Synthesis, transport activity, membrane localization, and dynamics of oligoester ion channels containing diphenylacetylene units

Article abstract of DOI:10.1039/c0ob00194e

Four new linear oligoesters containing a diphenylacetylene unit were prepared by fragment coupling sequences and the ion channel forming ability of the compounds was investigated. Activity in vesicles was very strongly controlled by overall length; the longest compound was effectively inactive. Planar bilayer studies established that all compounds are able to form channels, but that regular step changes in conductance depend on the location of the diphenylacetylene unit within the oligoester and on the electrolyte. The intrinsic fluorescence of the diphenylacetylene unit was used to probe aggregation and membrane localization. Both monomer (320 nm) and excimer (380 nm) emissions are quenched by copper ions; quenching of the excimer emission from an aqueous aggregate is very efficient. Time-dependent changes in the intensities of monomer and excimer emission show slow transfer of diphenylacetylene units from an aqueous aggregate to a membrane-bound monomer with subsequent growth of emission from a membrane-bound excimer. The latter species is not quenched by aqueous copper ions. The implications of these species and processes for the mechanism of ion channel formation by simple oligoesters are discussed.

Full text of DOI:10.1039/c0ob00194e

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Synthesis, transport activity, membrane localization, and dynamics of
oligoester ion channels containing diphenylacetylene units†
Joanne M. Moszynski and Thomas M. Fyles*
Received 1st June 2010, Accepted 10th August 2010
DOI: 10.1039/c0ob00194e
Four new linear oligoesters containing a diphenylacetylene unit were prepared by fragment coupling
sequences and the ion channel forming ability of the compounds was investigated. Activity in vesicles
was very strongly controlled by overall length; the longest compound was effectively inactive. Planar
bilayer studies established that all compounds are able to form channels, but that regular step changes
in conductance depend on the location of the diphenylacetylene unit within the oligoester and on the
electrolyte. The intrinsic fluorescence of the diphenylacetylene unit was used to probe aggregation and
membrane localization. Both monomer (320 nm) and excimer (380 nm) emissions are quenched by
copper ions; quenching of the excimer emission from an aqueous aggregate is very efficient.
Time-dependent changes in the intensities of monomer and excimer emission show slow transfer of
diphenylacetylene units from an aqueous aggregate to a membrane-bound monomer with subsequent
growth of emission from a membrane-bound excimer. The latter species is not quenched by aqueous
copper ions. The implications of these species and processes for the mechanism of ion channel
formation by simple oligoesters are discussed.
Introduction
Studies of the design and synthesis of synthetic ion channels
over the past two decades have uncovered numerous examples
of relatively simple compounds that can insert into a bilayer
membrane and open ion conducting states (reviews;1–3 recent
examples4–10). Larger, rigid, or conformationally defined synthetic
channel-forming compounds usually suggest a linkage between
structures and possible mechanisms11,12 and are amenable to
structure–function study as a route to determining mechanistic
details.4,6,13,14 Simpler or more flexible compounds typically offer
few structural clues to constrain mechanistic speculations.15–19
Simpler compounds also offer fewer opportunities for innocent
structural modifications to insert probes into the active com-
pounds. As a consequence mechanistic hypotheses—as plausible
as they might be—remain largely untested by direct experiment.
Such is certainly the case for the oligoester channel forming
compounds reported recently.20–22 A solid-phase synthesis gave
easy access to a suite of these compounds and a structure–
activity survey directed attention to a mechanism in which the
transport activity was controlled by aqueous-phase aggregation
Fig. 1 Working hypothesis for the formation of ion channels.
of the compound prior to membrane insertion (Fig. 1).20 The
aggregation of the compounds was probed indirectly and generally
supported the view that the most active compounds aggregated
to the least extent thereby increasing the available monomer for
partition to the membrane.20 Subsequent channel opening was
presumed to occur via a dimeric or oligomeric trans-membrane
species which stabilized an aqueous pore through which ions could
travel, in line with earlier proposals.23
The working hypothesis has much in common with proposals
for the channel-forming activity of conventional detergents,24 and
may be relevant to the modes of action of a number of other
very simple channel-forming compounds.10,25 The membrane-
spanning oligoesters are also inherently dipolar, a necessary
condition for voltage-response functions of synthetic channels.26–28
Consequently, a current focus is to directly test this mechanis-
tic framework in the expectation that higher-order membrane
functions, such as excitable membrane phenomena, will become
practicable based on inherently simple syntheses.
Department of Chemistry, University of Victoria, PO Box 3065, Victoria,
BC, Canada. E-mail: tmf@uvic.ca; Fax: +1 250 721 7147; Tel: +1 250 721
7192
† Electronic supplementary information (ESI) available: Details of syn-
thesis and purifications, 1H and 13C NMR, HPLC traces and MS spectra
for all new compounds; rate data determined from HPTS assay for final
oligomers; fluorescence and fluorescence quenching experimental data as
referred to in the text. See DOI: 10.1039/c0ob00194e
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5139
©

View Article Online
In an effort to find more active compounds, our focus was on
the assumed rate-limiting process of monomer partition from
aqueous solution to the bilayer. The rate of this step can be
controlled through the monomer concentration in equilibrium
with an aqueous aggregate as was inferred in the previous
report.20 It should also be possible to directly influence the
rate of partition via incorporation of rigid elements within the
oligoester; rigid units have previously been implicated as elements
of ion channel designs,2,3 and rigidification can contribute to
control of channel stability.13 In the oligoester system, a rigidifying
element would be expected to influence the activity in two ways:
acceleration of the partition step in the forward direction, and via
destabilization of the aggregate thereby increasing the equilibrium
monomer concentration.29 Introduction of a rigid unit also offers
the potential to incorporate an intrinsic environmental probe into
the target.30
The goals of this paper are to report the synthetic studies that
produce the required oligoesters containing the Dip subunit, to
establish the membrane activity of the new compounds relative
to more flexible relatives prepared previously, and to exploit the
fluorescence properties of the compounds to probe the mechanistic
framework in which these compounds act.
Synthesis
The diphenylacetylene derivative required for the projected targets
(3) was readily prepared by Sonogashira coupling34 of the known
iodo-acid 1 and the protected terminal acetylene 2 as shown
in Scheme 1. A modest excess of 1 with respect to 2 over
a Pd(PPh3)4/CuI/NEt3 catalyst mixture in DMF, followed by
extractive workup and purification by silica gel chromatography
gave the product in an average 85% yield (multiple preparations)
as clear, colourless crystals. However, all attempts to use 3 in a
solid-phase synthesis based upon previously established protocols
for coupling and cleavage gave complex mixtures with several
components containing chromophores. It was possible to establish
that 3 coupled very slowly with alcohols on the support resin, and
Accordingly we focused on a diphenylacetylene containing a,w-
hydroxy acid (Dip) subunit that we envisaged as being introduced
into the suite of known oligoester channels using the previously
reported solid-phase synthesis protocols and subunits. In addition,
the diphenylacetylene unit would provide a useful chromophore
to assist in purification and characterization of the final products,
and more importantly to provide a sensitive spectroscopic probe
of membrane localization and dynamics through fluorescence.
The origin of the luminescence of the parent diphenylacetylene
is complex31,32 but this has not inhibited its use as a spectroscopic
probe.33 Excitation of dilute solutions of diphenylacetylene at
280 nm results in a monomer emission maximum at 320 nm;
in more concentrated solutions a broad excimer emission about
380 nm can be observed.32,33 Both the monomer and excimer
emissions can be quenched, although the suite of known quenchers
is limited.32
The Dip subunit has a carbonyl to hydroxyl length of 1.43 nm,
midway between the estimated lengths of 1.14 nm and 1.65 nm
in the extended conformations of the 8-hydroxyoctanoic acid
and 12-hydroxydodecanoic acid subunits previously incorporated
into oligoester channels.‡ The previous work did not uncover
a marked dependence of activity on overall length, and flexible
structures have the possibility of overall shortening due to gauche
conformations within the chains. The homolog pictured above has
an approximate length of 4 nm, within the range required to span
a bilayer membrane.
‡ The trivial naming of the compounds follows from the linear structures
with the carboxy and hydroxy termini explicitly specified; named subunits
are assumed to be linked as esters. The G(n) unit indicates a diester of 3-
hydroxy glutaric acid with an n-carbon alkyl ester. The a,w-hydroxyacids
of 6, 8, 10, and 12 carbons are indicated as Hex, Oct, Dec, and Dod
respectively. In this paper only the final structures are named this way.
Scheme 1 Synthesis of HO2C-Dip-Hex-Hex-G(12)–OH.§
5140 | Org. Biomol. Chem., 2010, 8, 5139–5149
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
it also appears likely that a significant portion of the acetylene
does not survive the acidic cleavage conditions.
We therefore turned to a solution synthesis of the same overall
design in which the final oligoester would be built from the
carboxyl terminus towards the hydroxy terminus. The solution
synthesis follows a similar set of standard reaction protocols for
ester coupling and deprotection steps; the only exception is that
workup is compound specific as opposed to the filtration/washing
procedure of the solid-phase method. The solution synthesis
requires a carboxylic acid protecting group with properties or-
thogonal to tetrahydropyranyl (THP). The trimethylsilyloxyethyl
(TES) group35 appeared to fulfil the requirements; it is stable to
mild acid conditions but is cleaved by fluoride. An alternative
for carboxylic acid protection is the allylic ester derived from
isoprene (PRE) that is readily cleaved with trimethylsilyl triflate to
give volatile by-products on workup.35 In both cases the required
starting materials were prepared by coupling a THP-protected
hydroxy acid e.g. 6a with the required alcohol under standard
esterification conditions followed by standard THP deprotection.
A key to standard reaction conditions and protected species is
given in a footnote.§
Scheme 1 shows the synthesis of the tetramer HO2C-Dip-Hex-
Hex-G(12)–OH. This synthesis mimics the solid-phase method
through a sequential application of an acidic deprotection to
remove the THP followed by a standard coupling with a THP
protected hydroxy acid. Two deprotection-coupling cycles with
the six-carbon hydroxyacid 6a gave successively compounds 7/8
and 9/10. It was convenient to purify the doubly protected species
7 and 9 as deprotection is essentially quantitative and the products
8 and 10 could be used without purification.
Scheme 2 Synthesis of HO2C-Hex-Hex-Dip-G(12)–OH.§
The two-step sequences 6a to 8 and 8 to 10 occurred in yields of
76% and 72% respectively. The final coupling was accomplished
with the TBDMS protected hydroxyglutaric half ester 11 to give
the fully protected tetramer 12. A final deprotection with fluoride
cleaved both the TES and TBDMS groups to produce the required
hydroxy acid HO2C-Dip-Hex-Hex-G(12)–OH in 17% yield from 3.
This is a rigorous mimic of the unachievable solid-phase
synthesis. The eight chemical steps produce seven intermediates,
which unlike a solid phase synthesis are isolated and identified.
This additional effort eliminates any incomplete coupling products
and the final product is readily purified by preparative HPLC with
concurrent absorbance and fluorescence detection to produce the
tetramer in substantially higher purity than would have been pos-
sible with the solid-phase methodology. A comparable sequence
by solid-phase would take less than a week to complete including
final gel permeation chromatography; the manipulations required
for Scheme 1 take four weeks and involve six chromatographic
separations.
involving 3. A synthesis that explores improvements to the
synthesis of Scheme 1 is given in Scheme 2 for the isomeric HO2C-
Hex-Hex-Dip-G(12)–OH.
Any synthesis to place the Dip subunit at an internal oligoester
position requires a free hydroxyl group at some point. Although
this can be accomplished with the TES/THP combination, the
fluoride deprotection step is best followed by a chromatographic
purification due to impurities in the reagent. A cleaner depro-
tection process could potentially eliminate this chromatographic
step.
Starting from the THP protected hydroxyhexanoic acid 6a,
the PRE protecting group was installed in good yield using the
standard esterification protocol. A standard acidic cleavage of
the THP group in 6b gave the required alcohol 6c which was
extendedwiththe iodoacid 1 togive a PRE protected compound 13
(Scheme 2). Deprotection with TMSOTf was undertaken to allow
the extension of the ester chain from the carboxylic acid terminus.
As expected, the deprotection product 14 was suitable for the next
step without additional purification. The subsequent coupling
with 6e (from 6a by a TES protect/THP deprotect sequence) gave
the Sonogashira precursor 15.
Although a preformed diphenylacetylene such as 3 is required
for a solid-phase synthesis, it is not a requirement for a solution
synthesis. Moreover, the Sonogashira coupling offers an efficiency
advantage compared to the rather sluggish esterification reactions
The coupling with the acetylenic alcohol gave the hydroxy-
terminated trimer 17 that was extended to the fully protected
tetramer 18. As previously, both protecting groups were removed
in a single fluoride treatment to give the target HO2C-Hex-Hex-
Dip-G(12)–OH in an overall yield of 21% over 6 steps from 1.
This synthesis is one step shorter than the linear sequence of
§ Conditions: (i) Pd(0), CuI, NEt3, DMF or THF, rt → 50 ◦C; (ii) DIC,
DiPEA, HOBt, THF, rt → 50 ◦C; (iii) TsOH, CH2Cl2–CH3OH, rt; (iv)
TBAF, THF, rt; (v) TMSOTf, CH2Cl2, rt. Compound 6: RO2C-(CH2)5-
OR¢. 6a: R = H, R¢ = THP. 6b: R = PRE, R¢ = THP, 6c: R = PRE, R¢ =
H. 6d: R = TES, R¢ = THP, 6e: R = TES, R¢ = H. Compound 24: RO2C-
(CH2)9-OR¢: 24a: R = H, R¢ = THP, 24b: R = PRE, R¢ = THP, 24c: R =
PRE, R¢ = H
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5141
©

View Article Online
Scheme 4 Synthesis of HO2C-Dec-Dip-Hex-G(12)–OH.§
are the same; a G(12) unit is coupled with a hydroxy terminated
trimer 27 and the protected product 28 is deprotected in a single
step.
Scheme 3 Synthesis of HO2C-Hex-Dip-Hex-G(12)–OH.§
The earlier stages of the synthesis were complicated by the
apparent instability and poor reactivity of the THP-protected
ten carbon fragment 24a, a precursor of the required building
block 24c. However, once obtained, 24c was seen to undergo
efficient coupling with 1 to yield the aromatic iodide 25. In the
key event, the coupling with the terminal alkyne 19 produced the
expected internal alkyne 26 in 85% yield. The remaining steps of
the synthesis proceeded in good yields to give the target in 41%
overall yield from 1.
Although the manipulations in the solid-phase synthesis are
undoubtedly fewer and easier, the solution syntheses reported
here have significant advantages. These derive directly from the
incorporated chromophore/fluorophore that greatly improves the
detection of products of interest. All compounds are readily
purified by HPLC to give single component samples. This
eliminates the combination NMR/MS analysis used previously
to estimate sample purity. The structure of the synthesis varies, as
shown in the Schemes, but the repertoire of reactions is highly
standardized and reliable. Like the solid-phase, these methods
are sequence independent and applicable for the synthesis of any
desired homolog.
Scheme 1 by virtue of the Sonogashira coupling on an unprotected
alcohol. On a practical level it also reveals the advantage of the
PRE protecting group compared to TES. Nonetheless, this route
is also a linear sequence that proceeds from the carboxylic acid
towards the hydroxy terminus.
Scheme 3 shows a third strategy for the preparation of the final
sequence isomer HO2C-Hex-Dip-Hex-G(12)–OH containing the
Dip group in the central position. This route produces a trimer
strand for final coupling to the G(12) unit via chain growth from
the hydroxyl to the carboxyl terminus. In this sequence the hydroxy
protected acid 6a is converted to the terminal alkyne 19 which then
is coupled with the iodoacid 1 to give the free acid 20. The trimer is
completed through ester coupling with the PRE protected subunit
6c.
The final stages of the synthesis require THP cleavage to open an
alcohol for coupling with the protected G(12) subunit to give 23.
On treatment with TMSOTf, the fully protected product 23 lost
both the PRE and TBDMS protecting groups to give HO2C-Hex-
Dip-Hex-G(12)–OH in good yield and purity. The simultaneous
cleavage of PRE and TBDMS is general and provides a clean and
convenient means to complete the synthesis, leading to an overall
yield of 14% over 6 steps from 16.
Transport studies in vesicles
The routes explored in Schemes 1–3 all involve coupling of low
molecular weight acetylene precursors to ester chains of various
sizes. A more convergent strategy would involve Sonogashira
coupling uniting two similarly-sized ester components. This type
of route is given in Scheme 4. The target is the longer homolog
HO2C-Dec-Dip-Hex-G(12)–OH with the diphenylacetylene at the
centre of the trimer strand. The final stages of Schemes 3 and 4
The activity of the compounds was assessed using the HPTS assay
technique1 with the local variants reported previously.20 In this
method, vesicles are prepared with the pH-sensitive dye HPTS
entrapped. The initial mixture of vesicles and added transporter
shows stable fluorescence. Transport is initiated by the injection
of a pulse of base to create a transmembrane pH gradient that
5142 | Org. Biomol. Chem., 2010, 8, 5139–5149
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
collapses due to transport through the membrane mediated by
added transporter. Alternate excitation at 403 and 460 nm excites
the acid and conjugate base forms of the dye respectively which
both emit at 510 nm. The modulated output signal can be
converted to a normalized extent of transport using the procedure
described previously.20
Either zero- or first-order fitting results can be used to ex-
trapolate an apparent initial rate at the time of the pH pulse
(see the ESI†). This procedure allows comparison over a range
of concentrations for the various compounds. The results are
presented in Fig. 2(B) which emphasizes the point made above;
that the positional isomers of HO2C-Dip-Hex-Hex-G(12)–OH all
show a sharp increase in initial rate at a concentration between 5
and 20 mM. In contrast, the previously reported compound HO2C-
Oct-Dod-Oct-G(10)–OH still behaves as originally reported. The
transport rate constant derived at 32mM from these data is 8.4
¥ 10-4 s-1 in reasonable comparison to the value of (6.3 2.1) ¥
10-4 s-1 reported.20
Fig. 2(B) shows fits for the most active compounds to a logistic
function of the type used previously to characterize aggregation
behaviour.20 The fitting equation is functionally equivalent to the
Hill equation used to characterize the concentration dependence
of ion channels.1 The derived Hill coefficients range from 4 to 11,
indicating significant aggregation within the membrane as a rate-
limiting condition of transport activity. The apparent dissociation
constants of the active species range from 9–15 mM.
Fig. 2(A) shows the data for a transport experiment for
varying concentrations of HO2C-Dip-Hex-Hex-G(12)–OH. The
most striking result is the substantially higher activity of this
compound relative to all the compounds previously reported.
In the previous work, mid-range activity was reported for a
concentration of 32 mM; the activity of HO2C-Dip-Hex-Hex-
G(12)–OH is significant at a five-fold lower concentration. At
low concentrations, the transport curves are very similar to
those previously reported. There is an initial fast “jump” process
following the pH pulse followed by a linear section. The fast
process is the adjustment of the vesicles to ionic and osmotic
stresses via random pore formation within the bilayer that results
in some pH equilibration. The slower linear process is the solution-
diffusion regime which is accelerated by added transporter.36
However, at higher concentrations the curves are clearly never
linear functions of time, but can be fit with high significance to an
apparent first-order function (r-squared > 0.99). The transition
from zero- to first-order behaviour is a very abrupt function of
concentration.
Another striking result from Fig. 2(B) is the very low activity of
HO2C-Dec-Dip-Hex-G(12)–OH. In the previous study it would
have been at the borderline of significant activity. This is a
remarkable effect of length. Several reported systems reveal a
variation in activity as function of transporter length and this has
been exploited as a probe of membrane thickness.37 The very large
activity difference between homologs differing by four methylene
groups is unprecedented.
Transport in planar bilayers
The transport activity of the compounds was also assessed in
diphytanoylphosphatidylcholine (diPhyPC) planar bilayers by the
voltage-clamp technique.38–41 None of the compounds produced
detectible activity by direct injection of compound in methanol or
THF solution into the electrolyte solutions contacting the planar
bilayer. Rather, the compound, typically at 0.3 mol% in lipid, was
physically transferred to a formed bilayer. Alternatively, bilayers
could be formed from lipid containing compound. All bilayers
were monitored to ensure high capacitance/high resistance barri-
ers were continuously present.
The array of observed conductance behaviours is given in Fig. 3;
there is a range of behaviours represented, but each is reliably
produced by the compound/electrolyte combination presented.
Nearly ideally behaved single channels were observed for HO2C-
Hex-Dip-Hex-G(12)–OH with CsCl electrolyte (Fig. 3(A) and
3(B)). These channels are Ohmic with a conductance of 140
3 pS, a lifetime of the order of 700 msec, and an open probability
within the first 30 min of activity of 7%. The activity persists
for hours with increasing open probability that eventually results
in coalescence to a single persistent opening. Multiple levels are
not observed until the persistent opening is established. The longer
homolog HO2C-Dec-Dip-Hex-G(12)–OH is markedly different
but also very well-behaved as shown in Fig. 3(C). In this case
the channels have a conductance of 26 3 pS, a lifetime of 5.7
msec and an open probability of 56%. This flickering behaviour
persists for periods of hours and eventually progresses to an
essentially continuously open state and then to multiples of this
opening.
Fig. 2 Transport by Dip-containing oligoesters in vesicles containing
HPTS. A: Fraction of transport as a function of time for various con-
centrations of HO2C-Dip-Hex-Hex-G(12)–OH; B: Initial rate following
hydroxide gradient creation as a function of concentration for compounds
considered.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5143
©

View Article Online
Fig. 4 Fluorescence characteristics of Dip-containing oligoesters. (A)
Fluorescence spectra (lex = 305 nm) of HO2C-Dip-Hex-Hex-G(12)–OH
in THF (black line) or aqueous solution (10 mM phosphate buffer,
pH = 6.4, 100 mM NaCl, grey line). (B) Fluorescence spectrum of the
same compound in aqueous solution following addition of lipid vesicles
(after a 30 min equilibration period). (C) Time dependence of emission
intensity at 320 and 380 nm for HO2C-Dip-Hex-Hex-G(12)–OH; i marks
the introduction of lipid vesicles. (D) Same experiment as in (C) for
HO2C-Dec-Dip-Hex-G(12)–OH.
Fig. 3 Voltage-clamp traces of current as a function of time. A,
B: HO2C-Hex-Dip-Hex-G(12)–OH in diPhyPC/1M CsCl/-175 mV; C:
HO2C-Dec-Dip-Hex-G(12)–OH/1M CsCl/+175 mV; D, E: HO2C-Hex-
Hex- Dip-G(12)–OH/1M KCl/60 mV.
The other classes of behaviours observed are shown in Fig. 3
sections D and E. These were obtained with HO2C-Hex-Hex-
Dip-G(12)–OH and KCl as electrolyte, but all compounds show
this type of behaviour with KCl, and this isomer also shows
this behaviour with CsCl electrolyte. The dominant characteristic
of these records is a series of short “spikes” of highly variable
conductance. The spikes typically occur directly from the baseline
but their short duration, high conductance, and relatively high
frequency can result in an envelope of high conductance. This
is probably an artefact of the acquisition and hardware filtering
speed; it is highly likely the spikes are individually very short
duration (<msec) openings of highly varied conductance. In
some cases these appear to be associated with, or occurring
in addition to, a step-conductance state as illustrated in Fig. 3(E).
In these cases, the apparent underlying step conductance state
increases in duration and leads eventually to a continuously
open condition with spikes of variable magnitude superimposed.
The underlying level, if it is indeed a discrete state, can be
detected in an all-points histogram as a main feature with a
long tail to higher absolute conductance. The magnitude of
the underlying level varies considerably in the range of 150–
700 pS. There is no evidence of discrete values within this
range. Given that the electrolyte influences the type of activity
observed, a selectivity study is not possible in this system. Similarly,
the incorporation method precludes investigation of voltage-
gating.
Fluorescence studies
Diphenylacetylene is inherently fluorescent;32 in THF solution the
expected monomer emission at 320 nm is readily observed for
all four compounds (Fig. 4(A)). This band is absent in aqueous
buffer, and the emission shifts to 380 nm, a region previously
associated with excimer formation (Fig. 4(A)).32 The excimer arises
from an aggregate. The aggregate species can be detected from
the solvent polarity dependent pyrene emission spectrum as used
previously,20,43 which indicates that a hydrophobic aggregate forms
at a concentration below 20mM (see the ESI†). The overall excimer
emission intensity falls slowly with time as the aggregate species
initially formed by injection of a solution of compound in THF is
not in a stable state in aqueous solution.
The addition of a vesicle solution to the aggregate formed
by HO2C-Dip-Hex-Hex-G(12)–OH in aqueous buffer provokes
a time-dependent change in the emission spectra as illustrated in
Fig. 4(B) and 4(C). At early times following vesicle addition, the
excimer emission intensity decreases while the monomer emission
intensity abruptly increases. Later, the excimer band regains
intensity. These time dependent changes are plotted in Fig. 4(C)
for the emission maxima at 320 and 380 nm. The static spectrum
at 30 min (Fig. 4(B)) appears as a sum of the monomer and
excimer spectra. The alternate sequence (injection of compound
to a solution that already contains vesicles) results in spectra that
are comparable to Fig. 4(B) and evolve as the right-hand portion
of Fig. 4(C). Over long times, the final state is independent of the
sequence of mixing but progress to the final state is slow (minutes
to hours).
The Hille equation42 relates the observed conductance to an
apparent radius of the ion channel.39 Based on an assumed channel
length of 3.4 nm, the openings of Fig. 3(A) have an apparent radius
of 0.31 nm while the smaller conductance changes in Fig. 3(C)
have an apparent radius of 0.11 nm. On the same assumptions,
the spikes must have radii up to 1 nm. The underlying step-
conductance features give radii in the range 0.34–0.45 nm for the
four compounds in KCl electrolyte.
We interpret the time-dependent sequence of Fig. 4(C) as
follows. Before the addition of vesicles, the aggregate in water is
5144 | Org. Biomol. Chem., 2010, 8, 5139–5149
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
unstable and slowly evolves in structure which alters the emission
spectra. Upon vesicle addition, available aqueous monomer is
rapidly partitioned to the vesicle where it can emit; a portion
of the immediate intensity change is also due to a change in
the scattering of the solution. The monomer emission observed
must arise predominantly from the compound in vesicles, and
increases over time as compound migrates from the aqueous
phase. Partition of monomer depletes the amount of aqueous
aggregate thereby increasing the monomer emission intensity at
the expense of the excimer emission intensity. The slow decline
over the 200 s following vesicle addition, followed by reversal,
indicates that the rate is controlled by the disassembly of the
aqueous aggregate. Eventually, the monomer concentration in the
vesicle bilayer grows to a concentration where excimer within the
bilayer can form. Thereafter the monomer concentration in the
bilayer reaches a steady state while the excimer emission intensity
increases as further dissolution of aqueous aggregate proceeds.
In the limit, the excimer and monomer emissions will become
constant as the compound is fully dispersed between its various
states in the microheterogeneous mixture.
In contrast, the corresponding spectra for the longest compound
(HO2C-Dec-Dip-Hex-G(12)–OH; Fig. 4(D)) show a sharp change
on vesicle addition, but thereafter the migration dynamics are
either very much slower (or faster) as the system does not
appear to evolve at all; quenching studies (see below) establish
that HO2C-Dec-Dip-Hex-G(12)–OH undergoes very slow transfer
from aggregate to vesicle. The positional isomers of HO2C-Dip-
Hex-Hex-G(12)–OH lie between the two extremes of Fig. 4(C)
and 4(D); they appear to transfer somewhat more slowly than
HO2C-Dec-Dip-Hex-G(12)–OH, but clearly do transfer from an
aggregate in water to a vesicle environment (see the ESI†).
excimer/aggregate near neutral pH would be a polyanionic species,
so an electrostatic association of the divalent copper ion to
this aggregate is very likely. This would produce a very efficient
quenching at low concentrations of copper. A similar mechanism
has been reported for the efficient quenching of conjugated anionic
polyelectrolyte aggregates based on poly(phenylene ethynylene) by
micromolar cationic quencher concentrations.44 The quenching
by copper is pH dependent with higher copper concentrations
required as the pH and buffer concentrations increase (see the
ESI†).
Key experiments are illustrated in Fig. 5; compound (HO2C-
Dip-Hex-Hex-G(12)–OH in Fig. 5(A); HO2C-Dec-Dip-Hex-
G(12)–OH in Fig. 5(B)) was added to aqueous solution and
the aggregation process was allowed to occur over a 5 min
period, followed by injection of vesicles and a further period of
redistribution. At this point, aqueous CuSO4 was added to a final
concentration of 1 mM. In the case of HO2C-Dip-Hex-Hex-G(12)–
OH there is a small dilution perturbation to the emission spectrum,
but excimer fluorescence at 380 nm is virtually unchanged. In
contrast, the excimer emission from the system containing HO2C-
Dec-Dip-Hex-G(12)–OH is substantially quenched by the aqueous
copper. In both cases, the observed monomer emission is only
changed by a dilution factor as the copper concentration is well
below the level at which monomer quenching could be expected
even if the monomer were completely accessible to the quencher.
Fig. 5(C) and 5(D) illustrate the effect of added copper ion
on the full emission spectrum; while HO2C-Dip-Hex-Hex-G(12)–
OH is markedly quenched in water (4C(a)) and unquenched
in vesicles (4C(b)), HO2C-Dec-Dip-Hex-G(12)–OH undergoes
almost identical extent of quenching whether in water alone
(4D(a)) or in the presence of vesicles (4D(b)). The quenching
experiments can also be conducted under conditions where the
compounds and vesicles are mixed in the inverse order (see the
The rates of these redistribution processes are slow; do they have
any relevance to the transport processes detected using an HPTS
assay? The fluorescence of the Dip-containing compound can be
monitored concurrently with the HPTS assay. The results (see the
ESI†) confirm that the transport process for an active compound
such as HO2C-Dip-Hex-Hex-G(12)–OH and the redistribution
process occur concurrently. Redistribution is slower than the
collapse of the pH gradient reported by the HPTS assay; in a zero
order case the derived rate would not be significantly influenced
while the apparent first-order rate will incorporate the two but be
dominated by the transport.
An examination of the fluorescence quenching behaviours
was expected to yield information on the localization of the
fluorophore. If the Dip was located deep in a hydrophobic region
it should be relatively inefficiently quenched by a hydrophilic
quencher in the aqueous phase. After some initial trials, the
quenching by aqueous copper(II) was chosen for systematic
examination. In homogeneous methanol solution quenching of
the monomer (320 nm) emission occurs in the range of 0.1–1 mM
added CuSO4 to yield Stern–Volmer constants of about 1 ¥ 103 M-1
(see the ESI†). This order of magnitude is consistent with a simple
diffusion controlled quenching of an excited state with a lifetime of
the order of nanoseconds as previously determined.32 The differ-
ences between the compounds are negligible in comparison to the
dramatically more efficient quenching of the excimer emission of
aqueous aggregates by CuSO4 in the concentration range of 0.1–1
mM (see the ESI†). The system involving aggregates has complex
dynamics and a substantial static quenching contribution. The
Fig. 5 Upper panel: time-based emission at 320 and 380 nm for (A)
HO2C-Dip-Hex-Hex-G(12)–OH and (B) HO2C-Dec-Dip-Hex-G(12)–OH,
(lex = 305 nm). The compounds are initially in aqueous solution
(0.1 M NaCl, unbuffered), to which lipid vesicles (time i) and 1 mM
CuSO4 (time ii) are added. Lower panel: fluorescence spectra of (C)
HO2C-Dip-Hex-Hex-G(12)–OH and (D) HO2C-Dec-Dip-Hex-G(12)–OH,
(lex = 305 nm) in aqueous solution (a), or with the addition of lipid vesicles
(b), in the absence (black lines) or presence (grey lines) of 1 mM CuSO4.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5145
©

View Article Online
ESI†). The sequence homologs of HO2C-Dip-Hex-Hex-G(12)–
OH all behave similarly in that excimer emission after some period
in contact with vesicles cannot be significantly quenched by low
levels of copper. The compound HO2C-Dec-Dip-Hex-G(12)–OH
can be incorporated into vesicles during preparation. Under these
conditions there is strong monomer fluorescence at 320 nm (see the
ESI†). The intensity of the excimer emission from these vesicles is
too low for a copper ion quenching study; the monomer emission
is not quenched up to 1 mM added CuSO4.
A simple interpretation of these results is that the lack of
quenching by copper ion indicates a membrane location of the
compound such that the excimer is no longer accessible to aqueous
quencher, due to incorporation in a hydrophobic region of the
bilayer. In the case where quenching in the presence of vesicles is
observed, the excimer would arise from aqueous-phase aggregate
that has not redistributed to the vesicle bilayer. This interpretation
has clear mechanistic implications, but must be handled with
caution as the role of static quenching has not yet been fully
elucidated.
the intermediate is an aqueous monomer as given in Fig. 1, but
there is no direct spectroscopic evidence for this species.
Turning to the pores formed, there is evidence that oligomeric
species are required. The high Hill coefficients are consistent with
this interpretation, but would also be consistent with a rate-
limiting direct transfer of oligomers (not monomers) from an
aqueous aggregate to an active state in the bilayer. However,
this oligomer transfer is inconsistent with the voltage clamp
experiments in which the channels form from transporters held
within the bilayer. A pathway via monomer insertion to the bilayer
remains as both the simplest alternative and the one consistent
with all data available. The fluorescence results are consistent
with, but do not compel, a linkage between the observed excimer
emission and the transport activity. Certainly, excimer emission
is observed under conditions where transport can or does occur.
And transport is only observed under conditions where excimer
can also be observed. However, a direct correlation of excimer
emission intensity with transport activity in vesicles is not possible
in this system as the emission can also arise from the competing
aqueous aggregate.
The voltage clamp data indicate that a wide range of trans-
membrane conducting species can be formed and that these
species depend upon the electrolyte; more “regular” channels
are observed with CsCl as electrolyte than with KCl. This
suggests that transported ions are also essential in maintaining
pore structure. This finding precludes some types of experiments
that better defined ion channels will support. For example, the
ionic selectivity of ion channels is typically determined from the
“reversal potential”;1 this is not possible in this system as the
nature of the species being probed is a function of the bathing
electrolyte composition.
The location of the Dip unit plays a key role in channel
formation. The most regular and stable conducting states are
formed with the Dip unit located in the middle of the oligoester
strand HO2C-X-Dip-Y-G(12)–OH, followed by the carboxyl-
terminal location. A location of the Dip adjacent to the hydroxyl
terminus does not preclude the formation of conducting states,
but these are uniformly short-duration “spikes” of highly variable
conductance. The quenching data support a buried location for
the membrane-bound Dip rather than a surface accessible site. The
Dip could be accessible to water as part of a pore structure and
would still not show quenching since the ion currents supported
by these pores are high and the conjunction of an excitation at the
time of a quencher passing nearby has a low probability.
The properties of a bilayer environment as a solvent constrain
a discussion of the structures of the active species and their
precursors. Although crystal structures of lipids show neatly tilted
hydrocarbon chains arranged in bilayers,47 molecular dynamics
simulations and spectroscopic probes suggest a much more chaotic
environment.48 Nonetheless, individual lipid molecules retain a
tilt with respect to the membrane normal on a time-average, and
solutes with a long molecular axis dissolve in an orientation with
the long axis aligned with the average orientation of the lipids.49
Within the bilayer there is considerable motion, but headgroup,
midpolar, and hydrocarbon regions are maintained in a time
average.2,3 The compounds discussed here can certainly fold on
themselves, but in a membrane environment would be expected to
align with the lipids as illustrated in the very idealized sketch in
Fig. 6. The longest compound (HO2C-Dec-Dip-Hex-G(12)–OH) is
Discussion and conclusions
Mechanistic proposals for the formation of ion channels from
flexible acyclic compounds similar to the ones discussed here have
centred primarily on the structure of the active species.24,25,45,46
Although this is certainly an interesting issue, “mechanism” in
this context also involves a number of other aspects: which
species are present in the system, how do they interconvert,
what are the relative energies and rates of interconversion?1–3
The framework sketched in Fig. 1 is focused on this latter set of
questions and the collection of data presented in this report probes
these questions directly. Bilayer-bound states are required by the
transport results that show that all four compounds are capable of
provoking channel behaviour when premixed in lipid as detected
by the voltage clamp technique. The proposed20 aqueous phase
aggregate is directly observed in solutions without vesicles as well
as in contact with vesicles (pyrene assay, excimer fluorescence). A
membrane associated monomer can be formed by incorporation
during vesicle preparation for HO2C-Dec-Dip-Hex-G(12)–OH, or
by transfer from aqueous solution in the case of the other three
compounds. The aqueous aggregates in the absence of vesicles for
all four compounds show relatively weak monomer fluorescence;
therefore the monomer emissions observed in the presence of
vesicles must be due to a membrane-associated monomer. Finally,
the time-dependent and quenching experiments reported suggest
the rates at which the various equilibria can operate. Thus
the species and processes identified in Fig. 1 now have some
experimental basis.
The aqueous phase aggregate is in competition with, and
controls, the transport process. This is clear in the case of HO2C-
Dec-Dip-Hex-G(12)–OH in vesicles which stalls as the aqueous
aggregate and does not progress to membrane incorporation. This
results in this compound appearing to be inactive in the vesicle
transport assay, even though it is certainly capable of forming
channels once it is in a lipid bilayer. The other compounds are more
active in vesicles, but even in these cases the transfer of material
from an aqueous aggregate to a vesicle bilayer is a relatively slow
process as shown by the time dependence of the fluorescence
changes that accompany this process. We continue to assume that
5146 | Org. Biomol. Chem., 2010, 8, 5139–5149
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
is clear that the competing aqueous aggregate limits the activity
of the compounds. Ongoing efforts to produce highly active pore-
formers with this type of flexible lipophilic compound are directed
to avoiding this species.
Experimental
General synthesis procedures
Most chemicals and solvents were used as received from known
suppliers, except THF and DMF which were dried and distilled
before use. NMR spectra were collected on a 300 MHz Bruker or
500 MHz Varian instrument. UV spectra were run on a Cary 5 UV-
VIS spectrometer in a 10 ¥ 10 mm quartz cell. ESI Mass spectra
were recorded on a Waters MicroMass Q-TOF instrument running
in negative ion mode. HPLC was performed using an HP Series
1100 instrument, with either a Macherey-Nagel “Nucleosil” RP
C18 analytical (4 mm ¥ 250 mm) or a Grace Davison “Alltima” RP
C18 semi-prep (10 mm ¥ 150 mm) column. Solvents used (ACN,
CH3OH; HPLC-grade, H2O; Milipore) were filtered through a
Milipore sub-micrometre filter before use. HPLC elution was
monitored at various UV wavelengths (typically 254, 280 and 220
nm) and fluorometrically (lex = 310, lem = 330 nm). Fluorescence
spectra were run on a PTI QM-2 instrument at T = 20 ◦C, slit
width = 3 nm in 10 ¥ 10 mm quartz cells equipped with a micro
stir rod.
Sonogashira coupling (i): To a round bottomed flask equipped
with a septum, 1.2–1.5 equivalents (in relation to the alkyne
starting material) of the iodo-containing reactant and 6–10% CuI
were dissolved in dry DMF, which was then deoxygenated under
vacuum. The alkyne reactant was then added to the flask, which
was kept in the dark. Pd(PPh3)4 (3–5%) was then added, followed
by 2–5 equivalents of NEt3. The reaction was then stirred under N2
at temperatures ranging from rt to 50 ◦C, for 10–24 h, depending
on the reagents used. Reactions were monitored by TLC (silica
gel, EtOAc/hexanes as eluent, visualized by UV, p-anisaldehyde
and/or vanillin stain). Once complete, reactions were cooled if
necessary, diluted with EtOAc and washed with saturated EDTA
(2–3 times), H2O (once), saturated NaCl (once), dried over sodium
sulfate, and concentrated under vacuum. Unless noted otherwise,
the crude products were purified by column chromatography on
silica gel, typically using EtOAc/hexanes as eluent.
Ester coupling (ii): To a solution of 1.3–2 equivalents of either
the alcohol or acid building block in relation to 1 equivalent of
the other (excess reagent choice determined by ease of synthesis or
availability) in THF or DMF (depending on substrate) were added
1.3–2 eq. of DIC, HOBt and 2.6–4 eq. of DIPEA. The reaction
was sealed under an atmosphere of N2, and stirred for 16–24 h at
temperatures varying between rt and 50 ◦C. Reaction completion
was monitored by TLC. Once complete, the reaction was cooled
(if necessary), filtered to remove DIU, and diluted with DCM
or EtOAc. The organic phase was extracted with H2O (twice),
saturated NaHCO3 (twice), rinsed with sat NaCl (once), dried
with anhydrous sodium sulfate, and concentrated under vacuum.
Unless noted otherwise, the crude product was purified by column
chromatography on silica gel, typically using EtOAc/hexanes as
eluent.
Fig.
6
Schematic of a transmembrane orientation of HO2C-X-
Dip-Y-G(12)–OH in comparison to an idealized lipid bilayer.
approximately the correct length to span the bilayer hydrocarbon
region, The shorter HO2C-Hex-Dip-Hex-G(12)–OH (illustrated
in part in Fig. 6) would require some deformation and membrane
thinning to span the bilayer, as is proposed for the well-studied
gramicidin dimer channel.50,51 In this type of transmembrane
orientation, these positional isomers place the diphenylacetylene
crossing the bilayer midplane. The other positional isomers in a
similar orientation place the Dip within a single bilayer leaflet,
leaving the more flexible methylene chains to cross the bilayer
midplane. These transmembrane orientations are akin to lipid
interdigitation in which a longer lipid molecule intrudes into
the opposite leaflet.52 In lipid systems, interdigitation leads to
microphase separation and consequent rigidifying effects on lipid
dynamics.48
The observed channels could be produced from this trans-
membrane orientation of monomer through a similar phase-
separation mechanism. The separated domain would contain
monomer in both possible orientations together with water, ions,
and potentially a few lipid molecules. The differences between
channels as a function of overall length, Dip position and
electrolyte would then relate to the energetics and dynamics of
these domains. The CsCl/HO2C-Dec-Dip-Hex-G(12)–OH system
produced small, well-defined but short-lived channels (Fig. 3(C)).
This compound is the longest, so would require the least reorgani-
zation of the headgroups in order to attain a conducting state and
could form and dissociate readily from a small phase-separated
domain of a few molecules. The shorter homolog HO2C-Hex-
Dip-Hex-G(12)–OH produces longer-lived, and apparently larger
channels (Fig. 3(A)/(B)) with lower probability reflecting a more
challenging reorganization of the system before ionic conduction
can occur. The more irregular behaviours observed with KCl
electrolyte reflect the higher hydration energy of the cation in
addition to any membrane reorganizations required. The result
is a less defined domain that would allow highly variable transient
pores to open/close quickly.
This picture of the channels formed in this type of system is
more fluid than the proposals associated with larger and more
conformationally constrained examples. Although many details
remain obscure, the working hypothesis outlined in Fig. 1 appears
to be well-supported by a range of direct and indirect probes. It
THP removal (iii): pTsOH (5–25%) was added to a solution of
compound in ~10–30% CH3OH: DCM, which was stirred at rt
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5147
©

View Article Online
for 1–3 h, as monitored by TLC. Once complete, the reaction was
diluted with DCM, washed with H2O (once), saturated NaHCO3
(twice), and saturated NaCl (once), then dried over sodium sulfate
and concentrated under vacuum. If necessary, further purification
was carried out as noted.
compound was pre-incorporated into the vesicle bilayer by adding
a solution of the compound of interest to the lipid mix before
continuing the preparation as usual. Regardless of the identity of
the vesicles, the volume of vesicle stock solution added to the cell
was kept constant at 100 mL for each experiment.
Fluoride deprotection (iv): ~5–10 equivalents of 1.0 M TBAF in
THF were added to a solution of compound in THF, which was
then stirred at rt for 1–3 h, as monitored by TLC. Once complete,
the reaction was quenched by the addition of H2O, diluted into
EtOAc, washed with 10% aqueous HCl, (once), then H2O (until
neutral) and saturated NaCl (once), then dried over sodium sulfate
and concentrated under vacuum. Further purification was carried
out as noted.
Prenyl deprotection (v) (adapted53): 0.025–0.1 equivalents TM-
SOTf were added to the compound dissolved in DCM, which was
stirred at rt. Once complete (as monitored by TLC, generally <
1 h), the reaction mixture was diluted further into DCM, washed
with H2O, saturated NaHCO3 and saturated NaCl, then dried
over sodium sulfate and concentrated under vacuum. Further
purification was carried out as noted.
Acknowledgements
The ongoing support of the Natural Science and Engineering
Research Council of Canada is gratefully acknowledged.
Notes and references
1 S. Matile and N. Sakai, in Analytical Methods in Supramolecular
Chemistry, ed. C. A. Schalley, Wiley-VCH, Weinheim, 2007, pp. 381–
418.
2 G. W. Gokel and I. A. Carasel, Chem. Soc. Rev., 2007, 36, 378–389.
3 T. M. Fyles, Chem. Soc. Rev., 2007, 36, 335–347.
4 L. You, R. Ferdani, R. Li, J. P. Kramer, R. E. K. Winter and G. W.
Gokel, Chem.–Eur. J., 2008, 14, 382–396.
5 C. P. Wilson and S. J. Webb, Chem. Commun., 2008, 4007–4009.
6 L. Ma, M. Melegari, M. Colombini and J. T. Davis, J. Am. Chem. Soc.,
2008, 130, 2938–2939.
7 M. Woodbury, M. M. Mongare, C. D. Hall, Z. Wang, B. Draghici,
A. R. Katritzky, M. Tsikolia, A. C. Hall, C. Suarez, Z. O. Nylander,
S. M. Wardlaw, M. E. Gibson, K. L. Valentine, L. N. Onyewadume and
D. A. Ahove, Org. Biomol. Chem., 2009, 7, 3862–3870.
8 O. Lawal, K. S. J. Iqbal, A. Mohamadi, P. Razavi, H. T. Dodd, M. C.
Allen, S. Siddiqui, F. Fucassi and P. J. Cragg, Supramol. Chem., 2009,
21, 55.
9 O. V. Kulikov, R. Li and G. W. Gokel, Angew. Chem., Int. Ed., 2009,
48, 375–377.
10 N. Illy, L. Bacri, J. Wojno, D. Destouches, B. Brissault, J. Courty, L.
Auvray, J. Penelle and V. Barbier, Macromol. Symp., 2010, 287, 60–68.
11 A. Hennig, L. Fischer, G. Guichard and S. Matile, J. Am. Chem. Soc.,
2009, 131, 16889–16895.
12 B. Gong, A. J. Helsel, A. L. Brown, K. Yamato, W. Feng, L. Yuan, A. J.
Clements, S. V. Harding and Z. Shao, J. Am. Chem. Soc., 2008, 130,
15784–15785.
13 L. Ma, W. A. Harrell and J. T. Davis, Org. Lett., 2009, 11, 1599–1602.
14 S. Hagihara, L. Gremaud, G. Bollot, J. Mareda and S. Matile, J. Am.
Chem. Soc., 2008, 130, 4347–4351.
Spectroscopic and characterization data for all new compounds
is given in the ESI.†
Transport studies
All samples used in transport experiments were HPLC-purified.
The HPTS transport assay was done as previously reported;20
minor changes to the published procedure include the use of
a different buffer (100 mM NaCl, 10 mM Na3PO4·12H2O, pH
~6.4) and the injection of compound through a specially-designed
injection port, which allows constant monitoring without the need
for opening the instrument. Vesicles were made as reported, and
sized on a Brookhaven Instruments ZetaPALS instrument. Aver-
age vesicle size was found to be ~180–200 mm. The experimental
and data analysis systems for the voltage clamp experiments has
been previously described.40,41 All data were hardware filtered (8-
pole Bessel filter, 1 kHz) and in a survey mode were filtered during
acquisition at 1 kHz. The rapid spike and the flicker behaviors are
near the 1 kHz limit and data for more detailed examination did
not use the digital acquisition filter.
15 X. Li, B. Shen, X.-Q. Yao and D. Yang, J. Am. Chem. Soc., 2009, 131,
13676–13680.
16 W. Wang, R. Li and G. W. Gokel, Chem.–Eur. J., 2009, 15, 10543–10553.
17 W. Wang, R. Li and G. W. Gokel, Chem. Commun., 2009, 911–913.
18 X. Li, B. Shen, X.-Q. Yao and D. Yang, J. Am. Chem. Soc., 2007, 129,
7264–7265.
19 K. S. J. Iqbal, M. C. Allen, F. Fucassi and P. J. Cragg, Chem. Commun.,
Fluorescence studies
2007, 3951–3953.
20 T. M. Fyles and H. Luong, Org. Biomol. Chem., 2009, 7, 733–738.
21 T. M. Fyles and H. Luong, Org. Biomol. Chem., 2009, 7, 725–732.
22 T. M. Fyles, C. Hu and H. Luong, J. Org. Chem., 2006, 71, 8545–8551.
23 T. M. Fyles, D. Loock, W. F. van Straaten-Nijenhuis and X. Zhou,
J. Org. Chem., 1996, 61, 8866–8874.
24 T. K. Rostovtseva, C. L. Bashford, A. A. Lev and C. A. Pasternak,
J. Membr. Biol., 1994, 141, 83–90.
25 T. Renkes, H. J. Schafer, P. M. Siemens and E. Neumann, Angew. Chem.,
Int. Ed., 2000, 39, 2512–2516.
26 T. M. Fyles, D. Loock and X. Zhou, J. Am. Chem. Soc., 1998, 120,
2997–3003.
27 P. Reiss, L. Al-Momani and U. Koert, ChemBioChem, 2008, 9, 377–379.
28 G. A. Woolley, V. Zunic, J. Karanicolas, A. S. I. Jaikaran and A. V.
Starostin, Biophys. J., 1997, 73, 2465–2475.
29 J. M. Kuiper, R. T. Buwalda, R. Hulst and J. B. F. N. Engberts,
Langmuir, 2001, 17, 5216–5224.
30 E. Abel, G. E. M. Maguire, O. Murillo, I. Suzuki, S. L. De Wall and
G. W. Gokel, J. Am. Chem. Soc., 1999, 121, 9043–9052.
31 C. Ferrante, U. Kensy and B. Dick, J. Phys. Chem., 1993, 97, 13457–
13463.
All oligomers studied were purified by HPLC prior to use. Stock
concentrations of approximately 1–6 mM in THF were stored
under N2 in the freezer when not in use, and periodically re-checked
by HPLC. Volumes of compound in THF used in each experiment
did not exceed 25 mL added into typically 2 mL of aqueous
solution. The volume of solutions was kept approximately at 2 mL
for all experiments. For studies in organic solvents, the solvents
used were of spectral or higher quality, and were purged with N2.
For all aqueous studies except those involving CuSO4, the aqueous
buffer consisted of 10 mM Na3PO4·12H2O, 100 mM NaCl, pH =
6.4. For quenching studies with CuSO4 in aqueous solution, either
100 mM unbuffered NaCl or 10 mM Bis-Tris, 100 mM NaCl, pH =
6.4 were used (as noted). Vesicles were prepared and sized as above,
however, rather than always containing the HPTS dye, in certain
cases, thevesicles weremade withthe same buffer bothinternal and
external to the vesicle. Alternatively, for certain experiments, the
32 Y. Hirata, Bull. Chem. Soc. Jpn., 1999, 72, 1647–1664.
5148 | Org. Biomol. Chem., 2010, 8, 5139–5149
This journal is
The Royal Society of Chemistry 2010
©

View Article Online
33 R. L. Letsinger, T. Wu, J.-S. Yang and F. D. Lewis, Photochem.
Photobiol. Sci., 2008, 7, 854–859.
34 R. Chinchilla and C. Na´jera, Chem. Rev., 2007, 107, 874–922.
35 P. J. Kocien´ski, Protecting groups, 3rd edn, Georg Theime Verlag,
Stuttgart, 2005.
44 H. Jiang, X. Zhao and K. S. Schanze, Langmuir, 2007, 23, 9481–
9486.
45 Y. Kobuke and K. Morita, Inorg. Chim. Acta, 1998, 283, 167–174.
46 Y. Kobuke, K. Ueda and M. Sokabe, J. Am. Chem. Soc., 1992, 114,
7618–7622.
36 C. L. Kuyper, J. S. Kuo, S. A. Mutch and D. T. Chiu, J. Am. Chem.
47 H. Hauser and G. Poupart, in The Structure of Biological Membranes,
ed. P. L. Yeagle, CRC Press, Boca Raton, 2 edn, 2005, pp. 1–51.
48 K. Gawrisch, in The Structure of Biological Membranes, ed. P. L. Yeagle,
CRC Press, Boca Raton, 2 edn, 2005, pp. 147–171.
49 J. M. Pope, L. W. Walker and D. Dubro, Chem. Phys. Lipids, 1984, 35,
259–277.
Soc., 2006, 128, 3233–3240.
37 M. E. Weber, P. H. Schlesinger and G. W. Gokel, J. Am. Chem. Soc.,
2005, 127, 636–642.
38 Single-channel Recording, ed. B. Sakmann and E. Neher, New York,
London, 1983.
39 T. M. Fyles and C. C. Tong, New J. Chem., 2007, 31, 655–661.
40 M. B. Buchmann, T. M. Fyles and T. Sutherland, Bioorg. Med. Chem.,
2004, 12, 1315–1324.
50 G. A. Woolley and B. A. Wallace, Journal of Membrane Biology, 1992,
129, 109–136.
51 O. S. Andersen, R. E. Koeppe and B. Roux, IEEE Trans. NanoBiosci.,
2005, 4, 10–20.
52 J. L. Slater and C.-H. Huang, in The Structure of Biological Membranes,
ed. P. L. Yeagle, CRC Press, Boca Raton, 2nd edn, 2005, pp.
122–145.
41 P. K. Eggers, T. M. Fyles, K. D. D. Mitchell and T. Sutherland, J. Org.
Chem., 2003, 68, 1050–1058.
42 B. Hille, Ionic Channels of Excitable Membranes, 3rd edn, Sinauer
Associates, Incorporated, Sunderland, 2001.
43 K. Kalyanasundaraman and J. A. Thomas, J. Am. Chem. Soc., 1977,
99, 2039–2044.
53 M. Nishizawa, H. Yamamoto, K. Seo, H. Imagawa and T. Sugihara,
Org. Lett., 2002, 4, 1947–1949.
This journal is
The Royal Society of Chemistry 2010
Org. Biomol. Chem., 2010, 8, 5139–5149 | 5149
©