Identification of a small moleucle inhibitor of the IL-2/IL-2Rα receptor interaction which binds to IL-2

Full text of DOI:10.1021/ja970702x

J. Am. Chem. Soc. 1997, 119, 7589-7590
Identification of a Small Molecule Inhibitor of the
7589
IL-2/IL-2Rr Receptor Interaction Which Binds to
IL-2
Jefferson W. Tilley,* Li Chen, David C. Fry,
S. Donald Emerson, Gordon D. Powers, Denise Biondi,
Tracey Varnell, Richard Trilles, Robert Guthrie,
Francis Mennona, Gerry Kaplan, Ronald A. LeMahieu,
Mathew Carson, Ru-Jen Han, C.-M. Liu,
Robert Palermo, and Grace Ju
Figure 1. Schild regressions of the effects of unlabeled IL-2 and
compound 1 on [125I]IL-2 binding to soluble IL-2RR. Plots are derived
from Scatchard analyses of [125I]IL-2 binding to immobilized receptor
in the presence of the plotted concentrations of unlabeled IL-2 or
compound. DR represents the dose ratio and is defined as the Kd of
Roche Research Center, Hoffmann-La Roche Inc.
Nutley, New Jersey 07110
ReceiVed March 4, 1997
ReVised Manuscript ReceiVed May 27, 1997
[
125I]IL-2 binding in the presence of inhibitor divided by the Kd in the
Interleukin-2 (IL-2) is a 15.5 kDa cytokine that has a
predominant role in the growth of activated T cells. IL-2
stimulates T-cell proliferation by binding on the T-cell surface
with picomolar affinity to a heterotrimeric receptor complex
consisting of R, â, and γ chains.1 Antibodies that recognize
the R receptor subunit (IL-2RR) and block IL-2 binding have
proven clinically effective as immunosuppressive agents,1,2 and
thus, we have sought small molecules capable of blocking the
IL-2/IL-2RR interaction as potential orally active successors to
the antibody drugs. The design of such agents is based on a
combination of structural information obtained by X-ray crystal-
lographic3,4 and NMR studies5 of IL-2 which reveal the
backbone to include a bundle of four R-helixes (A, B, C, and
D) with an up-up-down-down arrangement and site-directed
mutagenesis of human IL-2 which identified residues in the AB
loop (K35, R38, T41, F42, K43, Y45) and the B helix (E62,
L72) as being critical for the binding of IL-2 to IL-2RR.6,7
One member of a series of acylphenylalanine derivatives
intended to mimic the R38-F42 region of IL-2 was found to
be a competitive inhibitor of IL-2/IL-2RR binding with a mid-
micromolar IC50. Structure-activity studies led to the synthesis
of 1 with an IC50 of 3 µM. Although the acylphenylalanine
absence of inhibitor.
previously identified by site-directed mutagenesis are selectively
perturbed in a concentration-dependent manner by 1, but not
by its inactive enantiomer 2, indicating that 1 inhibits IL-2/IL-
2RR binding by associating with IL-2. To our knowledge, this
represents the first well-characterized example of a small
molecule, nonpeptide inhibitor of a cytokine/cytokine receptor
interaction.
Compounds were evaluated for the ability to inhibit IL-2/
IL-2RR interactions using a scintilation proximity competitive
binding assay.8 The assay basically measures binding of 125I-
labeled IL-2 to immobilized soluble IL-2RR in the presence of
various concentrations of inhibitor compound. As determined
from Hill plots of the inhibition curves, compound 1 inhibited
binding of IL-2 to IL-2RR with an IC50 of 3 µM at pH 7.4. In
contrast, it’s enantiomer, 2, gave only 14% inhibition of binding
at the highest concentration tested (500 µM), while IL-2 had
an IC50 of 13 nM. To characterize the inhibitory activity of 1,
binding assays were carried out with varying concentrations of
125I-labeled IL-2 in the presence of varying concentrations of 1
or unlabeled IL-2. Scatchard analyses were performed and the
Kd values for [125I]IL-2 binding at each inhibitor concentration
were plotted versus inhibitor concentration in a Schild plot9
(Figure 1), which indicates that the observed inhibitory activity
of 1 is consistent with a competitive reversible inhibition.
To assess the mechanism of inhibition, NMR experiments
were undertaken with uniformly 15N-enriched IL-2.10 The 1H-
15N NMR assignments were obtained by analysis of 3D H-
1
1H-15N-TOCSY- and NOESY-HSQC spectra.11 These reso-
nance assignments were, in most cases, determined to be similar
to those previously published.12 Perturbations of the H-15N
1
chemical shifts were measured in 2D 1H-15N-HSQC spectra13
which were acquired in the presence of 0.0, 0.5, 1.0, and 2.0
derivatives were designed to complex with IL-2RR by emulating
residues R38 and F42 of the IL-2 ligand, we considered the
possibility that they might instead interact with the ligand. NMR
studies using uniformly 15N-labeled IL-2 indicate that the
resonances of residues in the vicinity of the binding epitope
(8) Udenfriend, S.; Gerber, L.; Nelson, N. Anal. Biochem. 1987, 161,
494-500.
(9) Ehlert, F. J. Mol. Pharm. 1988, 33, 187-194.
(10) 15N-labeled IL-2 was produced by metabolic labeling of IL-2
produced in the yeast Pichia pastoris. Briefly, Pichia cells were transformed
with an expression vector that directed the synthesis and secretion of human
IL-2 under the control of a methanol-inducible promoter. The cells were
grown in media containing [15N]ammonium sulfate for 2 days at 30 °C.
The cells were then concentrated, and IL-2 expression was induced by
addition of methanol to the culture. The culture medium containing secreted
IL-2 was harvested. IL-2 was purified by affinity chromatography over an
IL-2RR column as described previously: Bailon, P.; Weber, D. V.; Kenney,
R. F.; Fredericks, J. E.; Smith, C.; Familletti, P. C.; Smart, J. E.
Bio/Technology 1987, 5, 1195-1198.
(1) Waldmann, T. A. Immunol. Today 1993, 14, 264-270.
(2) Hakimi, J.; Mould, D.; Waldman, T. A.; Queen, C.; Anasetti, C.;
Light, S. In Antibody Therapeutics; Harris, W. J., Adair, J. R., Eds.; CRC
Press: Boca Raton, FL, 1997; pp 277-300.
(3) Brandhuber, B. J.; Boone, T.; Kenney, W. C.; MaKay, D. B. Science
1987, 238, 1707-1709. Bazan, J. F. Science 1992, 257, 410-412. McKay,
D. B. Science 1992, 257, 412-413.
(4) Hatada, M.; Surgenor, A.; Weber, D.; Danho, W.; Madison, V.
Hoffmann-La Roche internal communication of unpublished results.
(5) Mott, H. R.; Baines, B. S.; Hall, R. M.; Cooke, R. M.; Driscoll, P.
C.; Weir, M. P.; Campbell, I. D. J. Mol. Biol. 1995, 247, 979-994.
(6) Sauve´, K.; Nachman, M.; Spence, C.; Bailon, P.; Campbell, E.; Tsien,
W.-H.; Kondas J. A.; Hakimi, J.; Ju, G. Proc. Natl. Acad. Sci. U.S.A. 1991,
88, 4636-4640.
(11) Pulse sequences were modified relative to the published TOCSY-
HMQC and NOESY-HMQC sequences found in: Marion, D.; Driscoll, P.
C., Kay, L. E.; Wingfield, P. T.; Bax, A.; Gronenborn, A. M.; Clore, G. M.
Biochemistry 1989, 28, 6150-6156.
(12) Mott, H. R.; Driscoll, P. C.; Boyd, J.; Cooke, R. M.; Weir, M. P.;
Campbell, I. D. Biochemistry 1992, 31, 7741-7744.
(13) Bodenhausen, G.; Ruben, D. J. Chem. Phys. Lett. 1980, 69, 185-
189.
(7) Ju, G. Unpublished result.
S0002-7863(97)00702-6 CCC: $14.00 © 1997 American Chemical Society

7590 J. Am. Chem. Soc., Vol. 119, No. 32, 1997
Communications to the Editor
Figure 2. (A) Expansion of 15N-1H-HSQC spectra of [15N]IL-2. The
arrows indicate the chemical shifts observed for four backbone NH
resonances in the absence (dark signals) and presence (light signals)
of 2 equiv of 1. The absolute values of H and 15N shift changes are
1
Figure 4. Molscript16 rendering of the X-ray crystal structure of human
illustrated for F42. (B) Vector magnitude chemical shift changes
(((∆δ1H)2 + (∆δ15N)2)1/2) for the same NH signals shown in part A.
Here the chemical shift changes for 0.59 mM [15N]IL-2 NH signals
are calculated upon addition of 1.0, 1.5, and 2.0 equiv of 1.
IL-2.4 (A) Nitrogen atom positions are represented as dark spheres for
1
1
the highly perturbed H-15N resonances defined in Figure 2. H-15N
resonances which are highly perturbed upon addition of 1 include I28,
N30, Y31, K32, N33, K35, L36, T37, M39, L40, T41, F42, K43, F44,
E62, K64, E68, L72, Q74, S75, and A112 backbone NHs and the side-
chain Nδ2H of N33. (B) R-Carbon atom (CR) positions where
mutagenesis studies have identified IL-2 side-chain interactions im-
portant in the binding of Hu-IL-2 to IL-2RR.6,7
approximate Kd measurement by NMR at pH ) 7.5 clearly
illustrates a stronger complex (lower Kd) than that observed at
pH ) 4.6 (data not shown). Figure 3 illustrates the sum of
absolute chemical shift changes for both backbone and side-
chain 1H-15N resonances of [15N]IL-2. The horizontal line at
115 Hz represents an arbitrary threshold above which 1H-15N
resonances have been defined as “highly perturbed” by addition
of 2 equiv of 1. No effect on the [15N]IL-2 NMR resonances
was observed upon addition of up to 2 equiv of the enantiomer
2, indicating that the chemical shift changes observed in the
presence of 1 are the result of a specific interaction between 1
and IL-2.
The N atoms corresponding to the residues with highly
perturbed resonances are shown as dark spheres on the ribbon
diagram shown in Figure 4A. It is clear from this figure that
these residues form a contiguous patch which primarily involves
the N-terminal half of the AB loop and the carboxy-terminal
half of the B helix. The same orientation of IL-2 is shown in
Figure 4B, in which the dark spheres represent the CR positions
of residues identified as important for the binding of human
IL-26,7 to IL-2RR by site-directed mutagensis. The evident
correspondence between the region of IL-2 perturbed by
association with 1 (Figure 4A) and that shown to be involved
in the binding to IL-2RR (Figure 4B) suggests that 1 interferes
with IL-2/IL-2RR binding by competing with IL-2RR for its
binding site on IL-2. Further work including attempts to solve
the X-ray and NMR structures of the IL-2/1 complex are
currently underway in our laboratories.
Figure 3. Sum of the absolute value of ∆δ1H and ∆δ15N chemical
shift differences for [15N]IL-2 resonances in the presence and absence
of 2.0 equiv of 1. 1H shift differences are in black and 15N shift
differences are in white. Shift differences are calculated as ∆δ ) |δ2eq
- δ0eq|, where δ2eq represents the chemical shift (in hertz) for the [15N]-
IL-2 signal the presence of 2.0 equiv of 1, and δ0eq represents the
chemical shift for [15N]IL-2 alone. Chemical shifts were measured in
2D 1H-15N-HSQC spectra15 at 40 °C and pH 4.6, in a buffer containing
10 mM sodium acetate, 0.02% sodium azide, 10% D2O, and 90% H2O,
and using a Varian Unityplus 600 spectrometer. The largest combined
shift difference (∆δ1H + ∆δ15N) was measured to be 300 Hz for F42.
The shift differences for the residues I28, N33, K43, and F44 have an
estimated lower limit of 300 Hz for their combined (∆δ1H + ∆δ15N)
shift. Their shifts could not be accurately quantitated due to exchange
broadening of the HSQC signal in the presence of 1. The horizontal
line at 115 Hz is the arbitrary threshold above which 1H-15N resonances
are defined as “highly perturbed” by the addition of 2 equiv of 1 to
[15N]IL-2.
equiv of 1 to [15N]IL-2 as indicated in Figure 2. While titration
curves shown in Figure 2B correspond to dissociation constants
of 2-6 mM for the IL-2 complex with 1, the drop in the binding
constant is primarily due to the difference in pH between the
NMR solution conditions (pH ) 4.6) and the binding assay14
(pH ) 7.4). Accurate determination of Kd at pH ) 7.4 by NMR
is difficult because of limited solubility of the protein; however,
Supporting Information Available: Text describing the synthesis
of 1 and 2 and a supplement to Figure 2 showing the complete 15N-
1H-HSQC spectra of [15N]IL-2 in the presence of 0.0, 1.0, 1.5, and 2.0
equiv of 1 (7 pages). See any masthead page for ordering and Internet
access instructions.
(14) Hakimi, J.; Cullen, B.; Chizzonite, R.; Ju, G.; Bailon, P.; Tsien,
W.-H.; Schumann, M.; Liu, C. M. In Biology of Actinomycetes ’88; Okami,
Y., Beppu, T., Ogawara, H., Eds.; Japan Scientific Societies Press: Tokyo,
1988; pp 189-194.
(15) Kay, L. E.; Keifer, P.; Saarinen, T. J. Am. Chem. Soc. 1992, 114,
10663-10665.
(16) Kraulis, P. J. J. Appl. Crystallogr. 1991, 24, 946-950.
JA970702X