A Structural Core Within Apolipoprotein C-II Amyloid Fibrils Identified Using Hydrogen Exchange and Proteolysis

Abstract

Plasma apolipoproteins show α-helical structure in the lipid-bound state and limited conformational stability in the absence of lipid. This structural instability of lipid-free apolipoproteins may account for the high propensity of apolipoproteins to aggregate and accumulate in disease-related amyloid deposits. Here, we explore the properties of amyloid fibrils formed by apolipoproteins using human apolipoprotein (apo) C-II as a model system. Hydrogen-deuterium exchange and NMR spectroscopy of apoC-II fibrils revealed core regions between residues 19–37 and 57–74 with reduced amide proton exchange rates compared to monomeric apoC-II. The C-terminal core region was also identified by partial proteolysis of apoC-II amyloid fibrils using endoproteinase GluC and proteinase K. Complete tryptic hydrolysis of apoC-II fibrils followed by centrifugation yielded a single peptide in the pellet fraction identified using mass spectrometry as apoC-II_56-76. Synthetic apoC-II_56-76 readily formed fibrils, albeit with a different morphology and thioflavinT fluorescence yield compared to full-length apoC-II. Studies with smaller peptides narrowed this fibril-forming core to a region within residues 60–70. We postulate that the ability of apoC-II_60-70 to independently form amyloid fibrils drives fibril formation by apoC-II. These specific amyloid-forming regions within apolipoproteins may underlie the propensity of apolipoproteins and their peptide derivatives to accumulate in amyloid deposits in vivo.

Introduction

Current interest in the structure and assembly mechanisms for amyloid fibrils stems from their association with a wide range of debilitating diseases. In addition, amyloid fibrils are an alternative stable structural fold, the study of which could yield valuable insight into the fundamental rules of protein folding.¹ In this vein, over two decades of structural studies on amyloid fibrils have painted an informative, if low-resolution, picture of the common features and supramolecular organization of the amyloid fibril, detailing a cross-β core consisting of tightly packed, self-complementing β sheets.2., 3., 4. At the molecular level however, it is clear that amyloid fibrils formed by different proteins differ structurally in a number of ways, including the extent to which cross-β structure is formed along their sequences, strand orientation and the organization of the core amyloid structure within the protein.5., 6., 7., 8. These differences are likely to have significant effects on both the solution and functional properties of amyloid fibrils and could substantially impact on their in vivo processing and manifestation during different disease states. In the absence of a complete understanding of the general formation of amyloid structure, local structural information on protein and disease-specific amyloid fibrils remains critical.

Of the set of proteins that form amyloid fibrils in vivo, plasma apolipoproteins are over-represented. For instance, apolipoprotein (apo)¹ A-I mutants are associated with several hepatic and systemic amyloid disorders,9., 10. while apoA-II forms amyloid fibrils with a renal localization.¹¹ Serum amyloid A, an acute phase reactant of the apolipoprotein family, self-aggregates to form amyloid fibrils at various sites of inflammation.¹² A C-terminal fragment of apoE binds Aβ fibrils in neuritic plaques and itself forms amyloid fibrils in vitro.¹³ In addition, we have recently demonstrated that apoB in low-density lipoproteins (LDL) acquires amyloid-like structures under oxidizing conditions. Oxidation of LDL promotes their uptake by macrophages in a step that precedes foam cell formation and atherosclerosis.¹⁴ Furthermore, immunohistochemical studies of atherosclerotic plaques reveal the presence of apoA-I, apoB, apoC-II and apoE aggregates.15., 16. The aberrant deposition of these amyloid species in the artery wall may contribute to decreased elasticity of blood vessels, the induction of vascular inflammation, and alterations in lipid metabolism.¹⁷

Sequence comparisons of apolipoprotein genes indicate a multigene family with a common ancestry.¹⁸ A common property of the apolipoprotein family is the high proportion of class-A amphipathic helices implicated in lipid binding,¹⁹ and a limited conformational stability in the absence of lipid that is postulated to underlie their propensity to form amyloid.²⁰ In addition, proteases that are abundant in atherosclerotic lesions, cathepsins B, K, and L, cleave apolipoprotein A-I to generate intermediate-sized fragments that form amyloid.²¹ All of these parameters may contribute to the accumulation of apolipoprotein-derived amyloid in atherosclerotic lesions. However, despite their apparent prevalence in amyloid deposits, little is known about the mechanisms of amyloid formation by apolipoproteins or their detailed structure in the amyloid state.

Human apoC-II is a 79-residue component of very-low-density lipoproteins, where it plays an essential role in activating lipoprotein lipase during lipid metabolism. ApoC-II readily aggregates under lipid-poor conditions to form homogeneous fibrils²² and is one of the few amyloid systems to form fibrils at physiological pH without prolonged agitation. Plasma apoC-II accumulates in atherosclerotic plaques,²³ where it co-localizes with serum amyloid P component, a non-fibrillar marker of amyloid deposits.¹⁶ While this co-localization provides a prima facie case for the presence of apoC-II fibrils in vivo, definitive evidence requires the development of specific reagents that can distinguish fibrillar and non-fibrillar forms of this protein. The amyloidogenic properties of apoC-II have been extensively studied in vitro and these studies form the basis for much of the current knowledge of amyloid fibril formation by the apolipoproteins.¹⁷ ApoC-II fibrils initiate macrophage inflammatory responses via the CD36 receptor, including reactive oxygen production and TNF-α expression, which promote atherogenesis.¹⁶ Here, we study apoC-II as a model system for amyloid fibrils formed by the apolipoproteins. Using a combined approach of hydrogen/deuterium (H/D) exchange and proteolysis experiments we sought to identify the core region(s) within apoC-II fibrils that predispose to amyloid formation.