Formation of a macrobicyclic tris(disulfide) by molecular self-assembly

Full text of DOI:10.1021/jo982166c

334
J . Org. Chem. 1999, 64, 334-335
F or m a tion of a Ma cr obicyclic Tr is(d isu lfid e)
by Molecu la r Self-Assem bly
Suk-Wah Tam-Chang,* J effrey S. Stehouwer, and
J insong Hao
Department of Chemistry, University of Nevada,
Reno, Nevada 89557
Received October 28, 1998
Self-assembly is a fundamentally important process em-
ployed by nature in the construction of supramolecular
biological systems with a minimal expenditure of resources
and energy. Extensive effort has been spent in designing
components which mimic natural systems by undergoing
molecular self-organization through coordination of metals1
or selective noncovalent interactions such as hydrogen
bonding.2 In contrast, the area of self-assembly with covalent
modification is relatively little explored.3,4 One of the more
intriguing examples comes from nature: the thiol-disulfide
interchange reaction which is important in stabilizing
protein structure by disulfide bridge formation.4,5 This
reaction has potential applications in the self-assembly of
robust supramolecular arrays with proof-reading to remove
errors by its remarkable ability to effect the reversible
formation and cleavage of strong, covalent S-S bonds at
room temperature.6 The ability to control the formation of
disulfide bonds can be applied to the design of preorganized
monomers which may self-assemble to generate organized
entities with interesting architectures and properties. We
report herein an enhancement of the thermodynamic stabil-
ity of a tris(disulfide) dimer by molecular design and the
first example of the formation of tris(disulfide) dimer 1 from
its trithiol monomer 2 under equilibrium conditions (Scheme
1).
Previous attempts to form a macrocyclic dimeric bis-
(disulfide) or tris(disulfide) under equilibrium conditions7
were met with difficulty due to the unfavorable enthalpy
term originating in CSSC torsion angle strain. In addition,
the thiol groups in the monomer are not preorganized in a
geometry that resembles the disulfides in the dimer result-
ing in an unfavorable entropy change upon dimer formation.
Our approaches in molecular design were to preorganize
the starting trithiol into a conformation resembling that of
the tris(disulfide) and to eliminate as many rotational
degrees of freedom as possible, while allowing the formation
of a strain free CSSC dihedral angle of about 90°. We thus
F igu r e 1. Study of the thiol-disulfide interchange equilibrium
(1 mM MEox) at 298 K among trithiol 2, dimer 1, and oligomer 3
in DMSO-d6 (under argon) by 500 MHz 1H NMR spectroscopy after
17 days.
designed and synthesized trithiol 2 as shown in Scheme 2.
Nonbonded steric effects between adjacent groups on the
aromatic ring8 are expected to force the ethyl groups and
the thiol-terminated substituents to lie on opposite sides of
the ring. In this conformation, the thiol groups are pre-
organized into a geometry that resembles the conformation
of the monomeric units in dimer 1. Amide groups which have
a lower rotational degree of freedom than alkyl chains are
introduced (7 f 8) so as to constrain the CH2SH groups to
lie away from the aromatic ring, thereby enabling the
formation of disulfide bonds without CSSC angle strain.
Under high dilution kinetic conditions, trithiol 2 is
oxidized by iodine to tris(disulfide) dimer 1 (in 84% yield
after purification). Oligomer 3, which is insoluble in CHCl3,
was isolated as a byproduct in about 10% yield. Evidence
for the formation of dimer 1 includes the following: (1) a
dimer parent ion peak ([M + H+]) at m/z 937.2973 in the
1
HRMS-FAB mass spectrum, and (2) sharp peaks in the H
NMR spectrum in CDCl3 but no thiol proton peak (a trimer
would have shown thiol peaks). In contrast, oligomer 3
showed no peak below m/z 2000 in the HRMS-FAB mass
spectrum and broad peaks in the 1H NMR spectrum in
DMSO-d6. Dimer 1 is stable in solution up to 50 °C. A 1 mM
solution of 1 in DMSO-d6 showed no change in 1H NMR
signals after heating at 50 °C for 6 days.
(1) Stang, P. J .; Olenyuk, B. Acc. Chem. Res. 1997, 30, 502.
(2) (a) Tecilla, P.; J ubian, V.; Hamilton, A. D. Tetrahedron 1995, 51, 435.
(b) Yang, J .; Fan, E.; Geib, S. J .; Hamilton, A. D. J . Am. Chem. Soc. 1993,
115, 5314. (c) Lehn, J .-M. Pure Appl. Chem. 1994, 66, 1961. (d) Timmerman,
P.; Vreekamp, R. H.; Hulst, R.; Verboom, W.; Reinhoudt, D. N.; Rissanen,
K.; Udachin, K. A.; Ripmeester, J . Chem. Eur. J . 1997, 3, 1823. (e)
Whitesides, G. M.; Simanek, E. E.; Mathias, J . P.; Seto, C. T.; Chin, D. N.;
Mamen, M.; Gordon, D. M. Acc. Chem. Res. 1995, 28, 37. (f) Conn, M. M.;
Rebek, J ., J r. Chem. Rev. 1997, 97, 1647. (g) Lawrence, D. S.; J iang, T.;
Levett, M. Chem. Rev. 1995, 95, 2229.
Under equilibrium conditions in the presence of 2-hy-
droxyethyl disulfide (MEox), trithiol 2 exists in equilibrium
with dimer 1, oligomer 3, and unsymmetrical disulfide9
(Figure 1). Starting from a solution of 2 (1 mM) and MEox
(1 mM) in DMSO-d6 under argon in a sealed NMR tube,
equilibrium was achieved in several weeks.10 Integration of
1
the NH peaks in the H NMR spectrum shows that 30% of
the amide hydrogens are present as dimer 1. Calculations
(3) (a) Philp, D.; Stoddart, J . F. Angew. Chem., Int. Ed. Engl. 1996, 35,
1154. (b) Rowan, S. J .; Hamilton, D. G.; Brady, P. A.; Sanders, J . K. M. J .
Am. Chem. Soc. 1997, 119, 2578.
based on 1 mM MEox indicate that the theoretical amount
(4) Lindsey, J . S. New J . Chem. 1991, 15, 153.
(5) (a) Ziegler, D. M. Annu. Rev. Biochem. 1985, 54, 305. (b) Houk, J .;
Singh, R.; Whitesides, G. M. Methods Enzymol. 1987, 143, 129. (c) Gilbert,
H. F. Methods Enzymol. 1995, 252, 8.
(6) (a) Lees, W. J .; Whitesides, G. M. J . Org. Chem. 1993, 58, 642. (b)
Rosenfield, R. E. J r.; Parthasarathy, R.; Dunitz, J . D. J . Am. Chem. Soc.
1977, 99, 4860. (c) Pappas, J . A. J . Chem. Soc., Perkin Trans. 2 1979, 67.
(7) (a) Singh, R.; Whitesides, G. M. J . Am. Chem. Soc. 1990, 112, 1190.
(b) Houk, J .; Whitesides, G. M. J . Am. Chem. Soc. 1987, 109, 6825. (c) Houk,
J .; Whitesides, G. M. Tetrahedron Lett. 1989, 45, 91. (d) Burns, J . A.;
Whitesides, G. M. J . Am. Chem. Soc. 1990, 112, 6296.
(8) (a) Vogtle, F.; Weber, E. Angew. Chem., Int. Ed. Engl. 1974, 13, 814.
(b) Stack, T. D. P.; Hou, Z.; Raymond, K. N. J . Am. Chem. Soc. 1993, 115,
6466. (c) Metzger, A.; Lynch, V. M.; Anslyn, E. V. Angew. Chem., Int. Ed.
Engl. 1997, 36, 862. (d) Bisson, A. P.; Lynch, V. M.; Monahan, M.-K., C.;
Anslyn, E. V. Angew. Chem., Int. Ed. Engl. 1997, 36, 2340.
(9) The presence of unsymmetrical disulfide is due to the formation of
disulfide bonds between mercaptoethanol and the free thiol ends of monomer
2 and oligomer 3.
(10) Equilibrium can also be established by mixing mercaptoethanol with
a solution of dimer 1 or oligomer 3.
10.1021/jo982166c CCC: $18.00 © 1999 American Chemical Society
Published on Web 01/05/1999

Communications
J . Org. Chem., Vol. 64, No. 2, 1999 335
Sch em e 1
Sch em e 2
Sch em e 3
because the compounds do not significantly absorb in the
UV-vis spectral region.
Trithiol 10 (Scheme 3) was prepared and used as a control
to test the effect of CSSC dihedral angle strain on dimer
formation. CPK modeling of dimer 11 showed that it is
highly strained. The inability to form strain-free CSSC
dihedral angles presumably discourages the formation of 11.
Attempts to prepare 11 from trithiol 10 under conditions
identical to those used to prepare tris(disulfide) 1 yielded
only oligomeric product 12.
In conclusion, we have demonstrated that by appropriate
molecular design it is possible to enhance the stability of a
dimeric tris(disulfide) which we have observed in equilibri-
um with the oligomer. The result is a macrobicyclic capsule
having a cavity size large enough to accommodate small
guest molecules. We are currently working toward increas-
ing the stability of the self-assembled dimer through hydro-
gen bonds and/or electrostatic interactions as well as specific
guest encapsulation within the cavity.
of disulfide bond formation is 66%.11 Similar studies using
1.5 mM (stoichiometric) and 3 mM (excess) of MEox result
in 27% and 23%, respectively, of the amide hydrogens
present as dimer 1.12 We could not determine the exact
concentration of the oligomer, and therefore its relative
stability compared with the dimer because the amount of
monomeric units existing as trithiol 2, oligomer, or unsym-
metrical disulfide could not be determined precisely due to
overlapping peaks for their NH protons and overlapping
peaks for their CH2NH protons. In addition, the exact
formula of the oligomer is not known. An attempt to
determine product percentages with HPLC was unsuccessful
Ack n ow led gm en t. This research is supported by the
University of Nevada, Reno. We are grateful to Professor
George M. Whitesides (Harvard University) for his helpful
advice.
(11) The total theoretical amount of free thiol ends remaining in monomer
2 and oligomer 3 is 33%.
(12) The decrease in dimer formation is presumably due to competing
formation unsymmetrical disulfide. Attempts to minimize unsymmetrical
disulfide formation by using cyclic disulfide (e.g. trans-1,2-dithiane-4,5-diol)
instead of 2-hydroxyethyl disulfide (MEox) to induce equilibrium were
unsuccessful due to the weak oxidative strength of the cyclic disulfide.
Su p p or tin g In for m a tion Ava ila ble: Spectroscopic data for
compounds 1-3, and experimental procedures for equilibrium
studies.
J O982166C