A split ligand for lanthanide binding: Facile evaluation of dimerizing proteins

Article abstract of DOI:10.1039/c2cc17891e

Luminescence of lanthanides is attractive for biological applications due to its long lifetime and sharp emission profiles. We describe the split display of a lanthanide binding ligand that allows facile evaluation of dimerizing proteins. The split lanthanide ligand is cysteine reactive, and therefore should be readily applicable to a variety of protein systems.

Full text of DOI:10.1039/c2cc17891e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 2997–2999
www.rsc.org/chemcomm
COMMUNICATION
A split ligand for lanthanide binding: facile evaluation of dimerizing
proteinsw
Yue Zhao and Jianmin Gao*
Received 17th December 2011, Accepted 27th January 2012
DOI: 10.1039/c2cc17891e
Luminescence of lanthanides is attractive for biological applications
due to its long lifetime and sharp emission profiles. We describe the
split display of a lanthanide binding ligand that allows facile
evaluation of dimerizing proteins. The split lanthanide ligand is
cysteine reactive, and therefore should be readily applicable to a
variety of protein systems.
Protein dimerization is a ubiquitous phenomenon in biology and
plays a critical role in transcription regulation and various
signaling processes.1 Methods that allow facile detection and
quantification of protein dimers are highly desirable for further
elucidating the significance of protein dimerization in physiology
and disease. Herein, we describe a novel method that reports on
protein dimerization with lanthanide luminescence.
The unique properties of lanthanide luminescence, including
Fig. 1 Structure of the lanthanide ligand HIP4 (a) and schematic
long lifetime and sharp emission profiles, have stimulated
much interest in developing lanthanide-based imaging agents
and biosensors.2 A key feature of lanthanide luminescence lies
in the requirement of light-absorbing ligands as sensitizers
owing to the poor absorptivity of lanthanide ions per se.
Therefore, control of lanthanide–ligand association allows one
to switch on/off lanthanide emission, presenting a powerful and
versatile strategy for sensing biological transformations.3 In this
contribution, we demonstrate that split display of a high-affinity
lanthanide ligand on dimerizing proteins allows facile detection
and quantification of protein dimers with a luminescence readout.
Our design started with the tetrameric construct of 2-hydroxy-
isophthalamide (Fig. 1a, dubbed HIP4 for convenience of
discussion), a lanthanide sensitizer developed by Raymond
and co-workers.4 HIP4 binds terbium with femtomolar affinity
and elicits strong luminescence emission. In sharp contrast, a
substructure of HIP4 (HIP2, a dimer of 2-hydroxyisophthalamide)
shows no binding to lanthanides with concentrations as high
as 10 mM (Fig. S1, ESIw). We hypothesize that proteins
labelled with HIP2, upon dimerization, will reconstitute a
HIP4-like motif for lanthanide binding and consequently
induce luminescence emission (Fig. 1b). In other words, split
display of HIP4 should serve as a generally applicable platform
illustration of the split display of HIP4 for detecting protein dimers
with induced lanthanide luminescence (b).
to report on protein–protein interactions. Importantly, the
2-hydroxyisophthalamide based ligands display maximum
absorption at 350 nm.4 In comparison to other lanthanide
sensitizers (e.g., dipicolinic acid5), the long-wavelength absorption
makes them particularly suitable for biological applications.
Given the broad use of cysteines for protein labeling, we
derivatized the HIP2 structure with a thiol-reactive iodoacetamide
group (Scheme 1). Our synthesis started with 2-methoxyisophthalic
acid 1. The carboxylic acid groups were activated with
2-mercaptothiazoline and then treated with one equivalent of
methylamine to give compound 3. This singly activated HIP
Department of Chemistry and Merkert Center of Chemistry, Boston
College, Chestnut Hill, MA 02467, USA. E-mail: jianmin.gao@bc.edu
w Electronic supplementary information (ESI) available: Ligand
synthesis, peptide synthesis and characterization. See DOI: 10.1039/
c2cc17891e
Scheme 1 Synthesis of the thiol-reactive lanthanide ligand IA-HIP2.
Chem. Commun., 2012, 48, 2997–2999 2997
c
This journal is The Royal Society of Chemistry 2012

View Article Online
derivative was mixed with tris(2-aminoethyl)amine in 2 : 1
stoichiometry. Upon completion, the reaction mixture was treated
with (Boc)2O to protect all the remaining amino groups. Boc
protection was necessary for purification of the product via flash
chromatography. The HIP2 derivative 5 was treated with BBr3 to
simultaneously remove the methyl and Boc protecting groups. The
resulting compound 6 was purified via ion exchange chromato-
graphy, and then treated with iodoacetic anhydride to afford the
desired final product IA-HIP2 (compound 7).
With the small molecule in hand, we investigated its
potential as a dimerization indicator by using a well-characterized
model protein a2D. This thirty-five-residue polypeptide folds into
dimeric helix bundles (Fig. 2a) following a simple two-state
equilibrium: the unfolded monomer and folded dimer.6 Prior
work from our group has shown that the stability of a2D can be
tuned by fluorinating its core residues.7 For example, mutating
the phenylalanine residues in the hydrophobic core to
perfluorophenylalanines (Z) gives the mutant a2D-ZZ that is
more stable than a2D-WT by over 6 kcal molÀ1. Consequently,
at low concentrations (e.g., 2 mM) a2D-ZZ folds completely into
dimers while a2D-WT is largely unfolded (Fig. S2, ESIw). These
two a2D variants were synthesized through solid phase peptide
synthesis, through which we also introduced the H30C mutation
for labelling with IA-HIP2. Due to the C2 symmetry of the a2D
dimer structure, residue 30 of one monomer is positioned in close
proximity of the other. Thus dimerization of the HIP2-labelled
a2D is anticipated to reconstitute a HIP4-like motif for lanthanide
binding (Fig. 2a).
The cysteine mutants of a2D reacted readily with IA-HIP2
under mildly basic conditions. HIP2 labelling minimally
affected the structure and folding of the a2D variants as
revealed by circular dichroism spectroscopy (Fig. S2, ESIw).
The dimerizing behaviour of a2D was examined by monitoring
lanthanide luminescence. With an excess amount of terbium,8
a2D-WT at 2 mM gives slightly higher emission than the HIP2
control (compound 6), consistent with its predominantly
monomeric state under the experimental conditions. In sharp
contrast, a2D-ZZ elicits a strong terbium luminescence
(Fig. 2b), presumably due to the formation of a2D-ZZ dimers.
At equal concentrations, the luminescence emission afforded
by the a2D-ZZ dimer is B50% as strong as that of HIP4,
suggesting that the lanthanide binding motif reconstituted by
the peptide dimers gives good, albeit less than ideal, efficiency
for sensitizing terbium emission. A titration experiment
yields a linear relationship between the terbium luminescence
intensity and the concentration of a2D-ZZ (Fig. 2c), showcasing
the potential of using lanthanide luminescence to quantify the
amount of protein dimers. Importantly, the strong lanthanide
emission enables analysis of a2D dimerization even at sub-
micromolar concentrations (Fig. 2c), which is often challenging
for other techniques including CD and FT-IR. Furthermore,
the dynamic (un)folding behaviour of a2D can be readily
monitored by terbium luminescence. As shown in Fig. 2d,
similar structural transition profiles are displayed by the folding
(cooling down) and unfolding (heating up) curves. The slightly
reduced luminescence of the unfolding curve is presumably due
to photobleaching over time.
Fig. 2 Analyzing a2D dimerization with terbium luminescence. (a)
Cartoon representation of the a2D dimer (PDB: 1QP6); the monomers
are colored green and purple, respectively, and the cysteine residues for
HIP2 labeling are shown in sticks (sulfur atoms are depicted in yellow).
(b) Luminescence spectra of the a2D variants and small molecule
controls. The emission intensities are normalized against that of HIP4
at 545 nm. (c) Concentration profiles of terbium emission showing that
luminescence correlates linearly with peptide concentration. (d) Thermal
folding/unfolding curves monitored by terbium luminescence. The
luminescence intensities were normalized against that at 10 1C.
To test the general applicability of IA-HIP2, we evaluated
another dimerizing model protein GCN4, the coiled coil
c
2998 Chem. Commun., 2012, 48, 2997–2999
This journal is The Royal Society of Chemistry 2012

View Article Online
In summary, we have successfully developed a novel method
that robustly detects protein dimerization with a luminescence
readout. Our method builds on the split display of a high-affinity
lanthanide ligand on target proteins, which upon dimerization
induces lanthanide binding and luminescence emission. An
analogous strategy was recently reported that utilized the split
display of a tetracysteine motif to probe protein folding and
assembly processes.12 In comparison, the current lanthanide-
based method does not involve covalent bond formation as
does the tetracysteine–FlAsH system; the fast lanthanide–
ligand (dis)association allows one to monitor the dynamic
behaviour of target proteins. In addition, the high sensitivity
of lanthanide emission and the possibility of time-gated
measurement make this assay particularly attractive for applications
in complex biological systems. Although the current report focuses
on protein dimerization, the split lanthanide ligand display should
be extendable to the study of protein–protein interactions in general,
as well as protein aggregation processes.
Fig. 3 Analysis of GCN4 dimerization in the presence of bovine serum.
The luminescence data are normalized against the maximum emission
intensity of HIP4 (at 545 nm). Terbium luminescence effectively differentiates
the GCN4 monomer and dimer even in the presence of bovine serum.
Notes and references
1 N. J. Marianayagam, M. Sunde and J. M. Matthews, Trends Biochem.
Sci., 2004, 29, 618; G. Grigoryan, D. T. Moore and W. F. DeGrado,
Annu. Rev. Biochem., 2011, 80, 211; D. A. Guarracino, B. N. Bullock
and P. S. Arora, Biopolymers, 2011, 95, 1.
2 J. C. Bunzli, Chem. Rev., 2010, 110, 2729; K. N. Allen and
B. Imperiali, Curr. Opin. Chem. Biol., 2010, 14, 247; E. G. Moore,
A. P. Samuel and K. N. Raymond, Acc. Chem. Res., 2009, 42, 542;
C. P. Montgomery, B. S. Murray, E. J. New, R. Pal and D. Parker,
Acc. Chem. Res., 2009, 42, 925; H. E. Rajapakse, N. Gahlaut,
S. Mohandessi, D. Yu, J. R. Turner and L. W. Miller, Proc. Natl.
Acad. Sci. U. S. A., 2010, 107, 13582.
3 S. C. Zondlo, F. Gao and N. J. Zondlo, J. Am. Chem. Soc., 2010,
132, 5619.
4 S. Petoud, S. M. Cohen, J. C. Bunzli and K. N. Raymond, J. Am.
Chem. Soc., 2003, 125, 13324.
5 A. A. Hindle and E. A. H. Hall, Analyst, 1999, 124, 1599.
6 R. B. Hill and W. F. DeGrado, J. Am. Chem. Soc., 1998, 120,
1138.
7 H. Zheng and J. Gao, Angew. Chem., Int. Ed., 2010, 49, 8635;
C. J. Pace, H. Zheng, R. Mylvaganam, D. Kim and J. Gao, Angew.
Chem., Int. Ed., 2012, 51, 103.
domain from the bZip repressor.9 There has been long-standing
interest in deciphering the dimerization specificities of this
family of transcription factors.10 Truncated versions of GCN4
(B30 residues) were shown to fold to parallel dimers following
a
two-state mechanism.11 Through solid phase peptide
synthesis, we have prepared the wild type GCN4, as well as
the L19P mutant for comparison (Fig. 3). The incorporation
of proline completely abolishes the dimerization behaviour of
GCN4 as revealed by circular dichroism spectroscopy (Fig. S3,
ESIw). Instead of labelling at the dimer interface, a tripeptide
(Cys-Gly-Gly) fragment is added to the N-terminus of the
GCN4 peptides for HIP2 conjugation. Similar to the a2D system,
the dimerizing peptide GCN4-WT gives a strong luminescence
emission, while the monomeric mutant (GCN4-L19P) yields a
comparable readout as the HIP2 control (Fig. S5, ESIw). Again
the terbium luminescence increases linearly with the GCN4-WT
concentration (Fig. S5, ESIw), suggesting the promise of using
HIP2 to quantify protein dimers. As a preliminary test of this
assay in complex biological systems, we examined the GCN4
peptides in the presence of bovine serum. Although the addition
of 10% bovine serum causes significant quenching for all samples
(Fig. S6, ESIw), the normalized luminescence profiles of the
peptides with serum (Fig. 3) are essentially identical to those
without (Fig. S5, ESIw), suggesting that terbium luminescence
reliably reports on GCN4 dimerization even in the presence of
high concentrations of random proteins.
8 Based on a titration experiment, the terbium concentration is set at
500 mM, as this concentration gives saturating luminescence emission
of the peptide dimers (see Fig. S4, ESIw for details).
9 E. K. O’Shea, R. Rutkowski and P. S. Kim, Science, 1989, 243, 538.
10 L. Gonzalez, Jr., D. N. Woolfson and T. Alber, Nat. Struct. Biol.,
1996, 3, 1011; G. Grigoryan and A. E. Keating, Curr. Opin. Struct.
Biol., 2008, 18, 477.
11 J. W. Blankenship, R. Balambika and P. E. Dawson, Biochemistry,
2002, 41, 15676.
12 N. W. Luedtke, R. J. Dexter, D. B. Fried and A. Schepartz, Nat.
Chem. Biol., 2007, 3, 779; R. A. Scheck and A. Schepartz, Acc.
Chem. Res., 2011, 44, 654; B. Krishnan and L. M. Gierasch, Chem.
Biol., 2008, 15, 1104; C. J. Pace, Q. Huang, F. Wang, K. S. Palla,
A. A. Fuller and J. Gao, ChemBioChem, 2011, 12, 1018.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 2997–2999 2999