Site-specific tagging proteins with a rigid, small and stable transition metal chelator, 8-hydroxyquinoline, for paramagnetic NMR analysis

Article abstract of DOI:10.1007/s10858-016-0011-7

Design of a paramagnetic metal binding motif in a protein is a valuable way for understanding the function, dynamics and interactions of a protein by paramagnetic NMR spectroscopy. Several strategies have been proposed to site-specifically tag proteins with paramagnetic lanthanide ions. Here we report a simple approach of engineering a transition metal binding motif via site-specific labelling of a protein with 2-vinyl-8-hydroxyquinoline (2V-8HQ). The protein-2V-8HQ adduct forms a stable complex with transition metal ions, Mn(II), Co(II), Ni(II), Cu(II) and Zn(II). The paramagnetic effects generated by these transition metal ions were evaluated by NMR spectroscopy. We show that 2V-8HQ is a rigid and stable transition metal binding tag. The coordination of the metal ion can be assisted by protein sidechains. More importantly, tunable paramagnetic tensors are simply obtained in an α-helix that possesses solvent exposed residues in positions i and i + 3, where i is the residue to be mutated to cysteine, i + 3 is Gln or Glu or i - 4 is His. The coordination of a sidechain carboxylate/amide or imidazole to cobalt(II) results in different structural geometries, leading to different paramagnetic tensors as shown by experimental data.

Full text of DOI:10.1007/s10858-016-0011-7

J Biomol NMR (2016) 64:103–113
DOI 10.1007/s10858-016-0011-7
ARTICLE
Site-specific tagging proteins with a rigid, small
and stable transition metal chelator, 8-hydroxyquinoline,
for paramagnetic NMR analysis
1
1
2
1
Yin Yang
•
Feng Huang
•
Thomas Huber
•
Xun-Cheng Su
Received: 18 November 2015 / Accepted: 1 January 2016 / Published online: 6 January 2016
Ó Springer Science+Business Media Dordrecht 2016
Abstract Design of a paramagnetic metal binding motif in
a protein is a valuable way for understanding the function,
dynamics and interactions of a protein by paramagnetic
NMR spectroscopy. Several strategies have been proposed
to site-specifically tag proteins with paramagnetic lanthanide
ions. Here we report a simple approach of engineering a
transition metal binding motif via site-specific labelling of a
protein with 2-vinyl-8-hydroxyquinoline (2V-8HQ). The
protein-2V-8HQ adduct forms a stable complex with tran-
sition metal ions, Mn(II), Co(II), Ni(II), Cu(II) and Zn(II).
The paramagnetic effects generated by these transition metal
ions were evaluated by NMR spectroscopy. We show that
Keywords Protein labeling Á 8-Hydroxyquinoline Á
Paramagnetic NMR spectroscopy Á Pseudocontact shift Á
Paramagnetic relaxation enhancement
Introduction
One of the mainstream uses of transition metals in struc-
tural biology is exploiting the diverse paramagnetic prop-
erties of these ions (Bertini and Luchinat 1996, 1999; Clore
and Iwahara 2009) by high resolution NMR spectroscopy.
The intrinsic paramagnetic properties of transition metal
ions provide valuable structural restraints in biomolecules,
which are governed by the well-established theory (Abra-
gam and Bleaney 1970; La Mar et al. 1973; Bertini and
Luchinat 1996). These paramagnetic effects, including
paramagnetic relaxation enhancement (PRE), pseudocon-
tact shift (PCS) and residual dipolar coupling (RDC), are
experimentally observable by NMR spectroscopy. In par-
ticular, PCSs as the chemical shift differences between the
paramagnetic and diamagnetic species can be measured
with high accuracy. While PREs only depend on the dis-
tance between the paramagnetic center and the observed
nuclear spins, PCSs contain both distance and direction
information of the nuclear spins with respect to the para-
magnetic center (Bertini et al. 2002; Clore and Iwahara
2
V-8HQ is a rigid and stable transition metal binding tag.
The coordination of the metal ion can be assisted by protein
sidechains. More importantly, tunable paramagnetic tensors
are simply obtained in an a-helix that possesses solvent
exposed residues in positions i and i ? 3, where i is the
residue to be mutated to cysteine, i ? 3 is Gln or Glu or
i - 4 is His. The coordination of a sidechain carboxylate/
amide or imidazole to cobalt(II) results in different structural
geometries, leading to different paramagnetic tensors as
shown by experimental data.
Electronic supplementary material The online version of this
article (doi:10.1007/s10858-016-0011-7) contains supplementary
material, which is available to authorized users.
2009; Otting 2010). Because of their high chemical simi-
larity and large variations in the paramagnetic anisotropy,
lanthanide ions have received increasing interests in
structural biology by means of high resolution NMR
spectroscopy (Geraldes 1993; Bertini et al. 2002; Geraldes
and Luchinat 2003; Pintacuda et al. 2007; Otting 2008;
Bertini et al. 2008; Otting 2010; Liu et al. 2014). As many
proteins do not contain a paramagnetic center, strategies for
site-specific labeling of proteins with paramagnetic species
&
Xun-Cheng Su
xunchengsu@nankai.edu.cn
1
State Key Laboratory of Elemento-Organic Chemistry,
Collaborative Innovation Center of Chemical Science and
Engineering (Tianjin), Nankai University, Tianjin 300071,
China
2
Research School of Chemistry, Australian National
University, Canberra, ACT 0200, Australia
123

1
04
J Biomol NMR (2016) 64:103–113
Fig. 1 Molecular structure of
V-8HQ
have been developed (Rodriguez-Casta n˜ eda et al. 2006; Su
and Otting 2010; Koehler and Meiler 2011; Liu et al. 2014).
Paramagnetic transition metal ions have an even longer
history in paramagnetic protein NMR spectroscopy (La Mar
et al. 1973; Bertini and Luchinat 1984; Bertini et al. 1993;
Bertini and Luchinat 1996; Donaire et al. 1998; Ubbink et al.
2
1
998; Bertini et al. 2005; Jensen et al. 2004; Arnesano et al.
005), however, site-specific labeling of biomolecules with a
2
A28C, respectively. The combination of a small functional
tag and protein sidechain in immobilizing the transition
metal ions was analyzed. The rigidity of the transition metal
ion with respect to the protein and the affinity of sidechainsof
glutamate, glutamine and histidine for the metal ion were
evaluated by high resolution NMR spectroscopy and
isothermal titration calorimetry (ITC). In the structure of
ubiquitin, the solvent exposed sidechains of residues Glu24,
Gln31 and Asp32 are close to Thr22 and Ala28, respectively,
and can assist in binding a metal ion in addition to 8HQ. As
imidazole is a preferred ligand for transition metal ions, the
double-point mutant E24H/A28C was constructed. The
PCSs of A28C-2V-8HQ and E24H/A28C-2V-8HQ com-
plexed with Co(II) differed significantly, suggesting differ-
ent paramagnetic tensors. Tunable paramagnetic anisotropy
is highly valuable in structural biology, and we show that
different sets of paramagnetic susceptibility tensors can
simply be obtained by changing the metal ion coordination
by straight-forward protein sidechain mutagenesis.
rigid transition metal ion is less studied (Gochin 1998; Man
et al. 2010; Nguyen et al. 2011). Complexes of high-spin
Co(II) produce the largest paramagnetic anisotropy and
Mn(II) complexes exert strong PRE effects on nuclear spins
because Mn(II) has five unpaired electrons and usually a long
electronic relaxation time (Bertini and Luchinat 1996).
Cu(II) forms more stable complexes but generally generates
weaker PREs than Mn(II). In the case of Ni(II), its param-
agnetic effects are largely determined by its coordination
with the chelating ligand. In comparison with lanthanide
ions, transition metal ions have fewer coordination sites
(
generally up to six) and prefer to coordinate with histidine
sidechains. This offers additional advantages in the design of
smaller and more rigid paramagnetic tags, as the stability and
rigidity of the tag strongly influences the interpretation of
paramagnetic restraints from NMR spectroscopy.
8-Hydroxyquinoline (8HQ), a frequently used organic
chelator for metal ions, forms stable complexes with
transition metal ions in varying molar ratio (Johnston and
Freiser 1952). Incorporation of 8HQ into a protein via
bioorthogonal protein synthesis has recently been reported
Materials and methods
(
Lee et al. 2009; Liu et al. 2013; Park et al. 2015). How-
ever, 8HQ is a bidentate ligand, i.e. it has only two coor-
dinating atoms. Therefore, one 8HQ moiety alone is not
able to immobilize the metal ion without additional coor-
dination by the protein, and aggregation due to the unsat-
urated coordination sites of the metal ion should be
prevented in the design of paramagnetic tags for NMR or
other spectroscopic analysis in structural biology. Several
examples have shown that simultaneous coordination by a
protein sidechain and a small functional tag can avert the
tag-induced protein aggregation and significantly restrain
the metal ion, thus suppressing the positional fluctuations
of the paramagnetic center (Su et al. 2008; Swarbrick et al.
Protein expression and purification
The target proteins were expressed using an optimized high-
density method (Cao et al. 2014) according to the previ-
ously reported protocol (Marley et al. 2001) and typically
1
5
20 mg N-protein was obtained from 250 mL M9 media.
Synthesis of 2V-8HQ
N-oxide-8-hydroxyquinoline (1): 0.5 g 8-hydroxyquinoline
was dissolved in 4 mL dichloromethane and 0.69 g meta-
chloroperbenzoic acid (m-CPBA) was added stepwise in an
ice-cooled water bath. The resulting mixture was stirred at
room temperature overnight. 7.14 mL 2 M NaOH was
added slowly to the above solution. The organic layer was
separated and the water phase was washed with dichlor-
omethane. The organic fractions were combined and
washed with brine, dried with sodium sulfate, and then
filtered. The organic solvent was removed under reduced
2
011; Yagi et al. 2013; Huang et al. 2013).
The high reactivity and chemoselectiviy of a 4-vinyl
pyridine moiety to thiol groups of proteins has been
demonstrated in protein modifications for paramagnetic
NMR and EPR analysis (Li et al. 2012; Yang et al. 2013; Qi
et al. 2014; Martorana et al. 2015). We herein report an 8HQ
derivative, 2-vinyl-8-hydroxyquinoline (2V-8HQ), which
can be site-specifically attached to a protein via a Michael
addition-like thiol-ene reaction (Fig. 1 and Scheme 1). To
test the reactivity of 2V-8HQ to protein thiols, we made two
single-point cysteine mutants of human ubiquitin, T22C and
pressure, resulting in 0.39 g title compound as yellow solid
1
(yield, 70.3 %) H NMR (300 MHz, CDCl ) d 8.25 (dd,
3
J = 6.0, 0.8 Hz, 1H), 7.80 (d, J = 8.1 Hz, 1H), 7.50 (t,
J = 8.0 Hz, 1H), 7.25–7.21 (m, 2H), 7.08 (dd, 1H).
1
23

J Biomol NMR (2016) 64:103–113
105
Scheme 1 Reagents and conditions: i m-CPBA/DCM, ii Ac
TBAF/PPh /DMF
3
2
O/Ar/100 °C, iii POBr
3 3 2
/Ar/CHCl , iv hydrolysis, v triethoxyvinylsilane/Pd(OAc) /
Compounds 2, 3 and 4 were synthesized according to
the published methods with minor modifications (Petitjean
et al. 2005).
then five equivalents of 2V-8HQ (100 mM in ethanol stock
solution) were added to the protein solution. It should be
noted that excess of TCEP is not desirable as TCEP reacts
with vinyl tags. The pH was adjusted to 7.8 using 1.0 M
NaOH and the mixture was incubated at room temperature
for about 10 h. The ligation product was purified by gel
filtration. The overall yield of purified ligation product was
approximately 80 %.
2
-Vinyl-8-hydroxyquinoline (2V-8HQ): similar to the
previously reported coupling reaction (Alacid and Najera
008), the mixture of 1.2 g (0.03 mol) NaOH,20 mg
0.09 mmol) Pd(OAc) ,2.28 g (0.01 mol) compound 4,
2
(
2
2
.53 mL (0.012 mmol) triethoxyvinylsilane, 30 mL H O and
2
3
0 g PEG 2000 was stirred at 80–90 °C for 10 h and the
resulting mixture was extracted with diethyl ether. The
organic phase was combined and washed with brine. The
organic solution was dried with sodium sulfate and filtered.
The organic solvent was removed under reduced pressure,
resulting in 2.0 g title compound as white powder (yield,
Results and discussion
2V-8HQ is a thiol-specific reaction reagent
for labeling proteins
1
1
8
8.9 %). H NMR ( H 400 MHz, CDCl ) d 7.95 (d,
3
1
The ligation reaction was monitored by recording N-
5
J = 8.6 Hz, 1H), 7.43 (d, J = 8.6 Hz, 1H), 7.27 (q,
J = 8.3 Hz, 1H), 7.15 (d, J = 8.2 Hz, 1H), 7.05 (d,
J = 7.5 Hz, 1H), 6.86 (dd, J = 17.7, 10.9 Hz, 1H), 6.18 (d,
J = 17.7 Hz, 1H), 5.52 (d, J = 10.9 Hz, 1H).
1
5
HSQC spectra of the reaction mixture with the N-ubiq-
uitin mutant protein with 2V-8HQ. Quantitative ligation of
protein and 2V-8HQ was achieved by incubating the pro-
tein with five equivalents of tag at room temperature, pH
7.8 in 20 mM Tris buffer for 10 h. The reaction of 2V-8HQ
with the ubiquitin mutants T22C, A28C and E24H/A28C
was completed in 10 h at pH 7.8 in 20 mM Tris buffer
NMR experiments
All NMR spectra of protein samples were recorded at
1
5
2
98 K on a Bruker AV600 NMR spectrometer equipped
(monitored by N-HSQC spectra), and the reactivity of
2V-8HQ towards free thiols is significantly higher than the
previously reported 4V-PyMTA (Yang et al. 2013) and
also 4V-DPA (Li et al. 2012). The tagging reaction rate
increases with pH and the ligation can be completed within
5 h at pH 8.5.
with a QCI-cryoprobe. The protein samples were in 20 mM
2
-(N-morpholino)ethanesulfonic acid (MES) buffer at pH
6
.4 unless indicated otherwise. The recycle delay for PRE
measurements was 2 s for both diamagnetic and param-
agnetic samples.
Mass analysis by MALDI-TOF mass spectrometry con-
firmed that proteins were conjugated with only one tag (Fig
S2), suggesting no additional groups, including the amino
group of the N-terminal methionine and amino groups of
lysine sidechains, reacted with 2V-8HQ. This is consistent
Site-specific labeling of ubiquitin mutant with 2V-
8
HQ
1
.5 mM solution of N-labeled protein in 20 mM tris(hy-
5
0
1
5
droxymethyl)aminomethane (Tris) pH 7.8 was first mixed
with 0.2 mM tris(2-carboxyethyl)phosphine (TCEP), and
with NMR chemical shift mapping in N-HSQC spectra
using free protein and its 2V-8HQ conjugates (Fig. 2 and
123

1
06
J Biomol NMR (2016) 64:103–113
1
Fig. 2 a Superimposition of N-HSQC spectra of 0.1 mM uniformly
5
1
5
15
Fig. 3 a Superimposition of
N-HSQC spectra of a 0.1 mM
N-labeled ubiquitin E24H/A28C (black) and E24H/A28C-2V-8HQ
2
1
5
uniformly N-labeled ubiquitin E24H/A28C-2V-8HQ in the presence
of one equivalent Zn(II) (black) and one equivalent Co(II) (red).
b PCSs of protein backbone amide protons generated by Co(II) in
complex with ubiquitin 2V-8HQ adducts of T22C (circles), A28C
conjugate (red). b Chemical shift differences d = ((Dd
1
H
) ? (Dd
0) ) ) between free protein and its 2V-8HQ conjugate: T22C (red
M
/
2
1/2
circles); A28C (blue triangles); E24H/A28C (black squares). Signif-
icant chemical shift changes were also observed for residues 50–60
that are close to T22 in the T22C-2V-8HQ adduct
(
triangles), E24H/A28C (squares). The PCSs were calculated as the
chemical shift differences between paramagnetic and diamagnetic
samples using Zn(II) as the diamagnetic reference
Fig. S1), as the chemical shift perturbations were mainly
located around the ligation site of the protein thiol. The high
specificity of the reactions of 2V-8HQ with protein thiols
suggests that this small tag can be applied to label proteins
via the formation of a stable thioether tether.
increased. The association–dissociation rate between the
metal bound and unbound protein complexes was slow on
the NMR time scale, indicating the formed metal complex
was stable. There was only a single paramagnetic species
corresponding to the complex formed between the protein-
2V-8HQ adduct and Co(II) ion during the titration. Simi-
larly, only a single paramagnetic species was observed for
the T22C-2V-8HQ and E24H/A28C-2V-8HQ adducts upon
addition of Co(II). The A28C-2V-8HQ and E24H/A28C-
2V-8HQ adducts complexed with Co(II) experienced sig-
nificantly larger chemical shift perturbations than T22C-2V-
8HQ (Fig. 3 and Fig S3), and therefore we mainly focus on
the A28C-2V-8HQ and E24H/A28C-2V-HQ conjugates in
the following. Exchange between Mn(II), Co(II), Ni(II),
Cu(II) and Zn(II) complexes of ubiquitin-2V-8HQ was also
Interaction of 2V-8HQ protein adducts
with transition metal ions
The interactions of transition metal ions with ubiquitin-2V-
8
HQ adducts were first analyzed by titration of Co(II) into
the solution of protein conjugates, because Co(II) has a large
magnetic anisotropy among the transition metal ions and has
been shown previously to be an excellent metal probe in
metalloproteins (Bertini and Luchinat 1984). Addition of
1
5
CoCl2 into the solution of 0.1 mM N-A28C-2V-8HQ
1
5
generated a new set of signals in the N-HSQC spectrum
with increasing intensities as the concentration of cobalt ion
1
5
slow as observed in the N-HSQC spectra.
1
23

J Biomol NMR (2016) 64:103–113
107
Binding affinity measurement of ubiquitin-2V-8HQ
with transition metal ions
complexes were similar to that of a previously published
histidine-coordinated protein complex (Jones et al. 2005).
Figure 3 shows sizable PCSs induced by the Co(II)
complex that produces larger chemical shift changes (up to
0.8 ppm) than any of the other transition metal ions.
Similarly, a large number of residues are observable in the
The association constants between transition metal ions
and protein-2V-8HQ adducts are difficult to measure by
NMR due to slow exchange between the bound and
unbound states of the protein. Therefore the thermody-
namic stabilities of protein-2V-8HQ complexed with
transition metal ion were analyzed by isothermal titration
calorimetry (ITC) (Fig S4). In 20 mM MES buffer at pH
1
5
N-HSQC spectra of ubiquitin-2V-8HQ complexed with
other metal ions. In the case of the copper(II) complex,
˚
proton spins within 10 A of the paramagnetic center are
generally inaccessible to NMR measurement (Arnesano
et al. 2003), but nonetheless a large number of residues are
6
.4 and 289 K, the binding constants of Co(II) and Ni(II)
1
5
were determined by ITC measurement. The interactions of
Mn(II), Cu(II) and Zn(II) ions with protein-2V-8HQ did
not result in a sigmoidal titration curve and the titration
curves of Mn(II), Cu(II) and Zn(II) failed to yield the
binding constants under the conditions used for Co(II) and
Ni(II). Despite the fact that A28C-2V-8HQ and E24H/
A28C-2V-8HQ showed slow exchange between free pro-
tein and metal complex, they differed greatly in binding
affinity with Co(II) and Ni(II) ions. The determined asso-
ciation constants, Ka, of E24H/A28C-2V-8HQ for Ni(II)
observable in N-HSQC spectra of the ubiquitin-2V-8HQ-
Cu(II) complex (Fig. 4). In the case of Mn(II), which
1
5
causes stronger PRE than Cu(II), several peaks in the N-
HSQC spectra of the ubiquitin-2V-8HQ Mn(II) complex
display line broadening, yet the same signals as with the
copper complex can be observed clearly.
The paramagnetic anisotropy was analyzed for the
protein complexes of Co(II) as the PCSs produced by
Cu(II) are small. As shown in Fig. 3b, the T22C-2V-8HQ
conjugate in complex with cobalt(II) generated smaller
PCSs than those of the A28C-2V-8HQ and E24H/A28C-
2V-8HQ conjugates. The different sizes of PCSs in the
three conjugates indicate different degrees of averaging.
Since mobility of either the coordination to the paramag-
netic center or of the tether linking 2V-8HQ and protein
will result in paramagnetic averaging, the sizes of the PCSs
and RDCs are sensitive to this flexibility. In the structure of
ubiquitin, Thr22 resides at the start of the first a-helix and
Ala28 is in the middle of the same helix. Smaller PCSs in
the T22C-2V-8HQ-Co(II) complex indicate a stronger
averaging of paramagnetic effects due to flexible coordi-
nation of the cobalt ion, implying that coordination with
the protein sidechain probably does not sufficiently restrain
the paramagnetic center in this case.
5
5
-1
and Co(II) are 4.8 9 10 and 4.0 9 10 M , respectively,
and are almost ten times larger than those of A28C-2V-
4
-1
4
-1
8
HQ (6.7 9 10 M with Ni(II) and 4.9 9 10 M with
Co(II)). The change of Glu 24 to His 24 in the ubiquitin-
V-8HQ adducts thus increased the binding affinity for the
2
transition metal ions by almost an order of magnitude,
supporting the notion that the sidechain imidazole is a
preferred ligand over a carboxylate.
Paramagnetic anisotropy of Co(II) complexed
with ubiquitin-2V-8HQ
The paramagnetic properties of transition metal ions bound
to protein-2V-8HQ conjugates were analyzed by high res-
olution NMR spectroscopy. PCSs were determined from the
chemical shift differences between the paramagnetic and
diamagnetic species (Fig. 3, Fig. S3 and Fig. S5), using
Zn(II) as the diamagnetic reference. Ni(II) has been
extensively applied in structural biology due to its aniso-
tropic paramagnetic properties (Jensen and Led 2006; Jen-
sen et al. 2004), however, its complex with protein-2V-8HQ
produced only small chemical shift changes in the present
study. As expected, the NMR resonances were also much
less attenuated by PREs compared with Co(II), Mn(II) and
In general, the PCS of a nuclear spin is described by
Eq. 1,
1
2pr3
Â
À
Á
Ã
2
2
PCS ¼ 1
Dvax 3 cos h À 1 þ 1:5Dvrh sin h cos 2/
ð1Þ
where r, h, and / are the polar coordinates of the nuclear
spin relative to the principal axes of the magnetic suscep-
tibility anisotropy tensor Dv, and Dvax and Dvrh are the
axial and rhombic components of Dv-tensor, respectively.
It is evident that reliable tensors can be determined pro-
vided that the paramagnetic center is accurately defined.
Using the program Numbat (Schmitz et al. 2008), the
magnetic susceptibility tensor was determined by using the
PCS data of backbone amide protons and the crystal
structure of ubiquitin (Ramage et al. 1994) or the RDC
refined NMR structure (Maltsev et al. 2014), respectively.
Both structures gave similar results (Table 1). Excellent
1
5
Cu(II) because most cross-peaks were observable in N-
HSQC spectra even for residues close to the ligation site
(
Fig S5). The negligible chemical shift changes and lack of
cross-peak attenuation observed with Ni(II) in the ubiquitin-
V-8HQ conjugates is likely due to the formation of a
2
square-planar complex which is diamagnetic because the
electrons occupying the 3d orbitals are fully paired. The
diamagnetic properties of the protein-2V-8HQ-Ni(II)
123

1
08
J Biomol NMR (2016) 64:103–113
1
5
Fig. 4 Superimposition of N-HSQC spectra of ubiquitin-2V-8HQ (red) complexed with Mn(II) and Cu(II). a 0.1 mM A28C-2V-8HQ
complexed with 0.1 mM Mn(II) and Cu(II) (black). b 0.05 mM E24H/A28C-2V-8HQ complexed with 0.05 mM Mn(II) and Cu(II) (black)
Table 1 Dv-tensor parameters of ubiquitin-2V-8HQ complexes with
Co(II) ion
The magnetic anisotropy determined for Co(II) bound to
the ubiquitin A28C-2V-8HQ adduct (Table 1) is close to
that of high-spin cobalt(II) in the blue copper protein azurin
A28C
E24H/A28C
T22C
(
Donaire et al. 1998), where the cobalt forms a tetrahedral
coordination complex. The cobalt complex in A28C-2V-
HQ is also similar to the West Nile virus NS2B-NS3
Dvax
Dvrh
a
8.3 (7.5)
-10.3 (-8.6)
-3.7 (-2.8)
79.4 (56.9)
141.2 (90.1)
66.5 (31.9)
-2.1 (-2.2)
-0.6 (-0.8)
90.6 (134.8)
35.7 (78.3)
85.5 (174.6)
5.0 (4.4)
8
64.1 (117.3)
23.7 (92.3)
91.2 (167.8)
protease (Nguyen et al. 2011), where the cobalt binds to a
genetically encoded bipyridylalanine assisted by additional
protein sidechains. These two high-spin cobalt complexes
produce similar magnitude of the Dv-tensors.
The Co(II) complex of the E24H/A28C-2V-8HQ adduct
generated larger PCSs than ubiquitin A28C-2V-8HQ, and
correspondingly the determined Dv-tensor is also larger
(Table 1). More importantly, the two cobalt complexes
produced different Dv-tensors both in magnitudes and
orientations. This confirms that the single point mutation of
E24H is effective in changing the paramagnetic tensor
significantly, suggesting different coordinating modes of
b
c
-
32
3
m and were determined
The tensor parameters are in units of 10
by fitting the PCS data of backbone amide protons to the crystal
structure of ubiquitin (PDB code: 1UBI) using program Numbat
(
Schmitz et al. 2008). For comparison, the tensor parameters by using
the RDC refined ubiquitin structure (PDB code: 2MJB, first con-
former) are shown in brackets
correlations between the experimental PCS and back-cal-
culated data are observed (Fig S6), indicating that the
calculated tensors are reliable.
1
23

J Biomol NMR (2016) 64:103–113
109
Fig. 5 Structural view of
ubiquitin-2V-8HQ complexed
with Co(II). a A28C-2V-8HQ.
b E24H/A28C-2V-8HQ. The
calculated paramagnetic center
is identified by a sphere
(
magenta) and the distances of
metal ion to sidechain amide
oxygen and nitrogen of
imidazole is indicated by a solid
line. The structure was
determined by PyParaTool
Co(II) complexes in these two protein adducts. From
structure modeling of the tag based on the PCS data it is
evident that the imidazole of histidine is coordinated to
cobalt in the E24H/A28C-2V-8HQ conjugate (Table 1;
Fig. 5b).
To obtain a structural view of the metal complex with
respect to the protein, explicit tag coordinates were cal-
culated using PyParaTool (Stanton-Cook et al.). For the
cobalt complex of A28C-2V-8HQ, the determined metal
position resides close to the sidechain oxygen of Q31 with
˚
a distance of 3.0 A, suggesting Q31 instead of D32 is
Error analysis of the paramagnetic tensors was performed
by random selection of 80 % of the PCSs data to fit the crystal
structure (Ramage et al. 1994; PDB code: 1UBI) and RDC-
refined NMR structure (Maltsev et al. 2014; PDB code:
coordinated to the cobalt ion in addition to 8HQ (Fig. 5). In
the structure of E24H/A28C-2V-8HQ, the sidechain of
histidine is more favorable to bind the cobalt ion. The
calculated distance of the cobalt ion to the sidechain
2
MJB). Both A28C-2V-8HQ and E24H/A28C-2V-8HQ
˚
nitrogen of His24 is 2.4 A as shown in Fig. 5b, which is in
adducts complexed with Co(II) gave high precision in tensor
determinations, and the outcomes of 100 calculations are
shown in Fig. S7. The Sanson-Flamsteed projections
demonstrate that the paramagnetic tensors of Co(II) com-
plexes of A28C-2V-8HQ and E24H/A28C-2V-8HQ differ
greatly in orientations (Fig S7). The different coordination
modes of E24H imidazole and E24, Q31 or D32 sidechain
carboxylate or amide are likely to cause different paramag-
netic anisotropy and change the angles of the principal axes of
the Dv-tensors. In the case of T22C-2V-8HQ, the fitted
paramagnetic tensors are more spread, suggesting that the
paramagnetic tensors are less reliable. This is because in the
T22C-2V-8HQ conjugate the dual coordination by 8HQ and a
protein sidechain (probably E24) is ill-matched to restrict the
flexibility of paramagnetic center, resulting in a loosely
coordinated metal complex. In the structure of ubiquitin, A28
resides in the middle of the first a-helix and between the
acidic residues E24 and D32 that reside in the i - 4 and i ? 4
positions, respectively. In addition, Q31 is located in the
i ? 3 position. In the case of the A28C-2V-8HQ conjugate,
good agreement with the typical coordination distance of
Co(II). These results are in agreement with a fluorescence
analysis of the zinc complex. The preferential coordination
of Zn(II) to histidine over carboxylate or amide sidechains
results in stronger fluorescence in E24H/A28C-2V-8HQ
than A28C-2V-8HQ (Fig S8). It is noted that in the T22C-
2V-8HQ and A28C-2V-8HQ adducts the pair of residues at
positions i and i ? 4, where i is Cys and i ? 4 is Glu or
Asp, is not an ideal combination to restrain the transition
metal ion. However, a segment in the helix, where residue i
is Cys, i ? 3 is Gln or Glu or i - 4 is His, is preferred to
immobilize the 2V-8HQ tag.
Comparison of PCSs and RDCs measured for Co(II)
complexes of ubiquitin-2V-8HQ adducts
Mobility of a paramagnetic tag generally arises from
chemical bond reorientations of the metal-chelating groups
and from motions of the linker between protein and para-
magnetic tag. The flexibility of the tag will cause different
averaging in RDCs and PCSs, because PCSs are distance
dependent whereas RDCs are not. RDCs are more sensitive
to the tag mobility than PCSs (Shishmarev and Otting
2013). To elucidate the flexibility of the tag in ubiquitin-
2V-8HQ, RDCs of protein backbone amides were mea-
sured for the Co(II) complex of ubiquitin E24H/A28C-2V-
8HQ. The alignment tensor was calculated by fitting the
RDCs to the crystal structure of ubiquitin (PDB code:
˚
the calculated paramagnetic center has a distance of 5.4 A to
the sidechain oxygen of D32 but with a shorter distance of
˚
about 3.0 A to the sidechain oxygen of Q31. In contrast, the
cobalt ion is closer to the side-chain of residue E24 with a
˚
distance of 5.0 A in the complex with E24H/A28C-2V-8HQ.
In both protein conjugates, the Q-factors are very small (Fig
S6). These results indicate that Q31 and E24H are involved in
coordinating the cobalt(II) ion in the A28C-2V-8HQ and
E24H/A28C-2V-8HQ adducts, respectively.
123

1
10
J Biomol NMR (2016) 64:103–113
Table 2 Comparison of alignment tensor and Dv-tensor parameters
of ubiquitin E24H/A28C-2V8HQ complexed with Co(II)
4
a
4
a
4
b
4
b
1
0 Aax
10 Arh
10 Aax
10 Arh
Co(II)
2.2
1.5
2.6 (2.2)
0.9 (0.7)
1
All data recorded at 25 °C and 600 MHz H NMR frequency
a
Alignment tensor determined using Module (Dosset et al. 2001).
Only RDCs of residues in regular secondary structure elements were
used in the fit (residues 2–8, 12–17, 41–45, 47–50, 65–70)
b
Alignment tensor predicated from the Dv-tensor values of Table 1
2
B
0
using Aax;rh ¼ 1
Dvax;rh
5kTl0
1
UBI) using the program Module (Dosset et al. 2001)
(
Table 2). Similar results were obtained with the NMR
structure (Maltsev et al. 2014). The alignment tensors
determined from RDCs and back-calculated from PCSs
were quite similar (Table 2). The result is consistent with
the FANTEN calculation (Rinaldelli et al. 2014) that can
directly determine the Dv-tensor parameters from the input
of RDC values (Table S1). These data confirm that 2V-
8
HQ is a rigid cobalt binding tag, which immobilizes the
paramagnetic center in a similar way as the published
MMDPA (Su et al. 2008) and 3MDPA (Man et al. 2010)
4
tags.
PRE analysis of Mn(II) and Cu(II) complexed
with 2V-8HQ adducts
Fig. 6 PREs of protein backbone amide protons detected as peak
1
5
intensity ratios in the N-HSQC spectra of protein samples in the
presence and absence of one equivalent paramagnetic Mn(II) (circle)
or Cu(II) (triangle). a 0.1 mM A28C-2V-8HQ, b 0.05 mM E24H/
A28C-2V-8HQ. Ipara and Idia are the cross-peak intensites for the
backbone amide protons in the presence and absence of paramagnetic
metal ion. The ratio Ipara/Idia is shown on the left and the calculated
distances of backbone amide protons from the paramagnetic center,
determined by fitting PCS data to the crystal structure of ubiquitin
PREs are extensively used in structural biology (Clore and
Iwahara 2009), but caution must be taken when interpreting
PREs because transient non-specific encounters between
the paramagnetic center and the protein surface can con-
tribute significant effects, requiring an assessment of
whether the presence of free metal ion in the protein
sample and non-specific transient associations between the
paramagnetic tag and the surface of the protein that con-
tributes to the observations. Several studies have shown
that ubiquitin itself binds transition metal ions (Arena et al.
(PDB code: 1UBI), are plotted as a line with the scale shown on the
right
in Fig. 4. In contrast, addition of Mn(II) into the solution of
protein-2V-8HQ only produced PREs on NMR signals. In
general, the residues most affected by PREs were vicinal to
the mutation site of A28C, indicating that the paramagnetic
center was located near that site. Assuming that Mn(II) and
Cu(II) share similar structural modes as Co(II) bound to
ubiquitin-2V-8HQ, the metal positions of Mn(II) and
Cu(II) were assumed to be the same as that of Co(II) in
these complexes. Correlations of peak intensity ratios
between the paramagnetic and diamagnetic cross-peaks
with respect to the amino acid sequence are shown in
Fig. 6. For comparison, the cross-peak attenuations with
respect to the distances of backbone amide protons to the
paramagnetic center are also shown in Fig. 6. The amide
2
011; Falini et al. 2008; Arnesano et al. 2011), therefore
the PRE effects on ubiquitin-2V-8HQ conjugates were also
assessed in this study.
The PRE effects were evaluated by comparison of the
1
5
changes in cross-peak intensity in N-HSQC spectra
between the diamagnetic (in the absence of metal ion) and
paramagnetic samples. The ubiquitin-2V-8HQ conjugates
formed complexes with Mn(II) and Cu(II) ions that were in
slow exchange on the NMR time scale. Titration of the
paramagnetic ions Mn(II) and Cu(II) into the solution of
ubiquitin-2V-8HQ produced different effects on the protein
signals. For example, formation of the copper(II) complex
with protein not only generated significant PREs but also
sizable chemical shift changes for some residues as shown
˚
protons within a 10 A radius of the paramagnetic center
1
23

J Biomol NMR (2016) 64:103–113
111
1
5
were generally undetectable in the N-HSQC spectra of
the A28C and E24H/A28C-2V-8HQ adducts complexed
with Mn(II) and Cu(II) ions. In the case of the A28C-2V-
is a useful paramagnetic and fluorescent tag that can be
used in structural biology by NMR and fluorescence
analysis. Selective modification of a single thiol group may
also be possible for a protein containing two or more
cysteines, provided that the cysteines have different solvent
accessibility and local mobility as those effects will greatly
affect the reactivity of protein thiols (Ma et al. 2014).
8
HQ conjugate, significant peak attenuations were also
observed for the residues Gln2, Ile3 and Gly76 at the N-
and C-termini, which are probably due to intermolecular
PREs at the protein concentration of 0.1 mM. In general,
the peak attenuations were in good agreement with the
calculated distances of the amide protons from the para-
magnetic center for both Mn(II) and Cu(II) complexes.
For the metal complexes of E24H/A28C-2V-8HQ, the
PREs of backbone amide protons correlated particularly
well with the distances from the paramagnetic center
Conclusions
We show that 8HQ, a well-known transition metal chelator,
can be site-specifically attached to a protein in a Michael
addition-like thiol-ene reaction. The protein-2V-8HQ
conjugates contain a stable thioether tether and show high
binding affinities with transition metal ions. The deter-
mined paramagnetic tensors of cobalt possess the character
of high-spin cobalt(II) complexes. The metal binding
affinity can be further enhanced in a helix (Yagi et al.
2013), if i is histidine and i ? 4 the tagged cysteine. Dif-
ferent paramagnetic tensors can simply be achieved by
changing the coordination of the paramagnetic ion by
suitable amino acid sidechains in the 2V-8HQ labeled
protein. In summary, 2V-8HQ offers an attractive way of
coordinating transition metal ions, producing a rigid,
stable and small transition metal binding tag that can be
used in structural analysis by NMR and also fluorescence
spectroscopy.
(
Fig. 6b). Compared with A28C-2V-8HQ, the cross-peak
attenuations determined for the Mn(II) and Cu(II) complexes
with E24H/A28C-2V-8HQ agreed well with the paramag-
netic center determined by the PCSs of the Co(II) complex,
indicating more similar structural modes associated with the
coordination of histidine sidechain than with carboxylates or
amides. It should be noted that the C-terminal flexible
residues also experienced significant PREs for the complex
of Mn(II), whereas the peak attenuations in the Cu(II)
complex matched the expectations equally well for these
C-terminal residues as for those in the rest of the protein.
The effect may have arisen from solvent PREs due to an
excess of Mn(II) ion during the titration. In both protein
samples, no indication of free paramagnetic ion bound to the
protein surfaces as shown previously (Arena et al. 2011;
Falini et al. 2008; Arnesano et al. 2011) was found.
Acknowledgments Financial support by the 973 program
Compared with published small paramagnetic tags, such
as 4MMDPA (Su et al. 2008), 3MDPA (Man et al. 2010),
(
(
(
2013CB910200), the National Science Foundation of China
21473095 and 21273121), and the Australian Research Council
DP120100561 and DP150100383) is greatly acknowledged.
4
MDPA (Jia et al. 2011), NTA (Swarbrick et al. 2011) and
IDA tags (Loh et al. 2015), which are suitable for lan-
thanide ions, 2V-8HQ is of similar size but has high
binding affinity for transition metal ions. In addition, C–S
thioether protein conjugates are stable and resistant to
reducing reagent and therefore can be used in NMR anal-
yses under reducing conditions. The cobalt(II) coordinated
References
Abragam A, Bleaney B (1970) Electron paramagnetic resonance of
transition ions. Oxford University Press, Oxford
Alacid E, Najera C (2008) Aqueous sodium hydroxide promoted
cross-coupling reactions of alkenyltrialkoxysilanes under ligand-
free conditions. J Org Chem 73:2315–2322
Arena G, Fattorusso R, Grasso G, Grasso GI, Isernia C, MalgieriG, Milardi
D, Rizzarelli E (2011) Zinc(II) complexes of ubiquitin: speciation,
affinity and binding features. Chem Eur J 17:11596–11603
Arnesano F, Banci L, Bertini I, Felli IC, Luchinat C, Thompsett AR
(2003) A strategy for the NMR characterization of type II
copper(II) proteins: the case of the copper trafficking protein
CopC from Pseudomonas syringae. J Am Chem Soc 125:
3
MDPA protein adduct reported previously showed sig-
nificantly smaller paramagnetic tensors compared with
A28C-2V-8HQ and E24H/A28C-2V-8HQ, implying a
different coordination of 8HQ to cobalt. As lanthanide ions
generally have coordination numbers of eight to ten, small
lanthanide binding tags do not easily afford strong binding
capability and generally fast exchange between protein and
protein–metal complex will be observed in NMR spectra,
which can result in significant solvent PREs and non-
specific association of free metal ion to the protein surface.
7
200–7208
Arnesano F, Banci L, Piccoli M (2005) NMR structures of
paramagnetic metalloproteins. Q Rev Biophys 38:167–219
Arnesano F, Belviso BD, Caliandro R, Falini G, Fermani S, Natile G,
Siliqi D (2011) Crystallographic analysis of metal-ion binding to
human ubiquitin. Chem Eur J 17:1569–1678
2
V-8HQ is not suitable for lanthanide ions but its advan-
tage lies in its strong binding affinity for transition metal
ions and fluorescent property in complex with a zinc ion,
which can be used in fluorescent energy transfer (FRET)
measurements. Due to its small size and rigidity, 2V-8HQ
Bertini I, Luchinat C (1984) High spin cobalt(II) as a probe for the
investigation of metalloproteins. Adv Inorg Biochem 6:71–111
123

1
12
J Biomol NMR (2016) 64:103–113
Bertini I, Luchinat C (1996) NMR of paramagnetic substances. Coord
Chem Rev 150:1–243
La Mar GN, Horrocks WD, Holm RH (1973) NMR of paramagnetic
molecules. Elsevier, Amsterdam
Bertini I, Luchinat C (1999) New applications of paramagnetic NMR
in chemical biology. Curr Opin Chem Biol 3:145–151
Bertini I, Turano P, Vila AJ (1993) Nuclear magnetic resonance of
metalloproteins. Chem Rev 93:2833–2932
Bertini I, Luchinat C, Parigi G (2002) Magnetic susceptibility in
paramagnetic NMR. Prog NMR Spectr 40:249–273
Lee HS, Spraggon G, Schultz PG, Wang F (2009) Genetic incorpo-
ration of a metal-ion chelating amino acid into proteins as a
biophysical probe. J Am Chem Soc 131:2481–2483
Li QF, Yang Y, Maleckis A, Otting G, Su XC (2012) Thiol-ene
reaction: a versatile tool in site-specific labelling of proteins with
chemically inert tags for paramagnetic NMR. Chem Commun
48:2704–2706
Liu X, Li J, Hu C, Zhou Q, Zhang W, Hu M, Zhou J, Wang J (2013)
Significant expansion of the fluorescent protein chromophore
through the genetic incorporation of a metal-chelating unnatural
amino acid. Angew Chem Int Ed Engl 52:4805–4809
Liu WM, Overhand M, Ubbink M (2014) The application of
paramagnetic lanthanoid ions in NMR spectroscopy on proteins.
Coord Chem Rev 273–274:2–12
Loh CT, Graham B, Abdelkader EH, Tuck KL, Otting G (2015)
Generation of pseudocontact shifts in proteins with lanthanides
using small ‘‘clickable’’ nitrilotriacetic acid and iminodiacetic
acid tags. Chem Eur J 21:5084–5092
Ma FH, Chen JL, Li QF, Zuo HH, Huang F, Su XC (2014) Kinetic
assay of the Michael addition-like thiol-ene reaction and insight
into protein bioconjugation. Chem Asian J 9:1808–1816
Maltsev AS, Grishaev A, Roche J, Zasloff M, Bax A (2014) Improved
cross validation of a static ubiquitin structure derived from high
precision residual dipolar couplings measured in a drug-based
liquid crystalline phase. J Am Chem Soc 136:3752–3755
Man B, Su XC, Liang H, Simonsen S, Huber T, Messerle BA, Otting
G (2010) 3-Mercapto-2,6-pyridinedicarboxylic acid: a small
lanthanide-binding tag for protein studies by NMR spectroscopy.
Chem Eur J 16:3827–3832
Marley J, Lu M, Bracken C (2001) A method for efficient isotopic
labeling of recombinant proteins. J Biomol NMR 20:71–75
Martorana A, Yang Y, Zhao Y, Li QF, Su XC, Goldfarb D (2015)
Mn(II) tags for DEER distance measurements in proteins via C–
S attachment. Dalton Trans 44:20812–20816
Bertini I, Luchinat C, Parigi G, Pierattelli R (2005) NMR spec-
troscopy of paramagnetic metalloproteins. Chembiochem
6
:1536–1549
Bertini I, Luchinat C, Parigi G, Pierattelli R (2008) Perspectives in
paramagnetic NMR of metalloproteins. Dalton 3782–3790
Cao C, Chen JL, Yang Y, Huang F, Otting G, Su XC (2014) Selective
1
5
N-labeling of the side-chain amide groups of asparagine and
glutamine for applications in paramagnetic NMR spectroscopy.
J Biomol NMR 59:251–261
Clore GM, Iwahara J (2009) Theory, practice, and applications of
paramagnetic relaxation enhancement for the characterization of
transient low-population states of biological macromolecules and
their complexes. Chem Rev 109:4108–4139
Donaire A, Salgado J, Moratal JM (1998) Determination of the magnetic
1
axes of cobalt(II) and nickel(II) azurins from H NMR data:
influence of the metal and axial ligands on the origin of magnetic
anisotropy in blue copper proteins. Biochemistry 37:8659–8673
Dosset P, Hus JC, Marion D, Blackledge M (2001) A novel
interactive tool for rigid-body modeling of multi-domain
macromolecules using residual dipolar couplings. J Biomol
NMR 20:223–231
Falini G, Fermani S, Tosi G, Arnesano F, Natile G (2008) Structural
probing of Zn(II), Cd(II) and Hg(II) binding to human ubiquitin.
Chem Commun 5980–5962
Geraldes CF (1993) Lanthanide shift reagents. Methods Enzymol
2
27:43–78
Geraldes CF, Luchinat C (2003) Lanthanides as shift and relaxation
agents in elucidating the structure of proteins and nucleic acids.
Met Ions Biol Syst 40:513–588
Nguyen THD, Ozawa K, Stanton-Cook M, Barrow R, Huber T, Otting
G (2011) Generation of pseudocontact shifts in protein NMR
spectra with a genetically encoded cobalt(II)-binding amino acid.
Angew Chem Int Ed 50:692–694
Otting G (2008) Prospects for the lanthanides in structural biology by
NMR. J Biomol NMR 42:1–9
Gochin M (1998) Nuclear magnetic resonance characterization of a
paramagnetic DNA-drug complex with high spin cobalt; assign-
1
31
ment of the H and P NMR spectra, and determination of
electronic, spectroscopic and molecular properties. J Biomol
NMR 12:243–257
Huang F, Pei YY, Zuo HH, Chen JL, Yang Y, Su XC (2013)
Bioconjugation of proteins with a paramagnetic NMR and
fluorescent tag. Chem Eur J 19:17141–17149
Jensen MR, Led JJ (2006) Metal–protein interactions: structure
information from Ni(II)-induced pseudocontact shifts in a native
nonmetalloprotein. Biochemistry 45:8782–8787
Otting G (2010) Protein NMR using paramagnetic ions. Annu Rev
Biophys 39:387–405
Park SH, Wang VS, Radoicic J, De Angelis AA, Berkamp S, Opella
SJ (2015) Paramagnetic relaxation enhancement of membrane
proteins by incorporation of the metal-chelating unnatural amino
acid 2-amino-3-(8-hydroxyquinolin-3-yl)propanoic acid (HQA).
J Biomol NMR 61:185–196
Petitjean A, Kyritsakas N, Lehn JM (2005) Ion-triggered multistate
molecular switching device based on regioselective coordina-
tion-controlled ion binding. Chem Eur J 11:6818–6828
Pintacuda G, John M, Su XC, Otting G (2007) NMR structure
determination of protein–ligand complexes by lanthanide label-
ling. Acc Chem Res 40:206–212
Jensen MR, Lauritzen C, Dahl SW, Pedersen J, Led JJ (2004) Binding
studied by
2?
ability of a HHP-tagged protein towards Ni
paramagnetic NMR relaxation: the possibility of obtaining
long-range structure information. J Biomol NMR 29:175–185
0
Jia X, Maleckis A, Huber T, Otting G (2011) 4,4 -Dithiobisdipicolinic
acid: a small and convenient lanthanide binding tag for protein
NMR spectroscopy. Chem Eur J 17:6830–6836
Johnston WD, Freiser H (1952) Structure and behavior of organic
analytical reagents. III. Stability of chelates of 8-hydroxyquino-
line and analogous reagents. J Am Chem Soc 74:5239–5242
Jones CE, Klewpatinond M, Abdelraheim SR, Brown DR, Viles JH
Qi A, Gross A, Jeschke G, Godt A, Drescher M (2014) Gd(III)-
PyMTA label is suitable for in-cell EPR. J Am Chem Soc
136:15366–15378
Ramage R, Green J, Muir TW, Ogunjobi OM, Love S, Shaw K (1994)
Synthetic, structural and biological studies of the ubiquitin
system: chemically synthesized and native ubiquitin fold into
identical three-dimensional structures. Biochem J 299:151–158
Rinaldelli M, Carlon A, Ravera E, Parigi G, Luchinat C (2014)
FANTEN: a new web-based interface for the analysis of
(
2005) Probing copper binding to the prion protein using
1
diamagnetic nickel and H NMR: the unstructured N terminus
facilitates the coordination of six copper ions at physiological
concentrations. J Mol Biol 346:1393–1407
Koehler J, Meiler J (2011) Expanding the utility of NMR restraints
with paramagnetic compounds: background and practical
aspects. Prog NMR Spectr 59:360–389
magnetic anisotropy-induced NMR data.
61:21–34
J Biomol NMR
1
23

J Biomol NMR (2016) 64:103–113
113
Rodriguez-Casta n˜ eda F, Haberz P, Leonov A, Griesinger C (2006)
Paramagnetic tagging of diamagnetic proteins for solution NMR.
Magn Reson Chem 44:S10–S16
Shishmarev D, Otting G (2013) How reliable are pseudocontact shifts
induced in proteins and ligands by mobile paramagnetic metal
tags? A modelling study. J Biomol NMR 56:203–216
Schmitz C, Stanton-Cook MJ, Su XC, Otting G, Huber T (2008)
Numbat: an interactive software tool for fitting Dv-tensors to
molecular coordinates using pseudocontact shifts. J Biomol
NMR 41:179–189
Swarbrick JD, Ung P, Su XC, Maleckis A, Chhabra S, Huber T,
Otting G, Graham B (2011) Engineering of a bis-chelator motif
into a protein a-helix for rigid lanthanide binding and param-
agnetic NMR spectroscopy. Chem Commun (Camb) 47:
7368–7370
Ubbink M, Ejdeb a¨ ck M, Karlsson BG, Bendall DS (1998) The
structure of the complex of plastocyanin and cytochrome f,
determined by paramagnetic NMR and restrained rigid-body
molecular dynamics. Structure 6:323–335
Yagi H, Maleckis A, Otting G (2013) A systematic study of labelling
an a-helix in a protein with a lanthanide using IDA-SH or NTA-
SH tags. J Biomol NMR 55:157–166
Stanton-Cook MJ, Su XC, Otting G, Huber T. http://compbio.anu.edu.
au/mscook/PPT/
Su XC, Man B, Beeren S, Liang H, Simonsen S, Schmitz C, Huber T,
Messerle BA, Otting G (2008) A dipicolinic acid tag for rigid
lanthanide tagging of proteins and paramagnetic NMR spec-
troscopy. J Am Chem Soc 130:10486–10487
Yang Y, Li QF, Cao C, Huang F, Su XC (2013) Site-specific labeling
of proteins with a chemically stable, high-affinity tag for protein
study. Chem Eur J 19:1097–1103
123