Biochimica et Biophysica Acta 1854 (2015) 249–257

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbapap

Structural characterization of amyloid fibrils from the human parathyroid hormone Mohanraj Gopalswamy a,b,1, Amit Kumar a,b,1, Juliane Adler c, Monika Baumann a,b, Mathias Henze a,b, Senthil T. Kumar d, Marcus Fändrich d, Holger A. Scheidt c, Daniel Huster c, Jochen Balbach a,b,⁎ a

Institut für Physik, Biophysik, Martin-Luther-Universität Halle—Wittenberg, D-06120 Halle, Germany Mitteldeutsches Zentrum für Struktur und Dynamik der Proteine (MZP), Martin-Luther-Universität Halle—Wittenberg, D-06120 Halle, Germany Institut für Medizinische Physik und Biophysik, Universität Leipzig, D-04107 Leipzig, Germany d Institut für Pharmazeutische Biotechnologie, Universität Ulm, D-89081 Ulm, Germany b c

a r t i c l e

i n f o

Article history: Received 16 October 2014 Received in revised form 2 December 2014 Accepted 21 December 2014 Available online 30 December 2014 Keywords: Human parathyroid hormone Amyloid fibrils EGCG Electron microscopy NMR spectroscopy MALDI-TOF mass spectrometry

a b s t r a c t Amyloid deposits are common in various tissues as a consequence of misfolded proteins. However, secretory protein and peptides are often stored in membrane coated granules as functional amyloids. In this article, we present a detailed characterization of in vitro generated amyloid fibrils from human parathyroid hormone (hPTH(1–84)). Fully mature fibrils could be obtained after a short lag phase within less than one hour at 65 °C. These fibrils showed all characteristic of a cross-β structure. Protease cleavage combined with mass spectrometry identified the central region of the peptide hormone involved in the fibril core formation. EGCG, an inhibitor of amyloid fibril formation, showed binding to residues in the peptide monomers corresponding to the later fibril core and thus explaining the inhibition of the fibril growth. Conformational and dynamic studies by solid-state NMR further corroborated the cross-β core of the fibrils, but also identified highly mobile segments with a random coil structure not belonging to the rigid fibril core. © 2014 Elsevier B.V. All rights reserved.

1. Introduction Failure of proteins to reach a stable native structure may result in the population of misfolded or aggregated species [1–3]. Amyloid deposits are common in particular polypeptide producing tissues and typically represent a characteristic feature of diseases affecting these tissues including Alzheimer's or Parkinson's disease and type II diabetes [4]. On the other hand, because of their optimized packing properties, amyloids are found in nature for storage of proteins and peptides. Such amyloids have a non-pathological biological function and are called ‘functional amyloids’. These are present in Escherichia coli [5], silkworms [6], fungi [7], and mammalian skin [8]. Some functional amyloids of fungal prions are involved in prion replication, and the amyloid protein Pmel17 is involved in mammalian skin pigmentation. Formation of functional amyloids is also common to the secretory proteins such as hormones [9]. Hormones and secretory proteins stored in membrane coated secretory granules over long time periods at very high concentration before release into the blood stream occurs. Often, this high concentration

⁎ Corresponding author at: Institut für Physik, Biophysik, Martin-Luther-Universität Halle-Wittenberg, Betty-Heimann-Str. 7, D-06120 Halle (Saale), Germany. Tel.: +49 345 5528550; fax: +49 345 5527161. E-mail address: [email protected] (J. Balbach). 1 These authors contributed equally.

http://dx.doi.org/10.1016/j.bbapap.2014.12.020 1570-9639/© 2014 Elsevier B.V. All rights reserved.

leads to self-association of the peptide chains. It has been reported that peptide and hormones are found in the pituitary secretory granules of the endocrine system in the form of functional amyloids [9]. Few hormones including insulin and glucagon tend to form amyloid fibrils under certain in vitro conditions [10]. The basic architecture of selfassembled amyloid fibrils is the cross-β structure [11]. The cross-β motif is composed of intermolecular β-sheets arranged along the fibril axis with the β-strands aligned perpendicularly to the axis of fibrils. The amyloid-like cross-β-rich conformation was also observed for functional amyloids in secretory granules of the endocrine system [9]. Human parathyroid hormone is secreted from the parathyroid glands if the Ca2+ level in the blood drops or the blood phosphate concentration increases [12]. Human parathyroid hormone is translated as a 115 residues comprising pre-pro protein. The 25 N-terminal residues (pre sequence) are required as signal for efficient transport to the endoplasmic reticulum [13,14]. This signal sequence is rapidly [15] cleaved off by a signal peptidase and the resulting pro-hPTH peptide subsequently transferred to the Golgi apparatus [16]. The 6 residues of the pro sequence at the N-terminus are proteolytically removed [17] and the mature hPTH(1–84) is packaged and stored in secretory granules until the release into the blood. It has been reported that peptide and hormones found in the pituitary secretory granules of the endocrine system are stored in the form of functional amyloids [9]. Few hormones including insulin [10], glucagon [10], and PTH [18] tend to form amyloid

250

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

fibrils under certain in vivo and in vitro conditions [9]. Often, mutations result in the accumulation of a conformationally defective protein in the ER which contributes to diseases such as Alzheimer's, Parkinson's, Huntington's, or type 1 diabetes [19–21]. C18R mutation in the pre sequence of pre-pro-PTH disrupts the hydrophobic core of the signal sequence leading to intracellular accumulation and causes the familial isolated hypoparathyroidism (FIH) [22]. In the present study, we test the hypothesis whether hPTH(1–84) undergoes fibril formation under in vitro conditions. Indeed, we could explore in detail amyloid fibril formation of hPTH(1–84) and elucidate its structural properties. The fibrils are curvilinear, long structures and exhibited the characteristic of repeating cross-β motif. Primarily, the N-terminal residues 25R–37L form the cross-β core structure of the fibrils. Most of the diseases related to the secreted proteins are caused by accumulation of misfolded proteins and the progressive degeneration of the associated tissues. An amyloid-like structure of peptide and protein hormones could explain most of their properties. 2. Material and methods

the buffer at pH 7.0 in 20 mM Na2HPO4 and 25 mM NaCl. Subsequently, this sample was incubated at 37 °C for 24 h and supernatant was collected after centrifugation at 30,000 rpm, 4 °C for 30 min. Additionally, quick release of monomer was tested by mixing the buffer followed by immediate separation of solution and fibrils by centrifugation. The sample volume was kept constant throughout the experiment. The concentration of released hPTH(1–84) in the supernatant was monitored by absorption spectroscopy at 280 nm. 2.6. Mass spectrometry and fibril core identification The fibrils were prepared as mentioned in the previous section. These fibrils were mixed with chymotrypsin in the 1:100, enzyme to fibril (w/w) [27] ratio in 25 mM Tris–Cl, 50 mM NaCl at pH 7.8. Tosyllysine chloromethyl ketone hydrochloride (TLCK) treated MS grade chymotrypsin was used. The mixture was incubated at 37 °C and samples were taken at different time points. The fibrils were dissociated into monomers adding hexafluoro-2-propanol in 1:1 ratio by volume. Samples were analyzed by MALDI-TOF or ESI-MS-MS. For the mass spectrometry, samples were desalted by Pierce C18 Tips prior to analysis.

2.1. Protein expression and purification 2.7. Transmission electron microscopy (TEM) Human PTH(1–84) was purified by recombinant expression following the previously reported procedure [23,24]. In brief, pET SUMO adapt vector containing hPTH(1–84) with C-terminal His-tags was transformed in E. coli BL21codon + cells. Isotope labeling was achieved in M9 media with 15N NH4Cl as nitrogen source. Ni-NTA purified samples were mixed with SUMO protease (1:100 ratio) 50–150 μg/ml. Cleaved products were further purified by S-75 gel filtration chromatography. Aβ peptides were recombinantly expressed and purified according to reported procedure [25].

The samples were diluted to 50 μM and a 5 μl droplet of hPTH(1–84) fibril samples was pipetted onto a formvar carbon-coated copper grid (Ted Pella Inc.) and washed three times with 50 μl of water drops. The water was carefully removed with filter paper between washes. Grids were then stained with 50 μl of 2% (w/v) uranyl acetate which was then removed and air dried. Specimens were examined using Zeiss 900 transmission electron microscope. The microscope was operated at an acceleration voltage of 80 kV.

2.2. Thioflavin T (ThT) kinetic and hPTH(1–84) fibril preparation

2.8. Inhibition of fibril growth and solution NMR spectroscopy

Fibril formation of hPTH(1–84) was achieved by dissolving recombinant purified hPTH(1–84) in 50 mM borate buffer, pH 9.0 at a concentration 10 mg/ml and followed by incubation at 65 °C. ThT kinetics was performed in borate buffer and individual aliquots were prepared and incubated. The reaction was stopped by dilution with addition of pre-cooled buffer at 4 °C. The kinetics was followed at 0–3 h of incubation time. These samples were finally diluted to 35 μM of hPTH(1–84) and ThT was added in equal ratio. ThT fluorescence was monitored by excitation wavelength of 450 nm. For the EM analysis the hPTH(1–84) samples were incubated at least for 1 h.

The inhibition of hPTH(1–84) fibril growth was monitored by electron microscopy using 3000–30,000 magnification. For the EM analysis, 10 mg/ml of hPTH(1–84) was dissolved with and without EGCG in 50 mM borated buffer to the molar ratio of 1:2.4 and 1:10 (protein: EGCG). Inhibition was examined by microscopy at different time points from 1 to 48 h. Fluorescence titration was carried in the same buffer, while intrinsic fluorescence of Trp23 was utilized for binding analysis. The sample was excited at 280 nm with an addition of 40 μM of EGCG at each step. For the solution NMR analysis, 50 μM of 15N-hPTH(1–84) was titrated with EGCG to the molar ratio of 1:1 and 1:5, respectively, and monitored by 2D 1H-15N HSQC experiments. The NMR titration was carried out in 10 mM BisTris, 300 mM Na2SO4, 0.02% NaN3, pH 5.3 and at 25 °C.

2.3. Aβ fibril formation Aβ fibrils were grown in 50 mM HEPES buffer at pH 7 by dissolving 50 μM of the peptide followed by incubation at 37 °C [26]. 2.4. Attenuated total reflection–Fourier-transform infrared (ATR-FTIR) spectrometry ATR-FTIR spectra for hPTH(1–84), hPTH(1–84) fibrils and Aβ(1–40) fibrils were recorded with a Tensor 27 FTIR spectrometer (Bruker, Germany) equipped with a BIOATR II cell and a MCT detector that was cooled with liquid nitrogen. 15 μl of the samples (protein concentration: 5 mg/ml) was placed onto the crystal of the ATR cell and measured at room temperature. Collected spectra represent averages of 64 scans at 4 cm−1 resolution. 2.5. Monomer release assay Amyloid fibrils were grown as mentioned above at 918 μM of hPTH(1–84) monomers, pH 9.0. Amyloid fibrils were separated by centrifugation at 30,000 rpm, 4 °C for 30 min. These fibrils were mixed with

2.9. Solid-state NMR spectroscopy hPTH(1–84) fibril solutions were ultracentrifuged at 86,000 ×g for 2 h at 4 °C. The pellets were lyophilized, rehydrated to 50 wt.% H2O, homogenized by freezing the sample in liquid nitrogen and thawing it at 37 °C and centrifuged into 4 mm MAS rotors. MAS NMR spectra were acquired on a Bruker Avance III 600 MHz NMR spectrometer (BrukerBioSpin GmbH, Rheinstetten, Germany) at a resonance frequency of 600.1 MHz for 1H, 150.9 MHz for 13C and 60.8 MHz for 15N. A double channel 4 mm MAS probe was used in all NMR experiments and the temperature was set to 30 °C. NMR spectra were acquired using direct polarization, INEPT or CP excitation schemes. The typical 1H and 13C 90° pulse length were 4 μs. For 13C CP MAS experiments the contact times were 20 μs, 100 μs, 400 μs, 700 μs and 1500 μs and for 15N CP MAS experiments contact times were 100 μs, 500 μs, 1000 μs and 2000 μs with a relaxation delay of 2.5 s and a MAS frequency of 7 kHz. 1 H dipolar decoupling during acquisition with an rf amplitude of 65 Hz was applied using Spinal64. Two dimensional 13C-13C DARR

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

[28] spectra were acquired with mixing times of 20 ms, 100 ms, or 500 ms at a MAS frequency of 10 kHz to avoid folding of sidebands into the region of interest. Typically, 100–160 data points were acquired in the indirect dimension. The 1H-13C dipolar couplings were measured in constant time DIPSHIFT experiments [29] using direct polarization or CP with contact times of 20 μs, 100 μs, 400 μs and 700 μs [30]. The amplitudes of the dipolar dephasing over one rotor period provide a measure for the strength of the dipolar coupling. Homonuclear decoupling using the frequency switched Lee–Goldberg (FSLG) with an effective rf field of 80 kHz was applied during dipolar evolution. The MAS frequency for 13C DIPSHIFT experiments was 5 kHz. The strength of the dipolar coupling was determined from numerical simulations using an angle increment of 1° for powder averaging. Order parameters were calculated as the ratio of the motionally averaged dipolar coupling and the rigid limit value determined for crystalline amino acids [31,32]. 3. Results 3.1. hPTH(1–84) fibril formation and characterization The ability of hPTH(1–84) to form fibrils was examined under six different conditions, varying the pH value, temperature, and buffer type (Table S1). Formation of fibrils could be observed when hPTH(1– 84) was incubated at pH 9.0 and 65 °C within 1 h (Fig. 1). The time dependent aggregation of hPTH(1–84) was monitored by the fluorescent dye thioflavin T (ThT). Binding of ThT to aggregated amyloid species leads to a several fold increase in its fluorescence intensity at 480 nm [33,34]. A standard ThT assay could not be performed due to the instability of ThT at 65 °C. Therefore, 16 individual samples of the same concentrations were incubated in parallel one by one taken out between

0 and 20 min. ThT shows only a small fluorescence signal for hPTH(1– 84) monomers at time point 0. Upon longer incubation, the fluorescence intensity increased significantly up to 12 fold after a lag phase of about 1 min (Fig. 1A). The increased fluorescence indicates that the peptide hormone forms higher aggregated/fibrils species with a time course typical for amyloid fibril formation of other archetypical peptides. Surprisingly, the ThT kinetics was found to be very fast and completed within 5 min of incubation under the mentioned conditions. In order to investigate the stability of the fibrils, a monomer release assay was performed. The fibrils separated by centrifugation were exposed to physiological pH 7. Immediate mixing of buffer did not result in significant amounts of monomers in the supernatant. However, 24 h after mixing 10.1 ± 2.5% of monomer was released (Fig. S1). The N-terminal part of hPTH(1–84) contains ordered secondary structure elements. Both the free peptide hormone and the peptide bound to the extracellular domain of the PTH receptor 1 show αhelical structures for residues 15 to 34 (PDB code 1ZWB [35] and 3C4M [36]. According to the low dispersion of the backbone resonances in the 1H-15N HSQC spectrum of hPTH(1–84), the C-terminal part of the hormone is largely unstructured (Fig. S2). The CD spectrum shows the same signature by the negative ellipticity around 225 nm corresponding to the α-helical contribution and a large negative ellipticity around 200 nm indicative of an intrinsically disordered section (black in Fig. 1B). Fibril formation is accompanied by the formation of a βsheet-rich secondary structure regardless of the structure of the native protein [37]. The CD spectrum of hPTH(1–84) fibrils show a loss in negative ellipticity at 225 nm (red in Fig. 1B) corresponding to the β-sheet formation. The comparison with the CD spectrum of fibrils of the Alzheimer's Aβ(1–40) peptide (blue in Fig. 1B) shows that hPTH(1– 84) still contains significant contributions from random coil structures

B Elipticity (mdeg cm/decimol)

A ThT fluorescence (au)

60 50 40 30 20 10 0 0

251

1

2

3

4

5

10 12 15 20

Time (min)

0 -5 -10 -15 -20 -25 190

200

210

220

230

240

250

Wavelength (nm)

D

E

Absorbance (au)

C

5

1600

1625

1650

Wavenumber

1675

1700

(cm-1)

Fig. 1. Kinetic analysis and characterization of hPTH(1–84) fibrils. (A) Thioflavin T binding kinetics. (B) Secondary structural analysis upon fibril formation monitored by far-UV CD spectroscopy. (C) ATR-FTIR spectroscopy analysis of fibril formation. Only the amide I spectral regions are plotted. (D) Transmission electron micrograph of hPTH(1–84) fibrils. Scale bar: 200 nm. (E) X-ray diffraction pattern of the PTH fibrils. Colors in (B) and (C): black—hPTH(1–84) monomers, red—hPTH(1–84) fibrils, and blue—Aβ(1–40) amyloid fibrils as a comparison.

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

(ellipticity around 200 nm) indicating that only parts of the 84 residues end up in the cross-β structure. Changes in the secondary structures upon fibril formation were also characterized using ATR-FTIR spectroscopy. The amide I spectral region (1600–1700 cm−1) is informative about the stretching mode vibrations of the backbone carbonyl groups and is affected by the secondary structure of the polypeptide chain [38,39]. In the spectrum of hPTH(1–84), amide I vibrations evolved at 1650 cm−1 (black in Fig. 1C) which suggest the presence of random coil-like structure [38,39]. In addition, hPTH(1–84) fibrils showed an extra shoulder at 1615 cm−1 (red in Fig. 1C) that corresponds to the amyloid-like elements of β-sheet structure [40]. Yet, this band was far less pronounced than the large peak seen at this position in the spectrum of Alzheimer's Aβ(1–40) peptide fibrils (blue in Fig. 1C). This observation is consistent with the in silico predictions (see below), which suggested that only a small part of the hPTH(1–84) is involved in the formation of the fibril β-sheet structure. We carried out a transmission electron microscopy (TEM) analysis of hPTH(1–84) to see if fibrils were formed under six different experimental conditions, including varying the pH value, temperature, and the buffer type (Table S1). The screening of various conditions showed that filamentous morphologies of hPTH(1–84) were found within 1 h when the peptide was incubated at pH 9.0 and 65 °C (Fig. 1D). These fibrils had a length of up to several micrometers and featured an unbranched curvilinear morphology. Fibril diameters ranged between 10 and 20 nm. In order to analyze the fibril architecture we recorded X-ray diffraction patterns of the hPTH(1–84) fibrils. The X-ray diffractogram is shown in Fig. 1E. A characteristic fibril diffraction pattern was observed highly indicative of the characteristic cross-β motif. The obtained pattern is defined by meridional reflections at 4.7 Å which correspond to inter β-strand spacing and an equatorial reflection at 10 Å corresponding to the distance between stacked β sheets. The obtained data is in good agreement with the cross-β structure of other amyloid fibrils which show meridional and equatorial reflections at ~ 4.7 Å and ~ 6– 11 Å, respectively [41–45]. From these results one can conclude that the repeating substructure consists of β-strands that run perpendicular to the fibril axis, forming a typical cross-β sheet. Thus, hPTH(1–84) forms well-defined amyloid fibrils containing cross-β structures. 3.2. Identification of the core of the amyloid fibrils In silico predictions using the computer programs PASTA [46] and Waltz [47] indicate that to some extent in the region of residues S1– M10 and predominantly D30–G40 of hPTH(1–84) are prone to aggregation (Fig. 2). Resistance to proteolytic cleavage is another defining characteristic for the presence of amyloid fibrils [48,49]. It is expected that upon fibril formation certain residues are well protected against protease cleavage; hence this method is useful to identify the fibril core. Accordingly, hPTH(1–84) monomers and fibrils were subjected to limited proteolysis by chymotrypsin [27]. The enzyme cleaves predominantly at aromatic residues; however it has a reduced propensity to cleave at leucine and methionine [50]. Samples were collected at different time intervals from 0 to 22 h after incubation and analyzed by MALDI-TOF mass spectrometry. The spectrum indicated a dominant peak at m/z = 1565.52 Da, and this peak consistently appeared upon longer incubation times of 22 h (Figs. 2A and S3). The computer based online program (ProteinProspector) suggested that this m/z value corresponds to the 25R–37L (theoretical wt: 1567.8 Da). On the other hand, monomers of hPTH(1–84) did not result in any protected specific cleavage patterns and the corresponding peaks were absent in the spectrum. The most abundant mono MH+ peak which appeared (m/z = 1565.52 Da) in the MALDI-TOF analysis was further subjected to fragment base ESI-MS-MS for further analysis, which subsequently identified it as 25R–37L. ESI-MS-MS analysis is further confirming the fibril core formation in this region. Although the identified sequence consists of two cleavage sites for chymotrypsin, L28 and F23 (Fig. 2B),

1565.52

A

8

Intensity (au)

252

6 4 2

0

750

1000

1250

1500

1750

2000

m/z

B

Fig. 2. Identification of the fibril core of hPTH(1–84) by limited proteolysis. (A) MALDI-TOF analyses of hPTH(1–84) fibrils digested with chymotrypsin after 2 h of incubation. (B) Amyloidal regions are highlighted on the primary sequence: grey indicates the in silico predicted regions by Waltz and PASTA, cyan indicates the fibril core identified by MALDITOF and ESI-MS-MS and red indicates the common region identified by mass spectrometry and in silico program. Underlined are the possible cleavage sites for chymotrypsin.

the corresponding cleaved product could not be detected by MALDITOF or ESI-MS-MS, thus, confirming the protection against chymotrypsin. These data correlated well with the in silico prediction, where the involvement of the same residues ~ 30–40 was predicted to form the fibril core. However, as revealed by the MS data, the first 10 residues of hPTH(1–84) were not protected against digestion. 3.3. Inhibition of fibril formation The polyphenol (−)-epi-gallocatechine gallate (EGCG) is a very effective amyloid inhibitor suppressing or modulating amyloid formation of different polypeptides including α-synuclein, the Alzheimer's peptide Aβ, and other amyloidogenic peptides by redirecting protein aggregation to nontoxic species [51–53,62–64]. In the present study, we used EGCG to investigate the inhibition of hPTH(1–84) fibril growth. hPTH(1–84) and EGCG were incubated at a molar ratio of 1:2.4 or 1:10 for a period of 48 h at 65 °C. Aliquots were taken after various times for both with and without EGCG and examined by EM. hPTH(1–84) incubated with EGCG at 1:2.4 or 1:10 molar ratio completely inhibited the fibril growth and led to formation of amorphous protein aggregates at all analyzed times (Fig. 3). Small spherical particles up to a size of about 30 nm were observed at a mixing ratio of 1:10 after 1 h of coincubation (Fig. 3 C). The formed species at both ratios are distinctly different from mature fibrils, which were only detected in the samples without EGCG (Fig. 3A and D). Similar effects were reported for αsynuclein or Aβ when treated with EGCG [54]. Next, binding of EGCG to hPTH(1–84) monomers was studied by fluorescence spectroscopy. hPTH(1–84) features an internal tryptophan (Trp23), which could be used as a fluorescent probe to study the binding of EGCG. Upon titration, fluorescent intensity of hPTH(1–84) decreased as a function of EGCG concentration indicative of the interaction between the two molecules (Fig. S4). Data analysis resulted in a KD of about 49.3 μM at a stoichiometry of 1:1.6. To reveal residue resolution of this inhibition, NMR spectroscopy was employed using the backbone N-H resonance assignments of hPTH(1–84) from previous studies [24]. 15 N-hPTH(1–84) was titrated with EGCG and analyzed by 2D 1H-15N HSQC spectra (Fig. 4). Increasing concentrations of EGCG lead to intensity and chemical shift changes of resonances of those residues involved in the binding of EGCG to the hPTH(1–84). In the 2D spectra, only marginal chemical shift changes were observed at a 1:1 (protein: EGCG) ratio. When the molar ratio was increased above 1:3, significant

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

253

Fig. 3. Electron micrographs of EGCG-treated hPTH(1–84) fibrils. EGCG and hPTH(1–84) was co-incubated at the indicated hPTH(1–84):EGCG ratio and EM micrographs were taken after 1 h (upper panel) and 48 h (lower panel). The scale bars represent 200 nm.

A

B

changes were observed. The majority of residues which are located in the fibril core as predicted by the in silico tools and mass spectrometry data also show significant chemical shift changes in the NMR titration experiment (Fig. 4C) and thus binding of EGCG to these residues inhibits fibril growth. 3.4. Solid-state NMR investigations of hPTH(1–84) fibrils

C

Fig. 4. NMR analysis of hPTH(1–84)-EGCG interaction. Overlay of the 1H-15N HSQC spectrum of free 15N-hPTH(1–84) (black) and in complex with EGCG (red) at a 1:1 ratio (A) and a 1:5 ratio (B). (C) Residues are marked on the hPTH(1–84) primary sequence in grey with moderate chemical shift changes and in red with significant changes upon EGCG interactions.

Amyloid fibrils grown from hPTH(1–84) were investigated by solidstate NMR spectroscopy. The 13C magic angle spinning (MAS) NMR spectra are shown in Fig. 5. Three excitation schemes for the 13C NMR coherences were employed, i.e. cross polarization (CP), direct excitation of the 13C nuclei, and insensitive nuclei enhanced by polarization transfer (INEPT) excitation. The different molecular mobility of certain segments of hPTH(1–84) can be used to selectively detect rigid and mobile parts of the molecule just by applying the appropriate excitation scheme [55–57]. The 13C CP MAS NMR spectrum shown in Fig. 5A uses dipolar couplings for the polarization transfer from 1H to 13C. As these couplings are strongest in rigid solids, the resonances shown should belong to the rigid part of hPTH(1–84) fibrils and could be correlated with the fibrillar core [58]. In contrast, the directly polarized 13C MAS NMR spectrum shown in Fig. 5B excites rigid and mobile residues approximately equally, but as mobile sites show narrower NMR signals, these lines dominate the NMR spectrum. Clearly, the spectral line shapes are different and the directly excited 13C MAS NMR spectrum is dominated by narrower lines. In particular in the CO region, some prominent narrow lines are detected in spectrum B. The Arg Cζ side chain resonances at 157.8 ppm can directly be assigned to these residues. According to the proteolytic cleavage data, only Arg25 is close to the fibril core, which should refer to the signal shown in the CP spectrum. In the directly excited 13C NMR spectrum, the Arg Cζ provides a more prominent signal, which now contains the intensity from all 5 Arg residues of hPTH(1–84). The 13C INEPT spectrum shown in Fig. 5C uses scalar couplings for the polarization transfer from 1H to 13C. This polarization transfer

254

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

Fig. 5. 13C MAS NMR spectra of hPTH(1–84) measured at 30 °C using a MAS frequency of 7 kHz. NMR spectra were acquired using (A) CP with a contact time of 700 μs, (B) direct polarization, and (C) the INEPT sequence.

scheme only works efficiently for mobile sites, which are now displayed as narrow resonances. As expected, the general intensity distribution and line widths of the directly excited and the INEPT spectra is relatively similar indicating that both methods predominantly detect mobile sites. In contrast, the CP MAS NMR spectrum in Fig. 5A shows a much different intensity distribution especially in the Cα and Cβ region. In particular, the intensity of the Cα signals is shifted towards higher field, which in general indicates β-sheet structure [59]. This further supports the assumption that the CP MAS NMR spectrum predominantly reports the rigid fibrillar part of the hPTH(1–84) fibrils. 15 N CP MAS NMR spectra (Fig. S5) also confirm these conclusions. At a short CP contact time of 100 μs, the spectral signals of Lys and Arg side chains are detected in addition to the prominent peak form the backbone NH sites. These can be assigned to Arg25 as well as Lys26 and 27 and, which are flanking the fibril core of hPTH(1–84). The Lys and Arg side chains receive more intensity as the CP contact time is increased, as also these more mobile residues are excited. Some further assignments of the NMR signals of the fibrillar part of the hPTH(1–84) sample can be provided by two dimensional 13C-13C correlation spectroscopy. The contour plot of a 13C–13C dipolar assisted rotational resonance (DARR) NMR spectrum [28] using the CP excitation scheme and a mixing time of 20 ms is shown in Fig. 6. At this short mixing time, only correlations within the spin system of the respective amino acid are shown. According to the cross peak pattern, we can clearly identify several residues; most of them should be part of the fibrillar core. There are 7 Ala residues in the hPTH(1–84) sequence and we find a total of 5 cross peaks, three of which with βsheet like chemical shift, which we tentatively assign to Ala36, 39, and 42 (Table S2). However, the two other cross peaks are characteristic of α-helical/random coil chemical shifts and indicate that also other Ala residues are rigid enough to provide sufficient dipolar coupling strengths for efficient polarization transfer. The chemical shifts of the amino acid types that could be unambiguously assigned on the basis of the 2D 13C–13C correlation experiments are given in Table S2. In particular, we find two sets of cross peaks for the 8 Val residues in the hPTH(1–84) sequence, 2 of which should be part of the fibril (Val31 and 35) and the respective signals show the corresponding β-sheet conformation. The same result is found for the 10 Leu residues. Three out of 10 Leu should be part of the fibril and we find one population of Leu

Fig. 6. 13C–13C DARR NMR spectrum of hPTH(1–84) measured at 30 °C using a mixing time of 20 ms and a MAS frequency of 10 kHz. Amino acid type assignments are given for some clearly identifiable correlations. Spectral correlations that are indicative of β-sheet structures are indicated by the dashed lines.

with clearly β-sheet like chemical shifts (tentatively assigned to Leu28, 37, and possibly 41). Two very clear correlations in the DARR spectrum further confirm that also residues, which are not part of the fibril core, are sufficiently rigid to contribute intensity in the DARR spectrum. Clear cross peaks are found for the Ser residues (there is a total of 7 in the hPTH(1–84) sequence), which both show β-sheet like chemical shifts. Also, a characteristic cross peak for the Ile side chain (Cδ/Cγ) indicates that the only Ile 5 in hPTH(1–84) resides in a somewhat rigid conformation. These rigid sites can be attributed to the N-terminus of PTH, which was predicted to have a fibril forming tendency, which was, however, not found in the proteolysis data. Although there are much more cross peaks displayed in the DARR spectrum, given the line width of the signals, no further assignments of the uniformly 13C labeled hPTH(1–84) are possible with full confidence. To this end, selectively labeled hPTH(1–84) fibrils will need to be studied [60], which will be subject to further work. Here, we turn our attention to the dynamic aspects of the hPTH(1–84) fibrils and study the segmental order parameters of the preparations. As implied from Fig. 5, the molecular dynamics in the hPTH(1–84) sample is broadly distributed varying from very rigid sites to highly mobile segments that provide well resolved INEPT spectra. In order to get some quantitative information on this interesting mobility, we carried out 1H–13C dipolar coupling measurements using the dipolar chemical shift correlation (DIPSHIFT) pulse sequence. As an excitation scheme we used either CP with varying contact times or direct excitation [30,58]. Dipolar coupling values were converted into a segmental order parameter that is 1 for completely rigid and 0 for isotropically mobile sites. Order parameters for the hPTH(1–84) fibrils are shown in Fig. 7. At a short CP contact time, only the rigid (i.e. fibrillar) segments of the hPTH(1–84) preparation are excited and consequently very high order parameters between 0.7 for the side chains and 0.9 for the backbone are detected. Increasing the CP contact time to 0.7 ms gradually excites also more mobile segments of the hPTH(1– 84) fibrils and significantly lower order parameters of ~0.7 for the backbone and ~ 0.2–0.45 for the side chains are detected. However, if the DIPSHIFT experiment is carried out using direct excitation, which

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

Fig. 7. 1H–13C order parameters of the hPTH(1–84) peptide determined from DIPSHIFT experiment acquired using CP with varying contact times or direct excitation of the 13C nuclei.

emphasizes the highly mobile groups, the detected order parameters drop further to very low values of 0.48 for the backbone and 0.18– 0.33 for the side chains illustrating the tremendous difference in the dynamics of the fibrillar and non-fibrillar segments of the hPTH(1–84) preparation.

4. Discussion Formation of amyloid fibrils is quite common in neurodegenerative diseases [4]. Such amyloid deposits are the consequences of the failure of proteins to fold properly in a non-aggregation prone, native conformation [1]. On the other hand, proteins/peptides of secretory pathways can be deposited in densely packed membrane coated granules referred to as functional amyloids [9]. Although functional amyloids are distinctly different from the disease-causing amyloids, they exhibited the same basic architecture of cross-β-sheet-rich conformations [9]. In early 1976, formation of fibrils from hPTH(1–84) was reported, however, without a detailed molecular characterization [9,18]. In a systematic screening of fibril forming in vitro conditions, we identified high pH and temperature as suitable condition for amyloid formation. The fibril growth experiments showed a quite fast kinetics and fibril formation accomplished within the first 5 min after a short lag phase (Fig. 1A). Similarly to other amyloid forming peptides [3], we expect a nucleation process of higher oligomers during this lag phase before ThT binding associates can form. Addition of EGCG disturbs this process (Fig. 3) by binding already to the monomers as monitored by NMR (Fig. 4). Mature fibrils were found in unbranched, curvilinear morphology of several micro-meters length and 10–20 nm of diameters (Fig. 1D). Such a morphological shape has been observed for several other amyloidogenic peptides including Aβ and α-synuclein [3] and other peptide hormones [9]. The characteristic morphology of cross-β formation was confirmed by the meridional and equatorial reflections in the X-ray diffraction data (Fig. 1E). Often, fibrils show β-sheet rich structures regardless of the secondary structure composition of the native peptide/protein structure [37]. hPTH(1–84) contains an α-helical secondary structure at the N-terminus [35,36], while fibril formation showed the characteristics of a β-sheet conformation reflected in the CD spectrum (Fig. 1B). Aggregated or fibrilar hPTH(1–84) might correspond to a storage form of the hormone, which implies that monomers can dissociate from the associated forms. For somatostatin-14 and corticotropin-releasing factor for example dissociation from the fibrilar storage form in secretory granules into monomers has been reported [9,61]. And indeed, dissolving hPTH(1–84) fibrils in pure buffer resulted in the release of about 10% of monomers after 24 h (Fig. S1). Together with the fast fibrillation

255

kinetics, at least the thermodynamic properties of hPTH(1–84) to form functional amyloids are given. The programs PASTA and Waltz provided a commonly predicted region of residues 30–40 prone for amyloid formation and some tendency for residues 1–10 (Fig. 2B). Our mass spectrometry analysis identified that 13 residues (25R–37L) are involved in the fibril core formation, because this sequence contains two chymotrypsin cleavage sites at L28 and F34 which are well protected against proteolysis of the mature fibrils (Fig. 2A). That only a part of the PTH(1–84) residues forms βsheet structure was confirmed by CD or FT-IR spectroscopy and a comparison with the spectra of Aβ(1–40) fibrils helped to identify the corresponding spectral features (Figs. 1B and 2A). Solid-state NMR further confirmed these findings by observing both highly flexible and more rigid regions in hPTH(1–84) fibrils. The backbone chemical shifts of the rigid parts are biasing towards higher field, a general indication of β-sheet structure [59]. Few tentative assignments to the types of amino acid types in this region match with the immobile core region and directly flanking sections. The DIPSHIFT experiment highlights again the huge dynamic range of the fibrils and thus the existence of both mobile and rigid sections (Fig. 7). EGCG is a well known inhibitor for amyloid fibril formation [51–53] and indeed the growth of hPTH(1–84) fibrils was inhibited by this small chemical compound. In an analysis by solution NMR of monomers at residue resolution the majority of changes in chemical shifts were observed particularly for those residues involved in the core of mature fibrils (Fig. 4). This indicates that EGCG masks the aggregation prone sections of hPTH(1–84) and thus retards fibril formation. Similar results have been previously reported for human calcitonin where EGCG prevented association of the peptide hormone before fibril formation [62]. The NMR analysis revealed that the aromatic rings of EGCG and the side chain of aromatic and hydrophobic residues (Tyr12, Phe16, Phe22 and Ala31) of the hormone may play an important role in inhibiting fibril formation. Another study showed that the prostatic acid phosphatase peptide (PAP 248–286) forms so called SEVI amyloid fibrils, which enhance the HIV infectivity [63]. EGCG inhibited the SEVI formation by interacting at two regions (Lys251–Arg257 and Gln269– Ile277) of primarily charged residues, particularly lysine. Another example is diabetes-related islet amyloid polypeptide (IAPP), which fibrillizes without an appreciable buildup of nonfibrillar intermediates [64]. However, its growth could be retarded by EGCG that stabilizes nonfibrillar large aggregates during fibrillogenesis. For hPTH(1–84) we found that EGCG primarily interacts with aromatic residues (His14, His32, Phe34) and hydrophobic sections (Phe34, Val35, Ala36, Leu37, Leu42, Ala43) flanked by positively charged Lys13 and Arg44.

5. Conclusion Compared to other amyloid fibril forming peptide/proteins, hPTH(1–84) forms very fast amyloid fibrils within minutes under our in vitro conditions. Their thermodynamic stability is low enough to dissociate again into monomers after dilution. This low stability might result from the finding that only about 20% of the residues of hPTH(1– 84) form the core cross-β structure around the aggregation prone peptide section comprising residues between 25 and 40. Therefore we speculate, that these fibrils are the storage conformation of hPTH(1–84) as ‘functional amyloids’ in secretory granules. The Nand C-terminal stretches are highly flexible and in a more random coil conformation. A confirmation of these suggestions requires a respective biophysical analysis of deposits from natural sources, which is future work.

Conflict of interest The authors declare no conflict of interest.

256

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257

Acknowledgments This work has been supported by grants from the DFG (GRK 1026, SFB TRR102), the BMBF (ProNet-T3), the state Sachsen-Anhalt (Exzellenznetzwerk Biowissenschaften), and ERDF by the EU. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbapap.2014.12.020. References [1] M. Fändrich, M. Schmidt, N. Grigorieff, Recent progress in understanding Alzheimer's beta-amyloid structures, Trends Biochem. Sci. 36 (2011) 338–345. [2] T. Eichner, S.E. Radford, A diversity of assembly mechanisms of a generic amyloid fold, Mol. Cell 43 (2011) 8–18. [3] I.W. Hamley, The amyloid beta peptide: a chemist's perspective. Role in Alzheimer's and fibrillization, Chem. Rev. 112 (2012) 5147–5192. [4] M. Fändrich, Oligomeric intermediates in amyloid formation: structure determination and mechanisms of toxicity, J. Mol. Biol. 421 (2012) 427–440. [5] M.M. Barnhart, M.R. Chapman, Curli biogenesis and function, Annu. Rev. Microbiol. 60 (2006) 131–147. [6] V.A. Iconomidou, G. Vriend, S.J. Hamodrakas, Amyloids protect the silkmoth oocyte and embryo, FEBS Lett. 479 (2000) 141–145. [7] M.L. Maddelein, S. Dos Reis, S. Duvezin-Caubet, B. Coulary-Salin, S.J. Saupe, Amyloid aggregates of the HET-s prion protein are infectious, Proc. Natl. Acad. Sci. U. S. A. 99 (2002) 7402–7407. [8] D.M. Fowler, A.V. Koulov, C. Alory-Jost, M.S. Marks, W.E. Balch, J.W. Kelly, Functional amyloid formation within mammalian tissue, PLoS Biol. 4 (2006) e6. [9] S.K. Maji, M.H. Perrin, M.R. Sawaya, S. Jessberger, K. Vadodaria, R.A. Rissman, P.S. Singru, K.P. Nilsson, R. Simon, D. Schubert, D. Eisenberg, J. Rivier, P. Sawchenko, W. Vale, R. Riek, Functional amyloids as natural storage of peptide hormones in pituitary secretory granules, Science 325 (2009) 328–332. [10] P. Westermark, On the nature of the amyloid in human islets of langerhans, Histochemistry 38 (1974) 27–33. [11] M. Fandrich, M. Schmidt, N. Grigorieff, Recent progress in understanding Alzheimer's beta-amyloid structures, Trends Biochem. Sci. 36 (2011) 338–345. [12] T.B. Usdin, The PTH2 receptor and TIP39: a new peptide-receptor system, Trends Pharmacol. Sci. 21 (2000) 128–130. [13] M.W. Freeman, K.M. Wiren, A. Rapoport, M. Lazar, J.T. Potts Jr., H.M. Kronenberg, Consequences of amino-terminal deletions of preproparathyroid hormone signal sequence, Mol. Endocrinol. 1 (1987) 628–638. [14] A.C. Karaplis, S.K. Lim, H. Baba, A. Arnold, H.M. Kronenberg, Inefficient membrane targeting, translocation, and proteolytic processing by signal peptidase of a mutant preproparathyroid hormone protein, J. Biol. Chem. 270 (1995) 1629–1635. [15] J.F. Habener, M. Rosenblatt, P.C. Dee, J.T. Potts Jr., Cellular processing of preproparathyroid hormone involves rapid hydrolysis of the leader sequence, J. Biol. Chem. 254 (1979) 10596–10599. [16] D.V. Cohn, J. Elting, Biosynthesis, processing, and secretion of parathormone and secretory protein-I, Recent Prog. Horm. Res. 39 (1983) 181–209. [17] K.M. Wiren, J.T. Potts Jr., H.M. Kronenberg, Importance of the propeptide sequence of human preproparathyroid hormone for signal sequence function, J. Biol. Chem. 263 (1988) 19771–19777. [18] I. Kedar, M. Ravid, E. Sohar, In vitro synthesis of “amyloid” fibrils from insulin, calcitonin and parathormone, Isr. J. Med. Sci. 12 (1976) 1137–1140. [19] M.P. Mattson, D.S. Gary, S.L. Chan, W. Duan, Perturbed endoplasmic reticulum function, synaptic apoptosis and the pathogenesis of Alzheimer's disease, Biochem. Soc. Symp. (2001) 151–162. [20] H. Nishitoh, A. Matsuzawa, K. Tobiume, K. Saegusa, K. Takeda, K. Inoue, S. Hori, A. Kakizuka, H. Ichijo, ASK1 is essential for endoplasmic reticulum stress-induced neuronal cell death triggered by expanded polyglutamine repeats, Genes Dev. 16 (2002) 1345–1355. [21] S. Oyadomari, E. Araki, M. Mori, Endoplasmic reticulum stress-mediated apoptosis in pancreatic beta-cells, Apoptosis 7 (2002) 335–345. [22] T.G. Ahn, S.E. Antonarakis, H.M. Kronenberg, T. Igarashi, M.A. Levine, Familial isolated hypoparathyroidism: a molecular genetic analysis of 8 families with 23 affected persons, Medicine (Baltimore) 65 (1986) 73–81. [23] E. Bosse-Doenecke, U. Weininger, M. Gopalswamy, J. Balbach, S.M. Knudsen, R. Rudolph, High yield production of recombinant native and modified peptides exemplified by ligands for G-protein coupled receptors, Protein Expr. Purif. 58 (2008) 114–121. [24] A. Kumar, M. Gopalswamy, C. Wishart, M. Henze, L. Eschen-Lippold, D. Donnelly, J. Balbach, N-terminal phosphorylation of parathyroid hormone (PTH) abolishes its receptor activity, ACS Chem. Biol. (2014), http://dx.doi.org/10.1021/cb5004515. [25] T. Christopeit, P. Hortschansky, V. Schroeckh, K. Guhrs, G. Zandomeneghi, M. Fändrich, Mutagenic analysis of the nucleation propensity of oxidized Alzheimer's beta-amyloid peptide, Protein Sci. 14 (2005) 2125–2131. [26] C. Haupt, I. Morgado, S.T. Kumar, C. Parthier, M. Bereza, P. Hortschansky, M.T. Stubbs, U. Horn, M. Fandrich, Amyloid fibril recognition with the conformational B10 antibody fragment depends on electrostatic interactions, J. Mol. Biol. 405 (2011) 341–348.

[27] C.J. Sarell, L.A. Woods, Y. Su, G.T. Debelouchina, A.E. Ashcroft, R.G. Griffin, P.G. Stockley, S.E. Radford, Expanding the repertoire of amyloid polymorphs by copolymerization of related protein precursors, J. Biol. Chem. 288 (2013) 7327–7337. [28] K. Takegoshi, S. Nakamura, T. Terao, C-13-H-1 dipolar-assisted rotational resonance in magic-angle spinning NMR, Chem. Phys. Lett. 344 (2001) 631–637. [29] M.G. Munowitz, R.G. Griffin, G. Bodenhausen, T.H. Huang, Two-dimensional rotational spin-echo nuclear magnetic resonance in solids: correlation of chemical shift and dipolar interactions, J. Am. Chem. Soc. 103 (1981) 2529–2533. [30] P. Schmidt, L. Thomas, P. Muller, H.A. Scheidt, D. Huster, The G-protein-coupled neuropeptide Y receptor type 2 is highly dynamic in lipid membranes as revealed by solid-state NMR spectroscopy, Chemistry 20 (2014) 4986–4992. [31] P. Barre, O. Zschornig, K. Arnold, D. Huster, Structural and dynamical changes of the bindin B18 peptide upon binding to lipid membranes. A solid-state NMR study, Biochemistry 42 (2003) 8377–8386. [32] D. Huster, L. Xiao, M. Hong, Solid-state NMR investigation of the dynamics of the soluble and membrane-bound colicin Ia channel-forming domain, Biochemistry 40 (2001) 7662–7674. [33] H. LeVine III, Thioflavine T interaction with synthetic Alzheimer's disease betaamyloid peptides: detection of amyloid aggregation in solution, Protein Sci. 2 (1993) 404–410. [34] T. Eichner, A.P. Kalverda, G.S. Thompson, S.W. Homans, S.E. Radford, Conformational conversion during amyloid formation at atomic resolution, Mol. Cell 41 (2011) 161–172. [35] U.C. Marx, K. Adermann, P. Bayer, M. Meyer, W.G. Forssmann, P. Rosch, Structure– activity relation of NH2-terminal human parathyroid hormone fragments, J. Biol. Chem. 273 (1998) 4308–4316. [36] A.A. Pioszak, H.E. Xu, Molecular recognition of parathyroid hormone by its G protein-coupled receptor, Proc. Natl. Acad. Sci. U. S. A. 105 (2008) 5034–5039. [37] K.E. Marshall, L.C. Serpell, Structural integrity of beta-sheet assembly, Biochem. Soc. Trans. 37 (2009) 671–676. [38] H. Hiramatsu, T. Kitagawa, FT-IR approaches on amyloid fibril structure, Biochim. Biophys. Acta 1753 (2005) 100–107. [39] S. Krimm, J. Bandekar, Vibrational spectroscopy and conformation of peptides, polypeptides, and proteins, Adv. Protein Chem. 38 (1986) 181–364. [40] G. Zandomeneghi, M.R. Krebs, M.G. McCammon, M. Fandrich, FTIR reveals structural differences between native beta-sheet proteins and amyloid fibrils, Protein Sci. 13 (2004) 3314–3321. [41] W.T. Astbury, S. Dickinson, K. Bailey, The X-ray interpretation of denaturation and the structure of the seed globulins, Biochem. J. 29 (1935) 2351–2360. [42] J. Greenwald, R. Riek, Biology of amyloid: structure, function, and regulation, Structure 18 (2010) 1244–1260. [43] M. Sunde, C. Blake, The structure of amyloid fibrils by electron microscopy and X-ray diffraction, Adv. Protein Chem. 50 (1997) 123–159. [44] M. Sunde, L.C. Serpell, M. Bartlam, P.E. Fraser, M.B. Pepys, C.C. Blake, Common core structure of amyloid fibrils by synchrotron X-ray diffraction, J. Mol. Biol. 273 (1997) 729–739. [45] J. Adler, H.A. Scheidt, M. Kruger, L. Thomas, D. Huster, Local interactions influence the fibrillation kinetics, structure and dynamics of Abeta(1–40) but leave the general fibril structure unchanged, Phys. Chem. Chem. Phys. 16 (2014) 7461–7471. [46] A. Trovato, F. Seno, S.C. Tosatto, The PASTA server for protein aggregation prediction, Protein Eng. Des. Sel. 20 (2007) 521–523. [47] J. Reumers, J. Schymkowitz, F. Rousseau, Using structural bioinformatics to investigate the impact of non synonymous SNPs and disease mutations: scope and limitations, BMC Bioinf. 10 (Suppl. 8) (2009) S9. [48] O.V. Bocharova, L. Breydo, A.S. Parfenov, V.V. Salnikov, I.V. Baskakov, In vitro conversion of full-length mammalian prion protein produces amyloid form with physical properties of PrP(Sc), J. Mol. Biol. 346 (2005) 645–659. [49] D.M. Hartley, C. Zhao, A.C. Speier, G.A. Woodard, S. Li, Z. Li, T. Walz, Transglutaminase induces protofibril-like amyloid beta-protein assemblies that are protease-resistant and inhibit long-term potentiation, J. Biol. Chem. 283 (2008) 16790–16800. [50] W. Appel, Chymotrypsin: molecular and catalytic properties, Clin. Biochem. 19 (1986) 317–322. [51] D.E. Ehrnhoefer, J. Bieschke, A. Boeddrich, M. Herbst, L. Masino, R. Lurz, S. Engemann, A. Pastore, E.E. Wanker, EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers, Nat. Struct. Mol. Biol. 15 (2008) 558–566. [52] F. Meng, A. Abedini, A. Plesner, C.B. Verchere, D.P. Raleigh, The flavanol (−)-epigallocatechin 3-gallate inhibits amyloid formation by islet amyloid polypeptide, disaggregates amyloid fibrils, and protects cultured cells against IAPP-induced toxicity, Biochemistry 49 (2010) 8127–8133. [53] L.M. Young, P. Cao, D.P. Raleigh, A.E. Ashcroft, S.E. Radford, Ion mobility spectrometry-mass spectrometry defines the oligomeric intermediates in amylin amyloid formation and the mode of action of inhibitors, J. Am. Chem. Soc. 136 (2014) 660–670. [54] J. Bieschke, J. Russ, R.P. Friedrich, D.E. Ehrnhoefer, H. Wobst, K. Neugebauer, E.E. Wanker, EGCG remodels mature alpha-synuclein and amyloid-beta fibrils and reduces cellular toxicity, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 7710–7715. [55] H. Heise, W. Hoyer, S. Becker, O.C. Andronesi, D. Riedel, M. Baldus, Molecular-level secondary structure, polymorphism, and dynamics of full-length alpha-synuclein fibrils studied by solid-state NMR, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 15871–15876. [56] D. Huster, P.K. Madhu, Simplifying solid-state NMR spectra for biophysical studies on membrane proteins: selective targeting of sites and interactions, Biophys. J. 106 (2014) 2083–2084. [57] D. Huster, J. Schiller, K. Arnold, Comparison of collagen dynamics in articular cartilage and isolated fibrils by solid-state NMR spectroscopy, Magn. Reson. Med. 48 (2002) 624–632.

M. Gopalswamy et al. / Biochimica et Biophysica Acta 1854 (2015) 249–257 [58] M. Sackewitz, H.A. Scheidt, G. Lodderstedt, A. Schierhorn, E. Schwarz, D. Huster, Structural and dynamical characterization of fibrils from a disease-associated alanine expansion domain using proteolysis and solid-state NMR spectroscopy, J. Am. Chem. Soc. 130 (2008) 7172–7173. [59] S.B. Spera, A. Bax, Empirical correlation between protein backbone conformation and Cα and Cβ 13C nuclear magnetic resonance chemical shifts, J. Am. Chem. Soc. 113 (1991) 5490–5492. [60] F. Castellani, B. van Rossum, A. Diehl, M. Schubert, K. Rehbein, H. Oschkinat, Structure of a protein determined by solid-state magic-angle-spinning NMR spectroscopy, Nature 420 (2002) 98–102. [61] A. Anoop, S. Ranganathan, B. Das Dhaked, N.N. Jha, S. Pratihar, S. Ghosh, S. Sahay, S. Kumar, S. Das, M. Kombrabail, K. Agarwal, R.S. Jacob, P. Singru, P. Bhaumik, R. Padinhateeri, A. Kumar, S.K. Maji, Elucidating the role of disulfide bond on amyloid

257

formation and fibril reversibility of somatostatin-14: relevance to its storage and secretion, J. Biol. Chem. 289 (2014) 16884–16903. [62] R. Huang, S. Vivekanandan, J.R. Brender, Y. Abe, A. Naito, A. Ramamoorthy, NMR characterization of monomeric and oligomeric conformations of human calcitonin and its interaction with EGCG, J. Mol. Biol. 416 (2012) 108–120. [63] N. Popovych, J.R. Brender, R. Soong, S. Vivekanandan, K. Hartman, V. Basrur, P.M. Macdonald, A. Ramamoorthy, Site specific interaction of the polyphenol EGCG with the SEVI amyloid precursor peptide PAP(248–286), J. Phys. Chem. B 116 (2012) 3650–3658. [64] Y. Suzuki, J.R. Brender, K. Hartman, A. Ramamoorthy, E.N. Marsh, Alternative pathways of human islet amyloid polypeptide aggregation distinguished by (19)f nuclear magnetic resonance-detected kinetics of monomer consumption, Biochemistry 51 (2012) 8154–8162.