Conformer independent heterodimerisation of linear arrays using three hydrogen bonds

Article abstract of DOI:10.1039/b712603d

5-Membered heterocycles are employed to give a conformer independent DDA array of hydrogen bonds, resulting in enhanced binding affinity to a complementary AAD array in comparison to a DDA array employing a 6-membered ring. The Royal Society of Chemistry.

Full text of DOI:10.1039/b712603d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Conformer independent heterodimerisation of linear arrays using three
hydrogen bonds{
Andrea M. McGhee,a Colin Kilnera and Andrew J. Wilson*ab
Received (in Cambridge, UK) 10th September 2007, Accepted 17th October 2007
First published as an Advance Article on the web 29th October 2007
DOI: 10.1039/b712603d
intramolecular hydrogen bond, although two different tautomers
are possible only one of which presents the required AAD array.
Ureidoimidazole 1a18 was synthesised in one-pot on gram scale
without the need for chromatographic purification (see ESI{).
Crystals were grown from the slow evaporation of a methanolic
5-Membered heterocycles are employed to give a conformer
independent DDA array of hydrogen bonds, resulting in
enhanced binding affinity to a complementary AAD array in
comparison to a DDA array employing a 6-membered ring.
The design and synthesis of linear arrays of hydrogen bonds is a
key area of research1,2 since these arrays represent the building
blocks for non-covalent synthesis of stimuli responsive assemblies
and polymers.3,4 Most current designs use linear arrays comprising
four hydrogen bonds because of the enhanced strength of
association that may be achieved,5 although high-affinity triply
hydrogen bonded systems have been observed.6 Designing such
systems is challenging since intramolecular hydrogen bonding,7
pre-organisation,8 secondary interactions,9 tautomerisation10 and
electronic substituent effects11 can all have effects upon the
strength and fidelity12 of recognition; with increasing numbers of
hydrogen bonds these factors become difficult to mitigate against
whilst maintaining synthetic accessibility. We became interested in
this problem and observed that most systems that have been
studied to date employ 6-membered heterocycles within the
hydrogen bonding framework. Herein we describe how 5-mem-
bered rings can be useful components of linear arrays, leading to
conformer independent binding when compared against their
6-membered counterparts.
solution of 1a (Fig. 2). An intramolecular hydrogen bond is
…
present as proposed, however only the conformer with an NH
O
hydrogen bond is observed. This behaviour is not maintained in
DMSO-d6 solution: 2 broad NH signals, one integrating to two
protons (presumably the urea NHs) and the other to one
(presumably the imidazole NH) are indicative of interconversion
between different conformations. The crystal structure also
suggests that the molecule may tend towards intermolecular self-
association; the imidazole engages in bifurcated H-bonding to the
urea fragment and the imidazole NH participates in intermolecular
H-bonding to the urea carbonyl. We were unable to study the
dimerisation of 1a in CDCl3 due to poor solubility and therefore
synthesised 1b in 3 steps with a solubilising tert-butyl group by
extending previously reported methodology for the synthesis of
2-aminoimidazoles (see ESI{).19 For 1b in CDCl3 at room
temperature, a single set of sharp signals for the tert-butyl group
and the aromatic protons is observed along with 2 broad NH
signals, as is observed for 1a in DMSO-d6. This suggests fast
interconversion between all of the possible conformers/tautomers
in addition to any intermolecular association that is occurring.
Nevertheless, at room temperature, small shifts in the aromatic
resonances of 1b can be observed upon dilution of a sample in
Our strategy focuses on intramolecular hydrogen bonding.
Pyridylureas e.g. 2 are attractive DDA building blocks because of
their ease of synthesis, yet they exhibit poor association constants
with complementary AAD arrays since an intramolecular hydro-
gen bond must be broken at an entropic price in order to form
three intermolecular non-covalent contacts.7,13 We hypothesised
that this problem could be negated by use of an imidazole14,15 as in
1a and 1b rather than a pyridine heterocycle (Fig. 1a). The
intramolecular hydrogen bond would still be present in such a
system, however it should act to pre-organise it, as in either of two
possible conformers a DDA array would be presented (this
concept is similar in nature to degenerate tautomerism).16,17 As a
complementary AAD array we selected amidoisocytosine 3
(Fig. 1b). Surprisingly these synthons have not, to the best of
our knowledge, been studied before. In contrast to amidocytosines,
amidoisocytosines should be pre-organised as a consequence of an
aSchool of Chemistry, University of Leeds, Woodhouse Lane, Leeds, UK
LS2 9JT. E-mail: A.J.Wilson@leeds.ac.uk; Fax: +44 (0)113 3431409;
Tel: +44 (0)113 3436565
bAstbury Centre for Structural Molecular Biology, University of Leeds,
Woodhouse Lane, Leeds, UK LS2 9JT
{ Electronic supplementary information (ESI) available: Full details of
crystallographic structure determination, synthetic procedures, titration
details and curve fitting. See DOI: 10.1039/b712603d
Fig. 1 DDA and AAD arrays used in this study.
344 | Chem. Commun., 2008, 344–346
This journal is ß The Royal Society of Chemistry 2008

View Article Online
weak interaction. Curve fitting gave a dimerisation constant of
4 M21 (¡1 M21) (see ESI{).
We next turned our attention to heterodimerisation. Upon
titration of a solution of 1b into 3 (at 6 mM), significantly larger
complexation induced shifts in the aromatic resonances of 3 were
observed than had been observed for the dilution study (Fig. 4a).
Although the NH resonances failed to sharpen, we were still able
to derive an association constant of 8400 M21 (¡1000 M21) by
curve fitting20 to a 1 : 1 association model that accounts for
homodimerisation of both components (Fig. 4b). The stoichio-
metry was confirmed by JOB plot (Fig. 4b inset) and a similar
value for Ka was obtained when the titration was repeated at 2 mM
(see ESI{). For both experiments, it seems clear that the simple
1 : 1 complex is not the only species present—at least in the early
Fig. 2 X-Ray crystal structure packing diagram of compound 1a shown
in stick format (carbon in grey, hydrogen in white, oxygen in red and
˚
nitrogen in blue). Key distances: H4–O7 2.31 A (intramolecular), H4–O7
˚
˚
˚
2.11 A, N1–H6 2.09 A, N1–H8 2.25 A, (intermolecular).{
CDCl3. This indicates any self-association is weak and that the
formation of supramolecular polymers in solution that mirror the
solid state structure is unlikely. Curve fitting to a dimerisation
model using the HypNMR program20 gives a dimerisation
constant of 11 M21 (¡2 M21) (see ESI{). For comparison we
also synthesised pyridyl urea 2 (see ESI{). This compound gave
sharp resonances in CDCl3 indicative of a single well defined
intramolecularly hydrogen-bonded conformation in solution. A
dilution study and curve fitting gave a dimerisation constant of
56 M21 (see ESI{) in line with previous observations.13
Amidoisocytosine 3 was synthesised in one step on gram scale
(see ESI{). An X-ray diffraction study{ on single crystals grown
from the slow evaporation of a methanolic solution in chloroform
revealed the presence of two similar molecules of 3 in the
asymmetric unit (Fig. 3). For both, the proposed intramolecular
hydrogen bond is present, although the undesired tautomer is
observed. At this stage it is not clear if this arises due to crystal
packing or if this is the lowest energy conformer of 3. However,
acylation has a dramatic effect: isocytosine itself adopts a
heterodimeric complex comprising one 6[1H]pyrimidinone tauto-
mer and one 4[1H]pyrimidinone tautomer.21 In contrast, only
weak intermolecular hydrogen bonds are observed in the structure
of 3. This is mirrored in CDCl3: in this case only one broad signal
is observed for the NH resonances and upon dilution only small
changes are observed for the aromatic resonances indicative of a
Fig. 3 X-Ray crystal structure packing diagram of compound 3 shown
in stick format (carbon in grey, hydrogen in white, oxygen in red and
Fig. 4 (a) Partial 1H NMR spectra (500 MHz, CDCl3) of 3 upon
addition of 1b. (b) Chemical shift change of Hc upon addition of 1b to a
6 mM solution of 3 together; the line shows the best fit curve obtained via
curve fitting (inset: JOB plot).
˚
nitrogen in blue). Key distances: O2–H16 1.94(5) A, O22–H36 1.99(5) A
˚
˚
(intramolecular), H1–O35 2.11(6) A, N12–H36 2.54(5) A, O15–H21
˚
˚
2.07 (4) A, H16–N32 2.59(4) A (intermolecular).
˚
This journal is ß The Royal Society of Chemistry 2008
Chem. Commun., 2008, 344–346 | 345

View Article Online
3
U = 1335.37(5) A , T = 150(2) K, space group P1, Z = 2, m = 0.388 mm
21
˚
¯
,
part of the titration. The resonance for Hc becomes broad in the
titration, whilst the early part of the JOB plot is curved. This
suggests the presence of higher order structures; nevertheless the
1 : 1 complex is clearly the dominant species and stable. In contrast
upon titration of 3 into 2, despite larger complexation induced
shifts for the NH resonances, the variation with concentration was
much less dramatic indicative of much weaker association. Curve
fitting20 (allowing for self-dimerisation of both components) gave
an association constant of 84 M21 (¡10 M21), consistent with
previous observations for related compounds.13 Thus, exchanging
the pyridine group in 2 for the imidazole group as in 1 results in
nearly 3 orders of magnitude difference in binding affinity to
complementary partners.
˚
l = 0.71073 A [Mo-Ka], 25222 reflections measured, 5222 unique (Rint
=
0.1042), 3827 observed (I . 2s(I)). The final R1 was 0.0799 (observed
reflections 0.1071) and wR(F2) was 0.2069 (all data 0.2314) for
370 parameters. CCDC 660530 and 660531. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b712603d
1 A. J. Wilson, Soft Matter, 2007, 3, 409.
2 S. C. Zimmerman and P. S. Corbin, Struct. Bonding, 2000, 96, 63.
3 L. Prins, D. N. Reinhoudt and P. Timmerman, Angew. Chem., Int. Ed.,
2001, 40, 2382.
4 L. Brunsveld, B. J. B. Folmer, E. W. Meijer and R. P. Sijbesma, Chem.
Rev., 2001, 101, 4071.
5 R. P. Sijbesma and E. W. Meijer, Chem. Commun., 2003, 5.
6 S. Djurdejevic, D. A. Leigh, H. McNab, S. Parsons, G. Teobaldi and
F. Zerbetto, J. Am. Chem. Soc., 2007, 129, 476.
7 C.-H. Chien, M.-K. Leung, J.-K. Su, G.-H. Li, Y.-H. Liu and Y. Wang,
J. Org. Chem., 2004, 69, 1866.
8 P. Prabhakaran, V. G. Puranik and G. J. Sanjayan, J. Org. Chem.,
2005, 70, 10067.
9 W. L. Jorgensen and J. Pranata, J. Am. Chem. Soc., 1990, 112, 2008.
10 T. Park, E. M. Todd, S. Nakashima and S. C. Zimmerman, J. Am.
Chem. Soc., 2005, 127, 18133.
11 V. G. H. Lafitte, A. E. Aliev, H. C. Hailes, K. Bala and P. Golding,
J. Org. Chem., 2005, 70, 2701.
12 E. M. Todd, J. R. Quinn, T. Park and S. C. Zimmerman, Isr. J. Chem.,
2005, 45, 381.
In summary we have shown that conformer independent linear
arrays of hydrogen bonds can lead to significantly enhanced
association constants with complementary hydrogen bonding
partners. This concept represents a useful tool to employ in the
design of linear arrays of H-bonds. Our own studies will now focus
on a detailed structural investigation of this system and upon
enhancing affinity.
We would like to thank the EPSRC (EP/C007638/1) for
financial support of this research. We would also like to thank
Tanya Marinko-Covell for MS and Martin Huscroft for
combustion analyses.
13 P. S. Corbin, S. C. Zimmerman, P. A. Thiessen, N. A. Hawryluk and
T. J. Murray, J. Am. Chem. Soc., 2001, 123, 10475.
14 S. Watanabe, O. Onogawa, Y. Komatsu and K. Yoshida, J. Am. Chem.
Soc., 1998, 120, 229.
15 M. G. Fisher, P. A. Gale and M. E. Light, New J. Chem., 2007, 31,
1583.
16 P. S. Corbin and S. C. Zimmerman, J. Am. Chem. Soc., 1998, 120, 9710.
17 P. K. Baruah, R. Gonnade, U. D. Phalgune and G. J. Sanjayan, J. Org.
Chem., 2005, 70, 6461.
Notes and references
{ Crystal data for C10H10N4O, M = 202.22, crystal size 0.36 x 0.19 x 0.05,
˚
monoclinic, a = 11.7156(3), b = 6.0176(2), c = 13.3965(4) A, b =
3
˚
18 C. J. Paget, K. Kisner, R. L. Stone and D. C. DeLong, J. Med. Chem.,
1969, 12, 1010.
92.3690(10)u, U = 943.64(5) A , T = 150(2) K, P21/n, Z = 4, m =
0.098 mm21, l = 0.71073 A [Mo-Ka], 12808 reflections measured, 1854
˚
19 T. L. Little and S. E. Webber, J. Org. Chem., 1994, 59, 7299.
20 C. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M. S. Moruzzi and
A. Vacca, Anal. Biochem., 1995, 231, 374.
21 L. M. Toledo, K. Musa, J. W. Lauher and F. W. Fowler, Chem. Mater.,
1995, 7, 1639.
unique (Rint = 0.0978), 1533 observed (I . 2s(I)). The final R1 was 0.0430
(observed reflections 0.0528) and wR(F2) was 0.1091 (all data 0.1174) for
136 parameters. Crystal data for C51H50Cl6N12O8, M = 1171.73, crystal
size 0.26 x 0.23 x 0.22 mm, triclinic, a = 10.5748(2), b = 11.2109(2),
˚
c = 11.8373(3) A, a = 101.7370(8)u, b = 100.6000(8)u, c = 96.1140(13)u,
346 | Chem. Commun., 2008, 344–346
This journal is ß The Royal Society of Chemistry 2008