Identification of a penta- and hexapeptide of islet amyloid polypeptide (IAPP) with amyloidogenic and cytotoxic properties

Abstract

Pancreatic amyloid is found in more than 95 % of type II diabetes patients. Pancreatic amyloid is formed by the aggregation of islet amyloid polypeptide (hIAPP or amylin), which is a 37-residue peptide. Because pancreatic amyloid is cytotoxic, it is believed that its formation is directly associated with the development of the disease. We recently showed that hIAPP amyloid formation follows the nucleation-dependent polymerization mechanism and proceeds via a conformational transition of soluble hIAPP into aggregated beta-sheets. Here, we report that the penta- and hexapeptide sequences, hIAPP(23-27) (FGAIL) and hIAPP(22-27) (NFGAIL) of hIAPP are sufficient for the formation of beta-sheet-containing amyloid fibrils. Although these two peptides differ by only one amino acid residue, they aggregate into completely different fibrillar assemblies. hIAPP(23-27) (FGAIL) fibrils self-assemble laterally into unusually broad ribbons, whereas hIAPP(22-27) (NFGAIL) fibrils coil around each other in a typical amyloid fibril morphology. hIAPP(20-27) (SNNFGAIL) also aggregates into beta-sheet-containing fibrils, whereas no amyloidogenicity is found for hIAPP(24-27) (GAIL), indicating that hIAPP(23-27) (FGAIL) is the shortest fibrillogenic sequence of hIAPP. Insoluble amyloid formation by the partial hIAPP sequences followed kinetics that were consistent with a nucleation-dependent polymerization mechanism. hIAPP(22-27) (NFGAIL), hIAPP(20-27) (SNNFGAIL), and also the known fibrillogenic sequence, hIAPP(20-29) (SNNFGAILSS) exhibited significantly lower kinetic and thermodynamic solubilities than the pentapeptide hIAPP(23-27) (FGAIL). Fibrils formed by all short peptide sequences and also by hIAPP(20-29) were cytotoxic towards the pancreatic cell line RIN5fm, whereas no cytotoxicity was observed for the soluble form of the peptides, a notion that is consistent with hIAPP cytotoxicity. Our results suggest that a penta- and hexapeptide sequence of an appropriate amino acid composition can be sufficient for beta-sheet and amyloid fibril formation and cytotoxicity and may assist in the rational design of inhibitors of pancreatic amyloid formation or other amyloidosis-related diseases.

Amyloid Fiber Formation and Membrane Disruption are Separate Processes Localized in Two Distinct Regions of IAPP, the Type-2-Diabetes-Related Peptide

“…Previous studies have shown that the N-terminal 1-19 region, rather than the amyloidogenic 20-29 region, is primarily responsible for the interaction of the IAPP peptide with membranes. Liposome leakage experiments presented in this study confirm that the pathological membrane disrupting activity of the full-length hIAPP is also shared by hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] . The hIAPP 1-19 fragment at a low concentration of peptide induces membrane disruption to a near identical extent as the full-length peptide.…”

“…The hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] and hIAPP [20][21][22][23][24][25][26][27][28][29] peptides were synthesized using standard FMOC (9-fluornylmethoxycarbonyl) methods. A PAL-PEG resin was used to provide an amidated C-terminus.…”

“…The hIAPP 20-29 peptide was synthesized as described above but with both an amidated C-terminus and an acetylated N-terminus. After synthesis, the crude hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] peptide was subjected to air oxidation to form the disulfide bond. The identity of the hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] and hIAPP [20][21][22][23][24][25][26][27][28][29] sequences and formation of the disulfide bond for hIAPP [1][2][3][4]…”

“…Lyophilized hIAPP and hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] were dissolved in HFIP for 1 h to obtain a 4 mg/mL stock solution. Aliquots of the peptide stock solution were lyophilized again and then resolubilized in either 10 mM sodium phosphate buffer, pH 7.3 or 10 mM sodium phosphate buffer with 150 mM NaF, pH 7.3.…”

“…It has previously been shown through fluorescence anisotropy, infrared reflectionabsorption spectroscopy, and lipid monolayer expansion experiments that the N-terminal part of IAPP is the primary membrane binding site of IAPP.11,14,15 Indeed, the 1 -19 fragment of human IAPP (hIAPP [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19] ) inserts into lipid membranes with a higher efficiency than full-length IAPP.11 We report here that the N-terminal region of the peptide (residues 1 -19) can disrupt synthetic lipid vesicles to a similar extent as the full-length IAPP peptide without forming amyloid fibers and may therefore be an important model for the study of membrane disruption by IAPP and other amyloid proteins.…”