Dynamic octopus amphiphiles as powerful activators of dna transporters: Differential fragrance sensing and beyond

Article abstract of DOI:10.1002/chem.201001352

We report the design, synthesis and evaluation of dynamic "octopus" amphiphiles with emphasis on their efficiency as activators in synthetic membrane-based sensing systems. Previously, we found that the in situ treatment of charged hydrazides with hydrophobic aldehydes or ketones gives amphiphilic counterion activators of polyion transporters in lipid bilayers, and that their efficiency increases with the number of their hydrophobic tails. Herein, we expand this series to amphiphiles with one cationic head (guanidinium or ammonium) and four exchangeable hydrophobic tails. These results, with the highest number of tails reported to date, confirm that dynamic octopus amphiphiles provide access to maximal activity and selectivity. Odorants, such as muscone, carvone, or anisaldehyde are used to outline their usefulness in differential sensing systems that operate based on counterion-activated DNA transporters in fluorogenic vesicles. The enhanced ability of octopus amphiphiles to enable the discrimination of enantiomers as well as that of otherwise intractable ortho, meta, and para isomers and short cyclo-/alkyl tails is demonstrated. These findings identify dynamic octopus amphiphiles as being promising for application to differential sensing, "fragrant" cellular uptake, and slow release.

Full text of DOI:10.1002/chem.201001352

FULL PAPER
DOI: 10.1002/chem.201001352
Dynamic Octopus Amphiphiles as Powerful Activators of DNA Transporters:
Differential Fragrance Sensing and Beyond
Javier Montenegro, Pauline Bonvin, Toshihide Takeuchi, and Stefan Matile*[a]
Abstract: We report the design, synthe-
sis and evaluation of dynamic “octo-
pus” amphiphiles with emphasis on
their efficiency as activators in synthet-
ic membrane-based sensing systems.
Previously, we found that the in situ
treatment of charged hydrazides with
hydrophobic aldehydes or ketones
gives amphiphilic counterion activators
of polyion transporters in lipid bilayers,
and that their efficiency increases with
the number of their hydrophobic tails.
Herein, we expand this series to am-
phiphiles with one cationic head (gua-
nidinium or ammonium) and four ex-
changeable hydrophobic tails. These re-
sults, with the highest number of tails
reported to date, confirm that dynamic
octopus amphiphiles provide access to
maximal activity and selectivity. Odor-
ants, such as muscone, carvone, or anis-
that operate based on counterion-acti-
vated DNA transporters in fluorogenic
vesicles. The enhanced ability of octo-
pus amphiphiles to enable the discrimi-
nation of enantiomers as well as that of
otherwise intractable ortho, meta, and
para isomers and short cyclo-/alkyl tails
is demonstrated. These findings identi-
fy dynamic octopus amphiphiles as
being promising for application to dif-
ferential sensing, “fragrant” cellular
uptake, and slow release.
AHCTUNGTRENNUNG
Keywords: amphiphiles
· DNA ·
membranes · sensors · supramolec-
ular chemistry
Introduction
other hand, modulate the activity of the synthetic transport
system.[24,25] However, synthetic sensing systems in bilayer
membranes that operate with enzymes (or DNA apta-
Olfactory sensing systems operate in lipid bilayer mem-
branes.[1] However, the creation of synthetic sensing systems
that work in the same way as in nature has been trouble-
some.[2–4] The synthesis of transport systems[5–17] that respond
to chemical stimulation was the first key step to achieve this
goal.[18–20] In sensing systems, these stimuli-responsive trans-
porters function as signal transducers. Their combination
with enzymes as signal generators[21,22] afforded the first
functional sensors that are operational in complex matrices
from the supermarket to hospitals.[23,24] General applicability
as multianalyte sensing systems, finally, became possible
with the introduction of reactive in-/activators as bifunction-
al molecules that can, on the one hand, covalently capture
the product of enzymatic signal generation and, on the
AHCTUNGTRENNUNG
mers)[26] as signal generators will never reach the sensing
power of mammalian olfactory systems, where 350 receptors
suffice to sense more than 10000 odorants.[1]
To get there, we have recently introduced pattern recogni-
tion to synthetic systems that work in lipid bilayer mem-
branes (Figure 1).[27] In pattern recognition or “differential”
sensing approaches, single analytes are identified and quan-
tified not from molecular recognition by one specific sensor
but as a unique composite response or fingerprint.[28–40] Pat-
tern recognition is attractive for signal generation because
promiscuous behavior of individual sensors is not only toler-
ated but preferable, and an infinite number of analytes can
be detected with a relatively small number of individual
sensing units. Pattern-recognition approaches have been ap-
plied previously to several chemosensor systems[28–46] that
operate with indicator displacement assays, stimuli-respon-
sive dyes, polymer absorption, or electrochemical reac-
tions.[28–40] Topics addressed with differential sensing include
not only analyte families, such as anions, sugars, amino
acids, or nucleotides but also challenges, such as the discrim-
ination of stereoisomers or the creation of electronic, colori-
metric or fluorescent artificial tongues and noses.
[a] Dr. J. Montenegro, P. Bonvin, Dr. T. Takeuchi, Prof. S. Matile
Department of Organic Chemistry
University of Geneva, Geneva, Switzerland
Fax : (+41)22 379 3215
E-mail: stefan.matile@unige.ch
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem201002352.
Chem. Eur. J. 2010, 16, 14159 – 14166
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
14159

prints needed for differential
sensing.[27] Incubation of ana-
lytes with different reactive
counterions produces a small
collection of in-/activators of
synthetic transport systems
(Figure 1). From the fluorescent
response to increasing activator
concentrations, characteristics,
such as the maximum normal-
ized response as determined
after 190s (YMAX), the effective
concentration to reach 50% of
this maximal activity (EC50), or
the Hill coefficient (n) are ex-
tracted to generate analyte spe-
cific patterns. These multidi-
mensional patterns are then
subjected to principal compo-
nent analysis (PCA) and hier-
archical cluster analysis (HCA)
to convert n-dimensional space
into
three-dimensional
PC
space and 2D dendrograms, re-
spectively.
This approach to differential
sensing systems that work like
olfactory systems in lipid bilay-
er membranes has been exem-
plified with hydrophobic ana-
lytes containing a ketone or an
aldehyde, such as odorants O1–
O13 (Scheme 1).[27] Without
modification, these analytes are
essentially undetectable with
most synthetic sensing systems
that operate in lipid bilayer
membranes. However, their in-
cubation with cations G1H1–
Figure 1. Scheme for synthetic sensing systems in fluorogenic vesicles. Hydrophobic analytes (e.g., muscone
O1) are covalently captured by hydrophilic countercations (e.g., tetrahydrazides G1H4) to give four-tailed oc-
topus amphiphiles (e.g., G1H4O1) that can activate polyanions (e.g., ctDNA) as transporters in lipid bilayer
membranes. Transporter activation by hydrophobic analytes is reported as increase in the emission of internal
anionic fluorophores in response to the export of cationic quenchers (or both probes).
Differential sensing has not been applied to synthetic sys-
tems that operate in lipid bilayers because the number of
available cross-responsive sensor components has been in-
sufficient.[3,4] To overcome this limitation, we have consid-
ered the following sensing scheme.[27] Hydrophobic analytes,
otherwise undetectable with membrane-based sensing sys-
tems, are first covalently captured by reactive hydrophilic
cations (Figure 1). The resulting amphiphiles can act as
countercation activators for polyanions and generate active
polyion–counterion transport systems. Their change in activ-
ity caused by the hydrophobic analyte can be monitored by
several different methods. Arguably the simplest technique
is to follow fluorescent recovery while cationic quenchers
(and/or anionic fluorophores) are carried out of fluorogenic
vesicles.
G1H3 and A1H1–A1H3 that contain at least one hydrazide
gives hydrazones, such as G1H1O1–G1H3O13, and
A1H1O1–A1H3O13, which in turn can activate polyanions,
such as calf thymus DNA (ctDNA) as cation transporter in
fluorogenic vesicles (Figures 1 and Scheme 1). This new
sensing system can differentiate between enantiomers (e.g.,
enantioenriched (R)-(À)-muscone O1 (61% ee) versus race-
mic muscone O2, R-(À)-carvone O9 (caraway) versus S-
(+)-carvone O10 (spearmint)), cis–trans isomers, single
atom homologues (e.g., O5–O8) or perfumes, as examples
of more complex matrices used in daily life.
Sensing sensitivity is best with long, linear, or branched
alkyl or aryl tails including O6–O8, whereas shorter alkyl or
aryl tails, such as O3–O5, O11–O13, and cyclic ketones, such
as O1–O2, O9–O10, are more difficult to detect under stan-
dard conditions. However, such analytes could be sensed as
competitive inhibitors of other more easily detectable odor-
ants in inverse detection schemes.
For pattern generation, a small collection of reactive
counterions G1H1–G1H3 and A1H1–A1H3 has been de-
signed, synthesized, and used to produce the unique finger-
14160
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 14159 – 14166

Differential Fragrance Sensing
FULL PAPER
cost of more demanding syn-
thetic efforts. The synthesis
route to multihydrazide coun-
terions was designed as a mod-
ular approach based on routine
peptide chemistry to maximize
user-friendly simplicity. Tetra-
hydrazide G1H4 was synthe-
sized from Z-protected gluta-
mate 1 in six steps, overall, with
20% total yield (unoptimized)
and under mild conditions
(Scheme 2). Namely, treatment
of 1 with glutamate diester 2 af-
forded the desired tetravalent
scaffold in a single step. The
obtained tetraester 3 was treat-
ed with hydrazine to give
nearly intractable tetrahydra-
zides. However, conversion of
the crude product with Boc-an-
hydride readily afforded the
Boc-protected tetrahydrazides
4, and the desired peptide den-
dron[47–51] scaffold was purified
without problems under stan-
dard conditions. Selective Z re-
moval by hydrogenolysis gave
Scheme 1. New reactive counterions A1H4 and G1H4 for differential sensing systems in fluorogenic vesicles
with odorant analytes that are not well detectable with the previously reported A1H1–A1H3 and G1H1– amine 5, which was guanidiny-
G1H3, including enantioenriched (R)-(À)-muscone O1 (61% ee), racemic (Æ)-muscone O2, (R)-(À)-carvone
lated with N,N’-di-Boc-1H-pyr-
azole-1-carboxamidine (6).
Complete deprotection of the
obtained tripeptide under
O9 (caraway), (S)-(+)-carvone O10 (spearmint), ortho-, meta-, and para-anisaldehydes O11–O13, or butanal
O4.
7
As far as reactive cations G1H1–G1H3 and A1H1–A1H3
are concerned, guanidinium cations have been better than
ammonium cations, and efficiencies of signal generation
(e.g., YMAX/EC50) increased with increasing numbers of hy-
drazides. This observation called for the synthesis of charged
peptide dendrons[47–51] as modular scaffolds that offer multi-
ple hydrazides to attach hydrophobic aldehydes and thus
produce “octopus” amphiphiles with one head and many
sticky tails (“tentacles”). Herein, we report the modular syn-
thesis of dendrons A1H4 and G1H4, and show that most of
the guanidinium amphiphiles G1H4O1–G1H4O13 obtained
from incubation with odorants O1–O13 can activate DNA
transporters in bilayer membranes with unprecedented effi-
ciency. This result is of interest not only to expand the scope
of synthetic differential sensing systems in lipid bilayer
membranes but also for biological applications, such as cel-
lular uptake.[49–54]
acidic conditions gave the target molecule G1H4. Tetrahy-
drazide A1H4 with an ammonium cation in place of the
guanidinium cation was obtained by deprotection of inter-
mediate 5.
All new compounds were optically active, and their
1H NMR spectra showed no indication of epimerization.
This observation was further supported by a single peak in
the reverse-phase (RP) HPLC of the protected target mole-
cule 7.
Characterization: The activity of the new tetrahydrazide cat-
ions G1H4 and A1H4 was determined by comparison with
the previously reported trihydrazide cations G1H3 and
A1H3 (Scheme 1). However, the solubility of ammonium
tetrahydrazide A1H4 and most of its tetrahydrazone deriva-
tives was insufficient to encourage meaningful in-depth char-
acterization. In striking contrast, the solubility of the corre-
sponding guanidinium tetrahydrazide G1H4 and all tetrahy-
drazone derivatives was unproblematic. Octopus amphi-
philes G1H4O1–G1H4O13 were prepared individually by
incubation with 2 equiv odorants O1–O13 per hydrazide in
DMSO for 1 h at 608C (Figure 1).[27] Their ability to activate
polyanion transporters was explored under routine condi-
tions with ctDNA transporters in EYPC-LUVsꢀHPTS/DPX
Results and Discussion
Synthesis: Synthetic sensing systems that can really be of
use in practice are either very straightforward to prepare or
have exceptional characteristics that overcompensate the
Chem. Eur. J. 2010, 16, 14159 – 14166
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14161

S. Matile et al.
Hill analysis of the dose-re-
sponse curve gave the three
fundamental characteristics de-
scribing octopus amphiphile
*
G1H4O8 (Figure 3a, , a),
that is, EC50 =(3.9Æ0.2) mm,
Y
MAX =(62.2Æ1.5)% and the
Hill coefficient n=3.3Æ0.4, re-
porting an informative combi-
nation of stoichiometry, stabili-
ty and cooperativity of the
active transport system.[55] The
EC50 remained constantly low
for octyl, heptyl, and hexyl tails
*
in G1H4O8 (Figure 3a,
),
G1H4O7 (Figure 3a, &), and
!
G1H4O6 (Figure 3a,
), re-
spectively. This poor discrimina-
tion among long alkyl tails sug-
gests that the limit of stoichio-
metric binding[56] is reached and
activities cannot be further im-
proved in the given assay. The
pentyl tails in octopus amphi-
Scheme 2. a) O-(Benzotriazol-1-yl)tetramethyluronium hexafluorophosphate (HBTU), N,N-diethylamine
(DEA), CH2Cl2, room temperature; b) 1. N2H4 monohydrate, MeOH, room temperature; 2. Boc2O (Boc=tert-
butyloxycarbonyl), MeCN, H2O, room temperature, 34% (three steps); c) H2, Pd/C, room temperature, quant;
d) ethyldiisopropylamine (DIEA), MeCN, 508C, 64%; e) 1m HCl in Et2O, CH2Cl2, reflux, 91%; f) 1m HCl in
Et2O, CH2Cl2, reflux, 94%.
phile G1H4O5 produced
a
clearly reduced EC50 =(23.8Æ
1.5) mm at preserved high
(egg yolk phosphatidylcholine large unilamellar vesicles
loaded with the anionic fluorophore 8-hydroxy-1,3,6-pyrene-
trisulfonate and the cationic quencher p-xylene-bis-pyridini-
um bromide, Figure 1). In this assay, the ability of DPX and/
or HPTS to leave the vesicle is reported as fluorescence re-
covery independent of the mechanism of transport. Howev-
er, previous controls in the U-tube have confirmed that
counterion-activated DNA can act as cation carrier at low
concentrations.[20]
In a typical procedure, octopus amphiphile G1H4O8 was
added as concentrated stock solution in DMSO to the
Figure 2. Changes in fractional fluorescence intensity (IF) of HPTS (lex
=
413 nm, lem =510 nm) during addition of G1H4O8: a) 1, b) 2.5, c) 5,
d) 10, and e) 25 mm (final concentrations t<0 s), ctDNA (1.25 mgmLÀ1
final concentration, t=40 s) and Triton X-100 (0.024% final concentra-
tion, t=200 s) to EYPC-LUVsꢀHPTS/DPX.
stirred and thermostated suspension of fluorogenic vesicles
(Figure 2). Unchanged fluorescence at this point confirmed
that octopus amphiphiles G1H4O8 alone are not active in
lipid bilayer membranes. After a brief incubation period,
ctDNA was added as concentrated aqueous stock solution
to produce operational polyion–counterion transporters.
Their ability to mediate DPX export was then followed as
fluorescence recovery as a function of time. At the end of
each experiment, vesicles were lysed to reach the constant
final emission intensity needed for calibration. These experi-
ments were repeated for various concentrations of octopus
amphiphile G1H4O8 at constant DNA concentration. In-
creasing activity with increasing concentrations of G1H4O8
was found (Figure 2). This trend was quantified in the dose-
response curve of octopus amphiphile G1H4O8 (Figure 3a,
*
Y
MAX =(60.1Æ3.2)% (Figure 3a, ). The butyl tails in
G1H4O4 remained with EC50 =(63.9Æ2.8) mm detectable
below the 100 mm threshold, although the YMAX dropped to a
~
less important
Y
MAX =(14.9Æ1.0)% (Figure 3a,
).
G1H4O3 was not characterized because the detection limit
was reached with G1H4O4.
Compared to previously reported results with guanidini-
um amphiphiles with three (G1H3, Figure 3b), two (G1H2,
Figure 3c), and one tail (G1H1, Figure 3d),[27] the results for
the n-alkyl series of the new octopus amphiphile G1H4 (Fig-
ure 3a) revealed the following clear trends. Most important-
*
).
14162
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 14159 – 14166

Differential Fragrance Sensing
FULL PAPER
Figure 4. Dependence of: a) the effective concentration EC50, b) the max-
*
imal activity (YMAX), and c) the Hill coefficient (n) of G1H4 ( ), G1H3
*
^
(
), G1H2 (&), and G1H1 ( ) on the length of the n-alkyl tail of the al-
dehyde they are derivatized with (in number of carbons), with indication
of pertinent thresholds as dotted lines (EC50 =100 mm, YMAX =0.5, n=1).
Data from the Hill analysis of the curves in Figure 3.
able below the EC50 =100 mm threshold, whereas the corre-
sponding G1H3O4, G1H2O4 and G1H1O4 were at EC50 >
100 mm or fully inactive (Figure 4a, a).
The increasing efficiency of octopus amphiphiles with in-
creasing number of tails is likely to originate from improved
partitioning into the membrane. This interpretation was sup-
ported by low YMAX and low n with octopus amphiphiles
from G1H4 (Figure 4b and 4c). Low YMAX is commonly in-
terpreted as an early onset of precipitation of overly hydro-
phobic transporter components during their delivery to the
membrane at higher concentrations. However, YMAX never
dropped below the 50% threshold except for G1H4O4 with
an exceptionally low YMAX =(14.9Æ1.0)% (Figure 4b, a).
Decreasing n with increasing number and length of hydro-
phobic tails revealed increasing stability of the final poly-
ion–counterion transporters under these conditions.[55] Maxi-
mal n=6–9 indicated that the binding of at least 6–9 coun-
terions is needed to activate ctDNA transporters. Minimal
n>2 remained far above the n=1 threshold. This demon-
strated that also the most stable transport system is thermo-
dynamically unstable. As highly dynamic minority compo-
nents, all active structures are thus undetectable by routine
spectroscopic methods (Figure 4c, a).[55,56]
Figure 3. Dose-response curves for DNA activation by guanidinium hy-
Detection and discrimination of anisaldehyde odorants
O11–O13 was one of the problems faced with the previously
reported reactive cations G1H1–G1H3 and A1H1–A1H3
(Figure 5, c). This problem was rationalized with insuffi-
cient amphiphilicity. With octopus amphiphiles obtained
from peptide dendron G1H4, activities increased approxi-
mately 3 times with a lowest EC50 =(4.3Æ0.8) mm for the
meta isomer O12. Consistently weaker activity with the
ortho isomer O11 and para isomer O13 could possibly origi-
nate from incomplete hydrazone formation because of their
reduced electrophilicity (or from steric effects in the final
amphiphiles, etc.).
drazides: a) G1H4, b) G1H3, c) G1H2, and d) G1H1 after incubation
!
*
*
with octanal O8 ( ), heptanal O7 (&), hexanal O6 ( ), pentanal O5 ( ),
!
and butanal O4 ( ). Data for G1H3, G1H2, and G1H1 are adapted from
ref. [27]; for details, compare original data for amphiphile G1H4O8 in
Figure 2 and the Supporting Information.
ly, EC50 values with G1H4 improved by a factor of 2–4 com-
pared with G1H3, and ꢁ10 times compared with G1H2 and
G1H1 (Figure 4a and Table S1 in the Supporting Informa-
tion). The only exception from this trend was observed at
EC50 <10 mm, at which the onset of stoichiometric binding
processes obscured intrinsic differences between individual
transport systems.[56] As an interesting consequence, the ac-
tivity of tetrabutyl octopus amphiphile G1H4O4 was observ-
Enantiodiscrimination is of critical importance in all sens-
ing systems. With fragrance sensors, (R)-(À)-muscone O1
represents a benchmark challenge concerning enantiodiscri-
Chem. Eur. J. 2010, 16, 14159 – 14166
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14163

S. Matile et al.
tern generation with G1H2, G1H3, A1H2, and A1H3, and
pattern analysis with standard PCA or hierarchical cluster-
ing analysis (HCA) methods.
Attached to peptide dendrons, each 15-membered ring of
muscone appears to contribute two alkyl tails to the octopus
amphiphiles that are linked together at the end (Figure 1).
In amphiphile G1H4O1, this adds up to eight tails in total.
As a result, minor structural differences are multiplied in oc-
topus amphiphiles. For instance, the number of stereogenic
centers raises from one in muscone and five in amphiphile
G1H3O1 to seven in octopus amphiphile G1H4O1. This en-
hanced chirality suggests that octopus amphiphiles should
be ideal to enable enantiodiscrimination. Compared with
G1H3, the effective concentrations obtained for muscone
with G1H4 improved almost 10 times to EC50 =(2.8Æ0.1)
Figure 5. Dose-response curves for DNA activation by guanidinium hy-
drazides G1H4 (a) and G1H3 (c) after incubation with o-anisalde-
*
*
hyde O11 ( ), m-anisaldehyde O12 ( ), and p-anisaldehyde O13 (&).
Data for G1H3 are adapted from ref. [27]; for details, compare Figure 2
and the Supporting Information.
*
and (5.6Æ0.3) mm (Figure 6a, ). Together with clearer dif-
ferences in YMAX, the dose-response curves for enantioen-
*
mination. Muscone is extracted from a glandular secretion
by musk deer, and has been used for centuries in perfumery
and medicine.[57] Discrimination of enantioenriched (R)-(À)-
muscone O1 (61% ee) and racemic (Æ)-muscone O2 with
the previous trihydrazide countercation G1H3 and ctDNA
transducers was not clear considering the similarity of their
EC50 values of (23.0Æ2.1) and (33.6Æ4.2) mm, respectively,
and their YMAX (Figure 6a, &). In this case, reliable discrimi-
nation was possible only by differential sensing by using pat-
riched (À)-muscone O1 (Figure 6a, , a) and racemic
*
(Æ)-muscone O2 (Figure 6a, , c) became distinguishable
without PCA or HCA treatment. Because of high hydropho-
bicity leading to inefficient delivery, the absolute YMAX of
fragrant octopus amphiphiles G1H4O1 and G1H4O2
became, however, disappointingly low.
The ability of octopus amphiphiles to enable enantiodis-
crimination was very satisfactory with carvone, the leading
example for ring-contracted monoterpene analytes.[58] At-
tached to minidendron G1H3, the two enantiomers pro-
duced essentially the same dose-response curve at an inter-
mediate EC50 =31–33 mm (Figure 6b, &). Attached to the ex-
panded peptide dendron G1H4, both efficiency and discrim-
ination power improved to an excellent EC50 =(13.2Æ
1.6) mm for G1H4O9 and EC50 =(23.5Æ1.3) mm for
G1H4O10 without any complementary cuts in YMAX (Fig-
*
ure 6b, ).
To estimate the increase in enantiodiscriminatory power
mediated by octopus amphiphiles obtained from G1H4
more quantitatively, HCA was applied to the minimalist pat-
tern produced by all obtained Hill parameters measured
with G1H3 and G1H4 for muscones and carvones. HCA is
an unsupervised method of multivariant analysis that con-
verts all interpoint Euclidean distances in multidimensional
patterns into 2D dendrograms. The significant shortest dis-
tances in these dendrograms, measured in Eucliden units
(E.u.), can be cautiously used as a semiquantitative indica-
tion of sensing power. In the HCA dendrogram obtained
with G1H3 only, the carvone enantiomers were separated
by 26 E.u., whereas the enantiomer mixtures of muscone
were separated by 12 E.u. (Figure 7a). In the HCA dendro-
gram produced by G1H3 and G1H4, the carvone enantio-
mers were separated by 44 E.u., whereas enantiodiscrimina-
tion of muscones increased more than 3 times to 38 E.u.
(Figure 7b).
Figure 6. a) Dose-response curves for (À)-muscone O1 (a) and (Æ)-
*
muscone O2 (c) derivatized with G1H3 (&) and G1H4 ( ); O1 (ꢁ)
Compared with the many elegant, often much more prac-
tical differential chemosensors reported over the past two
decades,[28–46] sensing systems that work in lipid bilayers are
of interest because subtle differences in structure are magni-
fied by covalent capture of several analytes by the same
and O2 (+) alone were also measured. b) Dose-response curves for (À)-
carvone O9 (c) and (+)-carvone O10 (a) derivatized with G1H4
*
^ *
(
), G1H3 (&), G1H2 ( ) and G1H1 ( ); O9 (ꢁ) and O10 (+) alone
were also measured. Data for G1H3, G1H2, and G1H1 are adapted from
ref. [27]; for details, compare Figure 2 and the Supporting Information.
14164
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 14159 – 14166

Differential Fragrance Sensing
FULL PAPER
uptake.[49–54] RNAi, for example, which is a general approach
to regulate gene expression, operates with short oligonucleo-
tides, that is, siRNA that show less pronounced multivalency
effects and can thus respond less convincingly to more con-
ventional amphiphiles.[59] Moreover, dynamic octopus am-
phiphiles would be interesting to mix in a fluorescent probe
to image uptake pathways[52–54] or membrane domains[60]
without significantly disturbing intrinsic function.
Experimental Section
Details of all experimental procedures and characterizations can be
found in the Supporting Information.
Figure 7. HCA dendrogram showing the Euclidian distances between
average values for data obtained with: a) G1H3, and b) G1H3 and
G1H4; dotted line: 40 E.u.
Acknowledgements
We thank A. Hennig for help with data analysis, D. Jeannerat, A. Pinto,
and S. Grass for NMR measurements, G. Hopfgartner, S. Michalet, P.
Perrottet, and N. Oudry for MS measurements, L. Wunsche and A. Herr-
mann (Firmenich, Geneva) for odorant samples, and the University of
Geneva and the Swiss NSF for financial support.
counterion head-group and by the noncovalent assembly of
multiple counterion activators on one polyion transporter.
This is particularly true for the discrimination of enantio-
meric analytes, for which these effects are expressed in a
chiral environment. Covalent capture of pairs of enantio-
mers as analytes by chiral counterions produces diastereo-
meric amphiphilies, which will in turn interact differently
with chiral DNA transporters, and the resulting diastereo-
meric polyion–counterion complexes will interact differently
with the chiral phospholipids in the membrane. This situa-
tion implies that chiral reactive counterions are not necessa-
ry to enable enantiodiscrimination. The different binding of
enantiomeric amphiphiles to chiral polyions and membranes
should suffice to discriminate enantiomeric analytes. How-
ever, increasing the number of stereogenic centers in and
the number of chiral analytes attached to the reactive coun-
terions should naturally increase enantiodiscrimination. The
reported results with octopus amphiphile G1H4 are in
agreement with this interpretation.
[1] L. B. Buck, Angew. Chem. 2005, 117, 6283–6296; Angew. Chem. Int.
Ed. 2005, 44, 6128–6140.
[2] G. Das, S. Matile, Chem. Eur. J. 2006, 12, 2936–2944.
[3] S. Matile, H. Tanaka, S. Litvinchuk, Top. Curr. Chem. 2007, 277,
219–250.
[4] N. Sakai, J. Mareda, S. Matile, Acc. Chem. Res. 2008, 41, 1354–1365.
[5] S. Matile, A. Som, N. Sordꢂ, Tetrahedron 2004, 60, 6405–6435.
[6] A. L. Sisson, M. R. Shah, S. Bhosale, S. Matile, Chem. Soc. Rev.
2006, 35, 1269–1286.
[7] J. T. Davis, P. A. Gale, O. A. Okunola, P. Prados, J. C. Iglesias-San-
chez, T. Torroba, R. Quesada, Nat. Chem. 2009, 1, 138–144.
[8] R. J. Brea, C. Reiriz, J. R. Granja, Chem. Soc. Rev. 2010, 26, 1448–
1456.
[9] M. Febles, N. Pꢂrez-Hernandez, C. Pꢂrez, M. L. Rodrꢃguez, C.
Foces-Foces, M. V. Roux, E. Q. Morales, G. Buntkowsky, H. H. Lim-
bach, J. D. Martꢃn, J. Am. Chem. Soc. 2006, 128, 10008–10009.
[10] T. M. Fyles, Chem. Soc. Rev. 2007, 36, 335–347.
[11] A. P. Davis, D. N. Sheppard, B. D. Smith, Chem. Soc. Rev. 2007, 36,
348–357.
[12] G. W. Gokel, N. Barkey, New J. Chem. 2009, 33, 947–963.
[13] C. P. Wilson, S. J. Webb, Chem. Commun. 2008, 4007–4009.
[14] A. Satake, M. Yamamura, M. Oda, Y. Kobuke, J. Am. Chem. Soc.
2008, 130, 6314–6315.
[15] X. Li, B. Shen, X. Q. Yao, D. Yang, J. Am. Chem. Soc. 2009, 131,
13676–13680.
[16] M. Jung, H. Kim, K. Baek, K. Kim, Angew. Chem. 2008, 120, 5839–
5841; Angew. Chem. Int. Ed. 2008, 47, 5755–5757.
[17] P. L. Boudreault, N. Voyer, Org. Biomol. Chem. 2007, 5, 1459–1465.
[18] N. Sakai, B. Baumeister, S. Matile, ChemBioChem 2000, 1, 123–125.
[19] N. Sakai, S. Matile, J. Am. Chem. Soc. 2003, 125, 14348–14356.
[20] T. Takeuchi, V. Bagnacani, F. Sansone, S. Matile, ChemBioChem
2009, 10, 2793–2799.
[21] G. Das, P. Talukdar, S. Matile, Science 2002, 298, 1600–1602.
[22] T. Miyatake, M. Nishihara, S. Matile, J. Am. Chem. Soc. 2006, 128,
12420–12421.
Conclusion
The general conclusion of this study is simple but attractive
for several reasons. Cationic hydrazone amphiphiles with
four hydrophobic tails are the most powerful activators of
DNA transporters in lipid bilayers known today. Moreover,
increasing activity and discriminatory power with increasing
number of tails is confirmed. The emergence of stoichiomet-
ric binding with more hydrophobic analytes suggests that oc-
topus amphiphiles with one head and four tentacles could
be nearly ideal (Figure 3a). Preliminary results are in agree-
ment with the expectation that efficiencies are unlikely to
further increase with more than four analytes per amphi-
phile.
[23] S. Litvinchuk, N. Sordꢂ, S. Matile, J. Am. Chem. Soc. 2005, 127,
9316–9317.
[24] S. Litvinchuk, H. Tanaka, T. Miyatake, D. Pasini, T. Tanaka, G.
Bollot, J. Mareda, S. Matile, Nat. Mater. 2007, 6, 576–580.
The high activator efficiency found for octopus amphi-
philes is particularly interesting with regard to cellular
Chem. Eur. J. 2010, 16, 14159 – 14166
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
14165

S. Matile et al.
[25] S. M. Butterfield, T. Miyatake, S. Matile, Angew. Chem. 2009, 121,
331–334; Angew. Chem. Int. Ed. 2009, 48, 325–328.
[26] T. Takeuchi, S. Matile, J. Am. Chem. Soc. 2009, 131, 18048–18049.
[27] T. Takeuchi, J. Montenegro, A. Hennig, S. Matile, Chem. Sci. DOI:
10.1039/c0sc00386g.
[28] J. J. Lavigne, E. V. Anslyn, Angew. Chem. 2001, 113, 3212–3225;
Angew. Chem. Int. Ed. 2001, 40, 3118–3130.
[29] M. A. Palacios, R. Nishiyabu, M. Marquez, P. Anzenbacher, Jr., J.
Am. Chem. Soc. 2007, 129, 7538–7544.
[30] J. R. Epstein, D. R. Walt, Chem. Soc. Rev. 2003, 32, 203–214.
[31] C. Zhang, K. S. Suslick, J. Am. Chem. Soc. 2005, 127, 11548–11549.
[32] H. Zhou, L. Baldini, J. Hong, A. J. Wilson, A. D. Hamilton, J. Am.
Chem. Soc. 2006, 128, 2421–2425.
[33] A. Buryak, K. Severin, J. Am. Chem. Soc. 2005, 127, 3700–3701.
[34] S. H. Shabbir, L. A. Joyce, G. M. da Cruz, V. M. Lynch, E. Sorey,
E. V. Anslyn, J. Am. Chem. Soc. 2009, 131, 13125–13131.
[35] J. C. Jurs, G. A. Bakken, H. E. McClelland, Chem. Rev. 2000, 100,
2649–2678.
[36] J. W. Lee, J.-S. Lee, Y.-T. Chang, Angew. Chem. 2006, 118, 6635–
6637; Angew. Chem. Int. Ed. 2006, 45, 6485–6487.
[37] D. C. Gonzlez, E. N. Savariar, S. Thayumanavan, J. Am. Chem. Soc.
2009, 131, 7708–7716.
[38] O. R. Miranda, H.-T. Chen, C.-C. You, D. E. Mortenson, X.-C.
Yang, U. H. F. Bunz, V. M. Rotello, J. Am. Chem. Soc. 2010, 132,
5285–5289.
[42] A. Wada, S. Tamaru, M. Ikeda, I. Hamachi, J. Am. Chem. Soc. 2009,
131, 5321–5330.
[43] T. D. James, Top. Curr. Chem. 2007, 277, 107–154.
[44] R. Jelinek, S. Kolusheva, Top. Curr. Chem. 2007, 277, 155–180.
[45] L. Vial, P. Dumy, New J. Chem. 2009, 33, 939–946.
[46] M. Florea, W. M. Nau, Org. Biomol. Chem. 2010, 8, 1033–1039.
[47] K. Sadler, J. P. Tam, Rev. Mol. Biotechnol. 2002, 90, 195–229.
[48] T. Darbre, J.-L. Reymond, Acc. Chem. Res. 2006, 39, 925–934.
[49] D. A. Tomalia, A. M. Naylor, W. A. Goddard, Angew. Chem. 1990,
102, 119–157; Angew. Chem. Int. Ed. Engl. 1990, 29, 138–175.
[50] M. Liu, J. M. J. Frechet, Pharm. Sci. Technol. Today 1999, 2, 393–
401.
[51] K. T. Al-Jamal, T. Sakthivel, A. T. Florence, Coll. Surface A: Phys.
Eng. Asp. 2005, 268, 52–59.
[52] S. Bhattacharya, A. Bajaj, Chem. Commun. 2009, 4632–4656.
[53] J. Fernꢄndez-Carneado, M. Van Gool, V. Martos, S. Castel, P.
Prados, J. de Mendoza, E. Giralt, J. Am. Chem. Soc. 2005, 127, 869–
874.
[54] T. Takeuchi, M. Kosuge, A. Tadokoro, Y. Sugiura, M. Nishi, M.
Kawata, N. Sakai, S. Matile, S. Futaki, ACS Chem. Biol. 2006, 1,
299–303.
[55] S. Bhosale, S. Matile, Chirality 2006, 18, 849–856.
[56] F. Mora, D.-H. Tran, N. Oudry, G. Hopfgartner, D. Jeannerat, N.
Sakai, S. Matile, Chem. Eur. J. 2008, 14, 1947–1953.
[57] L. Ruzicka, Helv. Chim. Acta 1926, 9, 715–729.
[58] G. Wagner, Chem. Ber. 1894, 27, 2270–2276.
[39] N. T. Greene, K. D. Shimizu, J. Am. Chem. Soc. 2005, 127, 5695–
5700.
[40] Topics in Current Chemistry, Vol. 277 (Ed.: T. Schrader), Springer,
Heidelberg, 2007.
[59] A. Fire, S. Xu, M. Montgomery, S. Kostas, S. Driver, C. Mello,
Nature 1998, 391, 806–811.
[60] L. A. Bagatolli, Biochim. Biophys. Acta. 2006, 1758, 1541–1556.
[41] C. Guarise, L. Pasquato, V. De Filippis, P. Scrimin, Proc. Natl. Acad.
Sci. USA 2006, 103, 3978–3982.
Received: September 9, 2010
Published online: October 29, 2010
14166
www.chemeurj.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2010, 16, 14159 – 14166