Bioconjugation of proteins with a paramagnetic NMR and fluorescent tag

Article abstract of DOI:10.1002/chem.201302273

Site-specific labeling of proteins with lanthanide ions offers great opportunities for investigating the structure, function, and dynamics of proteins by virtue of the unique properties of lanthanides. Lanthanide-tagged proteins can be studied by NMR, X-ray, fluorescence, and EPR spectroscopy. However, the rigidity of a lanthanide tag in labeling of proteins plays a key role in the determination of protein structures and interactions. Pseudocontact shift (PCS) and paramagnetic relaxation enhancement (PRE) are valuable long-range structure restraints in structural-biology NMR spectroscopy. Generation of these paramagnetic restraints generally relies on site-specific tagging of the target proteins with paramagnetic species. To avoid nonspecific interaction between the target protein and paramagnetic tag and achieve reliable paramagnetic effects, the rigidity, stability, and size of lanthanide tag is highly important in paramagnetic labeling of proteins. Here 4′-mercapto-2, 2′: 6′,2′′-terpyridine-6,6′′-dicarboxylic acid (4MTDA) is introduced as a a rigid paramagnetic and fluorescent tag which can be site-specifically attached to a protein by formation of a disulfide bond. 4MTDA can be readily immobilized by coordination of the protein side chain to the lanthanide ion. Large PCSs and RDCs were observed for 4MTDA-tagged proteins in complexes with paramagnetic lanthanide ions. At an excitation wavelength of 340 nm, the complex formed by protein-4MTDA and Tb³⁺ produces high fluorescence with the main emission at 545 nm. These interesting features of 4MTDA make it a very promising tag that can be exploited in NMR, fluorescence, and EPR spectroscopic studies on protein structure, interaction, and dynamics. Rigid and small: Rigidity and small size of lanthanide-binding tags are critically important in site-specific labeling of proteins with lanthanide ions. 4′-Mercapto-2,2′:6′,2′′-terpyridine-6, 6′′-dicarboxylic acid (4MTDA) provides five coordination sites for lanthanide ions and can be readily immobilized by coordination of proteins (see figure).

Full text of DOI:10.1002/chem.201302273

FULL PAPER
DOI: 10.1002/chem.201302273
Bioconjugation of Proteins with a Paramagnetic NMR and Fluorescent Tag
Feng Huang, Ying-Ying Pei, Hui-Hui Zuo, Jia-Liang Chen, Yin Yang, and
Xun-Cheng Su*[a]
Abstract: Site-specific labeling of pro-
teins with lanthanide ions offers great
opportunities for investigating the
structure, function, and dynamics of
proteins by virtue of the unique prop-
erties of lanthanides. Lanthanide-
tagged proteins can be studied by
NMR, X-ray, fluorescence, and EPR
spectroscopy. However, the rigidity of
a lanthanide tag in labeling of proteins
plays a key role in the determination
of protein structures and interactions.
Pseudocontact shift (PCS) and para-
relies on site-specific tagging of the
target proteins with paramagnetic spe-
cies. To avoid nonspecific interaction
between the target protein and para-
magnetic tag and achieve reliable para-
magnetic effects, the rigidity, stability,
and size of lanthanide tag is highly
important in paramagnetic labeling
of proteins. Here 4’-mercapto-2,2’:
6’,2’’-terpyridine-6,6’’-dicarboxylic acid
(4MTDA) is introduced as a a rigid
paramagnetic and fluorescent tag
which can be site-specifically attached
to a protein by formation of a disulfide
bond. 4MTDA can be readily immobi-
lized by coordination of the protein
side chain to the lanthanide ion. Large
PCSs and RDCs were observed for
4MTDA-tagged proteins in complexes
with paramagnetic lanthanide ions. At
an excitation wavelength of 340 nm,
the complex formed by protein–
4MTDA and Tb3+ produces high fluo-
rescence with the main emission at
545 nm. These interesting features of
4MTDA make it a very promising tag
that can be exploited in NMR, fluores-
cence, and EPR spectroscopic studies
on protein structure, interaction, and
dynamics.
magnetic
relaxation
enhancement
(PRE) are valuable long-range struc-
ture restraints in structural-biology
NMR spectroscopy. Generation of
these paramagnetic restraints generally
Keywords: fluorescence
nides · NMR spectroscopy · N,O
ligands · proteins
· lantha-
AHCTUNGTRENNUNG
Introduction
larges the region of the paramagnetic center relative to the
protein surface and hence averages the paramagnetic ef-
fects.[2d] Immobilizing the paramagnetic tag is thus a key
issue in site-specific labeling of proteins.[2b,d] Several success-
ful strategies have been proposed for restriction of paramag-
netic tags, including increasing the size of the tag,[3] the
double-point ligation method,[3b,4] anchoring the tag by coor-
dination of a protein,[5] and preventing rotamer transitions
of a tag.[5c,6]
The DOTA-like paramagnetic tags and lanthanide-bind-
ing peptides usually generate large paramagnetic effects,[3]
but they are bulky and their molecular weights are generally
greater than 600 Da. 4-Methylmercapto dipicolinic acid
(4MMDPA),[5a] 3-mercapto dipicolinic acid (3MDPA),[5b]
and 4-mercapto dipicolinic acid (4MDPA)[5c] are small but
rigid paramagnetic lanthanide tags; however, they only have
three coordinating atoms. As lanthanide ions usually have
high coordination numbers (usually 8 or 9),[7] sufficient coor-
dination numbers are desirable in the design of lanthanide-
binding tags. Therefore, increasing the number of coordina-
tion sites and minimizing the size of lanthanide tags are still
challenging in tagging proteins with lanthanide ions.
Tagging proteins with lanthanide ions offers great opportu-
nities for studying biological structures by means of X-ray,
NMR, and fluorescence spectroscopy.[1] Lanthanide-tagged
proteins have attracted great interest in paramagnetic NMR
spectroscopy, because the dipolar interaction of unpaired
electron and nuclear spins of proteins are rich sources of
structural restraints.[1] These paramagnetic effects on pro-
teins usually include paramagnetic relaxation enhancement
(PRE), pseudocontact shift (PCS), and residual dipolar cou-
pling (RDC).[1a] Since most proteins have no natural binding
site for paramagnetic ions, generation of these paramagnetic
restraints generally relies on site-specific labeling of proteins
with paramagnetic tags.[2a–c] However, the flexibility of the
paramagnetic tag compromises the quality and accuracy of
the three-dimensional structures determined by paramagnet-
ic NMR spectroscopy, because the mobility of the tag en-
[a] F. Huang,+ Y.-Y. Pei,+ H.-H. Zuo, J.-L. Chen, Y. Yang, Prof. X.-C. Su
State Key Laboratory of Elemento-organic Chemistry
and College of Chemistry, Nankai University
Weijin Road 94, Tianjin 300071 (P. R. China)
Fax : (+86)22-23500623
E-mail: xunchengsu@nankai.edu.cn
[+] These authors contributed equally to this work.
Moreover, lanthanide complexes have unique photophysi-
cal properties that are widely used in biomedical anal-
AHCTUNGTRENNUNG
ysis,[1e,8a] and tagging a protein with a luminescent lantha-
nide complex is a valuable tool in studies on the functions
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302273.
and dynamics of proteins.[1e,8b] Attachment of a fluorescent
Chem. Eur. J. 2013, 19, 17141 – 17149
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
17141

tag to a protein proceeds by a similar chemical reaction to
paramagnetic labeling of proteins. However, the linker be-
tween protein and fluorescent tag is usually long (>10 ꢁ)
and flexible, which greatly decreases the precision of the dis-
tance between two fluorophores during protein interac-
tions.[9] Site-specific labeling of proteins with a rigid and
small tag can therefore improve the interpretation of mea-
sured results both in paramagnetic NMR and fluorescence
studies on proteins.
NMR and by fluorescence spectroscopy. These interesting
features make 4MTDA a very promising lanthanide tag in
studies on the structures, interactions, and dynamics of pro-
teins and protein complexes by NMR and fluorescence spec-
troscopy.
Results
2,2’:6’,2’’-Terpyridine and its
Synthesis of 4MTDA: The synthesis of the 4MTDA is de-
picted in Scheme 1. Starting from commercially available di-
picolinic acid, compound 5 was synthesized according to
a previous report.[13] The corresponding thiouronium bro-
mide 6 was obtained from 5 and thiourea in refluxing ace-
tone, similar to the published protocol.[14] The final product
4MTDA was obtained by hydrolysis of 6 in methanol/water.
derivatives are well-known
metal-binding ligands.[10] The
lanthanide complexes of terpyr-
idine derivatives have been
used as fluorescence probes or
tags.[11] Herein, we synthesized
4’-mercapto-2,2’:6’,2’’-terpyri-
dine-6,6’’-dicarboxylic
acid
Interaction of 4MTDA with paramagnetic metal ions: NMR
spectroscopy was used to explore the stoichiometry and sta-
bility of the complexes formed by 4MTDA and lanthanide
ions in aqueous solution. Addition of paramagnetic Nd3+ to
a solution of 4MTDA in D2O at pD 7.0 produced a new set
of signals indicating slow exchange between free ligand and
metal complex. Four sets of resonances were observed for
Figure 1. Structure of 4MTDA.
(4MTDA, Figure 1) and at-
tached it to human ubiquitin
mutants by formation of a disul-
fide bond.[12]
The molecular weight of 4MTDA of 353 Da is lower than
those of commonly used lanthanide-binding peptides and
DOTA-like tags[3] (Supporting Information, Figure S1). To
explore the rigidity and stability of the 4MTDA tag, two cys-
teine mutants of human ubiquitin, T22C and A28C, were
made and subsequently derivatized with 4MTDA. Signifi-
cant pseudocontact shifts (PCSs) and residual dipolar cou-
plings (RDCs) were observed in the 4MDTA-tagged pro-
teins in complexes with paramagnetic ions. Interestingly, the
lanthanide complex showed no heterogeneity in the protein
samples, since no diastereomers were observed in the 15N
HSQC spectra. More importantly, the paramagnetic ion is
immobilized by the five coordinating atoms of 4MTDA and
one carboxylate side chain of an acidic amino acid in a pro-
tein. In comparison with the published lanthanide tags,
4MTDA has several advantages: 1) 4MTDA is small and
rigid; 2) 4MTDA-tagged proteins can be studied both by
the complex [Nd
tion of Nd3+, four new peaks corresponding to [Nd-
(4MTDA]+ were generated. Interestingly, at a [Nd3+]/-
[4MTDA] molar ratio of about 0.44, a large amount of [Nd-
(4MTDA)2]ꢀ. With increasing concentra-
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
AHCTUNGTRENNUNG
ACHUTNGRENNUG CAHTUNGTRENNUGN
(4MTDA]+ coexists with [Nd(4MTDA)2]ꢀ (Figure 2). Addi-
tion of Yb3+ to a solution of 4MTDA resulted in opposite
chemical shifts in comparison with Nd3+. Slow exchange be-
tween free ligand, [YbACTHNUGTRENNUG ACHUTNGTRENNGUN
(4MTDA)2]ꢀ, and [Yb(4MTDA)]+
was also observed (Figure 3). These interesting results indi-
cate that lanthanides bind the first 4MTDA ligand more
tightly than additional 4MTDA in the formation of
[LnACHTNUTGRNEUNG
(4MTDPA)2]ꢀ complexes.
Accurate binding constants of 4MTDA and lanthanide
ions were not determined by NMR spectroscopy due to
slow exchange between the free ligand and its metal com-
Scheme 1. Synthesis of 4MTDA. i) SOCl2/EtOH; ii) acetone, NaH, THF, 40-458C; iii) NH4OAc/EtOH/rf; iv) PBr5/DMF/Ar/608C; v) thiourea/acetone/rf;
vi) 1) NaOH/MeOH, 2) HCl/H2O.
17142
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17141 – 17149

Bioconjugation of Proteins
FULL PAPER
previously established protocol.[12] Since the sulfhydryl
group of 4MTDA has a lower pKa value than 4MMDPA,
4MTDA is less reactive than the published lanthanide bind-
ing tag (LBT)[3a,4b,12] and DPA tags.[5] Therefore, eight equiv-
alents of 4MTDA were necessary to give a reasonable
amount of protein–4MTDA adducts. The general ligation
yield is about 50–70%, and it is lower than that of
4MMDPA, LBT, and 3MDPA.
Stability of protein–4MTDA and lanthanide complexes: To
evaluate the thermodynamic stability of 4MTDA-tagged
proteins and lanthanide complexes, we titrated dipicolinic
acid (DPA) into a mixture of ubiquitin–4MTDA and lantha-
nide ions. Addition of 0.12 mm DPA to a solution of 0.1 mm
ubiquitin A28C–4MTDA and 0.12 mm Tm3+ at pH 6.4 had
a negligible effect on the 15N HSQC spectrum of the protein
sample. The paramagnetic species corresponding to the
ubiquitin–4MTDA–Tm complex remained essentially un-
changed. When 0.20 mm DPA was added, the paramagnetic
species only decreased by 20% in peak intensity. A few
peaks showed obvious chemical-shift changes and decreased
peak intensity, which probably stem from noncovalent inter-
Figure 2. Interaction of 4MTDA with Nd3+ determined by NMR spec-
troscopy. A 10 mm aqueous solution of NdCl3 was titrated into a mixture
of 0.8 mm 4MTDA and 2.0 mm dithiothreitol (DTT) in D2O at pD 7.0.
The NMR spectra were recorded at a proton frequency of 600 MHz with
increasing NdCl3 concentration: a) 0, b) 0.18, c) 0.35, d) 0.53, e) 0.88, and
f) 1.2 mm NdCl3. The asterisk indicates signals of free 4MTDA, and cir-
cles indicate those of [NdACHTNUGRTNEUNG
(4MTDA)]+.
action of [TmACTHNUTRGNEUNG
(DPA)3]3ꢀ and the protein.[15] This competition
experiment suggests that 4MTDA-tagged protein forms
highly stable lanthanide complexes in aqueous solution
(Supporting Information, Figure S3).
Resonance assignments of the protein–4MTDA adducts:
The NMR resonances of protein backbone amido groups
were assigned on the basis of 3D 15N NOESY-HSQC and
15N HSQC spectra. Both protein constructs show high-quali-
ty NMR spectra. In comparison with the underivatized pro-
tein, the chemical-shift changes in ubiquitin–4MTDA con-
structs were limited to the amido groups in the vicinity of
T22C and A28C (Figure 4). The magnitudes of the chemi-
cal-shift changes were small and most changes were within
0.10 ppm. The secondary structural segments were retained
according to a comparison of NOESY-HSQC and protein
Figure 3. Interaction of 4MTDA with Yb3+ determined by NMR spec-
troscopy. NMR conditions were the same as in Figure 2. a) 0, b) 0.18,
c) 0.35, d) 0.53, e) 0.88, and f) 1.2 mm YbCl3. The asterisk indicates the
signals of free 4MTDA, and circles indicate those of [YbACTHNUTRGNEUNG
(4MTDA)]+.
plexes. Isothermal titration calorimetry (ITC) was carried
out to determine the association constants of 4MTDA and
lanthanides. At 286 K and pH 6.4 in 20 mm 2-(N-morpholi-
no)ethanesulfonic acid (MES) buffer, association constants
of Lu3+ for the first and second 4MTDA ligands of
(1.4ꢁ0.2)ꢂ107 mꢀ1 and (2.5ꢁ0.2)ꢂ106 mꢀ1 were determined,
respectively (Supporting Information, Figure S2).
Figure 4. Chemical-shift differences between ubiquitin and ubiquitin–
4MTDA. The chemical shifts were calculated as Dd=[(DdH)2 +(DdN/
10)2]1/2
A28C–4MTDA.
. Triangles: T22C and T22C–4MTDA; circles: A28C and
Ligation of proteins with 4MTDA: Chemical modification
of proteins with 4MTDA was carried out similarly to the
Chem. Eur. J. 2013, 19, 17141 – 17149
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17143

X.-C. Su et al.
structure in both protein–4MTDA samples, which suggests
that the introduction of 4MTDA causes negligible changes
in protein structure.
coordinates differ very little compared to the distance from
the paramagnetic center. The paramagnetic cross-peaks are
thus displaced along approximately parallel lines from their
diamagnetic partner. The PCSs were calculated as chemical
shifts of the backbone amide protons in the paramagnetic
samples minus those of diamagnetic sample.
PCS measurements and Dc tensors: Significant chemical-
shift changes were observed on addition of paramagnetic
Dy3+, Tb3+, Tm3+, or Yb3+ to a solution of 15N-ubiquitin-
4MTDA (Figure 5 and Supporting Information, Figures S4
and S5). All of the paramagnetic metal ions resulted in
high-quality NMR spectra. For each backbone amido group,
a single cross-peak was observed in the presence of diamag-
netic Y3+ or any of the paramagnetic ions. No heterogeneity
of conformational exchange in the lanthanide-loaded pro-
tein samples was determined.
Determination of Dc tensors: The PCSs measured for the
ubiquitin–4MTDA complexes with Tb3+, Dy3+, Tm3+, and
Yb3+ were used to determine the paramagnetic Dc tensors.
Four data sets of PCSs for the residues in well-structured re-
gions were applied in simultaneous fit of the paramagnetic
tensors to the crystal structure of ubiquitin. The calculated
Dc-tensor parameters are listed in Table 1. The NMR struc-
ture was used to evaluate the tensor calculation by using the
same PCSs data, and it gave very similar results. Notably,
PCSs including those of unstructured residues lead to larger
calculated paramagnetic tensors.
Table 1. Dc-tensor parameters [10ꢀ32 m3] of ubiquitin–4MDTA com-
plexed with Dy3+, Tb3+, Tm3+, and Yb3+
.
Dcax
Dcrh
a
b
g
T22C
A28C
Dy3+
Tb3+
Tm3+
Yb3+
Dy3+
Tb3+
Tm3+
Yb3+
31.6
19.3
14.0
ꢀ8.3
3.1
50.72
144.8
149.3
157.8
76.5
41.2
38.5
24.0
30.1
ꢀ13.6
5.4
38.9
23.9
51.2
56.7
117.7
46.2
38.1
37.5
46.2
167.6
168.4
62.7
5.2
17.0
15.2
ꢀ5.5
ꢀ1.6
174.0
168.8
12.1
ꢀ12.4
54.3
47.8
ꢀ5.6
To verify the calculated Dc tensors, Figure 6 depicts a plot
of experimental and calculated PCS. Excellent correlations
between the experimental and calculated data and very
small Q factors suggest that the calculated paramagnetic
tensors are of high quality.
RDC measurements and calculation of alignment tensors:
Significant PCSs observed for the lanthanide-loaded ubiqui-
tin–4MTDA protein samples suggest measurable RDCs,
since the anisotropic magnetic susceptibility tensor induces
partial alignment of lanthanide-bound proteins. The RDCs
for the lanthanide-loaded ubiquitin–4MTDA adducts were
Figure 5. A) Superposition of 15N HSQC spectra of 0.10 mm uniformly
15N labeled T22C–4MTDA in the absence (gray) and presence of
0.10 mm Tb3+ (black). B) Superposition of 15N HSQC spectra of 0.1 mm
uniformly 15N labeled A28C–4MTDA in the absence (gray) and presence
of 0.10 mm Tb3+ (black).
1
measured by comparison of the one-bond H–15N splitting
observed with paramagnetic lanthanides and diamagnetic
Y3+. At 298 K, HN–N RDCs of up to 13 Hz were deter-
mined for the Dy3+-loaded protein sample at a magnetic
field of 14.1 T. The axial and rhombic components of the
alignment tensors were determined by fitting the H–N
RDCs of residues in the well-structured regions to the crys-
tal structure of ubiquitin (Table 2). Figure 7 shows the corre-
lations between the experimental and back-calculated
RDCs. Larger Q factors than those of PCSs were obtained.
The lanthanide-bound and free protein samples show slow
exchange, as shown by 15N HSQC spectra. 15N-HSQC spec-
tra were recorded for the complexes of ubiquitin T22C–
4MTDA and A28C–4MTDA with Dy3+, Tb3+, Tm3+, and
Yb3+, by using the complex with Y3+ as the diamagnetic ref-
erence. The assignment of the paramagnetic signals was as-
sisted by the fact that the proton and nitrogen spins of
a backbone amido group have similar PCSs, because their
Fluorescence measurements on the ubiquitin–4MTDA
adduct in complex with Tb3+: The emission and excitation
17144
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17141 – 17149

Bioconjugation of Proteins
FULL PAPER
Figure 6. Correlations of the back-calculated PCS (PCScal) plotted against
the experimentally measured PCS (PCSexp) for the complex of 0.10 mm
ubiquitin–4MTDA and paramagnetic lanthanide ions, as well as Q fac-
tors. A) T22C–4MTDA. B) A28C–4MTDA.
Figure 7. Correlations of the back-calculated RDC (RDCcal
) plotted
against the experimentally measured RDC (RDCexp) for the complex of
0.10 mm ubiquitin–4MTDA and paramagnetic lanthanide ions, as well as
Q factor. A) T22C–4MTDA. B) A28C–4MTDA.
Table 2. Comparison of alignment tensors obtained by RDC fitting and
tin–4MTDA shows emission mainly at 545 nm but with dif-
ferent intensity.
back-calculated from Dc tensors of ubiquitin–4MTDA complexed with
[a]
Dy3+, Tb3+, Tm3+, and Yb3+
.
[b]
[b]
104 Arh
[c]
[c]
Ln3+
104 Aax
104 Aax
104 Arh
T22C
A28C
Dy3+
Tb3+
Tm3+
Yb3+
Dy3+
Tb3+
Tm3+
Yb3+
5.7
5.5
1.9
1.5
8.1
7.7
3.5
1.4
9.9
6.1
3.2
1.4
4.9
3.6
2.1
0.8
1.3
3.9
1.4
0.4
Discussion
ꢀ2.3
ꢀ0.9
Interaction of 4MTDA with paramagnetic lanthanide ions:
2,2’:6’,2’’-Terpyridine-6,6’’-dicarboxylic acid (TDA) and its
derivatives are excellent lanthanide chelators which have
been widely used in photophysical chemistry.[11a,b,f] However,
the thermodynamic stability of the complexes formed by
TDA and lanthanide ions has rarely been explored in aque-
ous solution. The binding constants determined by ITC for
1.3
0.2
6.5
5.1
3.1
1.1
ꢀ3.1
ꢀ1.0
ꢀ1.0
ꢀ0.3
[a] All data recorded at 258C and 600 MHz 1H NMR frequency. The
tensor parameters were obtained by fitting the experimental data to the
crystal structure of ubiquitin. [b] Alignment tensor determined with
Module program. [c] Aax and Arh back-calculated from the Dc tensors in
Table 1 by using Equation (2).
the complexes [Lu
(TDA)]+ and [Lu
(TDA)2]ꢀ are (1.4ꢁ
0.2)ꢂ107 mꢀ1 and (3.5ꢁ0.4)ꢂ1013 mꢀ2, respectively. They are
consistent with a recent report that showed that TDA forms
stable complexes with lanthanide ions in methanol.[11f]
Lanthanide ions bind the first DPA molecule with a high
association affinity.[16] To further compare the binding affini-
spectra of the 4MTDA-tagged protein in complex with Tb3+
are presented in Figure 8. The Tb3+ complexes with ubiqui-
Chem. Eur. J. 2013, 19, 17141 – 17149
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17145

X.-C. Su et al.
Ligation of proteins with 4MTDA: The lower yield of
4MTDA-tagged protein is probably due to the tautomeric
equilibrium between the sulfhydryl and thioketo forms of
4MTDA.[5c] In addition, the solvent susceptibility of cysteine
also affects the ligation yield since the less solvent exposed
A28C gives a lower yield of 4MTDA derivative than T22C.
PCS measurements and Dc-tensor determination: The para-
magnetic ions in complexes with 4MTDA-tagged protein
produce very large PCSs, as shown in Figure 5, which indi-
cate that the lanthanide ions are strictly immobilized in the
protein complexes. As a consequence, the calculated Dc ten-
sors of T22C–4MTDA and A28C–4MTDA adducts were
similar. The determined paramagnetic tensors are signifi-
cantly larger than those observed for DPA-tagged proteins.[5]
Figure 9 shows the calculated position of the paramagnetic
Figure 8. Emission (right) and excitation (left) spectra of 0.05 mm
4MTDA-tagged ubiquitin in complex with 0.06 mm Tb3+ in 20 mm MES
at pH 6.4 and 298 K; the wavelengths of the excitation and emission
spectra are 350 and 545 nm, respectively. A) T22C–4MTDA; B) A28C–
4MTDA.
ty of 4MTDA and DPA for lanthanide ions, Figure S6 (Sup-
porting Information) depicts a 1H NMR titration competi-
tion experiment in which DPA was titrated into a mixture of
Figure 9. Structural representation of ubiquitin–4MTDA. The calculated
paramagnetic centers are shown as black and gray spheres for T22C–
4MTDA and A28C–4MTDA, respectively. The calculated distance be-
tween the metal ion and the oxygen atom of the E24 side chain is shown.
Yb3+ and 4MTDA. At [DPA]/
dance of [Yb
(DPA)2]ꢀ and [Yb
(4MTDA)]+ remains about 72% of that in the absence of
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
C
center with respect to the structure of the protein, and dis-
tances of 3.86 and 4.16 ꢁ in Fig. 9, between the metal posi-
tion and the oxygen atom of E24 side chain were calculated
in the T22C–4MTDA and A28C–4MTDA adducts, respec-
tively. Evidently, the carboxylate group of the E24 side
chain is coordinated to the lanthanide ion in both protein
adducts, in consistence with the calculated large paramag-
netic tensors. In the A28C–4MTDA adduct, residue A28 sits
in an a helix and residue E24 is located at the iꢀ4 position
of the same helix. This in agreement with a recent report
that the i and iꢀ4/i+4 positions in an a helix are suitable to
anchor a paramagnetic tag by coordination of the protein.[17]
In the T22C–4MTDA adduct, the paramagnetic tag is also
rigid, despite T22C being located in a loop just before an
a helix. This suggests that the distance and orientation be-
tween the ligation site and the acidic side-chain residues are
favorable for cooperative restriction of lanthanide ions. In
both 4MTDA-derivatized T22C and A28C adducts, the dis-
tance between the b-carbon atom of each two residues and
DPA. The competition experiment suggests that 4MTDA
has higher binding affinity for lanthanides than DPA.
In comparison with DPA, 4MTDA forms more dynami-
cally stable complexes with lanthanide ions, as determined
by 2D H–H exchange spectroscopy (EXSY). At 298 K and
pH 7.0, no chemical exchange was observed between
[YACHTUNGTRENNUNG ACHTUNGTRENNUGN
(4MTDA)]+ and [Yb(4MTDA)]+ for mixing times of up
to 150 ms (data not shown) in a mixture of 1.0 mm 4MTDA,
0.6 mm YCl3, and 0.8 mm YbCl3. In contrast, significant
chemical exchange was determined for
a mixture of
[Y
N
ACHTUNGTRENNUNG
5 ms was used. This interesting observation is probably due
to slow conformational exchange of nonplanar 4MTDA on
coordination to lanthanide ions. The dynamic stability of
4MTDA and the lanthanide complexes offers advantages in
site-specific labeling of proteins, because chemical exchange
between ligand and metal ion averages the paramagnetic
effects.
17146
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17141 – 17149

Bioconjugation of Proteins
FULL PAPER
the oxygen atom of the E24 side chain is in the range of
6–8 ꢁ, which can be considered a favorable distance for the
design of a suitable cysteine mutation based on protein
structure.
rigid protein–tag constructs.[3abc,5] This observation may also
suggest different sensitivity of PCSs and RDCs to protein
mobility.[18] This observation can also be justified by compar-
ison of Q factors, as shown in Figures 5 and 6, in which PCS
fitting gives much lower Q values.
In the DPA-tagged protein adducts, lanthanide ions are
usually fourfold coordinated by three sites of the DPA tag
and one of a protein, and the remaining four to five coordi-
nation sites are generally occupied by water molecules. In
ubiquitin–4MTDA, the lanthanide ion is chelated by
4MTDA and one carboxylate group of the E24 side chain,
and two or three water molecules are engaged in the first
coordination sphere. This structural feature of 4MTDA-
tagged protein and lanthanide ion results in a thermodynami-
cally stable but kinetically inert complex, and thus repre-
sents a great advantage in retaining the paramagnetic effects
on proteins.
Fluorescence of ubiquitin–4MTDA in complex with lantha-
nide ions: The long-lived luminescence of Tb3+ and Eu3+
complexes has advantages in distance measurements on bio-
molecules by FRET[19] and also biomedical imaging and
analysis.[20] The complex of protein–4MTDA and lanthanide
ions is a rigid construct according PCS and RDCs analysis.
In comparison with published luminescent lanthanide com-
plexes,[9a,b,11c,d,21] 4MTDA is smaller and more rigid. The lu-
minescent Tb3+ complex of A28C–4MTDA shows higher in-
tensity than that of T22C–4MTDA, even though both com-
plexes show the main emission at 545 nm. This is probably
due to the mobility of the ligation site in the protein, which
may influence the fluorescence of the Tb3+ complex.
RDC measurements and assessment of 4MTDA rigidity:
Anisotropic paramagnetism partially aligns the molecules in
an external magnetic field. Sizable RDCs were determined
for the lanthanide complexes of 4MTDA-tagged ubiquitin at
a proton frequency of 600 MHz (Table 2). In the sample of
protein loaded with paramagnetic ions, the correlation be-
tween the alignment tensor determined from RDCs and the
Dc tensor determined from PCSs is valuable to assess the
mobility of the tag. In general, the PCS is written as Equa-
tion (1)[1a]
Conclusion
We have shown that 4MTDA can be ligated to a cysteine
residue at position i of an a helix and can be easily immobi-
lized by an aspartate or glutamate residue located in the iꢀ4
or i+4 position, as previously demonstrated.[17] For a protein
with known structure, the ligation site can be selected on
the basis of a distance of about 6–8 ꢁ between the b-carbon
atom of the cysteine residue to be mutated and the nearest
oxygen atom of an acidic residue (glutamate or aspartate)
side chain. 4MTDA is a small and rigid lanthanide binding
tag for site-specific labeling of proteins for structure deter-
mination and studies on protein interactions by NMR and
fluorescence spectroscopy[1,8,19,20,22] The rigid 4MTDA tag
can also be used in high-precision distance measurements
on proteins by EPR spectroscopy.[23]
1
ð1Þ
PCS ¼
½Dcaxð3 cos2 q ꢀ 1Þ þ 1:5Dcrh sin2 q cos 2ꢀꢂ
12pr3
where r, q, and f are the polar coordinates of the nuclear
spin relative to the principal axes of the Dc tensor and Dcax
and Dcrh the axial and rhombic components of the Dc
tensor. In a paramagnetically loaded sample, interconversion
between the Dc tensor (Dcax,rh) and alignment tensor (Aax,rh
)
can be written as Equation (2)[1a,5a]
B02
15kTm0
ð2Þ
Aax;rh
¼
Dcax;rh
Experimental Section
where B0 is the magnetic field strength, m0 the induction
constant, k the Boltzmann constant, and T the temperature.
The PCS can be precisely determined by means of the
chemical-shift differences between the paramagnetic and
Synthesis of 4MTDA: Starting from commercial material DPA (1), com-
pounds 2–5 were obtained by following the published protocols.[13]
Synthesis of 6: Acetone (3.0 mL) was added to a mixture of compound 5
(115 mg, 0.25 mmol) and thiourea (27 mg, 0.36 mmol). The mixture was
heated to reflux for 16 h. On cooling, the precipitated white solid was
collected by filtration and washed with acetone. The crude produce was
dried under reduced pressure (98.0 mg, 86.9%). 1H NMR (400 MHz,
[D6]DMSO): d=9.39 (s, 3H), 8.93 (d, J=7.6 Hz, 2H), 8.59 (d, J=
14.7 Hz, 2H), 8.38–8.11 (m, 4H), 4.45 (q, J=7.0 Hz, 4H), 1.41 ppm (t,
J=7.0 Hz, 6H); 13C NMR (101 MHz, [D6]DMSO): d=167.24, 164.29,
155.45, 153.90, 147.58, 139.33, 138.03, 125.98, 125.90, 124.45, 61.50,
14.17 ppm.
diaACHTUNGTRENNUNGmagnetic protein samples. However, in a given complex
formed by protein and paramagnetic lanthanide ion, the Dc
tensor calculated by PCS increases with increasing distance
between the nucleus and the paramagnetic metal ion. The
accurate position of the paramagnetic metal ion is therefore
important in evaluating the quality of simulated paramag-
netic tensors. In Table 2, the alignment tensors determined
from RDCs are close to those back-calculated from Dc ten-
sors, and this suggests that 4MTDA is rigid in the ubiquitin–
4MTDA construct complexed with lanthanide ions. General-
ly smaller RDC-determined alignment tensors compared to
the PCS-back-calculated ones are observed in a number of
Synthesis of 4MTDA (7): Sodium hydroxide solution (1.0 mL, 2.4 mmol)
was added dropwise to the suspension of compound
6 (110.0 mg,
0.24 mmol) in 5 mL of methanol under argon protection. The resulting
solution was stirred for 3 h and then the pH was adjusted to 2 with 6m
HCl. The orange precipitate was filtered off, washed with cold water, and
dried under reduced precessure (80.0 mg, 94.4%). 1H NMR (400 MHz,
Chem. Eur. J. 2013, 19, 17141 – 17149
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17147

X.-C. Su et al.
D2O, pD 10): d=8.19 (d, J=7.9 Hz, 1H), 8.11 (s, 1H), 8.01 (t, J=7.8 Hz,
1H), 7.92 ppm (d, J=7.7 Hz, 1H).
[2] a) F. Rodriguez-CastaÇeda, P. Haberz, A. Leonov, C. Griesinger,
Magn. Reson. Chem. 2006, 44, S10–S16; b) X. C. Su, G. Otting, J.
Biomol. NMR 2010, 46, 101–112; c) J. Koehler, J. Meiler, Prog.
Nucl. Magn. Reson. Spectrosc. 2011, 59, 360–389; d) D. Shishmarev,
G. Otting, J. Biomol. NMR 2013, 56, 203–216.
[3] a) X. C. Su, K. McAndrew, T. Huber, G. Otting, J. Am. Chem. Soc.
2008, 130, 1681–1687; b) P. H. Keizers, A. Saragliadis, Y. Hiruma,
M. Overhand, M. Ubbink, J. Am. Chem. Soc. 2008, 130, 14802–
1481; c) D. Hꢄussinger, J. R. Huang, S. Grzesiek, J. Am. Chem. Soc.
2009, 131, 14761–14767; d) B. Graham, C. T. Loh, J. D. Swarbrick, P.
Ung, J. Shin, H. Yagi, X. Jia, S. Chhabra, N. Barlow, G. Pintacuda,
T. Huber, G. Otting, Bioconjugate Chem. 2011, 22, 2118–2125; e) F.
Expression and purification of protein: Human ubiquitin T22C and
A28C mutants were prepared in M9 medium containing 15NHCl4 (Isotec)
as the sole nitrogen source. The mutant ubiquitin was purified from the
soluble fraction of the lysed cells by ammonium sulfate precipitation, fol-
lowed by chromatography on DEAE columns (GE Healthcare Biosci-
ences) and G50 (GE Healthcare Biosciences). Typically, 10 mg of protein
was obtained from 250 mL of medium.
Site-specific labeling of proteins with 4MTDA: The ligation experiment
was performed with uniformly 15N labeled human ubiquitin A28C and
T22C mutants. Similar to the previous ligation protocol,[12] the target pro-
tein at 0.5 mm was first activated with a fivefold excess of Ellmanꢃs re-
agent [5,5’-dithiobis(2-nitrobenzoic acid) (DTNB)] at pH 7.2. Excess
DTNB and produced 2-nibrobenzate (TNB) were removed with a pD10
column (GE), and then eight equivalents of 4MTDA were added to the
activated protein–TNB solution and the pH was adjusted to 7.4. The re-
sultant solution was incubated at room temperature for 6 h. The
4MDTA-tagged protein sample was purified with a PD10 column, and
the ligation yield was about 50–70%.
ˇ ˇ
Peters, M. Maestre-Martinez, A. Leonov, L. Kovacic, S. Beckers, R.
Boelens, C. Griesinger, J. Biomol. NMR 2011, 51, 329–337; f) C. T.
Loh, K. Ozawa, K. L. Tuck, N. Barlow, T. Huber, G. Otting, B.
Graham, Bioconjugate Chem. 2013, 24, 260–268; g) L. J. Martin,
M. J. Hꢄhnke, M. Nitz, J. Wçhnert, N. R. Silvaggi, K. N. Allen, H.
Schwalbe, B. Imperiali, J. Am. Chem. Soc. 2007, 129, 7106–7113;
h) K. Barthelmes, A.M. Reynolds, E. Peisach, H. R. Jonker, N. J.
DeNunzio, K. N. Allen, B. Imperiali, H. Schwalbe, J. Am. Chem.
Soc. 2011, 133, 808–819.
NMR spectra: NMR experiments were performed at 298 K on a Bruker
AV600 NMR spectrometer equipped with a QCI cryoprobe. All 15N
HSQC spectra in the presence of diamagnetic or paramagnetic metal
ions were recorded on 0.10 mm protein in 20 mm MES at pH 6.4 unless
described otherwise. Paramagnetic or diamagnetic protein samples was
prepared by titrating lanthanide ions (10 mm in stock solution) into the
solution of 0.10 mm protein. PCSs were calculated from 15N HSQC spec-
tra as differences in 1H chemical shifts between samples with paramag-
[4] a) T. Saio, K. Ogura, M. Yokochi, Y. Kobashigawa, F. Inagaki, J.
Biomol. NMR 2009, 44, 157–166; b) J. D. Swarbrick, P. Ung, X. C.
Su, A. Maleckis, S. Chhabra, T. Huber, G. Otting, B. Graham,
Chem. Commun. 2011, 47, 7368–7370.
[5] a) X. C. Su, B. Man, S. Beeren, H. Liang, S. Simonsen, C. Schmitz,
T. Huber, B. A. Messerle, G. Otting, J. Am. Chem. Soc. 2008, 130,
10486–10487; b) B. Man, X. C. Su, H. Liang, S. Simonsen, T. Huber,
B. A. Messerle, G. Otting, Chem. Eur. J. 2010, 16, 3827–3832; c) X.
Jia, A. Maleckis, T. Huber, G. Otting, Chem. Eur. J. 2011, 17, 6830–
6836.
[6] N. L. Fawzi, M. R. Fleissner, N. J. Anthis, T. Kꢅlai, K. Hideg, W. L.
Hubbell, G.M. Clore, J. Biomol. NMR 2011, 51, 105–114.
[7] a) D. Parker, R. S. Dickins, H. Puschmann, C. Crossland, J. A. K.
Howard, Chem. Rev. 2002, 102, 1977–2010; b) E. Pidcock, G. R.
Moore, J. Biol. Inorg. Chem. 2001, 6, 479–489.
netic lanthanides and diamagnetic Y3+
.
Calculation of the Dc tensors: The Dc-tensor parameters were deter-
mined using the Numbat program.[24] Only PCS data from residues locat-
ed in regions of well-defined secondary structure were included in the
fits to the crystal structure of ubiquitin (pdb code: 1ubi).[25] Q factors
were calculated by using Equation (3).[26]
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
2
ðPCSexp ꢀ PCScal
Þ
[8] a) J.-C. G. Bꢆnzl, Chem. Rev. 2010, 110, 2729–2755; b) P. R. Selvin,
Methods Enzymol. 1995, 246, 300–334.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
ð3Þ
Q ¼
2
ðPCSexp
Þ
[9] a) D. J. Posson, Neuron 2008, 59, 98–109; b) R. B. Best, K. A. Mer-
chant, I. V. Gopich, B. Schuler, B. A. Bax, Proc. Natl. Acad. Sci.
USA 2007, 104, 18964–18967; c) J. W. Taraska, M. C. Puljung, N. B.
Olivier, G. E, W. N. Zagotta, Nat. Methods 2009, 6, 532–537.
[10] A. Wild, A. Winter, F. Schlꢆltter, U. S. Schubert, Chem. Soc. Rev.
2011, 40, 1459–1511.
[11] a) A. K. Saha, K. Cross, E. D. Kloszewski, D. A. Upson, J. L. Toner,
R. A. Snow, C. D. V. Black, V. C. Desai, J. Am. Chem. Soc. 1993,
115, 11032–11033; b) A. P. de Silva, H. Q. N. Gunaratne, T. E. Rice,
Angew. Chem. 1996, 108, 2253–2255; Angew. Chem. Int. Ed. Engl.
1996, 35, 2116–2118; c) S. Poupart, C. Boudou, P. Peixoto, M. Mas-
sonneau, P.-Y. Renard, A. Romieu, Org. Biomol. Chem. 2006, 4,
4165–4177; d) M. Latva, H. Takalo, V.-M. Mukkala, C. Matachescu,
J.-C. Rodriguez-Ubis, J. Kankare, J. Lumin. 1997, 75, 149–169;
e) X. Y. Chen, Y. Bretonniere, J. Pecaut, D. Imbert, J. C. Bunzli, M.
Mazzanti, Inorg. Chem. 2007, 46, 625–637; f) E. S. Andreiadis, R.
Demadrille, D. Imbert, J. Pecaut, M. Mazzanti, Chem. Eur. J. 2009,
15, 9458–9476.
No data from the flexible loops and the flexible C-terminal residues were
used in simulations of Dc tensors.
RDC measurements and calculation of alignment tensors: Residual dipo-
1
lar couplings DNH were measurement as the 15N-doublet splitting of the
diamagnetic sample minus that of the paramagnetic sample by using the
IPAP pulse sequence.[27] Alignment tensors were determined by using
Module.[28] Only RDCs of residues in regular secondary-structure ele-
ments were used in fitting to the crystal structure of ubiquitin (pdb code:
1ubi),[25] and Q factors were calculated by using Equation (4).[29]
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
2
ðRDCexp ꢀ RDCcalc
Þ
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P
ð4Þ
Q ¼
2
ðRDCexp
Þ
Acknowledgements
[12] X. C. Su, T. Huber, N. E. Dixon, G. Otting, ChemBioChem 2006, 7,
1599.
[13] P. Kadjane, C. Plates-Iglesias, R. Ziessel, L. J. Charbonniere, J.
Chem. Soc. Dalton Trans. 2009, 5688–5700.
[14] A. A. Avetisyan, I. L. Aleksanyan, L. P. Ambartsumyan, Russ. J.
Org. Chem. 2005, 41, 769–771.
Financial support by 973 program (2013CB910200) and National Science
Foundation of China (21073101, 21273121, and 21121002) is greatly ac-
knowledged. We thank Dr. Thomas Huber at Research School of
Chemistry in Australian National University for discussion.
[15] Z. Wei, Y. Yang, Q. F. Li, F. Huang, H. H. Zuo, X. C. Su, Chem.
Eur. J. 2013, 19, 5758–5764.
[16] I. Grenthe, J. Am. Chem. Soc. 1960, 82, 6258.
[17] H. Yagi, A. Maleckis, G. Otting, J. Biomol. NMR 2013, 55, 157–166.
[18] M. Assfalg, I. Bertini, P. Turano, A. G. Mauk, J. R. Winkler, H. B.
Gray, Biophys. J. 2003, 84, 3917–3923.
[1] a) I. Bertini, C. Luchinat, G. Parigi, Prog. Nucl. Magn. Reson. Spec-
trosc. 2002, 40, 249–273; b) G. Otting, J. Biomol. NMR 2008, 42, 1–
9; c) G.M. Clore, J. Iwahara, Chem. Rev. 2009, 109, 4108–4139; d) I.
Bertini, C. Luchinat, G. Parigi, Coord. Chem. Rev. 2011, 255, 649–
663; e) K. N. Allen, B. Imperiali, Curr. Opin. Chem. Biol. 2010, 14,
247–254.
[19] P. R. Selvin, Ann. Rev. Biophys. Biolmol. Struct. 2002, 31, 275–302.
17148
www.chemeurj.org
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17141 – 17149

Bioconjugation of Proteins
FULL PAPER
[20] a) J. C. G. Bꢆnzli, Chem. Rev. 2010, 110, 2729–2755; b) S. Pandya, J.
Yu, D. Parker, Dalton Trans. 2006, 2757–2766; c) I. Hemmilꢄ, V.
Laitala, J. Fluoresc. 2005, 15, 529–542.
[21] a) M. Kawaguchi, T. Okabe, T. Terai, K. Hanaoka, H. Kojima, I.
Minegishi, T. Nagano, Chem. Eur. J. 2010, 16, 13479; b) K. Hashino,
K. Ikawa, M. Ito, C. Hosoya, T. Nishioka, M. Makiuchi, K. Matsu-
moto, Anal. Biochem. 2007, 364, 89–91; c) V. Laitala, I. Hemmilꢄ,
Anal. Chem. 2005, 77, 1483–1487.
Y. Song, T. J. Meade, D. Goldfarb, A. V. Astashkin, A. Raitsimring,
J. Magn. Reson. 2010, 205, 38–49.
[24] C. Schmitz, M. J. Stanton-Cook, X. C. Su, G. Otting, T. Huber, J.
Biomol. NMR 2008, 41, 179–189.
[25] R. Ramage, J. Green, T. W. Muir, O.M. Ogunjobi, S. Love, K. Shaw,
Biochem. J. 1994, 299, 151–158.
[26] I. Bertini, M. B. L. Janik, Y.M. Lee, C. Luchinat, A. Rosato, J. Am.
Chem. Soc. 2001, 123, 4181–4188.
[22] a) I. Bertini, C. Luchinat, M. Nagulapalli, G. Parigi, E. Ravera,
Phys. Chem. Chem. Phys. 2012, 14, 9149–9156; b) T. Saio, K. Ogura,
K. Shimizu, M. Yokochi, T. R., Jr. Burke, F. Inagaki, J. Biomol.
NMR 2011, 51, 395–408; c) C. Schmitz, R. Vernon, G. Otting, D.
Baker, T. Huber, J. Mol. Biol. 2012, 416, 668–677; d) H. Yagi, K. B.
Pilla, A. Maleckis, B. Graham, T. Huber, G. Otting, Structure 2013,
21, 883–890.
[27] M. Ottiger, F. Delaglio, A. Bax, J. Magn. Reson. 1998, 131, 373–378.
[28] P. Dosset, J.-C. Hus, D. Marion, M. Blackledge, J. Biomol. NMR
2001, 20, 223–231.
[29] A. Bax, G. Kontaxis, N. Tjandra, Methods Enzymol. 2001, 339, 127–
174.
[23] a) H. Yagi, D. Banerjee, B. Graham, T. Huber, D. Goldfarb, G.
Otting, J. Am. Chem. Soc. 2011, 133, 10418–10421; b) A. Potapov,
Received: June 15, 2013
Published online: November 7, 2013
Chem. Eur. J. 2013, 19, 17141 – 17149
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
17149