Reporter protein-targeted probes for magnetic resonance imaging

Article abstract of DOI:10.1021/ja206134b

Contrast agents for magnetic resonance imaging are frequently employed as experimental and clinical probes. Drawbacks include low signal sensitivity, fast clearance, and nonspecificity that limit efficacy in experimental imaging. In order to create a bioresponsive MR contrast agent, a series of four Gd(III) complexes targeted to the HaloTag reporter were designed and synthesized. HaloTag is unique among reporter proteins for its specificity, versatility, and the covalent interaction between substrate and protein. In similar systems, these properties produce prolonged in vivo lifetimes and extended imaging opportunities for contrast agents, longer rotational correlation times, and increases in relaxivity (r₁) upon binding to the HaloTag protein. In this work we report a new MR contrast probe, 2CHTGd, which forms a covalent bond with its target protein and results in a dramatic increase in sensitivity. A 6-fold increase in r₁, from 3.8 to 22 mM^-1s^-1, is observed upon 2CHTGd binding to the target protein. This probe was designed for use with the HaloTag protein system which allows for a variety of substrates (specific for MRI, florescence, or protein purification applications) to be used with the same reporter.

Full text of DOI:10.1021/ja206134b

COMMUNICATION
pubs.acs.org/JACS
Reporter Protein-Targeted Probes for Magnetic Resonance Imaging
Renee C. Strauch, Daniel J. Mastarone, Preeti A. Sukerkar, Ying Song, Jonathan J. Ipsaro, and
Thomas J. Meade*
Departments of Molecular Biosciences, Chemistry, Neurobiology & Physiology, and Radiology, Northwestern University,
2145 Sheridan Road, Evanston, Illinois 60208, United States
S Supporting Information
b
receptor-induced magnetization enhancement (RIME) effect in
ABSTRACT: Contrast agents for magnetic resonance
imaging are frequently employed as experimental and clinical
probes. Drawbacks include low signal sensitivity, fast clear-
ance, and nonspecificity that limit efficacy in experimental
imaging. In order to create a bioresponsive MR contrast
agent, a series of four Gd(III) complexes targeted to the
HaloTag reporter were designed and synthesized. HaloTag
is unique among reporter proteins for its specificity, versa-
tility, and the covalent interaction between substrate and
protein. In similar systems, these properties produce pro-
longed in vivo lifetimes and extended imaging opportunities
for contrast agents, longer rotational correlation times, and
increases in relaxivity (r1) upon binding to the HaloTag
protein. In this work we report a new MR contrast probe,
2CHTGd, which forms a covalent bond with its target
protein and results in a dramatic increase in sensitivity.
A 6-fold increase in r1, from 3.8 to 22 mMÀ1 sÀ1, is observed
upon 2CHTGd binding to the target protein. This probe
was designed for use with the HaloTag protein system which
allows for a variety of substrates (specific for MRI, flores-
cence, or protein purification applications) to be used with
the same reporter.
which the motion of a contrast agent is coupled to that of a target
protein, thereby increasing the τR and increasing relaxivity.4 One
example of such an agent is MS-325 (Ablavar), a small molecule
contrast agent targeted to human serum albumin (HSA). Upon
binding to HSA there is an increase in τR from 115 ps to 10 ns and
a concurrent increase in relaxivity from 5.5 to 25 mMÀ1sÀ1
(at 60 MHz, 37 °C).4 As a result of the improved relaxivity and
blood pool lifetime, MS-325 has recently been approved for
clinical MR angiography.5
Here, we present a series of MR contrast agents developed to
utilize a RIME strategy. These complexes target and covalently
bind a commercially available modified haloalkane dehalogenase
reporter (HaloTag) developed by Promega.6 Wildtype haloalk-
ane dehalogenase proteins are a class of bacterial enzymes that
use a catalytic triad active site to catalyze the hydrolysis of the
haloalkane substrate to a primary alcohol.7 The HaloTag active
site is located at the bottom of a 15 Å binding pocket where the
wildtype active site histidine has been mutated to a phenylala-
nine, leaving the enzyme unable to hydrolyze the covalent ester
intermediate.6 One haloalkane substrate covalently binds to
HaloTag under physiological conditions and acts as a suicide
inhibitor of the modified protein.6
A number of HaloTag-targeted substrates have been gener-
ated by conjugating various functionalities onto the distal end of
the haloalkane targeting group, including fluorophores, biotin,
quantum dots, and purification resins.6,8 These ligands have been
utilized in a variety of reporter experiments, including pulse chase,
time course, and fate mapping.9 All of these substrates utilize the
same 14 atom, chlorine-terminated haloalkane reactive moiety
designed to optimally bind the HaloTag protein.
In this report we describe the synthesis of a series of four MR
probes containing a Gd(III) chelate and a haloalkane substrate
connected by linkers of varying lengths (Figure 1). These derivatives
were designed to systematically shorten the distance between the
Gd(III) chelate and the surface of the protein in order to exploit the
RIME effect and to optimize binding to the protein. We hypothe-
sized that restricting the local motion of the Gd(III) chelate near the
flexible linker would result in longer rotational correlation times and,
therefore, increased relaxivity. The complexes are referred to by the
number of carbons in the linker excluding the carbonyl carbon,
the HaloTag targeting group, and the Gd(III) ion (1CHTGd,
2CHTGd, 3CHTGd, and 4CHTGd).
agnetic resonance imaging (MRI) is a diagnostic tool in
M
both research and clinical applications due to its capacity to
render images of high spatial and temporal resolution without the
need for ionizing radiation.1 Contrast agents, such as chelated
Gd(III), can be utilized in clinical MR imaging to improve
sensitivity and resolution as well as allow more rapid scanning
times.2 While valuable information can be attained using generic
contrast agents, their effectiveness is limited due to their non-
specificity, rapid clearance, and low relaxivity.2
In order to create new generations of MR agents that over-
come these shortcomings, a focus of our research has been the
development of bioresponsive contrast agents that bind to a target
protein, report on enzyme activity or cation binding.3 Here, we
describe a system to be used for tracking the expression of a
reporter protein using a MR contrast agent.
Contrast agent relaxivity is determined by the interaction
between a Gd(III) ion and nearby water protons.2 This interac-
tion is described by several parameters including the number of
coordinating water molecules (q), the exchange rate of the inner-
sphere water (kex or 1/τM), and the rotational correlation time
of the complex (τR).2 The fast molecular motion of small mole-
cule complexes is the factor most limiting the observed relaxivity.
A common strategy to increase relaxivity is to employ the
The synthesis of the amine-terminated HaloTag targeting moiety
is described in Supplemental Scheme 1. 1-Amino-2-ethoxyethanol
Received: July 1, 2011
Published: September 26, 2011
r
2011 American Chemical Society
16346
dx.doi.org/10.1021/ja206134b J. Am. Chem. Soc. 2011, 133, 16346–16349
|

Journal of the American Chemical Society
COMMUNICATION
Scheme 2. Synthesis of Macrocyclic Gd(III) Contrast Agents
Connected to Haloalkane Moieties by Linkers of Varying
Length
Figure 1. HaloTag-targeted contrast agents. A macrocyclic Gd(III)
chelator is connected to a haloalkane group via a flexible linker to
generate a HaloTag-targeted contrast agent. Four complexes, with
linkers of 1 to 4 carbons (shown), are described in this study.
Scheme 1. Two Synthetic Routes Coupling Gd(III) Chela-
tors and Variable Length Linkers Towards the Formation of a
HaloTag-Targeted Contrast Agent Series
The final products were prepared using the approach shown
in Scheme 2. The benzyl protecting group on the linker arm
was removed via hydrogenation using a palladium on carbon
catalyst to produce the free carboxylic acid. The haloalkane
moiety was peptide coupled to the free acid of the tert-butyl
DO3A-linker using DIC, DPTS, and TEA. The tert-butyl pro-
tecting groups were removed with triflouroacetic acid, and the
final complexes were produced by metalation with GdCl3 main-
taining the pH below 6. The metalated complexes were purified
by reverse-phase HPLC and characterized by MS. No aggrega-
tion was observed in relaxometric measurements.
was carbobenzyloxy (CBZ) protected using benzyl chlorofor-
mate and TEA to produce S1. 1-Bromo-6-chlorohexane was
coupled in 40% KOH in water using tert-butyl ammonium
hydroxide as a phase transfer catalyst to produce S2. The protect-
ing group was removed via hydrogenation with a palladium on
carbon catalyst to produce the amine-terminated haloalkane
species (S3) for peptide coupling to the Gd(III) chelate-linker
molecule.
The protected 3-carbon (1c) and 4-carbon linker (1d) arms
(benzyl 4-bromobutyrate and benzyl 5-bromovalerate re-
spectively) were synthesized by benzyl protecting 4-bromobu-
tyric acid and 5-bromovaleric acid using benzyl alcohol and
esterification agents DIC and DPTS (Scheme 1). These pro-
tected linkers were conjugated to the tert-butyl protected DO3A
macrocycle via an SN2 reaction using K2CO3 and reflux condi-
tions to produce 2a, 2c, and 2d.
Benzyl protection of 3-bromopropionic acid (1b) was
achieved in high yield using a previously described method
(Scheme 1).10 3-Bromopropionic acid was heated to reflux with
benzyl alcohol in toluene using p-toluene sulfonic acid mono-
hydrate as a catalyst. A new procedure for synthesis of a mono-
propionate derivative of DOTA is described in Scheme 2b.11,12
The propionate linker (1b) was coupled to the tert-butyl DO3A
macrocycle using an anion exchange resin (Amberlite IRA-410
resin) at room temperature to produce 2b.
The HaloTag protein was expressed recombinantly after
subcloning the target protein into the vector pMCSG7 to include
an N-terminal His6-Tag followed by a TEV protease cleavage
site. The protein was expressed in BL21(DE3) cells and purified
by Nickel affinity chromatography. The affinity tag was proteo-
lytically removed with TEV protease, and HaloTag protein was
further purified by ion exchange and gel-filtration chromatogra-
phy (the resulting protein was >99% pure by SDS-PAGE).
The HaloTag protein was labeled with a 5-fold molar excess of
the Gd(III) complex at 4 °C overnight. The excess was removed
via a desalting column followed by dialysis overnight into 10 mM
MOPS pH 7.4 resulting in a 1:1 Gd(III) complex-protein species.
The labeled proteins were concentrated to a starting concentra-
tion of ∼600 μM for all relaxivity analyses.
Longitudinal relaxivities of the unbound agents were deter-
mined in water and the protein-bound contrast agents in 10 mM
MOPS at 1.5 T, 37 °C (Table 1).13 The agents that possess
shorter linkers (1CHTGd and 2CHTGd) had relaxivities similar
to those of reported agents coordinating one water molecule.2
It is possible that an unfavorable geometry prevents the carbonyls
in 3CHTGd and 4CHTGd from coordinating the Gd(III)
16347
dx.doi.org/10.1021/ja206134b |J. Am. Chem. Soc. 2011, 133, 16346–16349

Journal of the American Chemical Society
COMMUNICATION
Table 1. Longitudinal Relaxivities of Free (F) and Protein-
Bound (PB) HaloTag Targeted Contrast Agents at 60 MHz,
37 °C
hydroxyl groups in serine and threonine, amines in lysine, and
carboxylic acids in glutamic and aspartic acids). It is possible for
nearby residues to displace the inner-sphere water of the bound
contrast agent causing coordinative saturation and a decrease
in observed relaxivity.4a,15
r1 À F
r1 À PB
Fold
Sample
(mMÀ1sÀ1
)
(mMÀ1sÀ1
)
Change
2CHTGd will facilitate magnetic resonance molecular ima-
ging of cellular events by binding to the HaloTag reporter protein.
Recently, an extracellular HaloTag-receptor protein was gene-
rated9b that precludes the need to develop a cell permeable
contrast agent and will allow systemic delivery of 2CHTGd. The
reduced molecular motion of a cell-membrane anchored Halo-
Tag is expected to improve bound relaxivity through an increase
in the τR of the complex. Increased signal-to-noise will be achieved
by the large increase in relaxivity of the protein-bound 2CHTGd
complex and rapid clearance of the low-relaxivity, free contrast
agent. The covalent bond formation between the contrast
agent and the protein will result in prolonged in vivo lifetimes
of the probe which in turn will extend imaging opportunities.
A potential limitation of a 1:1 contrast agent/target protein
model is that the low Gd(III) payload may limit our ability to
visualize the target by MRI. However, it has been shown that
monomeric contrast agents with high affinity and specificity for
their target protein have a lower detection limit if the target protein
is clustered in an area of high local concentration (microdomain).16
The Gd(III) concentration at the target site can be further increased
through several approaches, which include increasing the number of
HaloTag receptors through genetic means or utilizing several
Gd(III) complexes per haloalkane targeting group.17 Together,
2CHTGd and concurrent advancements in HaloTag technology
should allow for sensitive MR imaging of cellular processes at a
level that, to date, has been difficult to achieve.
1CHTGd
2CHTGd
3CHTGd
4CHTGd
3.6 ( 0.2
3.8 ( 0.1
5.1 ( 0.1
6.4 ( 0.5
13.4 ( 1.3
22.0 ( 2.2
7.6 ( 0.7
8.6 ( 1.1
3.7
5.8
1.5
1.3
Figure 2. MR images of HaloTag-targeted contrast agents. Gray-scale
(left panel) and intensity-scale (right panel) T1-weighted images of the
free (F) and protein-bound (PB) contrast agents (100 μM Gd(III))
obtained at 1.5 T (TR 150 ms, TE 3.6 ms).
providing an additional coordination site for water, resulting in
their slightly larger observed relaxivities.
The relaxivities of the protein-bound contrast agents do not
follow a trend with linker length and vary significantly, from 7.6
to 22.0 mMÀ1sÀ1 (Table 1). The bound 2CHTGd demonstrated
the largest difference in relaxivity with a 6-fold increase from 3.8 to
22.0 mMÀ1sÀ1. This increase in relaxivity is similar to that of
MS-325, which increases from 5.5 to 25 mMÀ1sÀ1 (1.5 T, 37 °C)
when bound to HSA.4b This result reflects an effective linker length
that couples the rotational correlation time of the Gd(III) chelator
to the HaloTag protein. Complexes 1-, 3-, and 4CHTGd, showed a
lesser increase in relaxivity upon binding to the protein; 3CHTGd
and 4CHTGd resulted in the smallest change in r1, while 1CHTGd
had an intermediate increase in r1 when protein-bound.
’ ASSOCIATED CONTENT
S
Supporting Information. Complete ref 6, synthesis of the
b
amine-terminated haloalkane targeting moiety; transverse relaxivities
of the free and protein-bound probes; high-field transverse and
longitudinal relaxivities of the HaloTag targeted probes; control
solutions imaged at 1.5 T; detailed experimental procedures
and characterization data for the synthesis of the complexes
1CHTGd, 2CHTGd, 3CHTGd and 4CHTGd. This material is
available free of charge via the Internet at http://pubs.acs.org.
This range in signal enhancement was visualized by acquiring
MR images at 1.5 T (Figure 2). The phantoms (100 μM solu-
tions of 1CHTGd, 2CHTGd, 3CHTGd, 4CHTGd, 1CHTGd-
Pro, 2CHTGd-Pro, 3CHTGd-Pro, 4CHTGd-Pro) confirmed
the longitudinal relaxivity results. A difference of a single carbon
had a significant effect on the relaxation properties of the protein-
bound agent.
The relatively low relaxivities of the bound 1-, 3-, and
4CHTGd complexes, compared to 2CHTGd, can be attributed
to one or a combination of unoptimized relaxation parameters,
τR, τM, or q. As discussed previously, relaxivity increases achieved
from binding a protein can be impacted by the rapid local motion
of the Gd(III) chelator. For example MP-2269, a serum albumin-
targeted contrast agent that is similar to MS-325, achieves only
half the protein-bound relaxivity of MS-325 due to its τR being an
order of magnitude shorter.14 The short τR of the protein-bound
MP-2269 is attributed to the internal flexibility which allows free
rotation of the Gd(III) complex.14 In addition, several amino acid
residues have the potential to coordinate Gd(III) (such as the
’ AUTHOR INFORMATION
Corresponding Author
tmeade@northwestern.edu
’ ACKNOWLEDGMENT
We thank Dr. Keith MacRenaris for ICP-MS analysis and
helpful discussions and Prof. Andy Larson and Dr. Wei Guo at
the CAMRI facility at Northwestern University for MR image
acquisition at 1.5 T. This work was supported by the National
Institutes of Health/National Institute of Biomedical Imaging
and Bioengineering Grant R01 EB005866 and National Cancer
Institute Grant U54 CA151800.
’ REFERENCES
(1) Frullano, L.; Meade, T. J. J. Biol. Inorg. Chem. 2007, 12, 939.
(2) Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B. Chem.
Rev. 1999, 99, 2293.
16348
dx.doi.org/10.1021/ja206134b |J. Am. Chem. Soc. 2011, 133, 16346–16349

Journal of the American Chemical Society
COMMUNICATION
(3) (a) Lee, J.; Burdette, J. E.; MacRenaris, K. W.; Mustafi, D.;
Woodruff, T. K.; Meade, T. J. Chem. Biol. 2007, 14, 824. (b) Louie, A. Y.;
Huber, M. M.; Ahrens, E. T.; Rothbacher, U.; Moats, R.; Jacobs, R. E.;
Fraser, S. E.; Meade, T. J. Nat. Biotechnol. 2000, 18, 321. (c) Urbanczyk-
Pearson, L. M.; Femia, F. J.; Smith, J.; Parigi, G.; Duimstra, J. A.;
Eckermann, A. L.; Luchinat, C.; Meade, T. J. Inorg. Chem. 2008, 47, 56.
(d) Li, W. H.; Parigi, G.; Fragai, M.; Luchinat, C.; Meade, T. J. Inorg.
Chem. 2002, 41, 4018. (e) Major, J. L.; Parigi, G.; Luchinat, C.; Meade,
T. J. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 13881. (f) Major, J. L.;
Meade, T. J. Acc. Chem. Res. 2009, 42, 893.
(4) (a) Aime, S.; Gianolio, E.; Terreno, E.; Giovenzana, G. B.;
Pagliarin, R.; Sisti, M.; Palmisano, G.; Botta, M.; Lowe, M. P.; Parker,
D. J. Biol. Inorg. Chem. 2000, 5, 488. (b) Caravan, P.; Cloutier, N. J.;
Greenfield, M. T.; McDermid, S. A.; Dunham, S. U.; Bulte, J. W.;
Amedio, J. C., Jr.; Looby, R. J.; Supkowski, R. M.; Horrocks, W. D., Jr.;
McMurry, T. J.; Lauffer, R. B. J. Am. Chem. Soc. 2002, 124, 3152.
(5) Pazdur, R. 2008; http://www.accessdata.fda.gov./drugsatfda_docs/
appletter/2008/021711s000ltr.pdf.
(6) Los, G. V.; et al. ACS Chem. Biol. 2008, 3, 373.
(7) Schindler, J. F.; Naranjo, P. A.; Honaberger, D. A.; Chang, C. H.;
Brainard, J. R.; Vanderberg, L. A.; Unkefer, C. J. Biochemistry 1999,
38, 5772.
(8) (a) Zhang, Y.; So, M. K.; Loening, A. M.; Yao, H.; Gambhir, S. S.;
Rao, J. Angew. Chem., Int. Ed. 2006, 45, 4936. (b) Lee, H. L.; Lord, S. J.;
Iwanaga, S.; Zhan, K.; Xie, H.; Williams, J. C.; Wang, H.; Bowman, G. R.;
Goley, E. D.; Shapiro, L.; Twieg, R. J.; Rao, J.; Moerner, W. E. J. Am.
Chem. Soc. 2010, 132, 15099.
(9) (a) Yamaguchi, K.; Inoue, S.; Ohara, O.; Nagase, T. Methods Mol.
Biol. 2009, 577, 121. (b) Kosaka, N.; Ogawa, M.; Choyke, P. L.;
Karassina, N.; Corona, C.; McDougall, M.; Lynch, D. T.; Hoyt, C. C.;
Levenson, R. M.; Los, G. V.; Kobayashi, H. Bioconjugate Chem. 2009,
20, 1367. (c) Huybrechts, S. J.; Van Veldhoven, P. P.; Brees, C.;
Mannaerts, G. P.; Los, G. V.; Fransen, M. Traffic 2009, 10, 1722.
(10) McCabe, P. H.; Milne, N. J.; Sim, G. A. J. Chem. Soc., Perkin
Trans. 2 1989, 1459.
(11) Initial tests with K2CO3 or Na2CO3 did not produce the
desired product (2b). A competing side reaction results in the formation
of an α,β-unsaturated product via the elimination of the β-bromide.
(12) (a) Jaszberenyi, Z.; Sour, A.; Toth, E.; Benmelouka, M.;
Merbach, A. E. Dalton Trans. 2005, 2713. (b) Balogh, E.; Tripier, R.;
Fouskova, P.; Reviriego, F.; Handel, H.; Toth, E. Dalton Trans.
2007, 3572. (c) Ferreira, M. F.; Martins, A. F.; Martins, J. A.; Ferreira,
P. M.; Toth, E.; Geraldes, C. F. Chem. Commun. (Camb) 2009, 6475.
(13) The transverse relaxivities at 60 MHz, 37 °C and transverse and
longitudinal relaxivities at 400 MHz, 37 °C for the free and HaloTag-
bound complexes are reported in the Supporting Information. As
expected, the transverse relaxivities increase with increasing magnetic
field strength.
(14) (a) Toth, E.; Connac, F.; Helm, L.; Adzamli, K.; Merbach, A. E.
J. Biol. Inorg. Chem. 1998, 3, 606. (b) Toth, E.; Helm, L.; Merbach, A. E.
Top. Curr. Chem. 2002, 221, 61.
(15) Zech, S. G.; Sun, W. C.; Jacques, V.; Caravan, P.; Astashkin,
A. V.; Raitsimring, A. M. ChemPhysChem 2005, 6, 2570.
(16) (a) Hanaoka, K.; Lubag, A. J. M.; Castillo-Muzquiz, A.;
Kodadek, T.; Sherry, A. D. Magn. Reson. Imaging 2008, 26, 608. (b)
Sukerkar, P. A.; MacRenaris, K. W.; Meade, T. J.; Burdette, J. B. Mol.
Pharmaceutics 2011, 8, 1390.
(17) Mastarone, D. J.; Harrison, V. S.; Eckermann, A. L.; Parigi, G.;
Luchinat, C.; Meade, T. J. J. Am. Chem. Soc. 2011, 133, 5329.
16349
dx.doi.org/10.1021/ja206134b |J. Am. Chem. Soc. 2011, 133, 16346–16349