized long helices.[34]

Simultaneous Multiple Sample Light Scattering (SMSLS)

Polypeptide aggregation was monitored in-situ by performing SMSLS experiments. These were carried out in an SMSLS instrument (Fluence Analytics, formerly Advanced Polymer Monitoring Technologies, Inc., New Orleans, LA). This instrument has 16 independent sample cells, each with its own 660 nm, vertically polarized 30 mW diode laser providing incident light, along with independent, programmable temperature control and stepper-motor-controlled stirring, and fiber optic-coupled light scattering detection at 90°. Unlike singe-cell instruments, the SMSLS has no moving optical components so that accurate alignment is permanently maintained for each cell, which further allows absolute measurements of polypeptide molecular weight (M_w) to be made over extended periods of time.[30] Experiments were performed at 25, 40, 60 and 80 °C for up to 72 hr, in small-volume glass cuvettes and a minimum sample volume of 150 μL. Cuvettes were calibrated using toluene (Fisher Scientific) and the stock potassium phosphate buffer prior to initiating the incubation of the polypeptide solutions.

Dynamic Light Scattering (DLS) and Transmission Electron Microscopy (TEM)

The aggregation behavior of the selected polypeptides was also characterized via DLS and TEM. Peptide solutions were incubated at 25, 40 and 60 °C in standard 2 mL HPLC glass vials. Aliquots were extracted very 24 hr for 72 hr and subject to DLS and TEM characterization. DLS measurements were performed on a Malvern Zetasizer Nano S (Malvern Instruments Ltd., Worcestershire, UK) at a scattering angle of 173° using a low volume quartz cuvette with a minimum sample volume of 12 μL. When quantification was performed, the data were processed using a cumulant analysis of the correlation function, and monomer or average aggregate size was measured from the computed diffusion coefficients using the Stokes–Einstein equation.[35] Each reported measurement was conducted from the average of at least three repeats from multiple different assessments of a given sample. TEM was conducted on a Tecnai 12 (FEI Company, Hillsboro, OR, USA) microscope using an accelerating voltage of 120 kV. TEM samples were prepared by adding a 5 μL drop of solution onto a 300-mesh ultrathin carbon-coated copper TEM grid (Electron Microscopy Sciences, Hatfield, PA). The sample was blotted with filter paper. Samples were stained with phosphotungstic acid (PTA).

Results and Discussion

1.1.1. Aggregation of alanine-rich peptides

Previously, we reported the synthesis of short, alanine-rich peptides and how their thermodynamic and unfolding behavior was accurately captured by a four-bead amino acid (4bAA) coarse grain molecular model.[18] These sequences were designed with the aim of resulting in soluble helical peptides which could be triggered into β-sheet-forming aggregates via external stimuli, such as temperature. Of the five previously reported peptides, three length variants were chosen for this work to characterize their aggregation: K(AAAQ)₄K (AQK18), K(AAAQ)₃K(AAAQ)₃K (AQK27), and K(AAAQ)₄K(AAAQ)₄K (AQK35) (see Table 1). These specific sequences were designed based on the hypothesis that increasing alanine stretches would promote aggregation. AQK18, AQK27, and AQK35 were successfully synthesized and purified according to our previous work,[18] and a suite of microscopic and spectroscopic techniques with online monitoring were used to characterize the aggregation.

Online monitoring of thermally induced peptide aggregation was achieved via isothermal SMSLS as a function of time. Solutions of AQK18, AQK27 and AQK35 were prepared in 10 mM potassium phosphate buffer at 4 mg/ml of peptide concentration. This peptide concentration was chosen based on preliminary trials between 0.5 mg/mL and 10 mg/mL. Characterization of solutions with concentrations below 2 mg/ml was hindered by the detection limits of SMSLS and DLS, and for solutions with peptide concentrations at and above 10 mg/ml, peptide aggregation was too rapid to consistently quantify across all techniques (data not shown). A peptide concentration of 4 mg/ml was chosen to enable comparison across the four characterization techniques as it produced consistent results while reducing the amount of peptide needed for experimental measurement. Data from the SMSLS characterization of the thermally induced aggregation as a function of time (up to 72 hours) for solutions of AQK18, AQK27 and AQK35 are shown in Figure 1. At 25 °C all peptide sequences remained monomeric in solution during the monitored time. When incubated at 40 °C, formation of small oligomeric species was detected within 24 hours for AQK18, with larger aggregates being formed in solution after ~65 hours. Samples of AQK27 remained monomeric for up to 72 hours. For AQK35, the initial presence of a smaller species may be the result of peptide association due to prior lyophilization. Their incubation at an elevated temperature may promote dissociation of these peptide oligomers yielding a monomeric solution. The apparent monomeric stability of these peptides at 25 °C and 40 °C (with the exception of the AQK18 at 72 hours) agree with trends reported on other alaninerich/polyalanine peptides. Marquesee et al. characterized alanine-rich peptides whose helical content was independent of peptide concentration (up to ~0.1 mg/ml), as expected for monomeric peptide.[19] In another study by Baldwin and co-workers, the thermal denaturation of short alanine-rich sequences was analyzed by helix-coil theory. At similar concentrations to those of Marquesee, these sequences displayed stable monomeric behavior, with the ability to retain the original α-helical conformation and monomeric state upon heating and subsequent cooling.[21] Bernecki et al. characterized the effect of polyalanine chain length on aggregation with sequences of YK(A)_nK (7 ≤ n ≤ 25; A7 to A25). All samples except for A25 were able to retain 90–95% monomer at concentrations of ~0.12–0.25 mg ml⁻¹ for up to 6 weeks of incubation at room temperature. The A25 peptide formed small oligomers upon dissolution, because of greater hydrophobicity and decreasing mean charge. Furthermore, in another study, monomeric polyalanine conversion to β-sheet complex was monitored as a function of peptide concentration.[22] At low concentrations (~0.7 mg/ml), solutions remained monomeric and partially α-helical after up to 200 hours of incubation at 65 °C. With increasing concentration (up to ~4.2 mg/ml), aggregation was observed, and β-sheet conversion was independent of incubation time (conversion did not increase after initial measurements after 3 hours).[22] It should be noted that aside from those reported by Blondelle et al, our concentrations are significantly higher than those reported in previous studies. The observation of stable monomeric solutions at such high concentrations is almost certainly attributable to the fact that the reported sequences are more hydrophilic due to the interspersed glutamines and central lysine. However, these previous studies do not offer insight as to why aggregation begins to occur for AQK18 and not the longer sequences after 72 hours at 40 °C. This will be investigated in future work.

Figure 1.

Online monitoring of peptide thermal-aggregation using isothermal SMSLS as a function of time for the AQK18 (panel A), AQK27 (panel B) and AQK35 (panel C) sequences at 25 °C (black solid lines), 40 °C (red dashed lines), 60 °C (blue dotted lines) and 80 (gray dash-dotted lines) for 24 hours (main panels) and up to 72 hours (insets) of incubation time. Data has been filtered for easy read.

At incubation temperatures of 60 °C and 80 °C, aggregation is observed for all sequences. In particular, similarities in aggregation behaviors are exhibited for AQK18 and AQK35, as compared to AQK27. With increasing temperature, both sequences undergo aggregation more rapidly and form larger aggregates in shorter time frames compared to incubation at lower temperatures. At 60 °C both sequences begin to aggregate within 8 hours of incubation, and at 80 °C the rate of aggregation is further increased, with pronounced increases in SMSLS intensity under 4 hours. A distinction is made between these two sequences at both temperatures as AQK35, the longer sequence, begins to aggregate within 1 and 2 hours at 80 °C and 60 °C, respectively. While a similar trend for aggregation is observed for AQK18, it begins detectable aggregation after 2 and 4 hours, at 80 °C and 60 °C, respectively. A further distinction is observed, as AQK35 forms larger aggregates more rapidly as compared to AQK18 at 60 °C and 80 °C as the observed intensity increases beyond expected values for small oligomers.[36] Though, in the case of AQK18 at 80 °C, the results might suggest that similar sized aggregates as compared to AQK35 could be achieved with continued incubation. The similar trends in the aggregation behavior for the AQK18 and AQK35 may be attributed to similar design in sequence. AQK35 is effectively two repeats of the AQK18 sequence separated by a central lysine residue. Consequently, they each have equivalent alanine-rich segments (four AAAQ repeats), with AQK35 having two of these stretches separated by the additional lysine, which likely increases the propensity for AQK35 to aggregate. As previously mentioned, AQK27 (three AAAQ repeats) remained monomeric at the two lowest temperature conditions. Aggregation for AQK27 at 60 °C and 80 °C occurs at similar times to that of AQK18 at ~2 and 4 hours, respectively. In contrast to the AQK18 and AQK35 solutions, the multimers formed in the AQK27 solutions do not immediately continue to aggregate and grow. Further aggregation is only observed upon continued incubation of up to 24 or 48 hours for 60 °C or 80 °C, respectively. The lower propensity for aggregation of the AQK27 may be the result of the shorter alanine-glutamine stretches (lower hydrophobicity) between the charged lysine residues as compared to AQK18 and AQK35. Additionally, the shorter lysine-lysine distance might energetically penalize the close contact of two or more AQK27 polypeptides to form stable aggregates. Therefore, although smaller oligomers are observed early on, the energetic penalty arising from charge repulsions presumably does not allow for aggregation to proceed readily. In general, these results agree with a high concentration polyalanine system incubated at elevated temperature. In their study, Blondelle et al observed an increase in β-sheet complexation with increasing peptide concentration (from ~1.4 – 4.2 mg/ml) at elevated temperature (65 °C). Additionally, they observed a correlation between polyalanine block length and β-sheet complexation. A series of polyalanine peptides were synthesized with the alanine stretch, n, ranging from 3 to 25 (n = 3 – 25). For n ≤ 9, no β-sheet complexation was observed, and peptides remained monomeric. For 10 ≤ n ≤ 14, a mixture of β-sheet complex and β-helical conformation was observed; for sequences n ≥ 15 interconversion to β-sheet complex from monomer was complete.[22] This effect of polyalanine block length is also consistent with our observation in the differences in aggregation behavior between AQK27 (three AAAQ repeats) and AQK18 and AQK35 (four AAAQ repeats). While the alanine-glutamine stretch in AQK27 is 12 residues long, as previously mentioned, the inclusion of the glutamine residues provides more hydrophilicity as compared to a straight chain of alanine residues. Therefore, the lower propensity for aggregation may also result from the shorter stretches, analogous to the polyalanine samples.

Further characterization of AQK18, AQK27 and AQK35 and their thermal aggregation was conducted via DLS. Samples were prepared at 4 mg/ml in pH 7.4 buffer and were incubated at 25 °C, 40 °C and 60 °C for 72 hours and aliquots were measured offline by DLS at 25 °C. Samples were measured at a scattering angle of 173° using a low volume quartz cuvette. Instrumentation did not allow to measure at multiple angles. Normalized autocorrelation functions of each peptide at 0, 24 and 72 hour incubation are presented in Figures 2–4 (data for incubation at 40 °C can be found in SI, SI Figure 1). These conditions were chosen based on the results observed via SMSLS. The normalized autocorrelation function for pre-incubated samples of AQK18 displayed a bimodal decay (Figure 2A). The earlier decay time (10 μs) is indicative of the monomeric peptide population in solution, in agreement with the results found via SMSLS. The decay at longer times may be the result of a polydisperse population of larger species. As DLS intensity increases nonlinearly with the diameter of the measured particles, a small molar concentration of large-size species will induce a bias on the measured DLS signal, leading to a pronounced second decay. The contribution of this second decay to the autocorrelation function decreases after 24 hours, which, in combination with the SMSLS results, suggests that the lyophilized powder might not fully dissolve in buffer initially, leading to the presence of undissolved polypeptide. Our SMSLS data under equivalent conditions did not show a rapid increase and then decrease in the normalized scattering signal, but an overall reduction in noise over time. Given the expected size of this measured particles (< 1 um) and the size of the materials used, non-specific adsorption or further aggregation in the form of large, insoluble particles that would precipitate over time is unlikely to be not present. Instead, these large species of “undissolved” or aggregated AQK18 might dissolve over time within the buffer increasing the weight percent of monomer in solution and decreasing the weight percent of un-dissolved or aggregated peptide. Similar results were observed for AQK27: a bimodal decay was observed in pre-incubated samples, with a decrease in the presence of larger particles after 24 hours (Figure 3A). Conversely, the results for AQK35 indicate that full monomer recovery (or peptide dissolution) is not observed (Figure 4A), suggesting the peptide particles/aggregates do not redissolve under these experimental conditions, and that polypeptide aggregates are more stable and likely to grow with the AQK35 sequence in comparison to the other 2 sequences. A similar behavior was observed by Bernecki et al., as multiple variations of their polyalanine sequences (alanine repeats of 7, 13 and 19) formed small oligomers upon dissolution into buffer.[25] Nevertheless, we do observe a more well-defined decay after 72 hours, demonstrating a decrease in large particle content. Samples incubated at 40 °C show an increased presence of monomer after 72 hours along with larger species. Again, we attribute this to a diverse population of particles which has not fully dissolved (SI, SI Figure 2), the larger species of which bias the recorded DLS signal. Ultimately, in comparing these results to that of the SMSLS, we observe similar trends for these peptides: samples remain predominantly monomeric for up to 72 hours at 25 °C and 40 °C (although AQK18 does aggregate at 72 hours at 40 °C as noted above).

Figure 2.

DLS results as a function of isothermal-incubation temperature and time for the AQK18 sequence at 25 °C (panel A), and 60 °C (panel B) for 0 hours(black line), 24 hours (gray line) and 72 hours (red line). Error bars represent the minimum and maximum values from independent replicas.

Figure 4.

DLS results as a function of isothermal-incubation temperature and time for the AQK35 sequence at 25 °C (panel A), and 60 °C (panel B) for 0 hours (black line), 24 hours (gray line) and 72 hours (red line). Error bars represent the minimum and maximum values from independent replicas.

Figure 3.

DLS results as a function of isothermal-incubation temperature and time for the AQK27 sequence at 25 °C (panel A), and 60 °C (panel B) for 0 hours(black line), 24 hours (gray line) and 72 hours (red line). Error bars represent the minimum and maximum values from independent replicas.

Upon incubation at 60 °C, it is apparent there is a shift to greater autocorrelation times for all three peptides, indicative of the formation of aggregates (Figures 2B–4B). After 24 hours, AQK18 presents a bimodal decay where both decays are shifted to greater autocorrelation times, and after 72 hours there is a well-defined decay which has increased again. These results show that AQK18 is aggregating within 72 hours, in agreement with the SMSLS results discussed above. Interestingly, incubation of AQK27 demonstrates higher monomer content in solution after 24 hours in comparison to the pre-incubated sample, showing that AQK27 does not aggregate within 24 hours even at the elevated temperature. Additionally, formation of aggregates in AQK27 after 72 hours is observed, as the early decay time has shifted to an increased decay time. Both results are in agreement with the SMSLS results. The normalized autocorrelation function for AQK35 at 24 hours shows multiple decays. These multimodal decays shift towards longer autocorrelation times upon incubation, suggesting the presence of aggregates within the solution. However, this result may also be due to the large population of particles present and DLS system limitations, as DLS does not capture size distributions on a number/concentration basis. As such, these DLS data are open to interpretation. When compared to the results of the SMSLS, the results are consistent with aggregation occurring over the first 24 hours of incubation. However, the population(s) suggested by the normalized DLS autocorrelation curve may not be an accurate distribution. Like AQK18 and AQK27, a well-defined autocorrelation curve was also observed for AQK35 after 72 hours, with the autocorrelation decay occurring at longer autocorrelation times, as compared to the non-incubated sample. Additionally, we note that the large error bars in these measurements may be indicative of unstable aggregates within solution. Taken together with the other characterization techniques used within this study, we argue that there is an on-going aggregation phenomenon that leads to a wide population of aggregate sizes as intermediate species. These results for AQK18, AQK27 and AQK35 all show that aggregation has occurred, and these observations will be further discussed in the context of the TEM results below.

To explore whether incubated samples recovered their initial α-helical conformation, or underwent a change in conformation, incubated peptide solutions were also characterized via CD at 0 °C (Figure 5). While these measurements by CD do not quantitatively characterize the aggregation of peptides, from a qualitative standpoint a lack of recovery of original conformation can be used to assess any irreversible conformation changes that occurred. Therefore, these data provide insight into the aggregation of these peptides when coupled with SMSLS and DLS. Solutions of each peptide were prepared and incubated in the same manner to those prepared for SMSLS and DLS. Pre-incubated solutions all showed α -helical conformation as previously reported.[18] Furthermore, all peptide samples incubated at 25 °C remained α-helical after 72 hours. Comparison of [Θ]MRE₂₂₂ between pre-incubated samples and incubated samples indicated nearly complete recovery of the conformation upon cooling to 0 °C. Along with the SMSLS and DLS results, the data show that these peptide solutions are stable at room temperature for up to 72 hours, with no detectable conformational changes. As previously mentioned, Blondelle et al. observed conformational stability for up to 200 hours of incubation for their polyalanine sequence under conditions which did not induce aggregation.[22] For shorter sequences studied by Bernacki et al., all were found to possess their original α -helical conformation for up to 14 days at room temperature.[25] In another study, one shorter polyalanine peptide did not undergo aggregation, which was attributed to it having α-helical conformation which was not likely to form oligomers.[24] Therefore, as AQK18, AQK27 and AQK35 remain helical at 25 °C, they are potentially less prone to aggregation as compared to their unfolded counterparts at elevated temperature and thus remain monomeric in solution.

Figure 5.

CD results of isothermal-incubation at 25 ºC and subsequent cooling for AQK18 (A), AQK27 (B), and AQK35 (C). For each peptide, solutions were incubated at 0-hour (black), 24-hour (red), and 72-hours (blue) and cooled to 0 ºC for characterization.

In contrast to samples incubated at ambient temperature, AQK18 and AQK35 incubated at 60 °C were not able to recover their original α-helicity after 72 hours of incubation (Figure 6). Additionally, the spectra for AQK18 and AQK35 resembled that of a mixture of α -helix and β-sheet CD signatures. The presence of β -sheet conformation was not observed for AQK27, but further loss of α -helical structure was, indicating a loss of soluble monomer (Figure 6). The presence of β -sheet conformation is not unexpected, given the similarity of these peptide sequences to those of alanine-rich polypeptides previously designed by our lab which underwent a β -sheet conformation transition with incubation at high temperature and low pH.[32] Furthermore, Blondelle and co-workers also observed β -sheet complexation with their polyalanine peptides at high temperature and concentration.[22] Additionally, just as was observed in the SMSLS results, these results also show a similar trend when comparing AQK18 and AQK35, as both show the presence of β -sheet conformation, which, as previously discussed, may be the result of increased hydrophobicity of the peptide due to the longer alanine-glutamine segments that each sequence possesses (16 residues). In contrast, AQK27, with shorter alanine-glutamine stretches (12 residues), does not aggregate as rapidly and does not undergo this conformational transition, consistent with the previous results from Blondelle et al. (discussed above) that indicate that long alanine-stretches (n ≥ 15) undergo aggregation, while shorter alanine-stretches (n ≤ 8) do not.[22] Although our reported sequences are not as hydrophobic as polyalanine sequences, relative to one another, the shorter alanine-glutamine stretches for AQK27 make it less hydrophobic compared to AQK18 and AQK35, and therefore it does not undergo aggregation as rapidly, nor is a conformation transition observed.

Figure 6.

CD results of isothermal-incubation at 60 ºC and subsequent cooling for AQK18 (A), AQK27 (B), and AQK35 (C). For each peptide, solutions were incubated at 0-hour (black), 24-hour (red), and 72-hours (blue) and cooled to 0 ºC for characterization.

The morphology of the aggregates formed after incubation at 60 °C was assessed by TEM; representative images are presented in Figures 7 and 8 (24 hour and 72 hour). The images were acquired after 24 and 72-hour incubation; samples were cooled to room temperature showing what remained after cooling. For all peptide samples, incubation for 24 hours yielded a mixture of small, spherical oligomers and fibrils (Figure 7). For all sequences after 72 hours, the aggregates predominantly appear as long fibrils (Figure 8). No fibrils or other aggregates were observed for samples incubated at 25 °C, in agreement with the SMSLS and DLS data. These fibrillization results agree with our previous study of alanine-rich polypeptides, which were found to undergo fibrillization after incubation at high temperature and low pH,[32] and with the previous work reported by Blondelle et al.[22] Additionally, Bernacki observed that initial spherical oligomers formed by two peptide sequences further associated into fibril aggregates with increased incubation time.[25] This is similar to the observations for the AQK series of peptides, which suggests that in the reported sequences, small, spherical oligomers are likely a precursor to the fibrillar aggregates. In another study, coarse-grained modeling was used to study the effects of hydrophobicity and temperature on the aggregation process and self-assembly of peptides with non-polar side chains.[37] Interestingly, they found that increased strength of hydrophobic interactions led to the formation of amorphous β-sheet aggregates, while very high temperatures produced disordered random coils independent of hydrophobic interactions. Only at intermediate hydrophobic interactions and intermediate temperatures were fibrils readily formed. These results may explain why AQK18 and AQK35 (more hydrophobic sequences) show the presence of more spherical aggregates as compared to AQK27 (Figure 7). We speculate, that the more hydrophobic sequences form more initial aggregates which form into fibrils with time. In contrast, AQK27 (less hydrophobic), is more easily able to form fibrils at earlier time points. The results presented in this study, as well as those found by Mu et al. are also in agreement with those of Bernacki et al., as their most hydrophobic sequence only formed amorphous aggregates at increased temperature, while a less hydrophobic sequence did indeed form fibrils. Though our sequences are not completely analogous to those studied by Mu et al. (all have polar glutamine side chains), we hypothesize that the intermediate hydrophobicity of our sequences, as well as incubation at 60 °C are consistent with expectation based on the predicted conditions from previous work under which fibrils are suggested to form.

Figure 7.

TEM micrographs of (A) AQK18, (B) AQK27 and (C) AQK35 after 24-hour incubation at 60 °C. All samples were stained with PTA. Scale bars are 100 nm.

Figure 8.

TEM micrographs of (A) AQK18, (B) AQK27 and (C) AQK35 after 72-hour incubation at 60 °C. AQK18 and AQK27 were both stained with PTA; AQK35 was unstained. Scale bars are 200 nm.

The observation of fibrils is consistent with the observed autocorrelation functions in our DLS studies presented above and may explain the well-defined autocorrelation function that is observed for all samples. Furthermore, the shift in the autocorrelation function to longer times after 24 hours may result from the presence of large aggregates or “clustering” of fibrils in solution, representing the large species detected by the DLS. Similar trends in the collected autocorrelation functions have been observed in other studies using DLS and microscopy to characterize protein/peptide fibrillization.[38–40] Autocorrelation functions were recorded for β-lactoglobulin protein at various heating times, and aliquots of these incubated samples were characterized by single molecule atomic force microscopy (AFM). The initial autocorrelation function of the protein decayed rapidly, indicative of a monomeric species. Similar to our peptides, with increased incubation time, the autocorrelation functions continued to shift to longer decay times, with some multimodal decays, indicating the formation of aggregates. Comparison of the autocorrelation functions to aliquots characterized by AFM showed protein monomer forming protofibrils, which finally matured into amyloid-like fibrils with increased incubation time.[40] In addition, Hill et al. characterized the fibrillization of hen egg-white lysosome (HEWL) via DLS and AFM. Similar results in the temporal evolution of correlation functions to that of Bolisetty et al. were observed. AFM characterization of aliquots showed the pathway from protein monomer, to oligomers which formed protofibrils, eventually self-assembling into mature fibrils.[39] Finally, DLS and TEM were used to characterize fibers formed by shorter peptides whose assembly was modulated by side-chain charges and pH of solution. The hydrodynamic radius of the fibrils calculated from the collected autocorrelation functions were in fair agreement with the diameters calculated by TEM.[38] The gradual evolution of autocorrelation functions over time observed in this present work are indeed in good agreement with these previous studies which have characterized fibril formation, with intermediate steps of non-fibril oligomeric aggregates as potential seeds for fibril nucleation.

For peptide systems that form fibrils, it is typical to observe a coinciding β-sheet conformation as well.[11,41,42] For AQK18 and AQK35, we see a mixture of α-helix and β-sheet signatures after 72 hours. For AQK27, we only observe a loss in helical content, while no β-sheet conformation is observed. Therefore, we hypothesize that the fibrils in solution are a small percentage (by mass) compared to the remaining monomeric peptide, which can refold to α-helical conformation upon cooling. This hypothesis is supported by the results observed by Blondelle, who showed that the conversion to β-sheet complexation was dependent on peptide hydrophobicity. Several moderately hydrophobic sequences (10 ≤ n ≤ 14) characterized by CD presented a coexistence of β-sheet and α-helical conformations after 24 hours of incubation at 65 °C. Furthermore, they did observe the presence of fibrils for a sequence of n = 13. At room temperature, Bernacki observed that their moderately hydrophobic sequences did undergo aggregation, however they found these solutions to remain 90–95% monomeric, and partially α-helical.[22] While the fibrils in this present study are formed at elevated temperature, it is possible that a similar solution condition is present: only a portion of the peptide monomer is aggregating into fibrils, while the rest remain unfolded, but monomeric. When cooled, the fibril formation is irreversible, but the unfolded peptide refolds into α-helix. Coupled with these observations, as well as the mix of conformations observed for sequences with longer alanine-glutamine stretches (AQK18 and AQK35), it is reasonable to conclude that the fibrils are indeed β-sheet stabilized.

Harnessing this conformational transition, and the association into fibrils may enable the production of useful hybrid materials. As previously discussed, many alanine-rich peptides have been shown to also associate into fibrils. In addition to this, peptides which associate into amyloid fibrils are also well studied and could be employed for the generation of unique materials. In particular, amyloid nanotubes are of particular interest for their ability to be modified, uniform dimensions and hollow architecture.[43] Zhao et al. has reported on an amphiphilic peptide (KI₄K) which is able to assemble into nanotubes with diameters of 80–160 nm.[44] Additionally, they found that the assembly could be tuned to form nanofibrils upon addition of acetonitrile. Lin and coworkers developed an amphiphilic nanotube system comprised of two oppositely charged drug-peptides. The peptide sequence GNNQQNY from yeast prion Sup35 was functionalized with camptothecin (CPT). The nanotubes had a diameter of about 123 nm and a 36% loading of CPT which could be used in drug delivery systems.[45] In addition to the potential uses of fibrils, the incorporation of PEG onto associating peptides has also enabled the generation of useful materials; it provides stability, solubility and increased circulation time.[1] Therefore, the conjugation of PEG to these fibril-forming, alanine-rich, peptides studied here is of interest for future work.

Conclusion

A series of new alanine-rich peptides have been synthesized and the thermodynamic unfolding and aggregation of a series of alanine-rich peptides has been characterized and described here. Light scattering techniques have characterized the aggregation of the peptides as a function of temperature. All sequences remain monomeric at 25 °C for up to 72 hours, while increased temperature induces aggregation within 2–4 hours. The rate of aggregation may also be tuned with increasing temperature, as aggregation was observed within an hour at 80 °C for AQK35. The increased length in alanine-glutamine stretches (increased hydrophobicity) also promoted aggregation at earlier times. Circular dichroism measurements revealed these sequences recover their original helicity after 72-hour incubation at 25 °C. However, when incubated at 60 °C, a loss of helical content (AQK27), and the presence of β-sheet conformation for AQK18 and AQK35 is present after 72 hours. TEM characterization revealed the presence of fibrils for samples incubated at 60 °C, however we postulate that these aggregates are a relatively small population of the peptide solution as compared to the monomeric peptide or non-fibril aggregates. In general, these alanine-rich peptides are of interest for further study as they provide soluble, monomeric peptides, whose aggregation into fibrils can be triggered by temperature. Their application in the generation of functional materials, or a platform to study protein aggregation will be explored in future studies.

HIGHLIGHTS:

• Aggregates of alanine-rich, helical peptides become monomers over time at ambient temperature

• Conversion of aggregates to fibrillar structures requires loss of α-helical structure

• A minimum uninterrupted sequence of (AAAQ)₄ is required for β-sheet and fibril formation

• Conformational and morphological changes are correlated with observed aggregation in light scattering