β-Cyclodextrin dimers as potential tumor pretargeting agents

Article abstract of DOI:10.1039/b102814f

A β-cyclodextrin dimer binds a di-tert-butylbenzyl-Cucyclen with high affinity, demonstrating potential as a receptor/ligand system for tumor pretargeting with monoclonal antibodies.

Full text of DOI:10.1039/b102814f

b-Cyclodextrin dimers as potential tumor pretargeting agents†‡
W. Barry Edwards,a David E. Reichert,a D. Andre´ d’Avignonb and Michael J. Welch*a
a Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 South
Kingshighway Boulevard, St. Louis, MO 63110, USA. E-mail: welchm@mir.wustl.edu
b Department of Chemistry, Washington University, St. Louis, MO 63110, USA
Received (in Columbia, MO, USA) 27th March 2001, Accepted 1st June 2001
First published as an Advance Article on the web 27th June 2001
A b-cyclodextrin dimer binds a di-tert-butylbenzyl–Cu–
cyclen with high affinity, demonstrating potential as a
receptor/ligand system for tumor pretargeting with mono-
clonal antibodies.
cyclodextrin dimer would be concentrated at the tumor site by
conjugation via a covalent bond to a Mab. Radiometals such
Cu-64 [t1/2 = 12.7 h; Ebmax = 0.653 MeV (17.4%); Ebmax
0.573 MeV (40%)] and Cu-67 [t1/2 = 62 h; Ebmax = 0.392 MeV
(56%)] have been used previously in imaging9 and therapy10,11
of tumors. To demonstrate the feasibility of this new approach,
Cu–BBC (Scheme 1) and a b-cyclodextrin dimer (Scheme 2)
were synthesized and affinities determined in an equilibrium
binding assay. A fluorescent dye, BNS, was synthesized as
previously described.12
=
Monoclonal antibodies (Mabs) have advantages for the delivery
of a radioisotope (or toxin) to tumor sites due to their high
affinity and specificity for their antigen.1,2 However, the results
of radioimmunotherapy of solid tumors have been disappoint-
ing because radiolabeled Mabs targeted to cell surface antigens
impart a high radiation dose to normal organs causing toxicity.3
In part, this is due to the long circulating times for Mabs.
Successful imaging or therapy with radiolabeled Mabs
depends not only on the concentration of the Mab at the targeted
site but on clearance rates from normal tissues relative to that
from the tumor. Imaging studies in humans have shown that
maximum concentrations of Mabs at the tumor site are
attainable within 24 h, but several more days are required before
the concentration of the radiolabeled Mab in the circulation
decreases to levels low enough for successful imaging to take
place.4 Pretargeting techniques seek to maintain high accumula-
tion of the radionuclide at the target site while minimizing non-
target tissue toxicity (see reviews1,2). One current approach at
pretargeting utilizes a Mab conjugated to a receptor with a high
affinity for a ligand bearing the radionuclide. The most often
utilized receptor/ligand system is avidin (or streptavidin/
biotin).5 After the Mab-receptor conjugate has localized at the
tumor site, a low molecular weight radiolabeled ligand, one that
is rapidly excreted via the kidneys, is administered to visualize
the tumor. To be effective, the ligand must be rapidly excreted
from the body to provide the desired high tumor accumulation
with relatively low non-target accumulation.
An alternative receptor for tumor pretargeting could be a b-
cyclodextrin dimer. Breslow et al. studied a series of b-
cyclodextrin dimers and achieved some of the highest affinities
reported to date.6,7 While the best match between a guest
molecule and single cyclodextrin subunit will provide affinities
on the order of 1024 M, cyclodextrin dimers can increase
affinities to 1028 to 10210 M.8 With affinities of this magnitude,
b-cyclodextrin dimers could serve as receptors for tumor
pretargeting. A potential strategy for pretargeting is the
development of a radiometal-binding macrocycle containing
pendant hydrophobic groups with the appropriate geometry for
inclusion into the cavity of the cyclodextrins. Ultimately, the b-
Breslow et al. reported binding affinities for a b-cyclodextrin
dimer with a single linkage in the low nanomolar region for
substrates bearing two tert-butylphenyl substituents.6 There-
fore, a b-cyclodextrin dimer with a single linkage, was
synthesized to serve as a host for a copper chelate-complex
bearing two tert-butylphenyl substituents. The dimer 1 was
prepared from commercially available tosyl-b-cyclodextrin and
2,6-naphthalene dithiol13 (Scheme 2). The cyclodextrin dimer
was purified by HILIC then by silica-gel chromatography.14
1
Analysis by H, 13C, HMQC, COSY NMR and MALDI-MS
confirmed the dimeric structure of the cyclodextrin.
Cyclen was chosen as the chelator because of its high stability
(log Kf = 24.6) for copper(II).15 Based on synthetic method-
ology for selectively protecting the 1 and 7 nitrogens of
tetraazamacrocycles,16 alkylation of the nitrogens with com-
mercially available tert-butylbenzyl bromide appeared to be the
most expedient way of introducing a pair of hydrophobic
substituents onto the chosen chelator (Scheme 1). Molecular
modeling showed that a distance of ca. 16 Å is attainable
between the two tertiary carbons of the tert-butyl groups and
between the midpoints of the b-cyclodextrin subunits of dimer
1 when the faces containing the primary hydroxyl groups are
opposed to each other.17
Equimolar amounts of copper(II) acetate and BBC resulted in
the quantitative formation of Cu–BBC. While Cu–BBC and
BBC co-eluted on a C-18 column, a diphenyl column resolved
a co-injected mixture of the two. Analysis of the reaction
Scheme 1 Reagents and conditions: i, pH 2.5, water, chloroethylformate; ii,
tert-butylbenzyl bromide, DMF; iii, hydrazine, ethylene glycol, KOH; iv,
copper(II) acetate, aq. NH4O2CMe, pH 6.4.
† Electronic supplementary information (ESI) available: experimental
section. See http://www.rsc.org/suppdata/cc/b1/b102814f/
‡ Abbreviations: Mabs, monoclonal antibodies; HILIC, hydrophilic inter-
action chromatography; HMQC, heteronuclear multiple quantam coher-
ence; COSY, H–H correlated spectroscopy; MALDI-MS, matrix assisted
laser desorption-ionization mass spectrometry; Kf, formation constant;
BBC, 1,7-(4-tert-butylphenylmethyl)cyclen; Cu–BBC, copper-1,7-(4-tert-
butylphenylmethyl)cyclen; ES-MS, electrospray mass spectrometry;
HPLC, high pressure liquid chromatography; BNS, 6-(4-tert-butylphenyl-
amino)naphthalene-2-sulfonic acid; KD, equilibrium dissociation constant;
Ki, the concentration of the competing ligand that will bind to half the
binding sites at equilibrium; IC50, concentration of competitive ligand that
inhibits half of the binding of a ligand; Cheng–Prusoff equation, Ki
=
Scheme 2 Reagents and conditions: i, KI, DMF, 85 °C; ii, 2,6-naphth-
alenedithiol, NH4HCO3, DMF, 85 °C.
21
IC50(1 + [ligand]/KD,ligand
)
1312
Chem. Commun., 2001, 1312–1313
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b102814f

Table 1 Affinities of BNS and Cu–BBC for b-cyclodextrin and dimer 1
success of dimer 1 and Cu–BBC as a receptor/ligand system for
tumor pretargeting.
Guest
BNS
BNS
Host
Calculated affinities
95% CI
In conclusion, a b-cyclodextrin dimer with a naphthalene
linker, dimer 1, was synthesized for evaluation as the receptor in
a new approach to tumor pretargeting. Dimer 1 showed a higher
affinity for a fluorescent dye, BNS, than did b-cyclodextrin.
BBC, a derivative of cyclen, was synthesized and Cu–BBC was
prepared. Cu–BBC was strongly bound to dimer 1 (Ki = 18 nM
relative to BNS). With affinities in the low nanomolar region,
Cu-64–BBC and dimer 1 could serve as a receptor/ligand
system for tumor pretargeting.
b-Cyclodextrin
Dimer 1
Cu–BBCa Dimer 1
KD1 = 14 mM
KD2 = 1.6 mM
KD = 388 nM
Ki = 18 nM
10–18 mM
0.6–2.6 mM
112–664 nM
12–28 nM
IC50 = 120 nM
76–189 nM
a BNS was included. Cu–BBC was a competitive ligand with varying
concentration.
This work was supported by the United States Department of
Energy (DE-FG02-87ER60512).
Notes and references
§ Determination of affinity constants: Utilizing GraphPad Prism® software
(GraphPad Software Inc., San Diego, CA), the data were fit by non-linear
regression to: F = (Fmax)[BNS]/(KD + [BNS]) where F is the observed
fluorescence intensity generated from incorporation of BNS into the
hydrophobic interior of b-cyclodextrin (corrected for background fluores-
cence from unincorporated BNS or dimer 1) to determine values for KD.21,22
To determine the IC50 value, the data were fit by non-linear regression to:
Fobs = T + (T 2 B)/1 + 10 exp(log[Cu–BBC] 2 log IC50), where T and B
are the top and bottom plateaus of the fitted curve and Fobs is the observed
fluorescence.21,22 To calculate Ki for Cu–BBC, the value obtained for the
KD of BNS and Cheng–Prussof equation were utilized.20–22 The data were
fit to both one or two binding sites in all experiments and the results for two
binding sites are reported when P < 0.05. The error is represented in terms
of 95% confidence interval (95% CI).
Fig. 1 Heterologous competitive binding assay between Cu–BBC and BNS
for dimer 1.
mixture on the diphenyl column showed that BBC accounted
for < 1% of the total peak area observed (l = 215 nm, Cu–BBC
and BBCs strongest absorbance). Analysis by ES-MS con-
firmed that the component observed by HPLC was Cu–BBC.
Affinity constants for b-cyclodextrin dimers and their
substrates are often ascertained by fluorescence spectroscopy§
since saturation binding will be attained at concentrations too
low to utilize NMR and UV spectroscopy. Therefore a
fluorescent dye, BNS, was synthesized. BNS fluoresces weakly
in water, but its inclusion into a hydrophobic environment, such
as that of the cavity of b-cyclodextrin, enhances its fluores-
cence. This dye has been used extensively in the character-
ization of other b-cyclodextrin dimers (see review8). Curve
fitting with commercially available software will then yield the
affinity of dye. After the affinity of dye for its host has been
quantified, a heterologous competitive binding assay between
the dye and another guest can provide the affinity of the guest.
These approaches were taken to evaluate the affinity of BNS
and Cu–BBC for dimer 1 (Table 1).
1 D. A. Goodwin and C. F. Meares, Cancer Biother. Radiopharm., 1999,
14, 145.
2 J. Barbet, F. Kraeber-Bodere, J. Vuillez; E. Gautherot, E. Rouvier and
J. Chatal, Cancer Biother. Radiopharm., 1999, 14, 153.
3 R. B. Wilder, G. L. DeNardo and S. J. DeNardo, J. Clin. Oncol., 1996,
14, 1383.
4 D. A. Goodwin, J. Nucl. Med., 1995, 36, 876.
5 D. B. Axworthy, P. L. Beaumier, B. J. Bottino, S. Goshorn, R. W.
Mallet, K. M. Stone, F. M. Su, L. F. Theodore, E. K. Yau and J. M.
Reno, Tumor Targeting, 1996, 2, 156.
6 R. Breslow, N. Greenspoon, T. Guo and R. Zarzycki, J. Am. Chem. Soc.,
1989, 111, 8296.
7 R. Breslow and S. Chung, J. Am. Chem. Soc., 1990, 112, 9659.
8 R. Breslow, S. Halfon and B. Zhang, Tetrahedron, 1995, 51, 377.
9 C. J. Anderson, L. A. Jones, L. A. Bass, E. L. Sherman, D. W.
McCarthy, P. D. Cutler, M. V. Lanahan, M. E. Cristel, J. S. Lewis and
S. W. Schwarz, J. Nucl. Med., 1998, 39, 194.
10 C. J. Anderson, F. Dehdashti, P. D. Cutler, S. W. Schwarz, R. Laforest,
L. R. Bass, J. S. Lewis and D. W. McCarthy, J. Nucl. Med., 2001, 42,
213.
11 G. R. Mirick, R. T. O'Donnell, S. J. DeNardo, S. Shen, C. F. Meares and
G. L. DeNardo, Nucl. Med. Biol., 1999, 26, 841.
12 S. Chung, Ph.D. Thesis, Columbia University, 1991.
13 V. N. Lisitsyn, L. S. Kukalenko and A. M. Tsatkis, Zh. Org. Khim.,
1985, 21, 161.
14 A. J. Alpert, J. Chromatogr., 1990, 499, 177.
15 R. M. Smith and A. E. Martell, Crictical Stability Constants, Plenum,
New York, 1989.
16 Z. Kovacs and A. D. Sherry, J. Chem. Soc., Chem. Commun., 1995,
185.
17 CambridgeSoft Corporation, CS Chem3D Std, 4.0, Cambridge MA,
1997.
18 R. Breslow and B. Zhang, J. Am. Chem. Soc., 1996, 118, 8495.
19 R. C. Petter, C. T. Sikorski and D. H. Waldeck, J. Am. Chem. Soc., 1991,
113, 2325.
20 Y. Cheng and W. H. Prusoff, Biochem. Pharmacol., 1973, 22, 3099.
21 L. E. Limbird, Cell Surface Receptors: A short course on theory and
methods, Kluwer Academic Publishers, Hingham, 1986.
22 H. J. Molutsky, Analyzing Data with GraphPad Prism, GraphPad
Software Inc., San Diego, 1999.
BNS exhibited an enhancement of fluorescence in aqueous
solution when interacting with the b-cyclodextrin subunits. The
interaction was characteristic of a receptor–ligand (host–guest)
complex because of its saturable and reversible nature. The
observed affinity of BNS agrees with the previously determined
value of KD = 20 mM.12 Previous affinities of BNS observed for
b-cyclodextrin dimers with a single linkage have ranged from
ca. 3 mM to 100 nM.18,19 The increase in affinity of BNS is
consistent with previously observed increases when two b-
cyclodextrin subunits are linked face to face.
Because the fluorescence of Cu–BBC is unaffected by the
presence of dimer 1, its affinity must be determined by
competition with BNS. The log-dose dependent displacement
of BNS from dimer 1 with Cu–BBC indicates that both Cu–
BBC and BNS have formed a complex with dimer 1 (Fig. 1). By
substituting the value of the IC50 of Cu–BBC and the KD of
BNS into the Cheng–Prussof equation, the Ki of Cu–BBC is
obtained.20 This remarkably strong affinity bodes well for the
Chem. Commun., 2001, 1312–1313
1313