A shape-dependent hydrophobic effect for tetrazoles

Article abstract of DOI:10.1039/b705256a

An adaptive tetrazole-derived host provides insight into tetrazolate-biomolecule interactions, and is the first member of a new family of receptors that function in pure water. The Royal Society of Chemistry.

Full text of DOI:10.1039/b705256a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A shape-dependent hydrophobic effect for tetrazoles{
Devin J. Mahnke,a Robert McDonaldb and Fraser Hof*a
Received (in Austin, TX, USA) 10th April 2007, Accepted 12th June 2007
First published as an Advance Article on the web 29th June 2007
DOI: 10.1039/b705256a
An adaptive tetrazole-derived host provides insight into
tetrazolate-biomolecule interactions, and is the first member
of a new family of receptors that function in pure water.
Tetrazoles have been used extensively in the creation of enzyme
inhibitors; they are routinely pulled from the medicinal chemist’s
toolkit to serve as pharmacophore replacements of carboxylic
acids, and are prized for their similar ionization and recognition
properties at physiological pH.1 In recent years, elegant studies on
tetrazole recognition have underlined the subtle differences in
hydrogen bonding between carboxylates and tetrazoles,2,3 and
tetrazoles have found use in such varied applications as stacked
supramolecular oligomers,4 functional polymers,5 and drug-
binding gels.6 We report here a compact tetrazole-derived receptor
that shows high selectivity in its binding of ammonium ions in
aqueous solutions. Binding and structural studies reveal that the
host’s charged tetrazolate rings have a significant hydrophobicity
only when their faces engage quaternary ammonium ions. Unlike
Scheme 1 Synthesis of hosts 1 and 2.
closely related non-tetrazole analogs, the tetrazole host retains the
ability to bind partners in pure buffered water.
groups cooperate to achieve effective binding of a single cation in
this competitive medium. In analogy with a previous report of
weak tetrazole–amidinium hydrogen bonding,2 host 1-Na3 binds
Host 1 was designed to present three tetrazole substituents on
the same face of its aromatic core,7–9 and was synthesized by
adaptation of a tetrazole-forming methodology that utilizes ZnII to
mediate the cycloaddition of nitriles with N32 (Scheme 1 — see the
ESI{ for synthetic procedures).10 The crystal structure{ of neutral
host 1 reveals a water molecule at the focal point of the host’s three
tetrazole binding elements: one tetrazole NH donates a hydrogen
bond to the water’s lone pair, while the other two tetrazoles are
oriented such that their lone pairs receive hydrogen bonds from the
water’s protons (Fig. 1). To study the cation-binding aptitude of 1
in water, the fully deprotonated host 1-Na3 is prepared by
treatment with NaOMe. NMR titrations in buffered 60 : 40
CD3OD–D2O (pD* 8.65) show a steady trend of decreasing
guanidinium cation below the limits of detection (Kassoc ,10 M21
)
in all of the solvent systems studied.
More interesting than the association constants themselves, a
key indicator of binding geometry is found in the movement of the
1H NMR signals for the host’s methylene protons upon addition
of different ammonium ions. Addition of MeNH3Cl causes an
upfield shift in the host’s CH2CH3 protons, while an opposing
downfield shift is observed upon binding of Me4NI (Fig. 2 and
+
binding constants in the series starting with primary MeNH3
(strongest) and proceeding one methyl group at a time to
quaternary Me4N+ (weakest) (Table, entries 1–4). Despite the
trianionic nature of the host under these conditions, Job plots for
both the primary ammonium salt MeNH3Cl and the quaternary
ammonium salt Me4NI show a 1 : 1 binding stoichiometry (Fig. 2,
inset—see the ESI{ for pKa determinations and additional Job
plots). This suggests that all three of the host’s convergent anionic
aDepartment of Chemistry, University of Victoria, PO Box 3065,
Victoria, Canada. E-mail: fhof@uvic.ca; Fax: 1 250 721 7147;
Tel: 1 250 721 7193
Fig. 1 X-Ray crystal structure of 1?H2O with thermal ellipsoids
displayed at 50% probability. Selected hydrogen bonding distances
and angles: N11–O2S 2.664(2) s; N11–H11–O2S 179(3)u; N24–O2S
2.9374(14) s; O2S–H2SO–N24 176.1(16)u. Primed atoms are disposed
about a crystallographically imposed plane of symmetry. A solvent MeOH
molecule on the exterior of the complex has been omitted for clarity.
bDepartment of Chemistry, University of Alberta, Edmonton, Canada.
E-mail: bob.mcdonald@ualberta.ca; Fax: 1 780 492 8231;
Tel: 1 780 492 2485
{ Electronic supplementary information (ESI) available: Synthetic proce-
dures and characterization, pKa determinations, supplementary Job plot,
NMR titration, and molecular modeling data. See DOI: 10.1039/b705256a
3738 | Chem. Commun., 2007, 3738–3740
This journal is ß The Royal Society of Chemistry 2007

View Article Online
Table 1 Binding constants and binding-induced chemical shifts for
hosts 1-Na3 and 2-Na3 determined by 1H NMR titration at 298 K
Entry Host
Guest
Solventa Kassocb/M21 Ddmaxc/ppm
1
2
3
4
5
6
7
8
9
10
a
1-Na3 MeNH3Cl M/W
1-Na3 Me2NH2Cl M/W
1030
180d
50
50
,10
65
1900
,10
,10
,10
20.29
0d
0.06
0.17
0
0.07
20.07
0
1-Na3 Me3NHCl
1-Na3 Me4NI
1-Na3 MeNH3Cl
1-Na3 Me4NI
2-Na3 MeNH3Cl
2-Na3 MeNH3Cl
2-Na3 Me4NI
M/W
M/W
W
W
M/W
W
M/W
W
0
0
2-Na3 Me4NI
M/W = 60 : 40 CD3OD–D2O, buffered with 10 mM Na2HPO4 at
pD* 8.65. W = D2O buffered with 10 mM Na2HPO4 at pD 7.4.
Values determined by fitting of 1H NMR titration data to a 1 : 1
b
binding isotherm. All values are the average of 2–3 runs with
[Host] = 1–2 mM, and include Kassoc values obtained from all host
protons that undergo chemical shift during the titrations. Estimated
uncertainty ¡20%. Values reported as ,10 are for titrations that
c
produced negligible chemical shift of any host protons. Maximum
binding-induced chemical shift of the diagnostic host CH2CH3
Fig. 3 HF energy-minimized stick diagram and space-filling models
obtained for (a) 132?MeNH3 and (b) 132?Me4N+ using the 6-31+G*
+
d
protons. The CH2CH3 protons did not shift during this titration.
Kassoc was determined by observation of other host protons.
basis set. The host’s methylene protons indicated with circles are those
whose 1H NMR signals shift upfield (for MeNH3+) or downfield (for
Me4N+) upon binding.
methylene protons is explained by their location at the deshielding
edge of the host’s tetrazolate (Fig. 3(b)).
Why does the host choose to direct the tetrazolate faces toward
the guest, and what is the nature of the interaction between an
anionic tetrazolate face and a quaternary ammonium ion?
Titrations in the more competitive pure D2O show that primary
+
MeNH3 is not bound to 1-Na3 at all, while Me4N+ forms a
complex slightly stronger than in CD3OD–D2O mixtures (Table 1,
entries 5 and 6). This result suggests that the tetrazolate rings can
participate in the hydrophobic effect (that is, experience favorable
dispersive interactions with the guest in place of weaker alternative
interactions with water) when they present their faces to the more
hydrophobic quaternary ammonium guest. Further evidence for
the significance of the hydrophobic nature of the tetrazolate faces
comes from the study of tricarboxylate host 2-Na3, an analog with
similar overall charge, geometry, and hydrogen bonding capability,
but lacking the faces of the tetrazolate host. As expected, the
higher charge density of 2-Na3 relative to tetrazole 1-Na3 provides
stronger electrostatic binding—in the less polar CD3OD–D2O
Fig. 2 1H NMR data showing divergent chemical shifts of the diagnostic
host CH2CH3 signal upon titration with MeNH3Cl (&) and Me4NI (m) at
298 K. [Host] = 1 mM. Solvent = 60 : 40 CD3OD–D2O (10 mM
Na2HPO4, pD* = 8.65). Fits of data to 1 : 1 binding isotherms are shown
as solid lines. Inset: Job plots for MeNH3Cl (&) and Me4NI (m)
demonstrate the formation of 1 : 1 complexes for both guests. [Host] +
[Guest] = 5 mM. See the ESI{ for representative titration curves in
buffered water.
+
mixtures, Kassoc for 232?MeNH3 is almost double that of
+
1
32?MeNH3 (Table, entries 1 and 7). However, in pure D2O,
host 2-Na3 does not bind the quaternary ammonium guest Me4N+
(Table, entry 10) despite its electrostatic superiority over host
1-Na3. Only the tetrazole host in its face-on orientation provides a
complementary hydrophobic surface for interaction with Me4N+.
It is well known to medicinal chemists that tetrazolate is
generally more hydrophobic than carboxylate, but tetrazolate-for-
carboxylate substitutions are only sometimes beneficial for
binding.1 Does the shape-dependent hydrophobic character of 1
help explain the interactions of tetrazolate ligands with biomole-
cules? Losartan, the first marketed AT1 receptor antagonist for the
treatment of hypertension, is an example of a tetrazolate-contain-
ing drug that displays enhanced potency (11-fold improved
IC501,12) relative to its carboxylate-containing analog.13 An early
study of losartan SAR/mutation data showed clear differences
between carboxylate- and tetrazolate-receptor interactions that the
Table).§ Molecular modeling11 reveals the source of these opposing
+
chemical shifts. The complex 132?MeNH3 assumes a geometry
similar to that observed in the crystal structure of 1?H2O; the edge
of each tetrazolate ring forms hydrogen bonds with the primary
ammonium ion, and the protons of the host’s adjacent methylene
group are thus forced into the (shielding) face of the aromatic
tetrazolate (Fig. 3(a)). Modeling suggests that in complex
1
32?Me4N+ the tetrazolate binding elements can adopt a variety
of stable edge-on and face-on geometries (see the ESI{), but
the NMR data is most consistent with a structure in which the
tetrazolate arms have rotated to present their aromatic faces to the
guest. In this geometry, the downfield shift of the host’s adjacent
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 3738–3740 | 3739

View Article Online
prism, 0.65 6 0.41 6 0.40 mm; orthorhombic, Cmc21 (no. 36); a =
3
˚
˚
11.9746(7), b = 18.4624(11), c = 10.3004(6) A, V = 2277.2(2) A ; Dc
=
˚
1.337 Mg m23; m = 0.094 mm21; Mo-Ka radiation, l = 0.71073 A;
T = 193 K; 2hmax = 52.76u; 8287 reflections measured (2437 independent);
structure solution and refinement using SHELXS-97 and SHELXL-97
(G. M. Sheldrick, University of Go¨ttingen, Go¨ttingen (Germany), 1997);
183 parameters, hydrogen atoms generated in idealized positions; R1
=
0.0262 (for 2386 reflns with I . 2s(I)), wR2 = 0.0709 (for all 2437
independent reflections). CCDC 637204. For crystallographic data in CIF
or other electronic format see DOI: 10.1039/b705256a.
§ As expected, the host’s other protons all move downfield upon binding
any cationic guest and the guests’ methyl protons move upfield upon
binding the anionic host.
1 R. J. Herr, Bioorg. Med. Chem., 2002, 10, 3379.
2 A. F. Tominey, P. H. Docherty, G. M. Rosair, R. Quenardelle and
A. Kraft, Org. Lett., 2006, 8, 1279.
3 L. Peters, R. Froehlich, A. S. F. Boyd and A. Kraft, J. Org. Chem.,
2001, 66, 3291.
4 A. Kraft, F. Osterod and R. Froehlich, J. Org. Chem., 1999, 64, 6425.
5 A. Taden, A. H. Tait and A. Kraft, J. Polym. Sci., Part A: Polym.
Chem., 2002, 40, 4333.
6 A. Tominey, D. Andrew, L. Oliphant, G. M. Rosair, J. Dupre and
A. Kraft, Chem. Commun., 2006, 2492.
Fig. 4 A portion of Type II dehydroquinase-inhibitor complex (PDB
code 2C4W18) showing edge-on hydrogen bonds (dashed lines) and face-
on hydrophobic contacts (solid lines) between the tetrazolate of the
˚
inhibitor and neighboring fragments. Selected distances (A) between heavy
7 T. D. P. Stack, Z. Hou and K. N. Raymond, J. Am. Chem. Soc., 1993,
115, 6466–6467.
8 G. Hennrich and E. V. Anslyn, Chem.–Eur. J., 2002, 8, 2219–2225.
9 A. Metzger, V. M. Lynch and E. V. Anslyn, Angew. Chem., Int. Ed.
Engl., 1997, 36, 862.
atoms are indicated.
authors could only attribute to a structurally undefined ‘‘non-
conventional salt bridge’’ between losartan’s tetrazolate and the
AT1 receptor’s Lys199.14 More than ten years later, the nature of
the interaction between the tetrazolate of losartan (and those of
newer tetrazolate-containing AT1 antagonists) and the protein is
still debated,15,16 but our current results are most consistent with a
recently published model15 in which the c, d, e methylenes of
Lys199 make a hydrophobic contact with losartan’s tetrazolate
face.17 Further evidence for the relevance of geometry-dependent
hydrophobicity for tetrazolates comes from one of the few crystal
structures of a bound anionic tetrazolate ligand in the Protein
Data Bank. The structure of a tetrazolate ligand bound to Type II
DHQase from H. pylori (PDB code 2C4W)18 reveals a significant
hydrophobic contact on each face of the inhibitor’s tetrazolate ring
and an equator of hydrogen bonding contacts to polar residues
positioned around the tetrazolate’s edges (Fig. 4).
10 Z. P. Demko and K. B. Sharpless, J. Org. Chem., 2001, 66, 7945.
11 ‘Spartan ‘04’, Wavefunction, Irvine, USA, 2004.
12 R. R. Wexler, W. J. Greenlee, J. D. Irvin, M. R. Goldberg,
K. Prendergast, R. D. Smith and P. B. M. W. M. Timmermans,
J. Med. Chem., 1996, 39, 625.
13 D. J. Carini, J. V. Duncia, P. E. Aldrich, A. T. Chiu, A. L. Johnson,
M. E. Pierce, W. A. Price, J. B. Santella, G. J. Wells, R. R. Wexler,
P. C. Wong, S.-E. Yoo and P. B. M. W. M. Timmermans, J. Med.
Chem., 1991, 34, 2525.
14 K. Noda, Y. Saad, A. Kinoshita, T. P. Boyle, R. M. Graham, A. Husain
and S. S. Karnik, J. Biol. Chem., 1995, 270, 2284.
15 A. Patny, P. V. Desai and M. A. Avery, Proteins: Struct., Funct., Bioinf.,
2006, 65, 824.
16 T. Tuccinardi, V. Calderone, S. Rapposelli and A. Martinelli, J. Med.
Chem., 2006, 49, 4305.
17 The structurally analogous interaction of Lysine’s e-CH2 group with
neutral aromatic side chains such as Trp is also energetically favorable.
See: C. D. Tatko and M. L. Waters, Protein Sci., 2003, 12, 2443–2452;
M. J. Rashkin, R. M. Hughes, N. T. Calloway and M. L. Waters, J. Am.
Chem. Soc., 2004, 126, 13320–13325; C. D. Tatko and M. L. Waters,
J. Am. Chem. Soc., 2004, 126, 2028–2034; R. M. Hughes and
M. L. Waters, J. Am. Chem. Soc., 2005, 127, 6518–6519.
18 D. A. Robinson, K. A. Stewart, N. C. Price, P. A. Chalk, J. R. Coggins
and A. J. Lapthorn, J. Med. Chem., 2006, 49, 1282.
19 J. Chin, C. Walsdorff, B. Stranix, J. Oh, H. J. Chung, S.-M. Park and
K. Kim, Angew. Chem., Int. Ed., 1999, 38, 2756.
20 J. Chin, J. Oh, S. Y. Jon, S. H. Park, C. Walsdorff, B. Stranix,
A. Ghoussoub, S. J. Lee, H. J. Chung, S.-M. Park and K. Kim, J. Am.
Chem. Soc., 2002, 124, 5374.
21 S.-G. Kim and K. H. Ahn, Chem.–Eur. J., 2000, 6, 3399.
22 K. H. Ahn, S.-G. Kim, J. Jung, K.-H. Kim, J. Kim, J. Chin and K. Kim,
Chem. Lett., 2000, 170.
In terms of binding capabilities in physiologically relevant
solutions, our tetrazole-derived host occupies a middle ground
between related hosts built using neutral heterocycles19–26 (which
rely on uncharged hydrogen bonds too weak to bind cationic
guests in pure water without extensive hydrophobic decoration27)
and carboxylate host 2-Na3 (which can employ electrostatic
attraction for binding in less polar solvents, but is too well solvated
to bind ammonium ion guests in pure water). We are currently
seeking hosts with higher affinity for quaternary ammonium ions
in order to take advantage of the attractive features of tetrazole-
derived hosts (low molecular weight, ease of synthesis, and good
pharmacological properties) not found in classical cyclophane-
derived quaternary ammonium ion receptors.28–30
23 S.-G. Kim, K.-H. Kim, J. Jung, S. K. Shin and K. H. Ahn, J. Am.
Chem. Soc., 2002, 124, 591.
24 S.-G. Kim, K.-H. Kim, Y. K. Kim, S. K. Shin and K. H. Ahn, J. Am.
Chem. Soc., 2003, 125, 13819.
The authors thank Tom Fyles for expert assistance with pKa
determinations. This research was funded by NSERC, CFI/
BCKDF, and the University of Victoria.
25 K. H. Ahn, H.-y. Ku, Y. Kim, S.-G. Kim, Y. K. Kim, H. S. Son and
J. K. Ku, Org. Lett., 2003, 5, 1419.
26 J. Kim, S.-G. Kim, H. R. Seong and K. H. Ahn, J. Org. Chem., 2005,
70, 7227.
27 J. Kim, B. Raman and K. H. Ahn, J. Org. Chem., 2006, 71, 38.
28 J. C. Ma and D. A. Dougherty, Chem. Rev., 1997, 97, 1303.
29 E. A. Meyer, R. K. Castellano and F. Diederich, Angew. Chem., Int.
Ed., 2003, 42, 1210.
Notes and references
{ Crystals of 1?H2O?MeOH grown by slow evaporation of a wet
methanol solution of 1: C19H30N12O2, Mr = 458.55 g mol21; colorless
30 P. Lhota´k and S. Shinkai, J. Phys. Org. Chem., 1997, 10, 273.
3740 | Chem. Commun., 2007, 3738–3740
This journal is ß The Royal Society of Chemistry 2007