7394
J. Am. Chem. Soc. 2000, 122, 7394-7395
Second-Site NMR Screening with a Spin-Labeled
First Ligand
Wolfgang Jahnke,* Lawrence B. Perez, C. Gregory Paris,
Andre´ Strauss, Gabriele Fendrich, and Carlo M. Nalin
NoVartis Pharma AG, Core Technologies
CH-4002 Basel, Switzerland
Oncology Research and Core Technologies
Summit, New Jersey
ReceiVed April 10, 2000
ReVised Manuscript ReceiVed June 2, 2000
NMR spectroscopy in drug discovery research has recently
widened its scope. Traditionally, biomolecular NMR has been
primarily used to support lead optimization by providing structural
information about lead compounds complexed to the target
molecule (often a protein) under investigation. More recently,
NMR has been recognized as a technique that is also valuable
for lead finding. In a pioneering technique termed “SAR by
NMR”,1 NMR screening2 is used to identify a ligand for a first
binding site on the target. At saturating concentrations of this
first ligand, NMR screening is then used to discover a second
ligand that binds to the target simultaneously and in the vicinity
to the first ligand (“second-site screening”). The structure of the
ternary complex is determined and used to guide chemistry to
connect both ligands.1,3 Due to additive binding energies and
favorable entropic effects, the resulting linked compound may
be a nanomolar ligand for the target, although both individual
fragments exhibited only millimolar or micromolar affinity. Here
we show that many of the exisiting problems in second-site
screening, such as the occurrence of false positives, excessive
protein demands, low sensitivity, insufficient compound solubility,
and difficult automation, can be eliminated by using a spin-labeled
analogue of the first ligand.
Spin-spin (R2) relaxation rates are proportional to the product
of the squares of the involved spins’ gyromagnetic ratio. The
gyromagnetic ratio of protons is small, and R2 relaxation rates
are thus relatively small. This is bliss for most aspects of NMR;
however, the method presented here uses spin-spin relaxation
to detect ligand binding, and thus spin-spin relaxation rates
should be as large as possible. The gyromagnetic ratio of an
unpaired electron is 658 times as large as that of a proton. Hence,
R2 relaxation effects on protons by an unpaired electron, called
paramagnetic relaxation enhancements, R2para, are dramatically
larger than the effects of a nuclear-nuclear interaction. Conse-
quently, NMR studies employing reagents with unpaired electrons
(“spin labels”) have been successful in measuring long distances,
dynamics, or surface accessibilities in proteins.4 Here we introduce
a new application of spin labels in biomolecular NMR to detect
simultaneous binding of two ligands to a target protein. The
method detects the paramagnetic relaxation enhancement on a
second ligand caused by a spin-labeled first ligand, if and only if
Figure 1. Principle of second-site screening using a spin-labeled first
ligand. The spin-labeled first ligand (1*) binds in the “southern” binding
site of Bcl-xL. If and only if a compound binds simultaneously at a
neighboring binding site, like compound 2, does it experience a
paramagnetic relaxation enhancement (thick arrows).
both ligands are bound to the target protein at the same time and
at neighboring binding sites (Figure 1).
The method is illustrated with the anti-apoptotic protein Bcl-
xL as an example. Bcl-2 and Bcl-xL are primarily responsible
for the reduced susceptibility of cancer cells to undergo pro-
grammed cell death (apoptosis) upon chemotherapy, and are
therefore interesting targets for cancer therapy.5 In-house high-
throughput screening using Bcl-2/Bax and Bcl-xL/Bax ELISA
assays identified compound 1 which has an IC50 value of 140
µM for disruption of the Bcl-xL/Bax interaction. The affinity of
this compound could not be optimized by traditional medicinal
chemistry. The binding site of 1 on Bcl-xL was determined by
NMR. It was found to overlap with the binding site of the
N-terminal part of the Bak peptide, the natural antagonist.5d,6 This
binding site is in the “southern“ part of the hydrophobic groove
and is formed by the side chains of residues Leu 130, Leu 108,
Ala 142, and Phe 105 (Figure 1). There is an adjacent binding
site, termed the “northern” binding site, where the C-terminus of
Bak peptide binds, and for which a ligand is sought by second-
site NMR screening. Second-site screening using existing meth-
ods2 was hampered by the inability to saturate the southern binding
site with 1, due to its low aqueous solubility.
Compound 1*, an adduct of 1 with a TEMPO7 spin label, was
synthesized by treatment of 4-bromo-1,1-di(4-chlorophenyl)-1-
* Address correspondence to this author at Novartis Pharma AG. E-mail:
(1) Shuker, S. B.; Hajduk, P. J.; Meadows, R. P.; Fesik, S. W. Science
1996, 274, 1531-1534.
(5) (a) Chao, D. T.; Korsmeyer, S. J. Annu. ReV. Immunol. 1998, 16395-
16419. (b) Reed, J. C. Oncogene 1998, 17, 3225-3236. (c) Muchmore, S.
W.; Sattler, M.; Liang, H.; Meadows, R. P.; Harlan, J. E.; Yoon, H. S.;
Nettesheim, D.; Chang, B. S.; Thompson, C. B.; Wong, S.-L.; Ng, S.-C.; Fesik,
S. W. Nature 1996, 381, 335-341. (d) Sattler, M.; Liang, H.; Nettesheim,
D.; Meadows, R. P.; Harlan, J. E.; Eberstadt, M.; Yoon, H. S.; Shuker, S. B.;
Chang, B. S.; Minn, A. J.; Thompson, C. B.; Fesik, S. W. Science 1997, 275,
983-986.
(2) (a) Fejzo, J.; Lepre, C. A.; Peng, J. W.; Bemis, G. W.; Ajay; Murcko,
M. A.; Moore, J. M. Chem. Biol. 1999, 6, 755-769. (b) Chen, A.; Shapiro,
M. J. Anal. Chem. 1999, 71, 669A-675A. (c) Hajduk, P. J.; Olejniczak, E. T.;
Fesik, S. W. J. Am. Chem. Soc. 1997, 119, 12257-12261. (d) Meyer, B.;
Weimar, T.; Peters, T. Eur. J. Biochem. 1997, 246, 705-709. (e) Moore, J.
M. Curr. Op. Biotechnol. 1999, 10, 54-58.
(6) Compound 1 was found to also bind weakly to the “northern” binding
site. Although this feature is interesting, it is not relevant in the present context
since relay or spin diffusion effects via the protein can be excluded. Compound
2 binds exclusively to the northern binding site.
(3) (a) Hajduk, P. J.; Sheppard, G.; Nettesheim, D. G.; Olejniczak, E. T.;
Shuker, S. B.; et al. J. Am. Chem. Soc. 1997, 119, 5818-5827. (b) Hajduk,
P. J.; Dinges, J.; Miknis, G. F.; Merlock, M.; et al. J. Med. Chem. 1997, 40,
3144-3150.
(7) TEMPO, 2,2,6,6-tetramethyl-1-piperidine-1-oxyl. TEMPO is com-
mercially available with a variety of functional groups.
(4) Kosen, P. A. Methods Enzymol. 1989, 177, 86-121.
10.1021/ja001241+ CCC: $19.00 © 2000 American Chemical Society
Published on Web 07/18/2000