Arginine magic with new counterions up the sleeve

Article abstract of DOI:10.1039/b501472g

The elusive questions how arginine-rich sequences allow peptides and proteins to penetrate cells or to form voltage-gated ion channels are controversial topics of current scientific concern. The possible contributions of exchangeable counterions to these

Full text of DOI:10.1039/b501472g

View Article Online / Journal Homepage / Table of Contents for this issue
A R T I C L E
Arginine magic with new counterions up the sleeve
Masamichi Nishihara,a Florent Perret,a Toshihide Takeuchi,b Shiroh Futaki,b,c
Adina N. Lazar,d Anthony W. Coleman,d Naomi Sakai*a and Stefan Matile*a
a Department of Organic Chemistry, University of Geneva, Geneva, Switzerland
b Institute for Chemical Research, Kyoto University, Japan
c PRESTO, JST, Kyoto, Japan
d Institut de Biologie et Chimie des Prote´ines, CNRS UMR 5086, 7 passage du Vercors, F69367,
Lyon, France. E-mail: naomi.sakai@chiorg.unige.ch, stefan.matile@chiorg.unige.ch;
Fax: +41(0)22 379 3215; Tel: +41(0)22 379 6523
Received 28th January 2005, Accepted 4th March 2005
First published as an Advance Article on the web 7th April 2005
The elusive questions how arginine-rich sequences allow peptides and proteins to penetrate cells or to form voltage-
gated ion channels are controversial topics of current scientific concern. The possible contributions of exchangeable
counterions to these puzzling processes remain underexplored. The objective of this report is to clarify scope and
limitations of certain counteranions to modulate cellular uptake and anion carrier activity of oligo/polyarginines.
The key finding is that the efficiency of counteranion activators depends significantly on many parameters
such as activator–membrane and activator–carrier interactions. This finding is important because it suggests that
counteranions can be used to modulate not only efficiency but also selectivity. Specifically, activator efficiencies are
found to increase with increasing aromatic surface of the activator, decreasing size of the transported anion, increasing
carrier concentration as well as increasing membrane fluidity. Efficiency sequences depend on membrane composition
with coronene > pyrene ꢀ fullerene > calix[4]arene carboxylates in fluid and crystalline DPPC contrasting
to fullerene > calix[4]arene ≈ coronene > pyrene carboxylates in EYPC with or without cholesterol or ergosterol.
In HeLa cells, the efficiency of planar activators (pyrene) exceeds that of spherical activators (fullerenes, calixarenes).
Polyarginine complexes with pyrene and coronene activators exhibit exceptional excimer emission. Decreasing
excimer emission with increasing ionic strength reveals dominant hydrophobic interactions with the most efficient
carboxylate activators. Dominance of ion pairing with the inefficient high-affinity sulfate activators is corroborated
by the reversed dependence on ionic strength. These findings on activator–carrier and activator–membrane
interactions are discussed as supportive of arene-templated guanidinium–carboxylate pairing and interface-directed
translocation as possible origins of the superb performance of higher arene carboxylates as activators.
their solubility by multiple counteranion exchange to adapt to
1
Introduction
changing environments.1,2 In support of this hypothesis, it was
not surprising to find that a) CPPs are anion carriers and b) that
bilayer penetration by CPPs can be regulated with amphiphilic
anions (Fig. 2).1,2 Counteranions (i.e., alignable dipoles of
hydrophobic ion pairs) might contribute to the mechanism of
voltage gating by R-rich motifs for the same reasons.1,3–8
Arginine-rich peptides and proteins seem to exist always in
“functional symbiosis” with tightly bound but exchangeable
counteranions. This “counteranion scavenging” could be the
trick of oligoarginines to minimize intramolecular charge re-
pulsion (Fig. 1A), because the weak acidity of the guani-
dinium cation hinders partial deprotonation under physiological
conditions.1 More acidic oligoamines such as K-rich peptides
do not show similar affinity to anions (Fig. 1B). Recently, we
suggested that counteranion scavenging may explain some of the
“mysterious” functions of arginine-rich peptides and proteins
in biomembranes.1 The movement of R-rich cell-penetrating
peptides (CPPs) like oligoarginines across bilayer membranes
can, for example, be understood as repeated alterations in
In our earlier studies, activator anions like EYPG are used
as a component of the model membranes.1,2 Very recently,
however, we found that externally added amphiphilic anions Y
can activate oligo/polyarginine-anion complexes in live cells as
well as otherwise inaccessible neutral EYPC bilayers (Fig. 3).9
Confirmed key characteristics of activators Y include one or
more negative charge for ion pairing with guanidinium cations
of the oligo/polyarginine.9 Moreover, all anion activators Y
known so far are amphiphiles with large hydrophobic domains
for favorable interactions with the bilayer membrane. Activators
Y with hydrophobic domains composed of higher aromatics
appear particularly promising because additional, more specific
activator–carrier, activator–activator and activator–membrane
interactions become possible (see below).
Fig. 2 summarizes how activating and inactivating coun-
teranions may contribute to the function of CPPs. In brief,
R-rich CPPs Rn will always exist as complexes Rn·(X)m to
minimize intramolecular charge repulsion (Fig. 1A). As an
illustrative example for the need for counterions, we recently
found that oligoarginines can scavenge inorganic phosphates
that are present in trace amounts in the environment.10 Ex-
change of scavenged hydrophilic counteranions X such as
inorganic phosphates in complexes Rn·(X)m with activators Y
Fig. 1 To minimize repulsion between proximal charges, (A) guani-
dinium-rich oligomers scavenge anions X− to give complexes Rn·(X)m
(Fig. 2), whereas (B) ammonium-rich oligomers release protons.
ꢁ
ꢁꢁ
is expected to yield active complexes Rn·(X)m ·(Y)m at the
T h i s j o u r n a l i s
T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 5
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 5 9
©

View Article Online
Fig. 2 Counteranion activators Y (red) of oligo/polyarginine carriers/CPPs Rn (black) are thought to act by replacing scavenged hydrophilic anions
ꢁ
ꢁꢁ
X (blue) in complexes Rn·(X)m to produce complexes Rn·(X)m ·(Y)m (B) that act as carriers to release entrapped anions like CF (orange→green, B→G)
or are directed by internal or external hydrophilic polyanions Xm (cyan) to accumulate inside (u, uptake) or outside a spherical bilayer membrane (i,
inhibition).1,2
of interfacial polyelectrolytes, particularly GAGs, is a topic of
current scientific concern.11–24 We reiterate that the synergistic
ꢁ
ꢁꢁ
carrier mechanism for CPP-activator complexes Rn·(X)m ·(Y)m
outlined in Fig. 2 is assumed in this report based on results
reported in references 1 and 2 that disfavor formation of ion
channels and pores as well as the occurrence of more drastic
events. Nevertheless, this assumption as well as its relevance for
oligo/polyarginine uptake by live cells remains to be verified.
Extensive experimental evidence from many groups in support
of a carrier but also of other, competing or coinciding uptake
mechanisms, particularly endocytosis, exists.7,8,11–24
The here envisioned role of CPP activators is reminis-
cent of the role of charge-reversal amphiphiles in gene
delivery.25–27 The rare occurrence of intrinsic cationic activators
in the biomembranes practically excludes spontaneous oligonu-
cleotide translocation, a fact that naturally directs scientific
attention to artificial countercation activators. In contrast,
ubiquitous anionic activators in biomembranes may have shifted
the focus of CPP research from counteranion activators to a
mysteriously “spontaneous” uptake1,2,7,8,11–24 that is sometimes
referred to as “arginine magic”.
This study focuses on nature and selectivity of the interactions
between anion activators, polyarginine (pR) carriers and lipid
bilayer membranes. The key question to address was why higher
arenes and carboxylates emerge as privileged motifs in efficient
activators. With regard to activator–membrane interactions, it
is well known that many membrane-spanning proteins exhibit
external surfaces with aromatic rings positioned to interact with
the interfacial region of bilayer membranes.28–37 This interface
contains the lipidic ester groups as well as their phosphates
paired with counterions such as the ammonium cation in the
case of phosphatidylcholine. Examples for membrane peptides
and proteins with aromatic rings at the interface include b-helices
with gramicidin A,36 a-helix bundles with potassium channels30
and b-barrels with maltoporin.28 Model studies confirmed
preferential location of indoles at the interface.31–35 Hydrogen
bonding to interfacial esters and cation–p interactions with
ammonium cations (but not yet arene-templated ion pairing)38
have been considered beyond simple preference for intermediate
dielectrics to explain interfacial accumulation of arenes.28–35
Model studies confirmed that appropriately positioned aro-
matics can stabilize transmembrane orientation of model
helices.32–34 A similar anchoring function has been proposed for
aromatics in synthetic ion channels and pores.36,37 Important for
this study, it has been suggested that the interfacial preference
of arenes may drive the translocation across bilayer membranes
to interfaces at the other side.29 For carriers, preferential
Fig. 3 Structure of potential anion activators Y (Fig. 2; anions 14, 15,
21 and 22 are inactive under the employed experimental conditions).
membrane–water interface (Fig. 2B). The resulting supramolec-
ꢁ
ꢁꢁ
ular carriers Rn·(X)m ·(Y)m may then shuttle across the bilayer,
pick up entrapped anions like CF, back-transfer and release
unquenched CF (Fig. 2). Therefore, an efficient activator Y
should have the ability to form stable but labile guanidinium–
anion interactions. Hydrophilic high-affinity polyanions Xm like
glycosaminoglycans (GAGs),1,2 oligonucleotides (RNA, DNA),
micelles (SDS)1,2 and presumably also nucleotides like ATP1
may play the interesting role to regulate carrier activity and cell
penetration. Namely, external Xm should inhibit both processes,
whereas internal Xm should inhibit back transfer and thus
facilitate the translocation. The possibly more complex role
1 6 6 0
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9

View Article Online
accumulation at the interface is essential for function because
this is the place where the ions are exchanged. Indeed, we
have recently found that interfacial preference is one of the
distinguishing characteristics of octaarginine, an excellent CPP,
compared to hexadecaarginine, a poor CPP that accumulates in
the hydrophobic core of EYPC/EYPG membranes.2 Therefore,
it was conceivable that fullerene, coronene and calixarene
activators may preferentially accumulate at the interfacial re-
gion and, therefore, mediate interface-directed translocation of
polyarginine carriers.
With regard to activator–carrier interactions, Thompson and
Smithrud pointed out recently that arene-templated ion pairing
occurs in nature and created an elegant synthetic model to
secure experimental evidence for relevance of this recognition
motif in water and in octanol.38 Important in the context of this
study, the Thompson–Smithrud model identified guanidinium–
carboxylate pairs39–45 as privileged motif; alkyl-templated ion
pairing was not observed.38 Arene-templated ion pairing was
explored in the context of a-helix recognition as well.45 We
have proposed the same effect early on as a possible origin
of activator–carrier interactions.1 With respect to the well-
recognized cation–p interaction,46,47 arene-templated ion pairing
may be considered as an extension that covers the counteranion
and the resulting network of possible interactions between
anion, cation and arene as well. Such an influence of a remote48
aromatic surface to the binding could be the key to achieve stable
Fig. 4 Activators with poor efficiency (A), superb EC50 (B), and superb
YMAX (C) exemplified with laurate 16 (A), fullerene malonate 1 (B) and
pyrene acetate 5 (C), measured at 25 ◦C. (D) Dose response curves for
16 (ꢀ), 1 (᭹) and 5 (ꢁ) with curve fit to Hill equation. (A–C): Changes
in CF emission I (kex 492 nm, kem 517 nm) as a function of time during
addition of activator (at t = 60 s; with increasing intensity at t = 360 s,
0, 10, 20, 30, 40, 50, 60, 70, 80 and 100 lM 16 in A, 0, 5, 35, 50, 100, 200,
300 and 400 nM 1 in B, 0, 10, 50, 60, 75, 100, 150 and 200 lM 5 in C)
and pR (at t = 120 s, 250 nM) to EYPC-LUVs⊃CF (250 lM EYPC),
calibrated by final lysis (I = 100% with excess triton X-100).
ꢁ
ꢁꢁ
complexes pR·(X)m ·(Y)m with kinetic lability.
Taken together, these considerations suggested that the most
efficient activators may bind to CPPs by arene-templated
carboxylate–guanidinium pairing to subsequently mediate the
interface-directing translocation of the resulting activator–
carrier complexes. The studies on activator–membrane and
activator–carrier interactions described in the following are in
support of this hypothesis, although the involvement of the other
mechanisms is likely.
concentrations (Fig. 4A and 4D, ꢁ). These diverse, in part
complementary characteristics of anion activators 1–22 were
classified using two parameters: The maximal activity YMAX and
the EC50 (Table 1). The effective activator concentration EC50
is defined as activator concentration needed to reach YMAX/2
(Table 1). As a general trend, EC50’s decreased with decreasing
2
Results and discussion
2.1 The efficiency sequence in EYPC membranes
Anions 1–22 were considered as activators Y of polyargi-
nine (pR) to give functional activator–carrier complexes
Y
MAX. This was regrettable because the perfect activator should
ꢁ
ꢁꢁ
pR·(X)m ·(Y)m (Fig. 3). Fullerene (2) and coronene (4) sulfates
were prepared in a few routine steps from commercial starting
materials (see Experimental section). Fullerene (1) and coronene
(3) carboxylates as well as calixarene (8–15) and naphthalene
19 activators were synthesized following reported procedures.9
Anions 5–7, 16–18, and 20–22 were commercially available.
The classical CF assay, in which the release of self-quenched
CF from EYPC-LUVs⊃CF is monitored continuously as an
increase of fluorescence, was used to assess the ability of
amphiphilic anions 1–22 to activate polyarginine carriers pR in
otherwise inaccessible, neutral EYPC bilayers.9 Conventionally,
the CF assay is used to identify large pores and membrane-
disruptive detergents.36 However, recent mechanistic studies
suggest that oligo-/polyarginines act as anion carriers under
the selected conditions,1,2 although contributions from transient
pores and other mechanisms cannot be excluded.36 In any
case, elucidation of the transport mechanism of activator–
achieve high YMAX at low EC50. To illustrate this dilemma,
activator efficiency E was defined as in eqn. (1) based on the
roughly exponential YMAX–EC50 profile9 of the pyrene series 5–
9. Namely,
E = YMAX × pEC50/f
(1)
where YMAX is the maximal activity in percent, pEC50 the
negative logarithm of the effective concentration in mM, and
f = 20.6 an arbitrary scaling factor to calibrate E in EYPC
bilayers between 0 and 10. This definition assigned E = 10 to
the best activator in EYPC bilayers, i.e., the Bingel fullerene 1.
The “pyrene threshold”9 was located between E = 4.1 for 5 and
E = 5.9 for 9, i.e., around E ≈ 5.0.
Only a few anion activators other than fullerene 1 (E =
10) exhibited efficiencies clearly above the pyrene threshold:
calix[4]arenes57 12 (E = 8.3), 11 (E = 7.2) and 10 (E = 6.5) as well
as coronenebutyrate58 3 (E = 7.5). Amphiphilic carboxylates
(fullerene ꢀ calix[4]arene ≈ coronene > pyrene ≈ steroid >
calix[6]arenes > naphthalene59,60 ≈ alkane) were clearly more
efficient than the corresponding sulfates (Table 1) and alcohols
(all inactive). Despite promising activity observed with alkyl
phosphate (Table 1, entry 17), phosphate activators were not
further investigated because of the inactivity of calix[4]arene 14
(Table 1, entry 14).
ꢁ
ꢁꢁ
polyarginine complexes pR·(X)m ·(Y)m is not the topic of this
study.
Amphiphilic anions 1–22 stimulated pR-mediated CF release
from EYPC-LUVs⊃CF in varying extent. Pyreneacetate49,50
ꢁ
ꢁꢁ
5, on the one hand, formed complexes pR·(X)m ·(5)m that
mediated up to 80% CF release at relatively high concentrations
ꢁ
ꢁꢁ
(Fig. 4C and 4D, ꢀ). Complexes pR·(X)m ·(1)m with the Bingel
fullerene27,51–56 1 as an activator, on the other hand, mediated
up to 50% CF release at exceptionally low concentrations
With 250 nM polymer, the average concentration of guani-
dinium cations calculated to 18 lM. This suggested that the
EC50 of most activators was below stoichiometric ion pairing.
8
ꢁ
ꢁꢁ
(Fig. 4B and 4D, ᭹). Laurate complexes pR·(X)m ·(16)m
,
finally, mediated up to 10% CF release only at relatively high
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 6 1

View Article Online
Table 1 Characteristics of anion activators 1–22 in EYPC-LUVs⊃CF
Table 2 Characteristics of coronene activators in EYPC-LUVs⊃CF at
at constant lipid and carrier concentrationa
constant lipid concentration as a function of carrier concentrationa
Entry
Anionb
YMAX (%)c,d
EC50/lMc,e
Ef
Entry
Anionb
pR/nMg
YMAX (%)c,d
EC50/lMc,e
Ef
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
1
48
30
65
50
80
78
55
48
46
46
65
68
33
—
—
10
61
27
30
48
—
—
1
2
2
3
3
2
4
3
2
3
2
2
3
0.051 0.002
0.11 0.006
4.3 0.3
10.0
5.8
7.5
5.2
4.1
5.1
5.4
5.1
5.9
6.5
7.2
8.3
3.7
0.0
0.0
0.7
5.1
2.4
2.5
4.7
0.0
0.0
01
02
03
04
05
06
3
3
3
4
4
4
250
125
50
250
125
50
65
51
43
50
42
22
2
2
2
3
1
1
4.3 0.3
3.3 0.2
2.6 0.2
7.5 0.7
7.6 0.2
4.6 0.1
7.5
6.1
5.4
5.2
4.3
2.5
2
3
4
7.5 0.7
5
86
44
3
2
6
7
9.3 0.7
6.7 0.6
2.2 0.2
1.2 0.2
5.3 0.3
3.1 0.2
5.1 0.4
a As in Table 1; measured at 25 ◦C. b As in Table 1. c As in Table 1. d As
in Table 1. e As in Table 1. f As in Table 1. g pR, poly-L-arginine.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
this reason, it was of interest to see whether or not complexes
ꢁ
ꢁꢁ
pR·(X)m ·(Y)m could mediate the release of anions larger than
CF from EYPC LUVs. This question was addressed using CFD,
i.e., CF-dextran, a classical probe in size exclusion experiments.61
CFD release from EYPC-LUVs⊃CFD by pR anion complexes
was tested at concentrations above their EC50 for CF release
from EYPC-LUVs⊃CF (Table 1). Overall poor efficiencies were
found for the release of CFD compared to CF (Fig. 6). Highest
efficiency to stimulate pR-mediated translocation of large anions
was found for planar coronene and pyrene carboxylates 3
and 6, whereas amphiphilic sulfates were inactive. Highest size
selectivity was observed for the more spherical calix[4]arene
activator 10 and pyrene hexanoate 7 with poor efficiency to
stimulate CFD release (Fig. 6) but acceptable efficiency with the
smaller CF (Table 1). The ability of activators such as 3 and 6
to mediate together with polyarginine the release of the large
CFD from vesicles suggested the involvement of mechanisms
other than carrier type,36 e.g., leakage through transient bilayer
defects (compare with below DPPC experiments). The different
kinetics observed for coronene and pyrene butyrates 3 and 6
hinted at the possibility that different activators may act by
different mechanisms (Fig. 6Aa and c).
g
g
>100
>100
34
19
16
18
1
3
1
1
1
1
1
1
1
9.3 0.3
>100
>100
g
g
a From CF efflux from LUVs⊃CF (250 lM lipid), 250 nM pR, 10 mM
NamHnPO4, 100 mM KCl, pH 6.5, 25 ◦C (if not otherwise indicated); part
of the data in Table 1 has been communicated previously in ref. 9. b See
Fig. 1 for structures. c From Hill analysis of dose response curves (e.g.,
Fig. 4D), see experimental part. d Maximal activity at saturation with
activator. e Effective concentration of anions 1–22 to mediate 50% of
maximal activity (YMAX/2). f Activator efficiency, eqn. (1). g No activity
observed.
Remarkably, the EC50 of fullerenes 1 (51 nM) and 2 (110 nM) was
below the binding of one activator per carrier (250 nM). In other
words, fullerenes activators appeared to act at “catalytic” con-
centrations. “Catalytic” activator efficiency demonstrated that
ꢁ
ꢁꢁ
activator–carrier complexes pR·(X)m ·(1)m form cooperatively
as well as that they are active at extremely low concentrations.
Activator efficiency decreased with decreasing carrier concen-
tration. The decrease in YMAX of coronene carboxylate 3 with
decreasing pR concentration, for instance, contributed more
to activator efficiency than the compensating decrease in EC50
(Fig. 5). The same trend was observed with the less efficient
coronene sulfate activator 4 (Table 2).
Fig. 6 (A) CF-dextran (CFD) efflux from EYPC-LUVs⊃CFD in the
presence of pR (250 nM) and activators 3 (a, 20 lM), 4 (b, 10 lM), 6 (c,
100 lM), 7 (d, 15 lM), 8 (e, 20 lM), 9 (f, 6 lM), 10 (g, 5 lM) and 18
(h, 30 lM, procedure as in Fig. 4), measured at 25 ◦C. (B) Comparison
of CF emission obtained for CF (ꢀ, Table 1) and CFD efflux (ꢁ, A)
ꢁ
ꢁꢁ
mediated by pR·(X)m ·(Y)m complexes at saturation with activators Y.
2.2 Activator–membrane interactions
Fig. 5 Dependence of activator efficiency on carrier concentration
at 25 ◦C. (A) Original curves for maximal activity at saturation with
coronene activator 3 (added at t = 60 s) for (a) 50 nM, (b) 125 nM and (c)
250 nM polyarginine (added at t = 120 s, procedure as in Fig. 4). (B) Dose
response curves for coronene activator 3 at polyarginine concentrations
of 50 nM (+), 125 nM (×) and 250 nM (ꢁ).
The main phase transition of DPPC-LUVs⊃CF occurs at
41.4 ◦C.62 Maximal efficiencies and EC50’s of pertinent activators
such as fullerene (1), coronene (3), pyrene (6) and calixarene
(10, 12) carboxylates plus cholesterol sulfate 20 were determined
below and above this temperature (Table 3). Efficiencies of all
tested activators decreased with decreasing membrane fluidity.
Whereas the EC50–T profile showed less significant changes
(Fig. 7B), the YMAX-T profile for pyrene sulfate 9 as represen-
tative activator exhibited the expected continuous increase with
increasing temperature (Fig. 7A).
Transport of anions as large as CF by supramolecular
ꢁ
ꢁꢁ
carriers Rn·(X)m ·(Y)m was remarkable. However CPPs have
been reported to mediate the uptake of even larger objects
such as proteins, liposomes or even nanoparticles.1,2,7,8,11–24 For
1 6 6 2
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9

View Article Online
Table 3 Characteristics of selected anion activators in DPPC-
LUVs⊃CF at constant lipid and carrier concentration as a function
of temperaturea,h
Table 4 Characteristics of selected anion activators in EYPC-
LUVs⊃CF at constant carrier concentration with and without choles-
terol and ergosterola
Entry
Anionb
T/◦Ci
YMAX (%)c,d EC50 (lM)c,e
Ef
Entry
Anionb
Sterolg
YMAX (%)c,d
EC50/lMc,e
Ef
g
01
02
03
04
05
06
07
08
09
10
11
12
13
14
1
25
60
25
60
25
60
25
60
25
60
25
60
25
60
—
40
70
83
65
91
18
48
—
79
29
90
3
>100
0.0
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
1
—
48
37
36
65
52
51
78
47
59
46
35
33
46
61
60
68
68
70
48
43
41
1
2
2
2
3
3
2
1
1
2
1
1
3
4
3
2
1
1
1
3
2
0.051 0.002
0.077 0.007
0.070 0.006
4.3 0.3
10.0
7.4
7.3
7.5
4.7
4.5
5.1
3.2
3.9
5.9
4.7
4.4
6.5
7.4
6.4
8.3
7.5
7.7
4.7
4.3
4.1
1
3
1
2
1
3
1
2
0.098 0.007
5.8 0.1
1.7 0.1
57
43
3.0 0.3
4.7 0.6
>100
0.80 0.04
38
15
26
13
7.8
7.6
11.2
3.9
6.0
2.2
5.4
0.0
11.9
2.0
8.0
0.3
4.7
1
chol
erg
—
chol
erg
—
chol
erg
—
chol
erg
—
chol
erg
—
chol
erg
—
3
1
3
3
6
1
2
3
14
15
44
39
43
1
1
2
1
2
6
3
9
6
9
6
g
10
10
12
12
20
20
6
2
1
1
1
3
9
2.2 0.2
1.6 0.1
1.7 0.1
1.2 0.2
3.1 0.4
6.3 0.8
3.1 0.2
5.5 0.1
5.4 0.1
9.2 0.3
8.6 0.7
9.1 0.5
1
1
1
1
9
9
10
10
10
12
12
12
20
20
20
51
a As in Table 1. b As in Table 1. c As in Table 1. d As in Table 1.
e As in Table 1. f As in Table 1. g As in Table 1. h DPPC: dipalmitoyl
phosphatidylcholine. i T: temperature.
chol
erg
a As in Table 1; measured at 25 ◦C. b As in Table 1. c As in Table 1. d As
in Table 1. e As in Table 1. f As in Table 1. g chol: 10 mol% cholesterol,
erg: 10 mol% ergosterol, total EYPC concentration constant.
activators. This interpretation was compatible with the results
from CF-dextran efflux (Fig. 6).
ꢁ
ꢁꢁ
The synergistic activity of complexes Rn·(X)m ·(Y)m did not
depend much on the presence of cholesterol (enriched in mam-
mals) and ergosterol (enriched in fungi)65,66 in EYPC-LUVs⊃CF
(Table 4). Activator efficiencies remained either nearly constant
(calixarenes 10 and 12) or decreased slightly. Reduced efficiencies
may, in part, originate from the increased order of sterol-rich
membranes. With respect to alternative modes of membrane
recognition, folate (21) was of interest because folate receptors
occur enriched at the surface of tumor cells.67,68 No activity
was found in sterol-free EYPC-LUVs⊃CF, presumably because
Fig. 7 (A) YMAX and (B) EC50 for pyrene sulfate 9 to activate pR
(500 nM) in DPPC-LUVs⊃CF as a function of temperature. The dotted
line indicates the phase transition of DPPC at 41.4 ◦C, data points were
taken at 25, 35, 45, 50 and 60 ◦C.
ꢁ
ꢁꢁ
the produced pR·(X)m ·(21)m complexes were too hydrophilic
The efficiency sequences in sol-phase DPPC, in gel-phase
DPPC and in sol-phase EYPC were not the same (Table 3).
The sequence fullerene > calix[4]arene ≈ coronene > pyrene in
fluid-phase EYPC at 25 ◦C (Table 1) changed to coronene >
calix[4]arene > fullerene > pyrene carboxylates in fluid-phase
DPPC at 60 ◦C (Table 3). The exceptional efficiencies of
coronene activator 3 and calix[4]arene activators 10 and 12
originated mainly from maximal activities YMAX ≥ 80% (Table 3,
entries 4 and 10). The EC50’s of coronene 3 and calix[4]arene 10
in DPPC bilayers at 60 ◦C were slightly lower than those in
EYPC bilayers at 25 ◦C.
(Table 1, entry 21).
Taken together, the efficiency of activator–carrier complexes
depended on membrane fluidity rather than composition.
This trend was more pronounced for spherical (fullerenes,
calixarenes) than for planar activators (coronenes, pyrenes).
These results are in agreement with the preferential interfacial
partitioning of aromatic amphiphiles, rather than molecular
recognition within the hydrophobic core of the bilayer.
The relevance of anion activation of oligo/polyarginines to
penetrate live cells was demonstrated previously by incubating
HeLa cells with pyrenebutyrate 6 before addition of R8-alexa,
a fluorescently labeled octa-L-arginine.9 According to flow
cytometry detection after removal of the cell surface-adsorbed
peptides by trypsin treatment, CPP uptake increased up to 4-
times in the presence of pyrenebutyrate 6. Measured under
nearly identical conditions, the maximal increase of CPP uptake
observed in the presence of less sterol-sensitive calix[4]arene 10
(maximal 1.5-fold increase, Fig. 8A) as well as the more sterol-
sensitive, EYPC-selective fullerene carboxylate 1 (maximal 1.2-
fold increase, Fig. 8B) were clearly below that of pyrenebutyrate
6. At high concentrations, both activators became inhibitors of
CPP uptake. The attractive implication that spherical activators,
particularly fullerene 1, may specifically mediate CPP-uptake in
bacteria will be investigated in due course.
Remarkably, the maximal activity of coronene-carrier com-
◦
ꢁ
ꢁꢁ
plexes Rn·(X)m ·(3)m in crystalline DPPC at 25 C (Table 3, entry
3) exceeded that in liquid-crystalline EYPC at 25 ◦C slightly
(Table 1, entry 3). This distinct DPPC > EYPC selectivity of
coronenebutyrate 3 was complementary to fullerene activator
1 with maximal EYPC > DPPC selectivity (Table 3, entries 1
and 2 versus Table 1, entry 1). Overall, the efficiency sequence
in crystalline DPPC bilayers at 25 ◦C was coronene > pyrene >
calix[4]arene > fullerene carboxylates. Calix[4]arene 10 and
fullerene 1 were completely inactive (Table 3, entries 1 and
9). High activities of coronenes and pyrenes in DPPC vesicles
might relate to the preference of polyaromatic hydrocarbons to
partition into liquid ordered phase rather than liquid disordered
phase of a bilayer.63,64 This sequence might also suggest that
mechanisms other than a carrier mechanism36 contribute to the
2.3. Activator–carrier interactions
ꢁ
ꢁꢁ
activity of complexes Rn·(X)m ·(Y)m with coronene and pyrene
ꢁ
ꢁꢁ
activators but, with all appropriate caution, presumably not
In water, pyrenebutyrate complexes pR·(X)m ·(6)m exhibited
ꢁ
ꢁꢁ
to that of Rn·(X)m ·(Y)m carriers with fullerene and calixarene
characteristic excimer emission49,50,69–72 at 470 nm with a maximal
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 6 3

View Article Online
carrier into account, EC50 values obtained from the excimer
emission (referring to pR concentration) and from functional
studies (referring to anion concentration) are in good agreement.
Contrary to that of pyrene carboxylate 6, the EC50 with respect
to maximal excimer emission of pyrene sulfate 9 increased
with increasing ionic strength (Fig. 10A). These differences
suggested that ion pairing is dominant to the stability of the less
ꢁ
ꢁꢁ
efficient complexes with sulfate activators like pR·(X)m ·(9)m
,
whereas hydrophobic interactions appear to contribute more
significantly to the more efficient complexes with carboxylates
like pyrene 6.
Fig. 8 Uptake ratios from flow cytometry for R8-alexa by HeLa cells
as a function of the concentration of activator 10 (A) and 1 (B) relative
to activator-free uptake.
excimer/monomer ratio of E/M = 0.07.1 This excimer emission
ꢁ
ꢁꢁ
ꢁ
ꢁꢁ
increased for complexes pR·(X)m ·(7)m and pR·(X)m ·(8)m to
ꢁ
ꢁꢁ
a spectacular maximal E/M = 4.3 for complex pR·(X)m ·(9)m
(Fig. 9 and Table 5).
Fig. 10 Dependence of the formation of activator–pR complexes with
pyrene carboxylate 6 (᭹), pyrene sulfate 9 (᭺) as well as coronene
carboxylate 3 (ꢁ) on (A) ionic strength (at 25 ◦C) and (B) temperature
(at 107 mM NaCl). EC50’s and DG’s were determined from Hill analysis
of the changes in activator emission with increasing pR concentration
as illustrated in Figs. 9 and 11.
This interpretation was supported by the dependence of
ꢁ
ꢁꢁ
complex pR·(X)m ·(6)m on temperature (Fig. 10B). As expected
for the exchange of high-affinity phosphate counteranions
with carboxylates of lower affinity,39–45 complex formation was
clearly endothermic (DH = +29.7 kJ mol−1). Overcompensation
of these losses by massive entropy gains corroborated the
importance of desolvation of ions as◦ well as large aromatic
surfaces (TDS = +64.2 kJ mol−1 at 25 C).
Fig. 9 Emission spectra of anion 9 (12 lM, kex = 340 nm) in water
in the presence of, with increasing intensity at 470 nm, 0, 50, 100, 150,
200, 250, 300, 350, and 400 nM pR (10 mM NamHnPO4, 107 mM NaCl,
pH 7.4, 25 ◦C); normalized to monomer emission at 375 nm.
E/M ratios as well as EC50’s decreased with decreasing
concentration of activator–pR-complexes. E/M–pR profiles
were recorded at the lowest detectable anion concentration to
obtain thermodynamic data (Table 5). Despite the difference
in the experiments, the obtained EC50’s at high dilution and
the EC50’s obtained from function (Table 1) showed the same
trend (9 ≤ 8 < 7 < 6). Taking the multivalent nature of the
Compared to the pyrene series, the emission spectra of
activator–pR complexes with coronene 3 exhibited quite strong
excimer emission,69 reaching E/M ratios up to 28 (Fig. 11).
Their “noisy” emission was indicative of scattering due to
ꢁ
ꢁꢁ
flocculation of the resulting complexes pR·(X)m ·(3)m . Even
stronger scattering was found in the emission spectra of the
ꢁ
ꢁꢁ
.
corresponding coronene sulfate complexes pR·(X)m ·(4)m
Table 5 Spectroscopic and thermodynamic data of activator–pR com-
plexes for coronene and pyrene anions 3 and 5–9 in watera
Entry
Anione
E/Mf
EC50/nMg
01
02
03
04
05
06
3
5
6
7
8
9
28.0c, 1.36d
20.3 1d
h
h
—
—
0.07b
920 60b
150 10c
122 4c
0.65b, 0.08c,
1.90b, 0.10c
4.27b, 3.91c, 1.88d
20.0 1d
a From fractional changes in emission intensities at high (pyrenes,
375 nm; coronenes, 427 nm) and low energy (pyrenes, 470 nm;
coronenes, ∼500 nm) at constant anion concentration and varied pR
concentration (107 mM NaCl, 10 mM NamHnPO4, pH 7.4, 25 ◦C).
b Anion concentration 120 lM. c Anion concentration 12 lM. d Anion
concentration 1.2 lM. e See Fig. 3 for structures. f Excimer/monomer
emission ratio at saturation with pR (entry 1: Fig. 11, entry 6: Fig. 9).
g Determined at minimal detectable anion concentration, from E/M-pR
profiles, analyzed with eqns. (4)–(5), see experimental part. h No excimer
emission observed.
Fig. 11 Emission spectra of coronene carboxylate 3 (12 lM, kex
=
340 nm) in water in the presence of, with increasing intensity at ∼500 nm,
0, 50, 100, 150, 200, 250, 500, 1200, 4000 and 5000 nM pR (10 mM
NamHnPO4, 107 mM NaCl, pH 7.4), normalized to maximal monomer
emission at 427 nm. Dotted spectrum: Magnified emission spectrum of
3 without pR (21×).
1 6 6 4
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9

View Article Online
According to their excimer emission, the EC50’s of coroneneb-
ꢁ
ꢁꢁ
utyrate complexes pR·(X)m ·(3)m decreased with increasing
ionic strength (Fig. 10A, ꢀ), nicely corroborating results with
ꢁ
ꢁꢁ
,
the corresponding pyrenebutyrate complexes pR·(X)m ·(6)m
this decrease implied dominance of hydrophobic interactions
in complexes with comparably weak carboxylate–guanidinium
pairs (Fig. 10A, ꢀ and ᭹). The values at high ionic strength
may contain contributions from self-assembly of coronene 3
observed under these conditions. Compared to that of pyrene
ꢁ
ꢁꢁ
carboxylate complex pR·(X)m ·(6)m , the lower EC50 of coronene
ꢁ
ꢁꢁ
carboxylate complex pR·(X)m ·(3)m corroborated the increased
activator affinity with increasing aromatic surface area (Table 5,
entry 1 versus entry 3).
The excimer emission of pyrene–polyarginine complexes
allowed us to detect the formation of “invisible” anion–pR
complexes as well. The excimer emission of pyrene complex
ꢁ
ꢁꢁ
pR·(X)m ·(8)m , for instance, decreased in the presence of in-
creasing concentrations of calixarene 12 (Fig. 12). This change
in emission reflected the transformation of the “visible” complex
ꢁ
ꢁꢁ
ꢁ
ꢁꢁ
pR·(X)m ·(8)m into the “invisible” complex pR·(X)m ·(12)m
.
Competition experiments with other anion activators gave
analogous results and could be used, if desired, to determine
global dissociation constants of any activator-carrier complex
of interest.73,74
Fig. 13 Arene-templated ion pairing exemplified with notional
ꢁ
ꢁꢁ
structures of activator–carrier complexes pR·(X)m ·(3)m (A) and
ꢁ
ꢁꢁ
pR·(X)m ·(1)m (B); additional possibilities like binding of phos-
phate–guanidinium pairs on p-surfaces are not depicted.
3
Summary and conclusions
The objective of this study was to elaborate on the nature
and selectivity of the interactions of selected anion activa-
tors with polyarginine (pR) carriers on the one hand and
lipid bilayer membranes on the other hand. We found that
activator–membrane interactions are dominated by the physical
properties of the bilayer membrane with little recognition of
specific components like ergosterol or cholesterol. This result
was consistent with interface-directed translocation28,29 as one
possible origin of the superb activator efficiency of higher
aromatics like amphiphilic fullerenes, calixarenes and coronenes.
Reduced dependence on membrane fluidity and on the size of the
transported anion with planar coronene and pyrene compared
to spherical fullerene and calix[4]arene activators suggested
more significant contributions from mechanisms other than
carriers with the former. In HeLa cells, planar pyrenes were
clearly more efficient CPP-activators than spherical fullerenes
and calixarenes, a trend that might be reversed in bacteria.
Ideally, activators should bind tightly to carriers without
sacrificing the anion exchange ability to retain carrier activ-
ities. Efficient activators are shown to fulfil this requirement
with large aromatic surfaces for, presumably, arene-templated
carboxylate–guanidinium pairing38 stabilized by multiple inter-
actions remote48 from the active sites. Maybe accounting for
interface-directed translocation as well, arene-templated ion
pairing supported by remote contacts on large surfaces seems
to reach perfection with fullerene 1 (Fig. 13B). The spectacular
spectroscopic characteristics discovered for certain activator–
carrier complexes open attractive perspectives with regard to
sensing applications.
Fig. 12 Counteranion exchange assay for adaptable fluorometric
detection of activator–carrier complexes like calix[4]arene complex
ꢁ
ꢁꢁ
pR·(X)m ·(12)m . Emission spectra of anion 8 (120 lM, kex = 340 nm) in
water with pR (2.5 lM) and, with decreasing intensity at 470 nm, 0, 1,
1.5, 1.75, 2, 2.5, 3, 5, 10, 15 and 30 lM calixarene 12 (10 mM NamHnPO4,
107 mM NaCl, pH 7.4); normalized to maximal monomer emission at
375 nm.
Overall, the outcome of spectroscopic studies of complexes
ꢁ
ꢁꢁ
pR·(X)m ·(Y)m detected by excimer emission of pyrene and
coronene activators confirmed that higher aromatics improve
and high-affinity ion pairing weakens activator efficiency. The
difference between the poor sulfate and the good carboxylate
activators turned out to originate indeed from the dominant
strong ion pairing which probably slows the anion exchange in
the former, and the availability of supportive interactions such
as hydrophobic contacts in the latter. This result supported the
importance of arene-templated ion-pairing for function on the
structural level. The increase in complex stability with increasing
aromatic surface from pyrene to coronene revealed the impor-
tance of non-interfering remote contacts, i.e., interactions far
from the active site,48 to support arene-templated ion pairing38
without loss in activity. The two carboxylates positioned above
an extended p-surface of the best activator, i.e., the Bingel
fullerene 1, appear perfectly preorganized for arene-templated
ion pairing (Fig. 13B). Possibilities for remote contributions to
complex stability include additional association of guanidinium
cations on the extended aromatic surfaces,75 p–p interactions
between proximal activators, and so on.
4
Experimental
4.1 Materials
Poly-L-arginine (HCl salt, MW 14 000, DP 72), sodium dodecyl
sulfate (SDS), triton X-100, buffers and salts were purchased
from Sigma, 4-pyrene-1-yl-butyric acid from Acros Organics,
coronene from Aldrich, fullerene C60 from Alfa Aesar, egg yolk
phosphatidylcholine (EYPC) and dipalmitoyl phosphatidyl-
choline (DPPC) from Avanti, cholesterol sulfate from Northern
Lipids, mono-n-dodecylphosphate from Lancaster and reagents
for synthesis, 5(6)-carboxyfluorescein (CF), 2-pyrene-1-yl-acetic
acid, 6-pyrene-1-yl-caproic acid, cholesterol, ergosterol and
lauric acid from Fluka. All reactions were performed under
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 6 5

View Article Online
N2 or argon atmosphere. Column chromatography was carried
out on silica gel 60 (Fluka, 40–63 lm), reverse phase column
chromatography on ODS (Fluka Silicagel 100 C18-Reverse
Phase), analytical thin layer chromatography (TLC) on silica gel
60 (Fluka, 0.2 mm) and preparative TLC (PTLC) on silica gel
GF (Analtech, 1000 lm). Melting points (mp) were recorded on
a heating table from Reichert (Austria). ESI-MS and APCI-MS
were performed on a Finnigan MAT SSQ 7000 instrument. 1H
and 13C spectra were recorded (as indicated) on either a Bruker
300 MHz, 400 MHz or 500 MHz spectrometer and are reported
as chemical shifts (d) in ppm relative to TMS (d = 0). Spin
multiplicities are reported as a singlet (s), doublet (d), triplet
(t), with coupling constants (J) given in Hz, or multiplet (m).
1H and 13C resonances were assigned with the aid of additional
information from 2D NMR spectra (H,H-COSY, DEPT 135,
HSQC and HMBC). UV-Vis spectra were measured on a Varian
Cary 1 Bio spectrophotometer. Fluorescence measurements
were preformed on either a FluoroMax-2 or a FluoroMax-
3, Jobin Yvon-Spex. The Mini-Extruder with a polycarbonate
membrane, pore size 100 nm, used for LUV preparation was
from Avanti.
(520 ll, 6.4 mmol) and DMAP (4.0 mg, 33 lmol) in dried
THF and malonyl chloride 26 (296 ll, 3.0 mmol) were added
at room temperature. The mixture was stirred overnight. After
evaporation of THF, CH2Cl2 (20 ml) was added, and the
organic layer was washed with water (20 ml), dried over MgSO4,
concentrated in vacuo and purified by column chromatography
(CH2Cl2–MeOH 20 : 1, Rf = 0.7) to give 25 (1.56 g, 71%) as a
yellow oil. 1H NMR (400 MHz, CDCl3): d 7.71 (d, 3J (H,H) =
7.6 Hz, 8H), 7.40–7.30 (m, 12H), 4.09 (t, 3J (H,H) = 6.6 Hz, 4H),
3.59 (t, 3J (H,H) = 6.3 Hz, 4H), 3.26 (s, 2H), 1.69 (tt, 3J (H,H) =
3
3
6.6 Hz, J (H,H) = 8.3 Hz, 4H), 1.53 (tt, J (H,H) = 6.3 Hz,
3J (H,H) = 8.3 Hz, 4H), 1.11 (s, 18H); 13C NMR (100 MHz,
CDCl3): 166.7 (s), 135.6 (d), 133.9 (s), 129.7 (d), 127.7 (d), 65.5
(t), 63.3 (t), 28.9 (t), 28.3 (t), 26.9 (q), 25.1 (t), 19.3 (s); MS (ESI,
MeOH): m/z (%) 726 (100) [M + H]+.
Fullerene 27. To a solution of 28 (50 mg, 69.4 lmol), 25
(55 mg, 76 lmol) and I2 (22 mg, 86 lmol) in dried toluene
(100 ml), DBU (27 ll, 170 lmol) was added. The mixture was
stirred for 12 hours at room temperature. After evaporation of
toluene, the crude product was filtered through a short plug
of silica, eluting first with toluene (to remove unreacted 28)
and then with CH2Cl2. Purification by PTLC (petroleum ether–
EtOAc 10 : 1, Rf = 0.5) gave pure compound 27 (26 mg, 26%) as
a gummy oil. 1H NMR (400 MHz, CDCl3): d 7.58 (d, 3J (H,H) =
7.6 Hz, 8H), 7.40–7.25 (m, 12H), 4.42 (t, 3J (H,H) = 6.6 Hz, 4H),
3.63 (t, 3J (H,H) = 6.6 Hz, 4H), 1.86 (tt, 3J (H,H) = 7.8 Hz, 3J
Abbreviations. CF: 5(6)-carboxyfluorescein; DBU: 1,8-diaza-
bicyclo[5.4.0]undec-7-ene; DIPEA: N,Nꢁ-diisopropylethyl-
enediamine; DMAP: 4-dimethylaminopyridine; DMF: N,N-
dimethylformamide; EYPC-LUVs: egg yolk phosphatidylcho-
line large unilamellar vesicles; TBDPS: tetrabutyldiphenylsilyl;
THF: tetrahydrofuran.
3
(H,H) = 6.6 Hz, 4H), 1.63 (tt, 3J (H,H) = 6.6 Hz, J (H,H) =
7.8 Hz, 4H), 1.00–0.98 (m, 18H); 13C NMR (100 MHz, CDCl3):
d 162.6 (s), 144.3 (s), 144.2 (s), 144.1 (s), 144.1 (s), 143.9 (s),
143.7 (s), 143.6 (s), 142.9 (s), 142.0 (s), 142.0 (s), 142.0 (s), 141.1
(s), 140.9 (s), 137.9 (s), 134.5 (d), 132.7 (s), 128.6 (d), 126.7 (d),
70.6 (s), 66.3 (t), 62.1 (t), 51.3 (s), 27.9 (t), 25.9 (q), 24.2 (d), 18.2
(s); MS (ESI, CH2Cl2): m/z (%) 1443 (45) [M + H]+, 722 (100)
[M + 2H]2+, 721 (10) [C60 + H]+.
4.2 Synthesis
Anion activators 1, 3, 8–15 and 19 were prepared following
previously reported procedures.9,51,58,60 Stock solutions were
prepared for all anions in buffer except for 1 and 2 (DMSO–
water 1 : 1) and 3 and 4 (DMSO/water 9 : 1). The pH
of all stock solutions was measured and accurately adjusted
if necessary; concentrations were confirmed, if applicable, by
UV/vis spectroscopy (if applicable).
Fullerene diol 29. A solution of 27 (25 mg, 17 lmol) in
THF (3 ml) was treated successively with acetic acid (20 ll,
0.35 mmol) and tetrabutylammonium fluoride (1 M in THF,
350 lL, 0.35 mmol) at room temperature. After stirring for
16 hours, the reaction mixture was diluted with CHCl3–MeOH
1 : 1 (40 ml), washed with water (20 ml), dried over MgSO4,
concentrated in vacuo and purified by column chromatography
(CH2Cl2–MeOH 8 : 1 then 4 : 1, Rf = 0.6) to give pure 29 (10 mg,
4.2.1 Fullerene disulfate 2 (disodium salt). This compound
was prepared in overall five steps as outlined in Scheme 1.
1
62%) as a brown oil. H NMR (300 MHz, CD3OD–CDCl3): d
3
3
4.05 (t, J (H,H) = 6.4 Hz, 4H), 3.55 (t, J (H,H) = 6.3 Hz,
4H), 1.70–1.55 (m, 4H), 1.55–1.40 (m, 4H); 13C NMR (75 MHz,
CD3OD–CDCl3): d 163.6 (s), 145.3 (s), 145.3 (s), 145.2 (s), 144.7
(s), 144.6 (s), 143.9 (s), 143.0 (s), 143.0 (s), 142.2 (s), 141.9 (s),
141.0 (s), 139.0 (s), 129.0 (s), 71.6 (s), 63.4 (t), 61.6 (t), 29.3 (t),
26.2 (d); MS (ESI, CH2Cl2): m/z (%) 967 (100) [M + H]+.
Fullerene disulfate 2 (disodium salt). Fullerene 29 (10 mg,
10 lmol) in anhydrous ether (4 ml) was added to chlorosulfonic
acid (14 ll, 21 lmol) in anhydrous ether (2 ml) at 0 ◦C. The
mixture was purged with nitrogen for an additional 30 minutes
to remove HCl, diluted with water (4 ml), 1 M NaOH (to reach
pH ∼7) and isopropanol (5 ml), washed with hexane (2 × 5 ml)
and concentrated in vacuo. Purification by reverse phase column
chromatography (MeOH, Rf = 0.5) yielded 2 (9 mg, 80%) as a
yellowish powder. mp: >230 ◦C; 1H NMR (300 MHz, CD3OD):
d 3.91 (t, 3J (H,H) = 6.4 Hz, 4H), 3.79 (t, 3J (H,H) = 6.3 Hz, 4H),
2.14 (tt, 3J (H,H) = 6.4 Hz, 3J (H,H) = 7.3 Hz, 4H), 1.89 (tt, 3J
(H,H) = 7.3 Hz, 3J (H,H) = 6.3 Hz, 4H); 13C NMR (75 MHz,
CD3OD): d 163.6 (s), 145.3 (s), 145.3 (s), 145.2 (s), 145.2 (s),
144.7 (s), 144.6 (s), 143.9 (s), 143.1 (s), 143.1 (s), 142.2 (s), 141.9
(s), 140.9 (s), 139.0 (s), 128.2 (s), 71.6 (s), 64.9 (t), 63.4 (t), 49.8
(s), 23.6 (t), 21.9 (d); MS (ESI negative mode, MeOH); m/z (%)
1147 (100) [M − Na]−, 562 (45), [M − 2Na]2−.
Scheme 1 Conditions: a) TBDPSCl, DIPEA, CH2Cl2, rt, 2 h, 95%; b)
26, pyridine, DMAP, THF, rt, 12 h, 71%; c) 28, I2, DBU, toluene, rt,
12 h, 26%; d) AcOH, Bu4NF, THF, rt, 16 h, 62%; e) 1. ClSO3H, ether,
0 ◦C, 10 min, 2. NaOH, 80%.
4-(tert-Butyl-diphenyl-silanyloxy)-butan-1-ol 23. To a solu-
tion of 24 (5.0 g, 55 mmol) in CH2Cl2 (10 ml) containing DIPEA
(10 ml), TBDPSCl (5 ml, 18 mmol) was added dropwise at room
temperature. The solution was stirred for 2 hours, concentrated
in vacuo and purified by column chromatography (hexane–
EtOAc 10 : 1, Rf = 0.8) to give 23 (5.63 g, 95%) as a colorless
oil. 1H NMR (400 MHz, CDCl3): d 7.74 (d, 3J (H,H) = 7.8 Hz,
4H), 7.60–7.40 (m, 6H), 3.76 (t, 3J (H,H) = 5.8 Hz, 2H), 3.72 (t,
3J (H,H) = 5.8 Hz, 2H), 2.20 (broad s, 1H, exchange with D2O),
1.80–1.69 (m, 4H), 1.12 (s, 9H); 13C NMR (100 MHz, CDCl3):
d 135.6 (d), 133.7 (s), 129.7 (d), 127.7 (d), 67.0 (t), 62.9 (t), 29.9
(t), 29.3 (t), 26.9 (q), 19.2 (s); MS (ESI, MeOH): m/z (%) 330
(100) [M + H]+.
4.2.2 4-Coronen-1-yl-butylsulfate
4 (sodium salt). This
Malonic acid bis-[4-(tert-butyl-diphenyl-silanyloxy)-butyl]
ester 25. To a solution of 23 (2.0 g, 6.1 mmol), pyridine
compound was prepared from 4-coronen-1-yl-butyric acid 3 in
overall two steps.
1 6 6 6
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9

View Article Online
4-Coronenyl-1-butanol. A solution of 3 (240 mg, 0.62 mmol)
in THF (16 ml) was transferred slowly to ◦a solution of LiAlH4
(186 mg, 5.00 mmol) in THF (3 ml) at 0 C. The mixture was
stirred for 2 hours at 60 ◦C. The reaction mixture was cooled at
0 ◦C and water was added drop by drop. The reaction mixture
was stirred for 20 min, filtered (washed with THF and CH2Cl2–
MeOH) and concentrated in vacuo. Purification of the crude
product by column chromatography (CH2Cl2, then CH2Cl2–
EtOAc 10 : 1, Rf = 0.4) yielded 4-coronenyl-1-butanol (128 mg,
55%) as a yellow solid. Mp: 205.0–206.3 ◦C; IR: m 3238 (m),
3019 (s), 1611 (s), 1461 (s), 1052 (s) cm−1; 1H NMR (400 MHz,
DMSO-d6): d 9.24–9.00 (m, 10H), 8.93 (s, 1H), 4.46 (t, 3J
7.4 mM HEPES, 37 mM NaCl, pH 7.4 (LUVs⊃CFD); outside:
10 mM NamHnPO4, 107 mM NaCl, pH 7.4.
4.4 Vesicle experiments
LUV stock solutions (1.3 mM lipid, above) were diluted with
buffer (10 mM NamHnPO4, 107 mM NaCl, pH 7.4) to give
≈13 lM lipid, placed in a thermostated fluorescence cuvette
(25 ◦C or as indicated) and gently stirred. CF efflux was followed
at kem 517 nm (kex 490 nm) as a function of time during the
addition of anion activators at varied concentration (1–22, 20–
40 ll from stock solutions, final concentrations as indicated),
polyarginine (250 nM) and aqueous triton X-100 at the end of
each experiment for calibration (0.024%). Data were normalized
to fractional emission intensity I using eqn. (2)
3
(H,H) = 5.3 Hz, 1H, exchange with D2O), 3.79 (t, J (H,H) =
3
3
7.9 Hz, 2H), 3.57 (dd, J (H,H) = 6.4 Hz, J (H,H) = 5.3 Hz,
2H), 2.13 (tt, 3J (H,H) = 7.9 Hz, 3J (H,H) = 7.5 Hz, 2H), 1.77
3
3
(tt, J (H,H) = 7.5 Hz, J (H,H) = 6.4 Hz, 2H); 13C NMR
(100 MHz, DMSO-d6): d 137.7 (s), 128.3 (s), 128.3 (s), 128.2 (s),
128.1 (s), 127.9 (s), 127.1 (s), 126.4 (d), 126.4 (d), 126.4 (d), 126.3
(d), 126.3 (d), 126.3 (d), 126.1 (d), 125.9 (d), 125.9 (d), 125.9 (d),
122.7 (d), 122.1 (s), 122.0 (s), 121.7 (s), 121.6 (s), 121.4 (s), 121.5
(s), 60.6 (t), 33.1 (t), 32.7 (t), 27.6 (t); MS (CI, DMF): m/z (%)
372 (100) [M]+.
I = (Ft − F0)/(F∞ − F0)
(2)
where F0 = Ft at polyarginine addition, F∞ = Ft at saturation
after lysis. Effective concentrations EC50 and Hill coefficients
were determined by plotting the fractional activity Y (= I just
before lysis) as a function of anion concentration and fitting
them to the Hill eqn. (3)
Y = Y0 + (YMAX − Y0)/{1 + (EC50/cANION)n}
(3)
4-Coronen-1-yl-butylsulfate 4 (sodium salt). A solution of 4-
coronenyl-1-butanol (91 mg, 0.24 mmol) in DMF (1.5 ml) was
transferred slowly to a solution of chlorosulfuric acid (35 mg,
0.30 mmol) in THF (1.5 ml) at 0 ◦C. The mixture was stirred for
1 hour at room temperature and purged with nitrogen to remove
HCl. The mixture was diluted with water (2 ml), 1 M NaOH (to
reach pH ∼7), water (50 ml) and 2-propanol (50 ml), washed
with petroleum ether (2 × 30 ml) and concentrated in vacuo.
Purification by reverse phase column chromatography (H2O–
acetone 1 : 1 Rf = 0.6, first with H2O then to H2O–acetone 5 :
1 to 3◦: 1) yielded pure 4 (70 mg, 63%) as a yellow solid. mp:
where Y0 is Y without anion, YMAX is a value with an excess
anion at saturation, cANION is the anion concentration in cuvette
and n is the Hill coefficient. Activator efficiencies E were
calculated using eqn. (1), where YMAX is the maximal activity
(%), EC50 the effective concentration (mM), and f = 20.6 an
arbitrary scaling factor. Pertinent E, YMAX, and EC50 values are
listed in Tables 1–4.
4.5 Flow cytometry with HeLa cells
1
>230 C. IR: m 3467 (m), 3016 (s), 1613 (s), 1216 (s) cm−1; H
The influence of counteranion 1 and 10 on the uptake of R8-
alexa by HeLa cells was determined using flow cytometry as
described previously.9 In brief, R8-alexa was preparedby reacting
H-(Arg)8-Gly-Cys-NH2 with 1.5 eq. alexa 488 C5 maleimide
(Molecular Probes) in DMF–MeOH (1 : 1, 1.5 h), purified
with RP-HPLC and characterized by MALDI-TOFMS (calcd.
2126.4, found 2127.2). Human cervical cancer-derived HeLa
cells (1.5 × 105 cells in 1.5 ml of a-MEM containing 10% (v/v)
calf serum) were cultured on 60 mm dishes (48 h) and washed
twice with PBS after removal of the medium. After incubation
of the cells in the presence or absence of fullerene 1 or calixarene
10 (300 ll PBS, 5 min, 37 ◦C), addition of R8-alexa (100 ll PBS,
final peptide concentration 5 lM), incubation (15 min, 37 ◦C),
washing (PBS, 3×), incubation with 0.01% trypsin (400 ll,
10 min, 37 ◦C), addition of PBS (600 ll) and centrifugation
(2,000 rpm, 5 min), the cell pellet was suspended, washed (2 ×
1 ml PBS) and finally resuspended in 1 ml of PBS for fluorescence
analysis with a FACScalibur (BD Biosciences) flow cytometer
using a 488 nm laser excitation and a 515–545 nm emission filter.
Each sample was analyzed for 10 000 events.
NMR (400 MHz, DMSO-d6): d 9.22–9.00 (m, 10H), 8.94 (s, 1H),
3.91 (t, 3J (H,H) = 6.5 Hz, 2H), 3.77 (t, 3J (H,H) = 7.8 Hz, 2H),
2.13 (tt, 3J (H,H) = 7.8 Hz, 3J (H,H) = 7.2 Hz, 2H), 1.87 (tt, 3J
(H,H) = 7.2 Hz, 3J (H,H) = 6.5 Hz, 2H); 13C NMR (100 MHz,
DMSO-d6): d 137.6 (s), 128.3 (s), 128.3 (s), 128.2 (s), 128.1 (s),
127.4 (s), 127.1 (s), 126.4 (d), 126.4 (d), 126.4 (d), 126.3 (d), 126.3
(d), 126.3 (d), 126.2 (d), 126.2 (d), 126.1 (d), 125.9 (d), 122.8 (d),
122.1 (s), 122.0 (s), 121.8 (s), 121.6 (s), 121.4 (s), 120.5 (s), 65.4
(t), 33.0 (t), 29.4 (t), 27.7 (t); MS (ESI negative mode, DMF):
m/z (%) 451 (100) [M − Na]−.
4.3 Vesicle preparation
Solutions of EYPC (for EYPC-LUVs), DPPC (for DPPC-
LUVs), and mixtures of EYPC and either ergosterol (erg, for
EYPC/erg-LUVs) or cholesterol (chol, for EYPC/chol-LUVs)
(25 mg) in CHCl3–MeOH 1 : 1 (2 ml) were dried under a
stream of N2 and then under vacuum (>2 h) to form thin
films. The resulting films were hydrated with 1 ml buffer A
(50 mM CF, 10 mM NamHnPO4, 10 mM NaCl, pH 7.4 (for
LUVs⊃CF) or 7.4 mM CF-dextran, 7.4 mM HEPES, 37 mM
NaCl, pH 7.4 (EYPC-LUVs⊃CFD)) for more than 30 min,
subjected to freeze–thaw cycles (5×) and extrusions (15×, Mini-
Extruder with two stacked polycarbonate membranes, pore
size 100 nm). Extravesicular CF/CFD was removed by gel
filtration (Sephadex G-50 for LUVs⊃CF, Sephacryl S300-HR
for LUVs⊃CFD) with buffer B (10 mM NamHnPO4, 107 mM
NaCl, pH 7.4). The LUV fractions were combined and diluted
to 6 ml with buffer B. Lipid concentrations were estimated
from amount of entrapped CF. The found values were in
agreement with earlier results from phosphate analysis (due to
10 mM phosphate in the buffer, phosphate analysis did not give
accurate results). The final stock solutions had the following
characteristics: ∼1.3 mM lipid; inside: 50 mM CF, 10 mM
NamHnPO4, 10 mM NaCl, pH 7.4, (LUVs⊃CF) or 7.4 mM CFD,
4.6 Pyrene excimers
Solutions of pyrene anions 5–9 (0.12–120 lM) in buffer (2 ml,
10 mM NamHnPO4, 0–2 M NaCl, pH 7.4) were placed in a cu-
vette. Then, pR was added stepwise (concentrations appropriate
to follow the appearance of excimer emission, see Table 5) and
the solutions were gently mixed. Fluorescence emission spectra
of the resulting solutions were measured with kex 340 nm at
25–60 ◦C. The obtained spectra were normalized to maximal
monomer emission intensity at 375 nm. E/M ratios IEM
470/I375 for given anions, NaCl concentrations and temperature
=
I
were plotted as a function of pR concentration and analyzed
using the eqn. (4)
EM
EM
IEM = I0 + (I∞ − I0EM)/[1 + (EC50/c)n]
(4)
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 6 7

View Article Online
where IEM is the response (relative excimer emission intensity),
12 G. Dom, C. Shaw-Jackson, C. Matis, O. Bouffioux, J. J. Picard, A.
Prochiantz, M. P. Mingeot-Leclercq, R. Brasseur and R. Rezsohazy,
Nucleic Acids Res., 2003, 31, 556–561.
13 P. E. G. Thore´n, D. Persson, E. K. Esbjo¨rner, M. Gokso¨r, P. Lincoln
and B. Norde´n, Biochemistry, 2004, 43, 3471–3489.
14 A. Ziegler, X. L. Blatter, A. Seelig and J. Seelig, Biochemistry, 2003,
42, 9185–9194.
EM
EM
I0 the initial value, I∞ the value at saturation, c the pR
concentration, and n the Hill coefficient, to obtain EC50’s. The
calculated apparent KD’s (EC50’s) were then plotted as a function
of the concentration of NaCl (Fig. 10A) or, converted into DG
using eqn. (5)
15 H. Binder and G. Lindblom, Biophys. J., 2003, 85, 982–995.
16 K. Takeshima, A. Chikushi, K. Lee, S. Yonehara and K. Matsuzaki,
J. Biol. Chem., 2003, 278, 1310–1315.
17 S. D. Kra¨mer and H. Wunderli-Allenspach, Biochim. Biophys. Acta,
2003, 1609, 161–169.
18 S. M. Fuchs and R. T. Raines, Biochemistry, 2004, 43, 2438–2444.
19 P. E. G. Thore´n, D. Persson, P. Isakson, M. Gokso¨r, A. Onfelt and
B. Norde´n, Biochem. Biophys. Res. Commun., 2003, 307, 100–107.
20 D. Persson, P. E. G. Thore´n, M. Herner, P. Lincoln and B. Norde´n,
Biochemistry, 2003, 42, 421–429.
21 J. L. Zaro and W. C. Shen, Biochem. Biophys. Res. Commun., 2003,
307, 241–247.
22 D. Seebach, K. Namoto, R. Yogesh, Y. R. Mahajan, P. Bindscha¨dler,
R. Sustmann, M. Kirsch, N. S. Ryder, M. Weiss, M. Sauer, C. Roth,
S. Werner, H.-D. Beer, C. Munding, P. Walde and M. Voser, Chem.
Biodiv., 2004, 1, 65–97.
23 T. B. Potocky, A. K. Menon and S. H. Gellman, J. Biol. Chem., 2003,
278, 50188–50194.
24 J. P. Richard, K. Melikov, E. Vives, C. Ramos, B. Verbeure, M. J.
Gait, L. V. Chernomordik and B. Lebleu, J. Biol. Chem., 2003, 278,
585–590.
DG = −RTln KD
(5)
as a function of temperature (R = 8.315 J K−1 mol−1, T = 298 K,
Fig. 10B). Additional thermodynamic data were obtained using
eqn. (6)
DG = DH − TDS.
Data analysis was done with Kaleidagraph, version 3.5
(Synergy Software).
(6)
4.7 Pyrene excimers, anion competition assay
To solutions of pyrene sulfate 8 (120 lM) and anions 1, 2, 10–
20 (appropriate concentration; c/lM) in buffer (2 ml, 10 mM
NamHnPO4, 107 M NaCl, pH 7.4), pR (2.5 lM) was added.
The solutions were gently mixed and the fluorescence emission
spectra were measured with kex 340 nm at 25 C (12, Fig. 12).
The obtained spectra were normalized as above and analyzed
using the eqn. (7)
◦
25 G. Zuber, E. Dauty, M. Nothisen, P. Belguise and J.-P. Behr, Adv.
EM
EM
IEM = I0 + (I∞ − I0EM)/[1 + (IC50/c)n]
(7)
Drug Delivery Rev., 2001, 52, 245–253.
26 A. D. Miller, Curr. Med. Chem., 2003, 10, 1195–1211.
27 E. Nakamura and H. Isobe, Acc. Chem. Res., 2003, 36, 807–815.
28 W. M. Yau, W. C. Wimley, K. Gawrisch and S. H. White, Biochem-
istry, 1998, 37, 14713–14718.
analogous to eqn. (4) to obtain inhibitory constants IC50 error.
Further analysis using the Cheng–Prusoff equation to estimate
KI values was possible.73,74
29 M. Schiffer, C. H. Chang and F. J. Stevens, Protein Eng., 1992, 5,
213–214.
4.8 Coronene fluorescence
30 R. MacKinnon, S. L. Cohen, A. Kuo, A. Lee and B. T. Chait, Science,
1998, 280, 106–109.
Solutions of coronene anion 3 (0.12–12 lM) in buffer (2 ml,
10 mM NamHnPO4, 100 mM NaCl, pH 7.4) were placed in a cu-
vette. Then, pR was added stepwise and the solutions were gently
mixed. Fluorescence emission spectra of the resulting solutions
were measured with kex 340 nm at 25 ◦C. The obtained spectra
were normalized to maximal monomer emission intensity at
427 nm (Fig. 11) and analyzed as described for pyrenes.
31 W. C. Wimley, Curr. Opin. Struct. Biol., 2003, 13, 404–411.
32 A. N. J. A. Ridder, S. Morein, J. G. Stam, A. Kuhn, B. de Kruijff and
J. A. Killian, Biochemistry, 2000, 39, 6521–6528.
33 M. R. R. De Planque, J. A. W. Kruijtzer, R. M. J. Liskamp, D. Marsh,
D. V. Greathouse, R. E. Koeppe II, B. de Kruijff and J. A. Killian,
J. Biol. Chem., 1999, 274, 20839–20846.
34 M. R. R. De Planque, B. B. Bonev, J. A. A. Demmers, D. V.
Greathouse, R. E. Koeppe II, F. Separovic, A. Watts and J. A. Killian,
Biochemistry, 2003, 42, 5341–5348.
35 J. M. Sanderson and E. J. Whelan, Phys. Chem. Chem. Phys., 2004,
6, 1012–1017.
36 S. Matile, A. Som and N. Sorde´, Tetrahedron, 2004, 60, 6405–6435.
37 G. W. Gokel and A. Mukhopadhyay, Chem. Soc. Rev., 2001, 30,
274–286.
Acknowledgements
We thank D. Jeannerat, A. Pinto and J.-P. Saulnier for NMR
measurements, P. Perrottet and the group of F. Gu¨lac¸ar for
MS, and the Swiss NSF (including National Research Program
“Supramolecular Functional Materials” 4047-057496, SM),
Japan Science and Technology Agency (JST) (SF), Delta
Proteomics (ANL) and the CNRS (AWC) for financial support.
38 S. E. Thompson and D. B. Smithrud, J. Am. Chem. Soc., 2002, 124,
442–449.
39 K. A. Schug and W. Lindner, Chem. Rev., 2005, 105, 67–114.
40 K. Ariga and T. Kunitake, Acc. Chem. Res., 1998, 31, 371–378.
41 M. D. Best, S. L. Tobey and E. V. Anslyn, Coord. Chem. Rev., 2003,
240, 3–15.
42 P. Breccia, M. Van Gool, R. Perez-Fernandez, S. Martin-Santamaria,
F. Gago, P. Prados and J. de Mendoza, J. Am. Chem. Soc., 2003, 125,
8270–8284.
References
1 N. Sakai and S. Matile, J. Am. Chem. Soc., 2003, 125, 14348–14356.
2 N. Sakai, T. Takeuchi, S. Futaki and S. Matile, ChemBioChem, 2005,
6, 114–122.
3 L. Monticelli, K. M. Robertson, J. L. MacCallum and D. P. Tieleman,
FEBS Lett., 2004, 564, 325–332.
4 Y. Jiang, A. Lee, J. Chen, V. Ruta, M. Cadene, B. T. Chait and R.
MacKinnon, Nature, 2003, 423, 33–41.
43 M. Berger and F. P. Schmidtchen, J. Am. Chem. Soc., 1999, 121,
9986–9993.
44 X. Salvatella, M. W. Peczuh, M. Giar´ı, R. K. Jain, J. Sa´nchez-
Quesada, J. de Mendoza, A. D. Hamilton and E. Giralt, Chem.
Commun., 2000, 1399–1400.
5 Y. Jiang, V. Ruta, J. Chen, A. Lee and R. MacKinnon, Nature, 2003,
423, 42–48.
45 B. P. Orner, X. Salvatella, J. Sa´nchez-Quesada, J. de Mendoza, E.
Giralt and A. D. Hamilton, Angew. Chem., Int. Ed., 2002, 41, 117–
119.
46 R. A. Kumpf and D. A. Dougherty, Science, 1993, 261, 1708–1710.
47 J. P. Gallivan and D. A. Dougherty, Proc. Natl. Acad. Sci. U. S. A.,
1999, 96, 9459–9464.
6 R. B. Bass, P. Strop, M. Barclay and D. C. Rees, Science, 2002, 298,
1582–1587.
7 D. Terrone, S. Leung, W. Sang, L. Roudaia and J. R. Silvius,
Biochemistry, 2003, 42, 13787–13799.
8 J. B. Rothbard, T. C. Jessop, R. S. Lewis, B. A. Murray and P. A.
Wender, J. Am. Chem. Soc., 2004, 126, 9506–9507.
9 F. Perret, M. Nishihara, T. Takeuchi, S. Futaki, A. N. Lazar, A. W.
Coleman, N. Sakai and S. Matile, J. Am. Chem. Soc., 2005, 127,
1114–1115.
10 N. Sakai, N. Sorde´, G. Das, P. Perrottet, D. Gerard and S. Matile,
Org. Biomol. Chem., 2003, 1, 1226–1231.
48 F. M. Menger, Biochemistry, 1992, 31, 5368–5373.
49 O. T. Jones and A. G. Lee, Biochemistry, 1985, 24, 2195–2202.
50 J. Martins and E. Melo, Biophys. J., 2001, 80, 832–840.
51 C. Bingel, Chem. Ber., 1993, 126, 1957–1959.
52 I. Lamparth and A. Hirsh, J. Chem. Soc., Chem. Commun., 1994,
1727–1728.
53 S. Foley, C. Crowley, M. Smaihi, C. Bonfils, B. F. Erlanger, P. Seta
and C. Larroque, Biochem. Biophys. Res. Commun., 2002, 294, 116–
119.
11 S. Futaki, Arginine-rich Peptides, Curr. Protein Pept. Sci., 2003, 4,
87–157.
1 6 6 8
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9

View Article Online
54 C. Wang, L. A. Tai, D. D. Lee, P. P. Kanakamma, C. K.-F. Shen,
T.-Y. Luh, C. H. Cheng and K. C. Hwang, J. Med. Chem., 1999, 42,
4614–4620.
64 B. Hoff, E. Strandberg, A. S. Ulrich, D. P. Tieleman and C. Posten,
Biophys. J., 2005, 88, 1818–1827.
65 B. De Kruijff and R. A. Demil, Biochim. Biophys. Acta, 1974, 339,
55 H. Park, M. Chhowalla, Z. Iqbal and F. Sesti, J. Biol. Chem., 2003,
57–70.
278, 50212–50216.
66 S. C. Hartsel and J. Bolard, Trends Pharm. Sci., 1996, 17, 445–449.
67 J. Turk, D. J. Waters and P. S. Low, Cancer Lett., 2004, 213,
165–172.
56 F. Rancan, S. Rosan, F. Boehm, A. Cantrell, M. Brellreich, H.
Schoenberger, A. Hirsch and F. Moussa, J. Photochem. Photobiol.
B, 2002, 67, 157–162.
68 X. B. Zhao and R. J. Lee, Adv. Drug. Delivery Rev., 2004, 56, 1193–
57 E. Da Silva, A. N. Lazar and A. W. Coleman, J. Drug Delivery Sci.
Technol., 2004, 14, 3–20.
58 Davenport, B. Shen, T. W. Joseph and M. P. Straher, Chem. Phys.
Lipids, 2001, 109, 145–156.
59 C. T. Miller, R. Weragoda, E. Izbicka and B. I. Iverson, Bioorg. Med.
Chem., 2001, 9, 2015–2024.
1204.
69 B. Stevens, Spectrochim. Acta, 1962, 18, 439–448.
70 D.Y. Sasaki, Cell Biochem. Biophys., 2003, 39, 145–161.
71 B. C. Roy, B. Chandra, D. Hromas and S. Mallik, Org. Lett., 2003,
5, 11–14.
72 B.-B. Wang, X. Zhang, X.-R. Jia, Z.-C. Li, Y. Ji, L. Yang and Y. Wei,
J. Am. Chem. Soc., 2004, 126, 15180–15194.
73 Y. Cheng and W. H. Prusoff, Biochem. Pharmacol., 1973, 22, 3099–
3108.
60 P. Talukdar, G. Bollot, J. Mareda, N. Sakai and S. Matile, J. Am.
Chem. Soc., in press.
61 S. Rex, Biophys. Chem., 1996, 58, 75–85.
62 P. Yeagle, The Structure of Biological Membranes, CRC Press, Boca
74 H. C. Cheng, Pharmacol. Res., 2004, 50, 21–40.
75 R. B. M. Ansem and L. T. Scott, J. Phys. Org. Chem., 2004, 17,
819–823.
Raton, Florida, 1992.
63 T. Baumgart, S. T. Hess and W. W. Webb, Nature, 2003, 425, 821–824.
O r g . B i o m o l . C h e m . , 2 0 0 5 , 3 , 1 6 5 9 – 1 6 6 9
1 6 6 9