Annexin A5 directly interacts with amyloidogenic proteins and reduces their toxicity

Abstract

Protein misfolding is a central mechanism for the development of neurodegenerative diseases and type 2 diabetes mellitus. The accumulation of misfolded alpha-synuclein protein inclusions in the Lewy bodies of Parkinson's disease is thought to play a key role in pathogenesis and disease progression. Similarly, the misfolding of the beta-cell hormone human islet amyloid polypeptide (h-IAPP) into toxic oligomers plays a central role in the induction of beta-cell apoptosis in the context of type 2 diabetes. In this study, we show that annexin A5 plays a role in interacting with and reducing the toxicity of the amyloidogenic proteins, h-IAPP and alpha-synuclein. We find that annexin A5 is coexpressed in human beta-cells and that exogenous annexin A5 reduces the level of h-IAPP-induced apoptosis in human islets by approximately 50% and in rodent beta-cells by approximately 90%. Experiments with transgenic expression of alpha-synuclein in Caenorhabditis elegans show that annexin A5 reduces alpha-synuclein inclusions in vivo. Using thioflavin T fluorescence, electron microscopy, and electron paramagnetic resonance, we provide evidence that substoichiometric amounts of annexin A5 inhibit h-IAPP and alpha-synuclein misfolding and fibril formation. We conclude that annexin A5 might act as a molecular safeguard against the formation of toxic amyloid aggregates.

Figure 1

(A) Mean number of apoptotic cells in human islets (n = 3 donors, 7–9 experiments per group, total islet number per group 61–97) after incubation for 48 h with vehicle (control), rat IAPP (40 μM), h-IAPP (40 μM), or h-IAPP (40 μM) with annexin A5 (1 μM). (B) Percentage of apoptotic RIN cells during a 24 h incubation (time-lapse microscopy) with vehicle (control), rat IAPP (40 μM), h-IAPP (40 μM), or h-IAPP (40 μM) with annexin A5 (1 μM). Data ( SEM. P-value derived by ANOVA. Group comparisons by *: p < 0.05 versus control. #: p < 0.05 versus h-IAPP.

Figure 2

Spinning-disc confocal microscopic images of 3 μM sections of three isolated human islets from two nondiabetic organ donors. Immunostaining was performed for annexin A5 (red), human IAPP (green), and nuclei with DAPI (blue). (A) Three-dimensional reconstruction of the full section. (B and C) Single-plane images. Annexin A5 is present in β- and non-β-cells in human islets. Magnification of 1000×.

Figure 3

Thioflavin T curves and EM images of amyloid fibrils with and without annexin A5. (A) Time-dependent normalized ThT emission curves of h-IAPP (20 μM) with (dashed line) and without (solid line) annexin A5 (4 μM). The solid and dashed lines are derived by nonlinear regression analysis of three (n) experiments per group. (B) Time-dependent normalized ThT emission curves of 115ter α-synuclein (100 μM) with (solid line) and without annexin A5 (20 μM) (dashed line). The ThT curves are normalized to the maximal observed intensity at the end of each aggregation reaction. Panels C and E show EM images of 100 μM h-IAPP (3000× magnification) and 115ter α-synuclein (6000× magnification) fibrils, respectively, grown in the absence of annexin A5. Panels D and F show h-IAPP (4000× magnification) and 115ter α-synuclein (2000× magnification) with annexin A5 in a 3:1 ratio. Amorphous protein aggregates and no fibrils are seen throughout the sample when amyloidogenic proteins are observed in the presence of annexin A5. The inset shows a 15000× magnification to illustrate the absence of fibrils at higher magnifications.

Figure 4

EPR spectra of R1-labeled h-IAPP and 115ter α-synuclein in the presence of annexin A5. To assess the effect of annexin A5 on amyloid misfolding, 20 μM annexin A5 was added to 100 μM spin-labeled (A) h-IAPP (21R1) or (B) 115ter α-synuclein (52R1) peptide and allowed to form fibrils. The black spectra are for control amyloid fibrils grown in the absence of annexin A5, and the gray spectra are for amyloid fibrils grown in the presence of annexin A5. A sample of 115ter α-synuclein was incubated with lysozyme (C) to control for structural effects caused by the presence of any large protein. All spectra were normalized to the same number of spins.

Figure 5

Annexin A5 expression decreases the number of α-synuclein inclusions in vivo. (A) Representative fluorescent images of head muscles of transgenic young adult animals expressing either α-synuclein::YFP alone (the pkIs2386 transgene) or α-synuclein:: YFP with human annexin A5 (pkIs2386; vjEx137). (B) False colored images from panel A with red and purple representing the highest and lowest pixel intensities, respectively. (C) Thresholded images from panel A showing pixel intensities above background in green. (D) Quantification of the number of inclusions above threshold per pharynx averaged over 20 animals. Data from two independently generated transgenic lines expressing annexin A5 are shown. The standard error is shown (p < 0.001).

Figure 6

Expression of α-synuclein in annexin A5 transgenic C. elegans. (A) Immunoblots of total protein extracts from control animals containing no α-synuclein or annexin A5 (lane 1), α-synuclein transgene (lane 2), and α-synuclein with the annexin A5 transgene (lane 3). The samples were subjected to SDS–PAGE and Western blotting with anti-α-synuclein and anti-actin antibodies to detect whether the annexin A5 transgene alters the expression of α-synuclein. Actin (43 kDa) was used as a control. (B) α-Synuclein protein levels in animals with and without the annexin A5 transgene. (C) Immunoblot of total protein extracts from animals containing YFP only (lane 1) and YFP with annexin A5 (lane 2). The samples were subjected to SDS–PAGE and Western blotting with anti-YFP and anti-actin antibodies to detect whether the annexin A5 transgene alters the expression of YFP. (D) YFP protein levels in animals with and without the annexin A5 transgene.

Figure 7

Model explaining the mechanism of annexin A5 action. Annexin A5 alters misfolding and interferes with fibril formation. This effect is caused by interaction with a subset of molecules, which are generated during the misfolding process rather than a direct interaction with the bulk monomer. The precise nature of the species with which annexin A5 interacts is not yet known, but it is likely that they represent misfolded species that are directly on the pathway to fibril formation.