PNAS Peer Review Sign up for PNAS Online eTocs
Link: Info for AuthorsLink: Editorial BoardLink: AboutLink: SubscribeLink: AdvertiseLink: ContactLink: Sitemap Link: PNAS Home
Proceedings of the National Academy of Sciences
Link: Current Issue "" Link: Archives "" Link: Online Submission ""  Link: Advanced Search

Published online on February 7, 2006, 10.1073/pnas.0507471103
PNAS | February 14, 2006 | vol. 103 | no. 7 | 2046-2051


This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supporting Information
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yan, L.-M.
Right arrow Articles by Kapurniotu, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yan, L.-M.
Right arrow Articles by Kapurniotu, A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg  
What's this?

 Previous Article  | Table of Contents |  Next Article 

BIOLOGICAL SCIENCES / BIOCHEMISTRY
Design of a mimic of nonamyloidogenic and bioactive human islet amyloid polypeptide (IAPP) as nanomolar affinity inhibitor of IAPP cytotoxic fibrillogenesis

Li-Mei Yan, Marianna Tatarek-Nossol, Aleksandra Velkova, Athanasios Kazantzis, and Aphrodite Kapurniotu*

Laboratory of Bioorganic and Medicinal Chemistry, Institute of Biochemistry, University Hospital of the Rheinisch–Westfälische Technische Hochschule Aachen, Pauwelstrasse 30, D-52074 Aachen, Germany

Edited by William F. DeGrado, University of Pennsylvania School of Medicine, Philadelphia, PA, and approved December 15, 2005 (received for review August 26, 2005)


   Abstract
Top
 Abstract
Results
 Discussion
 Conclusion
 Materials and Methods
 Acknowledgements
 References
 
Protein aggregation into cytotoxic oligomers and fibrils in vivo is linked to cell degeneration and the pathogenesis of >25 uncurable diseases, whereas the high aggregation propensity and insolubility of several bioactive polypeptides and proteins in vitro prevent their therapeutic use. Aggregation of human islet amyloid polypeptide (IAPP) into pancreatic amyloid is strongly associated with the pathogenesis of type II diabetes. IAPP is a 37-residue polypeptide that acts as a neuroendocrine regulator of glucose homeostasis. However, IAPP misfolds and self-associates into cytotoxic aggregates and fibrils even at nanomolar concentrations. Because IAPP aggregation causes beta-cell death and prohibits therapeutic application of IAPP in diabetes, we pursued a minimalistic chemical design approach to generate a molecular mimic of a nonamyloidogenic and bioactive IAPP conformation that would still be able to associate with IAPP and thus inhibit its fibrillogenesis and cytotoxicity. We show that the double N-methylated full length IAPP analog [(N-Me)G24, (N-Me)I26]-IAPP (IAPP-GI) is a highly soluble, nonamyloidogenic, and noncytotoxic IAPP molecular mimic and an IAPP receptor agonist. Moreover, IAPP-GI binds IAPP with low nanomolar affinity and completely blocks IAPP cytotoxic self-assembly and fibrillogenesis with activity in the low nanomolar concentration range. Importantly, IAPP-GI dissociates cytotoxic IAPP oligomers and fibrils and is able to reverse their cytotoxicity. Bifunctional soluble IAPP mimics that combine bioactivity with the ability to block and reverse IAPP cytotoxic self-assembly are promising candidates for the treatment of diabetes. Moreover, our amyloid disease inhibitor design concept may be applicable to other protein aggregation diseases.

amyloidogenesis inhibitor | chemical design | protein aggregation | diabetes | therapeutic compound

Folding, misfolding, and the aggregation and fibrillogenesis propensities of proteins or polypeptides appear to be intrinsic properties of the amino acid sequence (1). An increasing body of evidence suggests that aggregates that form during in vivo self-assembly and amyloidogenesis are cytotoxic. This offers a link between protein/polypeptide aggregation, cell degeneration, and the pathogenesis of a number of diseases called protein aggregation diseases (1, 2). In addition, the high aggregation and fibrillogenesis propensities of several bioactive polypeptides and proteins in vitro cause a major problem in their pharmaceutical use as therapeutics (3).

Compounds that block cytotoxic protein/polypeptide self-assembly and amyloidogenesis are therefore important targets of therapeutic intervention in disease (2, 4). However, the design of such compounds is a very demanding task (2, 4, 5). Short self-recognition regions of amyloidogenic polypeptides have been applied as scaffolds for low-molecular-weight peptidomimetics to disrupt amyloidogenesis (6–9). Although a number of such compounds have been generated, highly potent and pharmacologically tested agents are not yet available (2, 4–6). A potential hurdle in their development is to overcome the high affinity and cooperativity of the interactions of cytotoxic self-assembly of the full length amyloidogenic sequences (2, 4, 5).

The extremely high propensity of the pancreatic polypeptide human islet amyloid polypeptide (IAPP) to misfold into cytotoxic aggregates and fibrils is strongly associated with beta-cell degeneration in type II diabetes (10, 11). IAPP is a 37-residue polypeptide secreted by the beta-cells that acts together with insulin as a regulator of glucose homeostasis (12, 13). Clinical studies suggest that the use of IAPP or IAPP receptor agonists as glucose regulators might be beneficial in diabetes treatment (13). However, low solubility (11, 13), high fibrillogenesis propensity, and the cytotoxicity of misfolded and aggregated IAPP preclude its pharmacological use (12, 13). IAPP is a conformationally flexible polypeptide that misfolds into cytotoxic beta-sheet aggregates and fibrils via a multistep nucleation-dependent process (11, 14, 15). Soluble IAPP analogs that combine bioactivity with the ability to inhibit cytotoxic self-assembly of native IAPP are therefore of high biomedical interest. Such "bifunctional" analogs would be promising candidates for therapeutic application in diabetes (13) and tools for understanding IAPP amyloidogenesis and cytotoxicity.

Here, we use a minimalistic conformational restriction strategy to design a full length IAPP-derived molecular mimic of a nonamyloidogenic and bioactive IAPP conformation. According to our design concept, the mimic should have nearly the same sequence as IAPP but should be soluble, nonamyloidogenic, and noncytotoxic and would ideally exhibit agonistic activity. At the same time, the mimic should be able to bind IAPP and inhibit its cytotoxic amyloidogenesis process. Approaches to changing the biophysical properties of self-assembling proteins or polypeptides include redesigning self-recognition interfaces (16, 17). Our approach features the structure-based introduction of a minimum number of two N-methyl rests on the same side of the beta-strand in the IAPP "amyloid core" sequence IAPP [22–27] or NFGAIL of full length IAPP (Fig. 1A) (9, 18). Constraining the conformation of a conformationally flexible and bioactive polypeptide is a known strategy to both modulate its biophysical properties and generate high-affinity agonists of biological function (17, 19). Further, amide bond N-methylation is a minimally invasive method toward inhibiting peptide self-assembly in beta-sheets (17, 20).


Figure 1
View larger version (28K):
[in this window]
[in a new window]
  		Fig. 1. Primary structures of IAPP and the molecular mimic (both peptides are C-terminal peptide amides) (A) and expected mechanism of inhibition of IAPP fibrillogenesis and cytotoxicity by the mimic (B).

We envisioned that, because of its similarity to IAPP, the mimic should be able to interact with monomeric IAPP and kinetically stabilize it against cytotoxic misfolding and self-assembly (Fig. 1B). Binding to IAPP monomers might also shift the fibrillogenesis equilibrium toward dissociation of cytotoxic assemblies. Similarly to IAPP, the mimic should also be able to bind fibrillogenesis intermediates (Fig. 1B). A high affinity of biomolecular recognition, a strong kinetic stabilization of nontoxic IAPP monomers (or oligomers) against cytotoxic misfolding and self-assembly, and a strong inhibitory potency toward IAPP self-assembly should be provided by the maximized number of contact points between the full length IAPP-derived molecular mimic and IAPP, because of positive cooperativity, and by the constrained nature and the steric bulk of the molecular mimic (5, 16, 21).


   Results
Top
 Abstract
 Results
Discussion
 Conclusion
 Materials and Methods
 Acknowledgements
 References
 
Design of the Molecular Mimic. We have previously shown that N-methylation of the amide bonds at G24 and I26 in the amyloidogenic and cytotoxic hexapeptide NFGAIL generated the nonamyloidogenic and noncytotoxic peptide NF(N-Me)GA(N-Me)IL (18). Based on the hypothesis that NFGAIL is a crucial amyloid core and self-recognition sequence of IAPP (22), we decided to introduce the two N-methyl residues into the amides of G24 and I26 within the full length IAPP sequence and designed the double N-methylated IAPP analog [(N-Me)G24, N-Me)I26]-IAPP (IAPP-GI) (Fig. 1A). However, although the two N-methylations abolished amyloidogenicity of short IAPP fragments (18), their effect on full length IAPP was unpredictable because of the unknown role of other sequence parts in misfolding and self-assembly (11).

IAPP-GI Is a Highly Soluble Nonamyloidogenic Noncytotoxic and Yet Assembly-Competent IAPP Mimic. Sedimentation and transmission EM (TEM) studies at pH 7.4 confirmed that IAPP has a very low kinetic solubility and an extremely high amyloidogenic propensity (11) (Fig. 2A and B). IAPP aggregated into fibrils within 24 h at 1 µM, whereas at 10 and 100 µM, aggregates precipitated immediately, and fibrillogenesis was accomplished at 20 and 2 h, respectively (Fig. 2 A and B). By contrast, IAPP-GI was soluble even at a 100-fold higher concentration than IAPP (100 µM), and TEM indicated the absence of fibrillar aggregates (Fig. 2 A and B). To quantify amyloidogenicity, we incubated IAPP or IAPP-GI (pH 7.4) at various concentrations and followed fibril formation via the thioflavin T (ThT)-binding assay (23). IAPP aggregated into fibrils at submicromolar concentrations (625 nM) (Fig. 2C). By contrast, IAPP-GI did not bind ThT even when a 100-fold higher concentration than IAPP was applied (62.5 µM; 14 days) (Fig. 2C). Thus, IAPP-GI was at least 100-fold more soluble and 100-fold less amyloidogenic than IAPP at physiological pH.


Figure 2
View larger version (40K):
[in this window]
[in a new window]
  		Fig. 2. Solubility and amyloidogenicity of IAPP-GI and IAPP. (A) Solubilities of IAPP and IAPP-GI (1–100 µM) in aqueous buffer, pH 7.4, as assessed by a sedimentation assay. Results are means (±SEM) from three to five experiments. (B) TEM of aged incubations (4 days) of 1 µM IAPP and 100 µM IAPP-GI. (Scale bars, 100 nm.) (C) Fibrillogenesis of IAPP and IAPP-GI (625 nM–62.5 µM) as assessed by the ThT assay. Results are means (±SEM) from three experiments. (D) CD spectra of incubations of IAPP and IAPP-GI (5 µM) at various time points. (Insets) TEM at 24 h (IAPP) and 14 days (IAPP-GI). (Scale bars, 100 nm.)

We next followed the conformation of IAPP-GI and IAPP at pH 7.4 by far-UV CD for several days. Although IAPP (5 µM) aggregated into soluble beta-sheets and insoluble fibrils within 2.5 h, IAPP-GI (5 µM) was soluble, and its spectrum was nearly unchanged for 14 days (Fig. 2D). The spectrum of IAPP-GI was indicative of ordered, most likely beta-sheet- and/or beta-turn-containing, conformations, and TEM indicated a complete absence of fibrils (Fig. 2D Insets). These findings indicated that IAPP-GI was constrained in a nonamyloidogenic conformation.

IAPP-GI did have a pronounced self-association propensity. This was shown by TEM, size exclusion chromatography (SEC), and far-UV CD concentration dependence studies (not shown). TEM of IAPP-GI solutions (5–100 µM) revealed soluble spheroidal aggregates of different sizes (diameters up to 100 nm) (Fig. 2 B and D). SEC indicated that both IAPP-GI and IAPP formed monomers and dimers, when freshly dissolved (6.25 µM) in aqueous buffer, pH 7.4 (Fig. 7 A and B, which is published as supporting information on the PNAS web site). Kinetic follow-up of peptide association indicated that IAPP further associated into tetramers, various different oligomers, and large insoluble aggregates (Fig. 7B). IAPP-GI also formed soluble oligomers but no insoluble aggregates (Fig. 7A).

To quantify affinities of self-association, we used fluorescence spectroscopy (24). N{alpha}-amino-terminal fluorescein-labeled IAPP-GI (Fluos-IAPP-GI) and IAPP (Fluos-IAPP) were titrated at 1 nM with IAPP-GI and IAPP, respectively, in aqueous buffer, pH 7.4, and fluorescence emission spectra were recorded (9). Binding of IAPP to Fluos-IAPP and of IAPP-GI to Fluos-IAPP-GI at a 100-fold molar excess resulted in fluorescence enhancement of 260% and 190%, respectively (Fig. 3A and B). Sigmoidal titration curves were obtained and were consistent with cooperative self-assembly processes (Fig. 3 A and B). The determined apparent affinities (app. Kd) of interactions were 9.7 (±0.9) nM for Fluos-IAPP-IAPP and 4.2 (±0.8) nM for Fluos-IAPP-GI-IAPP-GI, demonstrating that both peptides self-associated with strong and nearly identical affinities.


Figure 3
View larger version (39K):
[in this window]
[in a new window]
  		Fig. 3. Self-assembly and effects on beta-cell viability of IAPP-GI and IAPP. (A and B) Self-association of IAPP (A) and IAPP-GI (B) studied by fluorescence spectroscopy: (A) Spectra of Fluos-IAPP (1 nM) alone and with IAPP (Left) and the fluorescence of Fluos-IAPP at 522 nm after titration with IAPP (Right). (B) Spectra of Fluos-IAPP-GI (1 nM) alone and with IAPP-GI (Left) and the fluorescence of Fluos-IAPP-GI at 522 nm after titration with IAPP-GI (Right). Fluos-IAPP/ligand or Fluos-IAPP-GI/ligand molar ratios are shown. Data are means (±SEM) from three binding curves. (C and D) Effects of IAPP and IAPP-GI on beta-cell viability. (C) MTT reduction assay. Data are means ± SEM of three to seven assays (n = 4 each). (D) Apoptosis assay. Data are means (±SEM) of six determinations (***, P < 0.001 by ANOVA).

We next studied the potential cytotoxic properties of IAPP-GI as compared with IAPP. Peptides were applied to cultured pancreatic rat insulinoma cells (RIN 5fm), and cell viability was determined by two different assays: first, by the 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) reduction assay (Fig. 3C) and secondly by assessment of apoptosis (Fig. 3D) (11). Both assays showed that, whereas IAPP was strongly cytotoxic and caused apoptotic cell death in beta-cells, IAPP-GI was completely devoid of any cytotoxicity (Fig. 3 C and D).

IAPP-GI Is a Full IAPP Receptor Agonist. IAPP exhibits several of its biological actions via binding to G protein-coupled receptors and activation of adenylate cyclase (12, 25, 26). To evaluate the receptor agonistic potency of IAPP-GI, we studied adenylate cyclase activation of high-affinity human IAPP receptors expressed on the human breast carcinoma cell line MCF-7 (25). Various concentrations of IAPP-GI or IAPP were incubated with the cells and generated cAMP was quantified. IAPP-GI exhibited the same maximum effect on adenylate cyclase stimulation as IAPP, which was consistent with IAPP-GI being a full agonist of the human IAPP receptor (Fig. 4A). However, the adenylate cyclase activation isotherms showed that IAPP-GI was a 7-fold less potent agonist than IAPP (IAPP, EC50 631 pM; IAPP-GI, EC50 4.4 nM). To determine the receptor-binding affinity of IAPP-GI, we next studied competitive inhibition of the specific receptor binding of radioactively labeled rat IAPP, the strongest ligand known, by IAPP-GI and IAPP on MCF-7 cells (Fig. 4B) (25). The binding isotherms showed that, whereas IAPP-GI was a potent IAPP receptor ligand, its receptor-binding affinity was significantly decreased as compared with IAPP (IAPP, IC50 2.3 nM; IAPP-GI, IC50 158.5 nM). Together, the above results showed that IAPP-GI is a full receptor agonist, albeit weaker than IAPP.


Figure 4
View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4. Human IAPP receptor activation and binding potentials of IAPP-GI and IAPP in MCF-7 cells. (A) Stimulation of adenylate cyclase after incubation with IAPP or IAPP-GI. Adenylate cyclase activity above basal levels is plotted as a function of peptide concentration. Results are means (±SEM) of three assays (n = 2 each). (B) Receptor-binding affinities as assessed by competitive inhibition of the specific receptor binding of 125I-rat IAPP by IAPP and IAPP-GI. Specific radioligand binding is plotted as a function of IAPP or IAPP-GI concentration. Data are means (±SEM) of two to three assays (n = 2 each).

IAPP-GI Is a Nanomolar Affinity IAPP Ligand. To address the question whether IAPP-GI is able to bind IAPP, we performed a pull-down assay in combination with NuPAGE gel electrophoresis and Western blotting (27) (Fig. 5A). Amino-terminal digoxigenin-labeled IAPP-GI (Dig-IAPP-GI) (2.5 µM) was incubated (1 h) with N{alpha}-amino-terminal biotin-labeled IAPP (Biotin-IAPP) (2.5 µM), and complexes were isolated by capturing Biotin-IAPP on streptavidin-coated beads. After dissociation, components were revealed by NuPAGE and Western blotting with antibiotin and antidigoxigenin antibodies (Fig. 5A). The antidigoxigenin Western blot showed the presence of significant amounts of Dig-IAPP-GI, demonstrating that IAPP-GI binds IAPP.


Figure 5
View larger version (46K):
[in this window]
[in a new window]
  		Fig. 5. Binding of IAPP-GI to IAPP and inhibition of IAPP misfolding and fibrillogenesis. (A) Binding of Dig-IAPP-GI to Biotin-IAPP assessed by a biotin pull-down assay. (Upper) Antidigoxigenin Western blot analysis of a mixture of Dig-IAPP-GI with Biotin-IAPP (2.5 µM) and suitable controls after pull-down and peptide dissociation. Lane Dig-IAPP-GI input: input (100%) [freshly dissolved (4 µg) not incubated with beads]; lane Biotin-IAPP: Biotin-IAPP alone; lane Dig-IAPP-GI: Dig-IAPP-GI alone [nonspecific binding (NSB) of Dig-IAPP-GI for lane "mixture"]; lane Mixture, mixture of Dig-IAPP-GI and Biotin-IAPP-GI; lane Biotin-IAPP input: Biotin-IAPP input (100%, 4 µg). (Lower) Antibiotin Western blot analysis of the same mixture as in Upper. Lanes are as in Upper. The band densities of the antidigoxigenin western are also shown. NSB of Dig-IAPP-GI was 4% of input. (B) Binding of IAPP to IAPP-GI as assessed by fluorescence spectroscopy: Spectra of Fluos-IAPP (1 nM) alone and after titration with IAPP-GI (Fluos-IAPP/ligand molar ratios as indicated) are shown (Left). (Right) Fluorescence at 522 nm of Fluos-IAPP after titration with IAPP-GI. Data are means (±SEM) from three binding curves. (C) Interaction of IAPP with IAPP-GI as followed by CD and TEM: The conformation of a mixture of IAPP and IAPP-GI (5 µM each) was followed by CD over 14 days. CD spectra at various time points up to 14 days are shown. (Inset) TEM at 14 days (Scale bar, 100 nm). In parallel, CD spectra of IAPP and IAPP-GI incubations (5 µM) were measured (Fig. 2D). These spectra (0 h) and their sum are also shown. (D) TEM of IAPP (Left) and the mixture of IAPP with IAPP-GI (1/1) (at 24 h) (incubation as in C) (Right).

The interaction of IAPP-GI with IAPP was then characterized by fluorescence spectroscopy. Fluos-IAPP (1 nM) was titrated in aqueous buffer, pH 7.4, with IAPP-GI. Binding of a 100-fold molar excess of IAPP-GI resulted in a 240% increase of fluorescence emission (Fig. 5B). A sigmoidal binding curve was obtained and an app. Kd of 17.5 nM (±3.2) was determined (Fig. 5B). Of note, a similar affinity was obtained when Fluos-IAPP-GI (1 nM) was titrated with IAPP [app. Kd = 8.9 nM (±1.5)] (not shown). A directly proportional concentration dependence was found for the fluorescence of Fluos-IAPP within 1–4 nM (9). This finding and the 10 nM affinity of Fluos-IAPP - IAPP assemblies were consistent with Fluos-IAPP being mainly monomeric at 1 nM. We also titrated a 10 nM Fluos-IAPP solution, where both Fluos-IAPP monomers and oligomers were present, with IAPP-GI or IAPP (not shown). The app. Kd were in the same concentration range as above. These findings suggested that IAPP-GI bound IAPP monomers and oligomers with very low nanomolar affinity and showed that the affinity of the IAPP-GI–IAPP interaction is very similar to the affinity of IAPP self-association.

IAPP-GI Inhibits IAPP Misfolding into beta-Sheets, Fibrillogenesis, and Cytotoxicity. To examine whether interaction of IAPP-GI with IAPP could inhibit IAPP misfolding and self-assembly into soluble beta-sheets and fibrils, we next used far-UV CD spectroscopy and TEM. IAPP (5 µM, pH 7.4) was incubated alone or in the presence of IAPP-GI (5 µM), and CD spectra were measured at various time points (Figs. 2D and 5C). Aliquots were subjected to TEM, whereas UV spectra were measured to control for peptide loss because of insolubilization. IAPP aggregated into soluble beta-sheets and insoluble fibrillar aggregates precipitated at 2.5 h (Figs. 2D and 5D Left). In the presence of IAPP-GI, however, no insolubilization of IAPP occurred, and the spectrum did not change over 14 days (Fig. 5C). The spectrum of the mixture differed markedly from the sum of the spectra of IAPP (at 0 h) and IAPP-GI, which indicated that IAPP-GI had interacted with IAPP. TEM between 24 h and 14 days showed soluble round oligomers and chain-like assemblies of spherical oligomers (Fig. 5D Right), but no fibrillar aggregates were detected (Fig. 5C Inset).

IAPP fibrillization in the absence or presence of IAPP-GI was next followed by the ThT assay (Fig. 6A). Conversion of IAPP (6.25 µM, pH 7.4) into fibrils was accomplished within 2–4 h. In the presence of IAPP-GI (1/1), however, fibril formation was completely suppressed (14 days) (Fig. 6A).


Figure 6
View larger version (30K):
[in this window]
[in a new window]
  		Fig. 6. Inhibition of IAPP fibrillogenesis and cytotoxicity, dissolution of preformed IAPP fibrils, and reversal of cytotoxic IAPP self-assembly and fibrillogenesis by IAPP-GI. (A) Fibrillogenesis of IAPP (6.25 µM), a mixture of IAPP and IAPP-GI (6.25 µM each), and IAPP-GI (6.25 µM) alone as assessed by ThT binding. (B) Effects of IAPP (16.5 µM) versus a mixture of IAPP and IAPP-GI (1/1) on beta-cell viability as assessed by MTT reduction. Data are means (±SEM) of three assays (n = 3 each). (C) Inhibition of IAPP (500 nM) mediated beta-cell apoptosis by IAPP-GI and determination of the IC50 of the inhibitory effect of IAPP-GI on IAPP cytotoxicity (Inset). Data are means (±SEM) of six determinations (***, P < 0.001 by ANOVA). (Inset) Titration of cytotoxic effect of IAPP (100 nM) with IAPP-GI using the MTT assay. Data are means (±SEM) of eight determinations. (D) Dissociation of IAPP fibrils by IAPP-GI as followed by ThT binding. IAPP (16.5 µM) was allowed to aggregate. At the indicated time points, aliquots were mixed with IAPP-GI (1/1), and fibrillogenesis of IAPP alone versus the mixtures was quantified. Data are means (±SEM) of three assays. (E) Reversal of cytotoxicity of IAPP oligomers and fibrils by IAPP-GI. Solutions of D were added to RIN 5fm cells (at 24 h). Cell viability was assessed by the MTT assay. Data are means (±SEM) of three assays (n = 3 each). (F) Inhibition of the nucleation effect of IAPP fibrillar seeds on IAPP fibrillogenesis by IAPP-GI as followed by the ThT assay. IAPP (6.25 µM) with or without IAPP-GI (1/1) was seeded with IAPP fibrils. Results are means (±SEM) of three to four assays.

Next, IAPP or 1/1 mixtures of IAPP with IAPP-GI were added to RIN 5fm cells, and cytotoxicities were assessed by MTT reduction (Fig. 6B) and by quantification of apoptotic cell death (Fig. 6C). IAPP-GI strongly suppressed IAPP-mediated apoptosis, and IAPP cytotoxicity was nearly completely inhibited over a 1,000-fold IAPP concentration range. Of note, rat IAPP, which is nonamyloidogenic and differs from IAPP in six residues, did not affect IAPP cytotoxicity (not shown). Titration of cytotoxic IAPP (100 nM) with IAPP-GI gave a sigmoidal titration curve, and a 1/1 molar ratio of IAPP-GI to IAPP caused maximal inhibition of cytotoxicity (Fig. 6C Inset). An IC50 of 20 nM was determined, which correlated very well with the 17 nM nanomolar affinity of the IAPP-GI–IAPP interaction.

IAPP-GI Redissolves Cytotoxic IAPP Aggregates and Fibrils and Reverses Their Cytotoxic Effects. We next asked whether IAPP-GI could block or reverse already-started IAPP fibrillogenesis processes. Using the ThT assay, we compared fibrillization kinetics of a solution of IAPP alone (16.5 µM, pH 7.4) with those of an identical solution to which IAPP-GI (16.5 µM) was added at various time points (Fig. 6D). IAPP-GI completely interrupted IAPP fibrillization (at 1/1) regardless of the stage of the process, i.e., when added both before and after nucleation. Most importantly, IAPP-GI (1/1) redissolved IAPP fibrils and reversed already-started fibrillogenesis (Fig. 6D).

To address whether IAPP-GI-mediated dissolution of fibrils resulted in reversal of cytotoxicity, solutions of the ThT assays were applied to RIN 5fm cells (at 24 h), and cell viability was assessed (Fig. 6E). In fact, IAPP-GI was able to nearly quantitatively reverse the cytotoxic effects of already-formed IAPP aggregates and fibrils (Fig. 6E). Of note, the same cytotoxic effect of the IAPP-alone incubations was obtained when IAPP was incubated for 4–24 h before addition to the cells (Fig. 6F), whereas cytotoxicity was somewhat reduced in the 0- or 30-min incubations (not shown).

We also tested whether IAPP-GI was able to inhibit the nucleating effect of exogenously added fibrillar seeds on IAPP fibrillogenesis. Under assay conditions where IAPP fibrillogenesis (6.25 µM) began after a lag of 48 h (9), seeding with preformed IAPP fibrils (10%) resulted in an immediate beginning of fibrillogenesis (Fig. 6F). In the presence of IAPP-GI (1/1), however, fibril formation was completely suppressed (Fig. 6F).


   Discussion
Top
 Abstract
 Results
 Discussion
Conclusion
 Materials and Methods
 Acknowledgements
 References
 
We have designed a molecular mimic of a nonamyloidogenic and bioactive IAPP conformation. The mimic, IAPP-GI, is a 37-residue polypeptide that exhibits nearly the same amino acid sequence as the extremely amyloidogenic and cytotoxic polypeptide IAPP. However, IAPP-GI is at least 100-fold more soluble and less amyloidogenic than IAPP under physiological pH and has no cytotoxic properties. Our results provide evidence that the gain-of-toxic activity of IAPP is directly linked to its ability to misfold and self-assemble in beta-sheet aggregates and fibrils (1, 2, 11).

Reported inhibitors of IAPP fibrillogenesis so far include aromatic compounds and short peptides active in the micromolar or submicromolar concentration range, at best (7–9, 28). These inhibitors can merely delay or suppress amyloidogenicity and cytotoxicity and are unable to block or reverse IAPP cytotoxic self-assembly after nucleation. For example, the hexapeptide NF(N-Me)GA(N-Me)IL, a recently described submicromolar affinity IAPP ligand, exhibits its strongest inhibitory effects in the micromolar concentration range, is unable to completely block IAPP fibrillogenesis and cytotoxicity, and does not block or reverse cytotoxic self-assembly after nucleation (9). By contrast, IAPP-GI binds IAPP with a {approx}10-fold higher affinity than NF(N-Me)GA(N-Me)IL, completely blocks IAPP cytotoxic misfolding and self-association with an IC50 of 20 nM both in pre- and postnucleation phase, and redissolves cytotoxic IAPP aggregates and fibrils. Because treatment of disease usually starts after the disease begins, these properties are very important with regard to a potential therapeutic use.

Our studies suggest that the high inhibitory potency of IAPP-GI might be mediated by the high affinity binding of IAPP-GI, (i) to IAPP monomers, their kinetic stabilization toward cytotoxic misfolding and self-assembly or their sequestering from the amyloid pathway and (ii) to IAPP fibrillogenesis intermediates and their sequestration from further fibrillogenesis. IAPP-GI binding to nontoxic IAPP monomers would shift self-assembly to the left and might underlie dissociation and reversal of cytotoxicity of oligomers and fibrils (2) (Fig. 1B).


   Conclusion
Top
 Abstract
 Results
 Discussion
 Conclusion
Materials and Methods
 Acknowledgements
 References
 
IAPP-GI is a designed low nanomolar affinity IAPP ligand. Furthermore, IAPP-GI is the only known compound that completely blocks with low nanomolar activity and reverses IAPP cytotoxic self-assembly and fibrillogenesis. Finally, IAPP-GI is the only known IAPP amyloidogenesis inhibitor that is at the same time a soluble at physiological pH, nonamyloidogenic and noncytotoxic, full length IAPP analog and an IAPP receptor agonist. For comparison, the best-known soluble IAPP receptor agonist, a triproline analog of full length IAPP ([P25,P28,P29]-IAPP), precipitates above pH 5.5, and no amyloid-inhibitory properties have been reported (13). In addition, because large parts of the IAPP sequence are required for receptor binding and activation, other reported inhibitors would be devoid of receptor agonistic properties (7–9, 13, 28). These points demonstrate the efficiency of our amyloid inhibitor design strategy and underline the so-far-unique bifunctional nature of IAPP-GI, i.e., its function as both a highly potent IAPP amyloid inhibitor and a soluble IAPP receptor agonist. Bifunctional IAPP mimics, such as IAPP-GI, could therefore be promising drug candidates for the treatment of diabetes (13).

Our results offer a proof of principle of a concept for designing potent amyloid disease therapeutics and of a chemical engineering approach to redesign a natively amyloidogenic and bioactive polypeptide sequence into a soluble, noncytotoxic, bioactive, and highly potent inhibitor of its own cytotoxic self-assembly. Therefore, our amyloid inhibitor design approach may be applicable to other disease-related self-associating polypeptides (3, 4, 29).


   Materials and Methods
Top
 Abstract
 Results
 Discussion
 Conclusion
 Materials and Methods
Acknowledgements
 References
 
Peptides. Synthetic IAPP (human and rat sequences) were obtained from Calbiochem, dissolved in 1,1,1,3,3,3-hexafluoro-2- propanol (HFIP) at 200–500 µM, and filtered (human sequence) over 0.2-µm filters (Millipore). Concentrations of stock solutions were determined by UV (9, 14). [125I]Bolton/Hunter-labeled rat IAPP for receptor-binding studies was from Amersham Pharmacia Biosciences (25).

Peptide Synthesis. Peptides, including the mimic and the labeled analogs, were synthesized by using previously published fluorenylmethoxycarbonyl solid-phase synthetic protocols (9, 30) (see Supporting Text, which is published as supporting information on the PNAS web site).

Sedimentation Assays. Assays were performed as described in Supporting Text.

TEM. Aliquots of incubations {results shown in Figs. 2 and 5 are from incubations in 10 mM phosphate buffer [1% 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP)], pH 7.4} were applied on carbon-coated grids, stained with uranyl acetate as described (14), and examined with a Philips (Eindhoven, The Netherlands) EM 400 T electron microscope at 60 kV.

ThT-Binding Assays. Assays were performed by using a recently published assay system (9) (see Supporting Text).

Far-UV CD Spectroscopy. CD spectra were measured as described in Supporting Text.

Size Exclusion Chromatography (SEC). SEC was performed with a TSK-GEL G2000SWXL (TosoHaas, Montgomeryville, PA) (300 x 7.8 mm), as described in Supporting Text.

Adenylate Cyclase Activation and Receptor-Binding Assays. Cell culture (MCF-7 cells) and assays were performed as described in Supporting Text.

Pull-Down Assays. Assays were performed as described in Supporting Text (27).

Fluorescence Spectroscopic Titration Studies. Measurements were performed with a Spex Fluorolog 2 fluorescence spectrophotometer, as described (9, 31) (see Supporting Text).

Assessment of Cytotoxicity via the MTT Reduction Assay. The rat insulinoma cell line RIN 5fm was used and cells were plated as described (9, 22) (see Supporting Text).

Assessment of Apoptosis via ELISA. RIN 5fm cells were plated at a density of 5 x 105 cells/ml, and the assay was performed as described (9) (see Supporting Text).


   Acknowledgements
Top
 Abstract
 Results
 Discussion
 Conclusion
 Materials and Methods
 Acknowledgements
References
 
We are grateful to J. Bernhagen for critical reading of the manuscript and helpful discussions, and to K. Alexandrov and R. S. Goody for help with the analysis of the fluorescence data. We thank K. Tenidis, M. Waldner, H. Vasen, and M. Müsken for technical assistance and M. Dewor, R. Wacker, and R. Fischer for MALDI. This work was partially supported by the Deutsche Forschungsgemeinschaft (DFG).


   Footnotes
 

Abbreviations: IAPP, human islet amyloid polypeptide; IAPP-GI, [(N-Me)G24, N-Me)I26]-IAPP; ThT, thioflavin T; MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; TEM, transmission EM; Fluos-IAPP, N{alpha}-amino-terminal fluorescein-labeled IAPP; Fluos-IAPP-GI, N{alpha}-amino-terminal fluorescein-labeled IAPP-GI; RIN 5fm, cultured pancreatic rat insulinoma cells; Dig-IAPP-GI, amino-terminal digoxigenin-labeled IAPP-GI; Biotin-IAPP, Na-amino-terminal biotin-labeled IAPP.

*To whom correspondence should be addressed. E-mail: akapurniotu{at}ukaachen.de

Author contributions: A. Kapurniotu designed research; L.-M.Y., M.T.-N., A.V., A. Kazantzis, and A. Kapurniotu performed research; L.-M.Y., M.T.-N., A.V., A. Kazantzis, and A. Kapurniotu analyzed data; and A. Kapurniotu wrote the paper.

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

© 2006 by The National Academy of Sciences of the USA


   References
Top
 Abstract
 Results
 Discussion
 Conclusion
 Materials and Methods
 Acknowledgements
 References
 

  1. Dobson, C. M. (2003) Nature 426, 884–890.[CrossRef][Medline]
  2. Soto, C. (2003) Nat. Rev. Neurosci 4, 49–60.[CrossRef][ISI][Medline]
  3. Zurdo, J. (2005) Protein Pept. Lett 12, 171–187.[CrossRef][ISI][Medline]
  4. Cohen, F. E. & Kelly, J. W. (2003) Nature 426, 905–909.[CrossRef][Medline]
  5. Kelly, J. W. (2005) N. Engl. J. Med 352, 722–723.[Free Full Text]
  6. Findeis, M. A. (2002) Curr. Top. Med. Chem 2, 417–423.[CrossRef][Medline]
  7. Scrocchi, L. A., Chen, Y., Waschuk, S., Wang, F., Cheung, S., Darabie, A. A., McLaurin, J. & Fraser, P. E. (2002) J. Mol. Biol 318, 697–706.[CrossRef][ISI][Medline]
  8. Gilead, S. & Gazit, E. (2004) Angew. Chem. Int. Ed 43, 4041–4044.
  9. Tatarek-Nossol, M., Yan, L. M., Schmauder, A., Tenidis, K., Westermark, G. & Kapurniotu, A. (2005) Chem. Biol 12, 797–809.[CrossRef][ISI][Medline]
  10. Hull, R. L., Westermark, G. T., Westermark, P. & Kahn, S. E. (2004) J. Clin. Endocrinol. Metab 89, 3629–3643.[Abstract/Free Full Text]
  11. Kapurniotu, A. (2001) Biopolymers 60, 438–459.[CrossRef][ISI][Medline]
  12. Wimalawansa, S. J. (1997) Crit. Rev. Neurobiol 11, 167–239.[ISI][Medline]
  13. Schmitz, O., Brock, B. & Rungby, J. (2004) Diabetes 53s3, 233–238.
  14. Kayed, R., Bernhagen, J., Greenfield, N., Sweimeh, K., Brunner, H., Voelter, W. & Kapurniotu, A. (1999) J. Mol. Biol 287, 781–796.[CrossRef][ISI][Medline]
  15. Padrick, S. B. & Miranker, A. D. (2002) Biochemistry 41, 4694–4703.[CrossRef][Medline]
  16. Stites, W. E. (1997) Chem. Rev 97, 1233–1250.[CrossRef][ISI][Medline]
  17. DeGrado, W. F. (1988) Adv. Prot. Chem 39, 51–124.[ISI][Medline]
  18. Kapurniotu, A., Schmauder, A. & Tenidis, K. (2002) J. Mol. Biol 315, 339–350.[CrossRef][ISI][Medline]
  19. Hruby, V. J. (2002) Nat. Rev. Drug. Discov 1, 847–858.[CrossRef][ISI][Medline]
  20. Sagan, S., Karoyan, P., Lequin, O., Chassaing, G. & Laviell, S. (2004) Curr. Med. Chem 11, 2799–2822.[ISI][Medline]
  21. Hunter, C. A. & Tomas, S. (2003) Chem. Biol 10, 1023–1032.[CrossRef][ISI][Medline]
  22. Tenidis, K., Waldner, M., Bernhagen, J., Fischle, W., Bergmann, M., Weber, M., Merkle, M.-L., Voelter, W., Brunner, H. & Kapurniotu, A. (2000) J. Mol. Biol 295, 1055–1071.[CrossRef][ISI][Medline]
  23. LeVine, H. III. (1999) in Methods in Enzymology: Amyloid, Prions, and Other Protein Aggregates ed. Wetzel, R. (Academic, San Diego), Vol. 309, pp. 274–284.[CrossRef]
  24. Eftink, M. R. (1997) in Methods in Enzymology: Fluorescence Spectroscopy eds. Brand, L. & Johnson, M. L. (Academic, San Diego), Vol. 278, pp. 221–257.
  25. Zimmermann, U., Fluehmann, B., Born, W., Fischer, J. A. & Muff, R. (1997) J. Endocrinol 155, 423–431.[Abstract]
  26. Poyner, D. R., Sexton, P. M., Marshall, I., Smith, D. M., Quirion, R., Born, W., Muff, R., Fischer, J. A. & Foord, S. M. (2002) Pharmacol. Rev 54, 233–246.[Abstract/Free Full Text]
  27. Kapurniotu, A., Buck, A., Weber, M., Schmauder, A., Hirsch, T., Bernhagen, J. & Tatarek-Nossol, M. (2003) Chem. Biol 10, 149–159.[CrossRef][ISI][Medline]
  28. Aitken, J. F., Loomes, K. M., Konarkowska, B. & Cooper, G. J. S. (2003) Biochem. J 374, 779–784.[CrossRef][ISI][Medline]
  29. Fowler, S. B., Poon, S., Muff, R., Chiti, F., Dobson, C. M. & Zurdo, J. (2005) Proc. Natl. Acad. Sci. USA 102, 10105–10110.[Abstract/Free Full Text]
  30. Kazantzis, A., Waldner, M., Taylor, J. W. & Kapurniotu, A. (2002) Eur. J. Biochem 269, 780–791.[ISI][Medline]
  31. Dursina, B., Thomä, N. H., Sidorovitch, V., Niculae, A., Iakovenko, A., Rak, A., Albert, S., Ceacareanu, A.-C., Kölling, R. & Hermann, C., et al. (2002) Biochemistry 41, 6805–6816.[CrossRef][Medline]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg    What's this?


This article has been cited by other articles in HighWire Press-hosted journals:


Home page
 DiabetesHome page
S. Zhang, H. Liu, H. Yu, and G. J.S. Cooper
Fas-Associated Death Receptor Signaling Evoked by Human Amylin in Islet {beta}-Cells
Diabetes, February 1, 2008; 57(2): 348 - 356.
[Abstract] [Full Text] [PDF]

Home page
 Biophys. JHome page
P. Soto, M. A. Griffin, and J.-E. Shea
New Insights into the Mechanism of Alzheimer Amyloid-{beta} Fibrillogenesis Inhibition by N-Methylated Peptides
Biophys. J., November 1, 2007; 93(9): 3015 - 3025.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow Supporting Information
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a colleague
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via ISI Web of Science (10)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Yan, L.-M.
Right arrow Articles by Kapurniotu, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Yan, L.-M.
Right arrow Articles by Kapurniotu, A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg  
What's this?