A. Kolocouris et al. / Bioorg. Med. Chem. Lett. 18 (2008) 6156–6160
6159
However, the replacement of the amino group with a hydroxyl
group (compounds 4, 6) abolished activity.
Anti-influenza virusal potency. The potencies of the aminoada-
mantane derivatives 1–3, 7, 8–10 and alcohols 4, 6 in inhibitoring
it therefore appears that compounds 2 and 3 bind to M2 protein in
a more favorable orientation. Of the 2-alkyl-2-aminoadamantane
0
analogues, spiro[piperidine-2,2 -adamantane] 3 had the stronger
binding and antiviral potency, which were similar to those of
amantadine 1. The present SAR study indicates that large and ex-
tended lipophilic moieties in the vicinity of adamantane carbon
C-2 are compatible with biological activity and suggests that there
is a complementary acceptor group/site within the lumen of the
M2 channel pore. Thus, the potency of the new compounds can be
rationalized in terms of a larger lipophilic cavity inside the lumen of
the M2 channel that can accommodate larger entities than the ada-
mantyl group. (b) It is apparent that for a series of aminoadaman-
tane compounds, the relative binding affinities to M2 protein are
not directly comparable to the relative antiviral potencies in cell
culture. (c) The aminoadamantane derivatives tested did not differ
from amantadine 1 in their NMDA antagonistic activity. (d) Ada-
mantanols 4, 6 were found to be active anti-influenza virus A
agents without having antagonistic activity against the NMDA
receptor in the cell-based assay. The simultaneous presence of
anti-influenza virus A activity with potentially no (or fewer) CNS
side effects are of particular interest and merit further
investigation.
in vitro replication of influenza virus A H
2 2 3 2
N and H N subtypes
and influenza B viruses, was determined using previously reported
2
1
methods (Table 3). All compounds were shown to be active
against the A/H and H strains, with corresponding low tox-
2
N
2
3 2
N
icity, and the amines 2, 3, 9, 10 were almost equipotent to aman-
tadine 1. All compounds tested were inactive at low
concentrations against influenza virus B in accordance with their
putative mode action, i.e., their interaction with influenza virus A
2
2
M2 protein, but inactivity against the influenza B BM2 protein.
Compound 3 reduced influenza virus B cytopathogenicity but only
at high concentrations. The latter effect is not specific since high
concentrations of amines generally inhibit the low pH-induced,
HA-mediated membrane fusion involved in uncoating of influenza
virus during endocytosis.23
Unexpectedly in view of their apparent lack of M2 binding in
our in vitro assay, the alcohols 4 and 6 exhibited marked inhibitory
activities against influenza virus A (but not B), which for A/Japan/
3
05/57(H2N2) were equivalent to those for the corresponding
amines, though lower for X-31(H3N2). These characteristics are
consistent with a similar specific mode of action of the alcohols
and amines against influenza virus A. It is therefore unlikely that
the alcohols 4 and 6 exert their inhibitory activity through a mech-
anism which is different from inhibition of the M2 channel, and it
is likely that the apparent inconsistencies in the results from the
two types of experiment are due to differences in the experimental
conditions. The fluorescence M2 binding assay was conducted at
pH 5 in the presence of low concentrations of detergent (0.1%
Acknowledgments
This research was in part supported by the Medical Research
Council (UK) and by the EU FP6 Programme VIRGIL, Contract Num-
ber 503359. F.X.S. gratefully acknowledges financial support from
the Spanish Ministerio de Educación y Ciencia (Project SAF 2006-
13092-C02-01).
1
3
LDAO), which may result in lower concentrations of the water
insoluble alcohols being available for interaction with the M2 pro-
tein. In this respect, recent preliminary results from electrophysio-
logical measurements have shown that the alcohol 4 does inhibit
proton currents in M2-expressing mouse erythroleukaemia (MEL)
References and notes
1.
(a) Taubenberger, J. K.; Reid, A. H.; Lourens, R. M.; Wang, R.; Jin, G.; Fanning, T.
G. Nature 2005, 437, 889; (b) Webby, R. J.; Webster, R. G. Science 2003, 302,
1519.
2. (a) Russell, C. J.; Webster, R. G. Cell 2005, 123, 368; (b) Webster, R. G.; Webby, R.
J.; Hoffmann, E.; Rodenberg, J.; Kumar, M.; Chu, H.-J.; Seiler, P.; Krauss, S.;
Songserm, T. Virology 2006, 351, 303; (c) Abdel-Ghafar, A. N.;
Chotpitayasunondh, T.; Gao, Z.; Hayden, F. G.; Nguyen, D. H.; de Jong, M. D.;
Naghdaliyev, A.; Peiris, J. S.; Shindo, N.; Soeroso, S.; Uyeki, T. M. N. Eng. J. Med.
2
4
cells, although the M2 specificity of the effect has yet to be
determined.
The antiviral activities (Table 3) indicate that the 2-ethyl, 2-pro-
pyl, spiropyrrolidine and spiropiperidine derivatives 9, 10, 2, 3
exhibited a small increase in potency with respect to the 2-methyl
derivative 8; it appears that enlargement of the alkyl group in-
creases anti-influenza virus A potency. However, the results in Ta-
ble 1 suggest that increase in size of the freely rotating alkyl group
reduces affinity of binding to M2. Thus, as for the alcohols, the rel-
ative binding affinities are not directly comparable to the relative
antiviral potencies in cell culture; while spiropyrrolidine 2 had a
2008, 358, 261.
3.
(a) Tumpey, T. M.; García-Sastre, A.; Mikulasova, A.; Taubenberger, J. K.;
Swayne, D. E.; Palese, P.; Basler, C. F. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 13849;
(b) Ilyushina, N. A.; Bovin, N. V.; Webster, R. G.; Govorkova, E. A. Antiviral Res.
2006, 70, 121.
4
5
.
.
(a) Hay, A. J.; Wolstenholme, A. J.; Skehel, J. J.; Smith, M. H. EMBO J. 1985, 4,
3021; (b) Hay, A. J. Semin. Virol. 1992, 3, 21.
Pinto, L. H.; Holsinger, L. J.; Lamb, R. A. Cell 1992, 69, 517.
6. In Burger’s Medicinal Chemistry, 5th ed.; Wolff, M. E., Eds.; John Wiley & Sons:
New York; Vol. 4, pp 590–591.
7.
(a) Kolocouris, A. Ph.D. Thesis, University of Athens, Greece, March 1995.; (b)
Kolocouris, N.; Foscolos, G. B.; Kolocouris, A.; Marakos, P.; Pouli, N.; Fytas, G.;
Ikeda, S.; De Clercq, E. J. Med. Chem. 1994, 37, 2896; (c) Kolocouris, N.;
Kolocouris, A.; Foscolos, G. B.; Fytas, G.; Neyts, J.; Padalko, E.; Balzarini, J.;
Snoeck, R.; Andrei, G.; De Clercq, E. J. Med. Chem. 1996, 39, 3307; (d) Fytas, G.;
Foscolos, G.; Stamatiou, G.; Kolocouris, A.; Kolocouris, N.; Witvrouw, M.;
Pannecouque, C.; De Clercq, E. Bioorg. Med. Chem. Lett. 1997, 1887; (e)
Kolocouris, A.; Tatarides, D.; Fytas, G.; Foscolos, G.; Mavromoustakos, T.;
Kolocouris, N.; De Clercq, E. Bioorg. Med. Chem. Lett. 1999, 3465; (f) Stamatiou,
G.; Kolocouris, A.; Kolocouris, N.; Fytas, G.; Foscolos, G. B.; Neyts, J.; De Clercq,
E. Bioorg. Med. Chem. Lett. 2001, 11, 2137; (g) Zoidis, G.; Kolocouris, N.; Foscolos,
G. B.; Kolocouris, A.; Fytas, G.; Karayannis, P.; Padalko, E.; Neyts, J.; De Clercq, E.
Antiviral Chem. Chemother. 2003, 14, 153; (h) Stamatiou, G.; Foscolos, G. B.;
Fytas, G.; Kolocouris, A.; Kolocouris, N.; Pannecouque, C.; Witvrouw, M.;
Padalko, E.; Neyts, J.; De Clercq, E. Bioorg. Med. Chem. 2003, 11, 5485; (i)
Stylianakis, I.; Kolocouris, A.; Kolocouris, N.; Fytas, G.; Foscolos, G. B.; Padalko,
E.; Neyts, J.; De Clercq, E. Bioorg. Med. Chem. Lett. 2003, 13, 1699; (j) Setaki, D.;
Tataridis, D.; Stamatiou, G.; Kolocouris, A.; Kolocouris, N.; Foscolos, G. B.; Fytas,
G.; Padalko, E.; Neyts, J.; De Clercq, E. Bioorg. Chem. 2006, 34, 248; (k) Tataridis,
D.; Fytas, G.; Kolocouris, A.; Fytas, C.; Kolocouris, N.; Foscolos, G. B.; Padalko, E.;
Neyts, J.; De Clercq, E. Bioorg. Med. Chem. Lett. 2007, 17, 692.
3
.6-fold lower binding constant than amantadine 1 their in vitro
potencies were equivalent. In mechanistic terms, in order to inhibit
virus replication amantadine 1 must first be solvated in the lipid
2
5
bilayer prior to the blockade of the M2 proton channel inside
the acidic endosome or trans Golgi in which, like other acidotropic
amines, amantadine 1 concentrates. Thus, the observed anti-influ-
enza virus A potencies of the aminoadamantane derivatives can be
considered to result from a combination of their ability to reach the
site of interaction with the receptor, via membrane penetration,
and binding affinity.
The major conclusions from this study can be summarized as
follows: (a) Transposition of the amino group from adamantane
C-1 in amantadine 1 to C-2 in 2-amantadine 7 resulted in a lower
binding affinity to the M2 channel. For the active 2-alkyl-2-amino-
adamantane analogues 8–10, which have an extended lipophilic
substituent, the binding affinity was reduced by increasing the size
of the rotatable alkyl group. In contrast, on going from 2-methyl-2-
adamantanamine 8 to the pyrrolidine 2 or the piperidine 3 deriva-
tives, the additional lipophilic moiety boosted the binding affinity;
8.
Recently an X-ray structure of M2TM and M2TM–amantadine complex was
published from the DeGrado’s group: (a) Stouffer, A. L.; Acharya, R.; Salom, D.;
Levine, A. S.; Di Costanzo, L.; Soto, C. S.; Tereshko, V.; Nanda, V.; Stayrook, S.;
DeGrado, W. Nature 2008, 451, 596; The first work on the amantadine location