Imaging β-amyloid fibrils in Alzheimer's disease: a critical analysis through simulation of amyloid fibril polymerization

https://doi.org/10.1016/j.nucmedbio.2005.02.003Get rights and content

Abstract

The polymerization of β-amyloid (Aβ) peptides into fibrillary plaques is implicated, in part, in the pathogenesis of Alzheimer's disease. Aβ molecular imaging probes (Aβ-MIPs) have been introduced in an effort to quantify amyloid burden or load, in subjects afflicted with AD by invoking the classic PET receptor model for the quantitation of neuronal receptor density. In this communication, we explore conceptual differences between imaging the density of amyloid fibril polymers and neuronal receptors. We formulate a mathematical model for the polymerization of Aβ with parameters that are mapped to biological modulators of fibrillogenesis and introduce a universal measure for amyloid load to accommodate various interactions of Aβ-MIPs with fibrils. Subsequently, we hypothesize four Aβ-MIPs and utilize the fibrillogenesis model to simulate PET tissue time activity curves (TACs). Given the unique nature of polymer growth and resulting PET TAC, the four probes report differing amyloid burdens for a given brain pathology, thus complicating the interpretation of PET images. In addition, we introduce the notion of an MIP's resolution, apparent maximal binding site concentration, optimal kinetic topology and its resolving power in characterizing the pathological progression of AD and the effectiveness of drug therapy. The concepts introduced in this work call for a new paradigm that goes beyond the classic parameters Bmax and KD to include binding characteristics to polymeric peptide aggregates such as amyloid fibrils, neurofibrillary tangles and prions.

Introduction

Alzheimer's disease (AD) is defined histologically by the presence of intraneuronal neurofibrillary tangles (NFTs) and extracellular β-amyloid (Aβ) plaques in cerebral cortex [1]. Neurofibrillary tangles are highly insoluble neurofibrils formed by hyperphosphorylation of the cytoskeletal protein tau [2]. Intraneuronal neurofibrillary tangles build-up renders cells ineffective resulting in complete degeneration of affected neurons and appearance of “tombstone” NFTs in the extracellular space [3]. On the other hand, Aβ plaques, are extracellular aggregates of polymeric fibrils which are in turn composed of Aβ peptide monomers. We and others have introduced molecular imaging probes (MIPs) in an effort to quantify Aβ burden in the living brains of individuals afflicted with AD [4], [5], [6], [7], [8], [9], [10], [11]. In this endeavor, it is assumed that the quantitation of Aβ load would parallel the quantitation of neuronal receptor density. In this work, we offer insights on the intricacies of imaging Aβ polymeric aggregates in vivo. We will start with a brief cell biology of Aβ processing as well as theories of Aβ polymerization, which will serve as a basis for the formulation of a mathematical model for the polymerization of Aβ. We use the model to generate tissue time activity curves (TACs) for hypothetical MIPs and draw insights from these simulations to address the efficacy of a particular probe in imaging the pathological progression and therapeutic interventions in AD. We note that the concepts presented in this work may be applicable to the prospect of imaging other polymeric aggregates such as NFTs and prions [12], [13].

One of the characteristics of brains afflicted with AD is the presence of extracellular structural elements referred to as amyloid plaques. Plaques are, in part, composed of masses of filaments which are in turn composed of the insoluble form of the Aβ peptide. The Aβ peptide is formed as a cleavage byproduct of a larger amyloid precursor protein (APP) (Fig. 1) [14], [15]. Amyloid precursor protein is a ubiquitous membrane glycoprotein encoded by a single gene on chromosome 21 [16], [17]. It is cleaved either via the α-secretase or the β-secretase pathway, often referred to as the amyloidogenic pathway (Fig. 1). When APP is cleaved by α-secretase, it produces a large amino-terminal fragment APPα destined for secretion and a smaller carboxyl-terminal fragment. Further processing of the carboxyl-terminal fragment by γ-secretase produces a 22- to 24-residue fragment termed P3, which may or may not be amyloidogenic. Alternatively, when APP is cleaved by β-secretase it produces a soluble amino-terminal fragment, APPβ, and a carboxyl-terminal fragment containing the Aβ peptide. Cleavage of the carboxyl-terminal fragment by γ-secretase results in the formation of multiple Aβ variants of 40–43 amino acids, which are prone to aggregate. Aggregation in the extracellular environment contributes to Aβ fibril (fAβ) assembly through a nuclear-dependent polymerization (NDP) reaction [18] or variations thereof [19], [20], [21]. The plaques seen in the brains of AD patients are structural assemblies made up of fAβ and other extracellular material such as astrocytes and microglial cell residues (see Refs. [22], [23], [24]).

Nuclear-dependent polymerization is characterized by a slow, rate-limiting, nucleation step (Fig. 2A). The formation of a nucleus requires a series of monomeric Aβ (mAβ) association steps, which are thermodynamically unfavorable. Once the nucleus is formed, it extends to form larger polymers. The addition of mAβs at this stage is thermodynamically more favorable. The extension of amyloid fibrils (fAβ) is characterized by a steady-state phase in which the polymeric aggregates (fAβ) and monomeric Aβs (mAβs) are in dynamic equilibrium [18], [25], [26]. The steady-state concentration of mAβ is defined as the critical concentration (XC) for polymer extension, and hence for fAβ formation. Below the critical concentration, there is a lag phase during which mAβ aggregate into β-sheet tapes [27]. With increased concentration of mAβ, the concentration of tapes reaches a level at which the tapes stack — forming ribbons — and extend by the deposition of mAβ [27]. At this stage, the Aβ structures are referred to as protofibrils (or protofilaments, Fig. 2B) [19], [28]. Above the critical concentration XC, protofibrils interweave to form fibrils (Fig. 2C) which further polymerize by the addition of mAβ [29], [30].

Section snippets

Imaging statement

An ideal MIP should be sensitive to changes in the underlying biological processes in the progression of the disease. In imaging neuronal death, for example, an MIP should be sensitive to changes in the concentration, or density, of neuronal receptor binding sites as a function of disease pathology. In contrast to neuronal receptors, however, Aβ formations undergo higher-order growth patterns — one in concentration and another in dimension (maturity) — due to the polymerization and deposition

A mathematical model for Aβ polymerization

In the spirit of previous models of particle polymerization [31], [32], and specifically of Aβ polymerization [33], [34], [35], let X denote the concentration Aβ particles, let N denote the concentration of nucleus particles with n Aβ monomers, and let Fi denote the concentration of fibrils with i monomers (i=n+1, n+2, …) (Fig. 2). The rate of change of X is dictated by the production, consumption and degradation of peptide monomers. Monomers are produced by the net influx of particles at a

Methods

In imaging AD, one would be interested in imaging the total Aβ load, or burden. We define amyloid load as the concentration of mAβ in fibrillary form in order to accommodate various interactions of probes with Aβ structures. By one account the concentration of mAβ in late AD patients can reach as high as 3000 nM (Table 1). The ability of a particular probe to measure the total Aβ burden is a function of the probe's resolution. A probe that binds to every fibrillary mAβ is defined to have a

Results

We will start with some insights drawn from the mathematical model for polymer growth which may parallel events in AD progression. Intuitively, it is expected that the deposition of Aβ peptides would contribute to an overall increase in Aβ burden or load. The question that we seek to address in the simulation of polymer growth is how the dynamics of deposition would influence the probe's binding site concentration as a function of total Aβ load or burden. In Fig. 5A, we depict the concentration

Amyloid load, resolution and the imaging unit

The term amyloid load or amyloid burden is often used to quantify Aβ density (diffuse, senile, hardcore plaques, etc.). At this juncture, quantitation of amyloid load becomes a philosophical matter for it depends on the imaging unit. In imaging neuronal receptor density, the imaging unit is the receptor and probes that target a given receptor share a similar target receptor Bmax. The structural complexity of fAβ, however, changes from immature fibrils to mature fibrils prevalent in dense core

Concluding remarks

The challenge in imaging AD is rooted in characterizing amyloid load in the living patient. Currently, an atomic resolution crystal structure of amyloid fibrils is not available; its availability would allow targeting of periodic fibrillary motifs, which in turn would standardize quantitation of amyloid load as a measure for the pathological progression of AD. In this work, we hypothesized four probes that exemplify possible interactions with fAβs. Interestingly, for a given amyloid pathology,

Uncited reference

[89].

Acknowledgments

This work was supported in part by DOE grant DE-FC0302ER63420. KSJ is a recipient of the UCLA Dissertation Fellowship Award.

References (89)

  • J.Q. Trojanowski et al.

    Relationship between plaques, tangles, and dystrophic processes in Alzheimer's disease

    Neurobiol Aging

    (1995)
  • S.H. Yen et al.

    Alzheimer neurofibrillary lesions: molecular nature and potential roles of different components

    Neurobiol Aging

    (1995)
  • J.M. Andreu et al.

    The measurement of cooperative protein self-assembly by turbidity and other techniques

    Methods Enzymol

    (1986)
  • D.M. Walsh et al.

    Amyloid beta-protein fibrillogenesis. Structure and biological activity of protofibrillar intermediates

    J Biol Chem

    (1999)
  • F. Ferrone

    Analysis of protein aggregation kinetics

  • R.M. Murphy et al.

    Probing the kinetics of beta-amyloid self-association

    J Struct Biol

    (2000)
  • H. Inouye et al.

    A beta fibrillogenesis: kinetic parameters for fibril formation from Congo red binding

    J Struct Biol

    (2000)
  • J. Masel et al.

    Quantifying the kinetic parameters of prion replication

    Biophys Chem

    (1999)
  • L.M. Shaffer et al.

    Amyloid beta protein (A beta) removal by neuroglial cells in culture

    Neurobiol Aging

    (1995)
  • S. Inoue et al.

    Basement membranes, microfibrils and beta amyloid fibrillogenesis in Alzheimer's disease: high resolution ultrastructural findings

    Brain Res Brain Res Rev

    (1999)
  • A.K. Chamberlain et al.

    Ultrastructural organization of amyloid fibrils by atomic force microscopy

    Biophys J

    (2000)
  • R.E. Mrak et al.

    Interleukin-1, neuroinflammation, and Alzheimer's disease

    Neurobiol Aging

    (2001)
  • W.C. Eckelman

    Sensitivity of new radiopharmaceuticals

    Nucl Med Biol

    (1998)
  • B.R. Zeeberg et al.

    Analysis of three- and four-compartment models for in vivo radioligand–neuroreceptor interaction

    Bull Math Biol

    (1987)
  • H. Braak et al.

    Frequency of stages of Alzheimer-related lesions in different age categories

    Neurobiol Aging

    (1997)
  • J.L. Price et al.

    The distribution of tangles, plaques and related immunohistochemical markers in healthy aging and Alzheimer's disease

    Neurobiol Aging

    (1991)
  • B.J. Cummings et al.

    Image analysis of beta-amyloid load in Alzheimer's disease and relation to dementia severity

    Lancet

    (1995)
  • J. Higaki et al.

    Inhibition of beta-amyloid formation identifies proteolytic precursors and subcellular site of catabolism

    Neuron

    (1995)
  • D.H. Silverman et al.

    Clinical value of neuroimaging in the diagnosis of dementia. Sensitivity and specificity of regional cerebral metabolic and other parameters for early identification of Alzheimer's disease

    Clin Positron Imaging

    (1999)
  • A. Cagnin et al.

    In-vivo measurement of activated microglia in dementia

    Lancet

    (2001)
  • A. Cagnin et al.

    In vivo imaging of neuroinflammation

    Eur Neuropsychopharmacol

    (2002)
  • V.M. Lee et al.

    Neurodegenerative tauopathies

    Annu Rev Neurosci

    (2001)
  • E. Braak et al.

    Neuropathology of Alzheimer's disease: what is new since A. Alzheimer?

    Eur Arch Psychiatry Clin Neurosci

    (1999)
  • W.E. Klunk et al.

    Imaging Abeta plaques in living transgenic mice with multiphoton microscopy and methoxy-X04, a systemically administered Congo red derivative

    J Neuropathol Exp Neurol

    (2002)
  • W.E. Klunk et al.

    Imaging brain amyloid in Alzheimer's disease with Pittsburgh Compound-B

    Ann Neurol

    (2004)
  • L. Cai et al.

    Synthesis and evaluation of two 18F-labeled 6-iodo-2-(4′-N,N-dimethylamino)phenylimidazo[1,2-a]pyridine derivatives as prospective radioligands for beta-amyloid in Alzheimer's disease

    J Med Chem

    (2004)
  • M. Sadowski et al.

    Targeting prion amyloid deposits in vivo

    J Neuropathol Exp Neurol

    (2004)
  • M. Bresjanac et al.

    Molecular-imaging probe 2-(1-[6-[(2-fluoroethyl)(methyl) amino]-2-naphthyl]ethylidene) malononitrile labels prion plaques in vitro

    J Neurosci

    (2003)
  • C.A. Wilson et al.

    Intracellular APP processing and A beta production in Alzheimer disease

    J Neuropathol Exp Neurol

    (1999)
  • V. Blanquet et al.

    The beta amyloid protein (AD-AP) cDNA hybridizes in normal and Alzheimer individuals near the interface of 21q21 and q22.1

    Ann Genet

    (1987)
  • J. Kang et al.

    The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor

    Nature

    (1987)
  • E. Zerovnik

    Amyloid-fibril formation. Proposed mechanisms and relevance to conformational disease

    Eur J Biochem

    (2002)
  • T.R. Serio et al.

    Nucleated conformational conversion and the replication of conformational information by a prion determinant

    Science

    (2000)
  • J.D. Harper et al.

    Models of amyloid seeding in Alzheimer's disease and scrapie: mechanistic truths and physiological consequences of the time-dependent solubility of amyloid proteins

    Annu Rev Biochem

    (1997)
  • Cited by (37)

    • Determination of critical nucleation number for a single nucleation amyloid-β aggregation model

      2016, Mathematical Biosciences
      Citation Excerpt :

      Secondly, within the narrow n* range determined by numerical analyses, the value of 12 shows good correlation with the experimental values. Shoghi-Jadid and co-workers approximately determined the n* value for Aβ to be 6 [38]. It is also noteworthy that Saric and co-workers elegantly demonstrated the significance of non-specific interactions during the pre-nucleation phase [11].

    • GMP-compliant automated synthesis of [<sup>18</sup>F]AV-45 (Florbetapir F 18) for imaging Β-amyloid plaques in human brain

      2010, Applied Radiation and Isotopes
      Citation Excerpt :

      Positron emission tomography (PET), a common nuclear imaging modality, can provide functional information on the pathophysiological processes of AD. Recently, successful launches of 18F- or 11C-labeled PET tracers for Aβ plaque imaging such as PIB (Klunk et al., 2004; Mathis et al., 2003), SB-13 (Ono et al., 2003; Verhoeff et al., 2004), FDDNP (Small, 2006; Shoghi-Jadid et al., 2005), AV-1 (BAY94-9172) (Rowe et al., 2008), AV-45 (Florbetapir) (Choi et al., 2009; Carpenter et al., 2009), Flutemetamol (Koole et al., 2009; Nelissen et al., 2009), BF-227 (Kudo et al., 2007) and AZD2184 (Nyberg et al., 2009) have demonstrated the potential usefulness of in vivo imaging. Encouraging imaging results for noninvasive assessment of Aβ deposition were obtained with [11C]PIB in patients with AD or mild cognitive impairment (Lopresti et al., 2005; Forsberg et al., 2008).

    • Surface-enhanced nucleation of insulin amyloid fibrillation

      2008, Biochemical and Biophysical Research Communications
      Citation Excerpt :

      When insulin solutions were incubated at pH 1.6 and 65 °C, the absorbance at 600 nm followed a characteristic sigmoidal curve, i.e., an initial lag phase, a subsequent growth phase and a final saturation phase (Fig. 1A). The curve is consistent with the nucleation-dependent elongation model [14,15]. The remaining soluble protein declined following an inverse sigmoidal curve as shown in Fig. 1B.

    View all citing articles on Scopus
    View full text