p-octiphenyl β-barrels with ion channel and esterase activity

Article abstract of DOI:10.1021/ol016914n

(Metrix Presented) Design, synthesis, and esterase and ion channel activity of a novel barrel-stave supramolecule with hydrophobic exterior and histidine-rich interior are reported. Voltage-dependent binding of pyrenyl-8-oxy-1,3,6-trisulfonates by histidi

Full text of DOI:10.1021/ol016914n

ORGANIC
LETTERS
2001
Vol. 3, No. 26
4229-4232
p-Octiphenyl â-Barrels with Ion Channel
and Esterase Activity
Bodo Baumeister, Naomi Sakai, and Stefan Matile*
Department of Organic Chemistry, UniVersity of GeneVa, CH-1211 GeneVa,
Switzerland
stefan.matile@chiorg.unige.ch
Received October 16, 2001
ABSTRACT
Design, synthesis, and esterase and ion channel activity of a novel barrel-stave supramolecule with hydrophobic exterior and histidine-rich
interior are reported. Voltage-dependent binding of pyrenyl-8-oxy-1,3,6-trisulfonates by histidines within p-octiphenyl â-barrels (and not monomers)
via ionic (and not hydrophobic) interactions (KD, K , KM < 1 µM) is the basis for superb esterolytic proficiency up to (kcat/KM)/kuncat ) 9.6 × 105
I
in water and bilayer membranes. The conductance of labile ion channels formed in planar bilayer membranes is shown to be reduced by
8-hydroxypyrene-1,3,6-trisulfonate on the single- and multichannel level.
The creation of “man-made enzymes” beyond the classical
setting has attracted scientific curiosity for decades.1 Pertinent
work in this dynamic area highlights binding proteins
(abzymes), oligonucleotides (ribozymes), organic polymers
(synzymes), de novo R-helix bundles, supramolecular ar-
chitectonics, steroids, cyclodextrins, crown ethers, and calix-
arenes as priviledged scaffolds.1,2 Most of these archetypal
motifs appear in artificial receptors1 and ion channels as
well.3,4 Synthetic supramolecular architectonics able to form
ion channels and to recognize and transform molecular guests
have, however, not yet been reported. Here, we describe
design, construction, and characteristics of the first synthetic
R. N. Biochemistry (Moscow) 2000, 65, 335. (g) Gokel, G. W. Chem.
Commun. 2000, 1. (h) Sakai, N.; Matile, S. Chem. Eur. J. 2000, 6, 1731.
(i) Seebach, D. Fritz, M. G. Int. J. Biol. Macromol. 1999, 25, 217. (j)
Hartgerink, J. D.; Clark, T. D.; Ghadiri, M. R. Chem. Eur. J. 1998, 4, 1367.
(k) Kobuke, Y. AdV. Supramol. Chem. 1997, 4, 163. (l) Koert, U. Chem.
Unserer Z. 1997, 31, 20. (m) Gokel, G. W.; Murillo, O. Acc. Chem. Res.
1996, 29, 425. (n) Fyles, T. M.; van Straaten-Nijenhuis, W. F. In
ComprehensiVe Supramolecular Chemistry; Reinhouldt, N. D.; Ed.; Elsevi-
er: Oxford, 1996; Vol. 10, p 53. (o) Voyer, N. Top. Curr. Chem. 1996,
184, 1. (p) Fuhrhop, J.-H.; Ko¨ning, J. Membranes and Molecular As-
semblies: The Cytokinetic Approach; The Royal Society of Chemistry:
Cambridge, U.K., 1994. (q) Nolte, R. J. M. Chem. Soc. ReV. 1994, 11. (r)
Åckerfeldt, K. S.; Lear, J. D.; Wasserman, Z. R.; Chung, L. A.; DeGrado,
W. F. Acc. Chem. Res. 1993, 26, 191. Recent publications: (s) Wang, D.;
Guo, L.; Zhang, J.; Jones, L. R.; Chen, Z.; Pritchard, C.; Roeske, R. W. J.
Peptide Res. 2001, 57, 301. (t) Bandyopadhyay, P.; Janout, V.; Zhang, L.-
H.; Regen, S. L. J. Am. Chem. Soc. 2001, 123, 7691. (u) Wright A. J.;
Matthews, S. E.; Fischer, W. B.; Beer, P. D. Chem. Eur. J. 2001, 7, 3474.
(v) Yoshino, N.; Satake, A.; Kobuke, Y. Angew. Chem., Int. Ed. 2001, 40,
457. (w) Waser, P.; Rueping, M.; Seebach, D.; Duchardt, E.; Schwalbe, H.
HelV. Chim. Acta 2001, 84, 1821. (x) Bong, D. T.; Clark, T. D.; Granja, J.
R.; Ghadiri, M. R. Angew. Chem., Int. Ed. 2001, 40, 988. (y) Fyles, T. M.;
Hu, C.; Knoy, R. Org. Lett. 2001, 3, 1335. (z) Biron, E.; Voyer, N.; Meillon,
J. C.; Cormier, M.-E.; Auger, M. Biopolymers 2001, 55, 364. (aa) Arndt,
H. D.; Knoll, A.; Koert U. Angew. Chem., Int. Ed. 2001, 40, 2076. (bb)
Pe´rez, C.; Esp´ınola, C. G.; Foces-Foces, C.; Nu´n˜ez-Coello, P.; Carrasco,
H.; Mart´ın, J. D. Org. Lett. 2000, 2, 1185.
(1) (a) Dugas, H. Bioorganic Chemistry; Springer: New York, 1996.
(b) Sazaki T.; Lieberman, M. Protein Mimetics. In ComprehensiVe
Supramolecular Chemistry; Murakami, Y.; Ed., Elsevier: Oxford, 1996;
Vol. 4, pp 193-243.
(2) (a) Kirby, A. J. Enzyme Mimics. In Stimulating Concepts in
Chemistry; Shibasaki, M., Stoddart, J. F., Vo¨gtle, F., Eds.; VCH: New York,
2000; pp 341-353. (b) Broo, K. S.; Nilson, H.; Nilson, J.; Balzer, L. J.
Am. Chem. Soc. 1998, 120, 10287. (c) Sanders, J. K. M. Chem. Eur. J.
1998, 4, 1378. (d) Kirby, A. J. Angew. Chem., Int Ed. Engl. 1996, 35,
707. (e) Pollack, S. J.; Schultz, P. G. J. Am. Chem. Soc. 1989, 111, 1929.
(f) Tramontano, A.; Janda, K. D.; Lerner, R. A. Science 1986, 234, 1566.
(g) Guthrie, J. P.; Cossar, J.; Dawson, B. A. Can. J. Chem. 1986, 64,
2456.
(3) Reviews on synthetic ion channels: (a) Gokel, G. W.; Mukhopadhyay,
A. Chem. Soc. ReV. 2001, 30, 274. (b) Bong, D. T.; Clark, T. D.; Granja,
J. R.; Ghadiri, M. R. Angew. Chem., Int. Ed. 2001, 40, 988. (c) Gokel, G.
W. Chem. Eur. J. 2001, 7, 33. (d) Matile, S. Chem. Rec. 2001, 1, 162. (e)
Kirkovits, G. J.; Hall, C. D. AdV. Supramol. Chem. 2000, 7, 1. (f) Reusch,
10.1021/ol016914n CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/07/2001

were so far unable to determine the precise barrel stoichi-
ometry experimentally, we decided to assume, supported by
molecular models,4f previous results,4 and the magnitude of
the single channel conductance (vide infra), a tetrameric
supramolecule (i.e., 14) throughout the manuscript with the
only intention to simplify calculations and text.
p-Octiphenyl peptide 1 was synthesized and characterized
following established protocols (see Supporting Information).
Structural studies of peptide rods 1 by atomic force micros-
copy suggested that barrel-stave supramolecule 1n self-
assembles spontaneously in aqueous media at concentrations
below 3 µM, whereas â-fibrillogenesis occurs above 3 µM.5
These insights limit functional studies on 14 to the nanomolar
range and hamper conventional structural studies at elevated
concentrations.
Esterolytic activity of barrel 14 was assessed using CB-
ester substrates (CB ) Cascade Blue ) pyrenyl-8-oxy-1,3,6-
trisulfonate,6 Scheme 1). The planar, trianionic CB motif was
Scheme 1
Figure 1. Structure of p-octiphenyl 1 and designed cutaway
suprastructure of rigid-rod â-barrel 14 with pertinent distances
estimated from molecular models. The extent of protonation of
internal histidines depends on pH and is indicated arbitrarily. Four
histidine residues from two â-strands forming a rectangle of ∼5 Å
× ∼7 Å are named “H-quartet” for convenience only (bottom).
envisioned for molecular recognition because the distances
between the three negative charges and those between the
three positive charges of protonated histidine residues in an
H-quartet are similar (Figure 1).
Esterolysis was monitored by sensitive fluorescence kinet-
ics following the increase in emission of product HPTS as a
function of time.7 The original trace with 1.0 µM CB-acetate
2 and 1.35 nM barrel 14 exemplifies turnovers of >120
substrates per barrel 14 (i.e., >30 per rod 1, >2 per histidine,
Figure 2a). The bell-shaped pH profile with maximal initial
velocities around pH 5.5, indicative for importance of partial
but not full histidine protonation for activity, was suggestive
for presence of nucleophilic or base catalysis (Figure 2b).
The esterolytic activity of rigid-rod â-barrel 14 at optimized
pH 5.5 was independent of presence and absence of spherical
bilayer membranes composed of egg yolk phosphatidylcho-
line (Figure 2c).8 Application of the Michaelis-Menten
model gave KM ) 0.7 µM, kcat ) 0.13 min-1 (0.0021 min-1
per histidine) and a catalytic efficiency kcat/KM ) 3064 M-1
s-1 (48 M-1 s-1 per histidine) for tetramer 14. Autohydrolysis
barrel-stave supramolecule with ion channel and esterase
activity, i.e., rigid-rod â-barrel 14, Figure 1.
The design of rigid-rod â-barrels, developed in our lab
during the past 2 years, is based on the self-assembly of n
rigid-rod staves (here 1) into barrel-stave supramolecules
(here 1n).4 Interdigitation of lateral pentapeptide strands gives
multiple antiparallel â-sheets. Their precise topology places
all hydrophobic leucine side chains at the outer barrel surface
to interact with lipid bilayers, while the hydrophilic histidine
residues point inward to form a transmembrane channel. The
hypothetical rectangle of ∼5 Å × ∼7 Å formed by four
histidine residues from two â-strands is in the following
referred to as “H-quartet” for convenience only. Because we
(4) (a) Matile, S. Chem. Soc. ReV. 2001, 30, 158. (b) Das, G.; Matile, S.
Chirality 2001, 13, 170. (c) Sakai, N.; Baumeister, B.; Matile, S.
ChemBioChem 2000, 1, 123. (d) Baumeister, B.; Sakai, N.; Matile, S.
Angew. Chem., Int. Ed. 2000, 39, 1955. (e) Baumeister, B.; Matile, S. Chem.
Commun. 2000, 913. (f) Baumeister, B.; Matile, S. Chem. Eur. J. 2000, 6,
1739. (g) Sakai, N.; Majumdar, N.; Matile, S. J. Am. Chem. Soc. 1999,
121, 1, 4294.
(5) Das, G.; Ouali, L.; Adrian, M.; Baumeister, B.; Wilkinson, K. J.;
Matile, S. Angew. Chem., Int. Ed. In press.
(6) Whitaker, J. E.; Haugland, R. P.; Moore, P. L.; Hewitt, P. C.; Reese,
M.; Haugland, R. P. Anal. Biochem. 1991, 198, 119.
(7) Wolfbeis, O. S.; Koller, E. Anal. Biochem. 1983, 129, 365.
4230
Org. Lett., Vol. 3, No. 26, 2001

influenced substrate binding (KM) rather than substrate
conversion (Vmax), i.e., acts as a competitive inhibitor with
KI ) 0.5 µM.
The influence of substrate hydrophobicity was determined
by comparing the characteristics of CB-butyrate 4, CB-
caprylate 5, and CB-laurate 6 with those of CB-acetate 2
(Figure 2e). Fractional changes of KM (9), kcat (b), and kcat/
KM (O) were plotted as a function of the number n of
additional methylene groups in the substrate (Scheme 1). All
observed changes were within 1 order of magnitude. Cor-
roborating the absence of hydrophobic substrate binding
pockets in barrel 14, the KM actually increased with increas-
ing substrate hydrophobicity up to 4.2 µM for CB-laurate 6.
Decreasing initial velocities of product formation in the
presence of increasing concentrations of KCl corroborated
the expected strong influence of ionic strength on dominant
electrostatic interactions (Figure 2f).
Ion channel formation was studied in planar EYPC-
bilayers doped with 0.06 mol % of rod 1. At pH 5.0, an
ohmic single-channel conductance level of g ) 1.23 nS,
compatible with the expected internal diameter of â-barrel
14, was dominant (Figure 3a). A mean lifetime τ ) 4.5 ms
Figure 2. Esterolytic activity of â-barrels 14. (a) Increase in product
concentration (HPTS, λem ) 510 nm, λex ) 415.5 nm; controls,
404 nm, 455 nm) as a function of time (1 µM 2, 1.35 nM 14, pH
5.5). (b) Initial velocities of product formation as a function of pH.
(c) Substrate concentration in the presence (b) and absence (O) of
EYPC bilayer membranes (10 mM EYPC) compared to 540 nM
MeIm (0). (d) Lineweaver-Burk plots in the presence of 0 µM
(b), 0.5 µM (0), and 1.0 µM (O) inhibitor 3, (e) hydrophobicity
plot of kinetic constants x (n) ) KM (9), kcat (b), and kcat/KM (O)
for substrates 2 (n ) 0, x (0)), 4 (n ) 2), 5 (n ) 6), and 6 (n )
10). (f) Initial velocities of product formation as a function of KCl
concentration. Conditions, if not varied as indicated: 135 nM
â-barrel 14, 2 µM 2, 10 mM MES, 100 mM KCl, pH 5.5, rt, dark.
Figure 3. Ion channel activity of â-barrels 14. Representative traces
at -25 mV before (a) and after (b) addition of 50 µM HPTS. (c)
Voltage dependence of the KD (µM) of HPTS from multichannel
dose response isotherms.
of CB-acetate 2 and catalysis by 4(5)-methylimidazole
(MeIm) at identical concentrations is very slow at pH 5.5.
Therefore, a second-order kMeIm ) 0.0032 M-1 min-1 was
measured at millimolar MeIm concentrations and used to
calculate comparable velocities (Figure 2c, 0) and the
proficiency (kcat/KM)/kMeIm ) 9.6 × 105 of tetramer 14 (1.5
× 104 per histidine). Lineweaver-Burk plots for esterolysis
of CB-acetate 2 catalyzed by barrel 14 in the presence of 0
µM (b), 0.5 µM (0), and 1.0 µM (O) pyrene-1,3,6,8-
tetrasulfonate 3 gave separate straight lines that intersected
on the y-axis (Figure 2d). This indicated that anion 3
was calculated. An additional, more stable low-conductance
level (g ) 0.26 nS) was observed occasionally at pH 5.0
and frequently at pH g 5.5 (not shown). Addition of 50 µM
HPTS reduced both the apparent conductance (g ) 0.69 nS)
and the mean lifetime (τ ) 0.69 ms) of the main single-
channel conductance level (Figure 3b). The periods of 50-
200 ms with predominant open channels remained, however,
similar in the presence and absence of HPTS. The lability
of the system was incompatible with quantitative analysis
of HPTS binding on the single-molecule level.9
(8) Control experiments without barrels 14 confirmed that spherical
EYPC-bilayers do not influence the rate of esterolysis. One referee suggested
that a 2-times reduced esterolytic activity in the presence of spherical EYPC-
bilayers could be expected because one barrel terminus is not as accessible
as in water. Compensation of this conceivable effect would then imply that
activity in bilayer membranes may be up to 2 times higher than in water
(because of, e.g., transmembrane substrate concentration gradients).
Addition of HPTS to multiple open channels caused
significant reduction of the current. Thus, apparent dissocia-
(9) (a) Bayley, H.; Cremer, P. S. Nature 2001, 413, 226. (b) Bayley, H.;
Martin, C. R. Chem. ReV. 2000, 100, 2575.
Org. Lett., Vol. 3, No. 26, 2001
4231

tion constants (KD’s) of HPTS were determined from dose
response isotherms, and log KD’s were plotted as a function
of applied voltage (Figure 3c). Although weak, the observed
voltage dependence indicated that the CB-binding site is
located within the ion-conducting pathway of 14. By applying
the Woodhull equation
Some interdependence of functional plasticity was ob-
served as well. Ion channel activity is affected on both the
single and multichannel level by the presence of CBs, and
CB-binding to ion channels 14 is affected by external voltage
applied to planar bilayer membranes. Evidently, ion channels
14 have esterolytic activity, because no reduction of catalytic
activity occurred under conditions where ion channel activity
is observed (Figure 2c). In principle, it should thus be
possible to observe hydrolysis of one single ester on the way
across an ion channel with esterase activity directly in planar
bilayer conductance experiments.11 However, the stability
of single channels formed by rigid-rod â-barrels 14 needs to
be increased to attempt meaningful approaches toward single-
molecule catalysis. Studies in this direction as well as on
the expansion of substrate diversity12 and the construction
of alternative internal recognition motifs are ongoing.13
log KD ) log KD (0 mV) - (δzFV)/(2.303RT)
a distance lA ) 2.7 Å was calculated (z ) charge of the
blocker (-3 for HPTS), δ ) lA/l, l ) channel length ) 34
Å, lA ) distance from entrance to active site).10 Although
the meaning of lA values smaller than the size of the blocker
can be questioned, it is intriguing to note that lA ≈ 2.7 Å is
exactly the value one would expect for CB-binding to the
first accessible H-quartets at the barrel entrance as rate-
limiting process.
Acknowledgment. We thank A. Pinto, J.-P. Saulnier, the
group of F. Gu¨lac¸ar, and H. Eder for NMR, MS, and
elemental analyses, respectively, and the referees for their
helpful suggestions. Financial support by the Swiss NSF (21-
57059.99 and National Research Program “Supramolecular
Functional Materials” 4047-057496) is gratefully acknowl-
edged.
In summary, rigid-rod â-barrel 14 can act as artificial ion
channel on one hand and as artificial esterase on the other.
Strong dependence of esterase activity on ionic strength
(Figure 2f), near independence on substrate hydrophobicity
(Figure 2e), competitive inhibition by tetraanion 3 (Figure
2d), submicromolar KM (Figure 2c) and bell-shaped pH-
profile maximal at pH 5.5 (Figure 2b) all argue in favor of
powerful electrostatic binding of planar CB-substrates by
topologically matching H-quartets within aqueous or mem-
brane-bound p-octiphenyl barrel 14. As a specific example
for this global interpretation, near independence on substrate
hydrophobicity is supportive for â-barrel 14 (rather than
monomer 1) as catalyst because only â-sheet conformation
permits proper separation of the hydrophobic leucine from
active, partially protonated histidine residues in LHLHL-
pentapeptides. Comparison of kcat and KM suggests that this
powerful molecular ground-state recognition (rather than
transition-state stabilization) accounts for the nearly 106-fold
increase of the second-order rate constant.2d Near indepen-
dence on presence and absence of bilayer membranes implied
existence of similar suprastructures in water and bilayers
(Figure 2c).8
Supporting Information Available: Experimental pro-
cedure and characterization for all new compounds and
protocols for ion channel and esterase assays. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL016914N
(10) (a) Hille, B. Ionic Channels of Excitable Membranes, 2nd ed.;
Sinauer: Sunderland, MA, 1992. (b) Woodhull, A. M. J. Gen. Physiol.
1973, 61, 687. It is assumed that the length of the ion channel corresponds
to the length of the p-octiphenyl stave. This assumption implies “free” access
of hydrated inorganic and organic ions to the barrel termini; this seems not
unreasonable in view of an outer barrel diameter of ≈30 Å estimated from
molecular models.4f
(11) (a) Xie, X. S.; Lu, H. P. J. Biol. Chem. 1999, 274, 15967. (b)
Mindell, J. A.; Zhan, H.; Huynh, P. D.; Collier, R. C.; Finkelstein, A. Proc.
Natl. Acad. Sci. U.S.A. 1994, 91, 5272.
(12) RNase activity of 1: Baumeister, B.; Matile, S. Unpublished work.
(13) p-Octiphenyl â-barrels from LDLDL-pentapeptide rods: Das, G.;
Matile, S. Unpublished work.
4232
Org. Lett., Vol. 3, No. 26, 2001