M. Koller et al. / Bioorg. Med. Chem. Lett. 21 (2011) 3358–3361
3361
boundaries, the oral ED50-values are not directly correlated to the
affinities of the compounds. For example, the weakest ligand 16
has the lowest ED50-value, whereas 9, with an IC50-value slightly
below 1 lM, requires the highest dose for anticonvulsant effects.
On the other hand, the nanomolar compound 28 is devoid of oral
activity in DBA/2 mice and requires high intraperitoneal doses to
inhibit E-shock induced convulsions in mice. This disappointing re-
sult may—at least in part—be explained by the higher polar surface
area (PSA) of 28 (164 Å2) compared to the PSA of the orally active
compounds (101–110 Å2). Given the strongly polar character of the
glutamate binding cavity of the AMPA receptor, such an increase in
PSA may strengthen the receptor interactions but also hamper
absorption/distribution, and in particular the brain penetration of
the compounds.22
In conclusion, the quinazolinedione sulfonamides represent a
novel class of competitive AMPA receptor antagonists, displaying
nanomolar affinities and providing examples—albeit of lower affin-
ity—with oral activity in animal models for anticonvulsant effects.
Whether these two properties—high receptor affinity and oral
in vivo activity—can be combined within a single compound, will
be the topic of further investigations.
Figure 1. X-ray structure at 2.1 Å resolution of the ligand binding domain of a
hGluA2 construct (carbons in yellow, nitrogens in blue, oxygens in red and sulfurs
in brown) bound to 28 (carbons in cyan). Selected interactions (distances in Å) and
water molecules are shown in white.
References and notes
1. Lees, G. J. Drugs 2000, 59, 33.
2. Mattes, H.; Carcache, D.; Kalkman, H. O.; Koller, M. J. Med. Chem. 2010, 53, 5367.
3. Newcomer, J. W.; Krystal, J. H. Hippocampus 2001, 11, 529.
4. Rosenmund, C.; Stern-Bach, Y.; Stevens, C. F. Science 1998, 280, 1596.
5. Leeson, P.; Baker, R.; Carling, R. W.; Curtis, N. R.; Moore, K. W.; Williams, B.;
Foster, A. C.; Donald, A. E.; Kemp, J. A.; Marshall, G. R. J. Med. Chem. 1991, 34,
1243.
Table 2
Anticonvulsant effects against E-shock induced sei-
zures after oral administration in mice (pretreatment
time 60 min); N = 5
6. Wilding, T. J.; Huettner, J. E. Mol. Pharmacol. 1996, 49, 540.
7. Ohmori, J.; Sakamoto, S.; Kubota, H.; Shimizu-Sasamata, M.; Okada, M.;
Kasasaki, S.; Hidaka, H.; Togami, J.; Furuya, T.; Murase, K. J. Med. Chem. 1994,
37, 467.
Compd
ED50 [mg/kg]
16
17
4
19
9
9
16
23
8. Suter, W.; Hartmann, A.; Poetter, F.; Sagelsdorff, P.; Hoffmann, P.; Martus, H.-J.
Mutat. Res., Genet. Toxicol. Environ. Mutagen. 2002, 518, 181.
9. Krishnamurthy, M.; Iyer, K. A.; Dogra, S. K. J. Photochem. 1987, 38, 277.
10. In the mean time, this chemotype was published by other groups: (a) Colotta,
V.; Catarzi, D.; Verano, F.; Calabri, F. R.; Filacchioni, G.; Costagli, C.; Galli, A.
Bioorg. Med. Chem. Lett. 2004, 14, 2345; (b) Colotta, V.; Catarzi, D.; Varano, F.;
Lenzi, O.; Filachioni, G.; Costagli, C.; Galli, A.; Ghelardini, C.; Galeotti, N.;
Gratteri, P.; Sgrignani, J.; Deflorian, F.; Moro, S. J. Med. Chem. 2006, 49, 6015.
11. Koller, M. WO Patent 19346, 1995.
29.5
54
12
28
60%a
40%b
a
Inhibition after oral administration of 30 mg/kg.
Inhibition after intraperitoneal administration of
30 mg/kg (28 is orally inactive against sound induced
seizures in DBA/2 mice).
b
12. All compounds were characterized by 1H NMR (360 MHz) and MS-
spectroscopy.
13. Honore, T.; Drejer, J.; Nielsen, E. O.; Nielsen, M. Biochem. Pharmacol. 1989, 38,
3207.
14. Baron, M. B.; Siegel, B. W.; Harrison, B. L.; Gross, R. S.; Hawes, C.; Towers, P. J.
Pharmacol. Exp. Ther. 1996, 279, 62.
15. Takano, Y.; Futoshi, S.; Asano, J.; Ando, N.; Uchiki, H.; Anraku, T. Bioorg. Med.
Chem. Lett. 2003, 13, 3521.
16. A similar idea was followed by attaching the phosphono glycine side chain of
the potent compound (V) (AMP397) in position 5 of compound 10; however,
the resulting product was only marginally active at the AMPA receptor (data
not shown).
therefore, selectivity is improved. On the other hand, the benzyl-
substituted sulfonamide 25 shows good selectivity for the glycine
receptor (about 14-fold), demonstrating that quinazolinedione sul-
fonamides could be modified towards glycine site selective ligands.
The promising oral activity found with 4 in the mouse E-shock
test led to further in vivo examination of the series. Several com-
pounds turned out to be orally active in the audiogenic seizure par-
adigm in DBA/2 mice20, and a selection of those, together with the
orally inactive compound 28, was then tested in the E-shock assay
in OF1 mice21 (Table 2). All compounds of the selection inhibited
E-shock induced seizures after oral administration, and only 28
was devoid of oral activity.
17. Armstrong, N.; Gouaux, E. E. Neuron 2000, 28, 165.
18. Kallen, J. et al. The details of this X-ray structure determination will be
published elsewhere. The coordinates have been deposited with PDB ID code
3R7X.
19. Furukawa, H.; Gouaux, E. EMBO J. 2003, 22, 2873.
20. DeSarro, G. B.; Meldrum, B. S.; Nistico, G. Br. J. Pharmacol. 1988, 93, 247.
21. Wamil, A. W.; Schmutz, M.; Portet, C.; Feldmann, A. F.; McLean, M. J. Eur. J.
Pharmacol. 1994, 271, 301.
The orally active compounds shown display IC50-values of
0.9–9.7 lM for the AMPA receptor. Within these narrow
22. Hitchcock, S. A.; Pennington, L. D. J. Med. Chem. 2006, 49, 7559.