Molecular medicine-The Formation of Amyloid Fibrils and its Pathological Consequences

The formation of amyloid fibrils plays a key
role in the origin of several neurodegenerative
pathologies, such as spongiform
encephalopathies and Alzheimer’s disease.
Historically, the term amyloid was
introduced to describe fibrillar protein
deposits associated with diseases known
as amyloidoses that involve the extracellular
deposition of amyloid fibrils and
plaques with the aspect of starch. For
many of these diseases, the major fibrillar
46 Aggregation, Protein
protein component has been identified.
In the 1970s, it was demonstrated that
lysosomal proteins under acidic conditions
could form amyloid fibrils. It was
generally accepted at this time that proteolysis
was the amyloidogenic determinant.
Twenty years later, it was shown that
purified transthyretin is converted into
amyloid fibrils via an acid-induced conformational
change in vitro, demonstrating
that conformational changes alone were
responsible for producing an intermediate
generating amyloid structure. These
aberrant protein self-assemblies are at
the origin of more than hundred human
amyloid diseases, some of them being
lethal.
Twenty unrelated protein precursors
are known to form amyloid fibrils,
among them transthyretin, lysozyme, immunoglobulin
light chain, β2 microglobulin,
Alzheimer Aβ1–40 and Aβ1–42
peptides, the mammalian prion protein,
and the yeast prion-like proteins (Table 1).
Since they are subjects of another chapter,
prion proteins will not be discussed
here. Although they have no homology
in sequence and structure, all form amyloid
fibrils with a similar overall structure,
suggesting a common self-assembly
Tab. 1 Amyloidogenic proteins and the corresponding diseases.
Clinical syndrome Precursor protein Fibril component
Alzeimer’s disease APP β-peptide 1–40 to 1–43
Primary systemic amyloidosis Immunoglobulin light chain Intact light chain or fragments
Secondary systemic amyloidosis Serum amyloid A Amyloid A (76-residue
fragment)
Senile systemic amyloidosis Transthyretin Transthyretin or fragments
Familial amyloid polyneuropathy I Transthyretin Over 45 transthyretin variants
Hereditary cerebral amyloid
angiopathy
Cystatin C Cystatin C minus 10 residues
Hemodialysis-related β2-microglobulin β2-microglobulin
amyloidosis Apolipoprotein A1 Fragments of Apolipoprotein
A1
Familial amyloid polyneuropathy III Gelsosin 71-amino acid fragment of
gelsosin
Finnish hereditory systemic
amyloidosis
Islet amyloid polypeptide
(IAPP)
Fragment of IAPP
Type II diabetes Calcitonin Fragments of calcitonin
Medullary carcinoma of the thyroid
Spongiform encephalopathies Prion Prion or fragments thereof
Atrial amyloidosis Atrial natriuretic factor
(ANF)
ANF
Hereditary nonneuropathic systemic
amyloidosis
Lysozyme Lysozyme or fragments
thereof
Injection-localized amyloidosis Insulin Insulin
Hereditary renal amyloidosis Fibrinogen Fibrinogen fragments
Parkinson disease α-synuclein∗
Source: (According to Kelly, J.W. (1996) Alternative conformations of amyloidogenic proteins govern
their behavior, Curr. Opin. Struct. Biol. 6, 11–17); ∗From J.C. Rochet & P.T. Lansbury (2000) Curr.
Opin. Struct. Biol. 10, 60–68.
Aggregation, Protein 47
pathway. In all proteins known to form
amyloid fibrils, there is a conversion of
α- to β-structure. Amyloid fibrils are abnormal,
insoluble, and generally proteaseresistant
structures. They were first recognized
by their staining properties. The
most commonly used method to detect
amyloid is staining by Congo Red,
which exhibits a green birefringence. Amyloid
fibrils are generally 60 to 100 A˚ in
diameter and of variable length. X-ray
diffraction data on fibrils, solid-state NMR
studies, cryoelectron microscopy, and infrared
Fourier transform experiments have
shown that amyloid fibrils aremade of two
or more β-sheet filaments wound around
one another. They have a characteristic
cross-β repeat structure, the individual β-
strands being oriented perpendicular to
the long axis of the fibril.
Recently, progress has been made in
the knowledge of the mechanisms involved
in the formation of amyloid fibrils.
Oligomeric prefibrillar intermediates have
been extensively characterized with respect
to their structure and temporal
evolution. A well-documented example is
provided by the studies on transthyretin.
The biological role of this protein is the
transport of thyroxin by direct binding
and the transport of retinol via the retinol
binding protein. The wild-type protein is
very stable at neutral pH. In certain individuals,
however, it is converted into
amyloid fibrils, and this is associated
with the disease, senile systemic amyloidosis.
Several variants are associated
with familial polyneuropathies. In vitro
biophysical studies have identified conditions
leading to amyloid formation. The
three-dimensional structure of the protein
is known. The wild-type protein is
a tetramer at pH ranging between 5 and 7;
the tetramer dissociates into a monomer
when the pH decreases. The dissociation
is the rate-limiting step of the process.
The monomer exhibits an altered tertiary
structure, which aggregates in amyloid
Native protein Amylogenic intermediate
Protofibrils
Fibrils
n
Unfolded protein
Fig. 10 Schematic representation of the formation of amyloid fibrils from a
partially folded intermediate.
48 Aggregation, Protein
protofilaments, and then forms amyloid
fibrils. This formation is at its maximumat
pH 4.4. Using deuterium–proton
exchange monitored by two-dimensional
NMR spectroscopy on transthyretin at pH
5.75 and 4.5, Liu et al. have shown a selective
destabilization of one half of the
β-sandwich structure of the protein, increasing
the mobility of this region. These
studies have identified the residues that
undergo increased conformational fluctuations
under amyloidogenic conditions.
The mutations in the pathological variants
responsible for familial amyloid polyneuropathies
are localized in this region. A
strategy to delay the formation of amyloid
fibrils proposed by Saccheti & Kelly was
to developmolecules capable of stabilizing
the tetramer.
Fig. 11 Molecular model of an amyloid
fibril derived from cryoelectron
microscopy analysis of fibrils grown
from an SH3 domain by incubation of a
solution containing the protein at low
pH (reproduced from Dobson, C. (1999)
TIBS 24, 331, with permission).
Two variants of human lysozyme,
Ile56Thr andAsp67His have been reported
to be amyloidogenic; they are responsible
for fatal amyloidoses. Pepys and colleagues
have determined the precise structures and
properties of these mutants. The native
fold of the two amyloidogenic variants,
as resolved by X-ray crystallography, is
similar to that of the wild-type protein.
Both variants are enzymatically active, but
have been shown to be unstable. The replacement
of an aspartate by a histidine
suppresses a hydrogen bond formed in
the wild-type protein with a tyrosine in a
neighboring β-strand. This rupture opens
a large gap between two β-strands. In
the other variant, the replacement of an
isoleucine by a threonine suppresses a van
der Waals contact with a neighboring helix.
Consequently, changes in the interface
between the α- and β-domains occur in
both variants, destabilizing the molecule.
The mutations leading to amyloid fibril
formation are observed to result in a
decreased stability of the native state. In
all cases, the formation of fibrils occurs
from a partially structured molecule via
nucleation-dependent oligomerization. It
was observed for several proteins that
fibrillation takes place only after a lag
phase, which is abolished upon seeding.
Nucleation is followed by the formation
of protofibrils whose characteristics have
been determined (Fig. 10). Atomic force
microscopy and fluorescence correlation
spectroscopy have been used to monitor
transitions among the different types
of assemblies.
Aggregation, Protein 49
Recent observations from Dobson and
his group have shown that several proteins
unrelated to amyloid diseases are
able to aggregate in vitro into amyloid
fibrils when exposed to mild denaturing
conditions. These fibrils are indistinguishable
from those found in pathological
conditions. It was demonstrated for different
proteins such as normal lysozyme,
an SH3 domain of a phosphatidyl inositol
protein kinase (Fig. 11), an acyl
phosphatase, and an α-helical protein,
myoglobin suggesting a common mechanism
for the formation of amyloid.
These findings clearly indicate that amyloid
formation is a general property of
polypeptide chains rather than one restricted
to definite sequences as occurs
with chameleon sequences capable of
adopting either a β- or an α-helicoidal
structure depending on their environment.
Furthermore, these aggregates exhibited
an inherent toxicity when incubated with
mouse fibroblasts. Several groups suggest
that oligomeric intermediates rather
than fibrils themselves are responsible for
pathogenicity.
Significant progress has been made in
understanding the mechanisms involved
in the formation of amyloid fibrils. This
is an important step in guiding research
into the discovery of molecules with
therapeutic efficiency.