Role of Cholesterol and Phospholipids in Amylin Misfolding, Aggregation and Etiology of Islet Amyloidosis

Abstract

Amyloidosis is a biological event in which proteins undergo structural transitions from soluble monomers and oligomers to insoluble fibrillar aggregates that are often toxic to cells. Exactly how amyloid proteins, such as the pancreatic hormone amylin, aggregate and kill cells is still unclear. Islet amyloid polypeptide, or amylin, is a recently discovered hormone that is stored and co-released with insulin from pancreatic islet β-cells. The pathology of type 2 diabetes mellitus (T2DM) is characterized by an excessive extracellular and intracellular accumulation of toxic amylin species, soluble oligomers and insoluble fibrils, in islets, eventually leading to β-cell loss. Obesity and elevated serum cholesterol levels are additional risk factors implicated in the development of T2DM. Because the homeostatic balance between cholesterol synthesis and uptake is lost in diabetics, and amylin aggregation is a hallmark of T2DM, this chapter focuses on the biophysical and cell biology studies exploring molecular mechanisms by which cholesterol and phospholipids modulate secondary structure, folding and aggregation of human amylin and other amyloid proteins on membranes and in cells. Amylin turnover and toxicity in pancreatic cells and the regulatory role of cholesterol in these processes are also discussed.

Fig. 4.1

Aggregation of human amylin and changes of its secondary structure coincides in time. (a) Primary structures of mature human (hIAPP) and rat (rIAPP) amylin are depicted. Species-specific aminoacids within the amyloidoigenic region (underlined) of the polypeptide chain are bolded for clarity. (b) Kinetics and extent of aggregation of human and rat amylin in PBS as a function of time. Thioflavin-T fluorescent assay reveals fibrilogenesis of 20 μM human amylin in solution (closed circles) and lack of aggregation of non-amyloidogenic rat amylin (20 μM; open circles). (c) Far-UV CD spectra of human amylin (solid line) and rat amylin (dashed line) taken after 20 min. in PBS solution in the presence of 2 % HFIP. Note the absorption minimum at ~220 nm for human but not rat amylin, typical for peptides and proteins adopting β-sheet conformation

Fig. 4.2

Dynamics of amylin aggregation on solid surface. (a) Structural intermediates, oligomers and fibrils, are resolved during amylin aggregation on mica by time-lapse AFM (tapping mode amplitude images). Note a time-dependent transition of human amylin from small round oligomers (0–10 min) into fibrils during early-mid stage of amylin aggregation (10–25 min). Late–stage of amylin aggregation (25–35 min) is characterized by accumulation of massive peptide deposits on the mica surface. All micrographs are 5 × 5 μm. (b) Fibril growth curves reveal two phases of amylin aggregation, an early oligomeric phase (0–10 min) characterized by oligomers formation, followed by oligomers incorporation into growing fibrils (10–25 min, second phase or fibril maturation). Note the significant increase in oligomer heights (inset) and widths during the first phase of amylin aggregation, and an abrupt increase in fibrils length following formation of full-size oligomers. Data represents mean particle size at each time point (mean ± SEM), obtained from three independent time-lapse AFM experiments. (c) 3-D AFM image of a single full-grown fibril on mica showing arrangement of several amylin oligomers and their bi-directional extension into a fibril (depicted by arrowheads). Micrograph is 800 × 800 nm scale

Fig. 4.3

Membrane cholesterol and anionic phospholipids oppositely regulate amylin aggregation and misfolding in solution. (a) Thioflavin-T fluorescent assay reveals slow aggregation of 10 μM human amylin in solution (circles). Presence of 100 μM anionic liposomes (PC:PS, 2.8:1.2 mol:mol, squares) in incubating solution (PBS, 1 % HFIP) accelerates amylin aggregation, the effect of which was reversed by inclusion of cholesterol in the lipid vesicles (PC:PS:Chol, 2.3:1:0.8 mol:mol:mol, triangles). (b) Dynamics of amylin secondary structural transitions in solution are regulated by membranes. CD spectras of human amylin (10 μM) incubated with 100 μM PC:PS liposomes (2.8:1.2 mol:mol, black trace) or PC:PS:Chol liposomes (2.3:1:0.8 mol:mol:mol, gray trace) were continuously acquired at 220 nm to monitor appearance of β-sheet conformation. Note that the onset of amylin aggregation and the transition from random coil to β-sheet coincide (hA+PC:PS, Fig. 4.3a, b). Inclusion of cholesterol prolonged amylin’s transition to β-sheets evoked by anionic liposomes (Fig. 4.3b), ultimately reducing kinetics and the extent of amylin aggregation (Fig. 4.3a)

Fig. 4.4

Dynamics and organization of amylin aggregates on planar membranes. (a) Amylin (20 μM) at time zero was injected into the imaging chamber and the peptide membrane assembly was monitored in real time by time-lapse AFM. All micrographs are 5 × 5 μm, and are taken at the same time intervals of 5 min. (b) Section analysis of amylin aggregates on anionic membranes. Channel-like topology of two amylin supramolecular complexes featuring a central pore is shown (b, PC:PS, 2.8:1.2 mol:mol, 10 min, inset). High-resolution 2D AFM micrographs of several amylin supramolecular complexes on PC:PS membranes are shown. Tetrameric and pentameric subunits are outlined (b, right panel). Bar is 50 nm. (c) Quantitative analysis of cholesterol-regulated amylin assembly on anionic membranes. Presence of cholesterol in planar membranes (PC:PS:Chol,2.3:1:0.8 mol:mol:mo l, circles) stimulates a significant increase in the size (height) of amylin aggregates over time when compared with cholesterol-depleted membranes (PC:PS, 2.8:1.2 mol:mol, filed squares, c, height plot). Cholesterol abrogates amylin deposition and overall membrane surface coverage with amylin (c, surface plot). Cholesterol also inhibits seeding of amylin aggregates on PC:PS membranes (c, particle plot). The mean (V_m) and the total particle volume (V_t) of amylin aggregates on PC:PS membranes are significantly different in the presence of cholesterol after 20 min (*p < 0.05 and **p < 0.01, n = 3 Student’s t-test) (c, volume plot)

Fig. 4.5

AFM analysis of membrane-directed amylin self-assembly. High-resolution 2D AFM micrographs (top panel) reveal distinct patterns of amylin aggregation and deposition on different surfaces. Note the clustering of amylin aggregates on cholesterol-containing membranes, PC:Chol (3.2:0.8 mol:mol) and PC:PS:Chol (2.3:1:0.8 mol:mol:mol) and their homogenous distribution/aggregation on cholesterol-free membranes, PC and PC:PS (2.8:1.2 mol:mol). In contrast to mica, no fibrils were detected on either membranes. Micrographs are 2 × 2 μm. Bottom panel: Proposed pathway of amylin polymerization and accumulation on different surfaces. The form and amount of amylin deposits correlate with the physicochemical properties of the supporting surface. On stiffpolar mica surface, amylin monomers (left) associate into spherical oligomers that align and elongate overtime to produce mature fibrils that randomly distribute across the surface (mica). On soft planar membranes, amylin self-assembles into globular highly symmetrical supramolecular structures featuring a central pore. Unstructured amorphous amylin aggregates are also formed on this surface (membrane). Incorporation of cholesterol into planar membranes redirects amylin surface deposition by stimulating formation of larger, but fewer amylin clusters (memb+chol). Consequently, the membrane surface area free of amylin deposits increases significantly, which diminishes amylin accumulation. Three major polymorphic forms, a fibril, a pore and a single cluster formed during amylin polymerization on different surfaces, are presented top to bottom (bottom panel)

Fig. 4.6

Binding and clustering of amylin oligomers into microdomains on the cell PM requires cholesterol. (a) Confocal microscopy analysis of amylin oligomer and cholera toxin (CTX) distribution on the cell PM. Characteristic binding profiles of amylin oligomers on the cell PM for each treatment (within boxes, right panel) are rendered in gray tones for easier particle comparisons, which are presented side by side with the original fluorescence images (left panels). Note the time-dependent increase in the number of internalized amylin oligomers (control, 30 min vs. 24 h, left panel), which prevents accumulation of amylin oligomers on the cell PM (control, 30 min vs. 24 h, right box). Single particle analysis demonstrates a threefold increase in the number of amylin oligomer clusters, or puncta, on the PM (particle no.), and their dispersion across the PM in cholesterol-depleted cells (mean particle area). Significance established at *p < 0.05, **P < 0.01 control vs. BCD/Lov, and #p < 0.05, ##p < 0.01 BCD/Lov vs. BCD/Lov/Chol. (b) Amylin monomer internalization is not blocked by cholesterol depletion. Confocal microscopy demonstrates internalization of amylin monomers both in controls (hA) and cholesterol-depleted cells (hA + BCD/Lov). No significant (NS) change in PM-binding pattern of amylin monomers (right boxes) was noticed upon cholesterol depletion. PM cholesterol does not modulate amylin deposition on the PM. The number of amylin puncta on PM (particle no.) and their area (mean particle area) did not change significantly upon depletion of PM cholesterol by BCD/Lov. Bars are 5 μm

Fig. 4.7

PM cholesterol prevents toxicity of soluble amylin oligomers in cultured human islet cells. Confocal microscopy analysis of phosphatidylserine (PS) externalization and caspase-3 proteolytic activation by amylin in cells with normal, depleted and enriched cholesterol levels (micrographs, left panel). Arrows depict non-apoptotic nuclei (blue) in viable cells, whereas arrowheads depict fluorogenic caspase-3 substrate found in the nuclei of apoptotic cells, giving these nuclei a green/blue appearance. The majority of PS-positive cells (red) show shrinkage and nuclear condensation (arrowheads) indicative of apoptosis (hA and hA+BCD/Lov). Bar is 10 μm. Linear regression analysis shows an inverse relationship between amylin-induced cell death and PM cholesterol levels. The extent of cell death evoked by amylin was plotted as a function of variable PM cholesterol levels in controls and treatments (apoptosis vs. cholesterol, graph). Quantitative analysis of amylin-induced apoptosis in human islets (caspase-3/annexin, graph) reveals a significant increase in amylin toxicity in cholesterol-depleted cells as compared to controls (@@p < 0.01 hA vs. control, **p < 0.01 hA vs. treatments, right panel). Replenishment of PM cholesterol levels significantly decreased amylin toxicity (##p < 0.01 hA+BCD/lov vs. hA+BCD/Lov/Chol)