A tryptophan-analog host whose interactions with ammonium ions in water are dominated by the hydrophobic effect

Article abstract of DOI:10.1016/j.tetlet.2009.09.161

The binding of quaternary ammonium guests to a flexible, indole-based host has been studied in both aqueous and organic solvents. Binding was shown to depend strongly on the hydrophobic effect and less on the cation-π interaction.

Full text of DOI:10.1016/j.tetlet.2009.09.161

Tetrahedron Letters 50 (2009) 7035–7037
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
A tryptophan-analog host whose interactions with ammonium ions in water
are dominated by the hydrophobic effect
*
Amanda L. Whiting, Nicole M. Neufeld, Fraser Hof
Department of Chemistry, University of Victoria, PO Box 3065, Victoria, BC, Canada V8W 3V6
a r t i c l e i n f o
a b s t r a c t
Article history:
The binding of quaternary ammonium guests to a flexible, indole-based host has been studied in both
aqueous and organic solvents. Binding was shown to depend strongly on the hydrophobic effect and less
on the cation–p interaction.
Received 23 June 2009
Accepted 28 September 2009
Available online 1 October 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Cation–p interaction
Hydrophobic effect
Quaternary ammonium ions
NMR titration
Aromatic cage
The combination of an electron-rich
p surface with a significant
hydrophobic character makes indoles key players in a variety of
biologically important recognition events. Surveys of protein struc-
tures have shown that over 25% of all tryptophan side chains are in
close contact with cationic neighbors,1 including a subset that are
particularly important for protein–protein interactions.2 Quater-
nary ammonium cations, including a variety of neurotransmitters,3
phospholipid head groups,4 and post-translationally methylated
lysine side chains,5 constitute a specific class of cations that are
routinely engaged by a family of protein-binding pockets called
‘aromatic cages’. The known examples of these pockets, expressed
in a variety of evolutionarily distinct domains, typically contain at
least one tryptophan indole ring alongside other aromatic residues.
Figure 1 shows two examples of such binding motifs, with one
example each of a cationic inhibitor binding to an enzyme (1a)
and a cation-mediated protein–protein interaction (1b).
Figure 1. Examples of quaternary ammonium binding in biological systems. (a)
+
RNMe3 group of an inhibitor bound in the S4-pocket of Factor Xa (PDB code
2BOK).16 (b) Histone H3 trimethylated Lys4 (Lys4me3) bound to the double tudor
domain of JMJD2A (PDB code 2GFA).19 Grey, carbon; red, oxygen; blue, nitrogen.
Taking inspiration from the aromatic cages of biology, tris(in-
dole) host 1 was designed while considering the general goals of
rapid synthesis, the presentation of multiple electron-rich indole
rings, and water solubility (Fig. 2). We report herein, the synthesis
of host 1 and studies on its binding with various quaternary
ammonium cations in pure water. Solvent effects were examined
by using the analogous chloroform-soluble host 2. Differing trends
were found in the two solvent systems, allowing us to comment on
the roles of cation–p interactions, the hydrophobic effect, and
preorganization.
* Corresponding author. Tel.: +1 250 721 7193.
E-mail address: fhof@uvic.ca (F. Hof).
Figure 2. Hosts studied in this work.
0040-4039/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2009.09.161

7036
A. L. Whiting et al. / Tetrahedron Letters 50 (2009) 7035–7037
Table 1
The desired water-soluble indole host 1 is synthesized in three
Cation affinities of hosts 1 and 2
steps starting from commercially available indole-3-propionic acid
3 (Scheme 1). After protection as the methyl ester,6 three equiva-
lents of indole 4 are coupled to 1,3,5-tribromomethylbenzene (5)
upon deprotonation with NaH in DMF in 36% yield. Basic hydroly-
sis of methyl ester 6 is followed by acidification and isolation of the
triacid product. Formation of the tri-sodium salt by treatment with
stoichiometric NaOMe gives 1 in an overall yield of 31%. Chloro-
form-soluble indole host 2 is similarly synthesized from 5 and in-
dole in 39% yield.7
Host 1 is soluble in water to a concentration greater than
15 mM, and NMR dilution studies revealed self-association in
phosphate-buffered D2O with a Khomodimer of 400 MÀ1. The degree
of homodimerization, though small at the concentrations of the
following experiments (1–2 mM), was taken into account when
calculating binding constants in water (see ESI for details). We
investigated the binding of various cationic guests by 1 by NMR
titration of solutions of ammonium salts (2–150 mM) into host
solutions (1–2 mM) in D2O buffered with sodium phosphate
(50 mM) at pD 7.6/pH 8.0 (Fig. 3).8 Host concentration was
matched in both titrant and receiving solution, and chemical shift
data were fit to a 1:1 binding isotherm using a standard protocol.9
The average Kassoc values resulting from three replicate titrations of
each host-guest pair are shown in Table 1. In all cases, the guest
signals moved upfield while the host signals moved downfield, as
expected for interactions between the cationic guests and the aro-
matic surfaces of the host.
b
Entry
Host
Guest
Solventa
Kassoc /MÀ1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
Me4NI
D2O
D2O
D2O
D2O
D2O
D2O
D2O
D2O
D2O
D2O
D2O
CDCl3
CDCl3
CDCl3
CDCl3
CDCl3
35
32
26
6
2
8
Me4NCl
Me4NOAc
LysMe3Cl
AChCl
MeNH3Cl
Me2NH2Cl
Me3NHCl
Et4NCl
63 13
42
7
5
5
20 18
34 16
66
7
Pr4NCl
450 250
Bu4NCl
AChCl
(2.7 3.5) Â 104
5
4
5
5
1
AChClc
Et4NCl
Pr4NCl
Bu4NCl
2
0d
0d
a
D2O = phosphate-buffered D2O (50 mM Na2HPO4/NaH2PO4 at pH 8.0).
Values determined at 295 K by fitting of 1H NMR titration data to a 1:1 binding
b
isotherm. All values are the average of 3–4 titrations. The errors reported are the
standard deviations of the averaged Kassoc values. In D2O, [host] = 1–2 mM,
[guest] = 2–150 mM. In CDCl3, [host] = 1–2 mM, [guest] = 45–400 mM.
c
Value determined by inverse titrations in CDCl3, where [guest] = 1–2 mM,
[host] = 20–300 mM.
d
Small chemical shifts and nonexistent curvature indicate that the Kassoc is below
the detectable limit for NMR.
In phosphate-buffered D2O, host 1 showed similar affinity for
Me4N+ ions regardless of counter anion (Table 1: entries 1–3),
which stands in contrast to the strong counterion effects typically
observed in CDCl3.10 Nevertheless, all subsequent titrations were
carried out with ClÀ salts to avoid the possibility of even small
counterion effects. Host 1 also displayed similar association con-
stants for the biologically relevant quaternary cations trimethylly-
sine and acetylcholine (LysMe3 and ACh, entries
4 and 5),
suggesting a similar mode of binding involving the engagement
of the RNMe3+ portion of each compound. A second series of guests
was investigated to examine the effect of increasing alkylation on
binding to 1 (entries 6–11). The stepwise increasing methylation
+
from MeNH3 (Kassoc = 7 MÀ1) to Me4N+ (Kassoc = 32 MÀ1) resulted
in a trend of increasing affinity with methylation. A small addi-
tional increase in affinity was found for Et4N+, while Pr4N+ and
Bu4N+ gave rise to much more dramatic increases in affinity, with
Bu4N+ (Kassoc = 2.7 Â 104 MÀ1) binding 1500-fold more strongly
than Me4N+.
Scheme 1. Synthesis of host 1.
NMR titrations on host 2 were performed in CDCl3 using proto-
cols similar to those reported for host 1. Acetylcholine chloride was
used as a replacement for Me4N+ ClÀ due to the insolubility of the
latter in CDCl3. Weak but consistently measurable binding con-
stants were observed (Table 1, entries 12–16). The strongest bind-
ing was found between
2 and acetylcholine chloride, with
progressively weaker binding observed upon increasing size of
quaternary ammonium ion.
The interactions of indoles (in the form of tryptophan side
chains) and quaternary ammonium ions (in a variety of neuro-
transmitters and in trimethyllysine side chains) are increasingly
recognized as critical components of a variety of biochemical sig-
naling pathways,11,12 and the cation–
p
interaction is a primary
driving force for these binding events.13,14 In general, larger cations
should produce weaker cation– interactions because (a) the lar-
ger cations have a larger radius of interaction between surface
p
Figure 3. (a) Representative 1H NMR titration data for the complexation of
acetylcholine chloride by host 1. Titrations carried out at 295 K in phosphate-
buffered D2O (50 mM Na2HPO4/NaH2PO4 at pH 8.0). Titrant containing guest
p
and positive charge and (b) larger cations bear a more diffuse po-
sitive charge, creating a weaker electrostatic interaction with elec-
(125 mM) and host (1 mM) was added to
a solution containing a matched
tron-rich
p surfaces. Yet it is often observed in protein binding
concentration of host. Tracked signals are j (6.72 ppm, PhH) and d (6.44 ppm,
indole 2-H). (b) Titration curve resulting from (a) using shifts from j. The line
represents fitted 1:1 binding isotherm for this equilibrium (Kassoc = 38 3 MÀ1).
events that more highly methylated ammonium ions bind more
strongly to their aromatic partners (indeed that is the basis for

A. L. Whiting et al. / Tetrahedron Letters 50 (2009) 7035–7037
7037
duces binding that is remarkably strong given the complete
flexibility of both host and guest. It is possible that this strong
binding motif, very easily prepared, may be exploited in other
unnatural aqueous-phase recognition systems. We are continuing
to explore the links between preorganization, hydrophobicity,
and high-affinity binding in water in this and other Trp-derived
systems.
Acknowledgments
A.W. thanks NSERC for PGS M and PGS D fellowships. F.H. is a
Career Scholar of the Michael Smith Foundation for Health Re-
search. This research was supported by NSERC and the University
of Victoria.
Figure 4. Optimized structures of host 2 bound to acetylcholine (left) and Bu4N+
(right) (DFT B3LYP/6-31G ). Key rotational degrees of freedom are indicated.
*
Supplementary data
selective signaling by trimethylated lysine residues15), and that
trend is reproduced in the current model system. The subtle in-
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2009.09.161.
+
crease observed upon moving from MeNH3 to Me4N+ and the
more dramatic increase in binding for the greasy Bu4N+ ion both
References and notes
run counter to the trends expected for cation–p interactions and
are better explained by the surface-area-dependent hydrophobic
effect. The complete lack of binding of Pr4N+ and Bu4N+ in
CDCl3—where no hydrophobic effect can participate—confirms
the importance of the hydrophobic effect for the binding of cations
by 1 in water.
1. Gallivan, J. P.; Dougherty, D. A. Proc. Nat. Acad. Sci. U.S.A. 1999, 96, 9459–9464.
2. Crowley, P. B.; Golovin, A. Proteins: Struct. Funct. Bioinf. 2005, 59, 231–239.
3. Zhong, W.; Gallivan, J. P.; Zhang, Y.; Li, L.; Lester, H. A.; Dougherty, D. A. Proc.
Nat. Acad. Sci. U.S.A. 1998, 95, 12088–12093.
4. Martin, S. F.; Follows, B. C.; Hergenrother, P. J.; Trotter, B. K. Biochemistry 2000,
39, 3410–3415.
5. Jacobs, S. A.; Taverna, S. D.; Zhang, Y.; Briggs, S. D.; Li, J.; Eissenberg, J. C.; Allis,
C. D.; Khorasanizadeh, S. EMBO J. 2001, 20, 5232–5241.
6. Pedras, M. S. C.; Jha, M. Bioorg. Med. Chem. 2006, 14, 4958–4979.
7. Rajakumar, P.; Swaroop, M. G. Tetrahedron Lett. 2006, 47, 3019–3022.
8. Glasoe, P. K.; Long, F. A. J. Phys. Chem. 1960, 64, 188–190.
9. Sanderson, J. M. http://www.dur.ac.uk/j.m.sanderson/science/downloads.html,
The Centre for Bioactive Chemistry, Department of Chemistry, University
Science Laboratories: Durham, UK.
Mutational studies of aromatic cage sites have demonstrated
that tryptophan side chains participate in cation–
p
interactions.3
Comparisons of t-butyl groups [RC(Me3)] and nearly isosteric
trimethylammonium groups [RN(Me3)+] have been elegantly used
to demonstrate the importance of the cation–p interaction in pro-
tein–protein and drug-protein contacts.13,16 The authors of these
studies conclude that the hydrophobic effect plays no role in the
observed binding events.13 Yet our studies of tris(indole) hosts 1
and 2 suggest that the hydrophobic effect can operate in this arti-
ficial tryptophan analog given a greasy enough guest, and that the
10. Bartoli, S.; Roelens, S. JACS 2002, 124, 8307–8315.
11. Xiu, X.; Puskar, N. L.; Shanata, J. A. P.; Lester, H. A.; Dougherty, D. A. Nature
2009, 458, 534–537.
12. Lin, Z.; Wang, W.; Uhl, G. R. Mol. Pharmacol. 2000, 58, 1581–1592.
13. Hughes, R. M.; Wiggins, K. R.; Khorasanizadeh, S.; Waters, M. L. PNAS 2007, 104,
11184–11188.
14. Zacharias, N.; Dougherty, D. A. Trends Pharmacol. Sci. 2002, 23, 281–287.
15. Jacobs, S. A.; Khorasanizadeh, S. Science 2002, 295, 2080–2083.
16. Schärer, K.; Morgenthaler, M.; Paulini, R.; Obst-Sander, U.; Banner, D. W.;
Schlatter, D.; Benz, J.; Stihle, M.; Diederich, F. Angew. Chem., Int. Ed. 2005, 44,
4400–4404.
17. Li, H.; Ilin, S.; Wang, W.; Duncan, E. M.; Wysocka, J.; Allis, C. D.; Patel, D. J.
Nature 2006, 442, 91–95.
18. Taverna, S. D.; Li, H.; Ruthenburg, A. J.; Allis, C. D.; Patel, D. J. Nat. Struct. Mol.
Biol. 2007, 14, 1025–1040.
19. Huang, Y.; Fang, J.; Bedford, M. T.; Zhang, Y.; Xu, R.-M. Science 2006, 312, 748–
751.
cation–p interaction, if operative, is scarcely measurable. What dif-
ferentiates the indoles of host 1 and those of proteins? Crystal
structures of cation-binding proteins in bound and free states re-
veal almost no movement of the aromatic cage side chains, sugges-
tive of a highly rigid and preorganized binding pocket.17 It has been
recently suggested that this rigidity is a hallmark of aromatic cage
binding sites.18 This rigidity is certainly lacking in the flexible hosts
1 and 2—DFT calculations of 2 bound to both acetylcholine and
Bu4N+ reveal that the host can adapt its shape to complement
the size of the guest cation (Fig. 4). The result is a system that com-
plements the larger hydrophobic surface area of Bu4N+ and pro-