Báo cáo khoa học: Stability and fibril formation properties of human and fish transthyretin, and of the Escherichia coli transthyretin-related protein potx

Stability and fibril formation properties of human and
fish transthyretin, and of the Escherichia coli
transthyretin-related protein
Erik Lundberg1, Anders Olofsson2, Gunilla T. Westermark3 and A. Elisabeth Sauer-Eriksson1
˚
1 Department of Chemistry, Umea University, Sweden
˚
2 Department of Medical Biochemistry and Biophysics, Umea University, Sweden
3 Division of Cell Biology, Diabetes Research Centre, Linkoping University, Sweden
ă

Keywords
amyloid; fibril formation; HIU hydrolase;
transthyretin; transthyretin-related protein
Correspondence
A. E. Sauer-Eriksson, Department of
˚
Chemistry, Umea University, SE-90187
˚
Umea, Sweden
Fax: +46 90 7865944
Tel: +46 90 7865923
E-mail: elisabeth.sauer-eriksson@chem.
umu.se
(Received 7 November 2008, revised 20
January 2009, accepted 26 January 2009)
doi:10.1111/j.1742-4658.2009.06936.x

Human transthyretin (hTTR) is one of several proteins known to cause
amyloid disease. Conformational changes in its native structure result in
aggregation of the protein, leading to insoluble amyloid fibrils. The transthyretin (TTR)-related proteins comprise a protein family of 5-hydroxyisourate hydrolases with structural similarity to TTR. In this study, we tested

the amyloidogenic properties, if any, of sea bream TTR (sbTTR) and
Escherichia coli transthyretin-related protein (ecTRP), which share 52%
and 30% sequence identity, respectively, with hTTR. We obtained filamentous structures from all three proteins under various conditions, but, interestingly, different structures displayed different tinctorial properties. hTTR
and sbTTR formed thin, curved fibrils at low pH (pH 2–3) that bound
thioflavin-T (thioflavin-T-positive) but did not stain with Congo Red (CR)
(CR-negative). Aggregates formed at the slightly higher pH of 4.0–5.5 had
different morphology, displaying predominantly amorphous structures.
CR-positive material of hTTR was found in this material, in agreement
with previous results. ecTRP remained soluble at pH 2–12 at ambient temperatures. By raising of the temperature, fibril formation could be induced
at neutral pH in all three proteins. Most of these temperature-induced
fibrils were thicker and straighter than the in vitro fibrils seen at low pH.
In other words, the temperature-induced fibrils were more similar to fibrils
seen in vivo. The melting temperature of ecTRP was 66.7 °C. This is
approximately 30 °C lower than the melting temperatures of sbTTR
and hTTR. Information from the crystal structures was used to identify
possible explanations for the reduced thermostability of ecTRP.

Transthyretin (TTR) is a homotetrameric plasma protein that binds and transports the thyroid hormones
3,5,3¢-triiodo-l-thyronine and 3,5,3¢,5¢-tetraiodo-l-thyronine (thyroxine) and retinol by binding to the retinolbinding protein when it is loaded with retinol [1]. TTR
is mainly expressed in the adult liver, the choroid
plexus of the brain, and the retina [2,3]. TTR is

involved in three amyloid diseases: familial amyloidotic
polyneuropathy, familial amyloidotic cardiomyopathy
(FAC), and senile systemic amyloidosis (SSA) [4,5].
Whereas SSA is associated with native TTR, point
mutations, of which more than 80 have been identified,
cause FAP and FAC [6]. TTR mutations associated
with familial amyloid diseases display a wide range of

Abbreviations
AFM, atomic force microscopy; BME, b-mercaptoethanol; CR, Congo Red; DSC, differential scanning calorimetry; ecTRP, Escherichia coli
transthyretin-related protein; EM, electron microscopy; FAC, familial amyloidotic cardiomyopathy; hTTR, human transthyretin; rTTR, rat
transthyretin; sbTTR, sea bream transthyretin; SSA, senile systemic amyloidosis; ThT, thioflavin-T; TLP, transthyretin-like protein; TRP,
transthyretin-related protein; TTR, transthyretin.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

1999


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

diversity in age of onset, penetrance, and tissues affected
[7,8]. SSA is a geriatric disease affecting approximately
25% of the European Caucasian population over
80 years of age [4]. Like FAC, SSA is characterized by
heavy deposits of amyloid fibrils in the heart.
Structures of TTRs from different species have been
studied [9], including human transthyretin (hTTR)
[10–12], rat TTR (rTTR) [13], chicken TTR [14], and sea
bream TTR (sbTTR) [15,16]. Within the TTR family,
fish TTR has the lowest sequence identity with hTTR
(e.g. sbTTR 52% [16,17], and lamprey TTR 47% [18]).
The transthyretin-related proteins (TRPs) comprise
a family of proteins recently shown to function as
5-hydroxyisourate hydrolases in the purine degradation
pathway [19–23]. These proteins are also referred to in

the literature as transthyretin-like proteins (TLPs).
However, to separate this family, whose members have
the characteristic sequence motif YRGS at their C-terminal end, from other protein sequences listed as
TTR-like, we prefer to refer to them as TRPs [19,24].
Sequence analysis of representative TTRs, TRPs and
TLPs suggests that the three protein groups are not
functionally related (Fig. 1).
The sequence identity between TRPs and TTRs is
relatively low; Escherichia coli TRP (ecTRP) shares
30% and 35% sequence identity with hTTR and
sbTTR, respectively (Fig. 1) [19]. Structures of TRP
from several species have been determined, and despite
their low sequence identity, the TTR and TRP structures were found to be very similar [24–27].
In TTR amyloidoses, the normally folded, secreted
protein cannot assemble into amyloid fibrils unless a
preceding partial unfolding event occurs [28]. Muta-

tions that destabilize the native structure of TTR are
known to lead to disease [29], but thermodynamic
stability alone does not reliably predict the severity of
the disease [30]. Instead, thermodynamic data,
combined with kinetic data, more reliably explain why
only some mutations lead to severe pathologies [31].
To understand the mechanism behind hTTR dissociation, misfolding, and amyloid formation, studies from
other species have provided valuable information. Like
hTTR, rTTR forms amyloid-like fibrils in vitro after
partial acid denaturation [32]. rTTR shares 85%
sequence identity with hTTR, which raises the question
of how important tertiary similarities are, as opposed
to sequence identity, for the ability of the protein to

form fibrils. In an attempt to answer this question, we
have investigated the fibril-forming properties of
sbTTR and ecTRP in vitro, and compared the results
with those of hTTR.
Our results showed that hTTR, sbTTR and ecTRP
can form fibrillar structures, but under different solution and temperature conditions. Furthermore,
depending on the conditions used, fibrils of different
morphology were obtained. Recent studies have shown
that sbTTR binds thioflavin-T (ThT) at low pH, suggestive of amyloid [33]. In our study, we verified that
sbTTR forms fibrillar structures at low pH that are
similar in shape to those of hTTR. We also found
that, even though  70% of the amino acids of ecTRP
are different from the respective amino acids in hTTR
and sbTTR, ecTRP has the ability to form Congo Red
(CR)-positive fibrils in vitro if the temperature is
increased sufficiently. Similar findings for ecTRP were
published while this work was in progress [34]. The

Fig. 1. Multiple sequence alignment of representatives of TRPs, TTRs, and TLPs. There are 57 gene clusters in Caenorhabditis elegans,
referred to as TTR-1 to TTR-57 in WORMBASE. Some of these sequences were originally identified as being structurally TTR-like by Sonnhammar & Durbin [68]. They seem to be functional and to influence aging in C. elegans, and are referred to as TRPs [69]. TTR-1 to TTR-57 are,
however, not related to the YRGS TRP family [19], which is why we prefer to refer to them as TLPs. The 57 TLP sequences present in
C. elegans are nematode-specific and share low sequence identity with each other; however, sequences TTR-18 to TTR-31 seem to
comprise a subgroup that is also found in other nematodes. (A) In this sequence alignment, TLP representatives from six nematodes have
been aligned with representatives from TTR and TRP. Identical and homologous residues within the three subgroups are marked in red and
pink, respectively. The first residue after signal sequence cleavage is highlighted in black. Sequence numbering refers to mature proteins.
Similarity is defined as amino acid substitutions within one of the following groups: FYW, IVLMF, RK, QDEN, GA, TS, or HNQ. Identical and
similar residues within the TRP family are shown in dark and light green, those within the TTR family in dark and light blue, and those within
the TLP subgroup in dark and light gray. The secondary structure elements are based on hTTR [12]. Residues lining the substrate-binding
channel in TTRs [13] and TRPs [26] are marked with blue stars. The assignment (*.:) shown below the sequences is directly from CLUSTALW2
[75] and refers to alignment of the six TLP sequences only. Only two amino acids (proline and glycine) are conserved throughout the three

groups. Whereas residues at the ligand-binding sites are almost completely conserved in the TTR and TRP families, they are not conserved
within the TLP group. In contrast, the most conserved regions are found at sites corresponding to b-strands B and E in TLPs, and to a-helix E
in TTRs and TRPs. (B) Phylogenetic tree of TRPs, TTRs, and TLPs. The tree was based on the multiple sequence alignment from (A). Caenorhabditis elegans TRP-R09H10.3 (GI:115532920); C. elegans TRP-ZK697.8 (GI:115534555); Mus Musculus TRP (GI:81916776); Bacillus subtilis
TRP (GI:3915561); Escherichia coli TRP (GI:3915454); Petaurus breviceps TTR (GI:1279636); Sminthopsis macroura TTR (GI:1279727); Gallus
gallus TTR (GI:45384444); Homo sapiens TTR (GI:114318993); Sparus aurata TTR (GI:6648602); Heterodera glycines TLP (GI:8571913); Radopholus similis TLP (GI:145279861); Brugia malayi TLP (GI:170583879); Xiphinema index TLP (GI:55724912); C. elegans TLP (TTR-30,
GI:25153261); Caenorhabditis briggsae TLP (TTR-18 GI:187037056).

2000

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

Stability and fibril formation of TTR and TRP

A

B

TLP--H. glycines
TLP--R. similis
TLP--B. malayi
TLP--X. index
TLP--C. elegans--TTR-30
TLP--C. briggsae--TTR-18
TTR--P. breviceps
TTR--S. macroura
TTR--G. gallus
TTR--H. sapiens

TTR--S. aurata
TRP--E. coli
TRP--C. elegans--R09H10.3
TRP--C. elegans--ZK697.8
TRP--M. musculus
TRP--B. subtilis

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2001


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

results emphasize the potential for amyloid formation
as a common property of all proteins, a feature that
can sometimes even bring new functionality [35–37].
The thermal stability of ecTRP was found by differential scanning calorimetry (DSC) to be approximately
30 °C lower than that of sbTTR and hTTR. Comparative studies of structures from homologous thermophiles
and mesophiles have revealed several factors that generally contribute to the intrinsic thermal stability of proteins [38–40]. These include tighter hydrophobic packing
of the protein core [41–43], increased electrostatic interactions on the surface of the protein [44,45], more prolines and alanines [46,47], and increased hydrogen
bonding of the polypeptide chain [48–50]. In addition,
improved intersubunit contacts within oligomeric proteins contribute to protein stability [51,52]. Here, we
analyze the structural basis for the differences in thermostability of hTTR, sbTTR, and ecTRP.

Results
Partial acid denaturation generates fibrils of hTTR
and sbTTR

Partial acid denaturation combined with turbidity assays
is a method frequently used to induce and monitor the
degree of protein aggregation and fibril formation of
hTTR [28,53]. Application of this method to the three
proteins in our study showed that hTTR displayed an
increase in turbidity at low pH, with a turbidity maximum at pH 4.5 (Fig. 2). This is in agreement with previous results [28,53]. For sbTTR, we measured only minor
turbidity increases at the low pH range of 4.0–5.5, and
for ecTRP, no effect was apparent over the entire pH
range (2–12) (Fig. 2). It should be noted that the pI of
ecTRP is estimated to be 8.2, in contrast to those of
hTTR and sbTTR, which are estimated to be 5.5–6.0.
The samples treated at low pH were visualized with
atomic force microscopy (AFM) to determine the

morphologies of the protein aggregates. Both hTTR and
sbTTR displayed fibrils at pH 2.0–3.5 that were very
similar in structure (Fig. 3). For hTTR, small amounts
of fibrillar structures were also formed at pH 4.0. The
thickness of these structures was found to be 1.3 nm,
which means that they were thinner than amyloid fibrils.
Their curved morphology agreed with previous observations of both hTTR and rTTR in vitro fibrils, suggesting
that they most likely represented protofibrils [32,54–56].
hTTR in vitro fibrils are reported to have widths varying
between 2.8 [55] and 10 nm [54], and the thicker in vitro
fibrils are believed to consist of up to five intertwined
protofilaments [54]. At the pH interval 4.0–5.5, predominantly amorphous aggregates, rather than fibrillar structures, were observed in the hTTR samples (Fig. 3A). It
is, however, not possible to quantitatively estimate the
ratio of aggregates to fibrillar structures from the AFM
images. The hTTR and sbTTR samples were also visualized with electron microscopy (EM). The EM images
were consistent with the morphologies that we elucidated from the AFM images, and verified that fibrillar

structures were present in the protein samples at pH 4.5
(Fig. 3B). To determine whether the lack of fibrillar
structures in the hTTR and sbTTR samples at pH 4.5
could be an effect of the technique used for analysis
(that is, the fibrils are unable to bind to mica gels at this
pH), fibril-containing samples of hTTR formed at
pH 2.0 were adjusted to pH 4.5 and incubated for various lengths of time. AFM images of this material
showed that the fibrils formed at pH 2.0 persisted at
pH 4.5, thereby verifying that these fibrillar forms can
bind to mica gels even at higher pH (data not shown).
Tinctorial properties of fibrillar structures formed
at low pH
The protein fibrils and aggregates obtained by the
partial acid denaturation experiments were tested for
ThT, which is a fluorescent dye commonly used to

Fig. 2. Turbidity assays for hTTR, sbTTR,
and ecTRP. The turbidity was measured at
330 nm after incubation of protein samples
at 37 °C for 72 h.

2002

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

Stability and fibril formation of TTR and TRP

A

B
a

b

a

b

C

Fig. 3. (A) AFM images of hTTR (left) and sbTTR (right). The samples were incubated at 37 °C for 72 h. Fibrils were present in samples incubated at pH 2.0–3.5, whereas aggregates were predominantly present in samples incubated at pH 4.5–5.5. No fibrils or aggregates were
detected in the ecTRP samples, at any of the pH intervals tested (pH 2.0–12.0; data not shown). The white scale bar is 500 nm. (B) EM
images of fibrils of hTTR (a) and sbTTR (b) incubated at pH 4.5. (C) CR staining of hTTR incubated for 3 days at pH 4.5. (a) shows fluorescence (at 594 nm) and (b) shows the characteristic apple-green birefringence with polarized light.

assess amyloid fibril formation [57,58], including hTTR
amyloid [59,60]. Pronounced emission at 482 nm, as a
result of ThT binding, was seen for the human and sea
bream samples incubated at pH 2.0 and 2.5, respectively (Fig. 4). This is in agreement with the presence
of the thin fibrils seen with AFM (Fig. 3A). At a pH
value around 3.0, both hTTR and sbTTR showed
significantly reduced ThT binding as compared to that
at lower pH. However, at the slightly higher pH range
of 3.5–4.5, hTTR formed structures that bound to
ThT (Fig. 4). This ThT-binding pattern correlates well
with the increase in turbidity of hTTR observed at
pH 4.5 (Fig. 2). The sbTTR samples did not display
increased ThT binding at pH 3.5–4.5, and demon-

strated only a very small increase in turbidity at this
pH range. ecTRP did not react with ThT at any pH
range tested.
The partially denaturated protein samples were also
tested for CR staining, visualized with EM. Generally,
the ability of fibrils to bind CR and to display a characteristic apple-green birefringence under polarized
light are the two most important criteria for detection
of amyloid fibrils in vivo. Such fibrils are called
CR-positive, but the specificity of this test has recently
been questioned [61,62]. The material from hTTR was
CR-positive at pH 4.0–4.5 (Fig. 3C), in agreement with
previous studies [28,53]. Aggregates from sbTTR, on
the other hand, were not found to be CR-positive at

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2003


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

To determine whether the thin and curved protofibrils formed at pH 2.0 by hTTR and sbTTR could be
induced to form amorphous aggregates or thicker filaments, the samples were readjusted to a pH of 4.5.
Interestingly, even after 6 weeks at 8 °C, these samples
contained the same type of protofibrils. This implies
that the low-pH-induced and ThT-binding fibrils are
not readily converted to CR-positive fibrillar aggregates or thicker filaments, or at least not under these

conditions.
Fiber formation induced by heating
Fig. 4. ThT-binding assays. Relative intensity of emission at
482 nm for samples incubated for 72 h at different pH. The excitation wavelength was 440 nm. The standard deviations between
triplet samples for pH intervals from 7.0 to 2.0 and between double
samples for pH intervals from 12.0 to 7.5 are shown. Strong ThT
binding is seen for hTTR and sbTTR at the lowest pH values, 2.0–
3.0. For hTTR, weaker ThT binding is also detected between
pH 3.5 and pH 4.5. As expected, neither hTTR nor sbTTR samples
incubated at pH 2.0 or 4.5 produce increased emission at 482 nm if
ThT is not added, which means that the increase in emission is
due to actual ThT binding. A local ThT-binding minimum is seen at
pH 3.0 for hTTR, even though fibrils are detected with AFM. The
reason for this is unknown. The fiber thickness at pH 3.0 generally
seems to be the same as for the fibers seen at lower pH. ecTRP
does not bind ThT at any pH level.

any pH range. Also, the large amounts of small fibrillar hTTR and sbTTR structures, observed with AFM
(Fig. 3A) at pH 2.0–3.0, did not stain with CR.
Apparently, ThT is more promiscuous than CR in
binding to thinner immature fibrillar structures of
TTR.
A

hTTR

B

0.2 µm

2004

sbTTR

Fibrillar structures of hTTR, sbTTR and ecTRP were
obtained by heating the protein samples for 72 h without stirring at different temperatures. Fibrils of hTTR
have previously been reported at 75 °C [63]. In this
study, we obtained thick fibrils at 55 °C for hTTR and
65 °C for sbTTR and ecTRP (Fig. 5A). Of these, only
fibrils from ecTRP showed a strong ThT response
(data not shown). These fibrils were verified to be
CR-positive (Fig. 5B).
Protein stability measured by SDS/PAGE and
DSC
The propensity for TTR amyloid formation is coupled
to tetramer dissociation. The stability of the tetrameric
structures of hTTR, sbTTR and ecTRP was analyzed
by gel electrophoresis according to the method of Lai
et al. [53] (Fig. 6). Unboiled samples from the acid
denaturation experiment were run on gel electrophoresis in the presence of SDS and b-mercaptoethanol
(b-ME). Whereas tetrameric hTTR dissociated at pH
ecTRP

Fig. 5. (A) AFM images of hTTR, sbTTR and
ecTRP heated at 55 °C (hTTR) or 65 °C
(sbTTR and ecTRP), respectively, for 72 h.
The fibril heights were estimated to be
 2.8 nm for hTTR,  3.5 nm for sbTTR,
and  4.0 nm for ecTRP. The white scale
bar is 500 nm. (B) Left: EM image of ecTRP

from the same sample as in (A). The material shows fluorescence (at 594 nm) (middle)
and apple-green birefringence (right) after
visualization with polarized light.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

Stability and fibril formation of TTR and TRP

Fig. 7. DSC profiles of hTTR, sbTTR, and ecTRP. The melting
temperatures (Tm) were 97.8 °C for hTTR, 93.0 °C for sbTTR,
and 66.7 °C for ecTRP.

Fig. 6. Analysis of tetramer stability by SDS ⁄ PAGE. The samples
were not boiled. The tetrameric structures of sbTTR and ecTRP
show increased stability at lower pH as compared to hTTR. Notably, both hTTR and sbTTR are unaffected by the SDS ⁄ BME treatment, and remain either in a monomeric or a tetrameric state,
depending on the pH of the protein buffer. In contrast, SDS ⁄ b-ME
treatment alone dissociates a fraction of the tetrameric ecTRP into
the monomeric state at all pH values.

values lower than 5.0, as shown previously [53], sbTTR
and ecTRP did not completely separate into their
monomeric units until the pH values were below 4.6
and 3.0, respectively (Fig. 6). In agreement with the
partial acid denaturation experiments, the ecTRP tetrameric structure seems to be more resistant to variations of pH. However, the SDS ⁄ b-ME treatment
generally dissociated a larger fraction of the ecTRP
protein material than of hTTR and sbTTR into monomeric structures. This behavior is likely to reflect
differences in the chemical composition of the proteins.

SDS is a detergent that readily dissolves hydrophobic
molecules, whereas acid denaturation affects electrostatic interactions rather than the hydrophobicity of
molecules. Gel filtration chromatography of ecTRP in
the presence and absence of 0.5% SDS in the running
buffer produced monomers only when the SDS was
included (data not shown).
The thermal stability of hTTR, sbTTR and ecTRP
was further studied by DSC (Fig. 7). The measured

thermal melting point (Tm) for hTTR was close to
values previously reported [64,65]. Even though tetrameric sbTTR seems to be more stable than hTTR at
lower pH, we found that it was less stable than hTTR
at physiological pH, with the Tm value for sbTTR
being approximately 4.5 °C lower than that for hTTR.
Gilthead sea bream is an ectotherm whose normal
habitat is the Mediterranean Sea. One feature that generally defines cold-adapted proteins and distinguishes
them from their mesophilic and thermophilic counterparts is their lower thermal stability [66]. This could
therefore be one explanation for the reduced stability
that we observed for sbTTR as compared to hTTR.
More unexpected, however, was the low thermostability of the ecTRP protein, with a Tm value approximately 26 °C below the values for both sbTTR and
hTTR (Fig. 7). The inability of ecTRP to form fibrils
at low pH can therefore not be directly correlated with
the thermostability of its tetrameric and monomeric
structures.
Thermostability and protein structures
We have analyzed the structures of hTTR (Protein
Data Bank code: 1F41 [12]), sbTTR (Protein Data
Bank code: 1SN2, [16]) and ecTRP (Protein Data
Bank code: 2G2N [24]) in an attempt to identify
factors that contribute to their profound differences in

stability. The results are summarized in Table 1.
Introduction of alanines, prolines and aromatic residues can contribute to protein stability [46,47,52].
hTTR and sbTTR have more alanine residues than
ecTRP, 12 versus eight and 13 versus eight, respectively, which might contribute to entropic stabilization.
On the other hand, ecTRP has more aromatic residues
than hTTR and sbTTR, 13 versus 12 and 13 versus

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2005


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

Table 1. Structural factors implicated in the thermostabilities of
hTTR, sbTTR, and ecTRP. Protein volume calculations were
obtained using VOIDOO [72], and hydrogen bonds were calculated
using WHATIF [73].
hTTR
Melting temperature (°C)
Amino acid composition
Ala
Phe, Tyr, Trp
Pro
Glu, Asp
Lys, Arg
Salt bridgesa
˚

Protein volume (A3)b
Monomer, Vm
Tetramer, Vt
DV = Vt ) 4Vm
Hydrogen bondsc
Monomer
Dimer

sbTTR

ecTRP

97.8

93.0

66.7

12
12
8
17
12
5

13
11
8
14
10

5

8
13
5
12
13
3

11 490
45 760
)200

11 340
45 190
)170

11 810
47 100
)140

75 (12)
165 (20)

70 (11)
163 (20)

64 (4)
141 (10)

a
˚
A distance less than or equal to 4 A between charged groups
defines an ion pair [74]. b The molecular volumes were calculated
˚
using a probe with radius 0 A, in order to obtain the protein volume
per se [52]. Before calculation, alternate conformations were
removed, and the structures were truncated at their N-termini and
C-termini, so that they structurally start and end at the same position. Residues included from the four chains, A, B, C, and D, of the
tetrameric structures were: hTTR, A10–A122, B10–B122, A¢11–
A¢122, and B¢10–B¢122; sbTTR, A10–A122, B10–B122, C11–C122,
and D10–D122; and ecTRP, A4–A114, B4–B114, C5–C114, and
D4–D114. c The numbers in parentheses are the numbers of buried
water molecules included in the calculation.

11, respectively. The backbone flexibility of the TTR
structures is probably also reduced because of a higher
proline content, which decreases the entropy of unfolding [46]. hTTR and sbTTR have eight proline residues
each, whereas ecTRP has only five.
The formation of salt bridges is another important
contributor to the temperature stability of proteins
[67]. In agreement with this, there is a higher number
of charged residues in the more thermostable TTR
protein structures (Table 1). There are five salt bridges
in hTTR and sbTTR, but only three in ecTRP.
Thermostable proteins are generally more tightly
packed than their less thermostable homologs [42]. The
structural homologs hTTR, sbTRP and ecTRP have
no internal cavities. However, hTTR and sbTRP have
two polar residues, Thr75 and His88 (hTTR numbering), buried within the hydrophobic core of the monomers. The side chains of these residues form, together

with the Ne1 atom of Trp79, hydrogen bonds with
three or four buried water molecules [12,16]. This
probably increases packing density and contributes to
stability [49]. In the ecTRP structure, two phenylala2006

nines occupy the same position as His88 and Trp79 in
hTTR, and consequently only one water molecule can
bind at this site in ecTRP [24].
The protein volumes of single subunits and tetrameric structures for the three proteins were investigated. Our calculations show that there is a decrease in
the protein volume of each monomer that is related to
an increase in thermal stability. Furthermore, there is
a clear correlation between thermostability and molecular volume occupied by the tetramers (Table 1). The
difference between the tetrameric volume and the
volume of the corresponding number of monomeric
units, DV, is negative in all cases, demonstrating that
the protein density increases slightly upon tetramerization.
Analysis of the number of hydrogen bonds formed
within monomers and dimers revealed pronounced differences between the three proteins. As previously
mentioned, the hTTR and sbTTR structures contain
buried polar residues and water molecules. These
waters allow the formation of 10 more hydrogen bonds
within their monomeric units, and about 20 more
hydrogen bonds within their dimeric units, than in the
ecTRP protein structure (Table 1). In addition, 14 and
15 hydrogen bonds are formed at the monomer–monomer interface of sbTTR and hTTR, whereas only eight
are formed in ecTRP. Five hydrogen bonds are formed
across the dimer–dimer interface of both hTTR and
sbTRP, whereas three are formed in ecTRP (Table 1).

Discussion

TTRs and TRPs are two protein families with similar
structures but different functions, due to divergent
evolution. The TTRs, found only in vertebrates, function as retinoic acid and thyroid hormone carriers,
whereas the TRPs, found predominantly in lower
eukaryotes and prokaryotes, are enzymes that hydrolyze
5-hydroxyisourate [19,21–23] in the purine catabolic
pathway. The structural similarity between several TTR
and TRP representatives has been verified by crystallographic studies [11,12,24–27]. Members from both
protein families have their active site positioned in the
hydrophobic channel formed at the dimer–dimer interface of their homotetrameric structure. The four amino
acid sequence motif YRGS at the C-terminal end of the
TRP sequences distinguishes them from other proteins
annotated as TTR-like or TTR-related [19]. These residues are involved in binding to substrate analogs [26].
Other proteins with sequence homology to TRPs and
TTRs exist, and we refer to these as TLPs. TLPs have so
far been found only in nematodes. Some of these were
identified as being structurally TTR-like in 1997 by

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

Sonnhammar & Durbin [68]. The function of the TLPs
is not known, but studies suggest that they are involved
in the regulation of the nematode lifespan [69]. From
their sequences, however, it is clear that their function
must be completely different from that of TTRs and
TRPs (Fig. 1).
In the current study, the fibril-forming and amyloidforming properties of three representatives from the

TTR and TRP families were investigated. Amyloidogenic properties of proteins are linked to their stability
and dissociation kinetics [31]. Therefore, the stabilities
of hTTR, sbTTR and ecTRP were analyzed by partial
acid and thermal denaturation. Some differences were
apparent. Analysis with SDS ⁄ PAGE showed that hTTR
is less stable than the other proteins under acidic conditions, and dissociates into monomers when the pH falls
below 5.0. sbTTR shows similar behavior, and is only
marginally less sensitive to acidic pH, dissociating at a
pH below 4.5. Interestingly, we found that ecTRP maintains its tetrameric structure even at very low pH values.
Different results were obtained when the protein
samples were analyzed with DSC. The melting point for
ecTRP was determined to be 66.7 °C, which is more
than 26 °C lower than those of both hTTR and sbTTR
(Fig. 7). Therefore, whereas previous SDS ⁄ PAGE
analysis suggested that the tetrameric structure of
ecTRP is more stable than that of hTTR [34], the DSC
results showed that ecTRP is significantly less thermostable than either hTTR or sbTTR.
Comparison of the crystal structures of hTTR,
sbTTR and ecTRP highlights a number of structural
differences that are consistent with the current explanations of thermal stability in proteins. Noteworthy is the
reduced number of negatively charged residues in the
ecTRP structure. This could possibly also explain its
structural stability at low pH. Furthermore, the thermostable TTR proteins have more hydrogen bonds and
ion pairs, and their structures are more densely packed
than that of ecTRP. Thus, it seems that the differences
in thermostability are mainly due to the presence of
specific polar and charged residues in hTTR and
sbTTR, which form additional hydrogen bonds that
stabilize their monomeric subunits as well as their
monomer–monomer and dimer–dimer interfaces.

In agreement with previous reports, TRP is amyloidogenic; fibrils do form upon heating of the protein
sample ([34] and this study). Whereas the previous study
reported an amyloid-inducing protocol for TRP that
involves heating at 24 °C at pH 5.8 with stirring, our
protocol involves heating at 65 °C at pH 7.4 for 3 days
without stirring. This temperature was chosen because
it was shown by DSC to be the Tm of the protein. These
fibrils are amyloidogenic, as determined from their

Stability and fibril formation of TTR and TRP

tinctorial properties in the ThT-binding and CR-binding assays. hTTR and sbTTR could also be induced to
form fibrils by heating (Fig. 5). Generally, fibrils
obtained from the heating procedure were thicker for
all proteins, and exhibited a straighter morphology than
that of the short and curved fibrils formed at low pH by
sbTTR and hTTR. We did not detect any fibrous material in samples incubated at 24 and 37 °C at pH 5.8,
with or without stirring. However, the discrepancy with
the previously described protocol [34] could be due to
subtle changes in sample preparation or experimental
procedure, which can drastically impact on fibril
formation. Interestingly, misfolded and aggregated
ecTRP material has been shown to be toxic for neuroblastoma cells, although the soluble protein is not [34].
In conclusion, sbTTR has properties similar to those
of hTTR in terms of tetramer dissociation and fibril
formation. Generally, where fibrils were observed for
hTTR, fibrils of similar morphology were observed
also for sbTTR, after some minor adjustments of the
fibrillization protocol. We did not detect any CR-positive fibrils of sbTTR at pH 4.5–5.5. This suggests that
hTTR forms amyloid fibrils by partial acid dissociation

more readily than sbTTR. The result does not exclude
the possibility that sbTTR can form CR-positive fibrils
at low pH, but more samples need to be examined, or
the concentration of protein needs to be increased.
ecTRP shares 30% sequence identity with hTTR. In
agreement with previous reports [34], we detected
CR-positive fibrils of ecTRP induced by heating. These
fibrils are similar both in shape and in dimension to
the fibrils of hTTR and sbTTR formed by heating.
The thick and straight morphology of heat-induced
fibrils of hTTR, sbTTR and ecTRP is similar to that
of amyloid fibrils in vivo. We have so far not been able
to convert thin and ThT-positive protofibrils of hTTR
and sbTTR, formed at low pH, to thicker and
CR-positive structures, suggesting that the kinetics are
very slow. This suggests that the TTR amyloid architecture is not the result of only one highly stringent
assembly of structures.
In the past, the propensity of proteins to form fibrillar
structures has most often been associated with disease.
Recently, however, examples have been presented where
conformational changes and fibril formation are
associated with an advantageous gain of function [37]. It
is not clear whether the fibril formation properties of the
TRP family are associated with any gain of function or
whether they even have any biological implications
whatsoever. In vivo fibrils have only been reported with
hTTR and rTTR, and it would be interesting to know
whether TTR from other species, as well as members of
the TRP family, can form fibrils in vivo.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2007


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

all samples were thoroughly vortexed to disperse aggregated
material before analysis by absorbance measurements at
330 nm in a standard UV cell.

Experimental procedures
Protein expression and purification
hTTR was expressed as previously described [70]. In brief,
competent E. coli BL21 cells were transformed with the
pET3a vector containing the hTTR construct, and plated
onto LB agar plates containing 50 lgỈmL)1 carbenicillin.
Bacteria were grown in LB medium supplemented with
50 lgỈmL)1 carbenicillin at 37 °C. At A600 nm = 0.4, cells
were induced with 0.2 mm isopropyl thio-b-d-galactoside
for 3 h, harvested by centrifugation for 20 minutes at
2800 g, and stored at )20 °C. The sbTTR gene placed in a
pET24d vector [16] was expressed using a similar protocol
as for hTTR, but with 50 lgỈmL)1 kanamycin as the antibiotic. After induction with isopropyl thio-b-d-galactoside,
the cells were grown overnight at 30 °C.
Similar purification protocols were used for hTTR and
sbTTR. Frozen cells were thawed in 20 mm Tris ⁄ HCl
(pH 8.0) and 50 mm NaCl, and lysed by sonication in the

presence of DNase I. Cell debris was removed by ultracentrifugation (120 000 g for 40 min) at 4 °C. The lysate was
filtered through a 0.2 lm syringe filter (Millipore Corporation, Bedford, MA, USA), and purified on a Q-Sepharose
Fast Flow anion exchange column (GE Healthcare,
Uppsala, Sweden) equilibrated with 20 mm Tris ⁄ HCl
(pH 8.0) and 50 mm NaCl, and eluted with a linear gradient
(0.1–1 m NaCl in 20 mm Tris ⁄ HCl, pH 8.0). TTR fractions
were pooled and concentrated (Centriprep-10; Amicon
Inc., Beverly, MA, USA), and then further purified by gel filtration on a HiLoad 16 ⁄ 60 Superdex-75 (GE Healthcare)
column with buffer containing 20 mm Tris ⁄ HCl (pH 6.8)
and 50 mm NaCl. Pure TTR fractions were pooled, concentrated to 5 mgỈmL)1 (Centriprep-3; Amicon), and stored at
)20 °C. ecTRP was cloned, expressed and purified as
previously described [19], using 50 mm Tris ⁄ HCl (pH 7.0)
and 200 mm NaCl as buffer in the final gel filtration step.
The pure ecTRP fractions were pooled and stored at )20 °C.

Partial acid denaturation
Denaturation studies were performed according to a previously described protocol for hTTR [53]. hTTR, sbTTR and
ecTRP were dialyzed against 2 mm NaH2PO4 ⁄ Na2HPO4
(pH 7.4) and 20 mm NaCl, and mixed to a final concentration of 0.2 mgỈmL)1 ( 3.5 lm tetramer), corresponding to
the TTR concentration in human plasma, at pH 2.0–12.0,
at intervals of 0.5 pH units. The buffers used gave a final
concentration of 50 mm glycine–HCl (pH 2.0–2.5), or
50 mm sodium acetate (pH 3.0–5.5), or 50 mm sodium
phosphate (pH 6.0–8.0), or 50 mm Hepes (pH 8.5–9.0), or
50 mm CAPSO (pH 9.5–10.0), or 50 mm CAPS (pH 10.5–
12.0). All buffers included 100 mm potassium chloride,
1 mm EDTA, and 1 mm dithiothreitol. After 72 h at 37 °C,

2008

Visualization
AFM
Following the turbidity measurements, the protein samples
were examined by AFM. Samples were diluted 10-fold with
H2O, and then 5 lL of the diluted sample solutions was
applied to freshly cleaved ruby red mica (Goodfellow,
Cambridge, UK). The material was allowed to adsorb
for 10 s, washed three times with 100 lL of distilled
water, and air dried. The surface was analyzed with
a Nanoscope IIIa multimode atomic force microscope
(Digital Instruments, Santa Barbara, CA, USA), using
Tapping Mode in air. A silicone probe was oscillated at
around 270 kHz, and images were collected at an optimized scan rate corresponding to 1–4 Hz. The images were
flattened and presented in height mode using nanoscope
software (Digital Instruments).

EM
Negative staining for EM was performed on the same samples used in the AFM studies. For this purpose, the material
was centrifuged at 16 000 g for 30 min, after which most of
the supernatant was removed and 200 lL of distilled water
was added. The material was vortexed, and aliquots of
3–5 lL were applied to Formvar-coated copper grids.
Contrast was achieved with 2% uranyl acetate in 50% ethanol, and the material was studied at 100 kV in a Jeol 1230
electron microscope (Jeol, Akishima, Tokyo, Japan).

CR-binding studies
For analysis with CR, 1–2 lL of diluted, vortexed samples
were applied to microscope slides and air dried. CR staining was performed according to Puchtler et al. [71], and
examined by light microscopy. The presence of amyloid
was verified by the green birefringence in polarized light

and with red fluorescence in a microscope equipped with
filters for wavelengths at 560 nm (excitation) and 590 nm
(emission).

ThT-binding studies
Protein samples incubated at 37 °C for 72 h were vortexed,
and 25 lL aliquots were mixed with 173 lL of a buffer
containing 100 mm sodium phosphate, 100 mm potassium
chloride (pH 7.6), and 2 lL of ThT stock solution (1 mm
ThT in 10 mm sodium phosphate, pH 7.4). The samples
were excited at 440 nm, and the emission at 482 nm was
recorded.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

Tetramer stability
SDS/PAGE
hTTR protein, sbTTR protein and ecTRP that had been
subjected to partial acid denaturation at various pH values
were analyzed by SDS ⁄ PAGE on 8–25% gradient polyacrylamide gels, using the Phast system (GE Healthcare),
and stained with Coomassie Brilliant Blue. The loading
buffer included 2.5% SDS, 5% b-ME, and 0.01% bromophenol blue.

DSC
Protein samples at a concentration of 1.5 mgỈmL)1
( 25 lm tetramer) of hTTR, sbTTR and ecTRP were
dialyzed against NaCl ⁄ Pi. Prior to DSC experiments, the

samples were vacuum degassed for 15 min at room temperature. DSC measurements were performed on a VP-DSC
calorimeter (MicroCal, Inc., Northampton, MA, USA) at a
heating rate of 10 °CỈh)1 from 30 to 100 °C. NaCl ⁄ Pi was
used as a control for these experiments.

Acknowledgements
The authors wish to thank Uwe H. Sauer and Tobias
Hainzl for valuable discussions, and Terese Bergfors
for critical reading of the manuscript. This work was
supported by grants from the Swedish Research Council, the FAMY ⁄ AMYL patients’ association, the
Kempe Foundation, and the Gustafsson Foundation.

References
1 Schreiber G & Richardson SJ (1997) The evolution of
gene expression, structure and function of transthyretin.
Comp Biochem Physiol B Biochem Mol Biol 116, 137–
160.
2 Cohen AS & Jones LA (1991) Amyloidosis. Curr Opin
Rheumatol 3, 125–138.
3 Benson MD & Kincaid JC (2007) The molecular biology and clinical features of amyloid neuropathy. Muscle
Nerve 36, 411–423.
4 Westermark P, Sletten K, Johansson B & Cornwell GG
III (1990) Fibril in senile systemic amyloidosis is derived
from normal transthyretin. Proc Natl Acad Sci USA 87,
2843–2845.
5 Saraiva MJ (1995) Transthyretin mutations in health
and disease. Hum Mutat 5, 191–196.
6 Connors LH, Lim A, Prokaeva T, Roskens VA &
Costello CE (2003) Tabulation of human transthyretin
(TTR) variants, 2003. Amyloid 10, 160–184.

7 Jacobson DR, Pastore R, Pool S, Malendowicz S, Kane
I, Shivji A, Embury SH, Ballas SK & Buxbaum JN
(1996) Revised transthyretin Ile 122 allele frequency in
African-Americans. Hum Genet 98, 236–238.

Stability and fibril formation of TTR and TRP

8 Afolabi I, Hamidi Asl K, Nakamura M, Jacobs P,
Hendrie H & Benson MD (2000) Transthyretin isoleucine-122 mutation in African and American blacks.
Amyloid 7, 121–125.
9 Hamilton JA & Benson MD (2001) Transthyretin: a
review from a structural perspective. Cell Mol Life Sci
58, 1491–1521.
10 Blake CC, Geisow MJ, Oatley SJ, Rerat B & Rerat C
(1978) Structure of prealbumin: secondary, tertiary and
quaternary interactions determined by Fourier refinement at 1.8 A. J Mol Biol 121, 339–356.
11 Hamilton JA, Steinrauf LK, Braden BC, Liepnieks J,
Benson MD, Holmgren G, Sandgren O & Steen L
(1993) The x-ray crystal structure refinements of normal
human transthyretin and the amyloidogenic Val30 fi Met variant to 1.7-A resolution. J Biol Chem 268,
2416–2424.
12 Hornberg A, Eneqvist T, Olofsson A, Lundgren E &
Sauer-Eriksson AE (2000) A comparative analysis of 23
structures of the amyloidogenic protein transthyretin.
J Mol Biol 302, 649–669.
13 Wojtczak A, Cody V, Luft JR & Pangborn W (2001)
Structure of rat transthyretin (rTTR) complex with thyroxine at 2.5 A resolution: first non-biased insight into
thyroxine binding reveals different hormone orientation
in two binding sites. Acta Crystallogr D Biol Crystallogr
57, 1061–1070.

14 Sunde M, Richardson SJ, Chang L, Pettersson TM,
Schreiber G & Blake CC (1996) The crystal structure of
transthyretin from chicken. Eur J Biochem 236, 491–499.
15 Folli C, Pasquato N, Ramazzina I, Battistutta R,
Zanotti G & Berni R (2003) Distinctive binding and
structural properties of piscine transthyretin. FEBS Lett
555, 279–284.
16 Eneqvist T, Lundberg E, Karlsson A, Huang S, Santos
CR, Power DM & Sauer-Eriksson AE (2004) High
resolution crystal structures of piscine transthyretin
reveal different binding modes for triiodothyronine and
thyroxine. J Biol Chem 279, 26411–26416.
17 Funkenstein B, Perrot V & Brown CL (1999) Cloning
of putative piscine (Sparus aurata) transthyretin:
developmental expression and tissue distribution. Mol
Cell Endocrinol 157, 67–73.
18 Manzon RG, Neuls TM & Manzon LA (2007)
Molecular cloning, tissue distribution, and developmental expression of lamprey transthyretins. Gen Comp
Endocrinol 151, 55–65.
19 Eneqvist T, Lundberg E, Nilsson L, Abagyan R &
Sauer-Eriksson AE (2003) The transthyretin-related
protein family. Eur J Biochem 270, 518–532.
20 Lee Y, Lee DH, Kho CW, Lee AY, Jang M, Cho S,
Lee CH, Lee JS, Myung PK, Park BC et al. (2005)
Transthyretin-related proteins function to facilitate the
hydrolysis of 5-hydroxyisourate, the end product of the
uricase reaction. FEBS Lett 579, 4769–4774.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2009


Stability and fibril formation of TTR and TRP

E. Lundberg et al.

21 Lee Y, Park BC, Lee do H, Bae KH, Cho S, Lee CH,
Lee JS, Myung PK & Park SG (2006) Mouse transthyretin-related protein is a hydrolase which degrades 5-hydroxyisourate, the end product of the uricase reaction.
Mol Cells 22, 141–145.
22 Hennebry SC, Wright HM, Likic VA & Richardson SJ
(2006) Structural and functional evolution of transthyretin and transthyretin-like proteins. Proteins 64, 1024–
1045.
23 Ramazzina I, Folli C, Secchi A, Berni R & Percudani R
(2006) Completing the uric acid degradation pathway
through phylogenetic comparison of whole genomes.
Nat Chem Biol 2, 144–148.
24 Lundberg E, Backstrom S, Sauer UH & Sauer-Eriksson
AE (2006) The transthyretin-related protein: structural
investigation of a novel protein family. J Struct Biol
155, 445–457.
25 Hennebry SC, Law RH, Richardson SJ, Buckle AM &
Whisstock JC (2006) The crystal structure of the
transthyretin-like protein from Salmonella dublin, a
prokaryote 5-hydroxyisourate hydrolase. J Mol Biol
359, 1389–1399.
26 Jung DK, Lee Y, Park SG, Park BC, Kim GH & Rhee
S (2006) Structural and functional analysis of PucM, a
hydrolase in the ureide pathway and a member of the
transthyretin-related protein family. Proc Natl Acad Sci

USA 103, 9790–9795.
27 Zanotti G, Cendron L, Ramazzina I, Folli C, Percudani
R & Berni R (2006) Structure of zebra fish HIUase:
insights into evolution of an enzyme to a hormone
transporter. J Mol Biol 363, 1–9.
28 Colon W & Kelly JW (1992) Partial denaturation of
transthyretin is sufficient for amyloid fibril formation
in vitro. Biochemistry 31, 8654–8660.
29 McCutchen SL, Lai Z, Miroy GJ, Kelly JW & Colon
W (1995) Comparison of lethal and nonlethal transthyretin variants and their relationship to amyloid disease.
Biochemistry 34, 13527–13536.
30 Jiang X, Buxbaum JN & Kelly JW (2001) The V122I
cardiomyopathy variant of transthyretin increases the
velocity of rate-limiting tetramer dissociation, resulting
in accelerated amyloidosis. Proc Natl Acad Sci USA 98,
14943–14948.
31 Hammarstrom P, Jiang X, Hurshman AR, Powers ET
& Kelly JW (2002) Sequence-dependent denaturation
energetics: a major determinant in amyloid disease
diversity. Proc Natl Acad Sci USA 99(Suppl. 4), 16427–
16432.
32 Tajiri T, Ando Y, Hata K, Kamide K, Hashimoto M,
Nakamura M, Terazaki H, Yamashita T, Kai H, Haraoka K et al. (2002) Amyloid formation in rat transthyretin: effect of oxidative stress. Clin Chim Acta 323,
129–137.
33 Morgado I, Melo EP, Lundberg E, Estrela NL,
Sauer-Eriksson AE & Power DM (2008) Hormone

2010

34

35

36

37

38

39
40

41

42

43

44

45

46

47

48

affinity and fibril formation of piscine transthyretin: the
role of the N-terminal. Mol Cell Endocrinol 295(1-2),

48–58.
Santos SD, Costa R, Teixeira PF, Gottesman M,
Cardoso I & Saraiva MJ (2008) Amyloidogenic
properties of transthyretin-like protein (TLP) from
Escherichia coli. FEBS Lett 582, 2893–2898.
Guijarro JI, Sunde M, Jones JA, Campbell ID &
Dobson CM (1998) Amyloid fibril formation by an SH3
domain. Proc Natl Acad Sci USA 95, 4224–4228.
Chiti F & Dobson CM (2006) Protein misfolding,
functional amyloid, and human disease. Annu Rev
Biochem 75, 333–366.
Fowler DM, Koulov AV, Balch WE & Kelly JW (2007)
Functional amyloid – from bacteria to humans. Trends
Biochem Sci 32, 217–224.
Fersht AR & Serrano L (1993) Principles of protein
stability derived from protein engineering experiments.
Curr Opin Struct Biol 3, 75–83.
Matthews BW (1995) Studies on protein stability with
T4 lysozyme. Adv Protein Chem 46, 249–278.
Jaenicke R (1996) Stability and folding of ultrastable
proteins: eye lens crystallins and enzymes from
thermophiles. FASEB J 10, 84–92.
Eriksson AE, Baase WA, Wozniak JA & Matthews
BW (1992) A cavity-containing mutant of T4
lysozyme is stabilized by buried benzene. Nature 355,
371–373.
Eriksson AE, Baase WA, Zhang XJ, Heinz DW, Blaber
M, Baldwin EP & Matthews BW (1992) Response of a
protein structure to cavity-creating mutations and its
relation to the hydrophobic effect. Science 255, 178–

183.
Chen J & Stites WE (2001) Packing is a key selection
factor in the evolution of protein hydrophobic cores.
Biochemistry 40, 15280–15289.
Serrano L, Horovitz A, Avron B, Bycroft M & Fersht
AR (1990) Estimating the contribution of engineered
surface electrostatic interactions to protein stability by
using double-mutant cycles. Biochemistry 29, 9343–
9352.
Vogt G, Woell S & Argos P (1997) Protein thermal
stability, hydrogen bonds, and ion pairs. J Mol Biol
269, 631–643.
Matthews BW, Nicholson H & Becktel WJ (1987)
Enhanced protein thermostability from site-directed
mutations that decrease the entropy of unfolding. Proc
Natl Acad Sci USA 84, 6663–6667.
Li C, Heatwole J, Soelaiman S & Shoham M (1999)
Crystal structure of a thermophilic alcohol dehydrogenase substrate complex suggests determinants of
substrate specificity and thermostability. Proteins 37,
619–627.
Wang M, Wales TE & Fitzgerald MC (2006) Conserved
thermodynamic contributions of backbone hydrogen

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS


E. Lundberg et al.

49

50

51

52

53

54

55

56

57

58

59

60

61

bonds in a protein fold. Proc Natl Acad Sci USA 103,
2600–2604.
Schell D, Tsai J, Scholtz JM & Pace CN (2006) Hydrogen bonding increases packing density in the protein
interior. Proteins 63, 278–282.
Bolen DW & Rose GD (2008) Structure and energetics
of the hydrogen-bonded backbone in protein folding.

Annu Rev Biochem 77, 339–362.
Korkhin Y, Kalb AJ, Peretz M, Bogin O, Burstein Y &
Frolow F (1999) Oligomeric integrity – the structural
key to thermal stability in bacterial alcohol dehydrogenases. Protein Sci 8, 1241–1249.
Dalhus B, Saarinen M, Sauer UH, Eklund P, Johansson
K, Karlsson A, Ramaswamy S, Bjork A, Synstad B,
Naterstad K et al. (2002) Structural basis for thermophilic protein stability: structures of thermophilic and
mesophilic malate dehydrogenases. J Mol Biol 318,
707–721.
Lai Z, Colon W & Kelly JW (1996) The acid-mediated
denaturation pathway of transthyretin yields a conformational intermediate that can self-assemble into amyloid. Biochemistry 35, 6470–6482.
Cardoso I, Goldsbury CS, Muller SA, Olivieri V, Wirtz
S, Damas AM, Aebi U & Saraiva MJ (2002) Transthyretin fibrillogenesis entails the assembly of monomers: a
molecular model for in vitro assembled transthyretin
amyloid-like fibrils. J Mol Biol 317, 683–695.
Olofsson A, Ippel JH, Wijmenga SS, Lundgren E &
Ohman A (2004) Probing solvent accessibility of transthyretin amyloid by solution NMR spectroscopy. J Biol
Chem 279, 5699–5707.
Lindgren M, Sorgjerd K & Hammarstrom P (2005)
Detection and characterization of aggregates, prefibrillar amyloidogenic oligomers, and protofibrils using fluorescence spectroscopy. Biophys J 88, 4200–4212.
Naiki H, Higuchi K, Hosokawa M & Takeda T (1989)
Fluorometric determination of amyloid fibrils in vitro
using the fluorescent dye, thioflavin T1. Anal Biochem
177, 244–249.
LeVine H III (1993) Thioflavine T interaction with synthetic Alzheimer’s disease beta-amyloid peptides: detection of amyloid aggregation in solution. Protein Sci 2,
404–410.
Lashuel HA, Wurth C, Woo L & Kelly JW (1999) The
most pathogenic transthyretin variant, L55P, forms
amyloid fibrils under acidic conditions and protofilaments under physiological conditions. Biochemistry 38,
13560–13573.

Hammarstrom P, Schneider F & Kelly JW (2001)
Trans-suppression of misfolding in an amyloid disease.
Science 293, 2459–2462.
Khurana R, Uversky VN, Nielsen L & Fink AL (2001)
Is Congo red an amyloid-specific dye? J Biol Chem 276,
22715–22721.

Stability and fibril formation of TTR and TRP

62 Bousset L, Redeker V, Decottignies P, Dubois S, Le
Marechal P & Melki R (2004) Structural characterization of the fibrillar form of the yeast Saccharomyces
cerevisiae prion Ure2p. Biochemistry 43, 5022–5032.
63 Chung CM, Connors LH, Benson MD & Walsh MT
(2001) Biophysical analysis of normal transthyretin:
implications for fibril formation in senile systemic amyloidosis. Amyloid 8, 75–83.
64 Shnyrov VL, Villar E, Zhadan GG, Sanchez-Ruiz JM,
Quintas A, Saraiva MJ & Brito RM (2000) Comparative calorimetric study of non-amyloidogenic and
amyloidogenic variants of the homotetrameric protein
transthyretin. Biophys Chem 88, 61–67.
65 Takeuchi M, Mizuguchi M, Kouno T, Shinohara Y,
Aizawa T, Demura M, Mori Y, Shinoda H &
Kawano K (2007) Destabilization of transthyretin by
pathogenic mutations in the DE loop. Proteins 66,
716–725.
66 Feller G & Gerday C (1997) Psychrophilic enzymes:
molecular basis of cold adaptation. Cell Mol Life Sci
53, 830–841.
67 Yip KS, Britton KL, Stillman TJ, Lebbink J, de Vos
WM, Robb FT, Vetriani C, Maeder D & Rice DW
(1998) Insights into the molecular basis of thermal

stability from the analysis of ion-pair networks in the
glutamate dehydrogenase family. Eur J Biochem 255,
336–346.
68 Sonnhammer EL & Durbin R (1997) Analysis of
protein domain families in Caenorhabditis elegans.
Genomics 46, 200–216.
69 Hansen M, Hsu AL, Dillin A & Kenyon C (2005) New
genes tied to endocrine, metabolic, and dietary regulation of lifespan from a Caenorhabditis elegans genomic
RNAi screen. PLoS Genet 1, 119–128.
70 Karlsson A, Olofsson A, Eneqvist T & Sauer-Eriksson
AE (2005) Cys114-linked dimers of transthyretin are
compatible with amyloid formation. Biochemistry 44,
13063–13070.
71 Puchtler H & Sweat F (1965) Congo red as a stain for
fluorescence microscopy of amyloid. J Histochem
Cytochem 13, 693–694.
72 Kleywegt GJ & Jones TA (1994) Detection, delineation,
measurement and display of cavities in macromolecular
structures. Acta Crystallogr D Biol Crystallogr 50, 178–
185.
73 Hooft RW, Sander C & Vriend G (1996) Positioning
hydrogen atoms by optimizing hydrogen-bond networks
in protein structures. Proteins 26, 363–376.
74 Barlow DJ & Thornton JM (1983) Ion-pairs in proteins.
J Mol Biol 168, 867–885.
75 Larkin MA, Blackshields G, Brown NP, Chenna R,
McGettigan PA, McWilliam H, Valentin F, Wallace
IM, Wilm A, Lopez R et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23, 2947–2948.

FEBS Journal 276 (2009) 1999–2011 ª 2009 The Authors Journal compilation ª 2009 FEBS

2011