Synthesis and binding affinities of novel SRIF-mimicking β-D-glucosides satisfying the requirement for a π-cloud at C1

Article abstract of DOI:10.1021/ol050119i

(Chemical Equation Presented) The synthesis of four bioactive analogues of the somatostatin (SRIF-14) mimetic, β-D-glucoside (+)-2, in which the C1 indole side chain is replaced with indole surrogates, has been achieved. These congeners, possessing the na

Full text of DOI:10.1021/ol050119i

ORGANIC
LETTERS
2005
Vol. 7, No. 6
1121-1124
Synthesis and Binding Affinities of
Novel SRIF-Mimicking -Glucosides
Satisfying the Requirement for a
-Cloud at C1
â
-D
π
Angie R. Angeles,† Irina Neagu,† Elizabeth T. Birzin,‡ Edward R. Thornton,*,†
Amos B. Smith, III,*,† and Ralph Hirschmann*,†
Department of Chemistry, UniVersity of PennsylVania,
Philadelphia, PennsylVania 19104-6323, and Department of Biochemistry and
Physiology, Merck Research Laboratories, Rahway, New Jersey 07065-0900
rfh@sas.upenn.edu
Received January 19, 2005
ABSTRACT
The synthesis of four bioactive analogues of the somatostatin (SRIF-14) mimetic,
â
-D-glucoside (+)-2, in which the C1 indole side chain is
replaced with indole surrogates, has been achieved. These congeners, possessing the naphthyl, benzothiophene, benzyl, and benzofuran
substituents, were predicted to satisfy the electrostatic requirements of the tryptophan binding pocket of SRIF. Unlike the previously described
C4 picolyl and pyrazinyl congeners, these ligands bind the hSST4 receptor.
In 1987, we initiated a then conceptually novel approach
for the discovery of nonpeptidal mimetics of cyclic peptides,
as an alternative to random screening, that involved attaching
relevant amino-acid-mimicking side chains to â-D-glucose
via ether linkages.1 Others have subsequently employed
various monosaccharides as privileged scaffolds incorporat-
ing a â-turn.2
Early on, molecular modeling suggested that attachment
of the indole and the lysine-mimicking side chains at the
C1 and C6 positions of the sugar, respectively, would provide
good overlap with those of Trp8 and Lys9 of somatostatin-
14 (SRIF-14, 1, Figure 1) in the critical i + 1 and i + 2
positions of the â-turn.1a,3 Molecular modeling also demon-
strated that the 2-benzyl substituent mimics the important
Phe7 of SRIF-14.
(2) (a) Papageorgiou, C.; Haltiner, R.; Bruns, C.; Petcher, T. J. Bioorg.
Med Chem. Lett. 1992, 2, 135. (b) LeDiguarher, T.; Boudon, A.; Elwell,
C.; Paterson, D. E.; Billington, D. C. Bioorg. Med. Chem. Lett. 1996, 6,
1983. (c) Dinh, T. Q.; Smith, C. D.; Du, X. H.; Armstrong, R. W. J. Med.
Chem. 1998, 41, 981. (d) Moitessier, N.; Minoux, H.; Maigret, B.; Chretien,
F.; Chapleur, Y. Lett. Pept. Sci. 1998, 5, 75. (e) Sofia, M. J.; Hunter, R.;
Chan, T. Y.; Vaughan, A.; Dulina, R.; Wang, H. M.; Gange, D. J. Org.
Chem. 1998 63, 2802. (f) Sofia, M. J. Med. Chem. Res. 1998, 8, 362. (g)
Wunberg, T.; Kallus, C.; Opatz, T.; Henke, S.; Schmidt, W.; Kunz, H.
Angew. Chem., Int. Ed. 1998, 37, 2503. (h) Kallus, C.; Opatz, T.; Wunberg,
T.; Schmidt, W.; Henke, S.; Kunz, H. Tetrahedron Lett. 1999, 40, 7783.
(i) Schweizer, F.; Hindsgaul, O. Curr. Opin. Chem. Biol. 1999, 3, 291. (j)
Sofia, M. J.; Silva, D. J. Curr. Opin. Drug DiscoVery DeV. 1999, 2, 365.
(k) Murphy, P. V.; O’Brien, J. L.; Smith, A. B., III. Carbohydr. Res. 2001,
334, 327. (l) Nakayama, K.; Kawato, H. C.; Inagaki, H.; Ohta, T. Org.
Lett. 2001, 3, 3447. (m) Le, G. T.; Abbenane, G.; Beker, B.; Gatwohl, M.;
Halliday, J.; Tometzki, G.; Zuegg, J.; Meutermans, W. Drug DiscoVery
Today 2003, 15, 701-709. (n) Gruner, S. A. W.; Locardi, E.; Lohof, E.;
Kessler, H. Chem. ReV. 2002, 102, 491-514. (o) Timmer, M. S. M.;
Verdoes, M.; Sliedregt, L. A. J. M.; Van der Marel, G. A.; Van Boom, J.
H.; Overkleeft, H. S. J. Org. Chem. 2003, 68, 9406-9411. (p) Chery, F.;
Murphy, P. V. Tetrahedron Lett. 2004, 10, 2067-2069.
† University of Pennsylvania.
‡ Merck Research Laboratories.
(1) (a) Hirschmann, R.; Nicolaou, K. C.; Pietranico, S.; Salvino, J.; Leahy,
E. M.; Sprengeler, P. A.; Furst, G.; Smith, A. B., III; Strader, C. D.; Cascieri,
M. A.; Candelore, M. R.; Donaldson, C.; Vale, W.; Maechler, L. J. Am.
Chem. Soc. 1992, 114, 9217-9218. (b) Nicolaou, K. C.; Salvino, J. M.;
Raynor, K.; Pietranico, S.; Reisine, T.; Freidinger, R. M.; Hirschmann, R.
In Proceedings of the 11th American Peptide Symposium; Rivier, J. E.,
Marshall, G. R., Eds.; ESCOM: Leiden, The Netherlands, 1990; pp 881-
884.
(3) Prasad, V.; Birzin, E. T.; McVaugh, C. T.; van Rijn, R. D.; Rohrer,
S. P.; Chicchi, G.; Underwood, D. J.; Thornton, E. R.; Smith, A. B., III;
Hirschmann, R. J. Med. Chem. 2003, 46, 1858-1869.
10.1021/ol050119i CCC: $30.25
© 2005 American Chemical Society
Published on Web 02/23/2005

replaced by a methoxy substituent, affording (+)-7 without
loss of affinity at the SRIF receptors of transformed AtT-20
cells1a (Figure 2b).
Figure 1. Structure of somatostatin (SRIF-14),4 our initial designed
monosaccharide-based mimetic (+)-2, and our best ligand at the
hSST4 receptor (Ki ) 53 nM) (-)-3.3
The initial design of (+)-2 incorporated an ethylene rather
than a methylene linker between the C1 indole and the an-
omeric oxygen to avoid a gramine fragmentation.5 A reaction
of this type was indeed observed on treatment of (-)-4 with
6 M sodium hydroxide in ethanol at reflux (Scheme 1). All
of our ligands described herein, including the best ligand
(-)-3 (Figure 1), incorporate an ethylene linker.3
Figure 2. (a) Binding mode of (+)-2 (Ki ) 1490 nM). (b) The
established binding mode for ligand (+)-7, lacking an indole side
chain (Ki ) 2857 nM). (c) 1-Methoxy congeners of (+)-7, with a
C4 heterocyclic substituent, do not bind.3
Later, we observed only a modest loss in affinity for (+)-7
vis-a´-vis (+)-2 at the hSST4 receptor subtype (the highest
affinity human receptor subtype for our glycosides).6 We
explained the fact that a methoxy group can replace the Trp8-
mimicking indole side chain by suggesting that rotation of
the molecule places the 4-benzyl substituent into the Trp8
binding pocket of SRIF-14 (i.e., the benzene ring acts as an
indole surrogate; Figure 2b). Subsequently, we reported
structure-activity relationships that confirmed the above
proposition beyond reasonable doubt.7 We later termed this
phenomenon radial symmetry.8
Scheme 1. Gramine-Type Fragmentation of (-)-4
(4) (a) Rivier, J.; Brown, M.; Vale, W. Biochim. Biophys. Res. Commun.
1978, 7, 447-451. (b) Veber, D. F.; Holly, F. W.; Paleveda, W. J.; Nutt,
R. F.; Bergstrand, S. J.; Torchiana, M.; Glitzer, M. S.; Saperstein, R.;
Hirschmann, R. Proc. Natl. Acad. Sci. U.S.A. 1978, 75, 2636-2640.
(5) Snyder, H. R.; Eliel, E. L. J. Am. Chem. Soc. 1948, 70, 1703-1705.
(6) Hirschmann, R.; Nicolaou, K. C.; Pietranico, S.; Leahy, E. M.;
Salvino, J.; Arison, B. H.; Cichy, M. A.; Spoors, P. G.; Shakespeare, W.
C.; Sprengeler, P. A.; Hamley, P.; Smith, A. B., III; Reisine, T.; Raynor,
K.; Maechler, L.; Donaldson, C.; Vale, W.; Freidinger, R. M.; Cascieri, M.
A.; Strader, C. D. J. Am. Chem. Soc. 1993, 115, 12550-12558.
Early in the program, we serendipitously made the
surprising observation that the indole ring of (+)-2 can be
1122
Org. Lett., Vol. 7, No. 6, 2005

Calculations of electrostatic potentials of the aromatic
substituents next allowed us to explain why a benzene moiety
can replace the C1 indole of glycoside-based SRIF-14
peptidomimetics in the Trp8 binding pocket and why pyridine
and pyrazinyl rings cannot (Figure 2c).3 The electrostatic
potential is defined as the energy of interaction of a point
positive charge with the nuclei and electrons of the molecule
of interest.9 Thus, the electron distributions of the hetero-
cyclic aromatics at C4 in (+)-8a-c and (+)-9 lack a
significant negative potential in the region of the π-cloud
and thereby fail to bind the aromatic Trp8 binding pocket
(Figures 2 and 3).
relative ability of aromatic π-clouds to give effective π-donor
interactions. Although all aromatic rings show significant
π-clouds, the EP surfaces of pyridine and pyrazine differ
dramatically from those of indole and benzene.
Herein, we report extension of such EP calculations that
correctly predict that analogues based on (+)-2 can indeed
provide the π-cloud interactions required for significant
binding at hSST4. We computed the electrostatic potentials
of the relevant aromatic substituents at C1 of the glycosides
(+)-10-(+)-13 (Figure 4), using Spartan 6-311G**MP2
molecular orbital analysis.9
Figure 3. Congeners of desindole (+)-7 and their corresponding
EP calculations. Spartan 6-311G**MP2 MO analysis of aromatic
electrostatic potentials that involve interactions of a positive charge
not only with the π-cloud but also with all other electrons and nuclei
of the molecule. The mesh surfaces depicted are the surfaces upon
which the electrostatic potential (i.e., the attraction of the molecule
for a positive point charge) equals -10 kcal mol-1. Structures are
arranged in decreasing order of the measured distance from the
electrostatic potential surface to the molecular plane.9
Figure 4. Analogues of (+)-2 in conjunction with EP calculations.
Structures are arranged in decreasing order of the measured distance
from the electrostatic potential surface to the molecular plane.9
Electrostatic potential is a versatile property that can be
computed accurately because it does not require modeling
of binding sites or specific interactions. We choose to present
EP surfaces because they dramatically show the way in which
significant attraction for a positive charge extends well
beyond the van der Waals space-filling molecular surface.
These surfaces give an easily visualized comparison of the
The electrostatic potentials for indole, naphthalene, ben-
zothiophene, benzene, and benzofuran, illustrated in Figure
4, reveal that these surrogates of indole can provide for
effective π-donor interactions via, for example, an aromatic
edge to face interaction in the Trp8 binding pocket; a similar
interaction is not possible for pyridine and pyrazine sur-
rogates. The latter substituents have electron-deficient π-sys-
tems that differ spatially from the relatively electron-rich
π-systems of the other aromatic systems mentioned herein.3
The EP surfaces have shapes that depend on the structure
and architecture of the molecule. To provide a simplified
comparison of different molecules, we show in Table 1 the
maximum vertical distance of each surface perpendicular to
the molecular plane along with the dissociation constants of
all analogues. When an aromatic group has some confor-
mational flexibility, it is expected to orient itself so as to
place the maximum vertical π-donor interaction as closely
(7) (a) Hirschmann, R. Peptides 1996. Proceedings of the 24th European
Peptide Symposium; Ramage, R., Epton, R., Eds.; Mayflower Scientific:
Kingswinford, England, 1998; pp 3-17. (b) Hirschmann, R.; Hynes, J.,
Jr.; Cichy-Knight, M. A.; van Rijn, R. D.; Sprengeler, P. A.; Spoors, P. G.;
Shakespeare, W. C.; Pietranico-Cole, S.; Barbosa, J.; Yao, W.; Rohrer, S.
P.; Smith, A. B., III. J. Med. Chem. 1998, 41, 1382-1391.
(8) (a) Liu, J.; Underwood, D. J.; Casieri, M. A.; Rohrer, S. P.; Cantin,
L. D.; Chicchi, G.; Smith, A. B., III; Hirschmann, R. J. Med. Chem. 2000,
43, 3827-3831. (b) McVaugh, C. T.; Han, G.; Chicchi, G.; Underwood,
D. J.; Prasad, V.; Liu, J.; Kurtz, M.; Birzin, E. T.; Cascieri, M. A.; Rohrer,
S.; Smith, A. B., III; Hirschmann, R. Manuscript in preparation.
(9) (a) MacSpartan Pro/Macspartan Plus (Version 2.0) User’s Guide;
Wavefunction: Irvine, CA, 2000; pp 64-69. (b) Hehre, W. J.; Yu, J.;
Klunzinger, P. E.; Lou, L. A Brief Guide to Molecular Mechanics and
Quantum Chemical Calculations; Wavefunction: Irvine, CA, 1988; pp 138-
139. (c) Thornton, E. R. University of Pennsylvania. Unpublished results.
Org. Lett., Vol. 7, No. 6, 2005
1123

as possible to the group with which it interacts in the binding
pocket. Molecules with larger vertical distances (giving -10
kcal/mol interaction with a point positive charge) are thus
expected to have stronger π-donor interactions at any distance
from the molecular plane.10 Consequently, binding abilities
are expected to have a direct correlation to the vertical extent
of the EP surface.
install the lysine-mimicking side chain; yields in general were
good. Stereospecific epoxidation of the enol ether (-)-15
with dimethyldioxirane was then followed by treatment with
a series of alcohols in the presence of ZnCl2 to generate
compounds 16a-d.1a Benzylation of the resulting intermedi-
ates followed by Staudinger reduction of the azides com-
pleted construction of analogues (+)-10-(+)-13 in good
yield.
Table 1. EP-Calculated Vertical Distances and Dissociation
Constants of the Designed Congeners of (+)-2 at hSST4
Scheme 2. Synthesis of Congeners of (+)-2
compound EP calculation vertical distance,a Å Ki (nM), n ) 3b
(+)-2
3.61
3.12
3.05
3.02
2.96
1490 ( 749
1415 ( 633
1613 ( 494
1557 ( 200
1671 ( 220
(+)-10
(+)-11
(+)-12
(+)-13
1 Maximum distance from EP surface to molecular plane. bn ) 3 except
for (+)-2, where n ) 13.
Binding assays for (+)-2 and (+)-10-(+)-13 at the hSST4
receptor (Table 1) were performed using (3-125I-Tyr11)-SRIF-
14 as the radioligand. The Packard Unifilter assay for SRIF
subtype receptor binding was employed in these experiments
as described by Birzin and Rohrer.11 Importantly, the
analogues of (+)-2 do bind the receptor as predicted by the
EP calculations. In addition, the identical binding affinities
of (+)-2 and (+)-12 provide additional validation of our
rationalization for the observation that (+)-7 binds the SRIF
receptors.1a
The relative dissociation constants for (+)-2, (+)-10, (+)-
11, (+)-12, and (+)-13, however, do not parallel exactly the
corresponding electrostatic potentials. We attribute these
discrepancies to the large error bars in the Kis that are always
seen with relatively weak ligands in the binding assays. It is
pleasing, however, that the calculations of the EPs for these
ligands correctly predict that they would bind the receptor.
The syntheses of analogues (+)-10-(+)-13 began with
the known intermediate alcohol (-)-14 (Scheme 2),1a which
was deprotonated using sodium hydride, followed by cou-
pling with the triflate derived from 5-azido-1-pentanol, to
Taken together, the data described herein support the
previously reported3 proposition that the electrostatic poten-
tials provide information necessary to predict whether an
aromatic ring system possesses the π-cloud required for
interaction with the side chains of proteins typified by the
Trp8 receptor binding pocket of SRIF-14.
Acknowledgment. Financial support was provided by the
National Institutes of Health through Grant GM-41821. We
also thank Merck Research Laboratories for an unrestricted
grant.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for all new compounds and
electrostatic potential diagrams. This material is available
free of charge via the Internet at http://pubs.acs.org.
(10) This EP surface of -10 kcal/mol for interaction with a point positive
charge is also exactly the surface for attraction of a +0.1 partial positive
charge at an energy of -1 kcal/mol, expected to be relevant to aromatic-
aromatic interactions.
(11) Birzin, E. T.; Rohrer, S. P. Anal. Biochem. 2002, 159-166.
OL050119I
1124
Org. Lett., Vol. 7, No. 6, 2005