Using molecular symmetry to form new drugs: Hydroxymethyl-substituted 3,9-diazatetraasteranes as the first class of symmetric MDR modulators

Article abstract of DOI:10.1002/1521-3773(20021004)41:19<3623::AID-ANIE3623>3.0.CO;2-V

More is not always better: Until now it was considered that modulators of P-glycoprotein-associated multidrug resistance in cancer treatment showed greater inhibitory activity with an increasing number of moieties that could participate in hydrogen bonds.

Full text of DOI:10.1002/1521-3773(20021004)41:19<3623::AID-ANIE3623>3.0.CO;2-V

COMMUNICATIONS
desired inhibitory effect requires concentrations in ranges
where the native pharmacological action dominates or serious
side effects appear.[6]
Using Molecular Symmetry to Form New
Drugs: Hydroxymethyl-Substituted 3,9-
Diazatetraasteranes as the First Class of
Symmetric MDR Modulators**
Here we present the only known class of compounds that
are themselves symmetric modulators. Preliminary in vitro
assays with MDR-resistant cancer cells show that concen-
trations below toxic levels are sufficient for a reversal of the
MDR. These novel modulators belong to the group of
hydroxymethyl-substituted 3,9-diazatetraasteranes and, in
contrast to all other drugs known so far, display a symmetrical
form. Owing to the rigid cage structure, the positioning of the
functional groups for interaction with amino acids in the
potential binding region is exactly defined. Since the com-
pounds act as strong inhibitors, their rigid symmetrical
structure could provide evidence for a possible, as yet
unknown, symmetrical structure of the potential protein
binding region.
The synthesis of the symmetric inhibitors 6a c started from
the N-substituted 4-aryl-1,4-dihydropyridines 1a c. Com-
pounds 1a and 1b were obtained by the reaction of nicotinic
acid ethyl ester with the methyl and phenyl esters of
chloroformic acid followed by regioselective arylation at the
4-position catalyzed by copper(i) iodide (Scheme 1).[7] The N-
Boc derivative 1c was prepared from 1b with potassium tert-
butoxide.
Andreas Hilgeroth,* JÛsef Molnµr, and
Eric De Clercq
Integral proteins that function as ion channels, neural
receptors, or transport proteins pose problems as target
structures for the modern development of drugs, since the
region that is recognized or bound by functionally modified
pharmaceuticals is often not known or only insufficiently
characterized.[1] Nonetheless, a design on the basis of struc-
ture activity relationships requires exact knowledge of the
spatial positioning of the amino acids in the bonding region.[1]
Since the crystal structure analysis of integral membrane
proteins is extraordinarily difficult, theories on how these
proteins function and how they are affected by drugs are often
based on models.[1,2] Examples of target structures that have
so far only been described with models and that are currently
of great interest for the battle against multidrug resistance
(MDR) in the treatment of cancer include the transmembrane
molecular pump P-glycoprotein (P-GP) and the multidrug
resistance associated protein (MRP).[3,4] Owing to their
position in the fluid mosaic membrane of the cell, integral
proteins are difficult to crystallize. However, suitable crystals
are needed to elucidate the structure of the bonding region for
drugs that could modulate or reverse the multidrug resistance
by blocking the molecular pump. As a result, even today the
development of MDR modulators based on the target
structure is not possible. Nevertheless, numerous drugs from
various classes such as verapamil (VRP) or cyclosporin A
have been characterized as MDR modulators in in vitro
assays. These drugs inhibit the integral protein pumps and,
when given together with substrates for the P-GP and MRP,
reduce the expulsion of the latter from the tumor cells.[5]
However, such drugs display a pharmacological action of their
own, which is also shown by related compounds that have
been further developed and structurally modified, and as a
result cannot be used as MDR modulators in practice. The
[*] Priv.-Doz. Dr. A. Hilgeroth
Martin-Luther-Universit‰t
Scheme 1. Synthesis of the monomeric 4-aryl-1,4-dihydropyridines. a) CuI
(5 mol%), ROCOCl (1 equiv), 10 min, À88C, THF; b) PhMgCl (1 equiv),
30 min, RT, 1a (75%), 1b (83%); c) tBuOK (2 equiv), 30 min, À188C,
THF, 1c (69%).
Institut f¸r Pharmazeutische Chemie
Wolfgang-Langenbeck-Strasse 4
06120 Halle (Germany)
Fax : (þ 49)345-55-27026
E-mail: hilgeroth@pharmazie.uni-halle.de
Interestingly, irradiation of solutions of 1a c with an Ultra-
Vitalux lamp at l > 270 nm resulted in excitation of the 1,4-
dihydropyridine chromophore between 296 and 305 nm and
formation of the 3,9-diazatetraasteranes 3a c as head-to-tail,
centrosymmetric dimerization products. In addition, the
corresponding novel anti dimers 4a c were also formed
(Scheme 2).
Prof. Dr. J. Molnµr
University of Szeged
Department of Medical Microbioloy
Dom ter 10, 6720 Szeged (Hungary)
Prof. Dr. E. De Clercq
Katholieke Universiteit Leuven
Rega Institute for Medical Research
3000 Leuven (Belgium)
Neither form of the dimerization products has yet been
described as the direct reaction product for 5-unsubstituted 4-
aryl-1,4-dihydropyridines. With the corresponding N-alkyl-
substituted monomeric, 1,4-dihydropyridine derivative, tetra-
[**] This work was supported by the Deutsche Pharmazeutische Gesell-
schaft and the European Union as part of the COST-B16 action.
Supporting information for this article is available on the WWW under
http://www.angewandte.com or from the author.
Angew. Chem. Int. Ed. 2002, 41, No. 19
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4119-3623 $ 20.00+.50/0
3623

COMMUNICATIONS
Table 1. Effects of the hydroxymethyl-3,9-diazatetraasteranes 6a d on the
MDR of tumor cells (mouse T-lymphoma).[a]
Activity ratio R[b,c]
Compound
3 mm
30 mm
6a
6b
6c
6d
1.11
28.10
6.81
18.62
30.87[d]
7.42[d]
5.40
0.97
[a] VRP control with 11 mm: 7.27. According to ref. [8] 11 mm is the relevant
maximum concentration for MDR modulation. The compounds are active
(A) when theðrMaDtRio; triesatRed=>MD1R.1; ,unatrneadtedvceornytroalÞctive (AA) for R values above 10.[8]
[b] R ¼ ðparent cell line; treated=parent cell line; untreated controlÞ. [c] Average value from two
determinations. [d] Saturation concentration for P-GP inhibition.
In the case of the hydroxymethyl cage compounds 6a c, the
N-methoxycarbonyl-substituted compound 6a showed an
activity similar to that of VRP; a linear relationship between
the concentration and the inhibitory activity was determined.
Structural characteristics required for sufficient activity of
MDR modulators include the presence of two lipophilic
moieties and a basic nitrogen atom, which was replaced for
the first time by Ecker et al. by an ester group that functioned
as a hydrogen bonding acceptor.[9,10] In our compounds the
nitrogen atom, a hydrophilic structural element, is replaced by
the hydroxymethyl group. The theory that sufficient lipo-
philicity is requirement for good MDR modulators was
supported by the fact that compounds 6b and 6c, which
contain more lipophilic groups at the N-substituents, showed
increased activity at low concentrations.[11] It is interesting
that an increase in the total number of potential hydrogen
bond donor and acceptor functionalities from four (two
hydroxymethyl and two carbamide ester groups) in 6b to six
(four hydroxymethyl and two carbamide ester groups) in
6d[12] is accompanied by complete loss of activity at the same
low concentration (3 mm).
Until now the presence of a large number of moieties that
can participate in hydrogen bonding, for interaction with the
amide groups that are in the proposed bonding site, was
considered to be important for an optimal bonding affinity to
the P-GP.[11,13,14] Initial studies on a series of 3,9-dialkyl-3,9-
diazatetraasteranes[15] provide inhibitory activity ratios of
30.19 at 3 mm and 24.83 at 30 mm for a 3,9-dibenzyl derivative
with four hydroxymethyl groups. This corresponds to the
activity of compound 6b, which contains two hydroxymethyl
and two carbamide ester groups and therefore also presents a
total of four hydrogen bonding sites. A 3,9-dimethyl-3,9-
diazatetraasterane with four hydroxymethyl groups provides
R values of 0.88 (3 mm) and 4.7 (30 mm) and thus does not
display any noteworthy P-GP inhibitory activity until a
concentration of 30 mm. This underlines the importance of a
bulky lipophilic N-substituent for inhibitory activity, as shown
by the comparison of the 3,9-dimethoxycarbonyl derivative
6a with the 3,9-diphenoxycarbonyl derivative 6b. A decrease
in the number of hydroxymethyl groups from four to two
leads to a decrease in the inhibitory activity ratio R to 16.42
(3 mm) and 10.69 for the 3,9-dibenzyl derivative (30 mm) and to
0.67 (3 mm) and 2.7 (30 mm) for the 3,9-dimethyl derivative. In
the case of four hydroxymethyl groups, similar results are
obtained as for the combination of two hydroxymethyl and
Scheme 2. Synthesis of the target structures 6a c starting from the
monomeric 4-aryl-1,4-dihydropyridines. a) hn, l > 270 nm, 4 weeks, 278C,
THF, 3a (44%), 4a (33%), 3b (46%), 4b (32%), 3c (41%), 4c (38%);
b) LiAlH4 (1 equiv), 2 h, À88C, THF, 6a (78%), 6b (75%), 6c (65%).
kishomocubane 5 was formed.[7] We are currently investigat-
ing the causes of the unexpected outcome of this reaction. In
the case of the anti dimers we propose a stabilizing conjugated
effect of the N-oxycarbonyl group on a hypothetical inter-
mediate biradical 2, which then undergoes bond rotation to
form the anti dimer. Subsequent selective reduction of the
ester groups at the substituted cyclobutane ring of the dimers
is accomplished with lithium aluminum hydride at low
temperature and provides the target compounds 6a c.
The in vitro assay to determine the inhibition of the activity
of the molecular pump made use of fluorescence measure-
ments of untreated cells and cells that had been treated with
P-GP modulators. The ratio of the uptake of a P-GP
transporter substrate that shows specific binding and fluores-
cence, for example, rhodamine 123, by tumor cells and cells
that have become resistant through MDR gene transfer was
calculated as the ratio of P-GP inhibitory activity R (Ta-
belle 1).[8] This direct comparison of the fluorescence data
ensures that the observed activities observed for the specific
P-GP inhibitors is based on the MDR modulators.
3624
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4119-3624 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 19

COMMUNICATIONS
[12] A. Hilgeroth, A. Billich, Arch. Pharm. Pharm. Med. Chem. 1999, 332,
380 384.
[13] A. Seelig, L. Blatter, F. Wohnsland, Int. J. Clin. Pharmacol. Ther. 2000,
38, 111 121.
[14] Attempts to determine the activities of the precursors 3a c, which
contain ester groups, failed due to insufficient solubility in the assay
system.
[15] A. Hilgeroth, M. Wiese, A. Billich, J. Med. Chem. 1999, 42, 4729
4732.
two carbamide ester groups. In contrast, the presence of four
hydroxymethyl and two carbamide ester groups, and there-
fore a total of six hydrogen bond donor and acceptor
functionalities, is considered to be less favorable. Our results
suggest a limit on the number of hydrogen bond donor and
acceptor groups, at least for this first series of symmetric
modulators, and therefore provide an important contribution
to the discussion on the role of hydrogen bonding with respect
to activity. A comparison of the concentrations required for
sufficient inhibitory activity (< 3 mm) and the CC50 values for
the cyctotoxicity of 6b and 6c (> 93 and 125 mm, respectively),
obtained from MTT assays[16] for the determination of lactate
dehydrogenase (LDH) activity in metabolically active normal
cells (MT-4), makes clear that the novel MDR modulators
evoke convincing inhibitory activities at concentrations below
those causing cytotoxic effects.
In conclusion, hydroxymethyl-substituted 3,9-diazatetraas-
teranes represent an independent class of novel MDR
modulators that show activity at concentrations below cyto-
toxic levels. A highly interesting characteristic of these
compounds is the rigidly symmetric form, which, owing to
the extraordinary effect on the target structure, could be a
reflection of the composition of the molecular pump or the
potential bonding region, whose structure is as yet unknown.
The goal of further work is to examine this ™symmetry
hypothesis∫ through the synthesis and investigation of unsym-
metrical compounds. In order to estimate the clinical potential
experiments are planned on additional resistant cell lines,
whose MDR is based on the expression of molecular pumps
other than the P-GP.
[16] Assay with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-
mide (MTT) as substrate: S. N. Pandeya, D. Sriram, E. De Clercq, C.
Pannecouque, M. Witvrouw, Indian J. Pharm. Sci. 1998, 60, 207 212.
Fluorescence Spectroscopic Quantification of
the Release of Cyclic Nucleotides from
Photocleavable
[Bis(carboxymethoxy)coumarin-4-yl]methyl
Esters inside Cells**
Volker Hagen,* Stephan Frings, J¸rgen Bendig,
Dorothea Lorenz, Burkhard Wiesner, and
U. Benjamin Kaupp
Caged compounds are an elegant means to produce rapid
jumps in the concentration of chemical messenger molecules
inside cells.[1] Caged compounds are photolabile inactive
derivates of biologically active molecules. The biologically
active substance is rapidly released by a photochemical
reaction. Caged compounds allow the elucidation of complex
intracellular processes and their resolution in time and space.
For many applications it would be useful to determine
quantitatively the degree of photolysis of the caged compound
inside cells. In a few cases this has been possible by a rather
complicated calibration of the cellular reaction. Here we show
that the determination of the amount of released cyclic
nucleotides is feasible by fluorescence measurements, using
the axial and equatorial diastereomers of [6,7-bis(carboxy-
methoxy)coumarin-4-yl]methyl (BCMCM) esters of cyclic
adenosine-3’,5’-monophosphate (cAMP) 1 and cyclic guano-
sine-3’,5’-monophosphate (cGMP) 2,[2] which are excellent
phototriggers for cyclic nucleotides.
Received: March 7, 2002
Revised: July 5, 2002 [Z18849]
[1] D. A. Dougherty, H. A. Lester, Angew. Chem. 1998, 110, 2463 2466;
Angew. Chem. Int. Ed. 1998, 37, 2329 2331.
[2] D. Rees, A. J. Chirino, K.-H. Kim, H. Komiya in Membrane Protein
Structure (Ed.: S. H. White), Oxford University Press, New York,
1994, pp. 1 26.
[3] a) R. L. Juliano, V. Ling, Biochim. Biophys. Acta 1976, 455, 152 162;
b) M. M. Gottesmann, I. Pastan, Annu. Rev. Biochem. 1993, 62, 385
427; c) K. C. Almquist, D. W. Loe, D. R. Hipfner, J. E. Mackie, S. P.
Cole, R. G. Deeley, Cancer Res. 1995, 55, 102 110.
[4] R. Preiss, Int. J. Clin. Pharmacol. Therapeut. 1998, 36, 3 8.
[5] a) T. Tsuruo, H. Iida, S. Tsukagoshi, Y. Sakurai, Cancer Res. 1982, 41,
1967 1972; b) N. J. Chao, M. Aihara, K. G. Blume, B. I. Sikic, Exp.
Hematol. 1990, 18, 1193 1198; c) J. S. Kukl, B. I. Sikic, K. G. Blume,
N. J. Chao, Exp. Hematol. 1992, 20, 1048 1054.
Upon irradiation with UV light in aqueous buffer solution
and inside cells, 1 and 2 rapidly produce (within 2 5 ns) cAMP
[6] a) H. Nawrath, M. Raschack, J. Pharmacol. Exp. Ther. 1987, 242,
1090 1097; b) R. Pirker, D. J. Fitzgerald, M. Raschack, Z. Frank,
M. C. Willingham, I. Pastan, Cancer Res. 1989, 49, 4791 4795; c) T.
Watanabe, M. Naito, T. Oh-hara, Y. Itoh, D. Cohen, T. Tsuruo, Jpn. J.
Cancer Res. 1996, 87, 184 193.
[*] Dr. V. Hagen, Dr. D. Lorenz, Dr. B. Wiesner
Forschungsinstitut f¸r Molekulare Pharmakologie
Robert-Rˆssle-Strasse 10, 13125 Berlin (Germany)
Fax : (þ 49)30-94793-159
[7] A. Hilgeroth, U. Baumeister, Angew. Chem. 2000, 112, 588 590;
Angew. Chem. Int. Ed. 2000, 39, 576 578.
E-mail: hagen@fmp-berlin.de
[8] a) D. Kessel, Cancer Commun. 1989, 1, 145 149; b) J. L. Weaver, D.
Szabo, P. S. Pine, M. M. Gottesmann, S. Goldberg, A. Aszalos, Int. J.
Cancer 1993, 54, 456 461; c) D. Sharples, G. Hajos, Z. Riedl, D.
Csanyi, J. Molnar, D. Szabo, Arch. Pharm. Pharm. Med. Chem. 2001,
334, 269 274.
Priv.-Doz. Dr. S. Frings, Prof. Dr. U. B. Kaupp
Institut f¸r Biologische Informationsverarbeitung,
Forschungszentrum J¸lich (Germany)
Doz. Dr. J. Bendig
Institut f¸r Chemie
Humboldt-Universit‰t Berlin (Germany)
[9] a) H. L. Pearce, M. A. Winter, W. T. Beck, Adv. Enzyme Regul. 1990,
30, 357 373.
[10] G. Ecker, M. Huber, D. Schmidt, P. Chiba, Mol. Pharmacol. 1999, 56,
791 796.
[11] A. Seelig, E. Landwojtowicz, Eur. J. Pharm. Sci. 2000, 12, 31 40.
[**] This work was supported by the Deutsche Forschungsgemeinschaft
and the Fonds der Chemischen Industrie. We thank B. Dekowski und
Dr. S. Helm for technical assistance.
Angew. Chem. Int. Ed. 2002, 41, No. 19
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4119-3625 $ 20.00+.50/0
3625