Amyloid Polymorphism of α-Synuclein Induced by Active Firefly Luciferase

Abstract

Amyloidogenic proteins often exhibit fibrillar polymorphism through alternative assembly processes, which has been considered to have possible pathological implications. Here, firefly luciferase (LUC) is shown to induce amyloid polymorphism of ${\alpha}$-synuclein, the major constituent of Lewy bodies found in Parkinson's disease, by acting as a novel template. The drastically accelerated fibrillation kinetics of ${\alpha}$-synuclein with LUC required the nucleation center produced by the active enzyme of LUC. Fluorescent dye binding, transmission electron microscopy, and Fourier transformed infrared spectroscopy revealed the morphologically distinctive amyloid fibrils of ${\alpha}$-synuclein prepared in the absence or presence of LUC. As the altered morphological characteristics became inherent to the mature fibrils, those properties were inherited to next-generations via nucleation-dependent fibrillation process. The seed control, therefore, would be an effective means to modify amyloid fibrils with different biochemical characteristics. In addition, the LUC-directed amyloid fibrillar polymorphism also suggests that other cellular biomolecules including enzymes in general are able to diversify amyloid fibrils, which could be self-propagated with diversified biological activities, if any, inside cells.

Introduction

Amyloidosis, a clinical phenomenon observed in several neurodegenerative disorders including Parkinson’s and Alzheimer’s diseases, is caused by abnormal accumulation of insoluble protein fibrils of amyloids.1 Amyloid fibrils are a product of disorder-to-order transition of innocuous soluble proteins under pathological conditions such as oxidative stress and genetic mutations, which results in cross β-sheet conformation with the hydrogen bonded amino acids aligned perpendicular to the fibrillar axis.2 The amyloid fibrils of α-synuclein in particular are pathologically implicated in the neurodegenerative diseases including Parkinson’s diseases, dementia with Lewy bodies, and multiple system atrophy collectively known as α-synucleinopathies.3,4

According to the most widely accepted amyloidogenesis mechanism of nucleation-dependent fibrillation, α-synuclein monomers are assembled into amyloid fibrils through three stages comprising lag, exponential, and stationary phase. Once nucleus forms during the lag phase, the fibrillar growth is drastically accelerated through selective accretion of the monomeric protein to the nucleus as they experience structural transition to an amyloidogenic conformation.5-7 There-fore, the nucleus formation is not only the rate-determining step of entire fibrillation process,8 but it also determines eventual properties of amyloid fibrils such as fibrillar morphology, stability, and possible toxicity.9-13

α-Synuclein is a natively unfolded acidic protein with Mr of 14 kDa that is primarily found at pre-synaptic terminals.14 The protein has three distinctive segments in its primary structure including amphipathic N-terminal region (residues 1-60), central hydrophobic region (residues 61-95), and highly acidic C-terminal region (residues 96-140), which are demonstrated to be responsible for membrane interaction, protein aggregation, and multiple ligand interactions, respectively.15-25 Various α-synuclein interactive molecules have been reported to either inhibit or enhance the fibrillation of α-synuclein. For instance, the fibril formation was accele-rated by metal ions such as manganese, copper, iron, zinc, and aluminum, which altered the conformation of α-syn-uclein.26-31 We also reported that dye molecules such as eosin and coomassie brilliant blue interacted with α-syn-uclein, which resulted in selective α-synuclein self-oligo-merization.32,33 A recent report described that an antioxidant enzyme of glutathione peroxidase-1 also stimulated the α-synuclein amyloid fibril formation.34

Firefly luciferase (LUC) is an oxidative enzyme that catalyzes generation of light from chemicals of luciferin and ATP: luciferin + ATP + O2 → oxyluciferin + AMP + light (550 nm). α-Synuclein was previously reported to enhance the activity of LUC with a dissociation constant (Kd) of 8.1 μM,35 which led us to suggest α-synuclein as a novel protein regulator for LUC. Alternatively, LUC in turn could influence the fibrillation of α-synuclein. In this study, we have investigated LUC’s effects on the fibrillation of α-synuclein with respect to the protein assembly kinetics and the final fibrillar morphology, which would contribute to not only improving our under-standing toward the fibrillation mechanism but also develop-ment of controlling agents for α-synuclein fibrillation. The information obtained, therefore, could be employed for the eventual development of diagnostic and/or therapeutic strategies against the α-synucleinopathies.

 

Experimental Section

Materials. Human recombinant α-synuclein was prepared according to the procedure previously described.36 Firefly luciferase (Photinus pyralis) and beetle luciferin were from Promega (Madison, WI). Thioflavin-T (Th-T), 5,5',6,6'-tetrachloro-1,1,3,3'-tetraethylbenzimidazolyl carbocyanine iodide (JC-1), 8-anilino-1-naphthalenesulfonic acid (ANS), ATP, and MgSO4 were purchased from Sigma (St. Louis, MO).

Amyloid Fibril Formation. α-Synuclein was incubated at 0.5 mg/mL in 20 mM Mes (2-(N-morpholino)ethane-sulfonic acid, pH 6.5) at 37 °C under a shaking condition at 200 rpm. During the incubation, α-synuclein was mixed with 2.5 μM Th-T in 50 mM glycine (pH 8.5) at room temper-ature. The fluorescence was monitored with chemilumine-scence spectrophotometer (LS-55B, Perkin-Elmer, Waltham, MA). The resulting fibrillar formation of α-synuclein was estimated with Th-T binding fluorescence at 485 nm with an excitation at 450 nm.

JC-1 Binding Assay. α-Synuclein and its amyloid fibrils were incubated with 1.2 μM JC-1 in 20 mM Mes (pH 6.5) containing 5% DMSO. The JC-1 binding fluorescence spectra were obtained between 500 and 600 nm with an excitation at 490 nm. The fluorescence was measured with chemilumine-scence spectrophotometer (LS-55B, Perkin-Elmer, Waltham, MA).

Quantification of the Remaining Monomers. After incubation for 70 hours, the α-synuclein amyloids were precipitated with centrifugation at 13,200 rpm for 30 min. The amount of soluble α-synuclein monomers remaining in the supernatant was measured with BCA (bicinchoninic acid) protein assay (Thermo Fisher Scientific Inc., Waltham, MA).

Transmission Electron Microscopy (TEM). An aliquot (20 μL) containing amyloid fibrils of α-synuclein was ad-sorbed onto a carbon-coated copper grid (200-mesh, Electron Microscopy Sciences, Hatfield, PA) and dried in the air for 3 min. Following negative staining of the air-dried sample with 2% uranyl acetate (Electron Microscopy sciences, Hatfield, PA) for 5 min, the grids containing the amyloid fibrils were visualized with TEM (JEM1010, JEOL, Japan).

Fourier Transformation-infrared Spectrophotometer (FT-IR). FT-IR spectra of amyloid fibrils were measured with Nicolet 6700 FT-IR spectrometer equipped with DTGS detector (Thermo Fisher Scientific Inc., Waltham, MA). Data fitting and curve deconvolution for the FT-IR spectra were performed using OriginPro 8.0 software (OriginLab, Northampton, MA).

Firefly Luciferase (LUC) Assay. Assay mixture was made up with 1 mM ATP, 2 mM MgSO4, and 188 μM luciferin in 25 mM Tricine (pH 7.8). Bioluminescence of the LUC activity was monitored with chemiluminescence spectrophotometer (LS-55B, Perkin-Elmer, Waltham, MA). Light emission was measured at 550 nm with an integration time of 0.02 sec.

ANS Binding Assay. Pre-incubated LUC at 2.0 μM was incubated with 20 μM ANS in 20 mM Mes (pH 6.5) at room temperature. The ANS binding fluorescence spectra were obtained between 400 and 600 nm with an excitation at 350 nm. The fluorescence was measured with chemiluminescence spectrophotometer (LS-55B, Perkin-Elmer, Waltham, MA).

Self-propagation Experiment. Amyloid seeds were pre-pared with ultrasonication of the mature amyloid fibrils using a micro-tip-type ultrasonic processor (VCX 750, Sonics & Materials, Inc., Newtown, CT). Sonication was performed repeatedly for total of 10 cycles with the fibrils located on ice. Each cycle consisted of an on/off cycle for 15 sec each. After each cycle of sonication, the sample was quiescently placed on ice for 60 sec. The resulting amyloid fragments were examined with TEM and used as the seeds. Daughter fibrils were developed by seeding monomeric α-synuclein with the amyloid fragments at 5% (w/w).

 

Results

Firefly Luciferase (LUC) Accelerates the Fibrillation of α-synuclein and Induces the Morphologically Distinc-tive Fibrils. Kinetics study monitored by thioflavin-T (Th-T) binding fluorescence of amyloid fibrils showed that the fibrillation of α-synuclein was drastically accelerated by LUC (Fig. 1(A)). The lag phase reduced from 20 h to 6 h and the final intensity of Th-T binding fluorescence increased by one and a half-fold in the presence of LUC at 0.1 μM. According to a previous study, JC-1 (5,5',6,6'-tetrachloro-1,1,3,3'-tetraethylbenzimidazolyl carbocyanine iodide) can also be used as a fluorescent probe to classify α-synuclein according to its assembled states between the intermediate oligomers and the final fibrils.37 The JC-1 binding fluorescence spectra were monitored between 500 nm and 600 nm with an excitation at 490 nm for the amyloid fibrils prepared in the presence or absence of LUC after 60 h of fibrillation (Fig. 1(B)). The JC-1 binding fluorescence of the α-synuclein amyloid fibrils prepared in the absence of LUC exhibited the maximum at 539 nm, which was dramatically increased by 4-fold for the fibrils obtained with LUC. In addition, the JC-1 binding fluorescence for the LUC-induced fibrils showed another significant shoulder peak culminated at 561 nm, indicating a distinctive feature of the fibrils derived with LUC.

In order to clarify whether the increases in the final Th-T and JC-1 binding fluorescence intensities were attributable to an increase in the actual amount of fibrils, the monomer levels left behind following 70 h of the extensive fibrillation in the presence and absence of LUC were monitored with bicinchoninic acid (BCA) protein assay (Fig. 1(C)). The data showed that the remaining monomer levels were rather com-parable for both conditions with and without LUC, suggest-ing that LUC presumable yielded a different type of amyloid fibrils. Examination with transmission electron microscopy (TEM) confirmed the distinctive fibrillar structures between the final fibrils obtained in the absence and presence of LUC (Fig. 1(D)). In the presence of LUC, the fibrils formed densely populated fibrillar meshwork reflecting rather associative nature of the LUC-induced fibrils which were thicker than the fibrils prepared with α-synuclein alone. The former showed an average width of 16 nm while the latter was 8 nm.

Figure 1.Enhanced amyloid fibril formation with LUC and the fibrillar polymorphism of α-synuclein in the presence and absence of LUC. (A) Fibrillation kinetics of α-synuclein in the presence of LUC. α-Synuclein (35 μM) was incubated with 0.1 μM LUC (red) at 37 °C with shaking at 200 rpm. α-Synuclein (black) and LUC (gray) were separately incubated as controls. Amyloid fibrillation was monitored by Th-T binding fluorescence at 482 nm with an excitation at 450 nm. (B) JC-1 binding fluorescence of the α-synuclein amyloid fibrils prepared without (black) and with LUC (red). Dotted lines indicate the respective spectra prior to the incubation without (black) and with LUC (red). Fluorescence spectra of JC-1 were monitored between 500 and 600 nm with an excitation at 490 nm (C) Amounts of soluble α-synuclein left in the supernatant before (black) and after (gray) the fibrillation for 70 hours at 37 °C. (D) TEM images of amyloid fibrils obtained in the absence (left) and presence (right) of LUC. Magnied (× 4) images are also presented in the insets. Scale bar, 200 nm. (E) FT-IR spectra of the fibrils prepared in the absence (a) and presence (b) of LUC at amide I region (1600-1700 cm−1). The colored lines obtained via the spectral deconvolution represent protein secondary structures constituting the amyloid fibrils (see Table 1).

To elaborate the fibrillar polymorphism, FT-IR spectro-photometer was employed to assess the secondary structures of the fibrils (Fig. 1(E)). In particular, amide I band (1600-1700 cm−1) has been employed to evaluate structural and conformational characteristics of proteins.38-40 Spectra of the α-synuclein fibrils produced without (Fig. 1(E)-(a)) and with (Fig. 1(E)-(b)) LUC were deconvoluted at 1619, 1630, 1650, and 1668 cm−1 using OriginPro 8.0 program, which repre-sent anti-parallel β-sheets, parallel β-sheets, α-helices/random coils, and β-turns, respectively.41-43 The fibrils obtained in the absence of LUC exhibit two major peaks at 1630 cm−1 and 1668 cm−1 with relative contributions of 50.2% and 29.8% and two minor peaks at 1619 cm−1 and 1650 cm−1 with 16.3% and 3.7%, respectively (Table 1). The fibrils with LUC, on the other hand, gave rise to the two major peaks at 1630 cm−1 and 1668 cm−1 with 58.0% and 29.7% and the two minor peaks at 1619 cm−1 and 1650 cm−1 with 7.6% and 4.7%, respectively. The results explain that the fibrils with LUC contain higher level of parallel β-sheets with decreased anti-parallel β-sheet content although this shift in the secondary structures could also be caused by the inter-fibrillar interactions responsible for the meshwork formation.

Table 1.Relative contribution of protein secondary structures to the FT-IR spectra

Active Conformation of LUC Plays a Critical Role on the Accelerated Fibrillation of α-Synuclein. In order to confirm whether the enzymatic activity of LUC was essen-tial for the accelerated fibrillation of α-synuclein, the fibrillation kinetics were monitored with and without the LUC inactivated via pre-incubation at 37 °C. As the pre-incubation proceeded, LUC was gradually inactivated to 30% of the original activity after 4 h (Fig. 2(a)). The decrease in LUC activity was accompanied with its structural change as assessed with 8-anilino-1-naphthalenesulfonic acid (ANS) binding fluorescence44 (Fig. 2(b)). As the enzymatic activity diminished, the fluorescence maximum at 460 nm increased, indicating an exposure of hydrophobic regions of LUC during the pre-incubation. In the presence of the 4 h pre-incubated LUC, the enhanced fibrillation observed with the fully active LUC was significantly suppressed even though the suppressed fibrillation was apparently faster than the kinetics observed with α-synuclein alone, indicating that the accelerated fibrillation by LUC was dependent upon the biological activity and/or active conformation of the enzyme (Fig. 2(c)).

The LUC activity-dependent accelerated fibrillation was further investigated in the presence of LUC’s substrates such as luciferin and ATP-Mg2+. In the absence of LUC, neither luciferin nor ATP-Mg2+ at 0.5 μM affected the fibrillation kinetics of α-synuclein. With LUC at 0.1 μM, however, luciferin augmented the Th-T binding fluorescence at the stationary phase twice as much as the fluorescence obtained without the luciferin (Fig. 2(d)). ATP-Mg2+, on the other hand, decreased the Th-T binding fluorescence by 0.8-fold from its absence. From these results, the active conformation of LUC which could be influenced by the substrates during the initial period of fibrillation appears to be crucial for the enhanced amyloidogenesis.

Figure 2.Requirement of LUC activity for the accelerated α-synuclein fibrillation. (a) LUC activity remaining after incubations at 37 °C for 0 (blue), 30 (green), 60 (purple), and 240 (red) min was examined with an emitted light at 550 nm. (b) ANS binding fluorescence of LUC after incubations at 37 °C for 0 (blue), 30 (green), 60 (purple), and 240 (red) min. Fluorescence spectra of ANS was monitored between 400 and 600 nm with an excitation at 350 nm. (c) Fibrillation kinetics of α-synuclein in the presence of active (blue) and inactive (red) LUC. α-Synuclein (black) was separately incubated at 35 mM as a control. (D) Fibrillation kinetics of α-synuclein with LUC and substrates (luciferin and ATP-Mg2+). In the absence of LUC, α-synuclein (35 μM) was fibrillated without (black open dots) and with the substrate of either 0.5 μM luciferin (blue open dots) or 0.5 μM ATP-Mg2+ (red open dots). The fibrillation kinetics of α-synuclein was also monitored in the presence of LUC (0.1 μM) without the substrates (black closed dots) and with either 0.5 μM luciferin (blue closed dots) or 0.5 μM ATP-Mg2+ (red closed dots).

Self-propagation of the Altered Morphology of α-Synuclein Fibrils Prepared with LUC. To clarify whether the newly developed morphology of the LUC-induced amyloid fibrils of α-synuclein was transient or inherent to the fibrils, the self-propagation experiment of amyloid fibrils was carried out. Previous studies demonstrated that the properties of amyloid fibrils can be transferred to subsequent daughter fibrils by introducing a small amount of amyloid fragments derived from their parent strands to a fibrillation process of amyloidogenic protein.45-47 α-Synuclein fibrillation was seeded at 5% (w/w) with the amyloid fragments obtain-ed from the parent fibrils prepared in the presence and absence of LUC with sonication (Fig. 3). In the presence of the seeds derived from the fibrils prepared without LUC, the lag period completely disappeared from the 30 h obtained in the absence of the seeds while the final Th-T binding fluorescence reached to the same level as the fibrillation obtained without the seeds. In the presence of the LUC-induced seeds at 5% (w/w), on the other hand, the final Th-T binding fluorescence increased drastically by more than 4-folds, which was reminiscent of the accelerated LUC-induced α-synuclein fibrillation as observed in Figure 1(a). The lag phase, however, was also shortened significantly from the unseeded fibrillation, but it was definitely longer than the kinetics obtained with the seeds prepared in the absence of LUC. To confirm whether the increase in the final Th-T binding fluorescence was due to the altered morpho-logy of amyloid fibrils instead of their actual quantity, the fluorescence intensity was divided by the mass of fibrils.48 The result indicated that the fibrils derived with LUC-induced seeds yielded higher Th-T binding fluorescence per mass of fibrils by two-fold than that of the fibrils obtained with the seeds prepared with α-synuclein alone (Fig. 3(b)). The data suggest that the increased Th-T binding fluorescence was not due to the actual amount of the fibrils but their altered properties reflected in the fibrillar morphology.

Figure 3.Self-propagation of the two amyloid polymorphs. (a) Fibrillation kinetics of α-synuclein with the seeds at 5% (w/w). α-Synuclein at 35 μM was incubated in the absence (black) and presence of the seeds prepared from either the α-synuclein fibrils (blue) or the LUC-induced fibrils (red). (b) The Th-T binding fluorescence per mass of fibrils self-propagated in the presence of the seeds derived from either the α-synuclein fibrils (blue) or the LUC-induced fibrils (red). (c) TEM images of the α-synuclein amyloid fibrils (top row) and the LUC-induced fibrils (bottom row) from the parent amyloid brils to their daughter brils. Scale bars, 200 nm.

As morphologies of the final amyloid fibrils were examin-ed with transmission electron microscope, structural charac-teristics of the parental strands were transferred to their respective daughter fibrils. While rather straight and un-tangled fibrils were obtained for the daughter fibrils with the seeds prepared with α-synuclein alone, the LUC-induced seeds produced the associative daughter fibrils which re-sembled the fibrils of α-synuclein prepared in the presence of LUC (Fig. 1(d)). This self-propagation property of the LUC-induced fibrils would suggest that the unique morpho-logy of α-synuclein fibrils derived with LUC became inherent to the fibril

Scheme 1.Amyloid polymorphism of α-synuclein with and without LUC.

 

Discussion

This study has shown that a single amyloidogenic protein turns into two different fibrils with distinctive morphologies. This bifurcation of protein assembly has been achieved with a biologically active enzyme of luciferase. The process can be hypothesized that α-synuclein exists in a kinetically trapped state from which the protein escape by demolishing certain part(s) of the surrounding energy barrier, which might provide more than one paths for the protein to be stabilized into thermodynamically stable states by yielding various forms of amyloid fibrils (Scheme). Both LUC and a trans-formed aggregative α-synuclein might be able to lower the activation barrier by acting as a template for the pristine α-synuclein. The resulting two distinctive fibrils could direct the fibrillation process into their respective states by provid-ing the seeds with which the corresponding amyloid fibrils can be amplified.

In fact, all the data shown in this study illustrate that the fibrillar polymorphism has been directed by the seed gene-rated by specific mutual interaction between α-synuclein and the biologically active LUC. The observations including the drastically reduced lag phase of α-synuclein fibrillation in the presence of LUC, the altered dye binding properties of the LUC-induced amyloid fibrils with thioflavin-T and JC-1, and the biochemical and morphological analyses of the fibrils with FT-IR and TEM support the seed-dependent alternative amyloid fibril formation. This phenomenon, there-fore, emphasizes the significance of nucleation centers for amyloid fibril formation in terms of not only facilitating the fibrillation process but also endowing those fibrils with certain biochemical properties. Those fibrils with unique properties, then, could be further amplified via the self-propagation procedure since the morphological characteri-stics become inherent to the newly developed fibrils.

This study, therefore, opens up a possibility that enzymes or even other biomolecules could have potentials to act as templates for amyloidogenic proteins like α-synuclein to exhibit the fibrillar polymorphism. By assuming that the biological activity of cellular toxicity could be associated the morphological and thus biochemical characteristics of amyl-oid fibrils in addition to the fact that those fibrils could be self-propagated, the diversified fibrils produced by a myriad of biological partners could be considered as independent toxic causes. Since cells would have various partners for amyloidogenic proteins and the resulting polymorphic fibrils would have distinctive cytotoxicities, if any, the diseases related to amyloidogenesis should be re-evaluated by scrutinizing those diversified toxic candidates under pathological conditions.