Báo cáo khoa học: FH8 – a small EF-hand protein from Fasciola hepatica docx

FH8 – a small EF-hand protein from Fasciola hepatica
Hugo Fraga
1
, Tiago Q. Faria
2
, Filipe Pinto
1
, Agostinho Almeida
3
, Rui M. M. Brito
2,4
and
Ana M. Damas
1,5
1 IBMC, Institute for Molecular and Cell Biology, University of Porto, Portugal
2 Center for Neuroscience and Cell Biology, University of Coimbra, Portugal
3 REQUIMTE, Faculdade de Farma
´
cia, Departamento de Quı
´
mica-Fı
´
sica, University of Porto, Portugal
4 Chemistry Department, Faculty of Science and Technology, University of Coimbra, Portugal
5 ICBAS, Instituto de Cie
ˆ
ncias Biome
´
dicas de Abel Salazar, University of Porto, Portugal
Introduction
Fasciola hepatica is a trematode parasite that is respon-

sible for fascioliasis. Although traditionally regarded
as a parasite of livestock, resulting in a large economic
loss to the agricultural community, it remains, in sev-
eral countries, an important human parasite, and it is
estimated that 2.4 million people are infected with liver
fluke worldwide [1]. Infection occurs when the larvae
adhering to vegetation are ingested and become infec-
tive juveniles in the duodenum. Then, infection pro-
ceeds with the rapid penetration of the parasite into
the intestinal wall and their entry into the peritoneal
cavity, where they break through the liver capsule.
Keywords
calcium binding protein; fasciolasis; FH8;
Fasciola hepatica; sensor protein
Correspondence
A. M. Damas, IBMC, Institute for Molecular
and Cell Biology, University of Porto,
R. Campo Alegre 823, 4150-180 Porto,
Portugal
Fax: +351 226099157
Tel: +351 226074900
E-mail:
(Received 15 July 2010, revised
24 September 2010, accepted 11 October
2010)
doi:10.1111/j.1742-4658.2010.07912.x
Vaccine and drug development for fasciolasis rely on a thorough under-
standing of the mechanisms involved in parasite–host interactions. FH8 is
an 8 kDa protein secreted by the parasite Fasciola hepatica in the early
stages of infection. Sequence analysis revealed that FH8 has two EF-hand

Ca
2+
-binding motifs, and our experimental data show that the protein
binds Ca
2+
and that this induces conformational alterations, thus causing
it to behave like a sensor protein. Moreover, FH8 displays low affinity for
Ca
2+
(K
obs
=10
4
M
)1
) and is highly stable in its apo and Ca
2+
-loaded
states. Homology models were built for FH8 in both states. It has only one
globular domain, with two binding sites and appropriate groups in the
positions for coordination of the metal ions. However, an unusually high
content of positively charged amino acids in one of the binding sites, when
compared with the prototypical sensor proteins, potentially affects the
protein’s affinity for Ca
2+
. The only Cys present in FH8, conserved in the
homologous proteins of other helminth parasites, is located on the surface,
allowing the formation of dimers, detected on SDS gels. These findings
reflect specificities of FH8, which are most probably related to its roles
both in the parasite and in the host.

Structured digital abstract
l
MINT-8041757: F8 (uniprotkb:Q9NIG5) and F8 (uniprotkb:Q9NIG5) bind (MI:0407)by
affinity chromatography technology (
MI:0004)
l
MINT-8041770: FH8 (uniprotkb:Q9NIG5) and FH8 (uniprotkb:Q9NIG5) bind (MI:0408)by
cross-linking study (
MI:0030)
Abbreviations
ANS, 8-anilonaphthalene-1-sulfonate; BS3, suberic acid bis(3-sulfo-N-hydroxysuccinimide) ester; CaBP, calcium-binding protein;
CaM, calmodulin; DLS, dynamic light scattering; R
H,
hydrodynamic radius; RLU, relative luminescence units; TECP, triphenyl phosphine;
T
m
, melting point; TnC, troponin C.
5072 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
After 8–12 weeks within the liver, they move to the
bile ducts, where they mature and produce eggs [1].
Recently, there has been an increase in the number
of liver fluke infections of livestock in countries with
temperate climates, owing to weather changes that
support one of the intermediate hosts of the parasite
(Galba truncatula), and the emergence of strains that
are resistant to benzimidazole compounds, which are
widely used for the treatment of fascioliasis [2].
Further contributing to the spread of this disease is its
difficult diagnosis, which is often based on the detec-
tion of the worm or lesions in liver sections at the

slaughterhouse, hampering systematic diagnosis of the
disease in farm animals [1].
Excreted–secreted antigens have been shown to be
useful in the diagnosis of human fascioliasis. Besides
their importance in screening, the secreted proteins are
essential for pathogenesis, as they are involved in
several physiological processes of the parasite [3–6].
Indeed, F. hepatica secretes a large array of proteins
into the host, and transcriptomic and proteomic
approaches, during the different life stages of the para-
site, have been applied to investigate the importance of
these proteins for the parasite–host interaction [7]. In
one of those studies, an ORF corresponding to a 69
amino acid protein was isolated from a screen of an
F. hepatica cDNA bank [8]. This protein was called
FH8, because of its molecular mass of 8 kDa, and was
detected in the early stages of infection (1–3 weeks
postinfection) [8]. Immunofluorescence studies demon-
strated that it was present on the surface of the para-
site and in secreted fluids, probably resulting from the
shedding of the worm glycocalyx. As FH8 is expressed
on the surface of the parasite, during its cercarial
stage, it is a good candidate for vaccine and drug
development. Three proteins homologous to FH8 were
also described in the fluke parasites Schistosoma man-
soni (SM8) [9], Clonorchis sinensis (CH8) and Schisto-
soma japonicum (SJ8) [10,11]. Besides FH8, two
calmodulin (CaM)-like proteins from F. hepatica have
been identified [2]. One of them (FhCaM1) is highly
similar to mammalian CaM (98% identity), whereas

the other (FhCaM2) has only 41% identity. Both of
them bind Ca
2+
, and homology models have been
obtained [2].
An analysis of FH8 amino acid sequence revealed
the presence of two EF-hands, which are helix–loop–
helix structural motifs involved in Ca
2+
coordination.
The most common EF-hand motif, also called the
canonical EF-hand, is present in CaM and troponin C
(TnC), and contains a 12 amino acid binding loop
that provides most of the oxygens that coordinate
Ca
2+
. However, the composition and length of the
Ca
2+
-binding loops can vary among EF-hand proteins
[12,13].
EF-hand proteins are organized into structural
domains, containing two or more EF-hands, which form
highly stable helical bundles. The minimum functional
unit present in EF-hand calcium-binding proteins
(CaBPs) is a domain with two EF-hands, whose stability
is maintained by a small antiparallel b-sheet (EF-hand
b-scaffold), formed by two stretches of the Ca
2+
-bind-

ing loops. The two EF-hands are covalently bonded via
the B ⁄ C linker, which connects the exiting helix (B) of
the first EF-hand to the first helix (C) of the second
EF-hand. Despite the similarities in sequence and three-
dimensional structure of EF-hands, it is known that
CaBPs perform a diverse range of functions [13,14], and
they are normally classified into two groups: Ca
2+
sen-
sors, represented by CaM and TnC, and Ca
2+
buffers,
such as calbindin D
9k
and parvalbumin. The sensor
proteins display Ca
2+
-dependent conformational
changes, whereas Ca
2+
buffers, which are involved in
Ca
2+
signal modulation, undergo minimal structural
changes upon Ca
2+
binding. It was reported that when
Ca
2+
binds to sensor proteins, it triggers a switch from

a closed to an open conformation, owing to the reorien-
tation of the four helices of each functional domain,
exposing a hydrophobic region that acts as a target-
binding surface. The molecular and structural features
responsible for the differences between the two classes
of CaBP, although studied by several groups, are not
completely understood [13–16]. Some researchers refer
to the importance of the B ⁄ C linker in distinguishing
between the sensor and buffer proteins [13,15]. More-
over, it has been reported that CaM and TnC have the
shortest B ⁄ C linkers, and also that the N-terminal
domain of CaM is less hydrophobic than that of calbin-
din D
9k
[13,15].
FH8 is one of the smallest CaBPs described until
now, and this makes it a very particular case study
protein for the as yet unclear structure–function
relationships in the EF-hand family of proteins.
Here, we report the cloning, expression and initial
biochemical and structural characterization of FH8
from F. hepatica.
Results
FH8 cloning and purification
Because of the small size of the protein and its hypo-
thetical Ca
2+
-binding properties, the expression and
purification of recombinant FH8, with the use of
conventional affinity tags, was not appropriate. As an

alternative, a construct was prepared with the
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5073
IMPACT system [17]. This system uses the inducible
self-cleavage activity of inteins to purify recombinant
native proteins by a single affinity step. Although the
protein eluted from the chitin column contained high
molecular mass impurities, FH8 was purified to homo-
geneity with an Amicon 30 kDa molecular filter
(Fig. 1A). In addition to the FH8 band, a band
around the 15 kDa marker was observed. This band
was stronger when the SDS sample buffer contained
no reducing agents. As the FH8 sequence has one Cys
(Cys36), this was an indication that the 15 kDa band
corresponded to a dimer formed with the oxidation of
the FH8 single Cys, and a reducing agent was there-
fore included in the assays [2 mm triphenyl phosphine
(TECP), unless otherwise indicated].
Results from MS analysis led to the identification of
a major peak corresponding to the product of the
N-terminal hydrolysis of FH8 Met (7532.7 gÆmol
)1
).
The cleavage of the N-terminal Met is a common post-
translational modification catalyzed by Escherichia coli
aminopeptidades when the side chain in the penulti-
mate residue is Ala, Cys, Pro, Ser, Thr, and Val. FH8
contains a Pro at position 2, and this process was
therefore very likely to occur during the overnight
cleavage with dithiothreitol. Because of its position,

the removal of this Met will not influence any of the
biochemical data. MALDI-TOF MS also allowed the
identification of a small peak of the FH8 dimer
(15 076 gÆmol
)1
). All of the purification steps were per-
formed in the presence of 1 mm EDTA, and the puri-
fied FH8 was free of Ca
2+
as confirmed by atomic
absorption spectroscopy.
Sequence analysis
The FH8 sequence was initially compared with those
of the two other CaM-like proteins from F. hepatica,
FhCaM1 and FhCaM2. Identities were only 19% and
26%, and 22% and 33%, for the N-termini and C-ter-
mini of FhCaM1 and FhCaM2, respectively (Fig. 2A).
Moreover, sequence similarity revealed that other
helminths have CaBPs similar to FH8 (Fig. 2B),
namely C. sinensis (CH8, 55% identity), S. mansoni [9]
(SM8, 40% identity) and S. japonicum [11] (SJ8, 37%
identity). Interestingly, just like FH8, the S. japonicum
protein is localized in the parasite surface, and is
expressed at the initial stages of infection [9,11].
In order to obtain an indication of the type of CaBP
that FH8 might be, several sequence alignments were
performed, using CaM and TnC as representatives of
the sensor proteins, and calbindin D
9k
, as a model for

the buffer CaBPs. They contain one (calbindin D
9k
)or
two (CaM and TnC) globular domains, each of them
with two EF-hand Ca
2+
-binding motifs linked by an
EF-hand b-scaffold.
Figure 2C shows the alignment of FH8 with the
N-terminal and C-terminal fragments of CaMs and
also with calbindin D
9k
. The C-terminal domain of
TnC is also shown, because of its close similarity to
CaMs and because some amino acids in the FH8
sequence are different from those in CaM but are
homologous to TnC residues.
The Ca
2+
-binding sites are presented in red. Whereas
the CaM and TnC families of proteins are characterized
by a binding loop with 12 amino acids, calbindin D
9k
has 14 amino acids in the corresponding loop.
In most EF-hand proteins, Ca
2+
is coordinated to
seven oxygen atoms, arranged in a pentagonal bipyra-
mid; six are provided by the protein, and one by a
water molecule. Positions X, Y and Z indicate the first

three Ca
2+
ligands of the loop, each of them contrib-
uting one oxygen; the Glu in the last position of the
loop ()Z) contributes two oxygens of its c-carboxyl
group, and the central residue of the loop ()Y) binds
Ca
2+
with the main chain carbonyl oxygen. In most
structures, the seventh ligand is a water molecule, in
position )X, provided indirectly by the protein. Next
to residue )Y, there is a hydrophobic amino acid, Ile
in most CaMs and Val and Leu in FH8, whose main
chain forms two hydrogen bonds with the equivalent
residue of the paired EF-hand, forming the EF-hand
b-scaffold. The structural integrity of the two-EF-hand
domain is maintained by this short b-sheet and by
hydrophobic contacts between the protein helices. The
last three residues of each Ca
2+
-binding loop are
helical, and form the first turn of the exiting helix.
AB
EDTA Ca
2+
Fig. 1. (A) Purification of recombinant FH8 with the IMPACT
expression system. Ext, E. coli extract after overnight induction of
the 63 kDa FH8–intein tag construct (*); FT, flowthrough of the chi-
tin column; Elu, eluted protein after overnight incubation in reducing
buffer (dithiothreitol, 50 m

M); 30 K, purified FH8 after 30 kDa
molecular exclusion. (B) Native gel retardation assay. FH8 displays
the characteristic mobility shift observed for EF-hand CaBPs in the
presence of Ca
2+
. Runs were performed in the presence of 1 mM
EDTA or 5 mM Ca
2+
.
FH8 from Fasciola hepatica H. Fraga et al.
5074 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
The results show that FH8 contains two EF-hand
motifs, which most probably form a globular domain.
The size and amino acid content of the Ca
2+
-binding
sites are very similar to those in CaM and TnC, but
not to those in calbindin D
9k
.
The conserved amino acids that coordinate Ca
2+
(positions X, Y, Z and )Z) are preserved between
CaM and FH8, with the exception of amino acids at
position Y in both loops, Asn17 ⁄ Asp and Asn53 ⁄ Asp
(amino acids refer to FH8 ⁄ CaM). Interestingly, these
substitutions also occur in TnC and on the first EF-
hand of FhCaM2. It is known that the replacement of
Glu at position )Z with other amino acids causes a
dramatic decrease in Ca

2+
affinity, whereas mutations
at other Ca
2+
-coordinating positions do not have such
drastic consequences [18,19]. As reports on structural
and biochemical data indicate that the mechanisms of
Ca
2+
-induced conformational changes in CaM and
TnC are similar [20], we foresee that the Asn ⁄ Asp
modifications will not impair Ca
2+
binding.
In positions )Y are Lys21 and Lys57, which are
different from the amino acids in CaM and TnC.
However, the coordination from this position is per-
formed through the oxygen of the main chain carbonyl
group. Although the side chain is not directly involved
in Ca
2+
coordination, it is possible that the positively
charged side chain will influence Ca
2+
binding through
charge repulsion.
Finally, we decided to check the hydrophobicity of
the FH8 sequence, and compare it with the CaM
N-terminal and TnC C-terminal sequences, CaM and
TnC being the proteins used for the homology model-

ing studies. The N-terminal and C-terminal fragments
were the domains of each protein that were most simi-
lar to FH8. The hydropathy plots are presented in
Fig. 3, and they show small, but noticeable, differences
between FH8 and the other two domains. The N-ter-
minal region up to amino acid 31 is similar for the
three proteins; then there is a region in FH8, involving
Asp32 and Asp33, which is clearly negatively charged,
whereas CaM and TnC have identical and positive
charges. Cys36-Pro37-Leu38 from FH8 is less hydro-
philic than the corresponding aligned regions in CaM
and TnC; it includes the only Cys present in FH8,
A
B
C
Fig. 2. Sequence alignment of FH8 with: (A) the CaM-like proteins FhCaM1 and FhCaM2; (B) the helminth proteins CH8, SH8 and SM8; and
(C) the prototypical Ca
2+
sensor protein CaM together with the C-terminal fragments of TnC and Ca
2+
buffer calbindin D
9K
. As CaM and TnC
have two domains, only the N-terminal or C-terminal fragments that had higher homology with FH8 are presented in the alignment. The
sequence numbering is shown at the beginning and at the end of each fragment. The Protein Data Bank codes of the protein structures
used as templates for the molecular modeling of FH8 are presented in parentheses. The Ca
2+
-binding sites are colored red, and positions
within sites I and II that are involved in chelating Ca
2+

are labeled X, Y, Z, )Y, )X and )Z. The consensus symbols denoting the degree of
conservation in each column between FH8 and CaMs and FH8 and calbindin D
9K
proteins are colored blue and have the following meaning:
‘*’, identical in all sequences in the alignment; ‘:’, conserved substitutions; ‘.’, semiconserved substitutions.
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5075
which is conserved in similar proteins from fluke para-
sites (CH8, SH8, SM8; see Fig. 2B). Clearly, Cys36 is
a valid target for mutagenesis in future functional
studies. Finally region 47–60, which belongs to the sec-
ond EF-hand and corresponds to four amino acids of
the first helix plus the Ca
2+
-binding site without the
two last residues, is substantially different from that in
CaM and remarkably similar to that in TnC. These
differences may influence protein stability and ⁄ or
Ca
2+
-binding affinity.
Ca
2+
-induced conformational alterations
FH8 sequence homology revealed the presence of two
EF-hand motifs, and therefore the first question to be
explored was the protein’s ability to bind Ca
2+
. This
was initially tested using the PAGE mobility assay, as

it has been reported that, in the presence of Ca
2+
,
functional CaBPs display significant mobility shifts in
electrophoresis [2]. In fact, we observed retardation in
FH8 migration in a 10% native gel with 5 mm Ca
2+
(Fig. 1B). This preliminary result, demonstrating the
functionality of FH8 EF-hand motifs, was further
substantiated with an equilibrium dialysis experiment
with Ca
2+
(5 mm), where a ratio of 2.0 ± 0.3 Ca
2+
per FH8 macromolecule was observed, indicating that
both Ca
2+
-binding sites are functional.
As FH8 was able to bind Ca
2+
, possible structural
modifications associated with Ca
2+
coordination were
explored. This question was particularly relevant, as it
is associated with the functionality of FH8, as a buffer
or sensor protein.
To probe for possible conformational modifications,
intrinsic amino acid fluorescence was used. FH8 does
not contain any Trp or Tyr residues, which are

commonly used for this approach, but it has two Phe
residues (Phe30 and Phe46). Phe has relatively weak
fluorescence, which is negligible in the presence of Trp
or Tyr, but it was previously used to monitor Ca
2+
binding [21]. Accordingly, Phe fluorescence was mea-
sured in the presence of EDTA and Ca
2+
, and we
observed an increase of 30% in the Ca
2+
-loaded state
(Fig. 4A), indicating a conformational alteration.
Figure 5 shows the molecular model obtained for FH8
by homology modeling. The side chains of Phe30 and
Phe46 are represented in the model, and they are
located close to binding site I (Phe30) and the
EF-hand b-scaffold (Phe46), which are regions that
undergo large conformational alterations as a result of
Ca
2+
binding (Figs 5 and 6). As can be seen in Fig.6,
a hydrophobic patch becomes exposed upon Ca
2+
binding in the case of CaM, and for FH8 the solvent-
exposed hydrophobic area is also larger in the Ca
2+
-
loaded state.
Dynamic light scattering (DLS) is another technique

that can be used as a screen for major conformational
changes in proteins, and has previously been used by
other researchers to monitor CaM conformational
alterations upon Ca
2+
binding [22]. We measured the
hydrodynamic radius (R
H
) of FH8 in the presence and
absence of Ca
2+
. Despite the large standard deviation
of the results, it is clear that FH8 displays a larger R
H
Fig. 3. Hydropathy profiles of FH8, the CaM N-terminus and the
TnC C-terminus.
Fig. 4. (A) Phe fluorescence. Emission spectrum of FH8 Phe in the
presence of 20 m
M Ca
2+
(circles) and 3 mM EDTA (crosses). (B) R
H
of FH8 in the presence of 20 mM Ca
2+
(circles) and 1 mM EDTA
(crosses) determined by DLS. FH8 shows an increase in R
H
that is
consistent with a more elongated shape in its Ca
2+

-loaded state.
FH8 from Fasciola hepatica H. Fraga et al.
5076 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
(2.7 ± 1.1 nm) in the Ca
2+
-loaded state than in the
apo state (1.8 ± 0.7 nm; Fig. 4B). Interestingly, these
results are in agreement with the data reported for
CaM [22], supporting the idea that Ca
2+
coordination
results in alterations in FH8 structure.
After determining that FH8 undergoes conforma-
tional changes upon Ca
2+
binding, we decided to test
whether these alterations resulted in an increase in the
hydrophobicity of the protein surface, which is a char-
acteristic of Ca
2+
sensors. The binding of the hydro-
phobic probe 8-anilonaphthalene-1-sulfonate (ANS) to
FH8 was monitored by fluorescence spectroscopy.
ANS binds noncovalently to hydrophobic segments of
proteins, and it was reported that Ca
2+
sensors show
strong enhancement of ANS fluorescence upon Ca
2+
binding [15]. Consistent with the previous hypothesis,

Ca
2+
was added to FH8, and ANS fluorescence detec-
tion resulted in a blue shift and enhancement of ANS
fluorescence (Fig. 7A). Moreover, no changes in ANS
fluorescence emission were observed in the presence of
200 mm NaCl, which excludes the effect of ionic
strength, or 20 mm Mg
2+
, demonstrating the specific-
ity of the conformational change for Ca
2+
(Fig. 7B).
The failure of Mg
2+
to induce hydrophobic residue
exposure may result from the noncoordination of this
ion or just its inability to promote a conformational
change, as previously reported for CaM [23].
In order to determine whether the increase in hydro-
phobic residue exposure in FH8 was a consequence of
the formation of dimers or aggregates induced by
Ca
2+
, the primary amine cross-linker suberic acid
bis(3-sulfo-N-hydroxysuccinimide) ester (BS3) was used
A
B
B
A

C
Site I
Site II
Site II
Site I
D
D
C
Fig. 5. The overall modeled structure of FH8 in the open (green) and closed (yellow) conformations, represented by cartoon ribbons. The
protein has two EF-hand Ca
2+
-binding sites (represented by site I and site II). The Ca
2+
-binding loops are presented in more detail; the side
chains of atoms that participate in Ca
2+
coordination are shown as sticks and labeled; the calcium ions are represented by gray spheres. In
site II, FH8 Asp59 ()X position) is also shown. The backbone of the residues forming the EF-hand b-scaffold is shown in a ball-and-stick rep-
resentation. The positions of Cys36, Phe30 and Phe46 are highlighted. The four helices are labeled A, B, C and D. The molecular models
were obtained by homology modeling, using the
SWISS-MODEL and SWISS-PDB VIEWER programs [41]. The figures were prepared with PYMOL
().
A
a
b
d
c
a
b
d

c
a
b
d
c
a
b
d
c
B
Fig. 6. Molecular surface representation of CaM and FH8 in the
closed and open conformations. The hydrophobic accessible sur-
faces, as defined by the side chains of Val, Ile, Leu and Phe, are in
green. (A) The N-terminal domain of R. norvegicus CaM in the apo
(Protein Data Bank code: 1QX5) and Ca
2+
-loaded (Protein Data Bank
code: 3B32) conformations. (B) FH8 modeled structures for both
conformations. The location of Cys36 is shown as yellow sticks. The
figures were prepared with
PYMOL ().
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5077
in the presence of different concentrations of Ca
2+
.As
shown in Fig. 7C, it is clear that the FH8 apo state is
a monomer, although some residual dimerization was
observed for concentrations of Ca
2+

above 1 mm.
The conformational alterations induced by Ca
2+
and, in particular, the sensitivity of the ANS assay
were used to assess FH8 affinity for this ion, as previ-
ously described for other proteins [24,25]. Figure 7D
shows the fluorescence of ANS titrated with Ca
2+
in
the presence of FH8. The Ca
2+
titration curve is sig-
moidal, an indication of cooperative binding, and fit-
ting of the experimental data to a simple allosteric
model gave a Hill coefficient of 1.6 ± 0.09 and a K
obs
of 590 ± 20 lm. This result is in agreement with the
findings of experiments on several EF-hand CaBPs,
where cooperativity in Ca
2+
binding was observed.
However, the K
obs
was unusually high in comparison
with the canonical proteins of the sensor family. In
order to corroborate this result, equilibrium dialysis
with 250 lm Ca
2+
was performed. Consistent with
the low affinity suggested by the Ca

2+
titration curve,
a ratio of 0.2 Ca
2+
per FH8 was determined, confirm-
ing that FH8 has a low affinity for Ca
2+
(data not
shown). In fact, EF-hand CaBPs have binding con-
stants for Ca
2+
that span a wide range (10
3
–10
9
m
)1
),
with no obvious correlation with the type or arrange-
ment of the Ca
2+
ligands [14], although it is known
that exposure of hydrophobic residues results in a
lower affinity for Ca
2+
, and sensor proteins invariably
show lower affinities than buffer proteins.
Sequence alignments show that the EF-hand coordi-
nation loops of FH8 have two (Arg16 and Lys21) and
four (Lys52, Lys54, Lys57 and Lys61) positively

charged amino acids, representing considerable electro-
static force repulsion, particularly for site II, when
Ca
2+
approaches the loops. This is probably one of
the main reasons for the low affinity observed for
FH8. In fact, it has been reported that increasing the
negative charge in the loop by replacing some amino
acids with Asp increases the Ca
2+
content, even if the
residue is not in one of the coordinating positions; in
contrast, removal of negatively charged side chains
causes a decrease in Ca
2+
affinity [26,27]. Curiously,
the other fluke CaBPs (see Fig. 2) do not contain this
positively charged stretch of amino acids, and it would
be interesting to compare their respective Ca
2+
affini-
ties, but no data are currently available.
Besides FH8, several other low-affinity EF-hand
proteins (K
d
=10
3
–10
4
m

)1
) have been described
[14,28–30], including a-spectrin [31] and multiple
AB
CD
Fig. 7. (A) ANS fluorescence spectroscopy. Changes in ANS fluorescence emission indicate a Ca
2+
-dependent increase in hydrophobic resi-
due exposure. ANS emission in the presence of Ca
2+
-loaded FH8 (20 mM Ca
2+
, circles) was 2.8-fold more intense than in its apo state
(1 m
M EDTA, crosses). The addition of ANS to FH8 did not result in a significant change in fluorescence as compared with ANS only (not
shown). (B) The changes in ANS fluorescence are specific for Ca
2+
; no changes in ANS fluorescence emission were observed with 200 mM
NaCl and 20 mM Mg
2+
(C) Chemical cross-linking of FH8. Purified recombinant FH8 was cross-linked using BS3 and the indicated Ca
2+
con-
centrations. The lower lane corresponds to FH8 monomer, and the upper lane corresponds to FH8 dimer. (D) Titration of FH8 with Ca
2+
,
using ANS fluorescence as reporter. Data were fitted to the Hill equation.
FH8 from Fasciola hepatica H. Fraga et al.
5078 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
members of the S100 [28] and CREC families [32].

Like FH8, all of these proteins can be found extra-
cellularly or in the endoplasmic reticulum secretory
pathways, where the Ca
2+
concentration is high.
Secondary structure and thermal stability
Information related to protein secondary structure and
stability in the presence and absence of Ca
2+
was
obtained by CD spectroscopy. The spectra are pre-
sented in Fig. 8, and they reveal that FH8 has a signif-
icant content of ordered secondary structure. The
far-UV CD spectrum of FH8 at 20 °C (Fig. 8A),
contains the shapes and amplitudes characteristic of
proteins with a high percentage of helical structure. In
the presence of 20 mm Ca
2+
, no significant changes in
the CD spectrum were observed. This is in contrast to
what is seen with the typical sensor proteins, CaM and
TnC, which display shifts in the CD spectrum as a
result of the reorganization of the helical packing
within the protein [33]. However, CD studies on the
N-terminal half of TnC also do not demonstrate
changes upon the addition of Ca
2+
[34].
The CD spectra for FH8 also show that Ca
2+

bind-
ing results in stabilization of the protein structure.
Indeed, as shown in Fig. 8B, the thermal stability of
the Ca
2+
-loaded FH8 is very high, and FH8 therefore
behaves like the Ca
2+
-loaded states of other EF-hand
CaBPs, namely CaM, TnC and calbindin D
9K
. In all
of these proteins, the denaturation temperatures of the
Ca
2+
-loaded states are so high that they are not exper-
imentally accessible [35,36]. Figure 8b shows that there
is no observable loss in FH8 secondary structure up to
98 °C. The FH8 apo state is less stable, and we were
able to determine its melting point (T
m
)as
74.0 ± 0.3 °C (Fig. 8C). Although not comparable to
its Ca
2+
-loaded state, apo-FH8 is still a remarkably
stable protein.
It is known that the two EF-hand domains are sta-
bilized by backbone hydrogen bonds connecting the
Ca

2+
-binding loops in a short stretch of antiparallel
b-sheet, as well as by numerous hydrophobic contacts
between the helices. In the Ca
2+
-loaded state, the
CaBPs are further stabilized by Ca
2+
ligand interac-
tions, and are normally more stable [13,37]. Curiously,
apo-FH8 is substantially more stable towards thermal
denaturation than CaM or TnC (T
m
$ 55 °C) [35],
although apo-calbindin D
9k
does demonstrate even
higher T
m
values (85 °C or higher [36]).
Homology models
The N-terminal fragment of CaM from Rattus norvegi-
cus was used for homology modeling of FH8, as this is
the only organism for which the X-ray crystallographic
structures for the apo (Protein Data Bank: 1QX5) and
Ca
2+
-loaded (Protein Data Bank: 3B32) states are
available [38,39]. swiss-model in the alignment mode
was used to produce the models for FH8 in the Ca

2+
-
loaded and Ca
2+
-free conformations. The 69 amino
acids of FH8 were modeled without any insertions or
deletions included in the sequence alignment. Analyses
of the models were performed by Anolea and Gromos
[40], revealing a favorable energy environment for all
of the amino acids, with the exception of a few resi-
dues belonging to the Ca
2+
-binding sites. The final
total energy for the models was approximately
A
B
C
Fig. 8. Far-UV spectra of FH8. (A) FH8 at 20 °C in the absence
(thin line) and in the presence (thick line) of 20 m
M Ca
2+
. (B) FH8 at
98 °C in the absence (thin line) and in the presence (thick line) of
20 m
M Ca
2+
. (C) Thermal denaturation of the apo state of FH8 was
followed by CD absorbance measurements at 210 nm. Unfolding
was shown to be reversible, and T
m

was calculated on the assump-
tion of a two-state model; details are given in Experimental
procedures.
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5079
)1750 kJÆmol
)1
, according to swiss-model calcula-
tions. Both models were checked with swiss-pdbviewer
[41], in order to verify the conservation of structural
features with a functional role, namely those implicated
in Ca
2+
binding. The FH8 amino acid side chains
directly implicated in Ca
2+
binding and that were
different from those present in R. norvegicus corre-
sponded to the amino acids that did not present a
good local conformation when the quality of the
models was evaluated. As the sequence alignments had
shown that these few amino acids were identical to
those present in TnC from Mus musculus (Protein
Data Bank: 1A2X), we manually modeled the side
chains of Asn17, Asn53 and Asp55, according to the
orientations observed for TnC.
The model structures in the open (Ca
2+
-loaded) and
closed (apo) conformations are presented in Fig. 5.

Closer views of Ca
2+
-binding sites I and II in the open
conformation are also shown. In site II, Asp59 ()X
position) is shown, as sequence alignments showed
that, although it is different in CaM, it is identical in
TnC and therefore could be modeled with a high level
of confidence. As happens for TnC, Asp59 is probably
hydrogen-bonded to a water molecule that belongs to
the coordination sphere of Ca
2+
. FH8 has four helical
regions, and the corresponding residues for the Ca
2+
-
loaded ⁄ Ca
2+
-free conformations are as follows: A,
2–14 ⁄ 2–13; B, 24–34 ⁄ 24–33; C, 40–48 ⁄ 40–50; and D,
60–68 ⁄ 58–68. Ca
2+
-binding sites I and II have the
appropriate amino acids and geometry to coordinate
the metal ions, and are located in loop regions (15–26)
and (51–62) that are flanked by helices on either side.
In fact, the last three residues at the end of each loop
form the first turn of exiting helices B for site I and D
for site II.
In the FH8 Ca
2+

-loaded protein, the antiparallel
b-sheet that links Ca
2+
-binding sites I and II has only
two hydrogen bonds, which are established between
the main chain carbonyls and amides of Val22 and
Leu58. In the case of the closed conformation, there is
an additional hydrogen bond established between the
carbonyl of Gly20 and the amide group of Leu60
(Fig. 5).
If helices A and D are kept roughly with the same
orientation in the images of Fig. 5, which represents
both states, it can be seen that the helical packing is
different. When Ca
2+
is bound, helices B and C open
up slightly and expose a number of hydrophobic side
chains, which were kept away from the solvent in the
apo state. Figure 6 shows a comparison of the surface
hydrophobicity for CaM and FH8 in the apo and
Ca
2+
-bound states. Amino acids such as Val, Ile, Leu
and Phe, which are very important in defining the
protein hydrophobic regions, are in green, and the
others are in gray. FH8 shows, in both states, larger
hydrophobic surfaces than are seen in CaM. The sig-
nificant exposed hydrophobic surface in the FH8
Ca
2+

-free state indicates a possible area for dimeriza-
tion or a region of interaction with other proteins. As
in CaM, the hydrophobic patch that becomes exposed
upon Ca
2+
-binding is probably a region of interaction
between FH8 and other proteins.
Figures 5 and 6B also show the position of Cys36,
whose oxidation we observed in SDS gels. It belongs
to the B ⁄ C linker, a region that was proposed to be
crucial for explaining the difference in behavior of the
sensor and buffer proteins [15]. Moreover, Cys36,
besides being very close to the protein exposed hydro-
phobic patch, even in the apo form, is on the surface,
and therefore may easily become oxidized through
covalent binding to small or larger molecules, such as
another FH8 macromolecule. However, if two FH8
macromolecules are covalently linked through the
Cys36 residues, this complex will not acquire the
dumbbell shape presented by CaM and TnC, both
containing four EF-hand motifs, organized into two
domains linked by a long helix.
It has been reported that CaM has several Met resi-
dues on the surface, which are close to each other in
the apo conformation and not so close in the Ca
2+
-
loaded (open) conformation, and these residues were
considered to be important because, when they were
mutated, activation of the ligands was impaired [42].

Moreover, they were proposed to have a role in stabi-
lizing the open conformation [43]. FH8 has one unique
Met (Met1), which sits at the N-terminal and obvi-
ously will not be involved in ligand binding.
Reinforcing the idea of the movement between heli-
ces, we also found, in the modeled FH8 structure, sev-
eral hydrophobic interactions between the amino acids
of the four helices. Table 1 shows the hydrophobic
interhelical contacts, and it is obvious that the interac-
tions within the pairs of helices A ⁄ D and B ⁄ C are
the least affected by Ca
2+
binding, whereas in the
Table 1. The hydrophobic interhelical contacts present in FH8.
Helices Ca
2+
-free state Ca
2+
-loaded state
A ⁄ D Val7, Leu11 ⁄ Leu63, Val64, Leu67 Val7, Leu10,
Leu11 ⁄ Leu60,
Leu63, Leu67
A ⁄ B Val7, Leu10, Leu11 ⁄ Phe30 Val13 ⁄ Phe30
B ⁄ C Ala24, Leu27 ⁄ Ile43, Phe46, Ile47 Ala24, Leu27 ⁄ Ile43,
Phe46
B ⁄ D Leu27, Phe30 ⁄ Leu63, Leu67
C ⁄ D Phe46, His50 ⁄ Leu63, Ile66 Phe46 ⁄ Ile66
FH8 from Fasciola hepatica H. Fraga et al.
5080 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
Ca

2+
-loaded state, the interhelical interfaces concern-
ing B ⁄ D disappear completely, and they are reduced in
number for A ⁄ B. In fact, the B ⁄ C pair of helices
swings away from the A ⁄ D pair. This is in agreement
with findings reported for CaM [43].
Figures 5 and 6 also show the positions of the two
Phe residues that were responsible for the detection of
structural modifications upon Ca
2+
binding by fluores-
cence. Phe30 belongs to helix B and is close to binding
site I, whereas Phe46 belongs to helix C and is part of
the hydrophobic patch that becomes exposed when the
protein has an open conformation. Our molecular
models show that when Ca
2+
binds, helices B and C
move in relation to their previous positions. In the apo
state, Phe30 is surrounded by hydrophobic side chains
belonging to helices A (Val7, Leu10, Leu11 and
Leu14), B (Val22 and Leu27) and D (Leu63 and
Leu67). Upon binding of Ca
2+
, Phe30 moves far away
from helices A and D, and only Leu27 remains close
to it. Phe46 in the closed conformation has several
aromatic side chains around it, namely Leu27 from
helix B, Ile43, Ile47 and His50 from helix C, and
Leu58, Leu63 and Ile66 from helix D; however, in the

open conformation, only Leu27 and Ile66 remain close
to Phe46. These modifications explain the observed dif-
ferences in Phe fluorescence for the Ca
2+
-loaded and
Ca
2+
-free FH8 states.
Discussion
FH8 is an EF-hand CaBP with only one globular
domain and the characteristics of a sensor protein, dis-
playing increases in radius and hydrophobic residue
exposure with Ca
2+
loading that are typical of this
class of proteins. Interestingly, the only Cys residue
present in FH8 is on the protein surface, where it is
able to be oxidized and form the dimers that we
observed in SDS gels. As Cys36 is located in the B ⁄ C
linker, the three-dimensional arrangement of the FH8
dimer will be significantly different from the CaM or
TnC overall structures. Curiously, Cys36 is strictly
conserved in this group of helminth proteins, suggest-
ing a specific role for this amino acid in these proteins.
FH8, like CaM and TnC, contains a short B ⁄ C linker,
a region that allows adjustment of the relative posi-
tions of these two helices in different conformations.
Interestingly, both CaM and TnC are also the proteins
that exhibit the largest domain opening.
As occurs for other members of the EF-hand family,

FH8 demonstrates an unusual thermal stability
(T
m
= 74.0 ± 0.3 °C), and in the Ca
2+
-loaded state,
FH8 is even more stable, as it is further stabilized by
Ca
2+
ligand interactions.
FH8 displays a low affinity for Ca
2+
(K
obs
=10
4
m
)1
),
probably because of charge repulsion between the metal
ion and the binding site, as loop I has two positively
charged side chains and loop II has four, which is an
extremely high number in comparison with canonical
CaBPs. This may be in accordance with its extracellular
location (1.2 mm Ca
2+
) and F. hepatica migration to
the gallbladder, where the Ca
2+
concentration is partic-

ularly high (18 mm). Interestingly, very high concentra-
tions of Ca
2+
were reported for the Schistosoma
glycocalyx, where the homologous SJ8 and SM8
proteins were found [11]. Ca
2+
and CaBP are well-
known mediators of changes in cell morphology, and
modifications in the glycocalyx have been suggested to
be one mechanism for host immune system suppression
by F. hepatica [44].
Using homology modeling, we were able to observe
that the structural integrity of the two-EF-hand
domain is maintained by a short stretch of antiparallel
b-sheet connecting the Ca
2+
-binding loops and by
numerous hydrophobic contacts between the helices.
We confirmed that, like other sensor proteins, FH8 in
the Ca
2+
-loaded state has a reduced number of interh-
elical contacts relative to those present in the
Ca
2+
-free state, allowing movement of the B⁄ C pair of
helices relative to the A ⁄ D pair.
Although the extent and nature of the Ca
2+

-induced
conformational changes can only be determined
through full structural characterization, FH8 repre-
sents, in our opinion, a case study protein for the EF-
hand family, owing to its solubility, size (69 amino
acids) and sensor properties, which allow a good con-
trast with the similar-sized calbindin D
9k
(76 amino
acids), the prototypical Ca
2+
buffer protein.
In summary, we present here a new EF-hand protein
that behaves like a sensor CaBP, displays low affinity
for Ca
2+
, and is highly stable in its apo and Ca
2+
-
loaded states. Our work will proceed with the determi-
nation of the three-dimensional structures of FH8 in
the apo and Ca
2+
-loaded states, using X-ray crystal-
lography and ⁄ or NMR, and mutagenesis of key amino
acids in order to study their influence on Ca
2+
affinity
and conformational changes. This will contribute to a
complete understanding of the main aspects that drive

conformational changes and affinity in CaBPs.
Experimental procedures
Unless indicated otherwise, all reactions were performed at
22 °Cin10mm Tris (pH 8.0) containing 2 mm TECP,
1mm EDTA and 150 mm KCl. ANS, phenylmethanesulfo-
nyl fluoride, Hepes, EDTA, ampicillin, Tris, BS3 and TECP
were obtained from Sigma. pTYB1 plasmid, NdeI, SapIand
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5081
chitin beads were obtained from New England Biolabs.
Isopropyl-thio-b-d-galactoside was obtained from Fermentas.
FH8 cloning and purification
FH8 was cloned into the pTYB1 expression vector, using
NdeI and SapI restriction sites. Expression and purification of
the tagged protein were performed as described by the manu-
facturer, using E. coli BL21(DE3) Star K cells, following
overnight induction with 0.3 mm isopropyl-thio-b-d-galacto-
side at 18 °C. The eluate from the chitin column was loaded
onto an Amicon ultra 30 kDa molecular filter (Millipore,
Billerica, MA, USA), and the protein recovered was homoge-
neous on a 10% SDS ⁄ Tricine gel. The protein concentration
was measured with the Bradford method [45]. The purified
protein was further analyzed by MALDI-TOF MS.
Native gel electrophoresis was performed with 5 lgof
purified FH8, which was loaded in a Tris ⁄ glycine 10%
native gel in the presence of 1 mm EDTA or 20 mm CaCl
2
.
SERVA native markers were used as reference.
Atomic absorption spectroscopy and equilibrium

dialysis
The Ca
2+
content in the purified FH8 solutions was deter-
mined with a Perkin Elmer 3100 flame atomic absorption
spectrophotometer. Equilibrium dialysis was performed with
800 lm FH8 (400 lL) in a 2 kDa dialysis membrane
(Thermo Scientific, Rockford, IL, USA) equilibrated over-
night with 1 L of buffer or buffer containing 250 lm or
5mm Ca
2+
.
Fluorescence spectroscopy
Phe fluorescence was measured in a Jasco Spectrofluorime-
ter with 300 lm FH8. The experiments were performed in
the 260–280 nm interval, with a 250 nm excitation wave-
length (10 nm excitation and emission slits). ANS fluores-
cence was measured in a Horiba spectramax fluorimeter
with 20 lm FH8 and 300 lm ANS in the presence of 1 mm
EGTA or 20 mm Ca
2+
. Fluorescence emission spectra were
measured in the 400–700 nm interval, with 385 nm as the
excitation wavelength for ANS (5 nm excitation and emis-
sion slits, 0.1 s integration time). ANS fluorescence titration
was performed with 30 lm FH8 in the presence of 300 lm
ANS, to which aliquots of Ca
2+
were added. The emission
at 485 nm was plotted as a function of the free Ca

2+
calcu-
lated using the apparent association constant for EGTA
(1 mm). Data were fitted to the Hill binding model [46] with
the equation F ¼ F
max
 Ca
n
=ðK
n
obs
þ Ca
n
Þ, using graphpad
prism 5.0c. In this equation, F is the ANS fluorescence at
485 nm, Ca is the concentration of free Ca
2+
, K
obs
is the
association constant, which is assumed to be equal to all
Ca
2+
-binding sites, and n is the Hill coefficient, which
reveals the type of binding cooperativity for the multiple
sites.
DLS
DLS measurements were obtained in Malvern Zetasizer
Nano ZS. The FH8 concentration was 125 lm (100 lL),
and protein samples were filtered and ultracentrifuged

before measurements to remove any aggregates. Measure-
ments were performed in the presence of 1 mm EDTA,
20 mm Ca
2+
and 20 mm Mg
2+
.
CD spectroscopy
Far-UV CD spectra were acquired on an Olis DSM 20 CD
spectropolarimeter continuously purged with nitrogen,
equipped with a Quantum Northwest CD 150 temperature-
controlled cuvette, and controlled by globalworks soft-
ware. A 40 lm protein solution was prepared in 10 mm
Tris ⁄ HCl buffer (pH 8.8), containing 2 mm TCEP and
1mm EDTA. To evaluate the effect of Ca
2+
on the pro-
tein, assays were carried out in the absence and in the pres-
ence of 20 mm Ca
2+
, which is about 35-fold the K
d
,
ensuring full occupancy. Scans were collected with a
0.2 mm path length cuvette, between 190 and 260 nm, at
1 nm intervals. Three scans with an integration time of 8 s
were averaged for each measurement. Spectra were acquired
at 20 and 98 °C. For the thermal unfolding experiments,
the ellipticity at 210 nm was monitored for 3 min at several
temperatures, ranging from 20 to 98 ° C. At each data

point, the sample temperature was allowed to equilibrate
for 2 min. The thermal unfolding of the protein was totally
reversible, allowing thermodynamic analysis of the tempera-
ture dependence of the ellipticity at 210 nm, in accordance
with a two-state model, considering a linear dependence of
the ellipticity values of the native and unfolded protein.
The results are expressed in terms of mean residue ellip-
ticity [Q]
MRW
in deg cm
2
dmol
)1
, according to the equation
[Q]
MRW
= Q
obs
· 100m ⁄ (lcN), where Q
obs
is the observed
ellipticity in mdeg, m is the protein molecular mass (g ⁄ mol),
l is the cuvette path length (cm), c is the protein concentra-
tion (g ÆL
)1
) and N is the number of residues of the protein.
Chemical cross-linking
Cross-linking was performed as described previously [47],
and with 50 lm FH8 in Hepes (pH 7.5), 150 mm KCl,
10 mm mercaptoethanol and the indicated Ca

2+
concentra-
tions to a final volume of 50 lL. Cross-linking was initiated
by the addition of freshly prepared BS3 to a final concen-
tration of 10 mm. The reaction mixtures were incubated at
room temperature for 30 min, and the reaction was
quenched by the addition of 1 m Tris (50 lL); the solutions
were then loaded onto a 10% SDS ⁄ Tricine gel.
FH8 from Fasciola hepatica H. Fraga et al.
5082 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
Sequence analysis and homology modeling
Hydropathy plots for analysis of the FH8 hydrophobic
character were obtained with the Kyte and Doolitle scale
[48], with a window size of 5.
The conserved regions of different CaMs and calbin-
din D
9k
proteins were detected by sequence homology, with
clustalw2 [49] and blast [50]. These regions were then
assigned to the corresponding regions along the FH8
sequence.
Finally, the crystal structures of CaM N-terminal frag-
ment from R. norvegicus in the apo (Protein Data Bank
code: 1QX5 [38]) and Ca
2+
-loaded (Protein Data Bank
code: 3B32 [39]) states were used as templates for homology
modeling of FH8 with the swiss-model and swiss-
pdb viewer programs [41]. We used CaM from R. norvegi-
cus, as X-ray crystallographic coordinates for both states,

apo and Ca
2+
-loaded, were available. The molecular model
of TnC from M. musculus (Protein Data Bank code: 1A2X
[51]) was also employed for modeling of the side chains
belonging to the Ca
2+
-binding sites that were different in
CaM but identical to those present in TnC. The figures
with the molecular models were prepared with pymol
().
Acknowledgements
The authors thank A. Castro and P. Moradas Ferreira
for fruitful discussions and the generous gift of the
FH8 plasmid. We also acknowledge financial support
from Fundac¸ a
˜
o para a Cieˆ ncia e Tecnologia (FCT),
Lisbon, Portugal, for projects CONC-REEQ ⁄ 564 ⁄
BIO ⁄ 2001 (FEDER & FCT), PTDC ⁄ SAU-NEU ⁄
69123 ⁄ 2006 and grant SFRH ⁄ 26490 ⁄ BPD ⁄ 2006 to
H. Fraga.
References
1 Mas-Coma S, Bargues MD & Valero MA (2005) Fasci-
oliasis and other plant-borne trematode zoonoses. Int J
Parasitol 35, 1255–1278.
2 Russel SL, McFerran NV, Hoey EM, Trudgett A &
Timson DJ (2007) Characterization of two calmodulin-
like proteins from the liver fluke, Fasciola hepatica. Biol
Chem 388, 593–599.

3 Berasain P, Goni F, McGonigle S, Dowd A, Dalton JP,
Frangione B & Carmona C (1997) Proteinases secreted
by Fasciola hepatica degrade extracellular matrix and
basement membrane components. J Parasitol 83, 1–5.
4 Maizels RM & Yazdanbakhsh M (2003) Immune regu-
lation by helminth parasites: cellular and molecular
mechanisms. Nat Rev Immunol 2, 499–511.
5 Donnelly S, O’Neill SM, Stack CM, Robinson MW,
Turnbull L, Whitchurch C & Dalton JP (2010)
Helminth cysteine proteases inhibit TRIF-dependent
activation of macrophages via degradation of TLR3.
J Biol Chem 285, 3382–3392.
6 Hewitson J, Grainger J & Maizels RM (2009) Helminth
immunoregulation: the role of parasite secreted proteins
in modulating host immunity. Mol Biochem Parasitol
167, 1–11.
7 Robinson MW, Menon R, Donnelly SM, Dalton JP &
Ranganathan S (2009) An integrated transcriptomic
and proteomic analysis of the secretome of the helminth
pathogen, Fasciola hepatica: proteins associated with
invasion and infection of the mammalian host. Mol Cell
Proteomics 8, 1891–1897.
8 Silva E, Castro A, Lopes A, Rodrigues A, Dias C,
Conceic¸ a
˜
o A, Alonso J, Costa JMC, Bastos M, Parra F
et al. (2004) A recombinant antigen recognized by
Fasciola hepatica-infected hosts. J Parasitol 90, 746–751.
9 Ram D, Grossman Z, Markovics A, Avivi A, Lantner F
& Schechter I (1989) Rapid changes in the expression of a

gene encoding a calcium-binding protein in Schistosoma
mansoni. Mol Biochem Parasitol 34, 167–175.
10 Hu S, Law PK, Lv Z, Wu Z & Fung MC (2008) Molec-
ular characterization of a calcium-binding protein SjCa8
from Schistosoma japonicum. Parasitol Res 103, 1047–
1053.
11 Lv Z, Yang L, Hu S, Sun X, He H, Li Z, Zhou Y,
Fung M, Yu X, Zheng H et al. (2009) Expression pro-
file, localization of an 8-kDa calcium-binding protein
from Schistosoma japonicum (SjCa8), and vaccine
potential of recombinant SjCa8 (rSjCa8) against infec-
tions in mice. Parasitol Res 104, 733–743.
12 Francis P & Bickle Q (1992) Cloning of a 21.7-kDa
vaccine-dominant antigen gene of Schistosoma mansoni
reveals an EF hand-like motif. Mol Biochem Parasitol
50, 215–224.
13 Grabarek Z (2006) Structural basis for diversity of the
EF-hand calcium-binding proteins. J Mol Biol 359,
509–525.
14 Gifford JL, Walsh MP & Vogel HJ (2007) Structures
and metal-ion-binding properties of the Ca
2+
binding
helix–loop–helix EF-hand motifs. Biochem J 405, 199–
221.
15 Bunick CG, Nelson MR, Mangahas S, Hunter MJ,
Sheehan JH, Mizoue LS, Bunick GJ & Chazin WJ
(2004) Designing sequence to control protein in an
EF-hand protein. J Am Chem Soc 126, 5990–5998.
16 Mouawad L, Isrovan A, Quiniou E & Craescu CT

(2009) What determines the degree of compactness of a
calcium-binding protein? FEBS J 276, 1082–1093.
17 Chong S, Montello GE, Zhang A, Cantor EJ, Liao W,
Xu M-Q & Benner J (1998) Utilizing the C-terminal
cleavage activity of a protein splicing element to purify
recombinant proteins in a single chromatographic step.
Nucleic Acids Res 26, 5109–5115.
H. Fraga et al. FH8 from Fasciola hepatica
FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5083
18 Evenas J, Malmendal A, Thulin E, Carlstrom G &
Forsen S (1998) Ca
2+
binding and conformational
changes in a calmodulin domain. Biochemistry 37,
13744–13754.
19 Evenas J, Thulin E, Malmendal A, Forsen S & Carl-
strom G (1997) NMR studies of the E140Q mutant of
the carboxy-terminal domain of calmodulin reveal glo-
bal conformational exchange in the Ca
2+
-saturated
state. Biochemistry 36, 3448–3457.
20 Herzberg O, Moult J & James MN (1986) A model for
the Ca
2+
-induced conformational transition of tropo-
nin C. A trigger for muscle contraction. J Biol Chem
261, 2638–2644.
21 Vanscyoc WS & Shea MA (2001) Phenylalanine fluores-
cence studies of calcium binding to N-domain fragments

of Paramecium calmodulin mutants show increased cal-
cium affinity correlates with increased disorder. Protein
Sci 10, 1758–1768.
22 Papish AL, Tari LW & Vogel HJ (2002) Dynamic light
scattering study of calmodulin–target peptide com-
plexes. Biophys J 83, 1455–1464.
23 Ohki S, Ikura M & Zhang M (1997) Identification of
Mg
2+
-binding sites and role of Mg
2+
on target
recognition by calmodulin. Biochemistry 36, 4309–
4316.
24 Gopal B, Swaminathan CP, Bhattacharya A,
Murthy MRN & Surolia A (1997) Thermodynamics of
metal ion binding and denaturation of a calcium
binding protein from Entamoeba histolytica.
Biochemistry 36, 10910–10916.
25 Braga CACA, Pinto JR, Valente AP, Silva JL, Soren-
son MM, Foguel D & Suarez MC (2006) Ca
2+
and
Mg
2+
binding to weak sites of TnC C-domain induces
exposure of a large hydrophobic surface that leads to
loss of TnC from the thin filament. Int J Biochem Cell
Biol 38, 110–122.
26 Martin SR, Linse S, Johansson C, Bailey PM &

Forse
´
n S (1990) Protein surface charges and Ca
2+
binding to individual sites in calbindin D9K, stopped
flow stuties. Biochemistry 29, 4188–4193.
27 Black DJ, Tikunova SB, Johnson JD & Davis JP (2000)
Acid pairs increase the N-terminal Ca
2+
affinity of
CaM by increasing the rate of Ca
2+
association.
Biochemistry 39, 13831–13837.
28 Franz C, Durussel I, Cox JA, Schafer BW &
Heizmann CW (1998) Binding of Ca
2+
and Zn
2+
to
human nuclear S100A2 and mutant proteins. J Biol
Chem 273, 18826–18834.
29 Yang A, Miron S, Duchambon P, Assairi L, Blouquit Y
& Craescu CT (2006) The N-terminal domain of human
centrin 2 has a closed structure, binds calcium with a very
low affinity, and plays a role in the protein self-assembly.
Biochemistry 45, 880–889.
30 Lakowski T, Lee G, Okon M, Reid R & Mcintosh L
(2007) Calcium-induced folding of a fragment of
calmodulin composed of EF-hands 2 and 3. Protein Sci

16, 1119–1132.
31 Trave
´
G, Pastore A, Hyvonen M & Saraste M (1995)
The C-terminal domain of alpha-spectrin is structurally
related to calmodulin. Eur J Biochem 227, 35–42.
32 Honore
´
B & Vorum H (2000) The CREC family, a
novel family of multiple EF-hand, low affinity Ca
2+
-
binding proteins localised to the secretory pathway of
mammalian cells. FEBS Lett 466, 11–18.
33 Gagne
´
SM, Tsuda S, Li MX, Chandra M, Smillie LB &
Sykes BD (1994) Quantification of the calcium-induced
secondary structural changes in the regulatory domain
of troponin-C. Protein Sci 3, 1961–1974.
34 Leavis PC, Rosenfeld SS & Gergely J (1978) Proteo-
lytic fragments of troponin C. J Biol Chem 253, 5452–
5459.
35 Brzeska H, Venyaminov SV, Grabarek Z & Drabikow-
ski W (1983) Comparative studies on thermostability of
calmodulin, skeletal muscle troponin C and their tryptic
fragments. FEBS Lett 153, 169–173.
36 Julenius K, Thulin E & Finn BE (1998) Hydrophobic
core substitutions in calbindin D9K: effects on stability
and structure. Biochemistry 37, 8915–8925.

37 Strynadka NCJ & James MNG (1989) Crystal struc-
tures of the helix–loop–helix calcium-binding proteins.
Annu Rev Biochem 58, 951–998.
38 Schumacher MA, Crum M & Miller MC (2004) Crystal
structures of apocalmodulin and apocalmodulin ⁄ SK
potassium channel gating domain complex. Structure
12, 849–860.
39 O’Donnell SE, Newman RA, Witt TJ, Hultman R,
Froehlig JR, Christensen AP & Shea MA (2009) Ther-
modynamics and conformational change governing
domain–domain interactions of calmodulin. Methods
Enzymol 466, 503–526.
40 Arnold K, Bordoli L, Kopp J & Schwede T (2006) The
SWISS-MODEL workspace: a web-based environment
for protein structure homology modelling. Bioinformat-
ics 22 , 195–201.
41 Guex N & Peitsch MC (1997) SWISS-MODEL and the
Swiss-PdbViewer: an environment for comparative pro-
tein modeling. Electrophoresis 18, 2714–2723.
42 Zhang M, Li M, Wang JH & Vogel HJ (1994) The
effect of Met-Leu mutations on calmodulin’s ability to
activate cyclic nucleotide phosphodiesterase. J Biol
Chem 269, 15546–15552.
43 Nelson MR & Chazin WJ (1998) An interaction-based
analysis of calcium-induced conformational changes in
the Ca
2+
sensor proteins. Protein Sci 7 , 270–282.
44 Hanna R (1980) Fasciola hepatica: glicocalyx replace-
ment in the juvenile as a possible mechanism for

protection against host immunity. Exp Parasitol 50,
103–114.
45 Bradford MM (1976) A rapid and sensitive method for
the quantitation of microgram quantities of protein
FH8 from Fasciola hepatica H. Fraga et al.
5084 FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS
utilizing the principle of protein-dye binding. Anal Bio-
chem 72, 248–254.
46 Hill AV (1910) The possible effects of the aggregation
of the molecules of haemoglobin on its dissociation
curves. J Physiol 40, 4–7.
47 Hunter MJ & Chazin WJ (1998) High level expression
and dimer characterization of the S100 EF-hand pro-
teins, migration inhibitory factor-related proteins 8 and
14. J Biol Chem 273, 12427–12435.
48 Kyte J & Doolittle RF (1982) A simple method for dis-
playing the hydropathic character of a protein. J Mol
Biol 157, 105–132.
49 McWilliam H, Valentin F, Goujon M, Li W, Naray-
anasamy M, Martin J, Miyar T & Lopez R (2009) Web
services at the European Bioinformatics Institute –
2009. Nucleic Acids Res 37 , w6–w10.
50 Altschul SF, Gish W, Miller W, Myers EW & Lipman DJ
(2000) Basic local alignment search tool. J Mol Biol 215,
403–410.
51 Vassylyev DG, Takeda S, Wakatsuki S, Maeda K &
Maeda Y (1998) Crystal structure of troponin C in
complex with troponin I fragment at 2.3-A resolution.
Proc Natl Acad Sci USA 95, 4847–4852.
H. Fraga et al. FH8 from Fasciola hepatica

FEBS Journal 277 (2010) 5072–5085 ª 2010 The Authors Journal compilation ª 2010 FEBS 5085