Protein misfolding and aggregation play an important role in many human diseases including Alzheimer's, Parkinson's and type 2 diabetes mellitus (T2DM). The human islet amyloid polypeptide (hIAPP) forms amyloid plaques in the pancreas of T2DM subjects (>95%) that are involved in deteriorating islet function and in mediating β-cell apoptosis. However, the detailed mechanism of action, structure and nature of toxic hIAPP species responsible for this effect remains elusive to date mainly due to the high cost associated with the chemical synthesis of pure peptide required for these studies. In the present work, we attempted to obtain structural and mechanistic insights into the hIAPP aggregation process using recombinant hIAPP (rhIAPP) isolated from Escherichia coli. Results from biophysical and structural studies indicate that the rhIAPP self-assembled into highly pure, β-sheet-rich amyloid fibrils with uniform morphology. rhIAPP-mediated apoptosis in INS-1E cells was associated with increased oxidative stress and changes in mitochondrial membrane potential. The transcript levels of apoptotic genes - Caspase-3 and Bax were found to be up-regulated, while the levels of the anti-apoptotic gene - Bcl2 were down-regulated in rhIAPP-treated cells. Additionally, the expression levels of genes involved in combating oxidative stress namely Catalase, SOD1 and GPx were down-regulated. rhIAPP exposure also affected glucose-stimulated insulin secretion from isolated pancreatic islets. The aggregation of rhIAPP also occurred significantly faster when compared with that of the chemically synthesized peptide. We also show that the rhIAPP fibrils were shorter and more cytotoxic. In summary, our study is one among the few to provide comprehensive evaluation of structural, biophysical and cytotoxic properties of rhIAPP.

Introduction

The incidence of type 2 diabetes mellitus (T2DM) — a serious, chronic metabolic disorder — is at all-time high, with ∼9% of the adult population affected worldwide. According to the estimates of the International Diabetes Federation (IDF), the prevalence of T2DM is expected to double with ∼552 million people to suffer from diabetes by 2030 [1]. Initially considered as an old-age disease, the incidence of pre-diabetes and T2DM is increasing in children, adolescents and young adults and has been linked to changes in lifestyle, over nutrition, stress, obesity, etc [2,3]. T2DM is caused mainly due to dysfunction and/or death of pancreatic β-cells and has been associated with defects in insulin production, secretion and signaling [4]. Among the various factors identified for the cause and progression of T2DM, the accumulation of aggregates of the amyloidogenic peptide — human islet amyloid polypeptide (hIAPP) — has been reported in the pancreatic islets of ∼95% of T2DM patients [5]. hIAPP or amylin is a 37-amino acid long peptide that is co-secreted with insulin by the pancreatic β-cells [6,7]. Under normal conditions, hIAPP plays a key role in regulating glucose homeostasis by slowing gastric emptying [8]. However, during T2DM progression, hIAPP aggregates into a variety of amyloidogenic states, a few of which are highly toxic and have been linked to the death of insulin-producing β-cells [9,10].

Several research groups have tried to elucidate the molecular mechanism and the nucleation pathway involved in the formation of hIAPP aggregates both in vitro [11,12] and in vivo [13]. Results from these studies, mostly performed using peptides prepared by conventional solid-phase chemical synthesis methods, suggest that the hIAPP aggregation follows a sigmoidal nucleation-dependent pathway wherein fibril formation is accompanied by a transition from unstructured conformation into a β-sheet-rich structure [14]. A debate exists about whether the mature fibrils, other ‘on-pathway’ oligomers or small protofilaments are responsible for β-cell apoptosis. The possible mechanisms proposed for hIAPP-mediated cytotoxicity include: physical damage to the cellular membranes induced by the growing aggregates; changes in cellular permeability; generation of endoplasmic reticulum (ER) stress, oxidative stress and increased mitochondrial dysfunction [15]. However, the exact mechanism of hIAPP-induced amyloidogenicity and β-cell apoptosis still remains elusive in vivo. Therefore, to combat T2DM by targeting hIAPP, a detailed characterization of structural and cellular properties of hIAPP amyloid assemblies and the ‘on-pathway’ intermediates is crucial to gain insights into the mechanism of hIAPP aggregation and to comprehend its exact role in T2DM progression.

In the present study, using a novel, efficient and cost-effective method, mature recombinant hIAPP (rhIAPP) peptides/fibrils were prepared from Escherichia coli and morphologically characterized using transmission electron microscopy (TEM) and atomic force microscopy (AFM). The mature hIAPP peptide self-assembled into β-sheet-rich, highly pure amyloid fibrils with uniform morphology. The uniformly labeled 13C–15N hIAPP fibrils were also used to record ssNMR (solid-state NMR)-MAS (magic angle spinning) spectrum. Cellular assays indicate that rhIAPP mediates cytotoxicity in INS-1E cells via induction of oxidative stress that was associated with lipid peroxidation, changes in mitochondrial membrane potential (MMP) and DNA damage. rhIAPP exposure also affected glucose-stimulated insulin secretion (GSIS) from pancreatic islets. Additionally, rhIAPP produced by the current protocol mediated higher cytotoxicity in INS-1E cells when compared with the synthetic hIAPP possibly due to the formation of shorter aggregates. We also observed that the aggregation of rhIAPP was significantly faster when compared with that of the synthetic peptide. The protocol developed and described here for the production of rhIAPP could further be extrapolated for the overexpression of other amyloidogenic peptides or proteins that are difficult to be synthesized chemically.

Materials and methods

Materials

All tissue culture reagents were purchased from Thermo Fisher Scientific, and all fine chemicals were purchased from Himedia. Synthetic hIAPP peptide (98.74% purity) was purchased from Peptide 2.0 (Chantilly, VA). Full-length hIAPP peptide was synthesized using standard FMOC (9-fluornylmethoxycarbonyl) solid-phase peptide synthesis, and the purity was checked with HPLC and MALDI-TOF (matrix-assisted laser desorption-time of flight; data not shown). For experiments, the stock solution of synthetic peptide was maintained as 1 mM in hexafluoroisopropanol (HFIP) (Himedia). Thioflavin T (ThT), Thioflavin S (ThS), crystal violet and other chemicals were obtained from Himedia. JC-1 was procured from Thermo Fisher Scientific. β-Actin, Poly (ADP-ribose) polymerase (PARP), GRP-78, Caspase-3 and p-eIF-2α antibodies were purchased from Abcam (U.S.A.). The mouse insulin ELISA kit was purchased from Mercodia (Sweden). Double-distilled water was filtered and deionized through a Millipore water purification system (Merck Millipore, U.S.A.).

Bacterial expression and purification of rhIAPP

The pGEV1-hIAPP1–37 construct expressing the hexahistidine-tagged hIAPP1–37 fused with GB1 (Protein G, B1 domain) and flanked by methionine at N- and C-termini (Supplementary Material and Figure S1B) was transformed into E. coli BL21 (DE3) cells, overexpressed in 2× YT medium and was purified using the protocol described in Figure 1A. The GB1, 6×-His tag and the extra methionine residue from both the terminal ends of recombinant hIAPP1–37 were cleaved by adding cyanogen bromide (CNBr) to the eluted protein, as described previously [16]. Post-cleavage, the sample was subjected to ultracentrifugation (Beckman Coulter Optima™ MAX-XP, U.S.A.) at 88 300 g for 55 min at 10°C. The resulting pellet was further washed 4–5 times with phosphate buffer and processed further for fibril formation (refer to Supplementary Methods for further details).

Expression and purification of recombinant hIAPP.

Figure 1.
Expression and purification of recombinant hIAPP.

(A) A flowchart depicting the protocol for expression and purification of recombinant hIAPP. (B) SDS–PAGE analysis of protein samples electrophoresed on 15% acrylamide gel after Ni-NTA column chromatography. Lane 1: flow-through of the Gdn–HCl-treated lysate after passing through the Ni-NTA column; lane 3: impurities obtained after denaturation buffer wash; lanes 4 and 5: eluted fractions after wash with native buffer containing 25 and 50 mM imidazole, respectively; lanes 6 and 7: eluted GB1–hIAPP (fusion protein) containing impurities. (C) Gel filtration chromatogram of the fusion protein. (D) SDS–PAGE analysis of purified fusion protein (after gel filtration chromatography) on a 15% acrylamide gel. (E) MALDI-TOF analysis of purified mature hIAPP showed a single peak at 4.037 kDa corresponding to Z = +1 state of hIAPP1–37 with one extra amino acid (methionine) at C-terminus after CNBr cleavage. (F) Tris–Tricine SDS–PAGE (10% C, 3% T) analysis of purified rhIAPP (lane 1 corresponds to supernatant obtained after CNBr digestion, while lane 3 corresponds to purified rhIAPP in 7 M Gdn–HCl).

Figure 1.
Expression and purification of recombinant hIAPP.

(A) A flowchart depicting the protocol for expression and purification of recombinant hIAPP. (B) SDS–PAGE analysis of protein samples electrophoresed on 15% acrylamide gel after Ni-NTA column chromatography. Lane 1: flow-through of the Gdn–HCl-treated lysate after passing through the Ni-NTA column; lane 3: impurities obtained after denaturation buffer wash; lanes 4 and 5: eluted fractions after wash with native buffer containing 25 and 50 mM imidazole, respectively; lanes 6 and 7: eluted GB1–hIAPP (fusion protein) containing impurities. (C) Gel filtration chromatogram of the fusion protein. (D) SDS–PAGE analysis of purified fusion protein (after gel filtration chromatography) on a 15% acrylamide gel. (E) MALDI-TOF analysis of purified mature hIAPP showed a single peak at 4.037 kDa corresponding to Z = +1 state of hIAPP1–37 with one extra amino acid (methionine) at C-terminus after CNBr cleavage. (F) Tris–Tricine SDS–PAGE (10% C, 3% T) analysis of purified rhIAPP (lane 1 corresponds to supernatant obtained after CNBr digestion, while lane 3 corresponds to purified rhIAPP in 7 M Gdn–HCl).

Matrix-assisted laser desorption-time of flight analysis

Aliquots of GB1–hIAPP fusion protein and HPLC-purified hIAPP (obtained after CNBr cleavage) were further characterized by MALDI-TOF mass spectrometry on Autoflex Speed (Bruker, Germany). Briefly, 2 μl of the protein (∼10 pmol) was mixed with 2 μl of matrix solution [sinapinic acid — 50% acetonitrile, 0.1% trifluoroacetic acid in Milli-Q water] and was placed on the sample plate to dry. The calibration of the instrument in low molecular mass range was carried out using 2 μl calibration mixtures of standards ranging from 3 to 20 kDa.

Thioflavin T assay

The kinetics of fibril formation was monitored by measuring the increase in fluorescence intensity of amyloid-specific benzothiazole dye — ThT. ThT binds exclusively to cross β-sheets of amyloid fibrils and thereby results in an increase in the fluorescence intensity [17]. Briefly, protein was diluted in phosphate buffer [20 mM NaH2PO4, 100 mM NaCl, and 0.01% NaN3 (pH 7.4)] to a final concentration of 20 μM; mixed with 5 μl of 1 mM ThT and incubated at room temperature for 30 min. Fluorescence spectra were recorded on a spectrofluorophotometer (RF-5301PC, Shimadzu, Japan) by exciting the samples at 450 nm and measuring the emission spectra in the range of 460−500 nm. The excitation and emission slit widths were kept at 5 nm. For time-dependent aggregation studies, stock solutions of recombinant and synthetic hIAPP were prepared in 100% HFIP, further diluted to a final concentration of 20 μM in 20 mM phosphate buffer saline (pH 7.4) and allowed for fibrillization at 37°C with agitation (170 rpm). The lag time (tlag) was calculated according to the published report [18] using the following equation:

 
formula

where y is the ThT fluorescence at a particular time-point, ymax is the maximum ThT fluorescence and y0 is the ThT fluorescence at t0.

The lag time (tlag) is defined as using the following equation:

 
formula

8-Anilinonaphthalene-1-sulfonate fluorescence measurement

An increase in the ANS (8-anilinonaphthalene-1-sulfonate) fluorescence intensity has generally been attributed to the increase in hydrophobicity of the sample [19]. Protein samples were diluted to a final concentration of 20 μM using phosphate buffer [20 mM NaH2PO4, 100 mM NaCl and 0.01% NaN3 (pH 7.4)]; mixed with 60 µM ANS dye and were further incubated at RT for 10 min in dark. The samples were excited at 370 nm and emission spectra were recorded between 400 and 600 nm using a spectrofluorophotometer (RF-5301PC, Shimadzu, Japan). The excitation and emission slit widths were kept at 3 nm. The stock solutions of recombinant and synthetic hIAPP were prepared in 100% HFIP, further diluted to a final concentration of 20 μM in 20 mM phosphate buffer saline (pH 7.4) and allowed for fibrillization at 37°C with agitation (170 rpm). The ANS fluorescence spectrum at the beginning and after 24 h was recorded in a similar way as discussed above.

Circular dichroism spectroscopy

Far UV-CD (circular dichroism) measurements were performed on a JASCO J-810 spectropolarimeter (JASCO, Japan) using a 0.1 cm path length cuvette (Hellma, Forest Hills, NY). Briefly, fusion protein and mature rhIAPP fibrils were diluted to a final concentration of 20 µM in 20 mM phosphate buffer (pH 7.4). All measurements were performed at 25°C. Raw data were smoothened and measurements from the buffer were subtracted from those of the protein samples. CD spectra have been represented as a plot of molar ellipticity (θ) in deg cm2 dmol−1 versus the wavelength.

Fourier-transform infrared spectroscopy

To determine the different secondary structural component of proteins, Fourier-transform infrared (FTIR) spectroscopy was performed. In the IR spectrum, amide I (1600–1700 cm−1) is a major class of band for the characterization of the protein [20]. Briefly, protein sample (20 µM) was spotted on the potassium bromide (KBr) pellet and dried immediately under IR lamp. 10 μl buffer spotted on another KBr pellet was used for determining the background spectra. Spectra were acquired in the region of 1600–1800 cm−1 as an average spectrum of 32 scans using a Vertex-80 FTIR system (Bruker, Germany). To analyze the secondary structural components of rhIAPP fibrils, spectra were subjected to Fourier self-deconvolution of amide I region (1600–1700 cm−1) followed by Lorentzian curve fitting using Opus 65 software (Bruker, Germany).

Atomic force microscopy

The mature rhIAPP fibrils were observed using an atomic force microscope (Asylum Research, CA, U.S.A.). Briefly, 40–50 μM proteins were deposited on freshly cleaved mica sheets, air-dried, washed twice with double-distilled water and were kept for drying at room temperature for ∼40 min. Scanning was performed in tapping mode using silicon nitride cantilever at a scan rate of 1 Hz. Random portions of the mica sheet were scanned to get desired images and to see the uniformity of the sample application. For AFM imaging of the synthetic peptide, 1 mM stock solution was prepared in HFIP, centrifuged to avoid any pre-aggregation, finally diluted to a final concentration of 40 μM concentration using 20 mM sodium phosphate saline buffer (pH 7.4) and was allowed to form aggregates for 48 h. The aggregates formed here were subjected to AFM imaging as discussed above.

Transmission electron microscopy

The morphology of rhIAPP fibril was analyzed under TEM. Samples were prepared by diluting protein to ∼40 μM final concentration in phosphate buffer [20 mM NaH2PO4, 100 mM NaCl and 0.01% NaN3 (w/v) (pH 7.4)]. The diluted sample was spotted on a carbon-coated grid (Electron Microscopy Sciences, Fort Washington, PA) and further incubated for 3 min. The grid was subsequently washed three times with distilled water and then stained with 1% (w/v) freshly prepared uranyl acetate solution. Images were taken using a FEI Tecnai G2 12 electron microscope at 120 kV with nominal magnifications of 260 000× and 60 000× and were recorded digitally using the SIS Mega view III imaging system (Olympus, U.S.A.).

ssNMR spectroscopy

For ssNMR spectroscopy, overexpression of protein was performed in M9 minimal media supplemented with 15N-labeled NH4Cl and 13C-labeled glucose as a sole source of nitrogen and carbon, respectively. More than 25 mg of uniformly labeled 13C–15N rhIAPP fibrils was collected by ultracentrifugation and washed 4–5 times with a buffer [20 mM NaH2PO4, 100 mM NaCl and 0.01% NaN3 (pH 7.4)]. The obtained pellet was packed into a 3.2 mm ZrO2 MAS rotor and stored at 4°C until further use. All the ssNMR experiments were carried out on a 750 MHz Bruker Ascend spectrometer (Bruker Biospin, Germany) using a 3.2 mm H/C/N E-free triple resonance probe-head operating at 10 kHz MAS with a sample temperature at 10°C. One-dimensional (1D) cross-polarization (CP) and two-dimensional (2D) Proton-Driven Spin Diffusion (PDSD), hNCA and hNCO, 1H/13C heteronuclear correlation (HETCOR) spectra of rhIAPP fibrils were acquired. 2D [13C–13C] PDSD was obtained using a 10 ms mixing period, spectral widths of 200 ppm (t1max ∼7.7 ms and t2max ∼13.05 ms) with 80 scans. The 2D [15N–13C] hNCA and hNCO were obtained by the transfer of the magnetization from 1H to 15N with a contact time of 1100 µs using 256 scans and 15N–13C (either Cα or CO) transfer was achieved using SPECIFIC CP with contact times of 3 ms. The flexible regions in the protein were identified by recording J-based 13C–1H correlation spectrum (HETCOR). In HETCOR, magnetization was transferred from 1H to 13C using the INEPT transfer scheme under the MAS condition. The spectra for 1H were referenced with respect to DSS (sodium salt of 2,2-dimethyl-2-silapentane-5-sulphonic acid), whereas those of 13C and 15N were referenced indirectly using the BMRB protocol [21]. The 1D spectrum was plotted in MestRa-C and 2D spectrum was processed in Topspin 3.2 (Bruker, Germany).

Cell culture

INS-1E cells (obtained as a kind gift from Prof. Claes Wollheim and Prof. Pierre Maechler; University; University Medical Centre, Geneva, Switzerland) between passages 56 and 65 were grown in monolayer cultures in a humidified 5% CO2 atmosphere at 37°C in RPMI 1640 media supplemented with 10 mM HEPES, 1 mM pyruvate, 50 μM 2-mercaptoethanol, 10% (v/v) heat-inactivated FBS, 100 units/ml penicillin and 100 μg/ml streptomycin [22].

Cytotoxicity assay

To determine rhIAPP-induced cytotoxicity, INS-1E cells were seeded onto 96-well plates (Corning) at an initial density of 1 × 104 cells/well and were allowed to adhere at 37°C for 24 h. Post-incubation, the medium was replaced with fresh culture medium (100 µl) containing increasing concentrations of recombinant and synthetic hIAPP (μM). Untreated wells containing only cells in culture media and those treated with PBS were evaluated as controls. After 24 h of treatment, the plates were incubated with 10 µl of MTT solution (5 mg/ml) for 4 h. Post-incubation, the MTT-containing media were carefully removed and formazan crystals were dissolved with 200 µl of DMSO for 10 min on a shaker. Absorbance was measured at 540 nm using a plate reader (Thermo Scientific). The reduction in cell viability was expressed as the percentage for each treatment relative to the control set (set as 100%). The experiments were performed three times with three replicate wells for each treatment.

Measurement of cell growth and loss in viability

For relative quantification of cell death and apoptosis, flow cytometry was performed using the Annexin V-FITC apoptosis detection kit (Invitrogen, U.S.A.), as described in Supplementary Material. Samples were acquired using FACS Verse (BD Biosciences, San Jose, CA) and analyzed using the BD FACS Diva software. Twenty thousand cells were analyzed for each sample. Cellular proliferation was assessed using a crystal violet assay [23], as described in Supplementary Methods. The relative surviving fraction was determined by measuring the absorbance of solubilized crystal violet staining in the rhIAPP-treated well normalized to the corresponding untreated well.

Mitochondrial membrane potential measurement

To determine the changes in mitochondrial membrane permeability, MMP was assessed using JC-1 dye (Thermo Fisher Scientific), as described in Supplementary Material. JC-1-stained cells were acquired using FACS Verse (BD Biosciences, San Jose, CA) and analyzed using the BD FACS Diva software. JC-1 forms J-aggregates that emit red fluorescence at 590 nm in healthy mitochondria, while J-monomers emit green fluorescence at 490 nm in depolarized mitochondria. An increased ratio of J-monomers indicates mitochondrial damage.

Pancreatic islet isolation and glucose-stimulated insulin secretion assay

Pancreatic islets were isolated from 8- to 10-week-old Swiss Albino mice maintainied under a standard 12 h light/dark cycle. Animals were housed in polypropylene cages at 25 ± 2°C and were given animal feed and water ad libitum. The study was conducted in accordance with the guidelines of the Committee for the purpose of Control and Supervision of Experiments on Animals (CPCSEA), and the experimental protocol was approved by the Institutional Animal Ethics Committee, SPPU, Pune (Registration no. 538/02/c/CPCSEA).

For islet isolation, mice were killed by cervical dislocation. Pancreatic tissue was dissected, minced thoroughly and digested with collagenase for 10 min at 37°C. The cellular digest was subsequently centrifuged at 2000 rpm for 4 min. Supernatant was discarded, and the pellet was resuspended in RPMI supplemented with 10% FBS and maintained at 37°C in CO2 incubator for 48 h. The isolated islets were treated with freshly sonicated 1.2 and 2.4 μM rhIAPP fibrils for 24 h. Following this incubation, the functionality of islets was checked by performing the GSIS assay. Briefly, ∼200 islets were placed in a single well of a 24-well plate containing a Krebs-Ringer bicarbonate HEPES (KRBH) buffer without glucose for 1 h. Post-incubation, the islets were treated with a KRBH buffer containing basal (5 mM) and stimulated (20 mM) glucose for 1 h. Supernatant was collected for estimating insulin concentrations using the Mouse insulin ELISA kit (Mercodia, Sweden). GSIS was expressed as the amount of insulin secreted under basal (5 mM glucose) and stimulated (16 mM glucose) conditions.

Immunofluorescence

INS-1E cells (1 × 105) were grown on coverslips in six-well plates (Corning) and were treated with freshly sonicated 1.2 and 2.4 μM rhIAPP for 24 h. Post-incubation, the cells were washed twice 1× PBS, fixed and stained with ThS and γH2AX, as described in Supplementary Material. The coverslips were mounted onto slides with ProLong Gold Antifade Mountant containing DAPI (Thermo Fisher). Imaging was performed on a confocal laser microscope (Nikon A1R), and the images were processed using the NIS elements Viewer (Nikon).

Reactive oxygen species measurement

INS-1E cells were plated on 96-well plates (Eppendorf) at an initial density of 1 × 104 cells/well and were allowed to adhere at 37°C for 24 h. Post-incubation, the medium was replaced with a fresh culture medium (100 μl) containing increasing concentrations of freshly sonicated rhIAPP. Following exposure to the peptides, reactive oxygen species (ROS) levels were assessed by incubating the cells with either DCFH-DA (10 μM) or MitoSox (5 μM) for 30 min at 37°C. Simultaneous detection of cell viability was also carried out using the MTT assay (as described above). The data were normalized to the number of viable cells.

Lipid peroxidation

INS-1E cells were seeded onto six-well plates at a density of 1 × 105 and were treated with freshly sonicated rhIAPP at the final concentration of 1.2 and 2.4 μM for 24 h. Lipid peroxidation was determined using the protocol, as described by Chen et al. [24]. After treatment with the mentioned doses of rhIAPP, the cells were lysed and boiled with 500 μl of TBA reagent (0.8% TBA in 20% TCA) at 95°C for 20 min. Tetramethoxy propane was used as a standard. The resulting fluorescence was measured at excitation 535 nm and emission 585 nm. The amount of protein was quantitated using BCA reagent and the results were expressed as nmol TBARs/mg protein.

Western blotting

INS-1E cells were seeded in six-well plates at a density of 1 × 105 and were subjected to different treatments for 24 h. Post-incubation, the cells were harvested and washed with 1× PBS, and total protein was extracted using a protein extraction buffer [20 mM Tris–HCl (pH 7.4), 1 mM EDTA, 1 mM PMSF and 0.1% Triton X-100, and 1× complete mini protease inhibitor cocktail (Roche)] and quantitated using BCA reagent (Thermo Fischer Scientific). An equal amount of proteins were analyzed by SDS–PAGE and electro-transferred to the polyvinylidene fluoride (PVDF) membrane (Bio-Rad) at 70 mA constant current. β-Actin, PARP, GRP-78, Caspase-3 and p-eIF-2α protein levels were determined by immunodetection using myECL Imager (Thermo Fisher Scientific).

Statistical analysis

All experiments were performed in triplicate, and the results have been presented as mean ± SD, unless stated otherwise. Statistical analysis was performed by one-way ANOVA and Tukey HSD post hoc test; a value of P < 0.05 was considered to be a significant difference between groups.

Results

Cloning, overexpression and purification of rhIAPP

A cost-effective protocol was developed in order to optimize a procedure for large-scale production of highly pure rhIAPP fibrils of uniform morphology (Figure 1A). The target peptide (hIAPP) was overexpressed as a fusion partner with GB1 protein and (His)6 tag in E. coli BL21 (DE3) cells (Supplementary Figure S1A). The GB1 protein tag (Protein G; B1 domain — 56 residues) has been extensively used as a solubility and stability enhancement tag to express insoluble proteins in E. coli that prevents the recombinant protein from forming inclusion bodies [25]. The (His)6 tag helps to facilitate protein purification using Ni-NTA affinity chromatography. In addition to these tags, two linkers and one additional methionine were added on the N- and C-termini of the peptide, respectively, to facilitate cleavage by CNBr. Briefly, protein expression was induced with 1 mM IPTG at 37°C for 5 h and the pellet was processed further. Initial attempts showed poor binding of the soluble protein onto Ni-NTA resin, possibly due to the incorporation of the (His)6 tag into the tertiary structure of the protein. Therefore, a hybrid approach (denaturation and column refolding) was used to expose the (His)6 tag that allowed binding of the overexpressed protein with Ni-NTA beads. Impurities were removed by subsequent washes in denaturating conditions (lane 3, Figure 1B). This was followed by on-column refolding to allow the protein to acquire its native state. The fused protein was eluted, dialyzed and subjected to gel filtration chromatography (Figure 1C). Around 73% (as quantitated by ImageJ 1.50b) of the protein after Ni-NTA column purification consisted of the fusion protein (lanes 6 and 7 in Figure 1B). Two major peaks were observed: peak I (90 ml elution volume) showed impurities along with the overexpressed fusion protein (Supplementary Figure S2A) and peak II (180 ml elution volume) showed pure fusion protein in the monomeric form (Supplementary Figure S2B). A sharp band corresponding to ∼12 kDa was observed for purified fusion protein (lane II, Figure 1D), which was further confirmed by MALDI-TOF mass spectrometry (Supplementary Figure S2D). The purified fusion protein was cleaved using excess of CNBr to remove the GB1 tag, linkers and (His)6 tag. The resulting full-length rhIAPP had one additional methionine at the C-terminus. It was observed that at high concentrations, the protein solution turned turbid possibly due to the rapid formation of rhIAPP aggregates upon cleavage from the fusion protein. Following digestion, the solution was subjected to ultracentrifugation. The pellet, thus obtained, was washed 4–5 times with a buffer and then subjected for the fibril formation. For fibrillization, the pellet was resuspended in a phosphate buffer and allowed to grow for 48 h at 37°C with continuous agitation.

After fibril formation, a small amount of the pelleted protein was further dissolved in 7 M Gdn–HCl and was subjected to MALDI-TOF and electrophoresis on 16.5% Tris–Tricine SDS–PAGE gel to confirm the purity of the preparation. A single peak was observed in the mass spectrum (at 4037 Da; Figure 1E) and on the Tris–Tricine SDS gel (∼4 kDa, lane 3; Figure 1F). The protocol yielded >25 mg of purified fibrils per liter of 2× YT medium.

Characterization of mature rhIAPP fibrils

The morphology and dimension of the mature fibrils thus formed was analyzed further using TEM (Figure 2A) and AFM (Figure 2B) imaging. A dense fibrillar network of rhIAPP, characteristic of amyloidogenic proteins, was observed (Figure 2A). The rhIAPP fibrils formed in the present study were homogeneous and consisted majorly of short, unbranched rhIAPP fibrils, with average length ranging 300–800 nm and width ranging 3–5 nm (Figure 2C). Even though a large majority of the fibrils acquired a straight ribbon-like morphology, a few twisted fibrils were also observed.

Morphological and structural characterization of mature rhIAPP fibrils.

Figure 2.
Morphological and structural characterization of mature rhIAPP fibrils.

(A) TEM analysis of negatively stained hIAPP preparation shows a dense network of fibrils. (B) Representative AFM image of hIAPP depicting morphology of the individual fibrils; uniform unbranched fibrils were observed. The red scale represents the width, while the black scale represents the length of the fibrils. (C) Analysis of fibril width (red) and length (black) as extracted from B. (D) Excitation spectrum for hIAPP fibril upon ThT binding (red). GB1 protein obtained after CNBr cleavage (blue) and buffer (green) were used as control. (E) Fluorescence spectrum of hIAPP fibril (red) upon ANS binding. GB1 proteins obtained after CNBr cleavage (blue) and buffer (green) were used as control. (F) Far UV-CD spectrum for hIAPP fibril (red). (G) FTIR spectrum of rhIAPP fibril (red: original spectra). Brown and blue represent spectra deconvoluted using Lorentzian fitted curve.

Figure 2.
Morphological and structural characterization of mature rhIAPP fibrils.

(A) TEM analysis of negatively stained hIAPP preparation shows a dense network of fibrils. (B) Representative AFM image of hIAPP depicting morphology of the individual fibrils; uniform unbranched fibrils were observed. The red scale represents the width, while the black scale represents the length of the fibrils. (C) Analysis of fibril width (red) and length (black) as extracted from B. (D) Excitation spectrum for hIAPP fibril upon ThT binding (red). GB1 protein obtained after CNBr cleavage (blue) and buffer (green) were used as control. (E) Fluorescence spectrum of hIAPP fibril (red) upon ANS binding. GB1 proteins obtained after CNBr cleavage (blue) and buffer (green) were used as control. (F) Far UV-CD spectrum for hIAPP fibril (red). (G) FTIR spectrum of rhIAPP fibril (red: original spectra). Brown and blue represent spectra deconvoluted using Lorentzian fitted curve.

The amyloidogenic propensity of rhIAPP was further validated by ThT binding assays. A high ThT fluorescence intensity indicated the presence of the cross β-sheet-rich amyloid fibril (Figure 2D, red). Therefore, the amyloidogenicity of the rhIAPP prepared by the current protocol was retained. A previous study by Krampert et al. [26] has reported the overexpression of fused hIAPP (His-tagged proIAPP; 67 residue protein); however, the amyloidogenicity of the proIAPP was lesser when compared with that of hIAPP (Table 1).

Table 1
Summary of results reported in the literature for recombinant hIAPP preparation
Author Cloning strategies Modification (amidation/non-amidation at C-terminus) Solubilization Yield (mg/l) Amyloidogenic properties Toxicity studies 
Krampert et al. [26pLysS-pET vector C-terminal S-tag ProIAPP or mutProIAPP Inclusion bodies 0.9 mg ProIAPP from 200 ml of bacterial culture ProIAPP (67 residue)-less amyloidogenic than native hIAPP Rat insulinoma cell line RIN5fm
ProIAPP—cytotoxic EC50 ∼—300 nm) 
Lopes et al. [32pLysS-pET, digestion Non-amidated Inclusion bodies 10 mg Amyloidogenic Primary islet cultures; cytotoxic (60 μM) 
Kosicka et al. [42pET32 vector Non-amidated Inclusion bodies 16 mg Amyloidogenic HEK293 cells
Cytotoxic(17.5 μM) 
Mirecka et al. [43]/Weirich et al. [44HI18-IAPP fusion construct ligated into pET302/NT-His vector Non-amidated (free-acid form) Total soluble protein fraction 3 mg Amyloidogenic NA 
Camargo et al. [41pGEX-TK-MIAPP Amidated Total soluble protein fraction 3 mg NA NA 
Current study pGEV1 vector the GB1 tag and (His)6 Non-amidated (one extra methionine at the C-terminus) Total soluble protein fraction >25 mg Amyloidogenic Pancreatic INS-1E cells
Cytotoxic (2.4 μM) 
Author Cloning strategies Modification (amidation/non-amidation at C-terminus) Solubilization Yield (mg/l) Amyloidogenic properties Toxicity studies 
Krampert et al. [26pLysS-pET vector C-terminal S-tag ProIAPP or mutProIAPP Inclusion bodies 0.9 mg ProIAPP from 200 ml of bacterial culture ProIAPP (67 residue)-less amyloidogenic than native hIAPP Rat insulinoma cell line RIN5fm
ProIAPP—cytotoxic EC50 ∼—300 nm) 
Lopes et al. [32pLysS-pET, digestion Non-amidated Inclusion bodies 10 mg Amyloidogenic Primary islet cultures; cytotoxic (60 μM) 
Kosicka et al. [42pET32 vector Non-amidated Inclusion bodies 16 mg Amyloidogenic HEK293 cells
Cytotoxic(17.5 μM) 
Mirecka et al. [43]/Weirich et al. [44HI18-IAPP fusion construct ligated into pET302/NT-His vector Non-amidated (free-acid form) Total soluble protein fraction 3 mg Amyloidogenic NA 
Camargo et al. [41pGEX-TK-MIAPP Amidated Total soluble protein fraction 3 mg NA NA 
Current study pGEV1 vector the GB1 tag and (His)6 Non-amidated (one extra methionine at the C-terminus) Total soluble protein fraction >25 mg Amyloidogenic Pancreatic INS-1E cells
Cytotoxic (2.4 μM) 

Binding of ANS — an extrinsic fluorescence probe to the exposed hydrophobic regions of protein was used as an additional tool to validate fibril formation. High ANS fluorescence was observed for the mature rhIAPP fibrils (red; Figure 2E). On the other hand, no ANS fluorescence was observed for GB1 (blue; Figure 2E). This confirmed the fact that ANS binds more appreciably to exposed hydrophobic surface of cross-β-sheet conformations of fibrils [27].

The secondary structure of mature rhIAPP fibrils was also analyzed by far UV-CD spectroscopy and FTIR spectroscopy. CD spectroscopy showed a negative peak ∼220 nm, suggesting a β-sheet conformation (Figure 2F). Similarly, the FTIR spectrum of rhIAPP fibril showed majority of a peak (amide 1 stretching frequency) at 1631 cm−1, thereby suggesting a predominant β-sheet conformation (Figure 2G). Consistent with our studies, experimental evidence from previous studies also suggests that the full length and shorter fragments of hIAPP fibrils show the presence of β-sheet structure [28].

ssNMR spectroscopy of rhIAPP fibril

The resolution and dispersion spectra obtained from the carbon–carbon and nitrogen–carbon correlation experiments, using MAS ssNMR experiments, provide first-hand information about molecular organization and quality of fibril produced for structural studies [29].

For ssNMR analysis, rhIAPP was expressed in the M9 minimal media, supplemented with 15N-labeled NH4Cl and 13C-labeled glucose as a source of nitrogen and carbon, respectively. The morphology of the purified fibrils was checked with AFM imaging. For structural investigation, 1D 13C CP and 2D homonuclear 13C–13C correlation [PDSD — proton-driven spin diffusion and heteronuclear 13C–15N correlation (hNCA and hNCO)] spectra were acquired. 1D CP spectrum showed sharp peaks in aliphatic region (15–80 ppm) (Figure 3A). 2D 13C–13C PDSD spectra showed well-resolved resonances at 10 ms mixing time, consisting of mostly intra-residue correlation cross-peaks. The line widths of cross-peaks were in the range of ∼1–1.5 ppm, thereby indicating for a well-ordered structure of the formed fibrils. Preliminary analysis of chemical shifts was performed based on the 13Cα and 13Cβ random coil chemical shift values. On comparison of chemical shift values of Cα and Cβ, we observed that the spin systems such as threonine, serine and valine in PDSD spectrum showed the characteristic of β-sheet conformation in the fibrillar state (Figure 3B; pink). Values of 13Cα for all these residues were lower than the random coil values, while values of 13Cβ were higher. The heteronuclear 13C–15N correlation, hNCO and hNCA spectrum, also showed well-resolved resonances in carbonyl and aliphatic regions (Figure 3C; cyan and Figure 3D; black). For complete structural characterization of rhIAPP fibrils, higher dimensional spectra on differently labeled samples will need to be acquired and analyzed. Subsequently, we compared the 2D homonuclear 13C–13C PDSD and heteronuclear 13C–15N correlation (hNCA) spectrum reported in the present study with the chemical shift reported by Weirich et al. For this, we extracted the 13Cα, 13Cβ and NH chemical shifts from the 13C–13C and 13C–15N correlation spectra, simulated and overlaid them with 13C–13C (right panel; purple in Figure 3B) and 13C–15N correlation spectra (right panel; red in Figure 3D) obtained from the fibrils prepared by the current procedure. We observed a similar pattern in the spectrum, but there was a difference in peak positions, possibly due to the difference in the morphology of the fibrils. We also performed 2D 1H–13C HETCOR to probe the flexible residues at the molecular level. We obtained well-resolved peaks with chemical shift values typical for random coil conformation (Supplementary Figure S3), indicating the presence of few monomeric species in the gel-like fibril matrix, as reported previously [30].

ssNMR spectrum of uniformly labeled 13C–15N hIAPP fibrils.

Figure 3.
ssNMR spectrum of uniformly labeled 13C–15N hIAPP fibrils.

The spectrum is acquired with 10 kHz MAS at 750 MHz spectrometer. (A) 1D 13C CP-MAS spectrum. (B) 2D 13C–13C PDSD correlation spectrum (mixing time 10 ms) showing the aliphatic and carbonyl regions (pink). (C and D) 2D 13C–15N hNCO (cyan) and hNCA (black) correlation spectrum showing carbonyl regions and Cα correlation with respective amide nitrogen, respectively. The purple and red spectra in panels B and D, respectively, indicates the overlaid spectra of 13C–13C (right panel; purple) and 13C–15N (right panel; red) of rhIAPP fibril, respectively, as reported by Weirich et al. [44] (BMRB entry 26706).

Figure 3.
ssNMR spectrum of uniformly labeled 13C–15N hIAPP fibrils.

The spectrum is acquired with 10 kHz MAS at 750 MHz spectrometer. (A) 1D 13C CP-MAS spectrum. (B) 2D 13C–13C PDSD correlation spectrum (mixing time 10 ms) showing the aliphatic and carbonyl regions (pink). (C and D) 2D 13C–15N hNCO (cyan) and hNCA (black) correlation spectrum showing carbonyl regions and Cα correlation with respective amide nitrogen, respectively. The purple and red spectra in panels B and D, respectively, indicates the overlaid spectra of 13C–13C (right panel; purple) and 13C–15N (right panel; red) of rhIAPP fibril, respectively, as reported by Weirich et al. [44] (BMRB entry 26706).

Recombinant hIAPP alters mitochondrial membrane potential, inhibits glucose-stimulated insulin secretion and induces cytotoxicity in pancreatic cells

It is believed that the amyloid aggregates formed by hIAPP induce apoptosis and dysfunction in pancreatic β-cells; shorter aggregates have been reported to be more cytotoxic [31]. However, the exact mechanism by which hIAPP mediates cytotoxicity is unknown. Previous studies have reported prominent cytotoxicity mediated by rhIAPP in purified human islets [32] and cell lines of diverse origins, including RIN5F and HEK293 (Table 1).

In the present study, the cytotoxic response of rhIAPP was investigated on INS-1E cells, a model system that has been used successfully for demonstrating the toxicity of synthetic hIAPP peptide in the β-cells [33,34]. The addition of recombinant peptide to INS-1E cells affected their survival and morphology at 24 h (Figure 4A; Supplementary Figure S4). Additionally, INS-1E cells treated with sub-optimal (viability ∼74.4 ± 5.6%; 1.2 μM rhIAPP) and IC50 (viability ∼50.54 ± 6.7%; 2.4 μM rhIAPP) concentrations of rhIAPP were also analyzed for Annexin V/PI staining on flow cytometer (Figure 4B). The cells undergoing apoptosis increased upon the addition of the recombinant peptide (Figure 4C; P < 0.05). In addition to this, the proliferation capacity of INS-1E cells decreased linearly with increasing concentration of the peptide, as was determined by the crystal violet assay (Figure 4D).

rhIAPP affects β-cell function and induces cytotoxicity.

Figure 4.
rhIAPP affects β-cell function and induces cytotoxicity.

(A) INS-1E cells were treated with different concentrations of rhIAPP for 24 h and cell viability was measured using the MTT assay (N = 4). (B) Representative images of Annexin V/PI staining of control and rhIAPP cells; the upper left quadrant (UL) represents dead cells (PI positive only); the upper right panel (UR) represents population of cells in late stages of apoptosis (Annexin V-FITC and PI positive); lower left panel (LL) represents live cells (Annexin V-FITC and PI negative); and the lower right (LR) panel represent cells in early stages of apoptosis (Annexin V-FITC positive). (C) Quantitation of Annexin V/PI-stained cells (N = 3). (D) Exposure of rhIAPP in INS-1E cells affects their proliferation ability (N = 4). (E) Representative histograms showing flow cytometric analysis of JC-1 staining in untreated (control) and rhIAPP-treated cells; the upper left quadrant (UL) represents red cells ( JC-1 aggregate only); the upper right panel (UR) represents the population of cells positive for JC-1 aggregate and monomer; lower left panel (LL) represents JC-1 negative and the lower right (LR) panel represents with loss in MMP ( JC-1 monomer positive). (F) Quantitation of JC-1-stained cells (N = 3): Population of JC-1 monomer (%) has been plotted along y-axis. (G) ThS staining in INS-1E cells exposed to rhIAPP. (H) Pancreatic islets (isolated from Swiss Albino mice) treated with different concentrations of rhIAPP showed significantly reduced insulin secretion in GSIS assay (N = 4). Basal glucose concentration was maintained at 5 mM (red), while islets were stimulated with 16 mM glucose (gray) under high glucose conditions.

Figure 4.
rhIAPP affects β-cell function and induces cytotoxicity.

(A) INS-1E cells were treated with different concentrations of rhIAPP for 24 h and cell viability was measured using the MTT assay (N = 4). (B) Representative images of Annexin V/PI staining of control and rhIAPP cells; the upper left quadrant (UL) represents dead cells (PI positive only); the upper right panel (UR) represents population of cells in late stages of apoptosis (Annexin V-FITC and PI positive); lower left panel (LL) represents live cells (Annexin V-FITC and PI negative); and the lower right (LR) panel represent cells in early stages of apoptosis (Annexin V-FITC positive). (C) Quantitation of Annexin V/PI-stained cells (N = 3). (D) Exposure of rhIAPP in INS-1E cells affects their proliferation ability (N = 4). (E) Representative histograms showing flow cytometric analysis of JC-1 staining in untreated (control) and rhIAPP-treated cells; the upper left quadrant (UL) represents red cells ( JC-1 aggregate only); the upper right panel (UR) represents the population of cells positive for JC-1 aggregate and monomer; lower left panel (LL) represents JC-1 negative and the lower right (LR) panel represents with loss in MMP ( JC-1 monomer positive). (F) Quantitation of JC-1-stained cells (N = 3): Population of JC-1 monomer (%) has been plotted along y-axis. (G) ThS staining in INS-1E cells exposed to rhIAPP. (H) Pancreatic islets (isolated from Swiss Albino mice) treated with different concentrations of rhIAPP showed significantly reduced insulin secretion in GSIS assay (N = 4). Basal glucose concentration was maintained at 5 mM (red), while islets were stimulated with 16 mM glucose (gray) under high glucose conditions.

Depolarization of MMP has been reported in cells undergoing apoptosis. JC-1 has the property of forming aggregates in cells with high MMP, while it exists as a monomer that emits green fluorescence at low MMP. Consistent with our earlier results, a progressive loss of red JC-1 aggregates and appearance of green JC-1 monomer was observed in INS-1E cells treated with increasing concentrations of rhIAPP (Figure 4E), thereby indicating a loss of MMP (Figure 4F). Indeed, the INS-1E cells exposed to rhIAPP showed positive staining with ThS (Figure 4G), thereby hinting toward the presence of amyloid structures in these cells.

The effect of rhIAPP exposure on the ability of pancreatic islets to secrete insulin, assessed using primary islet culture, has been depicted in Figure 4H. Islets were isolated from Swiss Albino mice and treated with different concentrations of rhIAPP. In the control islets, the amount of insulin release increased from 1.11 ± 0.01 μIU/ml under basal glucose conditions to 2.59 ± 0.47 μIU/ml under glucose-stimulated conditions (P < 0.05). On the other hand, in rhIAPP-treated islets, insulin released under stimulated conditions significantly reduced when compared with islets exposed to basal glucose concentrations (P < 0.05).

Recombinant hIAPP mediates its action via generation of oxidative stress and double-strand breaks in INS-1E cells

To further gain insights into the potential mechanism of cytotoxicity, we analyzed the generation of ROS at sub-optimal and IC50 concentrations of rhIAPP. As can be seen from Figure 5A, we observed a significant increase in intracellular ROS levels (determined using DCFH-DA that reacts with multiple ROS) at 24 h. A parallel increase in mitochondrial superoxide was also observed using the mitochondrial-specific probe — MitoSox. In addition to this, there was a significant increase in the lipid peroxides, as indicated by the thiobarbituric acid reactive substance (TBAR) assay (Figure 5B), in the rhIAPP-treated cells, further confirming that the damage occurred as a result of increased oxidative stress. Consistent with these findings, a significant increase in DNA double-strand breaks (as visualized by γH2AX foci; Figure 5C) was also observed in INS-1E cells treated with IC50 concentrations of rhIAPP (2.4 μM) for 24 h. The transcript levels of apoptotic genes — Caspase-3 and Bax — were found to be up-regulated, while the levels of the anti-apoptotic gene Bcl2 were down-regulated in rhIAPP-treated cells (Figure 5D). Also, the expression levels of gene involved in oxidative stress namely Catalase, SOD1 and GPx were down-regulated (Figure 5D). The protein levels of cleaved PARP and Caspase-3 were also up-regulated in rhIAPP-treated cells. rhIAPP-treated cells also underwent increased ER stress as can be seen by increased levels of GRP78 and increased phosphorylation of eIF-2α (Figure 5E). Collectively, our data indicate that oxidative stress-induced lipid peroxidation and corresponding DNA damage could be the major causes of cell death post-rhIAPP exposure in INS-1E cells.

Recombinant hIAPP mediates its action via generation of oxidative stress and double-strand breaks in INS-1E cells.

Figure 5.
Recombinant hIAPP mediates its action via generation of oxidative stress and double-strand breaks in INS-1E cells.

(A) Intracellular and mitochondrial ROS production measured in control and rhIAPP-treated cells as measured by DCFH-DA and Mitosox, respectively (N = 3). (B) Lipid peroxidation was measured in control and rhIAPP-treated cells in terms of nmol of TBARs/mg protein (N = 3). (C) Increased γH2AX staining was observed in rhIAPP-treated cells, thereby indicating increased DNA damage (N = 3). (D) Transcript levels of gene involved in apoptosis (Caspase 3, Bax and Bcl2) and oxidative stress (Catalase, SOD1 and GPx) (N = 3). (E) Western blot analysis of PARP, phospho-eIF2α, GRP-78, Caspase-3 and β-actin (N = 3). * indicates P < 0.05 with their respective control.

Figure 5.
Recombinant hIAPP mediates its action via generation of oxidative stress and double-strand breaks in INS-1E cells.

(A) Intracellular and mitochondrial ROS production measured in control and rhIAPP-treated cells as measured by DCFH-DA and Mitosox, respectively (N = 3). (B) Lipid peroxidation was measured in control and rhIAPP-treated cells in terms of nmol of TBARs/mg protein (N = 3). (C) Increased γH2AX staining was observed in rhIAPP-treated cells, thereby indicating increased DNA damage (N = 3). (D) Transcript levels of gene involved in apoptosis (Caspase 3, Bax and Bcl2) and oxidative stress (Catalase, SOD1 and GPx) (N = 3). (E) Western blot analysis of PARP, phospho-eIF2α, GRP-78, Caspase-3 and β-actin (N = 3). * indicates P < 0.05 with their respective control.

Comparison of cytotoxicity and morphological characteristics of synthetic and recombinant hIAPP

After getting an insight into the mechanism of rhIAPP-induced cell death, the cytotoxic behavior of synthetic and recombinant hIAPP was compared using MTT and ROS measurement assays. We would like to mention here that the synthetic hIAPP was solubilized using HFIP. On the other hand, rhIAPP was solubilized overnight in DMSO. The solution thus obtained was sonicated for 2 h with 2 min ON and 2 min OFF cycles in a water bath sonicator. All cellular studies were carried out after dilution of this solution in a phosphate buffer.

On comparing the cytotoxicity of the synthetic and recombinant hIAPP, it was observed that rhIAPP appeared to be more cytotoxic (IC50 = 2.4 μM) when compared with the synthetic peptide (IC50 = 4.8 μM) (Figure 6A). It was also observed that both recombinant and synthetic forms mediated cell death via increased ROS production in INS-1E cells (Figure 6B). MALDI-TOF analysis indicated almost similar molecular mass for the two preparations (Figure 6C). Also, the HPLC profile for the synthetic peptide (data not shown; provided in the Peptide 2.0 data sheet) indicated that the synthetic peptide was pure and free of impurities. Therefore, we speculated structural and morphological differences in the synthetic and recombinant hIAPP to be the reason for such differences in cellular toxicity. Indeed, upon comparison of AFM images, we observed that the synthetic hIAPP formed longer fibrils on dissolution and bigger aggregates upon sonication prior to addition to INS-1E cells (Figure 6E). On the other hand, solubilized rhIAPP fibrils were shorter and formed smaller aggregates upon sonication (Figure 6D). The average length of recombinant hIAPP fibril ranged 300–800 nm (Figure 6F). Conversely, the average length of the synthetic fibril ranged 1–7 μm, but the majority ranged 1–4 μm (Figure 6G). Hence, we could conclude that sonicated rhIAPP was more cytotoxic possibly due to the formation of shorter fibrils/oligomers when compared with the synthetic peptide (Figure 6H). Indeed, ThT fluorescence assay performed with mature and sonicated forms of recombinant and synthetic hIAPP also showed that the amyloid aggregates were smaller for the recombinant form (Figure 6I). However, no significant change was observed in terms of the hydrophobic surface for the two (Figure 6J).

Morphological properties and cytotoxic-mediated studies of recombinant and synthetic hIAPP.

Figure 6.
Morphological properties and cytotoxic-mediated studies of recombinant and synthetic hIAPP.

(A) INS-1E cells were treated with different concentrations of recombinant and synthetic hIAPP for 24 h and cell viability was measured using the MTT assay (N = 3). (B) Intracellular and mitochondrial ROS production measured in recombinant and synthetic hIAPP-treated cells as measured by DCFH-DA and Mitosox, respectively (N = 3). (C) MALDI-TOF spectrum of recombinant and synthetic hIAPP; AFM image of (D) recombinant hIAPP fibrils (before and after sonication). (E) Synthetic hIAPP1–37 fibrils (before and after sonication). The distribution of the number of fibrils obtained from AFM imaging (F) recombinant and (G) synthetic hIAPP fibrils. (H) Measurement of the height and length of the recombinant and synthetic hIAPP fibrils after sonication. Values for height and length extracted from AFM of D and E (after sonication), respectively. (I) ThT fluorescence assay for recombinant and synthetic hIAPP fibrils before and after sonication. (J) ANS fluorescence assay for recombinant and synthetic hIAPP fibrils before and after sonication.

Figure 6.
Morphological properties and cytotoxic-mediated studies of recombinant and synthetic hIAPP.

(A) INS-1E cells were treated with different concentrations of recombinant and synthetic hIAPP for 24 h and cell viability was measured using the MTT assay (N = 3). (B) Intracellular and mitochondrial ROS production measured in recombinant and synthetic hIAPP-treated cells as measured by DCFH-DA and Mitosox, respectively (N = 3). (C) MALDI-TOF spectrum of recombinant and synthetic hIAPP; AFM image of (D) recombinant hIAPP fibrils (before and after sonication). (E) Synthetic hIAPP1–37 fibrils (before and after sonication). The distribution of the number of fibrils obtained from AFM imaging (F) recombinant and (G) synthetic hIAPP fibrils. (H) Measurement of the height and length of the recombinant and synthetic hIAPP fibrils after sonication. Values for height and length extracted from AFM of D and E (after sonication), respectively. (I) ThT fluorescence assay for recombinant and synthetic hIAPP fibrils before and after sonication. (J) ANS fluorescence assay for recombinant and synthetic hIAPP fibrils before and after sonication.

Recombinant hIAPP aggregates faster than synthetic hIAPP

Several previous studies have used either DMSO or HFIP to dissolve synthetic hIAPP and have performed biophysical and cellular studies with this peptide [3436]. However, we refrained from using HFIP to dissolve the recombinant peptide and used DMSO in order to avoid high cytotoxicity in INS-1E cells with HFIP alone. In fact, biophysical studies with both ThT and ANS fluorescence spectroscopy revealed that rhIAPP spontaneously aggregated to cross β-sheet-rich amyloid fibrils and showed greater hydrophobicity in DMSO (Supplementary Figure S5). Therefore, to understand and compare the kinetics of aggregation and morphological properties of recombinant and synthetic hIAPP under similar solvent conditions in vitro, HFIP was used. The stock solution of synthetic hIAPP was prepared, as described earlier in the literature [37]. For characterization of rhIAPP, the pellets obtained after CNBr cleavage were washed several times with a phosphate buffer and dissolved in 100% HFIP (conditions identical with the synthetic peptide). The solution was subsequently centrifuged and the supernatant was used in both cases to ensure the presence of hIAPP monomer (as shown in MALDI-TOF spectrum; Figure 7A). Time-dependent aggregation kinetics for both these preparations was performed by diluting the peptide in a 20 mM phosphate buffer, 100 mM NaCl (pH 7.4). Our results showed that rhIAPP (lag time ∼30 min) aggregates significantly faster with higher ThT fluorescence intensities than the synthetic hIAPP (lag time ∼2.02 h) over the time period of 24 h (Figure 7B). Similarly, higher fluorescence intensity of ANS-bound rhIAPP also confirmed greater hydrophobic surface exposure when compared with synthetic hIAPP (Figure 7C). Far UV-CD spectra of recombinant and synthetic hIAPP confirmed the random coil conformation of hIAPP monomer at the beginning. These random coils acquired a β-sheet-rich structure after 24 h of incubation (Figure 7D). The morphology of recombinant and synthetic hIAPP after 48 h was also analyzed using AFM imaging. A dense network of fibrils with similar morphology for both peptides was observed. Interestingly, rhIAPP fibrils were shorter when compared with synthetic hIAPP (Figure 7E,F). The average length of rhIAPP fibrils generated by this procedure was also similar to the rhIAPP fibrils generated after CNBr cleavage (Figure 6D,F).

Comparison between recombinant and synthetic hIAPP prepared using identical solvent conditions.

Figure 7.
Comparison between recombinant and synthetic hIAPP prepared using identical solvent conditions.

(A) MALDI-TOF spectrum of recombinant (pellet obtained after CNBr cleavage digestion) and synthetic hIAPP dissolved in HFIP solvent. (B) Measurement of aggregation kinetics of recombinant and synthetic hIAPP using ThT fluorescence. (C) ANS fluorescence measurement of recombinant and synthetic hIAPP at initial time-point and after 24 h of incubation. (D) Far UV-CD spectroscopy of recombinant and synthetic hIAPP at the beginning and after 24 h of incubation. (E) AFM images of rhIAPP and synthetic hIAPP fibrils. (F) Measurement of height and length of the recombinant and synthetic hIAPP fibrils extracted from AFM images from (E).

Figure 7.
Comparison between recombinant and synthetic hIAPP prepared using identical solvent conditions.

(A) MALDI-TOF spectrum of recombinant (pellet obtained after CNBr cleavage digestion) and synthetic hIAPP dissolved in HFIP solvent. (B) Measurement of aggregation kinetics of recombinant and synthetic hIAPP using ThT fluorescence. (C) ANS fluorescence measurement of recombinant and synthetic hIAPP at initial time-point and after 24 h of incubation. (D) Far UV-CD spectroscopy of recombinant and synthetic hIAPP at the beginning and after 24 h of incubation. (E) AFM images of rhIAPP and synthetic hIAPP fibrils. (F) Measurement of height and length of the recombinant and synthetic hIAPP fibrils extracted from AFM images from (E).

Discussion

hIAPP is among the most amyloidogenic proteins known to date and has been implicated as the major cause of β-cell malfunction and apoptosis in T2DM [38]. Therefore, in-depth studies about the intermediates in the aggregation pathway and the associated cytotoxic responses need to be performed to gain deeper insights into the role of hIAPP in the pathophysiology of T2DM. Most aggregation studies with hIAPP reported so far have been performed with peptides prepared using solid-phase synthesis methods. In addition to increasing the overall cost of the experiment, solid-phase synthesis of amyloidogenic peptides has been reported to be difficult. For example, batch-to-batch variation has been reported with aggregation kinetics studies performed with another chemically synthesized amyloidogenic Aβ1–42 peptide [39,40]. Also, the peptides formed are often contaminated by the presence of a variable amount of intrinsic impurities incorporated during synthesis. As a result, it is often unlikely to reproduce the results obtained from such studies. It is also important to note that amyloidogenicity of the peptide can be sensitive to very small changes in the pH and the chemical composition of the buffer. For example, most studies reported so far have used buffers containing organic compounds like HFIP (1–2% v/v) for dissolution of the peptide which have been known to affect the rate of amyloid formation [15]. Therefore, in order to overcome these problems, a dire need exists to develop alternative strategies that would provide large quantities of purified recombinant peptides for performing structural and cellular studies. In this context, the production of rhIAPP in E. coli offers a suitable alternative for large-scale production of peptide.

A few protocols have been reported in the literature for the production of rhIAPP using E. coli, as pro-peptides and as fusion protein [26,32,4143]. Nonetheless, most of these protocols suffer from technical difficulties that include: low yield of peptide; expression of the peptide in inclusion bodies which makes the purification procedure cumbersome and time-consuming; and also low amyloidogenicity of the purified protein (Table 1). Therefore, it becomes imperative to develop a straightforward method to obtain high-quality, purified peptide in a large amount for structural and cellular studies. The studies by Kosicka et al. [42] and Lopes et al. [32] report the production of ∼16 and ∼10 mg/l of purified hIAPP fibrils from the inclusion bodies making the purification procedure cumbersome. Most other studies report the production of <5 mg/l of recombinant hIAPP (mature and/or fusion) from the soluble protein fraction. In contrast with these studies, the protocol used in the present study is relatively simple, straightforward and suitable for the production of large quantities of non-amidated rhIAPP from the soluble fraction. Most previous studies also report the production of non-amidated hIAPP in free-acid form [32,44]. Although the physiological form of the peptide is amidated and has been shown to be more potent than the non-amidated form, previous reports suggest that the amidation status does not affect its biological functions [32]. Similar to earlier reports, the rhIAPP prepared by our protocol mediated cytotoxicity in the rat insulinoma INS-1E and aggregated spontaneously into cross-β-sheet amyloid fibrils with higher hydrophobicity. These observations suggested that the additional methionine at the C-terminus did not significantly affect the structural properties and amyloidogenicity of rhIAPP and hence, its effect on mediating cytotoxicity. On the other hand, reports on Aβ42 being more cytotoxic than Aβ40 because of the addition of isoleucine and alanine at the C-terminus of the peptide cannot be ignored [45].

Multiple mechanisms have been proposed for hIAPP-induced cell death [9,15,46,47]. These include generation of oxidative stress, ER stress, mitochondrial membrane damage, permeabilization of cellular membranes and activation of cell death triggering pathways. Previous reports suggest that hIAPP amyloid fibrils tend to cluster on/near membranes, and exogenously added hIAPP has been shown to perturb cell membranes [31,4850]. In the present study, the mature rhIAPP fibrils formed a gel-like matrix which was difficult to use for cellular studies. Hence, rhIAPP fibrils were dissolved in DMSO, sonicated and then applied extracellularly to the INS-1E cells. Thiofalvin-S staining revealed the presence of amyloids in INS-1E cells exposed to rhIAPP. We observed that these species were shown to affect membrane permeability, mitochondrial membrane potential and induce oxidative stress in INS-1E cells. It was also observed that rhIAPP formed shorter oligomeric species when compared with the synthetic peptide in vitro (Figure 6D). In addition to this, the rhIAPP exposure to pancreatic islets affected their functionality by significantly reducing insulin secretion in response to glucose stimulation. Multiple apoptotic pathways have been shown to be affected upon hIAPP addition in immortalized β-cells or rodent islets [51]. We also observed an increase in DNA double-strand breaks and an up-regulation of apoptotic marker (including Caspase-3 and cleaved PARP) levels upon rhIAPP exposure in INS-1E cells. Also, more detailed studies would be required to comment on whether the aggregates/oligomeric species or the fibrillar species of hIAPP are involved in mediating cytotoxicity.

Electron micrographs of hIAPP fibrils from previous reports suggest that polymorphism exists among the fibrils [52,53]. The fibrils have been shown to organize as twisted fibrils that exhibit predominantly left-handed twisted [54] and straited ribbon-like morphology [52]. These fibrils were reported to be several nanometers long with width ranging from 5 to 15 nm [52,55]. The rhIAPP fibrils formed in the present study were typically short fibrils with a length of 300–800 nm and a width of 5 nm. In addition to this, experimental evidence from earlier studies also suggests that the full-length and shorter fragments of hIAPP fibrils show the presence of β-sheet structure [55,56]. These were confirmed by X-ray studies which indicated that the peptide chains were arranged in cross-β-sheet conformation, in which the individual β-strands were found to be perpendicular to the fibril axis [28]. Fluorescence studies conducted with fibrils formed in the currrent study confirmed the greater hydrophobicity and cross-β-sheet structure of the rhIAPP fibril. The presence of the β-sheet conformation was also confirmed by CD and FTIR spectroscopy. Additional data from preliminary ssNMR spectroscopy also confirmed the predominance of β-sheet in rhIAPP fibrils.

A recent study by Weirich et al. [44] reported a method for the expression of uniformly 13C–15N-labeled rhIAPP in the non-amidated, free-acid form and the corresponding ssNMR spectra of fibrils. The authors attempted to elucidate the structural basis of hIAPP amyloid formation and also investigated the morphology of the fibrils. We extracted the chemical shift values from their spectra, simulated and overlaid with the 13C–13C and 13C–15N correlation spectra obtained from the fibrils prepared by the current method. We observed a similar pattern; however, there was a significant difference in the individual peak positions. These differences in the peak positions could possibly be due to the difference in morphology of the rhIAPP fibrils obtained by the two protocols. We could detect distinct morphological differences in the fibrils obtained using the two distinct protocols. The AFM images obtained by their group showed laterally assembled bundles of hIAPP (designated as IAPPCOOH). They failed to detect any single fibril and/or amorphous aggregates. Hence, they could not comment on the sample homogeneity and polymorphism of the fibrils. On the other hand, rhIAPP fibrils formed instantly upon CNBr digestion using the current protocol. In the present study, we could detect homogeneous fibrils in the TEM and AFM images (Figure 2A,B). There were no amorphous aggregates detected in our preparation consistent with the earlier study [44]. However, the presence of monomeric species in the fibrillar matrix cannot be ruled out. Previously, it was reported on the basis of 2D 1H–13C hetronuclear correlation spectroscopy (HETCOR) experiments that monomers exist in the fibrillar matrix for other amyloidogenic protein [30]. It was shown that almost all amino acid types exhibited signals in the protein chain, indicating the incorporation of residual monomeric species into the gel-like fibril sample. In the current investigation as well, HETCOR spectrum under MAS condition showed all amino acid-type resonances, suggesting the presence of monomeric species embedded into the fibrillar matrix (Supplementary Figure S3).

In conclusion, the results from the present study describe an efficient and cost-effective production of rhIAPP fibrils in high yield that exhibit amyloidogenic and cytotoxic properties similar to the synthetic peptides in experiments conducted in vitro. These advancements in the methods would help gain further information about the mechanism of aggregation of hIAPP and the residue-specific structural details into the different intermediates along the aggregation pathway. Our results also suggest that changes in mitochondrial membrane permeability, oxidative stress-induced lipid peroxidation and corresponding DNA damage could be the major causes of cell death post-rhIAPP exposure in pancreatic INS-1E β-cells. In addition, we observed that rhIAPP is more cytotoxic (IC50 2.4 μM) due to the formation of smaller aggregates on comparison with synthetic peptide. In fact, rhIAPP aggregated faster than the synthetic peptide. The proposed protocol can further be extrapolated and can be used for studies of proteins associated with other amyloidogenic disorders.

Abbreviations

     
  • 1D

    one-dimensional

  •  
  • 2D

    two-dimensional

  •  
  • AFM

    atomic force microscopy

  •  
  • ANS

    8-anilinonaphthalene-1-sulfonate

  •  
  • CNBr

    cyanogen bromide

  •  
  • CP

    cross-polarization

  •  
  • CPCSEA

    Committee for the purpose of Control and Supervision of Experiments on Animals

  •  
  • DMSO

    dimethyl sulfoxide

  •  
  • ER

    endoplasmic reticulum

  •  
  • FTIR

    Fourier transforms infrared

  •  
  • GB1

    Protein G, B1 domain

  •  
  • GPx

    glutathione peroxidase

  •  
  • GSIS

    glucose-stimulated insulin secretion

  •  
  • HETCOR

    heteronuclear correlation

  •  
  • HFIP

    hexafluoroisopropanol

  •  
  • hIAPP

    human islet amyloid polypeptide

  •  
  • KBr

    potassium bromide

  •  
  • KRBH

    Krebs-Ringer biocarbonate HEPES buffer

  •  
  • MALDI-TOF

    matrix-assisted laser desorption-time of flight

  •  
  • MAS

    magic angle spinning

  •  
  • MMP

    mitochondrial membrane potential

  •  
  • PARP

    poly (ADP-ribose) polymerase

  •  
  • PDSD

    proton-driven spin diffusion

  •  
  • PMSF

    phenylmethylsulphonyl fluoride

  •  
  • rhIAPP

    recombinant hIAPP

  •  
  • ROS

    reactive oxygen species

  •  
  • SOD1

    superoxide dismutase 1

  •  
  • ssNMR

    solid-state NMR

  •  
  • T2DM

    type 2 diabetes mellitus

  •  
  • TBA

    thiobarbituric acid

  •  
  • TBARs

    thiobarbituric acid reactive substances

  •  
  • TEM

    transmission electron microscopy

  •  
  • ThT

    Thioflavin T

Author Contribution

R.D., A.K. and S.S. deigned the study and wrote the paper. R.D. performed the purification of the peptide and structural studies. P.M., D.M.S., S.H.K. and S.S. performed the cellular studies. J.D.A. performed the animal experiments. N.M. performed the AFM imaging. N.S.B. designed and constructed the vectors for expression. A.K. and S.S. supervised the study.

Funding

This work is supported by Grant-in-Aid [37(1509)/11/EMR-II] — CSIR-Government of India, Department of Biotechnology — DBT, Government of India [BT/RLF/Re-entry/11/2012], SERB Start-up Research grant [SB/YS/LS-23/2014] and SPPU Research and Development grant to the Department of Biotechnology.

Acknowledgments

The authors acknowledge HF-NMR facility and Bio-AFM funded by RIFC-IRCC, IIT Bombay; central instrumentation facility at Savitribai Phule Pune University (SPPU) and the flow cytometer facility at the Institute of Bioinformatics and Biotechnology, SPPU. R.D. is thankful for the financial assistance from UGC-JRF, Government of India, and A.K. acknowledges the funding from CSIR, Government of India. S.S. and P.M. acknowledge the funding from Ramalingaswami fellowship (Department of Biotechnology — DBT, Government of India) and SPPU Research and Development grant to the Department of Biotechnology. D.M.S. is thankful for the financial assistance from DBT-JRF programme and S.H.K. acknowledges DBT, GOI for her Masters in Biotechnology fellowship. J.D.A. acknowledges the funding from Start-up research grant by Science and Engineering Research board (SERB), Government of India. N.S.B. acknowledges the funding from ICGEB core funds. INS-1E cells were obtained as a kind gift from Prof. Claes Wollheim and Prof. Pierre Maechler, University Medical Centre, Geneva, Switzerland.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript. A patent has been filed for the protocol described in the paper for the production of hIAPP fibrils using recombinant DNA technology.

References

References
1
Whiting
,
D.R.
,
Guariguata
,
L.
,
Weil
,
C.
and
Shaw
,
J.
(
2011
)
IDF diabetes atlas: global estimates of the prevalence of diabetes for 2011 and 2030
.
Diabetes Res. Clin. Pract.
94
,
311
321
2
Chiti
,
F.
and
Dobson
,
C.M.
(
2006
)
Protein misfolding, functional amyloid, and human disease
.
Annu. Rev. Biochem.
75
,
333
366
3
Sánchez-Taínta
,
A.
,
Estruch
,
R.
,
Bulló
,
M.
,
Corella
,
D.
,
Gómez-Gracia
,
E.
,
Fiol
,
M.
et al. 
(
2008
)
Adherence to a Mediterranean-type diet and reduced prevalence of clustered cardiovascular risk factors in a cohort of 3204 high-risk patients
.
Eur. J. Cardiovasc. Prev. Rehabil.
15
,
589
593
4
Jaikaran
,
E.T.A.S.
and
Clark
,
A.
(
2001
)
Islet amyloid and type 2 diabetes: from molecular misfolding to islet pathophysiology
.
Biochim. Biophys. Acta, Mol. Basis Dis.
1537
,
179
203
5
Abedini
,
A.
and
Schmidt
,
A.M.
(
2013
)
Mechanisms of islet amyloidosis toxicity in type 2 diabetes
.
FEBS Lett.
587
,
1119
1127
6
Cooper
,
G.J.
,
Willis
,
A.C.
,
Clark
,
A.
,
Turner
,
R.C.
,
Sim
,
R.B.
and
Reid
,
K.B.
(
1987
)
Purification and characterization of a peptide from amyloid-rich pancreases of type 2 diabetic patients
.
Proc. Natl Acad. Sci. U.S.A.
84
,
8628
8632
7
Stridsberg
,
M.
and
Wilander
,
E.
(
1991
)
Islet amyloid polypeptide (IAPP) a short review
.
Acta Oncol.
30
,
451
456
8
Westermark
,
P.
,
Andersson
,
A.
and
Westermark
,
G.T.
(
2011
)
Islet amyloid polypeptide, islet amyloid, and diabetes mellitus
.
Physiol. Rev.
91
,
795
826
9
Gurlo
,
T.
,
Ryazantsev
,
S.
,
Huang
,
C.-j.
,
Yeh
,
M.W.
,
Reber
,
H.A.
,
Hines
,
O.J.
et al. 
(
2010
)
Evidence for proteotoxicity in β cells in type 2 diabetes: toxic islet amyloid polypeptide oligomers form intracellularly in the secretory pathway
.
Am. J. Pathol.
176
,
861
869
10
Haataja
,
L.
,
Gurlo
,
T.
,
Huang
,
C.J.
and
Butler
,
P.C.
(
2008
)
Islet amyloid in type 2 diabetes, and the toxic oligomer hypothesis
.
Endocr. Rev.
29
,
303
316
11
Shim
,
S.-H.
,
Gupta
,
R.
,
Ling
,
Y.L.
,
Strasfeld
,
D.B.
,
Raleigh
,
D.P.
and
Zanni
,
M.T.
(
2009
)
Two-dimensional IR spectroscopy and isotope labeling defines the pathway of amyloid formation with residue-specific resolution
.
Proc. Natl Acad. Sci. U.S.A.
106
,
6614
6619
12
Wei
,
L.
,
Jiang
,
P.
,
Xu
,
W.
,
Li
,
H.
,
Zhang
,
H.
,
Yan
,
L.
et al. 
(
2011
)
The molecular basis of distinct aggregation pathways of islet amyloid polypeptide
.
J. Biol. Chem.
286
,
6291
6300
13
Paulsson
,
J.F.
,
Andersson
,
A.
,
Westermark
,
P.
and
Westermark
,
G.T.
(
2006
)
Intracellular amyloid-like deposits contain unprocessed pro-islet amyloid polypeptide (proIAPP) in beta cells of transgenic mice overexpressing the gene for human IAPP and transplanted human islets
.
Diabetologia
49
,
1237
1246
14
Kayed
,
R.
,
Bernhagen
,
J.
,
Greenfield
,
N.
,
Sweimeh
,
K.
,
Brunner
,
H.
,
Voelter
,
W.
et al. 
(
1999
)
Conformational transitions of islet amyloid polypeptide (IAPP) in amyloid formation in vitro
.
J. Mol. Biol.
287
,
781
796
15
Cao
,
P.
,
Marek
,
P.
,
Noor
,
H.
,
Patsalo
,
V.
,
Tu
,
L.-H.
,
Wang
,
H.
et al. 
(
2013
)
Islet amyloid: from fundamental biophysics to mechanisms of cytotoxicity
.
FEBS Lett.
587
,
1106
1118
16
Andreev
,
Y.A.
,
Kozlov
,
S.A.
,
Vassilevski
,
A.A.
and
Grishin
,
E.V.
(
2010
)
Cyanogen bromide cleavage of proteins in salt and buffer solutions
.
Anal. Biochem.
407
,
144
146
17
Biancalana
,
M.
and
Koide
,
S.
(
2010
)
Molecular mechanism of Thioflavin-T binding to amyloid fibrils
.
Biochim. Biophys. Acta, Proteins Proteomics
1804
,
1405
1412
18
Willander
,
H.
,
Presto
,
J.
,
Askarieh
,
G.
,
Biverstål
,
H.
,
Frohm
,
B.
,
Knight
,
S.D.
et al. 
(
2012
)
BRICHOS domains efficiently delay fibrillation of amyloid β-peptide
.
J. Biol. Chem.
287
,
31608
31617
19
Gasymov
,
O.K.
and
Glasgow
,
B.J.
(
2007
)
ANS fluorescence: potential to augment the identification of the external binding sites of proteins
.
Biochim. Biophys. Acta, Proteins Proteomics
1774
,
403
411
20
Barth
,
A.
(
2007
)
Infrared spectroscopy of proteins
.
Biochim. Biophys. Acta, Bioenerg.
1767
,
1073
1101
21
Wishart
,
D.S.
,
Bigam
,
C.G.
,
Yao
,
J.
,
Abildgaard
,
F.
,
Dyson
,
H.J.
,
Oldfield
,
E.
et al. 
(
1995
)
1H, 13C and 15N chemical shift referencing in biomolecular NMR
.
J. Biomol. NMR.
6
,
135
140
22
Merglen
,
A.
,
Theander
,
S.
,
Rubi
,
B.
,
Chaffard
,
G.
,
Wollheim
,
C.B.
and
Maechler
,
P.
(
2004
)
Glucose sensitivity and metabolism-secretion coupling studied during two-year continuous culture in INS-1E insulinoma cells
.
Endocrinology
145
,
667
678
23
Taniguchi
,
T.
,
Garcia-Higuera
,
I.
,
Xu
,
B.
,
Andreassen
,
P.R.
,
Gregory
,
R.C.
,
Kim
,
S.-T.
et al. 
(
2002
)
Convergence of the fanconi anemia and ataxia telangiectasia signaling pathways
.
Cell
109
,
459
472
24
Chen
,
L.
,
Yang
,
X.
,
Jiao
,
H.
and
Zhao
,
B.
(
2002
)
Tea catechins protect against lead-induced cytotoxicity, lipid peroxidation, and membrane fluidity in HepG2 cells
.
Toxicol. Sci.
69
,
149
156
25
Zhou
,
P.
and
Wagner
,
G.
(
2010
)
Overcoming the solubility limit with solubility-enhancement tags: successful applications in biomolecular NMR studies
.
J. Biomol. NMR
46
,
23
31
26
Krampert
,
M.
,
Bernhagen
,
J.
,
Schmucker
,
J.
,
Horn
,
A.
,
Schmauder
,
A.
,
Brunner
,
H.
et al. 
(
2000
)
Amyloidogenicity of recombinant human pro-islet amyloid polypeptide (ProIAPP)
.
Chem. Biol.
7
,
855
871
27
Semisotnov
,
G.V.
,
Rodionova
,
N.A.
,
Razgulyaev
,
O.I.
,
Uversky
,
V.N.
,
Gripas
,
A.F.
and
Gilmanshin
,
R.I.
(
1991
)
Study of the ‘molten globule’ intermediate state in protein folding by a hydrophobic fluorescent probe
.
Biopolymers
31
,
119
128
28
Wiltzius
,
J.J.W.
,
Sievers
,
S.A.
,
Sawaya
,
M.R.
,
Cascio
,
D.
,
Popov
,
D.
,
Riekel
,
C.
et al. 
(
2008
)
Atomic structure of the cross-β spine of islet amyloid polypeptide (amylin)
.
Protein Sci.
17
,
1467
1474
29
Renault
,
M.
,
Cukkemane
,
A.
and
Baldus
,
M.
(
2010
)
Solid-state NMR spectroscopy on complex biomolecules
.
Angew. Chem. Int. Ed.
49
,
8346
8357
30
Heise
,
H.
,
Hoyer
,
W.
,
Becker
,
S.
,
Andronesi
,
O.C.
,
Riedel
,
D.
and
Baldus
,
M.
(
2005
)
Molecular-level secondary structure, polymorphism, and dynamics of full-length α-synuclein fibrils studied by solid-state NMR
.
Proc. Natl Acad. Sci. U.S.A.
102
,
15871
15876
31
Janson
,
J.
,
Ashley
,
R.H.
,
Harrison
,
D.
,
McIntyre
,
S.
and
Butler
,
P.C.
(
1999
)
The mechanism of islet amyloid polypeptide toxicity is membrane disruption by intermediate-sized toxic amyloid particles
.
Diabetes
48
,
491
498
32
Lopes
,
D.H.J.
,
Colin
,
C.
,
Degaki
,
T.L.
,
de Sousa
,
A.C.V.
,
Vieira
,
M.N.N.
,
Sebollela
,
A.
et al. 
(
2004
)
Amyloidogenicity and cytotoxicity of recombinant mature human islet amyloid polypeptide (rhIAPP)
.
J. Biol. Chem.
279
,
42803
42810
33
Cheng
,
B.
,
Gong
,
H.
,
Li
,
X.
,
Sun
,
Y.
,
Chen
,
H.
,
Zhang
,
X.
et al. 
(
2013
)
Salvianolic acid B inhibits the amyloid formation of human islet amyloid polypeptide and protects pancreatic beta-cells against cytotoxicity
.
Proteins: Struct., Funct., Bioinf.
81
,
613
621
34
Ma
,
L.
,
Li
,
X.
,
Wang
,
Y.
,
Zheng
,
W.
and
Chen
,
T.
(
2014
)
Cu(II) inhibits hIAPP fibrillation and promotes hIAPP-induced beta cell apoptosis through induction of ROS-mediated mitochondrial dysfunction
.
J. Inorg. Biochem.
140
,
143
152
35
Wang
,
L.
,
Li
,
Y.
,
Lu
,
T.
and
Li
,
F.
(
2014
)
The effects of organic solvents on the membrane-induced fibrillation of human islet amyloid polypeptide and on the inhibition of the fibrillation
.
Biochim. Biophys. Acta, Biomembr.
1838
,
3162
3170
36
Tatarek-Nossol
,
M.
,
Yan
,
L.-M.
,
Schmauder
,
A.
,
Tenidis
,
K.
,
Westermark
,
G.
and
Kapurniotu
,
A.
(
2005
)
Inhibition of hIAPP amyloid-fibril formation and apoptotic cell death by a designed hIAPP amyloid-core-containing hexapeptide
.
Chem. Biol.
12
,
797
809
37
Mishra
,
A.
,
Misra
,
A.
,
Vaishnavi
,
T.S.
,
Thota
,
C.
,
Gupta
,
M.
,
Ramakumar
,
S.
et al. 
(
2013
)
Conformationally restricted short peptides inhibit human islet amyloid polypeptide (hIAPP) fibrillization
.
Chem. Commun.
49
,
2688
2690
38
Konarkowska
,
B.
,
Aitken
,
J.F.
,
Kistler
,
J.
,
Zhang
,
S.
and
Cooper
,
G.J.S.
(
2006
)
The aggregation potential of human amylin determines its cytotoxicity towards islet β-cells
.
FEBS J.
273
,
3614
3624
39
May
,
P.C.
,
Gitter
,
B.D.
,
Waters
,
D.C.
,
Simmons
,
L.K.
,
Becker
,
G.W.
,
Small
,
J.S.
et al. 
(
1992
)
β-amyloid peptide in vitro toxicity: lot-to-lot variability
.
Neurobiol. Aging
13
,
605
607
40
Soto
,
C.
,
Castaño
,
E.M.
,
Kumar
,
R.A.
,
Beavis
,
R.C.
and
Frangione
,
B.
(
1995
)
Fibrillogenesis of synthetic amyloid-β peptides is dependent on their initial secondary structure
.
Neurosci. Lett.
200
,
105
108
41
Camargo
,
D.C.R.
,
Tripsianes
,
K.
,
Kapp
,
T.G.
,
Mendes
,
J.
,
Schubert
,
J.
,
Cordes
,
B.
et al. 
(
2015
)
Cloning, expression and purification of the human islet amyloid polypeptide (hIAPP) from Escherichia coli
.
Protein Expr. Purif.
106
,
49
56
42
Kosicka
,
I.
,
Kristensen
,
T.
,
Bjerring
,
M.
,
Thomsen
,
K.
,
Scavenius
,
C.
,
Enghild
,
J.J.
et al. 
(
2014
)
Preparation of uniformly 13C, 15N-labeled recombinant human amylin for solid-state NMR investigation
.
Protein Expr. Purif.
99
,
119
130
43
Mirecka
,
E.A.
,
Gremer
,
L.
,
Schiefer
,
S.
,
Oesterhelt
,
F.
,
Stoldt
,
M.
,
Willbold
,
D.
et al. 
(
2014
)
Engineered aggregation inhibitor fusion for production of highly amyloidogenic human islet amyloid polypeptide
.
J. Biotechnol.
191
,
221
227
44
Weirich
,
F.
,
Gremer
,
L.
,
Mirecka
,
E.A.
,
Schiefer
,
S.
,
Hoyer
,
W.
and
Heise
,
H.
(
2016
)
Structural characterization of fibrils from recombinant human islet amyloid polypeptide by solid-state NMR: the central FGAILS segment is part of the β-sheet core
.
PLoS ONE
11
,
e0161243
45
Uversky
,
V.
(
2011
)
Protein Chaperones and Protection From Neurodegenerative Diseases
.
John Wiley & Sons
46
Zraika
,
S.
,
Hull
,
R.L.
,
Udayasankar
,
J.
,
Aston-Mourney
,
K.
,
Subramanian
,
S.L.
,
Kisilevsky
,
R.
et al. 
(
2009
)
Oxidative stress is induced by islet amyloid formation and time-dependently mediates amyloid-induced beta cell apoptosis
.
Diabetologia
52
,
626
635
47
Casas
,
S.
,
Gomis
,
R.
,
Gribble
,
F.M.
,
Altirriba
,
J.
,
Knuutila
,
S.
and
Novials
,
A.
(
2007
)
Impairment of the ubiquitin-proteasome pathway is a downstream endoplasmic reticulum stress response induced by extracellular human islet amyloid polypeptide and contributes to pancreatic β-cell apoptosis
.
Diabetes
56
,
2284
2294
48
Mirzabekov
,
T.A.
,
Lin
,
M.-c.
and
Kagan
,
B.L.
(
1996
)
Pore formation by the cytotoxic islet amyloid peptide amylin
.
J. Biol. Chem.
271
,
1988
1992
49
Trikha
,
S.
and
Jeremic
,
A.M.
(
2011
)
Clustering and internalization of toxic amylin oligomers in pancreatic cells require plasma membrane cholesterol
.
J. Biol. Chem.
286
,
36086
36097
50
Pilkington
,
E.H.
,
Gurzov
,
E.N.
,
Kakinen
,
A.
,
Litwak
,
S.A.
,
Stanley
,
W.J.
,
Davis
,
T.P.
et al. 
(
2016
)
Pancreatic β-cell membrane fluidity and toxicity induced by human islet amyloid polypeptide species
.
Sci. Rep.
6
,
21274
51
Aston-Mourney
,
K.
,
Hull
,
R.L.
,
Zraika
,
S.
,
Udayasankar
,
J.
,
Subramanian
,
S.L.
and
Kahn
,
S.E.
(
2011
)
Exendin-4 increases islet amyloid deposition but offsets the resultant beta cell toxicity in human islet amyloid polypeptide transgenic mouse islets
.
Diabetologia
54
,
1756
1765
52
Goldsbury
,
C.S.
,
Cooper
,
G.J.S.
,
Goldie
,
K.N.
,
Müller
,
S.A.
,
Saafi
,
E.L.
,
Gruijters
,
W.T.W.
et al. 
(
1997
)
Polymorphic fibrillar assembly of human amylin
.
J. Struct. Biol.
119
,
17
27
53
Caillon
,
L.
,
Hoffmann
,
A.R.F.
,
Botz
,
A.
and
Khemtemourian
,
L.
(
2016
)
Molecular structure, membrane interactions, and toxicity of the islet amyloid polypeptide in type 2 diabetes mellitus
.
J. Diabetes Res.
2016
,
5639875
54
Bedrood
,
S.
,
Li
,
Y.
,
Isas
,
J.M.
,
Hegde
,
B.G.
,
Baxa
,
U.
,
Haworth
,
I.S.
et al. 
(
2012
)
Fibril structure of human islet amyloid polypeptide
.
J. Biol. Chem.
287
,
5235
5241
55
Luca
,
S.
,
Yau
,
W.-M.
,
Leapman
,
R.
and
Tycko
,
R.
(
2007
)
Peptide conformation and supramolecular organization in amylin fibrils: constraints from solid-state NMR
.
Biochemistry
46
,
13505
13522
56
Kajava
,
A.V.
,
Aebi
,
U.
and
Steven
,
A.C.
(
2005
)
The parallel superpleated beta-structure as a model for amyloid fibrils of human amylin
.
J. Mol. Biol.
348
,
247
252