Building thermodynamic combinatorial libraries of quinine macrocycles

Article abstract of DOI:10.1039/a703291i

Thermodynamically-controlled transesterification of a predisposed cinchonidine building block with a more flexible extended quinine monomer leads to a combinatorial mixture of 11 macrocyclic receptors which is analysed by electrospray mass spectrometry; s

Full text of DOI:10.1039/a703291i

View Article Online / Journal Homepage / Table of Contents for this issue
Building thermodynamic combinatorial libraries of quinine macrocycles
Stuart J. Rowan and Jeremy K. M. Sanders*
Cambridge Centre for Molecular Recognition, University Chemical Laboratory, Lensfield Road, Cambridge, UK CB2 1EW
Thermodynamically-controlled transesterification of a pre-
disposed cinchonidine building block with a more flexible
extended quinine monomer leads to a combinatorial mixture
of 11 macrocyclic receptors which is analysed by electro-
spray mass spectrometry; similar results with alkaloid–
cholate mixtures demonstrate the generality of the ap-
proach.
Under thermodynamic control, mixtures of cinchonidine
monomer 1 and preorganised xanthenes self-sort to give good
yields of cyclic trimer and dimer respectively because the
homo-products are stabilised.3 However, the relaxed constraints
of the extended monomer 2 combined with predisposed 1 lead
to quite a different outcome: thermodynamic transesterification
of 1 and 2 under the usual conditions led to a combinatorial
library of macrocycles, as observed by electrospray mass
spectrometry§ (Fig. 1). The mass spectrum contains two of the
three possible dimers, all possible trimers and small amounts of
all the possible tetramers. It is unwise to draw precise
quantitative conclusions from peak intensities since individual
species have different inherent detectabilities in electrospray
mass spectra, but there are systematic and interpretable
deviations from the expected statistical distribution of
hetero:homo oligomers. The extended quinine homo-dimer is
seemingly favoured over the hetero dimer, which is not
surprising given that 1 cannot form homo dimer at all. The
trimers show the expected weighting towards the cinchonidine
homo trimer but the more promiscuous monomer 2 gives access
to two new hetero trimers. To a first approximation, it appears
that the tetramers are formed in statistical proportions.
We have recently shown that thermodynamically-controlled
transesterification can lead to efficient synthesis of oligomeric
macrocycles,1,2 and that when the monomeric building blocks
are suitably predisposed a single product may dominate even
when other oligomers are kinetically accessible.3 For example,
the
cinchonidine
monomer
1
(HO–Cc–OMe)
is
O
O
MeO2C
MeO
11
N
N
HO
HO
MeO
HO
X
2′
N
N
1 HO–Cc–OMe
2 HO–Ce–OMe
N
N
predisposed to give almost exclusively the cyclic trimer under
thermodynamic control. We now report that the extended
quinine monomer 2 (HO–Ce–OMe), in which this predisposi-
tion is slightly relaxed, gives access to a wider range of homo-
macrocycles; more importantly it allows the development of
thermodynamically-controlled receptor libraries by combina-
tion with other building blocks such as 1.
ButMe2SiO
ButMe2SiO
MeO
i–iii
iv
HO–Ce–OMe
MeO
N
N
3
4 X = OTs
5 X = I
6 X = p-OC6H4CO2Me
In order to relax the predisposition of 1, we elected to extend
the cinchonidine monomer at the 11-position. The chosen
extension, a methyl 4-hydroxybenzoate moiety, serves several
functions: it is compatible with our transesterification condi-
tions; it programs into the molecule added length to increase the
size of any cavity formed and allow access to cyclic dimer;
introduces a little extra flexibility; and is still rigid enough to
prevent formation of cyclic lactone monomer. Synthesis of 2
started from the previously prepared 11-alcohol 3.2 This was
converted into the tosylate 4 with tosyl chloride and triethyla-
mine (77% yield), which on heating with NaI in acetone gave
9-O-tert-butyldimethylsilyl-10,11-dihydro-11-iodo quinine 5 in
86% yield. Compound 5 was converted into 6 (75%) by stirring
with methyl 4-hydroxybenzoate, 18-crown-6 and K2CO3, and
removal of the ButMe2Si group using Bu4NF–THF gave the
extended monomer 2 in 78% yield (Scheme 1).
The result of submitting monomer 2 to transesterification
under thermodynamic control2 was a mixture of mainly cyclic
dimer (Ce2):trimer (Ce3):tetramer (Ce4) and higher oligomeric
quinines, with a mass ratio of 30:41:29 respectively† (Scheme
2). Even though there is still a distinct preference for cyclic
trimer,‡ the extension unit has allowed access to cyclic dimer
(not previously seen even in kinetic cyclisations of 14) and
tetramer under thermodynamic control, demonstrating the
expected relaxation of predisposition.
Scheme 1 Reagents and conditons: i, TsCl, Et3N, room temp., 4 h; ii, NaI,
acetone, reflux, 16 h; iii, 4-HOC6H4CO2Me, 18-crown-6, K2CO3, room
temp., 16 h; iv, Bu4NF, THF, room temp., 4 h.
N
O
MeO
O
O
i
HO–Ce–OMe
N
N
O
O
O
OMe
N
Ce2
30%
Ce3
Ce4 + other
oligomers
29%
41%
Scheme 2 Reagents and conditions: i, KOMe, 18-crown-6, toluene, reflux,
1 h.
Chem. Commun., 1997
1407

View Article Online
To further examine the idea of relaxation of predisposition
monomers 1 and 2 were individually mixed with the cholate
monomer 7, which has enough flexibility to allow access to a
Selection
isolation and
release
O
OMe
But
But
R1
R2
Fig 2 Schematic representation of a thermodynamic selection approach to
receptors
O
CO2Me
HO
HO
obtained quickly from a small number of monomers. The way is
now open for creating thermodynamically controlled macro-
cyclic receptor libraries (Fig. 2), with the possibility of
thermodynamic templating to bias the library towards robust,
covalent receptors with good binding properties.5-7 If the guest
is tethered in some way it should also be possible to isolate the
receptors from an otherwise hopelessly complex mixture.
We thank Merck, Sharp and Dohme and the EPSRC for
generous financial support. We also thank Drs P. A. Brady and
D. G. Hamilton for cholate and xanthene samples.
7
8
R1 = OBn
R2 = OCH2OCH2CH2OMe
range of cyclic oligomers.1 Transesterification produced a
cocktail of all the possible hetero oligomers of both 1 and 2 with
7, which is consistent with the idea that if there is enough
flexibility in one component then that is generally sufficient to
give a range of macrocycles. We had already demonstrated that
mixing of 1 with the highly preorganised xanthene monomer 8
leads to self-sorting.3 To examine how powerfully preorganised
8 is we allowed it to react in the presence of extended quinine
monomer 2: the result was predominantly self-sorting, indicat-
ing that preorganised monomers such as 8 find it difficult to
form mixtures with monomers that do not have the correct ‘bite
size’ or flexibility.
These results demonstrate that it is possible to generate
combinatorial libraries of macrocyclic receptors even when one
of the building blocks is predisposed to produce just a single
product. By mixing only two monomers we observe 11 cyclic
products, showing how large libraries of macrocycles can be
Footnotes
* E-mail: jkms@cam.ac.uk
† Determined by 400 MHz 1H NMR spectroscopy of the reaction mixture
after work up and dissolution in CDCl3: The chemical shifts for H2A were:
dimer (Ce2: d 8.77), trimer (Ce3: d 8.74) and higher oligomers (Ce4+: d
8.70–8.73).
‡ In the absence of significant enthalpic differences in the stability of
oligomers, simple entropic considerations lead to an expected order of
abundance monomer > dimer > trimer > tetramer etc.
§ ES-MS were obtained on a VG BioQ triple quadrupole apparatus (VG Bio
Tech Ltd, Altrincham, UK). The electrospray source was heated to 70 °C
and the sampling cone voltage (Vc) was 105 V. Samples were introduced
into the mass spectrometer source with an LC pump (Shimadzu LC-9A LC
pump) at a flow rate of 4 ml min21 of MeCN–H2O (1:1). Calibration was
performed using protonated horse myoglobin. Scanning was performed
from m/z 200 to 2000 in 10 s. The data system was operated as a
multichannel analyser, and several scans were summed to obtain the final
spectrum. No hetero oligomers are detected in the electrospray mass spectra
of control mixtures of homo oligomers.
100
Ce
2
Cc
3
Cc Ce
2
References
CcCe
2
1 P. A. Brady, R. P. Bonar-Law, S. J. Rowan, C. J. Suckling and J. K. M.
Sanders, Chem. Commun., 1996, 319.
2 S. J. Rowan, P. A. Brady and J. K. M. Sanders, Angew. Chem., Int. Ed.
Engl., 1996, 35, 2143.
CcCe
3 S. J. Rowan, D. G. Hamilton, P. A. Brady and J. K. M. Sanders, J. Am.
Chem. Soc., 1997, 119, 2578.
4 S. J. Rowan, P. A. Brady and J. K. M. Sanders, Tetrahedron Lett., 1996,
37, 6013.
Ce
3
×
Cc
×
4
×
Cc Ce
3
Cc Ce
×
2
2
CcCe
Ce
4
3
×
0
5 W. C. Still, P. Hauck and D. Kempf, Tetrahedron Lett., 1987, 28,
2817.
6 A. V. Eliseev and M. I. Nelen, J. Am. Chem. Soc., 1997, 119, 1147.
7 I. Huc and J.-M. Lehn, Proc. Natl. Acad. Sci. USA, 1997, 94, 2106 and
references cited therein.
800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800
m / z
Figure 1 Electrospray mass spectrum of the thermodynamic cyclisation
reaction mixture of 1 (HO–Cc–OMe) and 2 (HO–Ce–OMe). Peaks marked
(3) are NH4+ adducts of molecular ions.
Received in Liverpool, UK, 13th May 1997; 7/03291I
1408
Chem. Commun., 1997