Redox-triggered contents release from liposomes

Article abstract of DOI:10.1021/ja8050469

An exciting new direction in responsive liposome research is endogenous triggering of liposomal payload release by overexpressed enzyme activity in affected tissues and offers the unique possibility of active and site-specific release. Bringing to fruitio

Full text of DOI:10.1021/ja8050469

Published on Web 10/08/2008
Redox-Triggered Contents Release from Liposomes
Winston Ong,† Yuming Yang,† Angela C. Cruciano,‡ and Robin L. McCarley*,†
Department of Chemistry and Center for Biomodular Multiscale Systems, Louisiana State
UniVersity, Baton Rouge, Louisiana 70803, and National Center for Macromolecular Imaging,
Department of Biochemistry and Molecular Biology, Baylor College of Medicine,
Houston, Texas 77030
Received July 1, 2008; E-mail: tunnel@lsu.edu
Abstract: An exciting new direction in responsive liposome research is endogenous triggering of liposomal
payload release by overexpressed enzyme activity in affected tissues and offers the unique possibility of
active and site-specific release. Bringing to fruition the fully expected capabilities of this new class of triggered
liposomal delivery system requires a collection of liposome systems that respond to different upregulated
enzymes; however, a relatively small number currently exist. Here we show that stable, ∼100 nm diameter
liposomes can be made from previously unreported quinone-dioleoyl phosphatidylethanolamine (Q-DOPE)
lipids, and complete payload release (quenched fluorescent dye) from Q-DOPE liposomes occurs upon
their redox activation when the quinone headgroup possesses specific substituents. The key component
of the triggerable, contents-releasing Q-DOPE liposomes is a “trimethyl-locked” quinone redox switch
attached to the N-terminus of DOPE lipids that undergoes a cleavage event upon two-electron reduction.
Payload release by aggregation and leakage of “uncapped” Q-DOPE liposomes is supported by results
from liposomes wherein deliberate alteration of the “trimethyl-locked” switch completely deactivates the
redox-destructible phenomena (liposome opening). We expect that Q-DOPE liposomes and their variants
will be important in treatment of diseases with associated tissues that overexpress quinone reductases,
such as cancers and inflammatory diseases, because the quinone redox switch is a known substrate for
this group of reductases.
Introduction
the unprecedented fabrication of complicated nanoscopic and
mesoscopic structures with unique physical and chemical
Stimuli-responsive macromolecules and self-assembled ar-
chitectures have piqued the interest of the basic science and
biomedical arenas due to the possibility of attaining nanometer-
level control over the construction and destruction of covalent
and noncovalent interactions.1-15 Systems that allow for
manipulation of these molecular-scale interactions may lead to
properties and unreported biological function. Precise control
of stimuli-responsive systems that unravel or self-destruct upon
application of a trigger (stimulus) has captured the interest of
the biomedical community because of the temporal and spatial
specificity it would afford the delivery of therapeutics.16,17
Recently, such self-immolative modalities have been engineered
into polymeric and dendritic constructs.18-21
Stimuli-responsive liposomes are highly attractive because
local control over payload release should be possible through
the use of either an exogenous or endogenous stimulus.22 The
lipids in these liposomes generally contain a triggerable subunit
that is responsible for gating the stability and/or permeability
of the lipid bilayer. These liposomes are sometimes referred to
† Louisiana State University.
‡ Baylor College of Medicine.
(1) Savariar, E. N.; Ghosh, S.; Gonzalez, D. C.; Thayumanavan, S. J. Am.
Chem. Soc. 2008, 130, 5416–5417.
(2) Liu, J. W.; Lu, Y. J. Am. Chem. Soc. 2007, 129, 8634–8643.
(3) Nijhuis, C. A.; Ravoo, B. J.; Huskens, J.; Reinhoudt, D. N. Coord.
Chem. ReV. 2007, 251, 1761–1780.
(4) Bellomo, E. G.; Wyrsta, M. D.; Pakstis, L.; Pochan, D. J.; Deming,
T. J. Nat. Mater. 2004, 3, 244–248.
(5) Dillenback, L. M.; Goodrich, G. P.; Keating, C. D. Nano Lett. 2006,
6, 16–23.
(14) Rui, Y. J.; Wang, S.; Low, P. S.; Thompson, D. H. J. Am. Chem. Soc.
1998, 120, 11213–11218.
(6) Wang, W.; Kaifer, A. E. Angew. Chem., Int. Ed. 2006, 45, 7042–
7046.
(15) Van Horn, B. A.; Wooley, K. L. Macromolecules 2007, 40, 1480–
1488.
(7) Serizawa, T.; Yamaguchi, M.; Akashi, M. Angew. Chem., Int. Ed. 2003,
42, 1115–1118.
(16) Allen, T. M.; Cullis, P. R. Science 2004, 303, 1818–1822.
(17) Dreher, M. R.; Chilkoti, A. J. Natl. Cancer Inst. 2007, 99, 983–985.
(18) Szalai, M. L.; Kevwitch, R. M.; McGrath, D. V. J. Am. Chem. Soc.
2003, 125, 15688–15689.
(8) Whitaker-Brothers, K.; Uhrich, K. J. Biomed. Mater. Res., Part A 2004,
70A, 309–318.
(9) Gugliotti, L. A.; Feldheim, D. L.; Eaton, B. E. J. Am. Chem. Soc.
2005, 127, 17814–17818.
(19) de Groot, F. M. H.; Albrecht, C.; Koekkoek, R.; Beusker, P. H.;
Scheeren, H. W. Angew. Chem., Int. Ed. 2003, 42, 4490–4494.
(20) Amir, R. J.; Shabat, D. In Polymer Therapeutics I: Polymers as Drugs,
Conjugates and Gene DeliVery Systems; Springer-Verlag Berlin:
Berlin, 2006; Vol. 192, pp 59-94.
(10) Eastoe, J.; Vesperinas, A. Soft Matter 2005, 1, 338–347.
(11) Goodwin, A. P.; Mynar, J. L.; Ma, Y. Z.; Fleming, G. R.; Frechet,
J. M. J. J. Am. Chem. Soc. 2005, 127, 9952–9953.
(12) Jiang, J. Q.; Tong, X.; Zhao, Y. J. Am. Chem. Soc. 2005, 127, 8290–
8291.
(21) Sagi, A.; Weinstain, R.; Karton, N.; Shabat, D. J. Am. Chem. Soc.
2008, 130, 5434–ff.
(13) Menger, F. M.; Gabrielson, K. J. Am. Chem. Soc. 1994, 116, 1567–
1568.
(22) Guo, X.; Szoka, F. C. Acc. Chem. Res. 2003, 36, 335–341.
9
10.1021/ja8050469 CCC: $40.75
2008 American Chemical Society
J. AM. CHEM. SOC. 2008, 130, 14739–14744 14739

A R T I C L E S
Ong et al.
as “smart” delivery systems, because unloading of the encap-
sulated payload requires stimuli activation.23-25 Ideally, a
stimulus triggers the onset of cargo unloading, thereby allowing
the carrier-cargo ensemble to be constructed without prema-
turely sacrificing or exposing the encapsulated cargo to the
external environment. Various physiological environmentsssuch
as low endosomal pH26,27 and elevated enzymatic activity28-31
or biomarker presence32-34sand external sources, including
radiation35-38 and hyperthermia,39,40 supply the necessary
stimuli that induce the unloading of the liposomal cargo; this
occurs either by perturbing the permeability or by completely
disrupting the noncovalent stability of the bilayer assembly. For
instance, it has been shown that the permeability of pH- and
radiation-sensitive liposomes can be perturbed by the acid-
triggered depegylation of PEG-conjugated lipids41 and photo-
chemical “uncorking” of o-benzyl-protected lipids,35 respec-
tively, leading to the release of the encapsulated dyes.
One of our long-term goals associated with the construction
of redox-sensitive liposomes is responsive liposomal carriers
that can deliver anticancer agents to tumor tissues. It is our
hypothesis that carefully designed redox-sensitive liposomes are
structurally optimized to preferentially accumulate (preconcen-
tration via enhanced permeability and retention (EPR) effect)42
and specifically respond to the high quinone reductase activi-
ties43 (localized release facilitated by complementary redox
potentials) in cancer tissues.
Figure 1. Structures of DOPE-quinones 1-Q1 and 1-Q3, the corresponding
model compounds 2-Q1 and 2-Q3, and lactones 3-HQ3 and 3-HQ1 (ref
48).
and electrochemically optimized for the delivery and triggered
release of anticancer drugs to cancer tissues, we herein report
on liposomes comprised of trimethyl-locked quinone lipids that
require a two-electron reductive activation to liberate the
liposomal payload. The integration of a trimethyl-lock quinone
switch within the liposome is critical, because such a quinone
switch has measurable activities toward several quinone
reductases44,45 that are upregulated in numerous types of cancer
tissues.43 In addition, due to the fact that liposome-encapsulated
drugsretainthepharmacokineticpropertiesofthecarrierssmeaning
that the drugs are not pharmacologically active until released
from the liposomessit is widely perceived that triggered release
of the active ingredients is necessary for the rapid delivery of
anticancer drugs.46,47 Thus, the development of new methods
for the specific, stimuli-triggered release of liposomal payloads
is extremely important. Herein, we demonstrate that liposomes
comprised of dioleoyl phosphatidylethanolamine (DOPE) lipids
having a trimethyl-locked quinone (Q3) headgroup (1-Q3 in
Figure 1) liberate their contents (encapsulated dyes) upon
reduction of Q3.
In order to gain an understanding of how to create a new
class of redox-responsive liposomal carriers that are structurally
(23) Sawant, R. M.; Hurley, J. P.; Salmaso, S.; Kale, A.; Tolcheva, E.;
Levchenko, T. S.; Torchilin, V. P. Bioconjugate Chem. 2006, 17, 943–
949.
(24) Andresen, T. L.; Jensen, S. S.; Kaasgaard, T.; Jorgensen, K. Curr.
Drug DeliVery 2005, 2, 353–362.
(25) Fattal, E.; Couvreur, P.; Dubernet, C. AdV. Drug DeliVery ReV. 2004,
56, 931–946.
(26) Gerasimov, O. V.; Boomer, J. A.; Qualls, M. M.; Thompson, D. H.
AdV. Drug DeliVery ReV. 1999, 38, 317–338.
(27) Zhang, J. X.; Zalipsky, S.; Mullah, N.; Pechar, M.; Allen, T. M.
Pharmacol. Res. 2004, 49, 185–198.
(28) Davidsen, J.; Jorgensen, K.; Andresen, T. L.; Mouritsen, O. G. Biochim.
Biophys. Acta 2003, 1609, 95–101.
(29) Meers, P. AdV. Drug DeliVery ReV. 2001, 53, 265–272.
(30) Huang, Z.; Szoka, F. C. In Liposome Technology, 3rd ed.; Gregoriadis,
G., Ed.; CRC Press: Boca Raton, FL, 2006; Vol. 3, pp 165-196.
(31) Sarkar, N.; Banerjee, J.; Hanson, A. J.; Elegbede, A. I.; Rosendahl,
T.; Krueger, A. B.; Banerjee, A. L.; Tobwala, S.; Wang, R. Y.; Lu,
X. N.; Mallik, S.; Srivastava, D. K. Bioconjugate Chem. 2008, 19,
57–64.
Results and Discussion
Quinone phospholipids 1-Q1 and 1-Q3 and the model
derivatives 2-Q1 and 2-Q3 in Figure 1 were successfully
synthesized by condensing the appropriate amines with the
N-hydroxysuccinimide esters of the quinone acids, NHS-Qx, x
) 1 or 3.48 The precursors of NHS-Q1 were synthesized using
strategies that paralleled the preparation of NHS-Q3.49 All
(32) Kirpotin, D.; Hong, K. L.; Mullah, N.; Papahadjopoulos, D.; Zalipsky,
S. FEBS Lett. 1996, 388, 115–118.
(33) Tang, F. X.; Hughes, J. A. Biochem. Biophys. Res. Commun. 1998,
242, 141–145.
(34) Zalipsky, S.; Qazen, M.; Walker, J. A.; Mullah, N.; Quinn, Y. P.;
Huang, S. K. Bioconjugate Chem. 1999, 10, 703–707.
(35) Chandra, B.; Mallik, S.; Srivastava, D. K. Chem. Commun. 2005,
3021–3023.
1
compounds yielded the expected outcomes from H and 13C
NMR spectroscopies and ESI-TOF or GC mass spectrometries.
(36) Eastoe, J.; Vesperinas, A.; Donnewirth, A. C.; Wyatt, P.; Grillo, I.;
Heenan, R. K.; Davis, S. Langmuir 2006, 22, 851–853.
(37) Shum, P.; Kim, J. M.; Thompson, D. H. AdV. Drug DeliVery ReV.
2001, 53, 273–284.
(44) Jaffar, M.; Abou-Zeid, N.; Bai, L.; Mrema, I.; Robinson, I.; Tanner,
R.; Stafford, I. J. Curr. Drug DeliVery 2004, 1, 345–350.
(45) Weerapreeyakul, N.; Visser, P.; Brummelhuis, M.; Gharat, L.;
Chikhale, P. J. Med. Chem. Res. 2000, 10, 149–163.
(46) Lim, H. J.; Masin, D.; Madden, T. D.; Bally, M. B. J. Pharmacol.
Exp. Ther. 1997, 281, 566–573.
(38) Zhang, Z. Y.; Smith, B. D. Bioconjugate Chem. 1999, 10, 1150–1152.
(39) Kono, K.; Murakami, T.; Yoshida, T.; Haba, Y.; Kanaoka, S.;
Takagishi, T.; Aoshima, S. Bioconjugate Chem. 2005, 16, 1367–1374.
(40) Needham, D.; Dewhirst, M. W. AdV. Drug DeliVery ReV. 2001, 53,
285–305.
(47) Ponce, A. M.; Wright, A.; Dewhirst, M. W.; Needham, D. Future
Lipidol. 2006, 1, 25–34.
(41) Boomer, J. A.; Inerowicz, H. D.; Zhang, Z. Y.; Bergstrand, N.;
Edwards, K.; Kim, J. M.; Thompson, D. H. Langmuir 2003, 19, 6408–
6415.
(48) The following nomenclature will be adapted throughout the article:
“Q” and “HQ” refer to the oxidized and reduced forms of the quinone,
respectively, regardless of whether “HQ” is generated in situ (e.g.,
1-HQ3) or synthesized (e.g., 3-HQ1). For instance, 1-HQ1 is the
reduced form of 1-Q1.
(42) Maeda, H.; Wu, J.; Sawa, T.; Matsumura, Y.; Hori, K. J. Controlled
Release 2000, 65, 271–284.
(43) Fitzsimmons, S. A.; Workman, P.; Grever, M.; Paull, K.; Camalier,
R.; Lewis, A. D. J. Natl. Cancer Inst. 1996, 88, 259–269.
(49) Carpino, L. A.; Triolo, S. A.; Berglund, R. A. J. Org. Chem. 1989,
54, 3303–3310.
9
14740 J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008

Redox-Responsive Liposomes
A R T I C L E S
Figure 2. Characterization of calcein-free liposomes: (A) cryo-TEM image
of 1-Q1 and (B) size distribution by DLS intensity of 1-Q3 at 25 °C.
Liposomes of 1-Q1 and 1-Q3 were prepared in pH 7.1/0.1 M
phosphate buffer/0.1 M KCl using the technique of hydrated
lipid film extrusion.50
Cryogenic transmission electron microscopy (cryo-TEM) and
dynamic light scattering (DLS) experiments reveal that 1-Q1
and 1-Q3 liposomes possess diameters in the 80-200 nm range
(representative data in Figure 2). When the liposomes were
preloaded with calcein dyes (5 × 10-2 M), the resulting DLS
intensity distributions were indistinguishable from those of the
calcein-free liposomes, a result which indicates that the presence
of dyes does not lead to significant changes in aggregate
morphology. The unchanged, quenched emission from quiescent,
aerobic solutions of calcein-loaded liposomes over a 2 day
period at room temperature reflects the stability of the liposomes
under ambient conditions; in addition, DLS-determined sizes
of calcein-free and loaded liposomes did not vary over a period
of a week. As expected, addition of excess Triton X-100
detergent to solutions of calcein-loaded liposomes caused release
of the encapsulated dye, as evidenced by the rapid increase in
emission intensity resulting from the dilution of calcein (Figure
3, trace B for 1-Q3 and trace D for 1-Q1).
Figure 3. Time-course dependence of the emission intensities of calcein-
loaded liposomes at 515 nm (λex ) 495 nm) under ambient conditions.
Traces A and B are for liposomes of 1-Q3, and traces C and D are for
liposomes of 1-Q1, respectively. Addition of 100 equiv of Na2S2O4 for
traces A and C occurred at 125 min.
Scheme 1. Reductive Lactonization of 1-Q3 Produces 3-HQ3 and
DOPE
We51,52 and others53,54 previously reported that the two-
electron reduction of various quinone trimethyl-lock (Q3)
systems by chemical agents (Na2S2O4 or NaBH4) or electrolysis
produces the corresponding hydroquinone, HQ3, possessing a
sterically congested configuration, and HQ3 spontaneously
dissociates from the parent structure as the lactone 3-HQ3. On
the basis of time-course data that exhibit an immediate increase
in fluorescence intensity of calcein following the addition of
the reducing agent Na2S2O4 to the calcein-loaded 1-Q3 lipo-
somes (Figure 3, Curve A), chemical reduction of liposomal
1-Q3 to 1-HQ3 resulted in the liberation of encapsulated calcein
from the carrier! Addition of simple salts, such as Na2SO4, at
equivalent concentrations does not lead to contents release. This
outcome is consistent with the dissociation of in situ generated
1-HQ3-containing liposomes into lactone 3-HQ3 and DOPE
inverted micelles (Scheme 1).
Unmodified DOPE is known to exist as hexagonal columnar
(HII) inverted micelles at pH 7 due to the head/tail volume ratio
of this lipid.55 On the other hand, N-acylated DOPE lipids, such
as 1-Q3 and 1-Q1, have larger head/tail volume ratios than
DOPE and can be made to readily form liposomes at pH 7.
Therefore, as depicted in Scheme 1, the structural transition of
lamellar (LR)/liposomal 1-Q3 to inverted micellar (HII) DOPE
(i.e., 1-Q3 (LR) f DOPE (HII) + 3-HQ3) was accompanied
by the release of calcein dyes. It is reasonable to assume that
the HII inverted micelles are incapable of encapsulating calcein
as efficiently as the lamellar liposomes. Thus, the inevitable
expulsion of the entrapped dyes during the LR f HII structural
transition is expected.
The results described herein are the first demonstration of
cargo unloading from a non-disulfide-based liposome that is
triggered by electron transfer. These outcomes will add to the
repertoire of methods for masking non-liposome-forming phos-
pholipids, including DOPE, with a liposome-forming, stimuli-
sensitive N-substituent to induce liposome formation, such as
those based on light,38 thiolysis,34 and pH41 stimuli. Further-
more, the potential for this system to be responsive to reductase
enzymes is high.43-45
(50) Szoka, F.; Papahadjopoulos, D. Annu. ReV. Biophys. Bioeng. 1980, 9,
467–508.
To assess the role of the LR f HII structural transition in the
liberation of the dyes, liposomal 1-Q1 was employed as a control
lipid because it lacks the geminal methyl (-CH3) groups that
are required to satisfy the trimethyl-lock configuration.56 Such
a geometric requirement is a prerequisite for fast lactonization
(51) Ong, W.; McCarley, R. L. Chem. Commun. 2005, 4699–4701.
(52) Ong, W.; McCarley, R. L. Macromolecules 2006, 39, 7295–7301.
(53) Yan, C.; Matsuda, W.; Pepperberg, D. R.; Zimmerman, S. C.;
Leckband, D. E. J. Colloid Interface Sci. 2006, 296, 165–177.
(54) Zheng, A.; Shan, D.; Wang, B. J. Org. Chem. 1999, 64, 156–161.
(55) Siegel, D. P.; Epand, R. M. Biophys. J. 1997, 73, 3089–3111.
9
J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008 14741

A R T I C L E S
Ong et al.
Scheme 2. Proposed Mechanism for the Conversion of Liposomal
1-Q3 from Lamellar (LR) Liposomes to Inverted Hexagonal (HII)
DOPE Micellesa
(t1/2 e few hours), and as a result, 1-HQ1 should not lactonize
within the time scale of the dye-liberation experiments (vide
infra). Thus, when the calcein-loaded liposomes of control lipid
1-Q1 were reduced, the dyes remained encapsulated inside the
liposomes (fluctuation of trace C in Figure 3 at 120 min), and
addition of Triton X-100 was required to release the dyes (trace
C at 1150 min). This observation is in sharp contrast with
liposomal 1-Q3, where the dyes were immediately liberated
following the addition of Na2S2O4 (compare traces A and C at
120 min in Figure 3). The inability of Na2S2O4 to liberate the
dyes from liposomal 1-Q1 remarkably suggests that the lamellar
structure was maintained after reduction and that the perme-
ability of the lipid barrier was unaffected by the redox
conversion of 1-Q1 to 1-HQ1. These phenomena would be
perhaps somewhat unexpected, considering that previous ac-
counts have demonstrated the dramatic effect of redox conver-
sion on the stability of some vesicles and micelles,57-60 but
the collapse mechanism of reduced 1-Q3 liposomes is different
than that of these redox-vesicle systems, vide infra. The ability
of liposomal 1-HQ1 to retain calcein further suggests that the
release of calcein from liposomal 1-HQ3 was not merely caused
by the redox conversion of the Q3 headgroup to HQ3 but was,
rather, imparted by the covalent disconnection of HQ3 from
DOPE. The observation that the emission intensities of the
liberated dyes remained similar, regardless of the lysing agent
used (compare traces A with B and C with D61 in Figure 3),
indicates that the release of dyes was effectively quantitative
in all cases.62
a Step A: Na2S2O4 reduces the 1-Q3 lipids that are located in the outer
layer of the liposome to 1-HQ3. Step B: 1-HQ3 dissociates into lactone
3-HQ3 and DOPE, as described in Scheme 1. Step C: release of
encapsulated calcein dye occurs as a result of liposome collapse caused by
DOPE layer-DOPE layer contact.
Q3) ) +0.12 V). Therefore, assuming that the ∆Ep value of
the model compounds is similar to the ∆Ep gap between lipids
1-Q1 and 1-Q363 (i.e., Ered (1-Q3) - Ered (2-Q3) ∼ Ered (1-
Q1) - Ered (2-Q1)), it is reasonable to conclude that Na2S2O4
reduces lipid 1-Q1 to the same extent as lipid 1-Q3. If anything,
the aforementioned potential gap suggests that reduction of 1-Q1
is thermodynamically easier than 1-Q3.
The dormancy of 1-HQ1 toward lactonization was verified
by employing compounds 2-Q1 and 2-Q3 in aqueous milieu to
model the relative reactivity, or lack thereof, of 1-HQ1 and
1-HQ3, respectively. Kinetic data from 1H NMR spectroscopy
experiments revealed that 2-HQ3, generated in situ by the
reduction of 2-Q3 using Na2S2O4, dissociated into lactone
3-HQ3 and 2-[2-(2-methoxy-ethoxy)-ethoxy]-ethylamine fol-
lowing first-order kinetics (k ∼ 0.02 min-1, t1/2 ∼ 30 min;
Supporting Information Figure S1). On the contrary, reduced
2-HQ1, also generated in situ by the reduction of 2-Q1,
remained stable over a 48 h period. In other words, within the
In Scheme 2 is outlined a proposed mechanism describing
the sequence of lipid reduction-headgroup cyclization/elimina-
tion and liposome destabilization events leading to the release
of calcein dyes from liposomal 1-Q3. The addition of Na2S2O4
reduces the 1-Q3 lipids that are located in the outer layer of
the liposome to 1-HQ3 (step A).64 On the basis of the results
gathered from control experiments that demonstrate the unper-
turbed fluorescence intensity of calcein following the reduction
of 1-Q1 to 1-HQ1 (trace C in Figure 3 at 120 min), it is evident
that the calcein dyes remained encapsulated even though the
outer lipid layer of the liposomes was already reduced to 1-HQ3.
Following reduction, and as described in Scheme 1, 1-HQ3 is
removed (cleaved) from the external lipids of the liposome to
yield lactone 3-HQ3 and DOPE (step B). As the 1-HQ3 f
DOPE + 3-HQ3 dissociation processscaused by intramolecular
cyclizationscontinues, the liposomal surface concentration of
DOPE eventually becomes high enough (HII-phase competent)65
to allow for DOPE layer-DOPE layer contact between indi-
vidual66 liposomes (aggregation);65,67 light scattering data
support aggregation of reduced 1-HQ3 liposomes (not shown),
as the average particle size increases almost 5-fold (∼500 nm)
1
detection limits of H NMR spectroscopy, release of lactone
3-HQ1 from 2-HQ1 was not observed. Therefore, it is indeed
reasonable to assume that liposomal 1-HQ1 did not lactonize
within the 1500-min window of the dye-liberation experiments
in Figure 3.
We conducted voltammetric experiments on the model
compounds 2-Q1 and 2-Q3 to provide support for the argument
that Na2S2O4 is capable of reducing lipid 1-Q1 as effectively
as 1-Q3. Voltammetric data (Supporting Information Figure S2)
reveal that two-electron reduction of 2-Q1 is more positive than
that of 2-Q3 by +0.12 V (i.e., ∆Ep ) Ered (2-Q1) - Ered (2-
(56) Milstien, S.; Cohen, L. A. J. Am. Chem. Soc. 1972, 94, 9158–9165.
(57) Abbott, N. L.; Jewell, C. M.; Hays, M. E.; Kondo, Y.; Lynn, D. M.
J. Am. Chem. Soc. 2005, 127, 11576–11577.
(58) Kakizawa, Y.; Sakai, H.; Yamaguchi, A.; Kondo, Y.; Yoshino, N.;
Abe, M. Langmuir 2001, 17, 8044–8048.
(59) Munoz, S.; Gokel, G. W. J. Am. Chem. Soc. 1993, 115, 4899–4900.
(60) Saji, T.; Hoshino, K.; Aoyagui, S. J. Am. Chem. Soc. 1985, 107, 6865–
6868.
(63) It is assumed here that the quinone units located in the outer layer of
the liposomes are kinetically and equally accessible, such that
probability of reduction of the liposomal lipids (1-Q1 and 1-Q3) can
be approximated by the thermodynamic reduction potentials of the
model quinones (2-Q1 and 2-Q3).
(64) Consistent with previous reports, it is assumed here that Na2S2O4 is
non-membrane-permeable and is therefore incapable of reducing the
lipids located in the interior layer of the liposome. For instance, see:
Angeletti, C.; Nichols, J. W. Biochemistry 1998, 37, 15114–15119.
(61) Calcein is known to be quenched by Triton, and this quenching is
Triton concentration dependent. See: Memoli, A.; Palermiti, L. G.;
Travagli, V.; Alhaique, F. J. Pharm. Biomed. Anal. 1999, 19, 627–
632, and references therein for more information on this topic.
(62) Percent release of calcein was estimated using the well-known
equation: % release ) [(I- I0)/(Imax- I0)] × 100, where I and Imax
are the emission intensities after the addition of Na2S2O4 and Triton
X-100, respectively, and I0 is the initial emission intensity.
9
14742 J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008

Redox-Responsive Liposomes
A R T I C L E S
Scheme 3. Synthesis of DOPE Quinones 1-Q1 and 1-Q3, and the Corresponding Model Compounds 2-Q1 and 2-Q3
(ESI) m/z (M + H)+, calcd ) 207.1021, obsd ) 207.1011, 4.8
ppm error. 4-Q1: reaction time was shortened to 5 min and
temperature was 0 °C. The crude product was recrystallized in
CH2Cl2/Et2O/n-hexanes to produce the product as dark-yellow
crystals. Yield ) 88%; 1H NMR (CDCl3, 300 MHz) δ 2.82 (t,
2H), 2.51 (t, 2H), 2.07 (s, 3H), 2.02 (s, 6H); 13C NMR (CDCl3, 75
MHz) δ 187.59, 187.00, 178.81, 141.93, 141.67, 140.87, 140.71,
32.77, 22.21, 12.54, 12.43, 12.37; HRMS (ESI) m/z (M + H)+,
calcd ) 223.0970, obsd ) 223.0955, 6.7 ppm error. 5-Q1: Yield
after addition of the reducing agent. Liposome collapse ensues
to yield DOPE micelles (HII) and released calcein dye; during
the collapse process, the remaining 1-Q3 lipids are quantitatively
reduced by dithionite via steps A and B to afford the lactone
3-HQ3 and micellar DOPE.
Conclusions
We have shown that lysis of liposomal 1-Q3 occurs upon
reduction of the Q3 head groups by Na2S2O4 to yield an
intermediate which rapidly forms 3-HQ3 and DOPE, the latter
being unable to support stable liposomes at pH 7, thereby
allowing release of liposome contents. The system presented
here has great potential in drug delivery systems that employ
bioreductive activation of liposomal anticancer drugs, as most
cancer tissues contain overexpressed quinone reductases that
are known to activate the Q3 headgroup.43,68 In addition, these
redox-triggered lipids and their variants may permit the study
of lipid dynamics in bilayers, either in free solution (liposomes)
or on surfaces (supported bilayers). Efforts to those ends are
currently underway.
1
) 84%; H NMR (CDCl3, 300 MHz) δ 2.89 (m, 2H), 2.83 (m,
6H), 2.07 (s, 3H), 2.02 (s, 6H); 13C NMR (CDCl3, 75 MHz) δ
187.38, 186.86, 169.17, 167.87, 142.39, 140.98, 140.66, 29.74,
25.69, 22.18, 12.54, 12.51, 12.40; HRMS (ESI) m/z (M + H)+,
calcd ) 320.1134, obsd ) 320.1145, 3.4 ppm error.
2-Q1 and 2-Q3. 2-[2-(2-Methoxy-ethoxy)-ethoxy]-ethylamine
was synthesized as reported.69 To a cooled solution (0 °C) of this
amine (1.31 mmol) in dry CH2Cl2 (20 mL) was added triethylamine
(3.9 mmol), then 5Qx (x ) 1 or 3, 1.31 mmol), and the reaction
was allowed to reach to completion (1 h) while being maintained
under Ar atmosphere. The crude reaction mixture was concentrated,
loaded onto a SiO2 column, and eluted using 7:1 ethyl acetate/
methanol. Rf values of 2-Q1 and 2-Q3 in SiO2 plate are ca.
0.75-0.80. The combined fractions were evaporated to dryness and
dried under vacuum to yield the products as yellow, viscous oils.
2-Q1: Yield ) 98%; 1H NMR (CDCl3, 300 MHz) δ 6.24 (bs, 1H),
3.65 (m, 6H), 3.56 (m, 4H), 3.43 (t, 2H), 3.38 (s, 3H), 2.81 (t,
2H), 2.31 (t, 2H), 2.07 (s, 3H), 2.01 (s, 6H); 13C NMR (CDCl3, 75
MHz) δ 187.65, 187.21, 171.76, 142.74, 141.41, 140.75, 140.53,
71.99, 70.58, 70.53, 70.26, 69.91, 59.09, 39.36, 35.08, 23.17, 12.52,
12.41, 12.31; HRMS (ESI) m/z ((M + H)+), calcd ) 368.2073,
Experimental Section
Synthesis. Compounds used here were made according to
Scheme 3. 3-HQ3, 4-Q3, and 5-Q3 were synthesized as reported.49
Their Q1 analogues (3-HQ1, 4-Q1, and 5-Q1) were prepared using
identical procedures, except as noted. 3-HQ1: reaction time was
extended to 14 h. For purification, the concentrated solution of the
crude product in ethyl acetate was diluted with n-hexanes, then
refrigerated overnight to produce the purified product as a brown
precipitate. Yield ) 46%; 1H NMR (CDCl3, 300 MHz) δ 4.53 (bs,
1H), 2.92 (t, 2H), 2.72 (t, 2H), 2.22 (s, 3H), 2.19 (s, 3H), 2.18 (s,
3H); 13C NMR (CDCl3, 75 MHz) δ 169.58, 148.43, 144.45, 123.12,
121.99, 119.47, 118.24, 29.27, 21.51, 12.34, 12.21, 11.90; HRMS
1
obsd ) 368.2076, 0.8 ppm error. 2-Q3: Yield ) 98%; H NMR
(CDCl3, 300 MHz) δ 6.07 (bs, 1H), 3.66-3.57 (m, 8H), 3.49 (t,
2H), 3.40 (s, 3H), 3.35 (m, 2H), 2.82 (s, 2H), 2.12 (s, 3H), 1.98 (s,
3H), 1.95 (s, 3H), 1.42 (s, 6H); 13C NMR (CDCl3, 75 MHz) δ
191.35, 187.74, 172.06, 153.93, 143.92, 137.84, 137.40, 72.09,
70.61, 70.33, 70.04, 59.18, 49.16, 39.10, 38.37, 28.89, 14.23, 12.91,
12.26; HRMS (ESI) m/z ((M + H)+), calcd ) 396.2386, obsd )
396.2388, 0.5 ppm error.
(65) Bentz, J.; Ellens, H.; Szoka, F. C. Biochemistry 1987, 26, 2105–2116.
(66) Intraliposomal bilayer-bilayer contact in oligolamellar Q-DOPE
liposomes may also contribute to liposome collapse and contents
release. The mechanism of Q-DOPE liposome collapse is currently
being investigated in detail.
1-Q1 and 1-Q3. The reaction conditions and reagent ratios were
identical to the preparation of 2-Qx (x ) 1 or 3), except that the
(67) Ellens, H.; Bentz, J.; Szoka, F. C. Biochemistry 1984, 23, 1532–1538.
(69) Bahr, J. L.; Yang, J. P.; Kosynkin, D. V.; Bronikowski, M. J.; Smalley,
(68) Ross, D. Drug Metab. ReV. 2004, 36, 639–654.
R. E.; Tour, J. M. J. Am. Chem. Soc. 2001, 123, 6536–6542.
9
J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008 14743

A R T I C L E S
Ong et al.
reaction times were extended to 4-6 h. For purification, the reaction
mixture was diluted with CH2Cl2 to 50 mL, then extracted with
5% NaHCO3 (1 × 50 mL). The organic layer was dried with
Na2SO4, concentrated, and loaded to a SiO2 column, then eluted
via a gradient of 1:1 CH2Cl2/EtOAc (to elute any unreacted 2-Qx,
if any) followed by 3:1:2 CH2Cl2/MeOH/n-hexanes. The combined
fractions were evaporated to dryness and dried under vacuum to
extruded solution through a column of Sephadex G-50 resin (GE
Healthcare BioSciences, Piscataway, NJ).
Calcein-Release Experiments. The purified calcein-encapsulated
liposomes were diluted with buffer to achieve an arbitrary
concentration in the 10 µg/mL range (based on lipid). This
concentration was determined spectrophotometrically using ε495,free
) ε495,encapsulated ) 7.0 × 104 M-1 cm-1 for calcein24 and an
estimated bilayer thickness ) 4 nm, cross-sectional area of lipids
) 0.7 nm2 and [encapsulated calcein] ) 5 × 10-2 M. Solid
Na2S2O4 (85%, Sigma-Aldrich) or/and an appropriate aliquot of
7% (w/v) Triton X-100 (Sigma-Aldrich) were added to the cuvettes
to attain the concentrations described in Figure 3. Fluorescence
intensities were recorded using a Perkin-Elmer LS 50 luminescence
spectrophotometer.
Liposome Characterization. Dynamic light scattering measure-
ments were obtained from backscatter intensities (173°, 633 nm
red laser) obtained at 25 °C on calcein-free and calcein-loaded
liposomes using a Zetasizer Nano ZS (Malvern Instruments,
Worcestershire, U.K.) particle size analyzer. Cryo-TEM images
were obtained using the method of vitrification of frozen hydrated
samples.25 Buffered liposomal solution (2.5 µL) was deposited on
a 400 mesh Cu grid covered with holey carbon film (Quantifoil
Micro Tools EmbH, Jena, Germany) and suspended in a Vitrobot
(FEI, Hillsboro, OR). The grid was blotted for 2 s with filter paper,
then the sample was vitrified using liquified ethane. A JEOL 2010F
TEM (Tokyo, Japan) was used to image the specimens at 200 kV
using a low-dose method, with the Gatan model 626 cryostage
(Gatan, Pleasanton, CA) maintained at 92 K during the whole
experiment. Images were recorded with a Gatan 4k × 4k CCD
camera (Gatan, Pleasanton, CA) and processed with custom-built
EMAN software.26
1
yield the products as yellow waxes. 1-Q1: Yield: 75%; H NMR
(CDCl3, 400 MHz) δ 7.38 (bs, 1H), 5.33 (m, 4H), 5.22 (m, 1H),
4.36 (m, 1H), 4.14 (m, 1H), 3.95 (m, 4H), 3.49 (m, 2H), 2.76 (t,
2H), 2.33 (t, 2H), 2.26 (m, 4H), 2.03-1.96 (m, 17H), 1.54 (m,
4H), 1.26 (m, 40H), 0.88 (t, 6H); 13C NMR (CDCl3, 75 MHz) δ
187.50, 187.33, 173.93, 173.68, 142.84, 141.54, 140.91, 140.53,
130.19, 129.80, 70.77, 64.94, 64.06, 63.17, 40.49, 34.78, 34.41,
34.24, 32.10, 29.98, 29.74, 29.52, 27.42, 25.13, 25.03, 22.88, 14.31,
12.57, 12.46, 12.31; HRMS (ESI) m/z (M - Na)-, calcd )
946.6173, obsd ) 946.5756, 44 ppm error. 1-Q3: Yield: 96%; 1H
NMR (CDCl3, 400 MHz) δ 7.65 (bs, 1H), 5.34 (m, 5H), 4.42 (m,
1H), 4.16 (m, 1H), 3.93 (m, 2H), 3.85 (m, 2H), 3.38 (m, 2H), 2.84
(s, 2H), 2.29 (m, 4H), 2.10 (s, 3H), 1.99 (m, 8H), 1.94 (s, 6H),
1.57 (m, 4H), 1.37 (s, 6H), 1.28 (m, 40H), 0.88 (t, 6H); 13C NMR
(CDCl3, 75 MHz) δ 191.49, 187.60, 174.15, 173.93, 172.86, 154.02,
143.73, 138.04, 137.54, 130.25, 129.75, 70.74, 65.21, 63.83, 63.07,
48.67, 39.84, 38.10, 34.50, 34.31, 32.10, 29.97, 29.74, 29.53, 28.84,
27.42, 25.08, 22.88, 14.32, 12.92, 12.29; HRMS (ESI) m/z (M -
Na)-, calcd ) 974.6486, obsd ) 974.6485, 0.1 ppm error.
Liposome Preparation. 1-Q1 or 1-Q3 (5-7 mg) was dissolved
in CHCl3 (10 mL) in a 25 mL round-bottom flask, and the lipid
solution was evaporated to a thin lipid film using a rotary
evaporator. The films were dried under vacuum for 1 h and
redissolved in pH 7.1 0.1 M phosphate buffer/0.1 M KCl at a
concentration of 1 mg/mL. The solution was aged (1 h) with
occasional vortexing (ca. 10-15 s at 20 min intervals), after which
it was freeze-thawed in a dry ice/acetone bath (7 times), followed
by extrusion (12 times) at ambient temperature through two stacked,
100-nm pore Whatman Nuclepore polycarbonate track-etched
membranes using a Lipex lipid extruder (Northern Lipids, Van-
couver, BC, Canada). The buffer solutions used to prepare
liposomes subjected to DLS measurements were filtered through a
Whatman Anotop 20 nm membrane. For the preparation of calcein-
loaded liposomes, the entire “extrusion” procedure described above
was used, except that the buffered solvent also contained dissolved
calcein (5 × 10-2 M, Sigma-Aldrich, Milwaukee, WI). Following
extrusion, the nonencapsulated dyes were separated from the
liposome-encapsulated dyes by gel filtration (two times) of the
Acknowledgment. This work was supported by the U.S. National
Science Foundation, the State of Louisiana, Louisiana State University,
and the U.S. National Institutes of Health (P41RR02250). We thank
Dr. R. Cueto at LSU for access to the DLS and luminescence
instruments and Nicole M. Hollabaugh for synthesis of 3-HQ1.
Supporting Information Available: Voltammetry of 2-Q1 and
2-Q3, the kinetics of reductive lactonization of 2-Q3, and the
NMR and mass spectra of compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
JA8050469
9
14744 J. AM. CHEM. SOC. VOL. 130, NO. 44, 2008