Bovine insulin filaments induced by reducing disulfide bonds show a different morphology, secondary structure, and cell toxicity from intact insulin amyloid fibrils

Abstract

Amyloid fibrils are associated with more than 20 diseases, including Alzheimer's disease and type II diabetes. Insulin is a 51-residue polypeptide hormone, with its two polypeptide chains linked by one intrachain and two interchain disulfide bonds, and has long been known to self-assemble in vitro into amyloid fibrils. We demonstrate here that bovine insulin forms flexible filaments in the presence of a reducing agent, Tris (2-carboxyethyl) phosphine. The insulin filaments, possibly formed due to partial reduction of S-S bonds in insulin molecules, differ from intact insulin fibrils in terms of their secondary structure. The insulin filaments were determined to have an antiparallel beta-sheet structure, whereas the insulin fibrils have a parallel beta-sheet structure. Of importance, the cell toxicity of the insulin filaments was remarkably lower than that of the insulin fibrils. This finding supports the idea that cell toxicity of amyloids correlates with their morphology. The remarkably low toxicity of the filamentous structure should shed new light on possible pharmacological approaches to the various diseases caused by amyloid fibrils.

Figure 1

ThT assay and CR assay of insulin aggregation in the presence of TCEP. (A) Time-course assay monitored by ThT fluorescence: bovine insulin was incubated in the absence (●) or presence (■) of 20 mM TCEP. ThT fluorescence at 485 nm was measured. The intensities were corrected to correspond to the equal amount (5 μM as monomer concentration) of the assembled insulin fibrils and filaments after 13 h incubation. (B) ThT assay: 10 μL of 1 mM insulin aggregates formed in the absence (broken line) and presence (solid line) of 20 mM TCEP dissolved in distilled water were added to a 2 mL solution containing 5 μM ThT in 10 mM phosphate, pH 7.4, 150 mM NaCl (dotted line), and fluorescence spectra with the excitation wavelength at 450 nm at 25°C were obtained by the spectrofluorometer. (C) CR assay: 10 μL of 1 mM insulin aggregates formed in the absence (broken line) and presence (solid line) of 20 mM TCEP dissolved in distilled water were added to a 2 mL solution containing 25 μM CR in 10 mM phosphate, pH 7.4, 150 mM NaCl (dotted line). After 30 min incubation at room temperature, absorbance spectra were obtained with an ultraviolet-visible spectrophotometer.

Figure 2

TEM analysis of insulin aggregates. Bovine insulin was incubated at a final concentration of 2 mg/mL at pH 1.6 and 70°C for 13 h in the absence (A) and presence (B) of 20 mM TCEP. Equal molar amounts of insulin A-chain and B-chain were co-incubated under identical conditions in the absence (C) and presence (D) of 20 mM TCEP. The scale bar represents 200 nm.

Figure 3

Far-UV CD spectra of insulin samples: 0.3 mg/mL insulin samples of native insulin (dotted line), insulin fibrils formed in the absence of TCEP (broken line), and insulin filaments formed in the presence of 20 mM TCEP (solid line) were assayed. The photomultiplier voltage is also shown in the figure.

Figure 4

Changes in the secondary structure composition of insulin samples with FTIR. ATR-FTIR spectra insulin samples: (A) native insulin, (B) insulin fibrils formed in the absence of TCEP, and (C) insulin filaments formed in the presence of 20 mM TCEP. The dotted line represents the raw data, and the solid line represents the fitted data used for deconvolution (broken line).

Figure 5

Raman studies of insulin fibrils and filaments. (A) Solid-state spectra of insulin fibrils and filaments formed in the absence and presence of 20 mM TCEP. Spectra were normalized relative to the 1673 cm⁻¹ amide I band as an internal standard. 1) A-chain and B-chain incubated with TCEP. 2) A-chain and B-chain incubated without TCEP. 3) Insulin incubated with TCEP. 4) Insulin incubated without TCEP. Peak frequencies are indicated in the figure. The broken lines indicate S-S stretching frequencies centered at 514 cm⁻¹ and S-H stretching frequencies at 2574 cm⁻¹, respectively. (B and C) Expansion of the amide I Raman region with Gaussian deconvolution of insulin fibrils formed in the absence of TCEP (B) and filaments formed in the presence of TCEP (C). (D) Expansion of the amide III Raman region of insulin fibrils formed in the absence of TCEP (broken line) and insulin filaments in the presence of TCEP (solid line).

Figure 6

Size exclusion chromatography of supernatants of incubated insulin. After incubation of insulin in buffer A in the presence and absence of 20 mM TCEP at a protein insulin concentration of 2 mg/mL at 70°C for 13 h, the samples were centrifuged (15,000 rpm for 60 min). The 100 μL aliquots of supernatant, including 0.2 mg/mL insulin samples treated with TCEP (solid line) and without TCEP (broken line) diluted in the running buffer (45% acetonitrile and 0.1% TFA), were applied to a G2000SWXL high-performance liquid chromatography column at 25°C at a flow rate of 1.0 mL/min, and the absorbance of the sample was monitored at 220 nm. The 100 μL aliquots of buffer A diluted (1/5 dilution) in the running buffer were applied as a control (dotted line). Native insulin (5734 kDa) and insulin B-chain (3399 kDa) were used for marker samples.

Figure 7

Cell toxicity of insulin fibrils and filaments. Cell viability was determined with the use of an MTT cell proliferation kit. PC12 cells were plated in 96-well plates at a density of 50,000 cells/well and grown overnight. Cells were then incubated in the absence (control) and presence of either insulin filaments (black) or insulin fibrils (white) dialyzed with PBS buffer at indicated concentrations.