A. Wissner et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1411–1416
1415
our earlier modeling studies of the quinazolines and
quinoline-3-carbonitriles. The complex was energy
minimized using the CharmM forcefield as implemented
in QuantaTM software.20 As in our previous work, we
allowed the entire system of protein, ligand, and water
molecules to move during minimization. In the final
model (Fig. 1) the N1 atom of 28 is hydrogen-bonded to
the backbone NH of Met 769. The 3-cyano group of 28
has displaced the water molecule that bridged the N3
atom of the original quinazoline ligand to the hydroxyl
group of Thr 766. The nitrogen atom of the 3-cyano
group now interacts with that same residue. The 4-(3-
bromoaniline) moiety lies in a hydrophobic pocket con-
taining Leu 820, Thr 830, Asp 831, Val 702, Ala 719, Ile
720, Lys 721, Thr 766, and Leu 764. Most significantly,
the b-carbon atom of the Michael acceptor side chain of
28 is located 4.13A from the sulfhydryl group of Cys
773, and should be easily accessible for covalent inter-
action. Additionally, the N-atom of the dimethylamino
group on 28 is located 3.56 A away from the sulfhydryl
hydrogen of Cys 773. In our earlier work on inhibitors
in the quinazoline and quinoline-3-carbonitrile series,
we proposed (and presented experimental evidence in
support of the fact) that this dimethylamino group can
serve as an intramolecular base catalyst for Michael
additions to these types of inhibitors. We are now pro-
posing that the dimethylamino group of 28 is playing a
similar role. Also, given the arrangement of functional-
ities predicted by this binding model, we suggest that
this intramolecular catalysis operates after 28 binds at
the active site of EGFR and that the Michael addition
reaction of the sulfhydryl group of Cys 773 to bound 28
is accelerated (relative to reactions of other nucleophiles
that may be in the cytosol) due to the close proximity of
the reactive center, nucleophile, and base catalyst.
interaction of the hydrogen atom attached to C8 and
the carbonyl oxygen of Met 769 is expected to be stabi-
lizing21À23 via a weak CH–O hydrogen bond. If the 1,8-
naphthyridine analogue 14 adopted a similar binding
orientation compared to 28, not only would the favor-
able aromatic hydrogen–carbonyl interaction be lost,
but it would be replaced by an unfavorable interaction
resulting from the repulsion of the lone pairs on N8 of
the inhibitor and the carbonyl oxygen of Met 769. In
reality, it is likely that 14 would bind in a manner that
would mitigate this unfavorable interaction, but this
might orient the molecule in a less favorable position for
the covalent interaction to occur at an efficient rate.
Additionally, one would expect that this aromatic
hydrogen–carbonyl interaction would be stronger the
more polarized the C–H bond. The C–H bond at the 8-
position is expected to be more polarized in 28 com-
pared to 32 and this might account for the slightly better
potency of 28 compared 32.
Acknowledgements
The authors wish to thank Drs. Philip Frost, Tarek
Mansour, and Janis Upeslacis for their support and
encouragement. We also would like to thank the mem-
bers of the Wyeth Discovery Analytical Chemistry
group for analytical and spectroscopic determinations.
References and notes
1. Hickey, K.; Grehan, D.; Reid, I. M.; O’Briain, S.; Walsh,
T. N.; Hennessy, T. P. J. Cancer 1994, 74, 1693.
2. Delarue, J. C.; Terrier, P.; Terrier-Lacombe, M. J.;
Mouriesse, H.; Gotteland, M.; May-Levin, F. Bull. Can-
cer 1994, 81, 1067.
3. Salomon, D. S.; Brandt, R.; Ciadiello, F.; Normanno, N.
Crit. Rev. Oncol. Haematol. 1995, 19, 183.
4. Gullick, W. J. Br. Med. Bull. 1991, 47, 87.
5. Woodburn, J. R. Pharmacol. Ther. 1999, 82, 241.
6. Bridges, A. J. Curr. Med. Chem. 1999, 6, 825.
7. Wissner, A.; Berger, D. M.; Boschelli, D. H.; Floyd,
M. B., Jr.; Greenberger, L. M.; Gruber, B. C.; Johnson,
B. D.; Mamuya, N.; Nilakantan, R.; Reich, M. F.; Shen,
R.; Tsou, H.-R.; Upeslacis, E.; Wang, Y.-F.; Wu, B.; Ye,
F.; Zhang, N. J. Med. Chem. 2000, 43, 3244.
Of particular significance tothe case at hand, the C8
atom of 28 is located 3.18 A from the backbone car-
bonyl oxygen of Met 769. We believe that this might
account for the large difference in activity we see
between the 1,7- and 1,8-naphthyridines. For 28, the
8. Wissner, A.; Overbeek, E.; Reich, M. F.; Floyd, M. B.;
Johnson, B. D.; Mamuya, N.; Rosfjord, E. C.; Discafani,
C.; Davis, R.; Shi, X.; Rabindran, S. K.; Gruber, B. C.;
Ye, F; Hallett, W. A.; Nilakantan, R.; Shen, R.; Wang,
Y.-F.; Greenberger, L. M.; Tsou, H. J. Med. Chem. 2003,
46, 49.
9. Wissner, A.; Floyd, M. B.; Rabindran, S. K.; Nilakantan,
R.; Greenberger, L. M.; Shen, R.; Wang, Y.-F.; Tsou, H.
Bioorg. Med. Chem. Lett. 2002, 12, 2893.
10. Torrance, C. J.; Jackson, P. E.; Montgomery, E.; Wissner,
A.; Kinzler, K. W.; Vogelstein, B.; Frost, P.; Discafani,
C. M. Nature Medicine 2000, 6, 1024.
11. For a review see: Boschelli, D. H. Curr. Topics Med.
Chem. 2002, 2, 1051.
12. Klunder, J. M.; Hargrave, K. D.; West, M.; Cullen, E.;
Pal, K.; Behnke, M. L.; Kapadia, S. R.; McNeil, D. W.;
Wu, J. C.; Chow, G. C.; Adams, J. J. Med. Chem. 1992,
35, 1887.
13. (a) Brechbuhler, H.; Buchi, H.; Hatz, E.; Schreiber, J.;
Figure 1. Binding model of 28 at the active site of EGFR kinase.