Amyloid Oligomers and Mature Fibrils Prepared from an Innocuous Protein Cause Diverging Cellular Death Mechanisms*

Background: Although oligomers are considered more important, mature fibrils also show evidence as cytotoxic agents in neurodegenerative diseases. Results: Oligomers and fibrils both kill PC12 cells albeit mechanistically differently. In vivo, only oligomers inhibit hippocampal long term potentiation. Conclusion: Protein aggregates, even those irrelevant to disease, are capable of inducing different toxic actions in neuronal cells. Significance: Understanding these toxic mechanisms is vital in improving amyloidosis therapy. Despite significant advances, the molecular identity of the cytotoxic species populated during in vivo amyloid formation crucial for the understanding of neurodegenerative disorders is yet to be revealed. In this study lysozyme prefibrillar oligomers and fibrils in both mature and sonicated states have been isolated through an optimized ultrafiltration/ultracentrifugation method and characterized with various optical spectroscopic techniques, atomic force microscopy, and transmission electron microscopy. We examined their level and mode of toxicity on rat pheochromocytoma (PC12) cells in both differentiated and undifferentiated states. We find that oligomers and fibrils display cytotoxic capabilities toward cultured cells in vitro, with oligomers producing elevated levels of cellular injury toward undifferentiated PC12 cells (PC12undiff). Furthermore, dual flow cytometry staining experiments demonstrate that the oligomers and mature fibrils induce divergent cellular death pathways (apoptosis and secondary necrosis, respectively) in these PC12 cells. We have also shown that oligomers but not sonicated mature fibrils inhibit hippocampal long term potentiation, a form of synaptic plasticity implicated in learning and memory, in vivo. We conclude that our in vitro and in vivo findings confer a level of resistance toward amyloid fibrils, and that the PC 12-based comparative cytotoxicity assay can provide insights into toxicity differences between differently aggregated protein species.

Despite significant advances, the molecular identity of the cytotoxic species populated during in vivo amyloid formation crucial for the understanding of neurodegenerative disorders is yet to be revealed. In this study lysozyme prefibrillar oligomers and fibrils in both mature and sonicated states have been isolated through an optimized ultrafiltration/ultracentrifugation method and characterized with various optical spectroscopic techniques, atomic force microscopy, and transmission electron microscopy. We examined their level and mode of toxicity on rat pheochromocytoma (PC12) cells in both differentiated and undifferentiated states. We find that oligomers and fibrils display cytotoxic capabilities toward cultured cells in vitro, with oligomers producing elevated levels of cellular injury toward undifferentiated PC12 cells (PC12 undiff ). Furthermore, dual flow cytometry staining experiments demonstrate that the oligomers and mature fibrils induce divergent cellular death pathways (apoptosis and secondary necrosis, respectively) in these PC12 cells. We have also shown that oligomers but not sonicated mature fibrils inhibit hippocampal long term potentiation, a form of synaptic plasticity implicated in learning and memory, in vivo. We conclude that our in vitro and in vivo findings confer a level of resistance toward amyloid fibrils, and that the PC 12-based comparative cytotoxicity assay can provide insights into toxicity differences between differently aggregated protein species.
The failure of specific proteins to correctly fold and adopt their native functional structures has been correlated with a vast range of debilitating diseases including Alzheimer and Parkinson diseases (1)(2)(3). Such proteinaceous fibrillar aggregates known as amyloid fibrils are currently implicated in scores of degenerative diseases that affect a variety of peripheral tissues as well as the central nervous system (4,5).
Although the dysfunctional assembly of peptides does not always carry a negative consequence, as seen with the tumoricidal molten globule-oleic acid complex HAMLET (Human Alpha-Lactalbumin Made LEthal to Tumor cells (6)), a group of ϳ25 unrelated proteins have been shown to be causative agents in the formation of a number of clinically distinct conditions (7). However, recent evidence suggests the ability to aggregate may be a generic property of possibly all polypeptide chains under specific denaturing conditions. Hen egg white lysozyme (HEWL) 2 can be engineered to aggregate at acidic pH and elevated temperatures (8) making it a very useful model to study protein misfolding and disease.
Despite the advancements in the field of protein aggregation and disease pathology, very little is known about the exact in vivo mechanisms of formation and cytotoxicity. Various studies involving animal models (9) and cell lines (10,11) provided early evidence that supported the idea that mature fibrils were solely cytotoxic. However, this hypothesis has been the subject of intense scrutiny in recent years, and a large number of experiments have shown that the prefibrillar aggregates, known as oligomers, are intrinsically involved in and may even be the sole and direct cause of cell damage observed in various disease * This work was supported in part by grants from the Health Research Board of Ireland (to I. K.) and Science Foundation Ireland (to M. J. R). The authors declare that they have no conflicts of interest with the contents of this article. 1  models (7,(12)(13)(14)(15). The growth in literature surrounding the role of oligomers in neurodegenerative disease pathology has coincided with increasing levels of controversy surrounding mature fibrils in these disorders. One argument suggests that fibrils may possess no cytotoxic abilities at all (13,15,16), whereas there is a substantial body of experimental evidence that fully demonstrates that amyloid fibrils are capable of causing cellular death in numerous situations (17)(18)(19). A related point of interest is whether cellular differentiation confers any significant biological resistance against amyloid fibrils and oligomers. Currently there is a substantial lack of knowledge regarding cell susceptibility and resistance to protein aggregates. Studies have presented evidence that suggests that toxic oligomers and fibrils display differing levels of toxicity toward particular cell lines (20,21). It has also been shown that differentiated cells exhibit elevated levels of resistance against amyloid injury (22).
In this article we investigate the neurotoxic effects of HEWL oligomers, sonicated fibrils, and mature fibrils on the rat neuronal cell line PC12 in both differentiated and undifferentiated states. Mature amyloid fibrils and prefibrillar aggregates were isolated and characterized using a battery of techniques including atomic force microscopy (AFM), thioflavin T (ThT) fluorescence, and Congo Red (CR) birefringence. Here, we report for the first time that both mature amyloid fibrils and prefibrillar oligomeric species, isolated after 21 days from wild type HEWL, are highly toxic to PC12 cells in vitro, and in addition, the separated fractions elicit different cellular death pathways, one apoptosis, the other a form of cell death associated with necrotic cell death or a late apoptotic mechanism. Differentiated PC12 cells (PC12 diff ) showed an elevated level of resistance toward oligomers only, while no such resistance was found for mature amyloid fibrils. We also investigated the effects of HEWL aggregates on long term potentiation (LTP), a model of cellular mechanisms underlying learning and memory formation. Interestingly, we found that LTP in vivo was inhibited by lysozyme oligomers but not sonicated fibrils.
These results indicate that although produced from the same original protein, amyloid bodies in different aggregated states represent individual cytotoxic entities possessing unique properties, potentially providing insight into the true injurious events observed in neurodegenerative diseases.

Experimental Procedures
Proteins and Reagents-All reagents were of analytical grade or the highest purity available. HEWL, propidium iodide (PI), 1-anilinonaphthalene-8-sulfonic acid (ANS), and Tris were purchased from Sigma. Alamar Blue and all cell culturing equipment were purchased from Invitrogen. RPMI and PBS were purchased from GIBCO. Congo Red and ThT were obtained from Acros. FITC-labeled annexin-V was purchased from Promokine.
HEWL Aggregation-HEWL was prepared in distilled water adjusted to pH 2 with HCl. The final protein concentration was 1 mM (molecular mass of HEWL ϭ 14.3 kDa, extinction coefficient ⑀ 1% ϭ 26.4 at 280 nm). Aliquots were incubated at 65°C for 21 days to allow sufficient fibrillization to occur (8). Control HEWL was prepared fresh at room temperature at pH 2.
HEWL Separation-Desired fractions of lysozyme were separated in relationship to particle mass after 21 days of incubation. Samples were ultracentrifuged for 1 h at 100,000 ϫ g using a Beckman Optima TLX Ultracentrifuge. The pellet, containing mature amyloid fibrils, was resuspended in pH 2 distilled water to a concentration of 1 mM. At this point sonicated samples were subjected to ultrasound power for 30 s periodically for 5 min. The supernatant was spun down via ultrafiltration, utilizing 100-kDa cutoff filters (Amicon Ultra-4 Centrifugal Filter Devices). The resulting filtrate was further separated by repeating the ultracentrifugation process with 30-kDa cutoff filters. The resulting retentate (Ͻ100 kDa and Ͼ30 kDa) constituted the oligomer solution. Aggregated samples were prepared fresh on the day diluted and a final concentration of 1 mM was determined using a NanoDrop ND-1000 Spectrophotometer (Thermo Scientific) averaging three readings for use in cytotoxic assays.
Thioflavin T Assay-ThT was carried out as described previously (23). HEWL samples were measured using a Jasco FP 6200 spectrofluorimeter. Fluorescence intensity was measured by excitation at 440 nm (slit width 5 nm) and emission at 482 nm (slit width 10 nm), averaging over 30 s.
Congo Red Spectroscopic and Birefringence Assay-Both Congo Red assays was carried out as described (23). UV-visible absorbance was measured using a NanoDrop ND-1000 Spectrophotometer between wavelengths 400 and 700 nm. A maximal spectral difference at 505 nm is evincive of amyloid fibrils. For the birefringence assay the Congo Red/HEWL solutions were analyzed using a Nikon Eclipse E400 POL with the polarizers crossed at a 90°angle to each other. Amyloid birefringence was identified by the presence of apple green coloration.
ANS Binding-ANS was carried out as described (24) using excitation at 390 nm, and emission between 410 and 600 nm.
AFM Imaging-HEWL samples were prepared for AFM analysis by depositing the fibril suspension onto freshly cleaved mica by spin coating. The spin recipe encompassed a dispersion routine (750 rpm for 45 s followed by 1000 rpm for 45 s) allowing the fibril suspension to be spread evenly across the mica surface. Samples were then dried (4000 rpm for 15 s) to remove any excess solvent. AFM measurements were performed in ambient conditions on an Asylum MFP-3D atomic force microscope (Asylum Research, Santa Barbara, CA). The probe normal spring constant was between 1.2 and 2.0 newton/m with a tip apex Ͻ5 nm. High resolution acoustically driven cantilevers (Nanosensors, SSS-FM) operating at a resonance frequency of 60 -80 KHz in AC-mode were used with a scan resolution of 1024 ϫ 1024 and scan frequency of 0.6 Hz. Structures were analyzed using MFP-3D IGOR PRO software.
Transmission Electron Microscopy-Protein samples were visualized using a JEOL 2100 Transmission Electron Microscope operating at 200 kV with a lanthanum hexaboride (LaB 6 ) emission source. Prior to microscopy, 10 l of HEWL sample was deposited onto carbon-coated grids. The solution was displaced with 0.5% glutaraldehyde and subsequently washed with dH 2 O. The sample was stained using uranyl acetate and allowed to air dry before transferring to microscope.
Cell Culture and Differentiation-Undiffereniated PC12 (PC12 undiff ) cells were maintained in RPMI 1640 medium/ GlutaMAX TM -1 supplemented with 10% FCS, and 1% penicillin/streptomycin. Culture medium was replaced every 3 days. For cell viability assays, cells were seeded at a density of 10 6 cells/well in 96-well plates. For NGF-induced differentiation, PC12 cells were seeded onto collagen-coated 96-well plates at a cell density of 1.5 ϫ 10 4 cells/well in RPMI 1640 medium/ GlutaMAX TM -1. Wells were supplemented with 1% FCS, 1% penicillin/streptomycin, and 0.01% NGF every 2 days for a period of 7 days to allow differentiation. Immediately prior to the experiment the 1 mM stock of the fibrillar and prefibrillar aggregates were diluted with distilled water to produce concentrations varying from 20 to 300 M. The cells, lysozyme aggregates, and media (volume 100 l) were added and incubated overnight prior to treatment with Alamar Blue. All tests were conducted in triplicate. All cells were maintained in a 95% air, 5% CO 2 humidified atmosphere at 37°C.
Alamar Blue Cell Viability Assay-Alamar Blue fluorescence measurements were carried out as described previously (25). Emission levels were measured using a SPECTRAmax Gemini XS Microplate Spectrofluorometer. Cell viability was expressed as a percentage of fluorescence absorbance in the aggregateexposed cells in relationship to untreated cells.
Flow Cytometry-Cells were seeded at 10 6 cells/ml/well the day before analysis in 12-well suspension plates. Wells were treated with 25 M of the aggregated solution and left for 24 h overnight at 37°C. Cells (1 ml) were removed and spun down at 1600 rpm for 5 min prior to washing in a 1ϫ annexin binding buffer (10.9 mM Hepes, 140 mM NaCl, and 2.5 mM CaCl 2 , pH 7.4). Cells were spun down and stained with 10 l of anti-annexin V antibody and left for 15 min on ice in the dark. Cells were spun down, washed, and resuspended in 500 l of 1ϫ annexin binding buffer. Immediately prior to FACS analysis, cells were treated with 10 l of 50 g/ml of PI. Flow cytometry was carried out using a FACSCalibur flow fluorocytometer (BD Biosciences). Compensation was carried out using control samples for untreated PI/annexin V only, double stained, and non-stained.
Electrophysiological Techniques-In vivo electrophysiology was performed using techniques described previously (26). Animal experiments were licensed by the Department of Health and Children, Ireland. Adult male Wistar rats were anesthetized with urethane (1.5 g/kg, intraperitoneally). Single pathway recordings of field excitatory postsynaptic potentials (EPSPs) were made from the stratum radiatum in the CA1 area of the dorsal hippocampus in response to stimulation of the ipsilateral Schaffer collateral/commissural pathways. Test EPSPs were evoked at a frequency of 0.033 Hz and at a stimulation intensity was adjusted to give an EPSP amplitude of 50% of maximum. The high-frequency stimulation (HFS) protocol for inducing LTP consisted of 10 trains of 20 stimuli, interstimulus interval of 5-ms, and inter-train intervals of 2 s. The intensity was increased to give an EPSP of 75% of maximum amplitude during the HFS. LTP is expressed as the mean Ϯ S.E. % baseline field EPSP amplitude recorded over at least a 30-min baseline period. Similar results were obtained when the EPSP slope rather than amplitude was measured.
For statistical analysis, EPSP amplitudes were grouped into 10-min epochs. Standard one-way analysis of variance was used to compare the magnitude of LTP between multiple groups followed by post hoc Tukey's tests. Unpaired Student's t tests were used for two-group comparisons. A p Ͻ 0.05 was considered statistically significant.
To inject samples into the rat brain a cannula was implanted in the lateral cerebral ventricle (coordinates: 1 mm lateral to the midline and 4 mm below the surface of the dura) just prior to electrode implantation. Injections (15 l over 10 min) were made via a Hamilton syringe connected to the internal cannula.

Isolation and Characterization of Prefibrillar Aggregates and Mature Amyloid Fibrils of Hen Egg White Lysozyme-HEWL
samples incubated for 21 days were separated through a series of ultracentrifugation and ultrafiltration steps (see "Experimental Procedures"). To eliminate any potential diffusion-limitation issues that could arise from using mature fibrils, samples were sonicated post-isolation to produce shorter fragments of mature fibrils that still maintain their general fibrillar structure (27). To validate the presence of amyloid structures conclusively and to avoid false negative results, newly isolated samples were subjected to a combination of Congo Red spectroscopic and birefringence assays as well as ThT (23). Both oligomers and sonicated fibrils showed no signs of birefringence, however, mature HEWL fibrils are clearly birefringent at ϫ4 magnification (Fig. 1A). For Congo Red spectral analysis all samples showed a shift in the maximum absorbance from 495 to 505 nm (Fig. 1B). HEWL was also examined daily for 21 days using the cationic benzothiazole dye ThT, which exhibits enhanced fluorescence upon binding to amyloid fibrils. HEWL at 65°C showed a marked daily increase in fluorescence intensity values when compared with values exhibited by control HEWL at room temperature (Fig. 1C). Furthermore, when treated with ANS ultracentrifuged/ultrafiltrated oligomers, sonicated fibrils, and mature fibrils all showed a blue shift in spectral maxima (Fig. 1D).
Microscopy Imaging of Amyloid Aggregates-As a direct visual confirmation of the existence of oligomeric and fibrils species, oligomers, sonicated fibrils, and mature fibrils were analyzed using AFM and TEM in conjunction with the above spectroscopic measurements (Fig. 2A). In contrast to the mature fibrils, both the oligomeric and sonicated samples exhibited a scattered array of spots providing direct evidence of the successful size-based fractionation through the combined ultracentrifugation/ultrafiltration methodology. For the mature fibril, the three-dimensional structure and the ␣ z-height line section was measured 3.26 Ϯ 0.65 nm, whereas the ␤ z-height line section at the crossover was 4.23 Ϯ 0.46 nm, indicative of helical twists in the mature fibril (Fig. 2, C and D) as classically found in typical cases (28). Oligomers, sonicated fibrils, and mature lysozyme fibrils were also visualized using TEM. All samples were negatively stained with the dye uranyl acetate (Fig. 2E). All samples show distinct differences in shape and size as a result of the destructive nature of sonication.
Oligomers and Fibrils Are Cytotoxic to PC12 Cells of Both Undifferentiated and Differentiated Morphologies-The toxicities of oligomeric and fibrillar species of both sonicated and mature states were assessed using the PC12 cell line. Cultured PC12 cells have the propensity to continuously divide and provide a model for tumor research. Upon continued treatment with NGF, PC12 undiff cells transform from a neoplastic morphology and begin to extend branching varicose processes (Fig.  3A). Different fibrillar and prefibrillar aggregates were added to both cell lines at concentrations in the 20 to 300 M range and incubated at 37°C for 24 h. Percentage cell death was calculated using the Alamar Blue viability assay. All lysozyme species show significant efficacy toward PC12 diff cells with oligomers yielding LC 50 values of 44 M and fibrils values of 91 M (Table 1). PC12 undiff cells were treated with identical concentrations. The LC 50 values for mature amyloid fibrils (94 M) showed no discernible difference in susceptibility. Interestingly, oligomertreated PC12 undiff cells give an LC 50 value of 29 M, considerably lower than that observed with the differentiated cell type. Sonicated fibrils, whereas not as potent as oligomers on both cell lines, were more potent than mature full fibrils. The monomeric state of HEWL was additionally examined and showed no cytotoxicity toward either cell line (Fig. 3, B and C).
HEWL Oligomers and Amyloid Fibrils Induce Cellular Death via Apoptotic and Secondary Necrotic Pathways, Respectively-During apoptosis, cells undergo highly specific morphological changes including plasma membrane blebbing, pyknosis, and exposure of the negatively charged phospholipid phosphatidylserine (PS) on the inner leaflet of the cell membrane (29,30). Necrotic cells undergo a different form of cell death characterized by swelling and sudden deflation of the dying cell, resulting in leakage of internal cellular contents into the surrounding milieu (31). PC12 undiff/diff cell death was analyzed by FACS using a combined stain for FITC-labeled annexin-V (apoptosis) and PI (necrosis or non-apoptotic). This experiment demonstrates that both fibrils and oligomers induce high levels of cell death within PC12 undiff/diff cells. Oligomer-treated PC12 undiff cells (Fig. 4B) show elevated annexin V-FITC staining when compared against the control group (Fig. 4A). Fibrils show increased staining for annexin V-FITC and PI (Fig. 4D). Similarly, PC12 diff cells also show differences in cell death mechanisms. Oligomeric treated cells show a propensity for binding annexin V-FITC staining (Fig. 4F), whereas fibrillar treated PC12 diff cells showed increased annexin V/PI double positive staining (Fig. 4H) when compared with control (Fig. 4D). PC12 undiff/diff cells were also analyzed using sonicated fibrils. PC12 undiff cells showed increased staining for annexin-V (Fig.  4C), whereas PC12 diff cells showed approximately equal stain- ing for annexin-V and dual annexin-V/PI (Fig. 4G). In addition cellular death in PC12 undiff/diff cells treated with HEWL samples was examined using Western blot analysis probing for PARP. In both cell lines native HEWL and the monomeric fraction showed no apoptotic death. Sonicated fibrils and mature fibrils elicited a slight apoptotic response although oligomers appeared to induce the highest level of apoptotic cell death (Fig.  5). These results suggest that PC12 undif/diff cells treated with HEWL-isolated oligomers, sonicated fibrils, and mature results in the cleavage of the apoptotic marker protein PARP, indicating that apoptosis has been activated in response to these proteinaceous fractions, which is in agreement with our flow cytometry data.
Inhibition of LTP by Lysozyme Oligomers-Amyloid neurodegenerative diseases commonly lead to cognitive decline and memory deterioration. To determine whether HEWL aggregates alter normal brain functions in vivo we examined their effects on hippocampal LTP, a well established correlate of learning and memory, which is potently inhibited by amyloid ␤-protein (A␤) (26). Because we had previously found that fibril preparations of A␤ are inactive in this model we compared the actions of oligomers and sonicated fibrils in the present experiments. Fig. 6 shows that in anesthetized control rats HFS induced robust and long lasting potentiation (Ͼ3 h) of excitatory synaptic transmission in the CA1 area of the hippocampus after intracerebroventricular injection of either 15 l of distilled water (n ϭ 3) or 1 mM non-aggregated HEWL (15 nmol, n ϭ 4) (combined group, 151 Ϯ 15%, n ϭ 7, p Ͻ 0.05, compared with pre-HFS baseline). In contrast, injection of the same volume of a 350 M HEWL oligomer sample completely FIGURE 2. AFM images of amyloid aggregates on mica formed from hen egg lysozyme. A, oligomers, sonicated fibrils, and amyloid mature fibrils visualized using AFM. B, shows a typical z-height AFM image of a fibril using a 1-m scan range. Fibril fine structure is evident from the periodic steps indicated by the white circles. The periodic steps can also be seen in the three-dimensional image found in C. D, represent z-height line sections on areas of the fibril that are between the step of the periodic structure (␣) and through the step itself (␤). E, transmission electron microscopy of HEWL samples. Samples represented are oligomers, sonicated fibrils, and mature fibrils. Samples were deposited onto carbon-coated grids and stained with 2% uranyl acetate prior to visualization. These images are representative of three separate experiments. NOVEMBER 20, 2015 • VOLUME 290 • NUMBER 47 inhibited LTP at 3 h post-HFS (5.25 nmol, 103 Ϯ 5%, n ϭ 7, p Ͼ 0.05, compared with baseline, p Ͻ 0.05, compared with control LTP) without affecting baseline synaptic transmission (102 Ϯ 9%, n ϭ 4, p Ͼ 0.05, compared with pre-injection recording, Fig.  7, A and B). We chose this concentration of HEWL oligomers based on the in vitro toxicity data and pilot in vivo experiments. Furthermore, a 10-fold lower dose of HEWL oligomers (35 M in 15 l, 0.525 nmol) did not inhibit LTP (145 Ϯ 10%, n ϭ 4, p Ͻ 0.05, compared with baseline, p Ͼ 0.05, compared with control LTP, Fig. 7C). Interestingly, short fibrillar fragments generated by sonication of mature fibrils failed to inhibit LTP even when a much higher dose was injected (1.5 mM in 15 l, 22.5 nmol, 137 Ϯ 10%, n ϭ 4, p Ͼ 0.05 compared with control LTP, Fig. 6). Of note, prior to HFS there was no significant difference in baseline excitability between any groups shown in Fig. 6 as measured by EPSP amplitude (2.7 Ϯ 0.2, 2.5 Ϯ 0.3, and 2.4 Ϯ 0.3 mV for vehicle, HEWL oligomers, and sonicated fibrils, respectively) or stimulation intensity (6.6 Ϯ 0.4, 7.0 Ϯ 0.5, and 7.2 Ϯ 0.6 mA for vehicle, HEWL oligomers, and sonicated fibrils, respectively).

Discussion
The current study demonstrates that HEWL amyloid fibrils and oligomers, judiciously isolated via ultracentrifugation and ultrafiltration, are highly toxic to cultured PC12 cells of both differentiated and undifferentiated states and that their cell death responses are specific to each. Both prefibrillar soluble aggregates and mature fibrillar aggregates displayed cytotoxicity toward rat PC12 cells, with differentiated cells showing an enhanced level of resistance against oligomers. Isolated fractions also induce differing cell death pathways in PC12 undiff/diff with oligomers activating an apoptotic response, whereas fibrils generate secondary necrotic, non-apoptotic death.
The prolonged incubation of native HEWL protein resulted in the formation of oligomers, which preceded the conversion of mature fibrils, consistent with current knowledge on amyloid formation (18,32). When compared with the oligomers, the mature amyloid fibrils isolated in this experiment showed all of the hallmark spectroscopic signatures, relatively more pronounced Congo Red birefringence, increased ThT binding, and a shift in UV-visual Congo Red (Fig. 1, A-C). All fractions also showed differences in their hydrophobic region exposure when measured using ANS (Fig. 1D). These results indicate that the transition from native HEWL protein toward fibrillar aggregate coincides with a change in the protein environment, whereby the protein loses its native three-dimensional structural packing and displays increased exposure of hydrophobic residues.
We also examined the separate fractions using high-resolution AFM and TEM (Fig. 2, A-D). Mature fibrils appear as long single-stranded protofilaments, displaying a characteristic left handed "twist." Both sonicated and oligomer samples appear smaller in size yet distinctively unique from each other proving that no further elongation of the aggregate was achieved and that the size-based purification procedure was successful.
There is considerable debate regarding the true pathogenic role of oligomers and mature fibrils in neurodegenerative diseases. Early studies suggested that mature amyloid fibrils were the main pathogenic entities involved in amyloidosis (9,33). In contrast there is equally strong evidence that oligomeric species may also have a pathogenic role in these disorders (15,34,35). The perception of the role for amyloid fibrils has changed over recent years and it has been suggested that their true function may be that of an evolutionary protection mechanism employed by the body after the fundamental damage was carried out at an earlier phase by the soluble oligomers (36).
Our study demonstrates that HEWL oligomers and fibrils display toxicity toward PC12 undiff/diff cells and that soluble oligomers may not be the only toxic component involved in neurodegenerative disease. The higher levels of HEWL needed to induce cell death are representational of the normally non-  : 100 m). B and C, Alamar Blue emission viability results for PC12 cell lines treated with HEWL-separated fractions (Log scale). Undifferentiated (B) and differentiated (C) PC12 cells were exposed to varying concentrations of prefibrillar, and fibrillar aggregates then assayed using the fluorescent staining compound Alamar Blue. Monomeric HEWL was also examined and displayed no decrease in viability. The error bar indicates the values of mean Ϯ S.E. of three experiments and were calculated using GraphPad software. This protective role is consistent with earlier work examining amyloid injury in differentiated and undifferentiated human neurotypic SH-SY5Y cells (22). A possible reason for this is that phenotypic alterations that are a consequence of differentiation modify aggregate binding to the cell surface (38).
Using dual apoptotic and necrotic staining, we have shown that oligomeric treated PC12 undiff/diff cells display increased annexin V staining indicative of apoptosis, although these cells treated with mature fibrils show annexin V/PI double positive staining, suggesting a late apoptotic or secondary necrotic death. FACS was also conducted on the sonicated fibrillar samples, using identical settings. Sonicated fibrils used to treat PC12 undiff cells showed increased apoptotic staining, whereas the PC12 diff cells showed equal amounts of staining for both apoptosis and late necrosis, suggesting that these cells are entering the later stages of the apoptotic cycle of cell death (Fig.  4, A-H). PC12 undiff/diff cells exhibited cleaved PARP, an apoptotic indicator, when treated with oligomers, sonicated fibrils, and mature fibrils (Fig. 5). Monomeric treated cells show no signs of apoptotic death, which is supported by the evidence that these samples are not cytotoxic to cells. HEWL oligomers produced the highest level of apoptotic death, whereas sonicated fibrils and mature fibrils gave slightly lower levels of cleaved PARP. This data in collaboration with FACS analysis suggests cells treated with HEWL fractions, in particular oligomers, are dying through an apoptotic cell death pathway. The lower levels of apoptotic cell death seen in fibrillar treated cells  NOVEMBER 20, 2015 • VOLUME 290 • NUMBER 47

Differences in Cell Death Caused by Amyloid Aggregates
suggest that in addition to apoptosis, non-apoptotic cell death is also occurring.
We found that HEWL oligomers inhibited hippocampal LTP in vivo (Fig. 6), a process linked with memory loss and function. Despite the fact that these oligomers are not as potent at inhibiting LTP as, for example, A␤ oligomers (39), the data provide new insight into a role of protein misfolding in neurodegenerative diseases as well as potential benefits of immunotherapies   using conformation selective antibodies. We have previously found that insoluble amyloid fibrils formed by A␤ did not affect LTP in vivo (40). One reason for the lack of effect is that, due to diffusional restrictions, such large aggregates are not reaching hippocampal parenchyma from the injection site in the ventricle. To address this issue we used a sonication approach to produce short fragments of mature fibrils whereas maintaining general fibrillar structure (27). The observed difference in the abilities of oligomers and fibrillar fragments to inhibit LTP in vivo (Fig. 6) is in accord with our cell death findings (Fig. 3, B  and C), where sonicated fibrils were less toxic than oligomers in vitro, and probably reflects the different nature of the interaction between neurons and these different HEWL aggregates. Indeed, it has been demonstrated that the pro-apoptotic signaling pathway is involved in inhibition of LTP by A␤ oligomers (41). Alternatively, intracerebroventricular injected sonicated fibrils may not be reaching hippocampal parenchyma and rather interact with cells at the ventricular surface.
It is important to note that whereas oligomers have been shown to universally display inherent toxicity through a shared membrane permeabilization mechanism of pathogenesis (34,42), fibrils exhibit much greater diversity with regard to pathogenicity and unlike oligomers may employ a number of various mechanisms of cytotoxicity (43,44). Fibrils formed from differing proteins and indeed from the same peptides have shown dramatic differences in terms of conformational variation and cytotoxicity (18,(45)(46)(47). This may be responsible for the variations observed in terms of cytotoxic capability and could be a reason for the inconsistency displayed by numerous experiments examining the cytotoxic ability of fibrils.
The true toxic component involved in the cell/tissue damage symptomatic of neurodegenerative diseases is still a proverbial gray area in the world of amyloid pathology (7,48). Our results demonstrate that HEWL aggregates exhibit toxicity toward PC12 cells, and that the exact method of toxicity may be unique for individual fractions as well as dependent on the differentiated state of the cell. Oligomers inhibit LTP in rats in vivo, although sonicated fibrils appear to have no effect. Although both oligomers and fibrils have been shown to be damaging in vitro, in vivo experiments have yet to show that fibrils possess the ability to disrupt LTP, further highlighting the complicated matrix that surrounds the efficacy of these protein species. Knowledge of these subtle differences would be very important when designing anti-fibrillogenesis therapeutics. Despite this, our findings confirm that innocuous hen lysozyme can be engineered to produce both cytotoxic soluble prefibrillar aggregates and mature amyloid fibrils, further strengthening the claim that fibrillar conformation, and not the identity of the protein, is key to cellular toxicity and the underlying specific cell death mechanism.