Master and Slave Relationship Between Two Types of Self

Nov 5, 2014 - Cross-seeding of fibrils of bovine insulin (BI) and LysB31-ArgB32 human insulin analog (KR) induces self-propagating amyloid variants wi...
6 downloads 7 Views 7MB Size
Article pubs.acs.org/JPCB

Master and Slave Relationship Between Two Types of SelfPropagating Insulin Amyloid Fibrils Weronika Surmacz-Chwedoruk,†,§ Viktoria Babenko,†,‡ and Wojciech Dzwolak*,†,‡ †

Institute of High Pressure Physics, Polish Academy of Sciences, Sokolowska 29/37, 01-142 Warsaw, Poland Department of Chemistry, Biological and Chemical Research Centre, University of Warsaw Ż wirki i Wigury 101, 02-089 Warsaw, Poland § Institute of Biotechnology and Antibiotics, Staroscinska 5, 02-516 Warsaw, Poland ‡

S Supporting Information *

ABSTRACT: Cross-seeding of fibrils of bovine insulin (BI) and LysB31-ArgB32 human insulin analog (KR) induces self-propagating amyloid variants with infrared features inherited from mother seeds. Here we report that when native insulin (BI or KR) is simultaneously seeded with mixture of equal amounts of both templates (i.e., of separately grown fibrils of BI and KR), the phenotype of resulting daughter fibrils is as in the case of the purely homologous seeding: heterologous cotemplates accelerate the fibrillation but do not determine infrared traits of the daughter amyloid. This implies that fibrillationpromoting and structure-imprinting properties of heterologous seeds become uncoupled in the presence of homologous seeds. We argue that explanation of such behavior requires that insulin molecules partly transformed through interactions with heterologous fibrils are subsequently recruited by homologous seeds. The selection bias toward homologous daughter amyloid is exceptional: more than 200-fold excess of heterologous seed is required to imprint its structural phenotype upon mixed seeding. Our study captures a snapshot of elusive docking interactions in statu nascendi of elongation of amyloid fibril and suggests that different types of seeds may collaborate in sequential processing of soluble protein into fibrils.



INTRODUCTION Growth of an amyloid fibril was conceptualized as “onedimensional crystallization” by Jarrett and Lansbury over two decades ago.1 Recent years have witnessed remarkable progress in understanding of the molecular and physicochemical basis of amyloidogenesis and of its biological context−including the pathogenic and biologically functional roles of amyloid fibrils.2,3 The amyloidogenic self-assembly is generally thought of as a thermodynamically downhill process,4−6 one aspect of which is its autocatalytic character. The one-dimensional crystallization concept adapts well to following findings such as the thermodynamic metastability of the native state relative to amyloid, and the self-propagating polymorphism of fibrils. The latter phenomenon consists in [i] existence of multiple kinetically stable structural variants of fibrils assembled from covalently identical polypeptide chains, and [ii] self-propagation of these variants (also termed “strains”) upon seeding− regardless of environmental biases favoring alternative modes of the amyloidogenic self-assembly. Such a conformation-imprinting effect has been observed in proliferation of amyloid fibrils from insulin,7,8 glucagon,9,10 Aβ-peptide,11,12 K3 fragment of β2-microglobulin,13 and parallels the proliferation scenario of prion strains (e.g., in yeast14,15). In contrast to proper threedimensional crystallization of globular proteins, conformations of soluble amyloidogenic precursors and insoluble amyloid fibrils are very different. The coupling of phase- and conformational transitions taking place upon amyloidogenesis results in a © XXXX American Chemical Society

dynamic complexity that is inaccessible to high resolution biophysical tools. Moreover, secondary nucleation events, fragmentation, and branching often accompany elongation of filaments which strongly complicates computational models of amyloid growth (excellently reviewed in16). As a consequence, the molecular-scale events accompanying integration of monomers with amyloid tips remain poorly understood. In one of models of amyloidogenesis, it has been proposed that amyloid elongation encompasses two distinct phases: docking followed by locking17 and this claim has been supported by in silico studies.18,19 Of these two, the locking phase is envisioned as markedly less reversible. On the other hand, reversible binding of monomers to amyloid tips has been observed for several proteins (e.g.,20−23) and was even exploited to assess relative levels of thermodynamic stability of amyloid variants.24 We have shown previously that a pair of amyloidogenic polypeptides: bovine insulin (BI) and LysB31-ArgB32 human insulin analog (KR), (Supporting Information, Figure S1), is an excellent model system to study self-propagating polymorphism of amyloid fibrils.25,26 The two types of insulin form fibrils with distinct infrared features allowing one to use FT-IR spectroscopy as a high throughput means to track propagation of the strains.27 When these fibrils are used as templates for crossseeding, the shape, absorption maximum and fine fingerprint Received: November 2, 2014

A

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Kinetic FT-IR measurements. To freshly prepared 1 wt % BI (or KR) in 0.1 M NaCl in D2O pD 1.9 proper amount of sonicated mother fibrils was added prior to measurements. The type of seed used was either [BI] or [KR] or their mixtures containing 5% of homologous and 95% of heterologous fibrils, as indicated in figure captions. The mass ratio of total amyloid: native protein at the initiation of seeding was either 1:100 or 1:2000. Briefly vortexed samples were swiftly transferred to a CaF2 transmission cell equipped with a 0.05 mm Teflon spacer. During 26 h-long measurements, the temperature in the cell (37 °C) was maintained using dedicated Peltier system while the sample chamber was continuously purged with CO2-free dry air. All time-lapse FT-IR spectra were collected on a Nicolet iS50 FT-IR spectrometer from Thermo. For a single spectrum 32 interferograms of 2 cm−1 resolution were coadded. Timedependencies depicting progress of α/β-transition accompanying insulin fibrillation plotted in Figure 4 correspond to relative transient spectral intensities at wavenumbers corresponding to the β-sheet-assigned components of the amide I′ band: 1622 cm−1 for fibrillation of KR, 1627 cm−1 for fibrillation of BI induced by mixed seeds, and 1628 cm−1 for homologous seeding of BI. In each case, the plots were calculated as degrees of α/β-transition (t) = (Io − I(t))/(Io − IF) where “Io” is the spectral intensity at such wavenumber in the first spectrum (corresponding to practically native insulin), “IF” is the corresponding intensity in the spectrum of completely aggregated insulin, and “I(t)” is transient spectral intensity at given time. Other details were the same as for acquisition of static FT-IR spectra. AFM. Samples of daughter fibrils for AFM imaging were prepared by seeding native insulin of either type (1 wt % in 0.1 M NaCl, pD 1.9) with sonicated mother fibrils (homologous or mixed) at the 1:2 amyloid:native protein mass ratio and subsequent incubation at 37 °C for 24 h. Collected samples of aggregates were diluted 60-times with deionized water. A small droplet (8 μL) of fibrils’ suspension was swiftly deposited onto freshly cleaved mica and left to dry overnight. AFM tappingmode measurements were carried out using Nanoscope III atomic force microscope (Veeco, USA) and TAP300-Al sensors, res. frequency 300 kHz (BudgetSensors, Bulgaria).

traits in the amide I′ band region are passed from mother seeds to daughter fibrils with high fidelity.25 Here, we report a counterintuitive outcome of seeding experiments in which insulin fibrillation is triggered by a mixture of preformed [BI] and [KR] fibrils. Instead of anticipated competition in recruiting soluble protein collaboration between the two structural variants has been observed: the heterologous seeds accelerate proliferation of the homologous phenotypes. The data has been discussed in the context of uncoupling of catalytic and structure-imprinting properties of elongating amyloid, and relative free energy levels of different structural variants of fibrils.



MATERIALS AND METHODS BI (Bovine Insulin) was purchased from Sigma (USA). KR (Recombinant LysB31-ArgB32 Human Insulin Analog) was manufactured by Institute of Biotechnology and Antibiotics (Warsaw, Poland) using recombinant DNA technology.28 D2O (“99.8 atom % D” grade) was from ARMAR Chemicals (Switzerland), and deuterium chloride (35 wt % DCl solution in D2O, 99 atom % D) was from Sigma (USA). Mother amyloid seeds were obtained through 65 °C/24 h quiescent incubation of 1 wt % insulin solutions in 0.1 M NaCl in D2O, pD 1.9 (where “pD” is pH-meter readout uncorrected for isotopic effects) as described earlier.25 Prior to seeding, mother fibrils were subjected to pulsed sonication (60 s in total) using Ultrasonic Processor VC130PB from Sonics & Materials, Inc. (USA) operating at 20 kHz frequency and 20% of nominal power of 130 W. The sonication was carried out in intervals to ensure that amyloid samples are not overheated. For seeding experiments, sonicated mother fibrils of [BI] and [KR] were mixed at desired ratios and subsequently added to freshly prepared 1 wt % insulin solutions in 0.1 M NaCl in D2O, pD 1.9 at specified total insulin fibrils: native insulin mass ratios (typically 1:100). In order to prevent competitive de novo nucleation, the temperature of the following incubation (carried out using Eppendorf Thermomixer Comfort accessory) was set at 37 °C. FT-IR measurements of ex-situ prepared samples. For FT-IR measurements carried out typically at 25 or 37 °C, a CaF2 transmission cell equipped with a 0.05 mm Teflon spacer was used. Temperature in the cell was controlled through an external water-circuit connected to a programmable thermostat. FT-IR spectra were collected on a Nicolet NEXUS FT-IR spectrometer equipped with a liquid nitrogen-cooled MCT detector. Typically, for a single spectrum 256 interferograms of 2 cm−1 resolution were coadded. During measurements the sample chamber was continuously purged with dry CO2depleted air. All insulin spectra were corrected by subtracting the proper amount of D2O and water vapor spectra prior to being baseline-adjusted. Data processing was performed using GRAMS software (Thermo, USA). The relative intensities at 1728 cm−1 plotted in panels C and D of Figure 3 were normalized vs the amide I′ intensity at maximum. The relative values were obtained as percent of the 1728 cm−1 band’s normalized intensity for homologously seeded [KR]KR amyloid. For [KR]KR and [BI]KR, the ratio of integral intensities of 1728 cm−1 band to the amide I′ band was 0.44% ± 0.02%, while for intensities measured at the spectra maxima, the ratio was 2.4% ± 0.1%. FT-IR measurements of all samples, including those collected from microplate reader after the kinetic scans, were carried out immediately after the specified incubation times. All further details have been described earlier.25



RESULTS AND DISCUSSION Infrared absorption spectra capture phenotypic differences between [BI] and [KR] insulin amyloid strains (Figure 1). The conformation-sensitive amide I′ band of spontaneously formed [BI] amyloid and of its homologously seeded daughter fibrils is centered at 1627 cm−1, whereas for [KR] absorption maximum is shifted to 1621 cm−1 with smaller peaks at 1652/1651 and 1728 cm−1 appearing as well.25 These spectral features have been attributed to subtle conformational variations distinguishing stacked β-sheets of the two strains, and are passed to daughter insulin fibrils either upon homologous or heterologous seeding (cross-seeding).25,26 The infrared spectra of insulin fibrils obtained through the two seeding scenarios (in black and blue in Figure 1) show that the type of mother template controls the conformation of daughter amyloid. The overlaid red spectra correspond to daughter fibrils composed of either BI or KR molecules (panels A and B, respectively), but induced through simultaneous seeding with mixed equal portions of [BI] and [KR] templates. In both cases, the spectra are practically identical with those of homologously seeded fibrils, i.e. no memory effect of the foreign template can be detected. Because this preliminary FT-IR data appears to B

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Results of the following infrared experiments designed to assess strength of the bias toward formation of homologous daughter fibrils are presented in Figure 3. Different mass ratios of [BI] and [KR] templates were used while the ratio of total amount of fibrils to native insulin to was maintained at 1:100. The stacked spectra of fibrils obtained from BI and KR insulins (panels A and B, respectively) retain a very high similarity to homologous mother seeds despite the dramatically decreasing ratios of homologous templates. The amide I′ band begins to depart from the mother seed’s infrared characteristics only when the percentage of heterologous templates approaches 90% for the seeding of BI, and 99% for the seeding of KR. The homologous seeds need to be even more diluted with foreign template to permit the latter one imprint structure through the memory effect. For seeding of BI, the threshold concentration of [KR] amyloid in seeds exceeds 99,9%, and for the seeding of KR this value appears to be even higher. The amyloid samples were sonicated briefly prior to infrared measurements in order to improve quality of spectra;25 however, exactly the same tendencies were observed for nonsonicated samples (Supporting Information, Figure S2). The unexpected perseverance of homologous phenotypes upon seeding with the mixed templates is illustrated by plots of the amide I′ band position and relative intensity of 1728 cm−1 peak (characteristic only for the [KR] phenotype) in spectra of daughter fibrils induced by varying ratios of [BI] and [KR] seeds (panels C and D of Figure 3). A simple conventional explanation of the reported experimental outcome would focus first on possible kinetic preferences for homologous seeds (e.g., higher elongation rates) or different fragmentation susceptibilities. However, we have shown previously that kinetics of homologous and heterologous seeding of BI and KR insulins are comparable25 (see also Supporting Information, Figures S3 and S4). In fact, the following experiment has shown that the heterologous component of the mixed seeds remains catalytically active and contributes strongly to the overall rate of the native-to-amyloid conversion, although with no impact on the daughter fibrils’ phenotype. Figure 4 depicts kinetics of insulin fibrillation induced by homologous and mixed seeds probed by FT-IR spectroscopy. Homologous seeds were used either at typical ratio to the native insulin (1:100) or diluted to one two-thousandth of the native insulin mass or mixed at such high dilution with 19-fold excess of heterologous templates. The data in Figure 4 correspond to time-lapse FT-IR measurements of insulin samples undergoing aggregation in situ in spectroscopic cell thermostated at 37 °C over the period of 26 h. The stacked infrared spectra in Figure 4A correspond to BI (top) or KR (below) insulin. The addition of homologous seeds at the 1:100 mass ratio to either type of native peptide triggers gradual spectral changes proceeding through single isosbestic points reflecting transitions to amyloid of the seed’s phenotype (panels labeled as [BI]100%BI and [KR]100%KR). According to the normalized time-dependencies of the β-sheet-assigned spectral components of the amide I′ band, fibrillation under these conditions is fastest out of all the considered scenarios (red points in panels B and C). When the mass ratio of homologous seeds to native insulin is reduced to 1:2000 (panels labeled as [BI]5%BI and [KR]5%KR) aggregation is markedly decelerated and is completed within the duration of experiment only in the case of KR.

Figure 1. Amide I/I′ infrared band region of [BI] (A) and [KR] (B) insulin fibrils obtained through 48h/37 °C incubations of respective 1 wt % native insulin solutions in 100 mM NaCl, D2O, pD 1.9 seeded with homologous mother fibrils (black), cross-seeded with the alternate type of fibrils (blue), or seeded with mixture of equal portions of [BI] and [KR] amyloid (red). Mass ratios of native insulin to total seed were fixed at 100:1.

suggest that heterologous seeds are not actively involved in cross-seeding in the presence of homologous seeds, we have attempted to capture such “arrested growth” effect using AFM. In Figure 2, AFM amplitude images of daughter fibrils induced by homologous- and mixed-seeding are presented. The seedings were carried out at untypical 1:2 ratio of total seed to native insulin which was designed to facilitate comparative statistical analysis of the elongation process. Under such conditions, an unperturbed growth should lead to, on average, extension of seed’s length by the factor of 3. On the other hand, arrested growth of a heterologous component of mixed fibrils should manifest through the presence of a fraction of seeds of the original averaged length. While the AFM data confirms amyloid character of daughter aggregates obtained through the different seeding scenarios, several factors preclude sound morphological analysis in the context of selectively arrested growth of heterologous seeds. First of all, [KR] and [BI] fibrils, while different in terms of infrared characteristics, are very similar in terms of diameter and surface morphological features (Figure 2) which is in unison with our previous study employing transmission electron microscopy25. Second, mother insulin amyloid seeds are pronouncedly polydispersed in terms of length even after extensive sonication. Finally, attempts at obtaining sound statistics of length distribution of single fibrillar specimen have proved practically unfeasible due to the strong tendency to aggregate which is visible in the AFM data and have been reported earlier.25 C

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 2. Amplitude AFM images of daughter [BI] (top) and [KR] (bottom) fibrils induced by homologous seeding (left) and seeding with both [BI] and [KR] templates mixed at the 1:1 ratio. The mass ratio of total amyloid seeds: native insulin was 1:2. Overlaid are cross sections of representative specimen (according to corresponding AFM height images).

The key result is the fibrillation kinetics upon seeding with mixed homologous and heterologous seeds (mass ratio to native insulin equal to 1:2000 for the homologous and 19:2000 for the heterologous component seed - panels labeled as [BI] 5%BI+95%KR and [KR]5%KR+95%BI ). The spectra show pronounced acceleration of BI fibrillation compared to the [BI]5%BI experiment yet, they clearly lead to the [BI] type of amyloid. The fibrillation is also accelerated for the seeding of KR, also without detectable changes in daughter fibrils phenotype. The control spectra in far-right panels show lack of aggregation in unseeded samples within the 26 h of duration of the experiment. The data are in very good agreement with parallel experiment in which Thioflavin T fluorescence and plate reader were used to probe kinetics of seed-induced insulin aggregation while infrared spectra of daughter fibrils were collected afterward (Supporting Information Figure S5). Hypothetically, dramatic increases in susceptibilities to spontaneous fragmentation of [BI]BI and [KR]KR vs [BI]KR and [KR]BI could explain the dominance of homologous seeds upon the mixed seeding. However, this would necessarily result in [i] much higher rates of homologous seeding compared to heterologous seeding, and [ii] morphological evidence of the higher brittleness of homologously seeded daughter fibrils. Our previous study25 as well as the kinetic and AFM data presented in this work does not support such scenario. Moreover,

conventional explanations based on the assumption that monomers reversibly detached from fibrils (during the docking phase) return immediately to the “bulk” dissolved state (i.e., the native form in the case of insulin under the experimental conditions used in this study) appear inadequate in the context of the observed capacity of heterologous seeds to enhance aggregation rate without conformational imprinting. Namely, according to such scenario, presence of heterologous seeds would only reduce the transient pool of insulin molecules available for homologous seeds and therefore decelerate overall fibrillation. However, the data shown in Figure 4 contradict this possibility. The herein reported results seem to be also incompatible with the full kinetic control over the self-propagating polymorphism of amyloid fibrils, i.e., complete irreversibility of monomer-tip docking events (koff = 0). In such case, relative concentrations of heterologous templates in mixed seeds should be reflected through “mixed” phenotypes of the daughter fibrils. According to study by Shammas et al. insulin fibrillation at low pH can become reversible with increasing pH, or upon addition of denaturants.29 On the other hand, our control experiment shown in Figure 5 which was carried out under the conditions used in this study (low pH/absence of chemical denaturants) suggests that mixtures of mature insulin fibrils of different types do not reach the state of D

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 3. Infrared spectra of daughter [BI] (A) and [KR] (B) amyloid obtained through seeding native insulin with mixed fibrils at different [BI]/ [KR] ratios. Percentages of homologous fibrils in seeds are indicated on the right side. The corresponding dependencies of amide I/I′ band position (black squares) and intensity of 1728 cm−1 peak (red triangles) on the fraction of homologous fibrils used for seeding BI and KR are plotted in panels (C) and (D), respectively. Region of extreme dilution of [KR] with [BI] upon seeding of KR (blue dots mark position of the lower intensity shoulder) is magnified in inset of panel D. Mass ratios of native insulin to total seed were fixed at 100:1.

in imprinting their conformational signatures in daughter amyloid generations while heterologous seeds are “enslaved”: they enhance the overall fibrillation rate but without much chance to propagate their phenotypes. Slightly different rates of elongation of [BI] and [KR] templates upon seeding and crossseeding (Supporting Information Figure S3, and ref 25), as well as possible differences in numbers of sonication-induced tips cannot account for the enormous and quasi-symmetric bias for homologous daughter phenotype regardless whether BI or KR insulin is seeded with the same mixture of templates (Figure 3). However, a modified dock-and-lock model17 could provide a phenomenological explanation for the data (Figure 6). First, we propose that the docking phase of the conformational memory effect responsible for the imprinting the heterologous seed’s structure in aggregating insulin (recapitulated in Figure 6A) occurs in quasi-reversible stages upon which intermediate monomer conformations are formed (I1···In)Figure 6B. Second, these early intermediates are structurally similar (“on pathway”) to intermediate states induced upon interaction of soluble insulin with homologous seeds while their free energy is lower during this interaction than energy of the native state. In the absence of homologous seeds, the relatively slow transition of soluble polypeptide into amyloid building block proceeds through intermediate states (in the docking phase) and subsequently enters the less reversible locking phase.17 Third, in the presence of both types of seeds, the intermediate states

thermodynamic equilibrium under the conditions, and on the time scales investigated in this study. That two different amyloid strains of a single insulin type have identical free energy levels (per monomer) is entirely unlikely. Therefore, in mixtures of amyloid variants composed of covalently identical insulin molecules (e.g., [BI]BI + [BI]KR, or [KR]KR + [KR]BI) and in dynamic equilibrium with soluble insulin, one should observe a slow transition toward the more stable strain. According to the data in Figure 5, the compositions of equimolar mixtures of sonicated homologous and heterologous daughter fibrils ([BI]BI + [BI]KR, or [KR]KR + [KR]BI), under the conditions favoring amyloid elongation, do not change over prolonged periods of time. This strongly suggests that, at least under such conditions, mixtures of separately grown fibrils of two different strains do not reach thermodynamic equilibrium within the experimental time scale. However, it must be stressed that the observed irreversibility of formation of mature [KR] and [BI] amyloid fibrils (under the conditions, and on the time scales of this study) does not rule out existence of transient reversible stages in elongation of these fibrils, as was proposed in the dock-and-lock model.17 According to the infrared data presented in this study the relationship between the two mixed types of insulin amyloid seeds reminds one of a “forced collaboration” rather than a strict competition in recruiting native insulin molecules. In other words, the homologous seeds (“masters”) are privileged E

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 4. (A) Time-lapse FT-IR spectra (amide I′ band range) of BI (top) and KR (bottom) insulins undergoing fibrillation induced by different types of seeds at 37 °C and over the period of 26 h. From the left: seeding with homologous templates at 100:1 native:amyloid mass ratio; seeding with diluted homologous seeds (effective seeding ratio 2000:1), seeding with mixed seeds at 100:1 (native:total amyloid) mass ratio where 95% of homologous seeds were replaced with heterologous templates. The far right column scorrespond to unseeded samples. The spectra were subsequently used for plotting time-dependent degrees of α/β-transition for seeding of BI (B) and KR (C).

[KR]KR + [KR]BI mixtures that would reflect slow transitions toward more stable strains. Our results shed light onto the elusive docking phase of elongation of fibrils where thermodynamic control plays a major role enabling elucidation of relative stabilities of amyloid variants. It should be stressed that this hypothetical model does not contain answers to many questions that must be asked: for example, whether intermediate I1···In states depart from heterologous seed’s surface and dockduring their lifetimesat the homologous seed’s tip. It is also unclear whether such intermediates would be mono- or oligomeric (in this respect, the helical entities described by Vestergaard et al.30

triggered by the heterologous amyloid tips (and then shed into the bulk solution) are recruited during their transient lifetimes by homologous seeds (Figure 6C). If the free energy level of homologous variant (i.e., spontaneously formed fibrils) is lower than of the heterologous one then this would provide driving force for the biased phenotype selection. Different free energy levels (per monomer) are expected for different amyloid strains which can remain kinetically stable due to huge barriers associated with locking of monomers within fibrils. This is clearly supported by the infrared data in Figure 5 showing no evidence of spectral drifts in coincubated [BI]BI + [BI]KR, or F

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B

Article

Figure 6. Hypothetical explanation of the strong selection bias for homologous phenotypes upon seeding with mixtures of two amyloid templates. (A) Both BI and KR insulins self-assemble into distinct fibrils depending on whether the amyloidogenesis is spontaneous or cross-seeded with an alternate template. (B) We propose that the docking and integration of insulin molecule with heterologous amyloid seed is a multistage process proceeding through intermediate states (In). (C) We further propose that free energy levels of these intermediate states are higher than of intermediate states induced upon docking at the tip of homologous amyloid seed (blue vs green trajectories). Hence homologous seeds may hijack intermediate conformations on the heterologous docking pathway (red line). Our model requires that preliminary interactions with heterologous seeds convert native insulin into amyloidogenic intermediate which is subsequently recruited by homologous seeds.

Figure 5. FT-IR spectra of mixtures of equal portions of separately grown, then sonicated daughter fibrils obtained through homologous and heterologous seeding (panel A, [BI]BI with [BI]KR, panel B, [KR]KR with [KR]BI) before (blue), and after (red) 1-week-long incubation at 37 °C/300 rpm, in 0.1 M NaCl D2O, pD 1.9.

could be candidates for the latter). Its main purpose is to provide a working model and to relate the observed phenomenon to the widely accepted dock-and-lock idea by Lee and Maggio.17 It should also be stressed that the preliminary interactions between dissolved insulin and heterologous seeds may be similar to those involved in secondary nucleation pathways which have been implicated in insulin amyloidogenesis.31,32 Further efforts will be needed to provide a deeper mechanistic explanation of the reported phenomenon. In summary, this study unveils nonadditive effects of mixed self-propagating amyloid variants used as seeds to induce insulin fibrillation. We have shown that the catalytic and structure-imprinting effects of heterologous seeds become uncoupled in the presence of homologous seeds. The observed effects could hypothetically be explained through hierarchical coprocessing of the soluble protein into amyloid in which different templates enhance fibrillation rate but only the homologous seeds determine the phenotype of daughter fibrils. Our findings suggest that seeding experiments employing mixed homologous and heterologous templates could illuminate obscure energy landscapes of amyloid fibrils.



ASSOCIATED CONTENT

S Supporting Information *

Additional FT-IR and kinetic data and primary structures of BI and KR. This material is available free of charge via the Internet at http://pubs.acs.org.



AUTHOR INFORMATION

Corresponding Author

*(W.D.) Telephone: +48 22 55 26 567. Fax: +48 22 822 5996. E-mail: [email protected]. Notes

The authors declare no competing financial interest. G

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry B



Article

(17) Esler, W. P.; Stimson, E. R.; Jennings, J. M.; Vinters, H. V.; Ghilardi, J. R.; Lee, J. P.; Mantyh, P. W.; Maggio, J. E. Alzheimer’s Disease Amyloid Propagation by a Template-Dependent Dock- Lock Mechanism. Biochemistry 2000, 39, 6288−6295. (18) Nguyen, P. H.; Li, M. S.; Stock, G.; Straub, J. E.; Thirumalai, D. Monomer Adds to Preformed Structured Oligomers of Aβ-Peptides by a Two-Stage Dock-Lock Mechanism. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 111−116. (19) Straub, J. E.; Thirumalai, D. Toward a Molecular Theory of Early and Late Events in Monomer to Amyloid Fibril Formation. Annu. Rev. Phys. Chem. 2011, 62, 437−463. (20) Carulla, N.; Caddy, G. L.; Hall, D. R.; Zurdo, J.; Gairí, M.; Feliz, M.; Giralt, E.; Robinson, C. V.; Dobson, C. M. Molecular Recycling within Amyloid Fibrils. Nature 2005, 436, 554−558. (21) Yamamoto, S.; Hasegawa, K.; Yamaguchi, I.; Goto, Y.; Gejyo, F.; Naiki, H. Kinetic Analysis of the Polymerization and Depolymerization of β2-Microglobulin-Related Amyloid Fibrils In Vitro. BBA-Protein Proteom. 2005, 1753, 34−43. (22) Binger, K. J.; Pham, C. L.; Wilson, L. M.; Bailey, M. F.; Lawrence, L. J.; Schuck, P.; Howlett, G. J. Apolipoprotein C-II Amyloid Fibrils Assemble via a Reversible Pathway that Includes Fibril Breaking and Rejoining. J. Mol. Biol. 2008, 376, 1116−1129. (23) Grüning, C. S. R.; Klinker, S.; Wolff, M.; Schneider, M.; Toksöz, K.; Klein, A. N.; Nagel-Steger, L.; Willbold, D.; Hoyer, W. The Offrate of Monomers Dissociating from Amyloid-β Protofibrils. J. Biol. Chem. 2013, 288, 37104−37111. (24) Qiang, W.; Kelley, K.; Tycko, R. Polymorph-Specific Kinetics and Thermodynamics of β-Amyloid Fibril Growth. J. Am. Chem. Soc. 2013, 135, 6860−6871. (25) 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, 9460−9469. (26) Dzwolak, W.; Surmacz-Chwedoruk, W.; Babenko, V. Conformational Memory Effect Reverses Chirality of Vortex-Induced Insulin Amyloid Superstructures. Langmuir 2013, 29, 365−370. (27) Jones, E. M.; Surewicz, W. K. Fibril Conformation as the Basis of Species- and Strain-Dependent Seeding Specificity of Mammalian Prion Amyloids. Cell 2005, 121, 63−72. (28) Bocian, W.; Borowicz, P.; Mikołajczyk, J.; Sitkowski, J.; Tarnowska, A.; Bednarek, E.; Głabski, T.; Tejchman-Malecka, B.; Bogiel, M.; Kozerski, L. NMR Structure of Biosynthetic Engineered Human Insulin Monomer B31(Lys)-B32(Arg) in Water/Acetonitrile Solution. Comparison with the Solution Structure of Native Human Insulin Monomer. Biopolymers 2008, 89, 820−830. (29) Shammas, S. L.; Knowles, T. P.; Baldwin, A. J.; MacPhee, C. E.; Welland, M. E.; Dobson, C. M.; Devlin, G. L. Perturbation of the Stability Of Amyloid Fibrils Through Alteration Of Electrostatic Interactions. Biophys. J. 2011, 100, 2783−2791. (30) Vestergaard, B.; Groenning, M.; Roessle, M.; Kastrup, J. S.; Van De Weert, M.; Flink, J. M.; Frokjaer, S.; Gajhede, M.; Svergun, D. I. A Helical Structural Nucleus Is the Primary Elongating Unit of Insulin Amyloid Fibrils. PLoS Biol. 2007, 5, e134. (31) Jansen, R.; Dzwolak, W.; Winter, R. Amyloidogenic SelfAssembly Of Insulin Aggregates Probed By High Resolution Atomic Force Microscopy. Biophys. J. 2005, 88, 1344−1353. (32) Foderà, V.; Librizzi, F.; Groenning, M.; Van De Weert, M.; Leone, M. Secondary Nucleation And Accessible Surface In Insulin Amyloid Fibril Formation. J. Phys. Chem. B 2008, 112, 3853−3858.

ACKNOWLEDGMENTS This work was supported by the National Science Centre of Poland, Grant DEC-2012/07/B/NZ1/02642 (W.D). Samples of human recombinant insulin analog were kindly provided by Dr. Piotr Borowicz from Institute of Biotechnology and Antibiotics, Warsaw.



ABBREVIATIONS: [X] mother fibrils of X formed spontaneously at 65 °C, and the corresponding amyloid phenotype; [X]Y daughter fibrils formed through seeding X with preformed fibrils of Y at 37 °C



REFERENCES

(1) Jarrett, J. T.; Lansbury, P. T., Jr. Seeding “One-Dimensional Crystallization” of Amyloid: A Pathogenic Mechanism in Alzheimer’s Disease and Scrapie? Cell 1993, 73, 1055−1058. (2) Chiti, F.; Dobson, C. M. Protein Misfolding, Functional Amyloid, and Human Disease. Annu. Rev. Biochem. 2006, 75, 333−366. (3) Fowler, D. M.; Koulov, A. V.; Balch, W. E.; Kelly, J. W. Functional Amyloid − from Bacteria to Humans. Trends Biochem. Sci. 2007, 32, 217−224. (4) Gazit, E. The ″Correctly Folded″ State of Proteins: Is It a Metastable State? Angew. Chem., Int. Ed. 2002, 41, 257−259. (5) Harper, J. D.; Lansbury, P. T., Jr. Models of Amyloid Seeding in Alzheimer’s Disease and Scrapie: Mechanistic Truths and Physiological Consequences of the Time-Dependent Solubility of Amyloid Proteins. Annu. Rev. Biochem. 1997, 66, 385−407. (6) Perczel, A.; Hudáky, P.; Pálfi, V. K. Dead-end Street of Protein Folding: Thermodynamic Rationale of Amyloid Fibril Formation. J. Am. Chem. Soc. 2007, 129, 14959−14965. (7) Dzwolak, W.; Smirnovas, V.; Jansen, R.; Winter, R. Insulin Forms Amyloid in a Strain-Dependent Manner: An FT-IR Spectroscopic Study. Protein Sci. 2004, 13, 1927−1932. (8) Dzwolak, W.; Jansen, R.; Smirnovas, V.; Loksztejn, A.; Porowski, S.; Winter, R. Template-Controlled Conformational Patterns of Insulin Fibrillar Self-Assembly Reflect History of Solvation of the Amyloid Nuclei. Phys. Chem. Chem. Phys. 2005, 7, 1349−1351. (9) Pedersen, J. S.; Dikov, D.; Flink, J. L.; Hjuler, H. A.; Christiansen, G.; Otzen, D. E. The Changing Face of Glucagon Fibrillation: Structural Polymorphism and Conformational Imprinting. J. Mol. Biol. 2006, 355, 501−523. (10) Andersen, C. B.; Hicks, M. R.; Vetri, V.; Vandahl, B.; RahbekNielsen, H.; Thøgersen, H.; Thøgersen, I. B.; Enghild, J. J.; Serpell, L. C.; Rischel, C.; et al. Glucagon Fibril Polymorphism Reflects Differences in Protofilament Backbone Structure. J. Mol. Biol. 2010, 397, 932−946. (11) Petkova, A. T.; Leapman, R. D.; Guo, Z.; Yau, W. M.; Mattson, M. P.; Tycko, R. Self-Propagating, Molecular-Level Polymorphism in Alzheimer’s Beta-Amyloid Fibrils. Science 2005, 307, 262−265. (12) Cloe, A. L.; Orgel, J. P. R. O.; Sachleben, J. R.; Tycko, R.; Meredith, S. C. The Japanese Mutant Aβ (ΔE22-Aβ(1−39)) Forms Fibrils Instantaneously, with Low-Thioflavin T Fluorescence: Seeding of Wild-Type Aβ(1−40) into Atypical Fibrils by ΔE22-Aβ(1−39). Biochemistry 2011, 50, 2026−2039. (13) Yamaguchi, K.; Takahashi, S.; Kawai, T.; Naiki, H.; Goto, Y. Seeding-Dependent Propagation and Maturation of Amyloid Fibril Conformation. J. Mol. Biol. 2005, 352, 952−960. (14) Tanaka, M.; Collins, S. R.; Toyama, B. H.; Weissman, J. S. The Physical Basis of How Prion Conformations Determine Strain Phenotypes. Nature 2006, 442, 585−589. (15) Foo, C. K.; Ohhashi, Y.; Kelly, M. J. S.; Tanaka, M.; Weissman, J. S. Radically Different Amyloid Conformations Dictate the Seeding Specificity of a Chimeric Sup35 Prion. J. Mol. Biol. 2011, 408, 1−8. (16) Gillam, J. E.; MacPhee, C. E. Modelling Amyloid Fibril Formation Kinetics: Mechanisms Of Nucleation And Growth. J. Phys.: Condens. Mater. 2013, 25, 373101. H

dx.doi.org/10.1021/jp510980b | J. Phys. Chem. B XXXX, XXX, XXX−XXX