Cross-talk between amyloidogenic proteins in type-2diabetes and Parkinson’s diseaseIstvan Horvatha and Pernilla Wittung-Stafshedea,1

aDepartment of Biology and Biological Engineering, Chalmers University of Technology, 412 96 Gothenburg, Sweden

Edited by Alan R. Fersht, Medical Research Council Laboratory of Molecular Biology, Cambridge, United Kingdom, and approved September 23, 2016(received for review June 29, 2016)

In type-2 diabetes (T2D) and Parkinson’s disease (PD), polypeptideassembly into amyloid fibers plays central roles: in PD, α-synuclein(aS) forms amyloids and in T2D, amylin [islet amyloid polypeptide(IAPP)] forms amyloids. Using a combination of biophysical meth-ods in vitro we have investigated whether aS, IAPP, and unpro-cessed IAPP, pro-IAPP, polypeptides can cross-react. Whereas IAPPforms amyloids within minutes, aS takes many hours to assembleinto amyloids and pro-IAPP aggregates even slower under thesame conditions. We discovered that preformed amyloids of pro-IAPP inhibit, whereas IAPP amyloids promote, aS amyloid forma-tion. Amyloids of aS promote pro-IAPP amyloid formation,whereas they inhibit IAPP amyloid formation. In contrast, mixingof IAPP and aS monomers results in coaggregation that is fasterthan either protein alone; moreover, pro-IAPP can incorporate aSmonomers into its amyloid fibers. From this intricate network ofcross-reactivity, it is clear that the presence of IAPP can accelerateaS amyloid formation. This observation may explain why T2D pa-tients are susceptible to developing PD.

α-synuclein | fluorescence | amyloid | atomic force microscopy | amylin

Parkinson’s disease (PD) is the second most common neuro-logical disorder and the most common movement disorder. It

is characterized by widespread degeneration of subcorticalstructures of the brain, especially dopaminergic neurons in thesubstantia nigra. These changes are coupled with bradykinesia,rigidity, and tremor, resulting in difficulties in walking and ab-normal gait in patients (1). The assembly process of the in-trinsically unstructured 140-residue protein α-synuclein (aS) intoamyloid fibers has been linked to the molecular basis of PD. aS isa major component of amyloid aggregates found in Lewy bodyinclusions, which are the pathological hallmark of PD, and du-plications, triplications, and point mutations in the aS gene arerelated to familial PD cases (2, 3). The exact function of aS isunknown, but it is suggested to be involved in synaptic vesiclerelease and trafficking, regulation of enzymes and transporters,and control of the neuronal apoptotic response (4, 5). aS ispresent at presynaptic nerve terminals (6–8) and, intriguingly,also in many cells outside the brain (e.g., red blood cells andpancreatic β-cells). aS can assemble via oligomeric intermediatesto amyloid fibrils under pathological conditions (9). Althoughsoluble aS oligomers have been proposed to be toxic (10, 11),work with preformed aS fibrils has demonstrated that the amy-loid fibrils themselves are toxic and can be transmitted from cellto cell and are also able to cross the blood–brain barrier (12–14).Type-2 diabetes (T2D) is another disease involving amyloid

formation. Here, the primary pathological characteristic is isletamyloid of the hormone amylin, also known as islet amyloidpolypeptide (IAPP), in pancreatic β-cells (15–18). The process ofislet amyloid formation (19–21) leads to pancreatic β-cell dys-function, cell death, and development of diabetes. IAPP (37residues, natively unfolded) is cosecreted with insulin after en-zymatic maturation of prohormones pro-IAPP (67 residues) andproinsulin in secretory granules. IAPP and insulin play roles incontrolling gastric emptying, glucose homeostasis, and in thesuppression of glucagon release. Although not understood on a

mechanistic level, impairment of prohormone processing hasbeen thought to play a role in initiation and progression of T2D(22, 23). Insulin and pro-IAPP (22, 24–26), but not proinsulin,can inhibit IAPP amyloid formation in vitro and in mice, sug-gesting that accumulation of unprocessed proinsulin may pro-mote IAPP amyloid formation (22, 24). Insulin-degradingenzyme (IDE) is a conserved metallopeptidase that can degradeinsulin and a variety of other small peptides including IAPP inthe pancreas (27, 28). Genome-wide association studies havelinked IDE to T2D (29, 30) and Ide mutant mice were found tohave impaired glucose-stimulated insulin secretion as well asincreased levels of IAPP, insulin, and, surprisingly, aS in pan-creatic islets (31, 32). Here, aS may be associated with insulinbiogenesis and exocytic release, as it was found to localize withinsulin-secretory granules in pancreatic β-cells (33). We recentlydemonstrated in vitro that IDE readily inhibits aS amyloid for-mation via C-terminal binding and, in parallel, IDE activity to-ward insulin and other small substrates increases (34, 35).Together, the key role of aS in PD and the inverse correlation

of impaired insulin secretion and increased aS levels in thepancreatic β-cells, imply that PD and T2D may be connected. Insupport, reports have suggested that patients with T2D arepredisposed toward PD (36, 37). For Alzheimer’s disease (AD),a direct link with T2D was found (15, 38). Amyloid fiber seeds ofthe AD peptide, amyloid-β, were shown to efficiently accelerateamyloid formation of IAPP in vitro (39, 40) and IAPP was part ofamyloid-β plaque found in mice brains (41). To address theunexplored question of cross-reactivity between the amyloido-genic peptides in PD and T2D, we here investigated cross-reactivity among aS, IAPP, and pro-IAPP using biophysicalmethods in vitro.

Significance

Protein assembly into ordered so-called amyloid fibers is aprocess that promotes several neurodegenerative disorders,such as Alzheimer’s and Parkinson’s disease (PD). Also type-2diabetes (T2D) is a disease involving amyloid formation, al-though it occurs in the pancreas. Since the protein that formsamyloids in PD, α-synuclein (aS), is also expressed in the pan-creas, we investigated whether it could affect aggregation ofthe peptide involved in T2D, and vice versa. Using in vitromethods and purified proteins, we here demonstrate that thetwo proteins cross-react and, importantly, the T2D amyloidprotein (both as monomer and amyloid seed) can accelerate aSaggregation. This result provides a possible explanation forwhy patients with T2D are more prone to getting PD.

Author contributions: I.H. and P.W.-S. designed research; I.H. performed research; I.H. andP.W.-S. analyzed data; and P.W.-S. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Freely available online through the PNAS open access option.1To whom correspondence should be addressed. Email: pernilla.wittung@chalmers.se.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1610371113/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1610371113 PNAS | November 1, 2016 | vol. 113 | no. 44 | 12473–12477

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

Results and DiscussionIndividual Protein Aggregation Reactions. IAPP, pro-IAPP, and aSall form amyloid fibers in vitro upon incubation at pH 7, 37 °C.With agitation, aS (70 μM) aggregates with a lag time of about20 h (Fig. 1A), whereas pro-IAPP (140 μM) has a longer lag timeof about 30 h (Fig. 1B). For IAPP, amyloid formation is muchfaster, with lag times of 5–10 min for 5 μM and 1–2 min for10 μM IAPP (Fig. 1C). One may speculate that variation in theintrinsic amyloid formation speed is dependent on the ratio ofamyloid-forming segment size over total length of the peptide(which will be highest for IAPP), as fluctuating peptidesextending from the amyloid fiber core may slow down secondarynucleation. The resulting amyloid fibers have cross-sections of

about 3–4 nm for IAPP and pro-IAPP, but 7–9 nm for aS (Fig. 1D–F and Table 1). Reported amyloid fiber dimensions for IAPPare ∼5 nm for single fibers (42–44). For aS, the just-releasedsolid-state NMR structure of full-length aS amyloid has a coredimension of ∼4.5 nm, surprisingly, with its β-strands arranged ina Greek Key motif (45). This information implies that our largeraS amyloids may be assemblies of more than one aS proto-fibril.We found that in all three cases (aS, IAPP, and pro-IAPP),preformed amyloid fiber seeds, speed up aggregation of mono-mers of the same protein (Fig. 1 B and C; not shown for IAPP asthe seeded reaction is too fast to capture by manual mixing).

Cross-Seeding of Amyloid Formation. Next we tested the possibilityof cross-seeding of one protein’s amyloid formation reaction withpreformed amyloid seeds of another protein. We found thatpreformed IAPP amyloid seeds speed up aS amyloid formation(Fig. 2A), and preformed aS amyloid seeds speed up pro-IAPPamyloid formation (Fig. 2B), whereas preformed aS amyloidseeds inhibit IAPP amyloid formation (Fig. 2 C and G), andpreformed pro-IAPP amyloid seeds inhibit aS amyloid formation(Fig. 2 D and H). The resulting amyloid fibers (from reactions inFig. 2 A and B) have the dimensions matching that of the ag-gregating protein when aggregating alone (Table 1 and Fig. 2 Eand F). This implies that added amyloid fiber seeds (when con-taining a protein that has an intrinsically faster amyloid forma-tion reaction than the monomeric protein it is mixed with)trigger the “normal” aggregation reaction. In contrast, whenamyloid fiber seeds of a protein that forms amyloid fibers slowerthan the protein it is mixed with (i.e., aS amyloid fiber seedsadded to IAPP monomers, and pro-IAPP amyloid fiber seedsadded to aS monomers; Fig. 2 C and D) are added, the outcomeis instead reduction of amyloid formation.

Coaggregation of aS and Pro-IAPP. When monomers of pro-IAPPare added to aS aggregation reactions, the pro-IAPP monomersinhibit aS aggregation in a concentration-dependent manner(Fig. 3A). Notably, under these conditions, pro-IAPP itself doesnot form amyloid fibers. Thus, like the pro-IAPP seeds (Fig. 2D),pro-IAPP monomers block aS from aggregation and the atomicforce microscopy (AFM) images lack amyloids or oligomers(Fig. 3D). In contrast, if aS monomers are added to high con-centrations of pro-IAPP (concentration that aggregates with lagtime of 30 h), the resulting thioflavin T (ThT) emission becomessignificantly increased but the lag time remains the same (Fig. 3B).This finding implies that aS monomers become incorporated intothe pro-IAPP amyloids without effect on formation speed butallowing for increased ThT emission. The increased ThT emissionmay be due to better ThT binding to resulting mixed amyloid fi-bers or simply more amyloids are made. Notably, as aS alone (at70 μM) would have a lag time of 20 h, and such a process is lackingin Fig. 3B, interactions of aS with pro-IAPP (monomers, oligo-mers, and protofibrils) must overrule independent aS amyloidformation. This conclusion correlates with the finding that pro-IAPP monomers (at concentrations not readily assembling intofibers) added to aggregating aS monomers, result in aS amyloid

0,0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,80,0

0,2

0,4

0,6

0,8

1,0

Nor

mal

ized

fluo

resc

ence

Time (h)

300nm

0 10 20 30 400,0

0,2

0,4

0,6

0,8

1,0

Nor

mal

ized

fluo

resc

ence

time (h)

300nm

0 10 20 30 40 50 60 700,0

0,2

0,4

0,6

0,8

1,0

Nor

mal

ized

fluo

resc

ence

time (h)

300nm

A

B

C

D

E

F

Fig. 1. Amyloid fiber formation of 70 μM aS alone (black) and in thepresence of 0.7 μM preformed amyloid fiber seeds (red) (A), 140 μM pro-IAPPalone (black) and with 7 μM preformed amyloid fiber seeds (red) (B), andIAPP at 5-μM (black) and 10-μM concentration (red) (C) as probed by ThTemission. AFM images of resulting amyloid fibers are shown for aS (D), pro-IAPP (E), and IAPP (F).

Table 1. Summary of key properties of amyloids formed in the different aS/IAPP/pro-IAPP mixtures based on AFManalysis

Amyloid fiber aS Pro-IAPP IAPPIAPP seeds

to aSaS seeds topro-IAPP

IAPP +aS monomers

Pro-IAPP +aS monomers

Fiber height (nm) 6–10 2.5–4.5 2.5–4.5 7–10 4–6 2–3 8–10Morphology Long >1 μm Short <1 μm Short

clusteringLong

clusteringShort

clusteringLong needle-

like softShort

clustering

Amyloid fiber height is a rough estimate that is based on Zmeasurements of 15–20 individual fibers in each case (distributions shownin Fig. S3). Morphology is based on visual inspection of a number of AFM images for each sample and reports the most commonfeatures observed throughout.

12474 | www.pnas.org/cgi/doi/10.1073/pnas.1610371113 Horvath and Wittung-Stafshede

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

inhibition (Fig. 3A). In essence, therefore, aS amyloid formationbecomes slowed down when monomers are added to high con-centrations of pro-IAPP and instead pro-IAPP is capable ofrecruiting the aS monomers into its amyloid fibers. AFM imagesof these mixtures imply that thicker fibrils are formed (8–10 nm)compared with the fibrils of pro-IAPP alone (Fig. 3E and Table 1).

Coaggregation of aS and IAPP. Similar to the mixing of aS and pro-IAPP monomers, we found increased ThT emission for IAPPamyloid formation reactions when monomeric aS was added(Fig. 3C). Notably, because IAPP amyloid formation occurswithin 10–20 min, the added aS will not aggregate at all by itselfunder these conditions. Intriguingly, not only did the ThTemission increase when aS monomers were added to IAPPmonomers, also at a ratio of aS to IAPP of 2:1 (5 μM IAPP), butthe lag time was also shortened to a few minutes (similar to whatwas observed when IAPP concentration was increased, compareFig. 1C with Fig. 3C). The increased ThT emission and reducedlag time strongly argue that IAPP and aS readily coaggregate andthat such reactions are favorable with respect to both speed andamount. In control experiments of IAPP and aS monomersmixed in the plate reader, we found the endpoint ThT signal(after 30 min) to follow the same trend as observed for the samemixtures in kinetic cuvette experiments (Fig. S1).AFM images of these mixtures suggest that the resulting am-

yloid fibers have a thickness of only ∼2 nm, which is thinner thanwhat we observe for the two homoprotein amyloids (Table 1).The resulting amyloid fibers in the aS–IAPP mixture appearneedle-like with rough, but sharp, edges according to thoroughAFM image inspection (Fig. 3F; we note that AFM quantifica-tion, apart from the Z height, is uncertain). In contrast to thetypical amyloids formed by IAPP and aS individually, the amy-loid fibers formed in mixtures of the two monomers are veryfragile: the AFM tip easily destroys these fibers. Therefore, thetapping mode imaging was possible only at instrument settingsthat corresponded to the exposure to weak forces. The differ-ences in amyloid fiber morphology, together with the altered

reaction kinetics, strongly argue for the formation of coassembledamyloids of aS and IAPP.Addition of aS monomers to aggregating IAPP after the ThT

emission plateau was reached (i.e., at 20 min; IAPP amyloids al-ready formed) did not result in further ThT increase within thetime frame of hours (Fig. S2). This result demonstrates that theprocess in Fig. 3C involves coaggregation of monomers, or earlyassemblies of the two proteins, and is not due to initial formation ofIAPP amyloids that subsequently template aS aggregation. Label-ing experiments will address the arrangement of IAPP and aSmonomers within the heterologous amyloids (work in progress).

ConclusionsAlthough most studies of amyloid formation have focused onindividual disease-specific peptides, interactions of peptides as-sociated with different amyloid diseases may modulate amyloidformation pathways and pathogenicity (38). IAPP and amyloid-βshare 50% sequence similarity and were found to cross-reactboth in vitro (40) and in vivo (41). Also peptides structurally andphysiologically unrelated, such as IAPP and the amyloidogenicdeterminant of the prion protein (PrP), can cross-react intomorphologically distinct amyloid aggregates (46) and amyloid-βfibers can seed aS amyloid formation (47). aS is present in thepancreas (33), its levels are inversely correlated with that of IDE(31), and IDE inhibits aS amyloid formation in vitro (34, 35).Despite these facts, putative cross-talk between aS and processedand/or unprocessed IAPP has not been addressed previously.The importance of possible cross-reactivity between these pro-teins is also underscored by recent reports demonstrating thatIAPP is present in brain cells and may form amyloidogenic de-posits there, although the IAPP peptide appeared to be synthe-sized elsewhere (41, 48–50).In addition to the intriguing coaggregation detected among

monomers, readily forming heterologous amyloid fibers, a trend wediscovered was that preformed amyloid fiber seeds from a faster/slower aggregating protein will accelerate/inhibit aggregation of anintrinsically slower/faster aggregating protein (summarized inFig. 4). This trend may be explained by a built-in “memory” in the

0 20 40 600,0

0,2

0,4

0,6

0,8

1,0

Nor

mal

ized

fluo

resc

ence

Time (h)

500nm

0 10 20 30 40 50 60 700,0

0,2

0,4

0,6

0,8

1,0

Nor

mal

ized

fluo

resc

ence

Time (h)

300nm

0 10 20 30 400

10000

20000

30000

40000

50000

Tht f

luor

esce

nce

(a.u

.)

Time (h)

A B

E

2.0µm

0,0 0,1 0,2 0,3 0,4 0,5 0,60

50

100

150

Tht F

luor

esce

nce

(a.u

.)

Time (h)

500nm

C D

F G H

Fig. 2. Cross-seeding of amyloid fiber formation of aS by IAPP amyloid seeds: 70 μM aS alone (black), with 3.5 μM (red) and 7 μM (blue) IAPP amyloid seedsadded (A) and of pro-IAPP by aS amyloid seeds: 140 μM pro-IAPP alone (black) and with 7 μM (red) aS amyloid seeds added (B). Inhibition of amyloid fiberformation of IAPP by aS amyloid seeds: 5 μM IAPP alone (black) and with 1.4 μM aS amyloid seeds (red) (C) and of aS by pro-IAPP amyloid seeds: 70 μM aSalone (black), with 7 μM (red) and 14 μM (blue) pro-IAPP fibrils (D). AFM images of the resulting amyloid fibers in A and B (shown in E and F) and dem-onstration of inhibitory effects in C and D (shown in G and H). In G, the large structures observed are not amyloids, but likely aggregates of protein.

Horvath and Wittung-Stafshede PNAS | November 1, 2016 | vol. 113 | no. 44 | 12475

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

amyloid seed (envisioned as fine structural differences) thatbecomes transmitted into newly formed amyloids and therebydictates the aggregation speed of the other protein. With morestructural studies of amyloids at atomic resolution, such micro-scopic variations may eventually be revealed.Finally, our findings provide a simple justification for why T2D

is a risk factor for PD (as IAPP amyloid seeds, present in T2D,trigger aS amyloid formation), whereas patients with PD arenot prone to getting T2D (aS amyloid seeds, present in PD, in-hibit IAPP amyloid formation) (36, 37). Because aS and IAPPmay come in contact with each other both in pancreatic β-cells(33) and in various brain cells (48–50), and amyloidogenic as-semblies may spread from cell to cell (51), in vivo studies are de-sired to reveal the biological consequences of the cross-reactivitydetected here.

Materials and MethodsProteins. aS was purified as described previously (52). The protein was freezedried and dissolved in the appropriate buffer before use, followed by fil-tration through a 0.2-μm filter. Pro-IAPP was purchased from Severn Biotech(product no. SBP0022). IAPP was obtained from Calbiochem (product no. 05–23-2540). Both pro-IAPP and IAPP were treated with hexafluoroisopropanol(105228; Sigma-Aldrich) before use to dissolve any preexisting aggregates.Amyloid seeds of each protein were prepared by incubating aggregatedprotein samples (i.e., the products of a normal amyloid formation reaction)for 10 s in a sonicating bath. The incubation time has to be optimized fordifferent sonicators because the power output can vary depending on themanufacturer (extended sonication can completely destroy fibrils). Amyloidseeds were characterized by AFM before use.

Amyloid Formation Assay. aS and pro-IAPP amyloid formation reactions wereconducted in a plate reader (Fluostar Optima; BMG Labtech) in TBS (0.05 M TrisHCl buffer, pH 7.4with 0.15MNaCl, 93318; Sigma-Aldrich) buffer in the presenceof 20 μM ThT (T3516; Sigma-Aldrich) at 37 °C with agitation and a 2-mm glassbead in each well. Samples were typically incubated for 60 h and fluorescencewas measured every 20 min. Due to its faster aggregation kinetics, IAPP amyloidformation was monitored in a 1-cm cuvette using a Cary Eclipse spectropho-tometer (with a magnetic stir bar). IAPP aggregation was initiated by pipetting asmall volume of IAPP from the 2.5 mM stock (in hexafluoroisopropanol) into thereaction buffer (10 μM ThT in TBS) followed by fluorescence measurementsevery 30 s. Comparison of endpoint ThT values for IAPP amyloid formation in acuvette and in the plate reader suggested that the IAPP aggregation process wassimilar in both setups. For both plate reader and cuvette experiments, ThT wasexcited at 440 nm and emission was recorded at 480 nm. All reported ThT curvesare averages of triplicate experiments, and all experiments were performed (intriplicate) three independent times. (For all of the mixture experiments ofmonomers/amyloid seeds of two proteins in various combinations, a range ofconcentrations was tested. The included figures show representative data thatillustrate the trends we discovered from a number of experimental conditions.)

AFM. Aggregated samples were diluted into Milli-Q water (10–20 times) anddeposited on freshly cleaved mica. After 10 min, the mica was rinsed with fil-tered Milli-Q water and dried under a gentle nitrogen stream. Images wererecorded on an NTEGRA Prima setup (NT-MDT) using a gold-coated singlecrystal silicon cantilever (NSG01, spring constant of ∼5.1 N/m; NT-MDT) and aresonance frequency of ∼180 kHz. The 512-pixel images were acquired with a0.5-Hz scan rate. Images were analyzed using the WSxM 5.0 software (53).

ACKNOWLEDGMENTS. Funding is acknowledged from the Knut and AliceWallenberg Foundation, the Swedish Research Council, and the ChalmersFoundation.

1. Chen AY, Wilburn P, Hao X, Tully T (2014) Walking deficits and centrophobism in anα-synuclein fly model of Parkinson’s disease. Genes Brain Behav 13(8):812–820.

2. Spillantini MG, et al. (1997) Alpha-synuclein in Lewy bodies. Nature 388(6645):839–840.3. Polymeropoulos MH, et al. (1997) Mutation in the alpha-synuclein gene identified in

families with Parkinson’s disease. Science 276(5321):2045–2047.4. Dev KK, Hofele K, Barbieri S, Buchman VL, van der Putten H (2003) Part II: Alpha-

synuclein and its molecular pathophysiological role in neurodegenerative disease.Neuropharmacology 45(1):14–44.

5. Lassen LB, Reimer L, Ferreira N, Betzer C, Jensen PH (2016) Protein partners of alpha-synuclein in Health and Disease. Brain Pathol 26(3):389–397.

6. Maroteaux L, Campanelli JT, Scheller RH (1988) Synuclein: A neuron-specific proteinlocalized to the nucleus and presynaptic nerve terminal. J Neurosci 8(8):2804–2815.

7. Eliezer D, Kutluay E, Bussell R, Jr, Browne G (2001) Conformational properties of alpha-synuclein in its free and lipid-associated states. J Mol Biol 307(4):1061–1073.

8. Iwai A, et al. (1995) The precursor protein of non-A beta component of Alzheimer’s diseaseamyloid is a presynaptic protein of the central nervous system. Neuron 14(2):467–475.

9. Uversky VN (2007) Neuropathology, biochemistry, and biophysics of alpha-synucleinaggregation. J Neurochem 103(1):17–37.

10. Xu J, et al. (2002) Dopamine-dependent neurotoxicity of alpha-synuclein: Amechanism for selective neurodegeneration in Parkinson disease. Nat Med 8(6):600–606.

11. Gosavi N, Lee HJ, Lee JS, Patel S, Lee SJ (2002) Golgi fragmentation occurs in the cellswith prefibrillar alpha-synuclein aggregates and precedes the formation of fibrillarinclusion. J Biol Chem 277(50):48984–48992.

IAPP amyloidforma�on

5-10 min

αS amyloidforma�on

~20 h

pro-IAPP amyloidforma�on

30-40 h

fiber seeds fiber seedsfiber seeds

FASTERINHIBITIONFASTER

fiber seedsINHIBITION

monomers no effecton speed

monomers

mix of monomers

INHIBITION

FASTERCO-AGGREGATION

FASTER

Fig. 4. Summarizing scheme of the discovered effects on amyloid forma-tion reactions among IAPP, aS, and pro-IAPP monomers and amyloid seeds.The statement “faster makes faster and slower makes slower” is a commonconclusion among the amyloid seeding reactions.

0 10 20 30 40 50 60 700

2000

4000

6000

8000

10000

12000

Tht

fluo

resc

ence

(a.u

.)

Time (h)

0,00 0,20 0,40 0,60 0,800

50

100

150

200

250

Tht f

luor

esce

nce

(a.u

.)

Time (h)2.0µm

1.0µm

0 10 20 30 40 50 60

5000

10000

15000

20000

25000

30000

35000

Tht f

luor

esce

nce

(a.u

.)

Time (h)

A

B

C

D

F

1.0µm

E

Fig. 3. Coaggregation upon mixing of monomers. (A) Effects on aS amyloidformation upon the addition of pro-IAPP monomers [70 μM aS alone (black)and with 17 μM (red) and 35 μM (blue) proIAPP added]. (B) Effects on pro-IAPP amyloid formation upon the addition of aS monomers [140 μM proIAPPalone (black) and with 35 μM (red) or 70 μM (blue) aS]. (C) Effects on IAPPamyloid formation upon the addition of aS monomers [5 μM IAPP alone(black) and with 10 μM aS (red)]. AFM images of the resulting amyloid fibersare shown in D (for A), E (for B), and F (for C).

12476 | www.pnas.org/cgi/doi/10.1073/pnas.1610371113 Horvath and Wittung-Stafshede

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1

12. Peelaerts W, et al. (2015) α-Synuclein strains cause distinct synucleinopathies afterlocal and systemic administration. Nature 522(7556):340–344.

13. Luk KC, et al. (2012) Pathological α-synuclein transmission initiates Parkinson-likeneurodegeneration in nontransgenic mice. Science 338(6109):949–953.

14. Paumier KL, et al. (2015) Intrastriatal injection of pre-formed mouse α-synuclein fibrilsinto rats triggers α-synuclein pathology and bilateral nigrostriatal degeneration.Neurobiol Dis 82:185–199.

15. Brender JR, Salamekh S, Ramamoorthy A (2012) Membrane disruption and earlyevents in the aggregation of the diabetes related peptide IAPP from a molecularperspective. Acc Chem Res 45(3):454–462.

16. DeToma AS, Salamekh S, Ramamoorthy A, Lim MH (2012) Misfolded proteins inAlzheimer’s disease and type II diabetes. Chem Soc Rev 41(2):608–621.

17. Kotler SA, Walsh P, Brender JR, Ramamoorthy A (2014) Differences between amyloid-βaggregation in solution and on themembrane: Insights into elucidation of themechanisticdetails of Alzheimer’s disease. Chem Soc Rev 43(19):6692–6700.

18. Patel HR, Pithadia AS, Brender JR, Fierke CA, Ramamoorthy A (2014) In search ofaggregation pathways of IAPP and other amyloidogenic proteins: Finding answersthrough NMR spectroscopy. J Phys Chem Lett 5(11):1864–1870.

19. Cao P, et al. (2013) Islet amyloid: From fundamental biophysics to mechanisms ofcytotoxicity. FEBS Lett 587(8):1106–1118.

20. Cao P, Abedini A, Raleigh DP (2013) Aggregation of islet amyloid polypeptide: Fromphysical chemistry to cell biology. Curr Opin Struct Biol 23(1):82–89.

21. Westermark GT, Gebre-Medhin S, Steiner DF, Westermark P (2000) Islet amyloiddevelopment in a mouse strain lacking endogenous islet amyloid polypeptide (IAPP)but expressing human IAPP. Mol Med 6(12):998–1007.

22. Jaikaran ET, Nilsson MR, Clark A (2004) Pancreatic beta-cell granule peptides formheteromolecular complexes which inhibit islet amyloid polypeptide fibril formation.Biochem J 377(Pt 3):709–716.

23. Marzban L, et al. (2006) Impaired NH2-terminal processing of human proislet amyloidpolypeptide by the prohormone convertase PC2 leads to amyloid formation and celldeath. Diabetes 55(8):2192–2201.

24. Wang H, Raleigh DP (2014) The ability of insulin to inhibit the formation of amyloidby pro-islet amyloid polypeptide processing intermediates is significantly reduced inthe presence of sulfated glycosaminoglycans. Biochemistry 53(16):2605–2614.

25. Krampert M, et al. (2000) Amyloidogenicity of recombinant human pro-islet amyloidpolypeptide (ProIAPP). Chem Biol 7(11):855–871.

26. Susa AC, et al. (2014) Defining the molecular basis of amyloid inhibitors: Human isletamyloid polypeptide-insulin interactions. J Am Chem Soc 136(37):12912–12919.

27. Kirschner RJ, Goldberg AL (1983) A high molecular weight metalloendoprotease fromthe cytosol of mammalian cells. J Biol Chem 258(2):967–976.

28. Duckworth WC, Bennett RG, Hamel FG (1998) Insulin degradation: Progress and po-tential. Endocr Rev 19(5):608–624.

29. Saxena R, et al.; Diabetes Genetics Initiative of Broad Institute of Harvard and MIT,Lund University, and Novartis Institutes of BioMedical Research (2007) Genome-wideassociation analysis identifies loci for type 2 diabetes and triglyceride levels. Science316(5829):1331–1336.

30. Scott LJ, et al. (2007) A genome-wide association study of type 2 diabetes in Finnsdetects multiple susceptibility variants. Science 316(5829):1341–1345.

31. Steneberg P, et al. (2013) The type 2 diabetes-associated gene ide is required forinsulin secretion and suppression of α-synuclein levels in β-cells. Diabetes 62(6):2004–2014.

32. Farris W, et al. (2003) Insulin-degrading enzyme regulates the levels of insulin,amyloid beta-protein, and the beta-amyloid precursor protein intracellular domainin vivo. Proc Natl Acad Sci USA 100(7):4162–4167.

33. Geng X, et al. (2011) α-Synuclein binds the K(ATP) channel at insulin-secretorygranules and inhibits insulin secretion. Am J Physiol Endocrinol Metab 300(2):E276–E286.

34. Sharma SK, Chorell E, Wittung-Stafshede P (2015) Insulin-degrading enzyme is acti-vated by the C-terminus of α-synuclein. Biochem Biophys Res Commun 466(2):192–195.

35. Sharma SK, et al. (2015) Insulin-degrading enzyme prevents α-synuclein fibril for-mation in a nonproteolytical manner. Sci Rep 5:12531.

36. Xu Q, et al. (2011) Diabetes and risk of Parkinson’s disease. Diabetes Care 34(4):910–915.

37. Santiago JA, Potashkin JA (2014) System-based approaches to decode the molecularlinks in Parkinson’s disease and diabetes. Neurobiol Dis 72(Pt A):84–91.

38. Morales R, Moreno-Gonzalez I, Soto C (2013) Cross-seeding of misfolded proteins:Implications for etiology and pathogenesis of protein misfolding diseases. PLoSPathog 9(9):e1003537.

39. O’Nuallain B, Williams AD, Westermark P, Wetzel R (2004) Seeding specificity inamyloid growth induced by heterologous fibrils. J Biol Chem 279(17):17490–17499.

40. Seeliger J, et al. (2012) Cross-amyloid interaction of Aβ and IAPP at lipid membranes.Angew Chem Int Ed Engl 51(3):679–683.

41. Oskarsson ME, et al. (2015) In vivo seeding and cross-seeding of localized amyloidosis:A molecular link between type 2 diabetes and Alzheimer disease. Am J Pathol 185(3):834–846.

42. Cho WJ, Jena BP, Jeremic AM (2008) Nano-scale imaging and dynamics of amylin-membrane interactions and its implication in type II diabetes mellitus. Methods CellBiol 90:267–286.

43. Sumner Makin O, Serpell LC (2004) Structural characterisation of islet amyloid poly-peptide fibrils. J Mol Biol 335(5):1279–1288.

44. Goldsbury CS, et al. (1997) Polymorphic fibrillar assembly of human amylin. J StructBiol 119(1):17–27.

45. Tuttle MD, et al. (2016) Solid-state NMR structure of a pathogenic fibril of full-lengthhuman α-synuclein. Nat Struct Mol Biol 23(5):409–415.

46. Gal N, et al. (2013) Lipid bilayers significantly modulate cross-fibrillation of two dis-tinct amyloidogenic peptides. J Am Chem Soc 135(36):13582–13589.

47. Tsigelny IF, et al. (2008) Mechanisms of hybrid oligomer formation in the patho-genesis of combined Alzheimer’s and Parkinson’s diseases. PLoS One 3(9):e3135.

48. Srodulski S, et al. (2014) Neuroinflammation and neurologic deficits in diabetes linkedto brain accumulation of amylin. Mol Neurodegener 9:30.

49. Schultz N, Byman E, Fex M, Wennström M (2016) Amylin alters human brain pericyteviability and NG2 expression. J Cereb Blood Flow Metab pii: 0271678X16657093.

50. Jackson K, et al. (2013) Amylin deposition in the brain: A second amyloid in Alzheimerdisease? Ann Neurol 74(4):517–526.

51. Espargaró A, Busquets MA, Estelrich J, Sabate R (2016) Key points concerning amyloidinfectivity and prion-like neuronal invasion. Front Mol Neurosci 9:29.

52. Chorell E, et al. (2015) Bacterial chaperones CsgE and CsgC differentially modulatehuman α-synuclein amyloid formation via transient contacts. PLoS One 10(10):e0140194.

53. Horcas I, et al. (2007) WSXM: A software for scanning probe microscopy and a tool fornanotechnology. Rev Sci Instrum 78(1):013705.

Horvath and Wittung-Stafshede PNAS | November 1, 2016 | vol. 113 | no. 44 | 12477

BIOPH

YSICSAND

COMPU

TATIONALBIOLO

GY

Dow

nloa

ded

by g

uest

on

Nov

embe

r 7,

202

1