Efficient synthesis and evaluation of bimodal ligand NETA

Article abstract of DOI:10.1016/j.bmcl.2008.03.084

The efficient and short synthetic route to the structurally novel bimodal ligand NETA for antibody-targeted radiation therapy (radioimmunotherapy, RIT) of cancer was developed. The structure of NETA was determined by X-ray crystallography. The arsenazo-ba

Full text of DOI:10.1016/j.bmcl.2008.03.084

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3436–3439
Efficient synthesis and evaluation of bimodal ligand NETA
Hyun-Soon Chong,* Hyun A. Song, Noah Birch, Thien Le, Sooyoun Lim and Xiang Ma
Chemistry Division, Biological, Chemical, and Physical Sciences Department, Illinois Institute of Technology,
3101 S. Dearborn Street, LS 182, Chicago, IL 60616, USA
Received 27 January 2008; revised 22 March 2008; accepted 31 March 2008
Available online 8 April 2008
Abstract—The efficient and short synthetic route to the structurally novel bimodal ligand NETA for antibody-targeted radiation
therapy (radioimmunotherapy, RIT) of cancer was developed. The structure of NETA was determined by X-ray crystallography.
The arsenazo-based UV spectroscopic complexation kinetics data suggest that NETA is a promising chelator for use in RIT appli-
cations of 212Bi, 213Bi, and 177Lu.
Ó 2008 Published by Elsevier Ltd.
Synthetic macrocyclic and acyclic ligands that possess
amino and carboxylate groups as metal-binding moieties
have been employed for biomedical and radiopharmaceu-
tical applications such as magnetic resonance (MR)1 and
positron emission tomography (PET)2 imaging and anti-
body-targeted radiation therapy (radioimmunotherapy,
RIT)3 of cancer. Research efforts have been directed
towards development of the adequate metal binding mac-
rocyclic and acyclic ligand, a critical component for
enhancing the efficacy of the therapeutic and diagnostic
applications.
kinetics of DOTA remains a limitation in RIT applica-
tions, particularly involving relatively short-lived radio-
active metals such as 212Bi (t1/2 = 60.6 m) and 213Bi (t1/2
=
45 m).9 Acyclic DTPA rapidly binds to the metals butpro-
duces relatively unstable complexes.10 NETA containing
both macrocyclic and acyclic moieties is proposed to bind
the metal ion more rapidly than DOTA8,11 via dynamic
binding while binding the metal stronger than DTPA10
via three-dimensional binding. NETA complexed with
various metals including lanthanides was very stable in
human serum or mice.4–6 NETA was found to bind
Y(III) within seconds, while macrocyclic DOTA displays
very slow complexation kinetics (>1 h) as determined by a
spectroscopic competing reaction with arsenazo III
(AAIII).4 Great potential of NETA for use in RIT
prompts us to further investigate the ligand in a unique
structural class. Herein, we report the efficient and short
synthetic route to NETA and its evaluation as RIT chela-
tors for 177Lu, 212Bi, and 213Bi therapeutic metals. The
structure of NETA was determined by X-ray crystallo-
graphic analysis.
RIT is a potent cancer therapeutic modality that employs
a tumor-targeting mAb for selective delivery of a cyto-
toxic radioisotope while minimizing exposure of healthy
cells.3 A variety of metallic radionuclides including
177Lu, 212Bi, and 213Bi have been proven effective for
RIT.3 We have recently developed the structurally novel
bimodal ligand NETA for use in RIT.4–7 NETA (2-(4,7-
biscarboxymethyl[1,4,7]triazacyclonona-1-yl-ethyl)car-
bonyl-methylamino]acetic acid) is a hybridized structure
of DOTA (1,4,7,10-tetraazacyclododecane-N,N0,N00,N000-
tetraacetic acid) and DTPA (diethylenetriaminepentaace-
tic acid), the most frequently explored polyaminocarb-
oxylates. Macrocyclic DOTA forms a stable complex
with various therapeutic metals and is a standard refer-
ence ligand for comparison.8 However, slow formation
It was necessary to demonstrate the hypothesized bimo-
dal binding of NETA with various metals by X-ray crys-
tallographic study and measurements of thermodynamic
constants that require significant amount of the chela-
tor. The original synthesis of NETA as previously re-
ported involves eight reaction steps including
a
Keywords: Radioimmunotherapy; NETA; Metal binding ligand; 177Lu;
convenient macrocyclization of the readily available
starting materials, ethanol amine and an N-tosyl pro-
tected ditosylate.4 We initially attempted preparation
of NETA in large quantities by modifying the original
212Bi; 213Bi.
*
Corresponding author. Tel.: +1 312 567 3235; fax: +1 312 567
3494; e-mail: chong@iit.edu
0960-894X/$ - see front matter Ó 2008 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2008.03.084

H.-S. Chong et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3436–3439
3437
Ts
Ts
H
H
i)
N
N
N
OH
H2N
N
Conc. H2SO4
115 oC, 72 h
TsO
N
N
OTs
N
N
ii) SOCl2
Ts
Ts
iii) H2, 10% Pd/C
iv) NaN3, DMSO
NH2
NH2
2
1
i) HCl
ii) DIPEA,
BrCH2CO2t-Bu
O
O
O
O
N
C
N
C
N
C
N
C
OH
O
t-Bu
HO
t
-BuO
HCl (g)
N
N
O
O
1,4-Dioxane
14 h
N
C
C
N
O
OH
O
t
-Bu
O
NETA
C
C
3
OH
Ot
-Bu
Scheme 1. Modification to original synthesis4 of NETA.
synthetic route (Scheme 1). We found that the original
synthetic route is quite lengthy and not efficient for
large-scale preparation of NETA. The detosylation
and alkylation reaction steps to produce 2 and 3, respec-
tively, significantly lower an overall yield (<20%) of the
synthetic route (Scheme 1). We observed that prepara-
tion of precursor molecule 1 in large quantities (>30 g)
was practically and efficiently carried out in excellent
yield (>86%). All of the compounds except the first mac-
rocyclization step were isolated after simple work-up or
recrystallization and did not require column chromato-
graphic purification. However, scale-up deprotection
of compound 1 was problematic providing 2 in a low
yield, despite working very well in small scale (<1 g)
and nearly 90% yield. Furthermore, detosylation of 1
in large quantities required tedious work-up involving
stepwise neutralization and extraction generating signif-
icant amount of toxic waste. We found that free amine 2
oxidizes in less than 12 h at both room temperature and
ꢀ20 °C as monitored by NMR, while the oxidation can
be prevented by immediately treating 2 with HCl to af-
ford an acidic salt (Scheme 1). The alkylation reaction of
2 as an acidic salt in large quantity produced large
amount of polyalkylated byproducts along with the de-
sired compound 3 in a very poor yield (11%). Optimiza-
tion of the alkylation reaction to exclusively produce 3,
which does not contain a UV chromophore, was quite
challenging. The polar-tailing compound 3 and byprod-
ucts possess concentration-dependent Rf values. There-
fore, TLC analysis of fractions eluted from flash
chromatography on normal phase silica gel is not very
informative in distinguishing the targeted product from
byproducts that possess very similar Rf value(s). This
also makes the desired 3 using flash silica gel (or alu-
mina) column chromatography extremely complicated.
This difficulty in producing NETA analogue 3 in large
quantities prompted us to develop more practical and
short synthetic route.
Since Prep-HPLC or other purification methods are not
helpful for purification of NETA without a strong UV
chromophore, we have attempted different reaction
routes in efforts to develop a synthetic route to NETA
avoiding harsh detosylation and complicated alkylation
reaction steps. An efficient and short method to prepare
NETA is shown in Scheme 2. The key step is the cou-
pling reaction of bisubstituted macrocyclic TACN
(1,4,7-Triazacyclononane) 5 with acyclic N-dialkylated
bromide 6 in an equimolar ratio. The starting material,
bisubstituted TACN (5),12 was efficiently prepared by
O
O
Br
N
CO2Bn
O
N
C
N
C
O
H
H
O
t
-Bu
O
t
-BuO
N
C
N
N
C
N
(6)
N
BrCH2CO2t-Bu
Ot-Bu
CO2Bn
CH3CN, 72 h, reflux
t
-BuO
N
N
H
4
C
N
CHCl3
rt, 24 h
OBn
O
H
5
7
C
OBn
6M HCl(aq)
Reflux, 5 h
O
O
N
C
N
C
O
t
-Bu
O
t
-BuO
N
O
O
N
C
N
4M HCl
OH 1,4-dioxane
rt, 12 h
C
C
O
N
OH
HO
N
O
8
C
C
O
N
OH
OH
C
NETA
OH
Scheme 2. Short and efficient synthesis of NETA.

3438
H.-S. Chong et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3436–3439
the reaction of TACN 4 with tert-butylbromoacetate. It
was necessary to develop a short and reproducible syn-
thetic route to 5 since the known procedures either in-
volve lengthy and poor total synthetic yield or require
column chromatographic purification of the polar tail-
ing macrocycle.12,13 In the present study, 5 was isolated
from a reaction mixture by pH controlled stepwise neu-
tralization without column chromatography. The best
yield of 5 was obtained from the reaction of TACN
(1 equiv) and tert-butylbromoacetate (2.2 equiv) without
the addition of base. In the presence of base (triethyl-
amine or diisopropylethylamine), the mixture of the de-
sired bisubstituted TACN and trisubstituted and/or
monosubstituted TACN was obtained (supplementary
data, Table 1). The coupling reaction of 5 with 616 pro-
vided the desired product 7 (Scheme 2). Since each of the
precursor molecules was alkylated prior to the coupling
reaction, the formation of polyalkylation byproducts
was minimized. Isolation of tailing polar macrocycle 7,
which can be monitored by HPLC and TLC analysis
due to the presence of the UV chromophore, was
achieved by flash column chromatography eluting with
5–6% CH3OH/CH2Cl2. Most importantly, with the
development of TLC and analytical and preparative
HPLC methods based on benzyl UV chromophore,
the reaction condition could be readily modified if an
improvement of the reaction yield is required. Com-
pound 7 was efficiently prepared in a good and highly
reproducible yield (>60%). Both benzyl and tert -butyl
protection groups in 7 were removed by refluxing 7 in
6 M HCl(aq). The benzyl groups in 7 were also selec-
tively removed by hydrogenation and provided tert -bu-
tyl group containing chelate 8. The partially deprotected
compound 8 may be a useful backbone molecule which
can be selectively reacted with a variety of amino-con-
taining peptides and antibodies to generate a viable che-
lator for biomedical applications. NETA was efficiently
prepared in two steps and large quantities and good
overall yield (61%).
Figure 1. XRD structure of NETA.
evaluate NETA for the complexation kinetics with
Lu(III) and Bi(III). The complexation kinetics of NETA
with Lu(III) and Bi(III) was determined using a well-
known spectroscopic competing reaction with AAIII
according to a modification of a previously reported
procedure.7 AAIII is known to form a weak complex
with many different metals, which produce a UV–Vis
absorbance maximum at ꢁ652 nm.14 However, uncom-
plexed AAIII absorbs weakly at this wavelength. When
introduced to a solution containing the AAIII–metal
complex, a chelate can compete with AAIII for the me-
tal. The idea is that if the chelate is more capable of
binding the metal than AAIII, the metal will dissociate
from the AAIII complex and form a complex with the
chelate leading to the decrease in the absorbance at
the wavelength. The absorbance (A652) for the Bi(III)–
AAIII or Lu(III)–AAIII complex was measured in the
absence and in the presence of the ligands over 1 h at
rt. The complexation kinetics of NETA with Lu(III)
and Bi(III) was determined at pH 4.5 or 4.0, respec-
tively, as hydrolysis occurs at a higher pH.7,15 The com-
plexation result of the new ligands studied herein was
compared to that of DOTA and DTPA, which are
known to form a complex with metallic radionuclides
with extremely slow and fast kinetics, respectively.7 A
plot of absorbance at 652 nm versus time is shown in
Figures 2 and 3. The data in Figure 2 indicate that
NETA displayed fast complexation kinetics with
Lu(III), while DOTA is sluggish in binding Lu(III). As
expected, DTPA was very fast in binding to Lu(III).
NETA was also shown to instantly bind to Bi(III),
and its complexation with Bi(III) was essentially com-
plete very shortly after the starting point of the measure-
ment (Fig. 3). DOTA consistently exhibited slow
complexation kinetics with Bi(III), and DTPA displayed
very fast complexation kinetics with Bi(III). The kinetics
data indicate that DOTA displayed sluggish complexa-
tion with Lu(III) and Bi(III), and the new ligand NETA
formed a complex with the metals at a much greater rate
than DOTA and a rate comparable to DTPA.
The structure of NETA as an acidic salt (HCl) was suc-
cessfully determined via X-ray crystallography (Fig. 1).
Crystals of NETA for X-ray analysis were obtained by
slow evaporation of H2O/EtOH/Et O in a 1:1:2 ratio.
2
Two of the carboxylate groups on the macrocyclic back-
bone are protonated (CO2H), while the other two car-
boxylate groups on the acyclic moiety exist as anion
(CO2ꢀ). The C–O bond distances of the protonated
and anionic carboxylates indicate that O5/O6 and
O7/O8 are in a resonance form (CO2ꢀ), while O1/O2 and
O3/O4 have one bond longer than the other (CO2H). It
is interesting to note that hydrogen bonding exists be-
tween the protonated nitrogen atoms N1, N2, and N4 in
both macrocyclic and acyclic moiety and the chlorine
atom Cl1; this may be a good entry point to demonstrate
the proposed bimodal binding hypothesis. Cl1 is bound to
the protonated nitrogens N1, N2, and N4 with respective
˚
bond distances of 3.22, 3.14, and 3.04 A.
We previously reported that NETA labeled with the
cancer therapeutic metal, 177Lu or 205/6Bi, was extremely
stable in human serum without leaking the radionu-
clide.5 With the promising data, we wanted to further

H.-S. Chong et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3436–3439
3439
0.06
0.05
0.04
0.03
0.02
0.01
0
Acknowledgments
We thank Benjamin E. Kucera and Victor G. Young, Jr.
(the X-ray Crystallographic Laboratory, University of
Minesota) for determination of the structure of NETA
via X-ray crystallography. We acknowledge financial
support from the National Institutes of Health
(K22CA102637 and R01CA112503-01A2).
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2008.03.084.
0
600
1200
1800
2400
3000
3600
Reaction Time (s)
Figure 2. Plot of absorbance (652 nm) versus time of Lu(III)–AAIII
(d), NETA (Ç), DOTA (–), and DTPA (D) at pH 4.5 (0.15 M
NH4OAc) and 25 °C.
References and notes
1. Caravan, P.; Ellison, J. J.; McMurry, T. J.; Lauffer, R. B.
Chem. Rev. 1999, 99, 2293.
2. Wadas, T. J.; Wong, E. H.; Weisman, G. R.; Anderson, C.
J. Curr. Pharm. Des. 2007, 13, 3.
3. Milenic, D. E.; Brady, E. D.; Brechbiel, M. W. Nat. Rev.
Drug Disc. 2004, 3, 488.
4. Chong, H. S.; Garmestani, K.; Millenic, D. E.; Brechbiel,
M. W. J. Med. Chem. 2002, 45, 3458.
5. Chong, H. S.; Milenic, D. E.; Garmestani, K.; Brady, E.
D.; Arora, H.; Pfiester, C.; Brechbiel, M. W. Nucl. Med.
Biol. 2006, 33, 459.
6. Chong, H. S.; Garmestani, K.; Bryant, L. H., Jr.; Milenic,
D. E.; Overstreet, T.; Birch, N.; Le, T.; Brady, E. D.;
Brechbiel, M. W. J. Med. Chem. 2006, 49, 2055.
7. Chong, H. S.; Mhaske, S.; Lin, M.; Bhuniya, S.; Song, H.
A.; Brechbiel, M. W.; Sun, X. Bioorg. Med. Chem. Lett.
2007, 17, 6107.
8. Kodama, M.; Koike, T.; Mahatma, A. B.; Kimura, E.
Inorg. Chem. 1991, 30, 1270.
9. Hassfjell, S.; Brechbiel, M. W. Chem. Rev. 2001, 101,
2019.
0.06
0.05
0.04
0.03
0.02
0.01
0
0
600
1200
1800
2400
3000
3600
Reaction Time (s)
Figure 3. Plot of absorbance (652 nm) versus time of Bi(III)–AAIII
(d), NETA (Ç), DOTA (–), and DTPA (D) at pH 4.0 (0.15 M
NH4OAc) and 25 °C.
10. Harrison, A.; Walker, C. A.; Parker, D.; Jankowski, K. J.;
Cox, J. P. Nucl. Med. Biol. 1991, 18, 469.
11. Camera, L.; Kinuya, S.; Garmestani, K.; Wu, C.; Brechb-
iel, M. W.; Pai, L. H.; McMurry, T. J.; Gansow, O. A.;
Pastan, I.; Paik, C. H.; Carrasquillo, J. A. J. Nucl. Med.
1994, 35, 882.
12. Huskens, J.; Sherry, A. D. J. Am. Chem. Soc. 1996, 118,
4396.
13. Li, C.; Wong, W-K. W.-K. J. Org. Chem. 2003, 68, 2956.
14. Savvin, S. B. Talanta 1961, 673.
In summary, the efficient and short synthetic route to
the bimodal ligand NETA and partially protected
NETA analogue 8, which allows for their preparation
in large quantities, is described. The structure of NETA
was determined via X-ray crystallographic analysis. The
efficient synthetic route to NETA can be applied to
preparation of various polyazamacrocyclic ligands for
biomedical and radiopharmaceutical applications. The
AA(III)-based spectroscopic complexation kinetics data
suggest that NETA is a viable chelator for use in RIT
applications of 177Lu, 212Bi, and 213Bi.
15. Pippin, C. G.; Parker, T. J.; McMurry, T. J.; Brechbiel, M.
W. Bioconjugate Chem. 1992, 3, 342.
16. Williams, M. A.; Rapport, H. J. Org. Chem. 1993, 58,
1151.