Multimerization of cRGD peptides by click chemistry: Synthetic strategies, chemical limitations, and influence on biological properties

Article abstract of DOI:10.1002/cbic.201000386

Integrin α_vβ₃ is overexpressed on endothelial cells of growing vessels as well as on several tumor types, and so integrin-binding radiolabeled cyclic RGD pentapeptides have attracted increasing interest for in vivo imaging of α_vβ₃ integrin expression by positron emission tomography (PET). Of the cRGD derivatives available for imaging applications, systems comprising multiple cRGD moieties have recently been shown to exhibit highly favorable properties in relation to monomers. To assess the synthetic limits of the cRGD-multimerization approach and thus the maximum multimer size achievable by using different efficient conjugation reactions, we prepared a variety of multimers that were further investigated in vitro with regard to their avidities to integrin α_vβ₃. The synthesized peptide multimers containing increasing numbers of cRGD moieties on PAMAM dendrimer scaffolds were prepared by different click chemistry coupling strategies. A cRGD hexadecimer was the largest construct that could be synthesized under optimized reaction conditions, thus identifying the current synthetic limitations for cRGD multimerization. The obtained multimeric systems were conjugated to a new DOTA-based chelator developed for the derivatization of sterically demanding structures and successfully labeled with ⁶⁸Ga for a potential in vivo application. The evaluated multimers showed very high avidities-increasing with the number of cRGD moieties-in in vitro studies on immobilized α_vβ₃ integrin and U87MG cells, of up to 131-and 124-fold, respectively, relative to the underivatized monomer.

Full text of DOI:10.1002/cbic.201000386

DOI: 10.1002/cbic.201000386
Multimerization of cRGD Peptides by Click Chemistry:
Synthetic Strategies, Chemical Limitations, and Influence
on Biological Properties
Carmen Wꢀngler,*[a, b] Simone Maschauer,[c] Olaf Prante,[c] Martin Schꢀfer,[d]
Ralf Schirrmacher,[a] Peter Bartenstein,[b] Michael Eisenhut,[d] and Bjçrn Wꢀngler*[b, d]
Integrin anb3 is overexpressed on endothelial cells of growing
vessels as well as on several tumor types, and so integrin-bind-
ing radiolabeled cyclic RGD pentapeptides have attracted in-
creasing interest for in vivo imaging of anb3 integrin expression
by positron emission tomography (PET). Of the cRGD deriva-
tives available for imaging applications, systems comprising
multiple cRGD moieties have recently been shown to exhibit
highly favorable properties in relation to monomers. To assess
the synthetic limits of the cRGD-multimerization approach and
thus the maximum multimer size achievable by using different
efficient conjugation reactions, we prepared a variety of multi-
mers that were further investigated in vitro with regard to
their avidities to integrin anb3. The synthesized peptide multi-
mers containing increasing numbers of cRGD moieties on
PAMAM dendrimer scaffolds were prepared by different click
chemistry coupling strategies. A cRGD hexadecimer was the
largest construct that could be synthesized under optimized
reaction conditions, thus identifying the current synthetic limi-
tations for cRGD multimerization. The obtained multimeric sys-
tems were conjugated to a new DOTA-based chelator devel-
oped for the derivatization of sterically demanding structures
and successfully labeled with 68Ga for a potential in vivo appli-
cation. The evaluated multimers showed very high avidities—
increasing with the number of cRGD moieties—in in vitro stud-
ies on immobilized anb3 integrin and U87MG cells, of up to
131- and 124-fold, respectively, relative to the underivatized
monomer.
Introduction
Integrin anb3 is overexpressed on activated endothelial cells of
growing vessels and also on tumor cells such as melanoma,
glioma, lung, ovarian and breast cancers,[1–5] and so it repre-
sents a promising target structure for the localization of tumor
cells, their angiogenesis, and their metastasis by molecular
imaging methods.[6] Because cyclic peptides containing the se-
quence arginine-glycine-aspartic acid (RGD) specifically bind to
anb3 integrin with high affinities, many examples of radiola-
beled cyclic RGD (cRGD) peptides, mainly used for imaging of
anb3 expression in vivo, can be found.[7–11] However, many
monomeric cRGD peptide derivatives show relatively low
tumor accumulation, low tumor-to-background ratios, and
rapid tumor washout, limiting their applicability.[12,13] This can
mainly be attributed to their fast blood clearance, limiting the
probability of a contact with cell-surface integrins, and their
suboptimal affinities to anb3 integrin.
Several groups have therefore investigated radiolabeled
cRGD multimers—containing mostly two to four cRGD moiet-
ies—as agents for molecular imaging applications. It could be
shown in most of these studies that the avidity of the multi-
meric system was directly proportional to the number of cRGD
peptides conjugated to the scaffold structure, resulting in im-
proved tumor accumulation and retention.[12,14,15,17–21]
The potentiated avidities are, however, not the result of
higher molecular weights or larger molecular size but of an in-
creased cRGD content per molecule. Wester and co-workers,
who synthesized multimers containing one cRGD and several
[a] Dr. C. Wꢀngler, Prof. Dr. R. Schirrmacher
McConnell Brain Imaging Centre
Montreal Neurological Institute, McGill University
3801 University Street, Montreal, H3A 2B4 QC (Canada)
[b] Dr. C. Wꢀngler, Prof. Dr. P. Bartenstein, Dr. B. Wꢀngler
University Hospital Munich, Department of Nuclear Medicine
Ludwig Maximilians-University Munich
One strategy for overcoming these limitations is the applica-
tion of constructs containing multiple cRGD moieties, improv-
ing the avidities (affinities of multimeric systems) of the multi-
mers to the anb3 integrin and resulting in decelerated blood
clearance. This results in higher tumor uptakes, improved
tumor-to-background ratios, and prolonged tumor reten-
tion.[14–16] By following this approach it is possible to develop
radiotracers for in vivo imaging of various diseases that exhibit
improved uptake and retention characteristics in target struc-
tures such as anb3-integrin-expressing tumors or metastatic or
angiogenic tissues.
Marchioninistraße 15, 81377 Munich (Germany)
E-mail: carmen.waengler@med.uni-muenchen.de
bjoern.waengler@med.uni-muenchen.de
[c] Dr. S. Maschauer, Dr. O. Prante
Friedrich-Alexander University
Clinic of Nuclear Medicine, Laboratory of Molecular Imaging
Schwabachanlage 6, 91054 Erlangen (Germany)
[d] Dipl.-Ing. M. Schꢀfer, Prof. Dr. M. Eisenhut, Dr. B. Wꢀngler
German Cancer Research Center
Department of Radiopharmaceutical Chemistry
Im Neuenheimer Feld 280, 69120 Heidelberg (Germany)
2168
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
cRAD moieties exhibiting affinities of the multimeric systems
similar to that of the cRGD monomer, were able to show that
the increase in avidity could not be achieved by enhanced mo-
lecular size, but only by multimerization of peptides that spe-
cifically bind to the integrin.[6,22] The avidity enhancement of
the cRGD multimers is therefore a result of more probable and
stronger binding to the anb3 integrin.
ing,[23,24] but the type and length of linker also have a signifi-
cant influence. Consequently, the linker must neither be too
short (a minimum distance between two cRGD moieties of 20–
25 bond lengths has to be achieved for simultaneous binding),
nor too long (long PEG linkers decrease the affinity to the in-
tegrin),[14,17,24,33] whereas the binding interaction of the cRGD
moieties with the integrin is also influenced by the level of
rigidity.[33,34]
Although these systems show such favorable properties,
with many published examples for dimeric and tetrameric sub-
stances,[12–21,23–27] only one example of an octameric system can
be found.[21] In general, no reasons have been given why no
larger systems have been synthesized, but a possible explana-
tion is the difficult purification of the multimers reported in
one case,[25] which can be attributed to incomplete derivatiza-
tion, it having reported for several multimers—among these
the only example of a cRGD-octamer—that the synthesis was
accomplished by the dimerization of the corresponding smaller
multimers.[12,14,21]
To obtain highly symmetric scaffold structures with opti-
mized properties for the multimerization reactions, we synthe-
sized PAMAM (polyamidoamine) dendrimers containing one,
two, four, eight, 16, 32, and 64 amino functionalities on a PEG
linker possessing a protected thiol functionality for final deriva-
tization and radioactive labeling (Scheme 1).[29]
Besides favorable properties for the synthesis of peptide
multimers, this class of dendrimers shows—in underivatized
form—very low toxicities and immunogenicities in relation to
other dendritic compounds.[35,36] Their structures are rather
rigid and highly homogeneous, resulting in symmetrically
branched arrangements with constant distances between two
adjacent peptide moieties. In addition, the unnatural structures
and enhanced molecular weights of the compounds should
result in greater metabolic stability and decelerated blood-
clearance of the multimers.
A reason for the difficulties in achieving homogeneous mul-
timers may be the acid–amide coupling reaction used for con-
jugation of the peptide moieties to the glutamic acid and
lysine dendrimer scaffolds that are used for almost all of the
reported cRGD multimers.[12–19,21,26,27]
One promising approach to overcome this problem of in-
complete conjugation reactions is the use of so-called “click
chemistry” coupling reactions, which proceed very efficiently
and so should be useful in the synthesis of very large and fully
derivatized multimeric systems on dendritic scaffolds of
increasing size.
The derivatization of the dendritic systems was adapted to
achieve a minimum distance of 22 bond lengths between two
cRGD moieties, which should allow simultaneous binding of
cRGD peptide moieties to the anb3 integrin on the cell surface
and so should result in high achievable avidities.[21,37]
We set out to determine the limits of cRGD multimerization
by this synthesis strategy and thus the achievable maximum
number of cRGD moieties per multimeric system. Our aim was
to find the most suitable coupling technique for the synthesis
of large cRGD multimers beyond tetramers. Subsequently, we
characterized the obtained multimers in vitro in terms of their
binding avidities to anb3 integrin to assess their potential for in
vivo imaging with positron emission tomography (PET).
To determine the chemical limit of the cRGD multimerization
approach on dendritic scaffolds, we investigated different cou-
pling strategies in terms of the multimer size that could be
achieved. A very powerful synthetic approach for the coupling
of multiple chemical moieties to a scaffold structure is the use
of click chemistry reactions, which proceed efficiently under
mild conditions in aqueous media, giving physiologically
stable coupling products in reproducibly high yields.
Currently, the three most effective and often used click
chemistry reactions are: 1) copper-catalyzed 1,3-dipolar cyclo-
additions between alkynes and azides, 2) oxime formations be-
tween aminooxy-derivatized compounds and aldehydes, and
3) Michael additions of thiols to maleimides.[38] A direct com-
parison of these approaches for the synthesis of peptide multi-
mers was therefore performed in order to determine the most
efficient coupling strategy for the synthesis of peptidic multi-
mers and the chemical limit for cRGD multimerization on
PAMAM dendrimer backbones.
Results and Discussion
Synthesis of derivatized dendritic scaffolds
Because the scaffold structures on which the multimerization
reactions are carried out have a critical impact on the avidities
and thus on the biological properties of the obtained multi-
mers, these points require careful consideration. Ideally, the
scaffold structure itself should be nontoxic and nonimmuno-
genic and should be highly homogenous and symmetrical. An-
other important factor is the constitution of the cRGD moiet-
ies. It could be demonstrated that a branched scaffold struc-
ture producing a spherical alignment of the peptide moieties
over the multimer is more favorable than a linear multimeric
architecture.[23] The choice of an appropriate linker between
the cRGD moieties is also an important requirement for high
avidities of multimeric systems. Generally, a linker is required
for enhanced avidities, because a minimum distance between
cRGD moieties has to be achieved for simultaneous bind-
Dendrimers of different size containing one, two, four, eight,
16, 32, and 64 amino functionalities were synthesized and de-
rivatized with hex-5-ynoic acid, bis-Boc-aminooxy acetic acid,
and maleimidohexanoic acid to obtain conjugates with func-
tionalities suitable for click chemistry reactions as described
above (Scheme 2).
The synthesized dendritic structures each contain—besides
functionalities applicable in click chemistry reactions—a thiol
functionality (initially S-trityl-protected) with enhanced reactivi-
ty, which is achieved by means of a PEG linker to facilitate the
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2169

C. Wꢀngler, B. Wꢀngler et al.
Scheme 1. Structures (PAMAM-amine1 to PAMAM-amine16) and schematic depictions (PAMAM-amine32 and PAMAM-amine64) of the synthesized PAMAM den-
drimers used as scaffold structures for peptide multimerization.
reactions of synthons with the bulky dendritic systems
(Scheme 1). This thiol is used for the efficient and site-specific
Synthesis of cRGD peptide derivatives
introduction of
a
DOTA (1,4,7,10-tetraazacyclodocecane-
To be applicable in the proposed click chemistry reactions with
the derivatized dendritic systems, the cRGD peptides had to
contain an azide, aldehyde, and thiol functionality and in addi-
tion needed to be based on the general sequence c(RGDfX),
where X is cysteine or lysine, because these two derivatives
have been demonstrated to exhibit high affinities to anb3 in-
tegrin.[39–41] The cyclic RGD peptides based on this general
N,N’,N’’,N’’’-tetraacetic acid) chelator, enabling one-step radio-
labeling of the resulting multimers with radiometal nuclides
such as 68Ga. These derivatized dendritic compounds therefore
represent an ideal scaffold platform for a peptide multimeriza-
tion approach.
2170
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
Scheme 2. Derivatization of the PAMAM dendrimer scaffold structures with A) hex-5-ynoic acid (HBTU, DIPEA, DMF, RT, 30 min; 63–95% yields), B) bis-Boc-ami-
nooxy acetic acid (HBTU, DIPEA, DMF, RT, 30 min; 67–95% yields), and C) maleimidohexanoic acid (HBTU, DIPEA, DMF, RT, 30 min; 36–89% yields).
structure that were synthesized (Scheme 3) were c(RGDfK)-
azide (18), c(RGDfK)-aldehyde (19), and c(RGDfC) (20).
c(RGDfC) (20) was obtained by standard solid-phase Fmoc-
peptide synthesis as shown in Scheme 3. In the case of
c(RGDfK)-azide (18), Fmoc-Lys(Mtt)-OH was used instead of the
standard amino acid Fmoc-Lys(Boc)-OH and after deprotection
of the lysine amino side chain functionality with TFA in DCM
(1.75%), azidopentanoic acid was coupled before cleavage of
Scheme 3. Synthesis pathways for the cRGD derivatives 18, 19, and 20 used for multimerization reactions on the functionalized PAMAM dendrimer scaffolds
by click chemistry.
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2171

C. Wꢀngler, B. Wꢀngler et al.
the peptide from the resin. The synthesis of c(RGDfK)-aldehyde
(19) was found to be complicated, because neither the use of
an aldehyde-releasing resin for peptide synthesis (H-Val-H No-
vaSynꢂ TG resin) nor the coupling of a protected aldehyde (3,3-
bis-methoxypropionic acid)—analogously to azidopentanoic
acid in the synthesis of 18—gave the desired peptidic alde-
hydes but only the corresponding alcohols.
Synthesis of multimers through copper-catalyzed 1,3-dipolar
cycloadditions
Different conditions were tested for the synthesis of cRGD mul-
timers in reactions of this type, because it has been shown
before that the 1,3-dipolar cycloadditions are extremely depen-
dent on the catalytic system used.[25,42,43] Interestingly, none of
these reaction conditions resulted in the formation of the de-
sired products, although some have been used for the synthe-
sis of peptide multimers before. One possible explanation for
this might be the application of conventional heating instead
of the described microwave system, or it might be attributable
to interactions of the cyclic peptides with the catalytic copper
ions, hampering the coupling reaction.
The cRGD-aldehyde 19 was therefore synthesized by treat-
ment of the amino functionality of purified c(RGDfK) (19a) in
solution with SFB (p-formylbenzoic acid N-hydroxysuccinimide
ester) in aqueous medium at pH 7.5 for one hour (Scheme 3).
Under these conditions, the amino function of the cRGD pep-
tide reacts selectively with the active ester of SFB to give the
product in good yields of 56% after HPLC purification.
Finally, a combination of several methods resulted in prod-
uct formation: the alkyne 1 (dissolved in H2O/MeCN 1:2,
1 equiv) and cRGD-azide 18 (1–4 equiv) were reacted by apply-
ing a catalytic system consisting of CuI (1 equiv), sodium ascor-
bate (10 equiv), copper powder (10 equiv) and DIPEA (2 equiv)
in 60 min at room temperature to give the cRGD-monomer 21
in 55% yield after HPLC purification (Scheme 5).
Synthesis of cRGD multimers by different coupling
strategies
For the multimerization reactions, the PAMAM dendrimers de-
rivatized with alkyne, aminooxy, and maleimide functionalities
(Scheme 2) were treated (Scheme 3) with c(RGDfK)-azide (18),
c(RGDfK)-aldehyde (19), and c(RGDfC) (20), respectively, in
aqueous media under mild conditions.
For radiometal labeling with 68Ga, the synthesized multimers
should be derivatized with the chelator DOTA, which forms
very stable complexes with many M2+ and M3+ radionuclides.
For efficient introduction of the chelator into the anticipated
sterically demanding multimers, a DOTA derivative containing
a maleimide function and a PEG spacer for enhanced reactivity
towards the voluminous cRGD multimers was developed
(Scheme 4). This DOTA-derivative—named DOTA-M-PEG3—was
synthesized by treating thiol-DOTA[28] with BM(PEG)3 (1,11-bis-
maleimido-triethylene-glycol) in aqueous medium at pH 7.4.
The triazole formation worked well for the synthesis of the
monomer 21, although only one equivalent of 18 was used.
Surprisingly, the alkyne dimer 2 and the alkyne tetramer 3
showed only incomplete reactions, of one and one and two re-
acted alkyne functions, respectively, with the azide 18. Even
the use of an excess of up to four equivalents of 18 per alkyne
function did not result in the formation of fully c(RGD)-derivat-
ized multimeric dendritic systems. Another chemical limitation
was the formation of by-products, restricting the reaction time
to 60 min. Because even the small multimeric alkynes could
not be derivatized completely with peptide moieties, the reac-
tions of the alkyne octamer 4 and the alkyne hexadecimer 5
with the cRGD-azide were not performed.
Scheme 4. Synthesis of DOTA-M-PEG3 (30). Conditions: phosphate buffer/MeCN 2:3, pH 7.4, RT, 5 min; 61% yield.
Scheme 5. Reaction between the alkyne-derivatized scaffold 1 and c(RGDfK)-azide (18). Conditions: CuI, Cu, sodium ascorbate, DIPEA, H2O/MeCN 1:2, RT,
60 min, 55% yield.
2172
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
The obtained monomer RGD1-triazole-S-trityl (21) was not
further derivatized with DOTA-M-PEG3 (30) because only a full
set of c(RGD)-modified dendrimers would warrant further eval-
uation.
Reactions between the cRGD-aldehyde 19 and the depro-
tected aminooxy-derivatized dendritic systems (6a–8a,
Scheme 6B) proceeded efficiently even when a low excess of
the peptide derivative 19 was used. However, the largest ho-
mogeneously derivatized multimeric system that could be syn-
thesized by this strategy was the cRGD tetramer RGD4-oxime-
SH (24), because the reaction between the octameric amino-
oxy-derivatized dendritic system 9a and 19 resulted in a mix-
ture of the seven- and eightfold derivatized dendrimer. This in-
completely derivatized multimer could neither be separated
from the fully derivatized compound nor be induced to react
further to afford the desired product cRGD8-oxime-SH, even
with the use of prolonged reaction times and excesses of up
to four equivalents of 19 per aminooxy function. Because it
had not been possible to obtain the fully derivatized cRGD-
oxime octamer from the reaction between the aminooxy-octa-
mer 9a and 19, the reaction between the aminooxy-hexadeci-
mer 10a and 19 was not examined.
The incompleteness of the oxime formation in the case of
9a might be caused by steric hindrance or intermolecular in-
teractions (H-bonding) between the aminooxy moiety and
other functionalities within the multimeric system, hampering
its reaction with the peptide aldehyde 19.
During the peptide multimerization in aqueous medium at
pH 4–5, no unintended side-reactions of the deprotected
thiols—oxidation, for example—took place, which was shown
with the example of the cRGD monomer 22 (Scheme 7). The
free thiol group of 22 efficiently
Synthesis of multimers by oxime formation and derivatiza-
tion with DOTA-M-PEG3 for 68Ga labeling
An advantage of chemoselective oxime formation between al-
dehydes and aminooxy-bearing compounds is the compatibili-
ty of this reaction with aqueous media (in the absence or the
presence of organic solvents) under mild coupling conditions
at pH 4–5. The reaction does not need the addition of a cata-
lyst—such as heavy metal ions—which would have to be re-
moved properly before consideration of an application of the
resulting compound as part of an in vivo imaging protocol.
For multimerization, the bis-Boc-protected aminooxy-derivat-
ized dendrimers 6–10 were deprotected with neat trifluoroace-
tic acid, which removed not only the Boc but also the S-trityl
protecting groups (Scheme 6A). The corresponding intermedi-
ates (6a–10a) were not further characterized, and their subse-
quent treatment with the aldehyde 19 occurred chemoselec-
tively with the aminooxy functions of 6a–10a at pH 4–5.
Scheme 6 depicts the reaction pathway for the synthesis of the
cRGD multimers 22–24 from the starting aminooxy-derivatized
dendrimers 6–10. The corresponding octameric and hexadeci-
meric compounds could not be synthesized by this approach.
reacted with DOTA-M-PEG3 (30)
at pH 7.4 to give compound 22a
in 73% yield after HPLC purifica-
tion.
Peptide
multimerization
through oxime formation—al-
though it proceeded efficiently
in the synthesis of the cRGD
mono- to tetramers 22–24—
therefore also did not result in
the formation of a whole set of
multimers for further 68Ga-radio-
labeling and in vitro evaluation.
Synthesis of cRGD multimers
by Michael addition and deri-
vatization with DOTA-M-PEG3
for 68Ga-labeling
Another method tested for cRGD
multimerization was the addition
of thiols to maleimides.
Coupling reactions between
the peptide thiol c(RGDfC) (20)
and the maleimide-derivatized
dendrimers 11–17, containing
one, two, four, eight, 16, 32 and
64 maleimide units, proceeded
very efficiently within five min-
Scheme 6. Reaction pathway for the cRGD multimers 22–24, starting from the aminooxy-derivatized dendritic sys-
tems (6–10). Conditions: A) TFA/TIS/H2O 95:2.5:2.5, RT, 60 min; B) phosphate buffer/MeCN 3:1, pH 4.5–4.8, RT,
60 min; 15–41% overall yield.
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2173

C. Wꢀngler, B. Wꢀngler et al.
Scheme 7. Reaction between monomeric cRGD1-oxime-SH (22) and DOTA-M-PEG3 (30), yielding conjugate RGD1-oxime-DOTA (22a) through thiol–maleimide
coupling. Conditions: phosphate buffer/MeCN 3:1, pH 7.4, RT, 10 min, 73% yield.
utes at room temperature in aqueous medium at pH 7.4, en
abling the synthesis of large multimeric systems (Scheme 8).
The largest peptide multimer that could be isolated and
characterized was the cRGD hexadecimer 35, because the
products of the reactions of the dendrimers containing 32 and
64 maleimide functionalities (16 and 17) with 20 were only
poorly soluble and thus difficult to analyze and most probably
not fully derivatized.
PAMAM-dendrimer scaffolds by application of the highly effi-
cient thiol–maleimide click chemistry coupling reaction. To the
best of our knowledge, the obtained hexadecimer 35 repre-
sents the largest cRGD multimer that has been synthesized up
until now. The presented cRGD multimers could easily be la-
beled with 68Ga in radiochemical yields between 95–98% and
radiochemical purities between 96–99%, a prerequisite for use
of these compounds as potential imaging agents for in vivo
imaging of integrin-positive tumors. To validate the assump-
tion that the avidity, even in large peptide multimers, is direct-
ly proportional to the number of cRGD peptides conjugated
(as a result of the rigid backbone structure and an optimal dis-
tance between two cRGD moieties allowing for simultaneous
binding of multiple cRGD moieties to the anb3 integrin), in
vitro competition binding experiments were conducted.
This observed solubility issues relating to peptide multimers
of very high molecular weight in aqueous solution not only
define a synthetic threshold but also point to limitations for
their in vitro evaluation and in vivo application.
The maleimido-thiol click reaction was therefore the only
synthesis pathway that finally provided a complete set of
cRGD multimers suitable for radiolabeling and in vitro evalua-
tion. These compounds provided the necessary contingent of
multimer homologues for investigation of the correlation be-
tween their integrin binding avidities and individual configura-
tions, based on the size and number of contained cRGD moiet-
ies.
For radiometal labeling with 68Ga, the synthesized cRGD
multimers 25–29 were derivatized with 30 after S-trityl remov-
al, accomplished with a mixture of TFA and TIS for 5 min to
yield the thiol intermediates 25a–29a for the maleimide-thiol
addition reaction with the chelator derivative 30.
Compound 30 was treated with the deprotected multimeric
intermediates 25a–29a in aqueous medium at pH 7.4 for
10 min at room temperature to give the DOTA-derivatized
cRGD multimers 31–35 in overall yields of 32–56% based on
the maleimide-derivatized PAMAM-dendrimer scaffolds 11–17
after HPLC purification (Scheme 8).
In vitro evaluation of the cRGD-MT-DOTA multimers 31–35
with immobilized integrin and U87MG cells
In order to assess the PET imaging potential of the synthesized
cRGD multimers obtained by maleimido-thiol click chemistry,
we determined their avidities to anb3 integrin in vitro. We de-
termined the affinities of the underivatized c(RGDfC) 20, the
cRGD monomer 31, and the multimers 32–35 by use of two
different competition receptor binding assays utilizing 125I-
echistatin: anb3-integrin-coated microplates and anb3-integrin-
expressing U87MG cells. The results of these binding experi-
ments are shown in Figure 1 and Table 1.
As a result of the presence of multiple integrins on U87MG
cells and the use of the highly potent but unselective disinte-
grin 125I-echistatin as a radioligand, the obtained Ki values differ
between the assay systems (Figure 1, Table 1). Congruently, in
both systems, the avidities are in principle potentiating with
increasing number of cRGD moieties and the magnitude of
avidity enhancement by multimerization is similar in the two
assay systems.
Scheme 9 shows the structure of the largest synthesized
cRGD multimer—cRDG16-MT-DOTA (35)—obtained from the
maleimido-thiol coupling approach.
These results show that cRGD multimers containing up to
sixteen c(RGDfC) moieties can successfully be synthesized on
2174
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
Scheme 8. Reaction pathway for the cRGD-MT-DOTA multimers 31–35 starting from the maleimido-derivatized dendritic systems 11–17. Conditions: A) phos-
phate buffer/MeCN 2:1, pH 7.4, RT, 10 min; B) TFA/TIS/H2O 95:2.5:2.5; RT, 5 min; C) phosphate puffer, pH 7.4, RT, 10 min; 32–56% overall yields.
However, the solid-phase binding assay on immobilized anb3
integrin-coated plates was found to reflect the true binding af-
finities of the larger sized cRGD multimers 34 and 35 inade-
quately, as a result of the two-dimensionality of the integrin
availability. This assay revealed no increase in avidity compar-
ing the cRGD octamer 34 with the cRGD hexadecimer 35, indi-
cating limitations of this assay system for multimeric ligands
such as 35, due to the lack of binding sites in all directions of
space. The binding assay with U87MG cells therefore more ac-
curately reflects the in vivo situation, especially because the
binding of multimeric cRGD peptides on surface integrins of
whole cells enables multivalent binding as well as clustering of
integrins.[12,44] This assay revealed a binding affinity of 35
almost five times higher than that of 34, in line with the antici-
pated amplified avidity as a result of the number of available
peptidic integrin binders.
These results are in line with those found for the only cRGD
octamer synthesized so far, which showed an anb3 integrin
avidity ten times higher than that of the corresponding
dimer.[21]
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2175

C. Wꢀngler, B. Wꢀngler et al.
Scheme 9. Structure of the largest synthesized cRGD multimer cRGD16-MT-DOTA (35). Violet: dendrimer core, blue: dendrimer derivatization with maleimido-
hexanoic acid, green: conjugated cRGD peptides, orange: linker to the chelator, red: DOTA chelator.
With the hexadecimer shown here, these favorably high
avidities obtained for the octameric system can even be in-
creased by a factor of five, resulting in an overall magnitude of
avidity enhancement of about 125-fold relative to the underiv-
atized c(RGDfC) binding with high affinity to the integrin.
The observed enhancement of avidity with increasing num-
bers of cRGD moieties can be attributed to several factors.
1) The first is multivalent binding, which is enabled because
the minimum distance between two cRGD moieties in the
cRGD multimers described here is 32 bond lengths.[20,21,37]
2) The higher concentration of peptide relative to the integrin
as a function of the degree of multimerization also leads to an
enhancement of avidity because the binding probability is in-
creased by constant rebinding, resulting in higher retention
and slower washout rates.[15] 3) The multivalent binding is pre-
sumably also a consequence of integrin clustering initiated by
cRGD binding to the cell surface.[15,20,23]
From these in vitro results, the cRGD hexadecimer 35 shows
the most favorable properties in terms of avidity to anb3 integ-
rin and corroborates the assumption that the available number
of cRGD moieties in a multimer determines its avidity as a
result of cooperative binding.
Conclusions
cRGD multimers containing one, two, four, eight, and 16
c(RGDfC) moieties could be synthesized on PAMAM-dendrimer
scaffolds of increasing size through the use of Michael addi-
tions of thiols to maleimides as the most efficient of the tested
click chemistry coupling reactions. Synthesis approaches apply-
ing oxime formation or 1,3-dipolar cycloaddition for multimeri-
zation yielded only smaller multimeric systems or the mono-
mer, respectively.
2176
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
ties with increasing number of cRGD moieties with a more
than 100-fold increase in avidity relative to the underivatized
peptide c(RGDfC).
The multimerization of cRGD peptides thus results in com-
pounds showing drastically enhanced avidities to the anb3 in-
tegrin. These multimeric systems are promising tracers for
study of their in vivo pharmacodynamic properties by small-
animal PET for application in tumor, angiogenesis, and meta-
stasis imaging. In addition, the chemistry presented and the in
vitro evaluation of the obtained cRGD multimers should set
the foundation for the development of novel potent peptide-
multimer-based imaging agents for PET.
Experimental Section
General: All commercially available chemicals were of analytical
grade and were used without further purification. 125I-Echistatin
was purchased from Perkin–Elmer (NEX4390). The resin for peptide
synthesis (Fmoc-Asp-(NovaSynꢂ TGA)-OAll resin), coupling reagents,
and Fmoc-protected amino acids were purchased from NovaBio-
chem. PAMAM-dendrimer scaffolds, maleimido-functionalized den-
drimers (11–17), and thiol-DOTA were synthesized by published
procedures.[28,29]
The analytical and semipreparative HPLC system used was an Agi-
lent 1200 system fitted with a Raytest Gabi Star radioactivity detec-
tor. The columns used for chromatography were a Chromolith Per-
formance (RP-18e, 100–4.6 mm, Merck, Germany) and a Chromolith
(RP-18e, 100–10 mm, Merck, Germany) column, operated with
flows of 4 and 8 mLminÀ1, respectively.
Figure 1. Results of in vitro binding experiments with underivatized
c(RGDfC) 20, the monomer 31, and the multimers 32–35 and immobilized
anb3 integrin and integrin-expressing human glioblastoma U87MG cells. Re-
sults were obtained from three independent experiments each performed in
triplicate (n=9).
ESI, MALDI, and NMR spectra were obtained with a Finnigan
MAT95Q, a Bruker Daltonics Microflex, and a Jeol AS500 spectrom-
eter, respectively.
Table 1. Results of in vitro binding assays with 125I-echistatin and immo-
bilized aνb3 integrin and U87MG cells.
Immobilized anb3 integrin
U87MG cells
The human glioblastoma cell line U87MG was purchased from the
American Type Culture Collection (ATCC, HTB-14). Fetal calf serum
(FCS), Dulbecco’s modified Eagle’s Medium (DMEM) and nonessen-
tial amino acids (NAAs) were obtained from Invitrogen (Karlsruhe,
Germany).
Ligand
Ki
Ligand
Ki
[nm]ÆSD[a]
[nm]ÆSD[a]
1237Æ240
c(RGDfC) (20)
34Æ10
c(RGDfC) (20)
cRGD1-MT-DOTA (31) 32Æ6
cRGD1-MT-DOTA (31) 1627Æ199
cRGD2-MT-DOTA (32) 522Æ57
cRGD4-MT-DOTA (33) 248Æ48
cRGD2-MT-DOTA (32) 13Æ2
General synthesis of alkyne-derivatized dendrimers (1–5): A so-
lution of HBTU (O-(benzotriazol-1-yl)-N,N,N’,N’-tetramethyluronium
hexafluorophosphate, 3.9 equiv per amino function of the corre-
sponding dendrimer) in DMF (250–500 mL) was added to hex-5-
ynoic acid (4 equiv per amino function) followed by DIPEA (diiso-
propylethylamine, 4 equiv per amino function) and allowed to
react for two minutes. Subsequently, the mixture was added to a
solution of the corresponding dendrimer (7.9–17.1 mmol) in DMF
(200 mL) and allowed to react for 30 min. The products were puri-
fied by semipreparative HPLC with 35–90% MeCN+0.1% TFA over
6 min as the gradient. The products were obtained as colorless to
brown oils after lyophilization in yields of 63 to 95%.
cRGD4-MT-DOTA (33)
1.3Æ0.1
cRGD8-MT-DOTA (34)
cRGD16-MT-DOTA (35) 0.26Æ0.04 cRGD16-MT-DOTA
0.32Æ0.04 cRGD8-MT-DOTA (34)
46Æ5
10Æ1
(35)
[a] Values are given as the meanÆstandard deviation (SD) of three inde-
pendent experiments performed in triplicate.
For radiometal labeling, the cRGD multimers were conjugat-
ed to DOTA-M-PEG3, a DOTA derivative containing a PEG
spacer developed to achieve enhanced reactivities with steri-
cally demanding systems and a maleimido moiety for chemo-
selective ligation to thiols.
PAMAM-alkyne1 (1):
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.44–7.42 (m, 6H;
H-2), 7.36–7.32 (m, 6H; H-3), 7.27–7.24 (m, 3H; H-1), 7.09 (brs, 1H;
The largest definite multimer that could be obtained—and
to the best of our knowledge the largest cRGD multimer so far
synthesized—was a hexadecimer. Solubility problems observed
for larger constructs result in a severe chemical limitation for
the synthesis of even larger multimers.
The dendritic cRGD multimers were studied in relation to
their avidities to anb3 integrin in vitro, showing enhanced avidi-
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2177

C. Wꢀngler, B. Wꢀngler et al.
H-20), 3.58–3.48 (m, 20H; H-8 to H-17), 3.43–3.41 (m, 2H; H-18),
3.35–3.30 (m, 4H; H-7+H-19), 2.38 (t, J(H,H)=6.7, 2H; H-6), 2.35
2.22 (m, 8H; H-36), 1.82–1.79 ppm (m, 8H; H-35); 13C NMR
(125 MHz, [D6]acetone, 258C, TMS): d=173.52 (C-33), 168.55+
168.23 (C-22+C-28), 146.41 (C-4), 131.14 (C-2), 129.83 (C-3), 128.75
(C-1), 86.27 (C-37), 71.02+70.98+70.89+70.86+70.55+70.71+
70.12+70.09+69.77+69.68 (C-7–C-18+C-38), 68.19 (C-5), 45.71
(C-30), 45.42 (C-31), 42.31 (C-26), 40.23 (C-19), 36.12 (C-21), 36.10
(C-27), 35.97 (C-34), 33.16 (C-6), 25.99 (C-35), 18.77 ppm (C-36); MS
(ESI): m/z: calcd: 1644.94 [M+H]+; found: 1644.95.
3
4
3
(t, J(H,H)=2.7, 1H; H-26), 2.27 (t, J(H,H)=7.4, 2H; H-22), 2.21 (dt,
3J(H,H)=7.2, J(H,H)=2.7, 2H; H-24), 1.77 ppm (p, J(H,H)=7.3, 2H;
H-23); 13C NMR (125 MHz, [D6]acetone, 258C, TMS): d=172.26 (C-
21), 145.91 (C-4), 130.41 (C-2), 128.75 (C-3), 127.54 (C-1), 84.54 (C-
25), 71.23+71.21+71.17+71.09+70.91+70.50+70.23+70.00 (C-
7–C-18+C-26), 67.22 (C-5), 39.79 (C-19), 35.20 (C-22), 32.64 (C-6),
25.42 (C-23), 18.36 ppm (C-24); MS (ESI): m/z: calcd: 700.33
[M+Na]+; found: 700.33.
4
3
PAMAM-alkyne8 (4): MS (ESI): m/z: calcd: 1467.37 [M+2H]2+
;
found: 1467.38.
PAMAM-alkyne2 (2):
PAMAM-alkyne16 (5): MS (ESI): m/z: calcd: 1378.59 [M+4H]4+
;
found: 1378.60.
General synthesis of aminooxy-derivatized dendrimers (6–10): A
solution of HBTU (3.9 equiv per amino function of the correspond-
ing dendrimer) in DMF (250–500 mL) was added to bis-(Boc-amino-
oxy)acetic acid (NovaBiochem, 4 equiv per amino function), fol-
lowed by DIPEA (4 equiv per amino function), and allowed to react
for 2 min. Subsequently, the mixture was added to a solution of
the corresponding dendrimer (7.9–17.1 mmol) in DMF (200 mL) and
allowed to react for 30 min. The products were purified by semi-
preparative HPLC with 35–90% MeCN+0.1% TFA in 6 min as the
gradient. The products were obtained as colorless oils after lyophi-
lization in yields of 67 to 95%.
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.44–7.41 (m, 6H;
H-2), 7.36–7.32 (m, 6H; H-3), 7.27–7.25 (m, 3H; H-1), 3.62–3.22 (m,
34H; H-7–H-21+H-24+H-25), 2.38 (t, 3J(H,H)=6.7, 2H; H-6), 2.36
PAMAM-bisBoc-aminooxy1 (6):
4
3
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.64 (brs, 1H; H-
20), 7.44–7.42 (m, 6H; H-2), 7.36–7.32 (m, 6H; H-3), 7.27–7.24 (m,
(t, J(H,H)=2.6, 2H; H-32), 2.27 (t, J(H,H)=7.4, 4H; H-28), 2.21 (dt,
4
3
3J(H,H)=7.3, J(H,H)=2.7, 4H; H-30), 1.78 ppm (p, J(H,H)=7.3, 4H;
H-29); 13C NMR (125 MHz, [D6]acetone, 258C, TMS): d=173.01 (C-
27), 167.88 (C-22), 145.73 (C-4), 130.11 (C-2), 128.67 (C-3), 127.51 (C-
1), 86.13 (C-31), 71.27+71.26+71.24+71.16+71.05+70.96+
70.82+70.33+69.89 (C-7–C-18+C-32), 68.17 (C-5), 48.23 (C-20),
46.52 (C-24), 46.06 (C-25), 39.99 (C-19), 35.20 (C-28), 34.56 (C-21),
33.04 (C-6), 25.71 (C-29), 18.39 ppm (C-30); MS (ESI): m/z: calcd:
1000.54 [M+H]+; found: 1000.55.
3H; H-1), 4.42 (s, 2H; H-22), 3.61–3.51 (m, 20H; H-8–H-17), 3.45–
3.42 (m, 4H; H-7+H-18), 3.31 (t, 3J(H,H)=6.7, 2H; H-19), 2.38 (t,
3J(H,H)=6.7, 2H; H-6), 1.54 ppm (s, 18H; H-25); 13C NMR (125 MHz,
[D6]acetone, 258C, TMS): d=164.69 (C-21), 148.19 (C-23), 143.24 (C-
4), 127.75 (C-2), 126.09 (C-3), 124.87 (C-1), 82.28 (C-24), 74.05 (C-
22), 68.58+68.55+68.54+68.53+68.50+68.39+68.37+68.23+
67.58+67.33 (C-7–C-18), 64.57 (C-5), 36.63 (C-19), 29.96 (C-6),
25.51 ppm (C-25); MS (ESI): m/z: calcd: 879.41 [M+Na]+; found:
879.41.
PAMAM-alkyne4 (3):
PAMAM-bisBoc-aminooxy2 (7):
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.44–7.40 (m, 6H;
H-2), 7.36–7.33 (m, 6H; H-3), 7.27–7.25 (m, 3H; H-1), 4.51 (s, 4H; H-
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.45–7.42 (m, 6H;
H-2), 7.37–7.33 (m, 6H; H-3), 7.28–7.24 (m, 3H; H-1), 3.58–3.18 (m,
42H; H-7–H-21+H-24–H-27+H-30+H-31), 2.40 (t, 3J(H,H)=6.6,
2H; H-6), 2.38–2.36 (m, 4H; H-38), 2.29–2.25 (m, 8H; H-34), 2.26–
2178
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
28), 3.62–3.29 (m, 26H; H-7–H-21+H-24+H-25), 2.42 (t, 3J(H,H)=
6.6, 2H; H-6), 1.52–1.48 ppm (m, 36H; H-31); 13C NMR (125 MHz,
[D6]acetone, 258C, TMS): d=166.89 (C-22), 164.55 (C-27), 147.79 (C-
29), 143.42 (C-4), 127.79 (C-2), 126.01 (C-3), 124.96 (C-1), 82.19 (C-
30), 74.10 (C-28), 70.21+70.19+70.15+70.02+69.89+69.71+
69.63+69.23+69.19+69.02 (C-7–C-18), 65.36 (C-5), 49.72 (C-20),
46.71 (C-24), 46.25 (C-25), 36.33 (C-19), 34.91 (C-21), 30.12 (C-6),
26.91 ppm (C-31); MS (ESI): m/z: calcd: 1358.70 [M+H]+; found:
1358.71.
was allowed to react with azidopentanoic acid, HBTU, and DIPEA
as described above. The peptide was cleaved from the resin by in-
cubation with a mixture of TFA (trifluoroacetic acid)/TIS (triisopro-
pylsilane)/H2O (95:2.5:2.5, 2 mL) for 90 min, suspended in diethyl
ether and purified by semi-preparative HPLC with use of a gradient
of 0–40% MeCN+0.1% TFA over 6 min. The product was isolated
as a white solid after lyophilization (12.8 mg, 17.6 mmol, 18%
yield). MS (ESI): m/z: calcd: 729.37 [M+H]+; found: 729.38.
Synthesis of c(RGDfK) (19a): The peptide was synthesized analo-
gously to 18 by use of Fmoc-Lys(Boc)-OH instead of Fmoc-Lys-
(Mtt)-OH. The peptide was cleaved from the resin as described
above, purified by semipreparative HPLC with use of a
gradient of 0–40% MeCN+0.1% TFA over 6 min and iso-
lated as a white solid after lyophilization (25.8 mg,
42.8 mmol, 43% yield). MS (ESI): m/z: calcd: 604.31
[M+H]+; found: 604.32.
PAMAM-bisBoc-aminooxy4 (8):
Synthesis of c(RGDfK)-aldehyde (19): A solution of SFB
(p-formylbenzoic acid N-hydroxysuccinimide ester,
Sigma–Aldrich, 4.1 mg, 16.6 mmol) in DMF (300 mL) was
added to a solution of 19a (10 mg, 16.6 mmol) in phos-
phate buffer (0.1m, pH 5.0, 500 mL) resulting in a pH of
the reaction mixture of 7.5. After 60 min, the product
was purified by semipreparative HPLC with 0–40%
MeCN+0.1% TFA in 6 min as the gradient and obtained
as a white solid after lyophilization (6.8 mg, 9.2 mmol,
56%). MS (ESI): m/z: calcd: 736.33 [M+H]+; found:
736.34.
Synthesis of c(RGDfC) (20): The peptide was synthesized
analogously to 18 by use of Fmoc-Cys(Trt)-OH instead of
Fmoc-Lys(Mtt)-OH. The resulting peptide was cleaved
from the resin as described above and was purified by
semipreparative HPLC with use of a gradient of 0–40%
MeCN+0.1% TFA over 6 min. The product was isolated
as a white solid after lyophilization (18.3 mg, 31.6 mmol,
32% yield). MS (ESI): m/z: calcd: 579.23 [M+H]+; found:
579.23.
1H NMR (500 MHz, [D6]acetone, 258C, TMS): d=7.45–7.42 (m, 6H;
H-2), 7.35–7.30 (m, 6H; H-3), 7.26–7.22 (m, 3H; H-1), 4.40–4.36 (m,
8H; H-34), 3.65–3.19 (m, 42H; H-7–H-21+H-24–H-27+H-30+H-
Synthesis of cRGD1-triazole-STrityl (21): Compound 18
(3.0 mg, 4.1 mmol), CuI (0.8 mg, 4.1 mmol), sodium ascorbate
(8.2 mg, 41.3 mmol), copper powder (2.6 mg, 41 mmol), and DIPEA
(1.4 mL, 8.2 mmol) were added to a solution of alkyne1 (1, 2.8 mg,
4.1 mmol) in H2O/MeCN 1:2 and allowed to react for 60 min. The
product was purified by semipreparative HPLC with 0–40%
MeCN+0.1% TFA in 6 min as the gradient and obtained as a
white solid after lyophilization (3.2 mg, 2.3 mmol, 55% yield). MS
(ESI): m/z: calcd: 1406.71 [M+H]+; found: 1406.72.
3
31), 2.33 (t, J(H,H)=6.6, 2H; H-6), 1.53–1.47 ppm (m, 72H; H-37);
13C NMR (125 MHz, [D6]acetone, 258C, TMS): d=166.69 (C-22),
166.36 (C-28), 163.94 (C-33), 147.63 (C-35), 143.61 (C-4), 127.62 (C-
2), 125.37 (C-3), 124.74 (C-1), 81.85 (C-36), 73.88 (C-34), 71.56+
71.52+71.49+71.40+71.29+71.14+71.10+71.00
(C-7–C-18),
64.78 (C-5), 49.16 (C-20), 45.48 (C-30), 44.94 (C-31), 43.19 (C-26),
36.17 (C-19), 29.76 (C-6), 27.03 ppm (C-37); MS (ESI): m/z: calcd:
1181.13 [M+2H]2+; found: 1181.14.
General synthesis of the multimers cRGD1-oxime-SH, cRGD2-
oxime-SH, and cRGD4-oxime-SH (22–24): The bis-Boc-aminooxy-
derivatized dendrimers (6–8, 5.3–16.2 mmol) were deprotected by
addition of a mixture of TFA/TIS/H2O (95:2.5:2.5, 3–5 mL) and sub-
sequent reaction for 60 min. The volatile components were evapo-
rated and the resulting solids were suspended in water (250–
500 mL). The insoluble trityl-silane adduct was removed by centrifu-
gation and the products were purified by semipreparative HPLC
with 0–30% MeCN+0.1% TFA in 6 min as the gradient. The HPLC
solvent was removed and the aminooxy-derivatized dendrimers
(3.7–5.9 mmol) were dissolved in H2O (250 mL). A solution of 19
(1.25 equiv per aminooxy function) in a mixture of MeCN (200 mL)
and phosphate buffer (0.5m, pH 4, 400 mL) was added to this solu-
tion. The resulting reaction mixtures exhibited pH values of 4.5 to
4.8 and were allowed to react for 60 min. The products were puri-
fied by semipreparative HPLC with 0–50% MeCN+0.1% TFA in
PAMAM-bisBoc-aminooxy8 (9): MS (ESI): m/z: calcd: 1456.13
[M+3H]3+; found: 1456.13.
PAMAM-bisBoc-aminooxy16 (10): MS (ESI): m/z: calcd: 1197.52
[M+7H]7+; found: 1197.50.
Synthesis of c(RGDfK)-azide (18): The peptide was synthesized on
solid support by standard Fmoc solid-phase peptide synthesis as
described by Wellings et al.[30] on a Fmoc-Asp-(NovaSynꢂ TGA)-OAll
resin together with Fmoc-Lys(Mtt)-OH as lysine derivative. Allyl de-
protection was carried out as published elsewhere[31] and the sub-
sequent cyclization of the peptide was accomplished by incubating
the peptide on resin (100 mmol) with HBTU (0.95 equiv, 95 mmol,
36.0 mg) and DIPEA (1 equiv, 100 mmol, 17.1 mL) in DMF (1.5 mL)
for 16 h. The unprotected lysine side chain amino functionality was
obtained by selective Mtt removal with TFA in DCM (1.75%), and
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2179

C. Wꢀngler, B. Wꢀngler et al.
6 min as the gradient and were obtained as white solids after lyo-
philization in overall yields of 15–41%.
pH 7.4, 250 mL). After 5 min the product was purified by semipre-
parative HPLC with 0–20% MeCN+0.1% TFA in 6 min as the gradi-
ent. The product was obtained as white, hygroscopic powder after
lyophilization in 61% yield.
MS (ESI) for cRDG1-oxime-SH (22): m/z: calcd: 1132.53 [M+H]+;
found: 1132.53.
MS (ESI) for cRDG2-oxime-SH (23): m/z: calcd: 1076.01 [M+2H]2+
found: 1076.02.
;
;
MS (ESI) for cRDG4-oxime-SH (24): m/z: calcd: 838.40 [M+5H]5+
found: 838.41.
Synthesis of cRGD1-oxime-DOTA (22a): cRGD1-oxime-SH (22,
2.4 mg, 2.1 mmol) was dissolved in a mixture of phosphate buffer
(0.1m, pH 5.0, 250 mL) and MeCN (250 mL) and added to a solution
of DOTA-M-PEG3 (30, 2.2 mg, 2.7 mmol) in phosphate buffer (0.1m,
pH 5.0, 500 mL). The pH of the reaction mixture was adjusted to 7.4
by the addition of phosphate buffer (0.5m, pH 7.4, 100 mL). After
10 min, the product was purified by semipreparative HPLC with 0–
60% MeCN+0.1% TFA in 6 min as the gradient. The product was
obtained as a white solid after lyophilization (3.0 mg, 1.5 mmol,
73% yield). MS (ESI): m/z: calcd: 993.42 [M+H+K]2+; found: 993.41;
calcd: 1004.41 [M+Na+K]2+; found: 1004.40.
1H NMR (500 MHz, [D6]DMSO, 258C, TMS): d=8.71 (brs, 1H; H-17),
7.03 (s, 2H; H-1), 4.05 (brs, 2H; H-16), 4.03 (dd, 3J1(H,H)=3.9,
3J2(H,H)=9.0, 1H; H-14), 3.91 (brs, 2H; H-19), 3.60 (brs, 6H; H-22),
3.58–3.31 (m, 32H; H-3–H-10+H-20+H-21), 3.21 (q, 3J(H,H)=9.1,
3
1H; H-13), 2.79–2.74 (m, 2H; H-15), 2.53 ppm (d, J(H,H)=3.9, 1H;
H-13); 13C NMR (125 MHz, [D6]DMSO, 258C, TMS): d=175.63 (C-23),
173.85 (C-18), 169.88 (C-2), 157.10 (C-11), 156.84 (C-12), 133.52 (C-
1), 68.63+68.60+68.35+68.31+65.90+65.14 (C-4–C-9), 53.60+
52.93+51.66+51.64+49.45+49.43+49.41+49.38+47.47+47.21
(C-19–C-22), 37.28 (C-3), 36.77 (C-10), 35.76 (C-13), 34.76 (C-15),
39.32 ppm (C-16); MS (ESI): m/z: calcd: 816.34 [M+H]+; found:
816.35.
General synthesis of multimer intermediates (25–29) and cRGD
multimers cRGD1-MT-DOTA, cRGD2-MT-DOTA, cRGD4-MT-DOTA,
cRGD8-MT-DOTA and cRGD16-MT-DOTA (31–35): A solution of 20
(1.5 equiv per maleimide) in phosphate buffer (0.1m, pH 7.4,
400 mL) was added to a solution of a maleimide-derivatized
PAMAM-dendrimer (11–17, 0.25–5.0 mmol) in MeCN (200 mL). The
pH values of the mixtures were adjusted to 7.4 by addition of
phosphate buffer (0.5m, pH 7.4, 100 mL). After 10 min, the conjuga-
tion reactions were finished and the solvent was evaporated. The
residual solids were dissolved in a mixture of TIS (200 mL) and TFA
(5 mL) and allowed to react for 5 min at room temperature. The
volatile components were evaporated and the products were puri-
fied by semipreparative HPLC with 0–40% MeCN+0.1% TFA in
6 min as the gradient. After evaporation of the HPLC solvent, the
obtained solids were allowed to react with DOTA-maleimide (30,
2 equiv) in phosphate buffer (0.1m, pH 7.2, 500 mL) for 10 min and
the products were purified by semipreparative HPLC with 0–40%
MeCN+0.1% TFA in 6 min as the gradient. The products were ob-
tained as white solids after lyophilization in overall yields of 32–
56%.
Radiolabeling reactions with 68Ga: An aliquot of 68GaCl3 (200 MBq
in 400 mL, 0.6m HCl), freshly eluted from a 68Ge/68Ga generator (IDB
Holland BV, The Netherlands/iThemba LABS, South Africa) was
added to the corresponding cRGD multimer (31–35, 10 nmol) in
acetate buffer (1.25m, 300 mL), resulting in a final pH of 4.5. After
incubation for 10 min at 958C, the solution was neutralized by the
addition of sodium bicarbonate (1m, 300 mL). The radiochemical
yields and purities of the 68Ga-labeled multimers were analyzed by
radio-HPLC with 10–50% MeCN+0.1% TFA in 5 min as the gradi-
ent.
All radiolabeled peptide multimers were found to be 96–99%
pure.
Cell culture: The human glioblastoma cell line U87MG (ECACC
N8 89081402) was grown in Dulbecco’s modified Eagle’s medium
(DMEM) containing glucose (4500 mgLÀ1) and supplemented with
non-essential amino acids (1%), sodium pyruvate (1 mm), and fetal
bovine serum (FBS, 10%) at 378C in a humidified CO2 (5%) atmos-
phere. Cells were routinely subcultured every 3–4 days. For anb3
cell binding assays, U87MG cells were harvested, suspended in
growth medium containing DMSO (10%) at a concentration of
500000 cells per tube and stored at À808C until experimental use.
MS (ESI) for cRDG1-MT-DOTA (31): m/z: calcd: 964.92 [M+2H]2+
found: 964.92; calcd: 643.61 [M+3H]3+; found: 643.62.
;
;
;
MS (ESI) for cRDG2-MT-DOTA (32): m/z: calcd: 976.76 [M+3H]3+
found: 977.11; calcd: 732.82 [M+4H]4+; found: 733.08.
MS (ESI) for cRDG4-MT-DOTA (33): m/z: calcd: 986.24 [M+5H]5+
found: 986.85; calcd: 822.03 [M+6H]6+; found: 822.54; calcd:
704.74 [M+7H]7+; found: 705.04.
anb3 solid-phase receptor binding assay: The receptor binding
assay with anb3-coated microplates was performed as described
previously.[32]
MS (ESI) for cRDG8-MT-DOTA (34): m/z: calcd: 1275.86 [M+7H]7+
found: 1276.73; calcd: 1116.51 [M+8H]8+; found: 1117.14; calcd:
992.56 [M+9H]9+ found: 993.13; calcd: 893.40 [M+10H]10+
found: 893.91; calcd: 812.28 [M+11 H]11 +; found: 812.38.
;
;
;
Cell binding assay: The determination of integrin binding affinity
with cell suspensions was performed by the procedure described
previously.[32] In brief, U87MG cells were quickly thawed and centri-
fuged at 210g for 4 min, and the cells were resuspended in block-
ing buffer [Tris·HCl (25 mm), NaCl (150 mm), CaCl2 (1 mm), MgCl2
(0.5 mm), MnCl2 (1 mm), pH 7.4, BSA (1%)]. Each tube was washed
twice by repeated centrifugation (210g for 4 min), removal of the
supernatant, and addition of fresh blocking buffer. After incubation
of the cell suspension with blocking buffer for 15 min at room
temperature, the cell pellets were washed twice with binding
MS (MALDI) for cRDG16-MT-DOTA (35): m/z: calcd: 16920 [M+H]+;
found: 16955.
Synthesis of DOTA-M-PEG3 (30): A solution of BM(PEG)3 (1,11-bis-
maleimido-triethyleneglycol, Thermo Scientific, 61 mg, 173 mmol) in
a mixture of phosphate buffer (0.1m, pH 6, 200 mL) and MeCN
(300 mL) was added to a solution of thiol-DOTA (72 mg, 156 mmol)
in phosphate buffer (0.1m, pH 7.4, 200 mL) and the pH of the mix-
ture was adjusted to 7.4 by addition of phosphate buffer (0.1m,
2180
www.chembiochem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemBioChem 2010, 11, 2168 – 2181

Multimerization of cRGD Peptides by Click Chemistry
[15] I. Dijkgraaf, J. A. W. Kruijtzer, S. Liu, A. C. Soede, W. J. G. Oyen, F. H. M.
Corstens, R. M. J. Liskamp, O. C. Boerman, Eur. J. Nucl. Med. Mol. Imaging
2007, 34, 267–273.
[16] X. Chen, M. Tohme, R. Park, Y. Hou, J. R. Bading, P. S. Conti, Mol. Imaging
2004, 3, 96–104.
[17] G. Thumshirn, U. Hersel, S. L. Goodman, H. Kessler, Chem. Eur. J. 2003, 9,
2717–2725.
[18] S. Liu, W.-Y. Hsieh, Y. Jiang, Y.-S. Kim, S. G. Sreerama, X. Chen, B. Jia, F.
Wang, Bioconjugate Chem. 2007, 18, 438–446.
[19] Z.-B. Li, K. Chen, X. Chen, Eur. J. Nucl. Med. Mol. Imaging 2008, 35,
1100–1108.
[20] L. Sancey, E. Garanger, S. Foillard, G. Schoehn, A. Hurbin, C. Albiges-
Rizo, D. Boturyn, C. Souchier, A. Grichine, P. Dumy, J.-L. Coll, Mol. Ther.
2009, 17, 837–843.
buffer [Tris·HCl (25 mm), NaCl (150 mm), CaCl2 (1 mm), MgCl2
(0.5 mm), MnCl2 (1 mm), pH 7.4, BSA (0.1%)]. Subsequently, the
U87MG cells were incubated on a shaker in the presence of varied
amounts of competing ligand (0.05 nm–50 mm cRGD multimers
(27–31)) with 370 Bq per tube (0.025 nm) 125I-echistatin in a total
volume of 100 mL binding buffer for one hour. The cell pellets were
rinsed three times with binding buffer, the supernatant was sucked
off, and the tubes containing the cell pellets were measured for ra-
dioactivity with a m-counter (Wallac Wizard). Each data point is an
average of values from triplicate wells and all experiments were in-
dependently repeated at least three times. Ki values were calculat-
ed by use of the software program GraphPad Prism with consider-
ation of the equation for “heterologous competitive binding with
ligand depletion”.
[21] Z.-B. Li, W. Cai, Q. Cao, K. Chen, Z. Wu, L. He, X. Chen, J. Nucl. Med.
2007, 48, 1162–1171.
[22] H.-J. Wester, H. Kessler, J. Nucl. Med. 2005, 46, 1940–1945.
[23] Y. Ye, S. Bloch, B. Xu, S. Achilefu, J. Med. Chem. 2006, 49, 2268–2275.
[24] Z. Liu, G. Niu, J. Shi, S. Liu, F. Wang, S. Liu, X. Chen, Eur. J. Nucl. Med.
Mol. Imaging 2009, 36, 947–957.
Acknowledgement
[25] I. Dijkgraaf, A. J. Rijnders, A. Soede, A. C. Dechesne, G. W. van Esse, A. W.
Brouwer, F. H. M. Corstens, O. C. Boerman, D. T. S. Rijkersa, R. M. J. Lis-
kamp, Org. Biomol. Chem. 2007, 5, 935–944.
[26] M. J. Janssen, W. J. Oyen, I. Dijkgraaf, L. F. Massuger, C. Frielink, D. S. Ed-
wards, M. Rajopadhye, H. Boonstra, F. H. Corstens, H. C. Boerman,
Cancer Res. 2002, 62, 6146–6151.
[27] T. Poethko, M. Schottelius, G. Thumshirn, U. Hersel, M. Herz, G. Henrik-
sen, H. Kessler, M. Schwaiger, H.-J. Wester, J. Nucl. Med. 2004, 45, 892–
902.
The authors would like to thank Dr. Werner Spahl (Department
of Organic Chemistry, Ludwig-Maximilians University Munich,
Germany) for his help with ESI mass spectrometry and Bianca
Weigel for excellent technical assistance. This study was support-
ed by the Friedrich-Baur-Stiftung (74/07), the Deutsche For-
schungsgemeinschaft (DFG MA 4295/1-2), the BMBF (MoBiMed,
01EZ0808) and the Fonds der Chemischen Industrie.
[28] C. Wꢀngler, B. Wꢀngler, M. Eisenhut, U. Haberkorn, W. Mier, Bioorg. Med.
Chem. 2008, 16, 2606–2616.
[29] C. Wꢀngler, G. Moldenhauer, M. Eisenhut, U. Haberkorn, W. Mier, Biocon-
jugate Chem. 2008, 19, 813–820.
Keywords: click chemistry · cRGD · dendrimers · imaging
agents · multimerization
[30] D. A. Wellings, E. Atherton, Methods Enzymol. 1997, 289, 44–67.
[31] B. Wꢀngler, C. Beck, U. Wagner-Utermann, E. Schirrmacher, C. Bauer, F.
Rçsch, R. Schirrmacher, M. Eisenhut, Tetrahedron Lett. 2006, 47, 5985–
5988.
[32] O. Prante, J. Einsiedel, R. Haubner, P. Gmeiner, H.-J. Wester, T. Kuwert, S.
Maschauer, Bioconjugate Chem. 2007, 18, 254–262.
[33] J. Vagner, H. L. Handl, R. J. Gillies, V. J. Hrubya, Bioorg. Med. Chem. Lett.
2004, 14, 211–215.
[34] M. Mammen, S.-K. Choi, G. M. Whitesides, Angew. Chem. 1998, 110,
2908–2953; Angew. Chem. Int. Ed. 1998, 37, 2754–2794.
[35] P. Rajananthanan, G. S. Attard, N. A. Sheikh, W. J. Morrow, Vaccine 1999,
17, 715–730.
[36] J. C. Roberts, M. K. Bhalgat, R. T. Zera, J. Biomed. Mater. Res. 1996, 30,
53–65.
[1] Y. Wang, X. Wang, Y. Zhang, S. Yang, J. Wang, X. Zhang, Q. Zhang, J.
Drug Targeting 2009, 17, 459–467.
[2] H. Zhao, J. C. Wang, Q. S. Sun, C. L. Luo, Q. Zhang, J. Drug Targeting
2009, 17, 10–18.
[3] L. M. Kenny, R. C. Coombes, I. Oulie, K. B. Contractor, M. Miller, T. J.
Spinks, B. McParland, P. S. Cohen, A. M. Hui, C. Palmieri, S. Osman, M.
Glaser, D. Turton, A. Al-Nahhas, E. O. Aboagye, J. Nucl. Med. 2008, 49,
879–886.
[4] O. Schnell, B. Krebs, E. Wagner, A. Romagna, A. J. Beer, S. J. Grau, N.
Thon, C. Goetz, H. A. Kretzschmar, J. C. Tonn, R. H. Goldbrunner, Brain
Pathol. 2008, 18, 378–386.
[5] X. Chen, E. Sievers, Y. Hou, R. Park, M. Tohme, R. Bart, R. Bremner, J. R.
Bading, P. S. Conti, Neoplasia 2005, 7, 271–279.
[6] M. Schottelius, B. Laufer, H. Kessler, H.-J. Wester, Acc. Chem. Res. 2009,
42, 969–980.
[7] R. Haubner, H.-J. Wester, W. A. Weber, C. Mang, S. I. Ziegler, S. L. Good-
man, R. Senekowitsch-Schmidtke, H. Kessler, M. Schwaiger, Cancer Res.
2001, 61, 1781–1785.
[8] T. H. Stollman, T. J. Ruers, W. J. Oyen, O. C. Boerman, Methods 2009, 48,
188–192.
[37] S. Liu, Mol. Pharm. 2006, 3, 472–487.
[38] C. Wꢀngler, R. Schirrmacher, P. Bartenstein, B. Wꢀngler, Curr. Med. Chem.
2010, 17, 1092–1116.
[39] O. Prante, J. Einsiedel, R. Haubner, P. Gmeiner, H.-J. Wester, T. Kuwert, S.
Maschauer, Bioconjugate Chem. 2007, 18, 254–262.
[40] X. Chen, Y. Hou, M. Tohme, R. Park, V. Khankaldyyan, I. Gonzales-Gomez,
J. R. Bading, W. E. Laug, P. S. Conti, J. Nucl. Med. 2004, 45, 1776–1783.
[41] S. Liu, Bioconjugate Chem. 2009, 20, 2199–2213.
[42] C.-B. Yim, O. C. Boerman, M. De Visser, M. De Jong, A. C. Dechesne,
D. T. S. Rijkers, R. M. J. Liskamp, Bioconjugate Chem. 2009, 20, 1323–
1331.
[9] I. Dijkgraaf, A. J. Beer, H.-J. Wester, Front. Biosci. 2009, 14, 887–899.
[10] R. Haubner, C. Decristoforo, Front. Biosci. 2009, 14, 872–886.
[11] E. Garanger, D. Boturyn, P. Dumy, Anticancer Agents Med. Chem. 2007, 7,
552–558.
[12] Y. Wu, X. Zhang, Z. Xiong, Z. Cheng, D. R. Fisher, S. Liu, S. S. Gambhir, X.
Chen, J. Nucl. Med. 2005, 46, 1707–1718.
[43] R. Schirrmacher, Y. Lakhrissi, D. Jolly, J. Goldstein, P. Lucas, E. Schirr-
macher, Tetrahedron Lett. 2008, 49, 4824–4827.
[44] R. O. Hynes, Cell 2002, 110, 673–687.
[13] X. Chen, S. Liu, Y. Hou, M. Tohme, R. Park, J. R. Bading, P. S. Conti, Mol.
Imaging Biol. 2004, 6, 350–359.
[14] Z. Wu, Z.-B. Li, K. Chen, W. Cai, L. He, F. T. Chin, F. Li, X. Chen, J. Nucl.
Med. 2007, 48, 1536–1544.
Received: July 6, 2010
Published online on September 9, 2010
ChemBioChem 2010, 11, 2168 – 2181
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chembiochem.org
2181