Hybrid bombesin analogues: Combining an agonist and an antagonist in defined distances for optimized tumor targeting

Article abstract of DOI:10.1021/ja4087648

Radiolabeled hybrid ligands with defined distances between an agonist and an antagonist for the gastrin-releasing peptide receptor were found to have excellent tumor-targeting properties. Oligoprolines served as rigid scaffolds that allowed for tailoring distances of 10, 20, and 30 A between the recognition elements. In vitro and in vivo studies revealed that the hybrid ligand with a distance of 20 A between the recognition elements exhibits the highest yet observed tumor cell uptake and retention time in prostate cancer cells.

Full text of DOI:10.1021/ja4087648

Communication
pubs.acs.org/JACS
Hybrid Bombesin Analogues: Combining an Agonist and an
Antagonist in Defined Distances for Optimized Tumor Targeting
Carsten Kroll,† Rosalba Mansi,‡ Friederike Braun,‡ Stefanie Dobitz,† Helmut R. Maecke,*,‡
and Helma Wennemers*,†
†Laboratory of Organic Chemistry, ETH Zurich, Zurich, Switzerland
̈
̈
‡Department of Nuclear Medicine, University Hospital Freiburg, Freiburg im Breisgau, Germany
S
* Supporting Information
ABSTRACT: Radiolabeled hybrid ligands with defined
distances between an agonist and an antagonist for the
gastrin-releasing peptide receptor were found to have
excellent tumor-targeting properties. Oligoprolines served
as rigid scaffolds that allowed for tailoring distances of 10,
20, and 30 Å between the recognition elements. In vitro
and in vivo studies revealed that the hybrid ligand with a
distance of 20 Å between the recognition elements exhibits
the highest yet observed tumor cell uptake and retention
time in prostate cancer cells.
eptidic, radiolabeled ligands that bind to receptors overex-
pressed on tumor cells are attractive for molecular imaging
P
and targeted therapy.1 The gastrin-releasing peptide receptor
(GRP-R) is one of the most interesting targets since it is
overexpressed on human prostate adenocarcinoma, the most
frequently diagnosed cancer among men in developed
countries.2,3 GRP-R can be targeted both by agonists and
antagonists that belong to the family of bombesin peptides.4−6
Whereas antagonists bind to GRP-R and remain on the surface of
tumor cells, typical agonists bind to GRP-R and are then
internalized into the cells.4−8 Among the largest challenges for
effective targeting in general is the development of radiolabeled
ligands that are (a) effectively taken up by the tumor, either by
binding to the cell surface or by internalization, (b) not washed
out of the tumor, and (c) have high tumor specificity.1 Toward
these goals the use of multivalent ligands is an attractive approach
since the entropic penalty for binding more than one binding
element is minimized.9 Several multivalent ligands have been
developed for tumor targeting that combine either identical
recognition elements or elements that bind to two different
receptors.10,11 Typical cores for multivalent ligands have a high
degree of conformational flexibility and therefore do not allow
for a precise control of the distance between the recognition
elements.10,11 We envisioned linking an agonist together with an
antagonist for GRP-R on a radiolabeled rigid scaffold in defined
distances as an attractive approach toward improved targeting
ligands. Such hybrid ligands12 may not only combine the potency
of agonists (internalization) with that of antagonists (high cell
surface binding), but also profit from an optimal distance
between the recognition elements and therefore lead to higher
tumor cell uptake and a longer retention within the tumor.
Within this manuscript we present radiolabeled oligoprolines
functionalized with bombesin derived antagonists and agonists in
Figure 1. (a) General concept of oligoproline-based hybrid ligands for
GRP-R targeting. (b) Hybrid ligands 1−3.
distances of 10, 20, and 30 Å as targeting ligands for GRP-R (1−
3, Figure 1). We demonstrate in vitro and in vivo that the hybrid
ligand 2 with a distance of 20 Å between the recognition
elements has superior tumor cell uptake properties compared to
mono- and bivalent analogues as well as hybrids with shorter or
longer spacings between the agonist and antagonist.
Recently we introduced azidoproline (Azp) containing
oligoprolines as conformationally well-defined molecular scaf-
folds that can be easily functionalized in defined spatial
orientations.13 In aqueous environments they adopt the highly
symmetric polyproline II (PPII) helix in which every third
residue is stacked on top of each other in a distance of ∼10 Å.13,14
The functionalization pattern of this molecular scaffold can be
Received: August 23, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja4087648 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
easily fine-tuned by a modular chemical synthesis. In addition,
oligoproline derivatives are very soluble in water under
physiological conditions. Oligoprolines were therefore envi-
sioned to be ideally suited scaffolds for the synthesis of hybrid
ligands bearing an agonist and an antagonist of GRP-R in defined
distances from each other. The agonist AMBA15 and the
antagonist RM15,16 are among the best so far developed
radiolabeled ligands for targeting GRP-R (Scheme 1b). Their
probe the influence of the oligoproline scaffold. CD-spectro-
scopic analyses of such oligoproline-bombesin conjugates
confirmed that they adopt as expected the PPII conformation
in water.21
To evaluate the tumor cell uptake properties of the ligands 1−
8 in vitro, GRP-R overexpressing human prostate adenocarci-
noma cells (PC-3) were used. Each of the hybrid, mono-, and
homobivalent ligands was labeled with the β−- and γ-emitter
177Lu (t1/2 = 6.7 d) to allow for a quantitative analysis of their
tumor cell uptake properties after incubation with PC-3 cells.21
In addition, to evaluate the importance of the covalent linkage
between the agonist and the antagonist, an experiment was
performed in which an equimolar mixture of the monovalent
ligands 4 and 5 was incubated with the PC-3 cells.
Scheme 1. (a) General Synthesis of the Oligoproline-Based
Ligands 1−8 and (b) Schematic Representation of
Monovalent Ligands 4−6, Bivalent Ligands 7 and 8, and RM1
and AMBA
All ligands were taken up into the tumor cells. Blocking of the
GRP-Rs on the PC-3 cells by an excess of the unlabeled agonist
or antagonist confirmed that the cellular uptake of ligands 1−8
was as expected specific and receptor mediated. The surface-
bound fraction and the internalized fraction of the ligands were
then determined by an established protocol.21 These experi-
ments revealed remarkable properties of the oligoproline-based
hybrid ligands (Table 1, Figure 2): (1) All hybrid ligands 1−3
Table 1. PC-3 Cell Uptake of 177Lu Labeled Ligands 1−8,
AMBA, and RM1
a
b
entry
ligand
surface binding [%]
internalization [%]
1
2
3
4
5
6
7
8
9
1
17.2 2.4
11.0 1.3
6.5 2.5
19.5 1.3
2.5 0.4
2.1 0.8
8.0 0.4
2.5 0.2
21.1 0.7
21.8 0.9
4.3 0.3
30.0 5.0
50.7 2.8
31.3 0.6
8.1 0.4
2
3
4
5
26.7 1.2
21.4 9.8
18.4 0.3
37.0 1.0
12.3 0.9
4.7 0.1
6
4 + 5
7
8
c
10
RM1
AMBA
c
11
29.0 2.3
a
b
Surface bound fraction of the ligand after 240 min. Internalized
c
fraction of the ligand after 240 min. Values taken from ref 5.
recognition elements QWAVGHLM and fQWAVGHXL (X =
Statin), respectively, were therefore chosen for the synthesis and
evaluation of hybrid ligands for GRP-R.17 The chelator DOTA18
was chosen since it is an established tight binder for radionuclides
such as 68Ga and 177Lu for in vitro and in vivo studies.19
For the synthesis of the hybrid ligands 1−3, the alkinylated and
side chain protected agonist and antagonist (Scheme 1a) were
prepared by regular solid phase peptide synthesis (SPPS) and
conjugated with the Azp containing trimer Fmoc-Pro-(4S)Azp-
Pro-OH using Cu(I)-catalyzed Huisgen’s 1,3-dipolar cyclo-
addition reactions (“click reaction”).20 The resulting building
blocks were then assembled by SPPS with the unfunctionalized
building block Fmoc-Pro-Pro-Pro-OH to the desired length.
Each of the coupling steps proceeded with satisfactory yields
using HATU/iPr2NEt as coupling reagent and piperidine in
DMF for the Fmoc deprotections. After coupling of β-alanine as
a short spacer and the chelator DOTA, the hybrids 1−3 were
released from the Rink amide resin along with all side chain
protecting groups under acidic conditions.21 The same modular
synthetic strategy was used to prepare the conjugates 4−8 that
bear either a single antagonist (4), agonist (5 and 6), two
agonists (7) or two antagonists (8) on the oligoproline scaffold
(Scheme 1b). These ligands were designed as mono- and
homobivalent reference compounds to the hybrids 1−3 and to
Figure 2. Time-dependent uptake of 177Lu-labeled hybrid ligands 1−3
in PC-3 cells.
have a higher uptake into the tumor cells compared to not only
the monovalent oligoproline-based controls (4−6), but also the
established ligands RM1 and AMBA (Table 1, entries 1−6, 10
and 11).22 (2) The monovalent ligands 4−6 have comparable
uptake properties as AMBA and RM1 demonstrating that the
oligoproline scaffold does not affect the uptake properties to a
significant extent. The tumor cell uptake of the divalent agonist 7
and divalent antagonist 8 is higher compared to that of the
monovalent ligands and in total comparable to those of hybrids 1
B
dx.doi.org/10.1021/ja4087648 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
and 3. (3) Among all ligands, hybrid 2 with a ∼20 Å distance
between the recognition elements stands out. More than 50% of
2 were internalized by the PC-3 cells, and 10% were bound on
their cell surface after 4 h (Table 1, entry 2 and Figure 2). This is
to the best of our knowledge the highest total cellular uptake yet
reported for bombesin-derived radiolabeled ligands.4−6,11a It is
twice as high as those of the monovalent ligands and significantly
higher compared to those of the homodivalent ligands 7 and 8
with the same distance between the recognition elements. It is
also higher than those of the other hybrid ligands 1 and 3 with
shorter or longer distances between the recognition elements.
Thus, not only the combination of an agonist with an antagonist
but also the distance between them is critical for optimal tumor
cell uptake.
A control experiment in which the tumor cell uptake of a
mixture of the radiolabeled monovalent ligands 4 and 5 was
examined (Table 1, entry 7) showed that the uptake of these
noncovalently linked ligands was significantly lower compared to
that of any of the hybrids and only comparable to those of the
monovalent agonists or antagonists. Thus, the covalent linkage of
the agonist and the antagonist in an optimal distance is critical for
the high tumor cell uptake.
These studies also revealed that the hybrid ligand is washed out
of the tumor in the mice very slowly (Figure 4).
Figure 4. In vivo tumor uptake and retention of 177Lu-labeled hybrid 2,
RM1, and AMBA determined by biodistribution studies in PC-3 tumor
implanted mice (data of RM1 and AMBA are from ref 5).
The tumor uptake was 18.6 3.3% injected activity per gram
(IA g−1) 4 h after the injection and was still 15.2 1.7% IA g−1
after 21 h. This wash-out is significantly lower compared to that
of the established antagonist RM1 where half of the ligand was
washed out of the tumor within the same time period.5 It is
comparable to the wash-out of the agonist AMBA with which,
however, the initial uptake after 4 h was only 3.7% IA g−1.5 This
data underlines that the hybrid ligand 2 combines the best
features of an agonist with those of an antagonist in vitro and in
vivo: it has the low wash-out of an agonist and the high tumor
uptake of an antagonist.
To evaluate whether hybrid ligand 2 exhibits also high tumor
uptake in living organisms, we performed in vivo studies with
nude mice in which PC-3 tumor xenografts had been implanted.
For these studies, hybrid ligand 2 was labeled with the positron
emitter 68Ga (t1/2 = 67.7 min) and was then injected into the
mice. PET and PET/CT images that were recorded one hour
after the injection showed a high uptake in the tumor (Figure 3a).
A detailed understanding of the importance of the distance
between the agonistic and antagonistic recognition elements for
the observed high synergistic effect is not trivial, in particular
since the structure of GRP-R has not yet been elucidated and the
understanding of the molecular mechanisms by which ligand-
GPCR interactions trigger signaling is still limited.23−25 In
addition, it is not yet clear whether the agonist and the antagonist
bind to the same site on the GRP-R.16,26 Our experimental
findings are in agreement with a mode of action that relies on the
binding of one of the two recognition elements to the GRP-R and
thereby predisposes the other into an optimal location to interact
with another binding site. This could involve enhanced rebinding
of the ligand with its second recognition element after
dissociation within the same GRP-R.27,9a Alternatively, the
hybrid ligand could bridge the distance between the binding sites
of a GRP-R dimer.25,28 Detailed pharmacological studies that will
shed more light on the uptake mechanism of the agonist−
antagonist hybrids are ongoing and will be reported in due
course.
In conclusion, we showed that a radiolabeled hybrid ligand
with a distance of 20 Å between an antagonist and an agonist for
the GRP-R has excellent prostate tumor uptake properties.
Comparison with other ligands revealed that not only the
covalent linkage between the agonist and the antagonist but also
the distance between them is critical for optimized tumor uptake.
In vitro and in vivo studies showed that the hybrid ligand
combines the high tumor uptake of the antagonist with the high
internalization of the agonist, which results in a long lasting
retention in the tumor. These features render hybrid ligands very
attractive for tumor targeting. Since the oligoproline-based
ligands are readily accessible by a modular synthesis, the concept
of hybrid ligands with controlled distances between the
recognition elements can be easily expanded to the targeting of
other receptors, e.g., somatostatin and gastrin receptors that are
Figure 3. PET/CT (a) and PET (b and c) images of [68Ga]-2 in PC-3
tumor implanted mice 1 h after injection. 4 mm coronal and transversal
slices: (a) unblocked, CT recorded on separate scanner; (b) blocked
with excess of unlabeled agonist 5; (c) blocked with excess of unlabeled
antagonist 4.
The tumor uptake is also in vivo receptor mediated as shown
by experiments with mice in which the GRP-Rs were first blocked
by an excess of either the unlabeled agonist 5 or antagonist 4
before injection of the radiolabeled ligand (Figure 3b and c). The
tumor delineation was very good and the background radio-
activity low. The only other organs that show uptake of the
radiolabeled ligand are as expected from studies with other
ligands: the pancreas that has high physiological GRP-R
expression as well as the kidneys and the bladder through
which the ligand is excreted. The high tumor uptake observed in
the PET images was verified by additional quantitative
biodistribution studies using the 177Lu-labeled hybrid ligand 2.
C
dx.doi.org/10.1021/ja4087648 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Communication
Handl, H. L.; Josan, J. S.; Morse, D. L.; Mash, E. A.; Gillies, R. J.; Hruby,
V. J. Angew. Chem., Int. Ed. 2008, 47, 1685−1688;(e) Angew. Chem.
2008, 120, 1709−1712. (f) Yan, Y.; Chen, X. Amino Acids 2010, 41,
1081−1092. (g) Liu, Z.; Shi, J.; Jia, B.; Yu, Z.; Liu, Y.; Zhao, H.; Li, F.;
Tian, J.; Chen, X.; Liu, S.; Wang, F. Mol. Pharm. 2011, 8, 591−599.
(h) Xua, L.; Josanb, J. S.; Vagnerc, J.; Capland, M. R.; Hruby, V. J.; Mash,
E. A.; Lynche, R. M.; Morsea, D. L.; Gillies, R. J. Proc. Natl. Acad. Sci. U. S.
A. 2012, 109, 21295−21300. For an example of the use of oligoprolines
as a scaffold for CXCR4 ligands, see: (i) Tanaka, T.; Nomura, W.;
Narumi, T.; Masuda, A.; Tamamura, H. J. Am. Chem. Soc. 2010, 132,
15899−15901.
overexpressed on neuroendocrine tumors, certain thyroid and
lung cancers, respectively.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details on the synthesis, analytical data, and in vitro and in vivo
studies. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
helmut.maecke@uniklinik-freiburg.de; helma.wennemers@org.
chem.ethz.ch
■
(11) (a) Abiraj, K.; Jaccard, H.; Kretzschmar, M.; Helm, L.; Maecke, H.
R. Chem. Commun. 2008, 3248−3250. (b) Liu, Z.; Niu, G.; Wang, F.;
Chen, X. Eur. J. Nucl. Med. Mol. Imaging 2009, 36, 1483−1494. (c) Li, Z.-
B.; Wu, Z.; Chen, K.; Ryu, E. K.; Chen, X. J. Nucl. Med. 2008, 49, 453−
461. (d) Carrithers, M. D.; Lerner, M. R. Chem. Biol. 1996, 3, 537−542.
(12) For examples of small molecule-based GPCR hybrid ligands, see:
Valant, C.; Lane, J. R.; Sexton, P. M.; Christopoulos, A. Annu. Rev.
Pharmacol. Toxicol. 2012, 52, 153−178 and references cited therein.
(13) (a) Kuemin, M.; Nagel, Y. A.; Schweizer, S.; Monnard, F. W.;
Ochsenfeld, C.; Wennemers, H. Angew. Chem., Int. Ed. 2010, 49, 6324−
6327. (b) Nagel, Y.; Kuemin, M.; Wennemers, H. Chimia 2011, 65,
264−267. (c) Kuemin, M.; Schweizer, S.; Ochsenfeld, C.; Wennemers,
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the University of Basel where large parts of this work
were performed. We are grateful to E. Gourni, Dr. M. Behe, P.
Wilhelm, and A. Jodal for experimental support and Prof. G.
Schertler, Prof. A. Beck-Sickinger, Prof. R. Schibli, and Dr. X.
Deupi for valuable input to the manuscript. This work was
supported by BACHEM and the Swiss National Science
Foundation.
H. J. Am. Chem. Soc. 2009, 131, 15474−15482. (d) Kumin, M.; Sonntag,
̈
L.-S.; Wennemers, H. J. Am. Chem. Soc. 2007, 129, 466−467.
(14) (a) Kakinoki, S.; Hirano, Y.; Oka, M. Polym. Bull. 2004, 53, 109−
115. (b) Rabanal, F.; Ludevid, M. D.; Pons, M.; Giralt, E. Biopolymers
1993, 33, 1019−1028.
(15) (a) Waser, B.; Eltschinger, V.; Linder, K.; Nunn, A.; Reubi, J. C.
Eur. J. Nucl. Med. Mol. Imaging 2006, 34, 95−100. (b) Bodei, L.; Ferrari,
M.; Nunn, A.; Llull, J.; Cremonesi, M.; Martano, L.; Laurora, G.;
Scardino, E.; Tiberini, S.; Bufi, G.; Eaton, S.; de Cobelli, O.; Paganelli, G.
Eur. J. Nucl. Med. Mol. Imaging 2007, 34, S221.
(16) Tokita, K.; Katsuno, T.; Hocart, S. J.; Coy, D. H.; Llinares, M.;
Martinez, J.; Jensen, R. T. J. Biol. Chem. 2001, 276, 36652−36663.
(17) In the present study, the oxidation sensitive methionin was
replaced by isosteric norleucin. Statin = (3S,4S)-4-amino-3-hydroxy-6-
methyl-heptanoic acid.
(18) 1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetraacetate.
(19) Wadas, T. J.; Wong, E. H.; Weisman, G. R.; Anderson, C. J. Chem.
Rev. 2010, 110, 2858−2902.
(20) (a) Tornøe, C. W.; Christensen, C.; Meldal, M. J. Org. Chem.
2002, 67, 3057−3064. (b) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.;
Sharpless, K. B. Angew. Chem., Int. Ed. 2002, 41, 2596−2599.
(c) Huisgen, R. Angew. Chem. 1963, 75, 604−637.
(21) For experimental details, see the Supporting Information.
(22) Binding affinity studies showed that the hybrid ligands bind 2−4-
fold tighter to GRP-R (1: Kd = 32 nM, 2: Kd = 24 nM) than the
monovalent antagonist (4: Kd = 100 nM) or agonist (5: 75 nM).
(23) (a) Kenakin, T. In GPCR Molecular Pharmacology and Drug
Targeting; Gilchrist, A., Ed.; Wiley: Hoboken, NJ, 2010; pp 1−26.
(b) Christopoulos, A. Nat. Rev. Drug Discovery 2002, 1, 198−210.
(24) Venkatakrishnan, A. J.; Deupi, X.; Lebon, G.; Tate, C. G.;
Schertler, G. F.; Babu, M. M. Nature 2013, 494, 185−194.
REFERENCES
■
(1) For selected recent reviews, see: (a) Lee, S.; Xie, J.; Chen, X. Chem.
Rev. 2010, 110, 3087−3111. (b) Ambrosini, V.; Fani, M.; Fanti, S.;
Forrer, F.; Maecke, H. R. J. Nucl. Med. 2011, 52, 42S−55S.
(2) Markwalder, R.; Reubi, J. C. Cancer Res. 1999, 59, 1152−1159.
(3) American Cancer Society. Global Cancer Facts & Figures, 2nd ed;
American Cancer Society: Atlanta, GA, 2011.
(4) (a) Erspamer, V. Ann. N. Y. Acad. Sci. 1988, 547, 3−9. (b) Sancho,
V.; Di Florio, A.; Moody, T. W.; Jensen, R. T. Curr. Drug Delivery 2011,
8, 79−134. (c) Breeman, W. A. P.; Hofland, L. J.; de Jong, M.; Bernard,
B. F.; Sinivasan, A.; Kwekkeboom, D. J.; Visser, T. J.; Krenning, E. P. Int.
J. Cancer 1999, 81, 658−665. (d) Zhang, H.; Chen, J.; Waldherr, C.;
Hinni, K.; Waser, B.; Reubi, J. C.; Maecke, H. R. Cancer Res. 2004, 64,
6707−6715.
(5) Mansi, R.; Wang, X.; Forrer, F.; Kneifel, S.; Tamma, M.-L.; Waser,
B.; Cescato, R.; Reubi, J. C.; Maecke, H. R. Clin. Cancer Res. 2009, 15,
5240−5249.
(6) Cescato, R.; Maina, T.; Nock, B.; Nikolopoulou, A.; Charalambidis,
D.; Piccand, V.; Reubi, J. C. J. Nucl. Med. 2008, 49, 318−326.
(7) (a) Ginj, M.; Zhang, H.; Waser, B.; Cescato, R.; Wild, D.; Wang, X.;
Erchegyi, J.; Rivier, J.; Macke, H. R.; Reubi, J. C. Proc. Natl. Acad. Sci. U.
̈
S. A. 2006, 103, 16436−16441. (b) Wadas, T. J.; Eiblmaier, M.;
Zheleznyak, A.; Sherman, C. D.; Ferdani, R.; Liang, K.; Achilefu, S.;
Anderson, C. J. J. Nucl. Med. 2008, 49, 1819−1827. (c) Perrin, M. H.;
Sutton, S. W.; Cervini, L. A.; Rivier, J. E.; Vale, W. W. J. Pharmacol. Exp.
Ther. 1999, 288, 729−734.
(8) On a given tumor cell, many more binding sites are available for
antagonists than agonists (see ref 7), presumably since the binding site is
only available for the agonist in a certain conformation of the GRP-R,
whereas the antagonist can bind to its binding site regardless of the
overall conformation of the GRP-R.
(9) (a) Kiessling, L. L.; Gestwicki, J. E.; Strong, L. E. Angew. Chem., Int.
Ed. 2006, 45, 2348−2368;(b) Angew. Chem. 2006, 118, 2408−2429.
(c) Mammen, M.; Choi, S.; Whitesides, G. M. Angew. Chem., Int. Ed.
1998, 37, 2754−2794.
(10) For selected examples, see: (a) Thumshirn, G.; Hersel, U.;
Goodman, S. L.; Kessler, H. Chem.Eur. J. 2003, 9, 2717−2725.
(b) Liu, W.; Hao, G.; Long, M. A.; Anthony, T.; Hsieh, J.; Sun, X. Angew.
Chem., Int. Ed. 2009, 48, 7346−7349. (c) Yim, C.-B.; Boerman, O. C.; de
Visser, M.; de Jong, M.; Dechesne, A. C.; Rijkers, D. T. S.; Liskamp, R.
M. J. Bioconjugate Chem. 2009, 20, 1323−1331. (d) Vagner, J.; Xu, L.;
(25) Urizar, E.; Yano, H.; Kolster, R.; Gales
A. Nat. Chem. Biol. 2011, 7, 624−630.
(26) Nakagawa, T.; Hocart, S. J.; Schumann, M.; Tapia, J. A.; Mantey,
S. A.; Coy, D. H.; Tokita, K.; Katsuno, T.; Jensen, R. T. Biochem.
Pharmacol. 2005, 69, 579−593.
́
, C.; Lambert, N.; Javitch, J.
(27) Vauquelin, G. Expert Opin. Drug Discovery 2010, 5, 927−941.
(28) The higher binding efficiency of antagonists compared to agonists
(refs 7 and 8) are likely the reason why hybrid 2 outperforms the
homodivalent agonist 7 with the same distance between the recognition
elements.
D
dx.doi.org/10.1021/ja4087648 | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX