Formulation of photocleavable liposomes and the mechanism of their content release

Article abstract of DOI:10.1039/b518359f

In pursuit of designing photocleavable liposomes as drug delivery vehicles, we synthesized several amphiphilic lipids by connecting stearyl amine (as the non-polar tail) and charged amino acids (as polar heads) via the o-nitrobenzyl derivatives. The lipid

Full text of DOI:10.1039/b518359f

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Formulation of photocleavable liposomes and the mechanism of their
content release†
Binita Chandra,a Rajesh Subramaniam,a Sanku Mallik*a and D. K. Srivastava*b
Received 4th January 2006, Accepted 22nd February 2006
First published as an Advance Article on the web 29th March 2006
DOI: 10.1039/b518359f
In pursuit of designing photocleavable liposomes as drug delivery vehicles, we synthesized several
amphiphilic lipids by connecting stearyl amine (as the non-polar tail) and charged amino acids (as
polar heads) via the o-nitrobenzyl derivatives. The lipids containing Glu, Asp, and Lys amino acids
were subjected to photocleavage reaction by UV light, and the overall spectral changes of the
chromophoric o-nitrobenzyl conjugates were determined as a function of time. The experimental data
revealed that the feasibility of the cleavage reaction, nature and magnitude of the spectral changes
during the course of the cleavage reaction, and their overall kinetic profiles were dictated by the type of
amino acid constituting the polar head groups. The cleavage reactions of the Asp and Glu containing
lipids were found to be more facile than that of the lysine-containing lipid. Using these lipids, we
formulated photocleavable liposomes, and investigated the photo-triggered release of an encapsulated
(within the liposomal lumen) dye as a function of time. The kinetic data revealed that the release of the
liposomal content conformed to a two-step mechanism, of which the first (fast) step involved the
photocleavage of lipids followed by the slow release of the liposomal content during the second step.
The overall mechanistic features intrinsic to the photocleavage of Asp, Glu and Lys containing
o-nitrobenzyl conjugated lipids, and their potential applications in formulating liposomes (whose
contents can be “unloaded” by the UV light) as drug delivery vehicles are discussed.
for effective drug release. In addition, the liposomal contents
(e.g., easily degradable drug molecules, nucleic acids, enzymes,
Introduction
Among drug carriers, liposomes have been considered to be an
etc.), entrapped by endosomes and subsequently exposed to the
efficient drug delivery vehicle. Presently, there are 13 liposome-
lysosomal system, are destroyed prior to eliciting their effects.
mediated drug delivery systems approved for the treatment of
These constraints could be circumvented by developing “triggered
a variety of human diseases (e.g., breast cancer, ovarian cancer,
release methodology” of the liposomal contents. The triggering
agents commonly employed are pH,5 mechanical stress,6 metal
ions,7 temperature,8 light,9 and enzymes.10
meningitis, fungal infections, leukemia, etc.).1 In addition, the
liposome-mediated delivery of about 30 other small molecule
drugs, DNA fragments, and diagnostic compounds are currently
at different stages of clinical trials.1
We recently demonstrated the triggered release of liposomal
contents by a matrix metalloproteinase, MMP-9.11 Since MMP-9
is overexpressed in a variety of cancerous tissues,12 the overall
methodology is likely to find applications in drug delivery.
However, if the cancerous tissues do not overexpress MMP-9,
this method cannot be used for triggering the release of liposomal
contents. In order to develop a more general “triggered” release
methodology, we noted that o-nitrobenzyl substituted compounds
are easily cleaved by near-UV radiation.13 This group is widely
used in organic synthesis as a photolabile protecting group14 and
as a photocleavable linker in solid phase synthesis15 because of its
high photocleavage efficiency in the near-UV range (wavelength
>320 nm). The above group has also been used for “caging”
a variety of biologically important molecules such as ATP,16
cAMP,17 cGMP,18 and neurotransmitters,19 as well as in designing
photo-prodrugs.20 With these applications in mind, we recently
synthesized a few o-nitrobenzyl group-containing lipids (with
acidic amino acids as the polar head groups), and investigated
their cleavage reactions in a preliminary manner.21
In liposome-mediated target-specific drug delivery systems,
three features need to be taken into consideration: (i) appropriate
coating of liposomes to circumvent their clearance by the reticu-
loendothelial system (RES), (ii) attachment of the target-specific
recognition moiety (e.g., antibodies, receptor agonist/antagonist),
and (iii) incorporation of the triggering mechanism. Polyethylene
glycol coating of liposomes precludes their recognition by the
RES,2 and attachment of receptor specific ligands ensures their
adhesion to the target cell surface.3 At the target sites, passive
release of the liposome contents can take place by the fusion
of liposomes with cell membranes.4 But this process is too slow
aDepartment of Pharmaceutical Sciences, North Dakota State Univer-
sity, Fargo, North Dakota, 58105, USA. E-mail: sanku.mallik@ndsu.edu;
Fax: 701-231-8333; Tel: 701-231-7661
bDepartment of Chemistry, Biochemistry and Molecular Biology, North
Dakota State University, Fargo, North Dakota, 58105. E-mail:
dk.srivastava@ndsu.edu; Fax: 701-231-7884; Tel: 701-231-7831
It should be mentioned that photolabile liposomes have
been frequently formulated, in recent years, using various
dithiane-based photocleavable lipids.9 In these formulations,
† Electronic supplementary information (ESI) available: spectra for the
liposomal release studies incorporating the photocleavable lipids. See DOI:
10.1039/b518359f
1730 | Org. Biomol. Chem., 2006, 4, 1730–1740
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
photocleavable amphiphilic lipids are synthesized by interfac-
ing a dithiane group between phosphocholine as the polar
“head” groups and fatty acids as the non-polar “tails”. Photo-
polymerization induced destabilization of liposomes and subse-
quent release of liposomal contents are also reported.9 Although
these lipids are well suited for formulating photocleavable li-
posomes, their synthetic schemes are rather elaborate and time
consuming. In contrast, our syntheses of o-nitrobenzyl containing
photocleavable lipids are fairly simple, and several variants of the
head and tail groups could be easily incorporated to produce
liposomes with diverse photo-triggered “unloading” features.
Herein, we elaborate on the syntheses of selected o-nitrobenzyl
group containing lipids, and investigate the influence of the
oppositely charged polar head groups (contributed by Asp, Glu
and Lys amino acids) on the spectral and kinetic features of the
overall photocleavage reaction. We further demonstrate that these
lipids are ideally suited for formulating photocleavable liposomes,
and discuss the underlying mechanism of their content release.
protected amino acids (aspartic acid, glutamic acid and lysine). In
the final step, the acid-sensitive protecting groups were removed
by treatment with trifluoroacetic acid to provide the lipids. The
overall yields for all the steps were very good and the synthetic
scheme is a significant improvement over that used by Nagasaki
et al. in their synthesis of a similar lysine derivative.22
To ensure that our synthesized lipids underwent photolysis,
and the overall process produced the expected intermediates,
we determined the time course of the spectral changes upon
irradiation of the individual lipid at appropriate wavelengths.
In a typical experiment, an ethanolic solution of a lipid was
irradiated for different time intervals, and the absorption spectra
were immediately recorded. It was observed that the Asp-lipid and
the Glu-lipid were cleaved upon irradiation at 365 nm, but the Lys-
lipid required irradiation at 254 nm for efficient cleavage. Besides,
there was a marked difference in the spectral and kinetic profiles
of these lipids during the course of the photocleavage reaction.
Fig. 2 shows the time dependent spectral changes upon ir-
radiation of Glu-lipid at 365 nm. Initially, Glu-lipid showed a
pronounced spectral band around 245 nm and a shoulder band
around 300 nm. Upon irradiation at 365 nm, the original shoulder
peak appeared to split into two predominant peaks at 290 and
320 nm, with a “valley” at 304 nm. These peaks are apparent in
the difference spectra (bottom left panel of Fig. 2), which were
generated by subtracting the first spectrum (i.e., the spectrum of
Glu-lipid prior to irradiation). The difference spectral data reveal
that as the irradiation time increases, the absorption bands at
all the above wavelengths (i.e., 290, 304, and 320 nm) increase
and the overall spectral transition conforms to common isosbestic
points. These spectral features led to the suggestion that no
stable intermediary species (with distinct spectral characteristics)
populated during the course of the photocleavage reaction. To
further ascertain whether the spectral transitions of Fig. 2 involved
kinetically predictable (metastable) intermediates, we analyzed the
time dependent increase in absorptions at 290 nm, 304 nm, and
320 nm, respectively. The right panels of Fig. 2 show the time
slices of the spectral data at the above wavelengths. These kinetic
data were best fitted by the single exponential rate equation, with
rate constants at 290, 304, and 320 nm of 0.36 0.014, 0.32
Results
The amphiphilic lipids were synthesized by incorporating
“charged” amino acids (viz., Asp, Glu and Lys) as “head” groups
and the alkyl chain of stearyl amine as the “tail” group. These
groups were conjugated to the photocleavable o-nitrobenzyl moi-
ety, such that the cleavage releases the polar head groups, resulting
in the “unloading” of the liposomal content due to destabilization
of the lipid domains. The structures of the synthesized lipids are
shown in Fig. 1.
Scheme 1 summarizes the synthetic details of the lipids in Fig. 1.
To synthesize these lipids, commercially available p-aminomethyl
benzoic acid was trifluoroacetylated with trifluoroacetic anhydride
to give the protected amine in 88% yield. This protected amine was
◦
subjected to regiospecific nitration at −10 C with fuming nitric
acid to give 2 in 92% yield. In the following two steps, the trifluo-
roacetyl group was efficiently replaced by a t-butyloxycarbonyl
(Boc) group (3). Conjugation of the hydrophobic tail to the
carboxylic acid moiety of 3 was performed by reaction with
stearyl amine in presence of the peptide coupling reagents
O-(benzotriazol-1-yl)-N,N,Nꢀ,Nꢀ tetramethyluronium hexafluo-
rophosphate (HBTU), and 1-hydroxybenzotriazole (HOBT). The
Boc group was removed with 4 M HCl in dioxane. The resultant
free amine 4 was then coupled to the acid groups of appropriately
0.009, and 0.33
0.015 min−1, respectively. The solid smooth
lines on the right hand (different wavelength) panels of Fig. 2 are
the calculated lines for the above rate constants. Note a marked
similarity in the above rate constants, attesting to the fact that
Fig. 1 Structures of the synthesized photocleavable lipids.
The Royal Society of Chemistry 2006 Org. Biomol. Chem., 2006, 4, 1730–1740 | 1731
This journal is
©

View Article Online
Scheme 1 The syntheses of the photocleavable lipids.
the spectral transition during the course of the photocleavage of
Glu-lipid proceeded without the formation of any spectral distinct
intermediary species.
Lys-lipid was not being readily photocleaved upon exposure
to 365 nm UV light, the source which was adequate for the
photocleavage of both Glu- and Asp-lipids. The photocleavage
of Lys-lipid required exposure of the sample to a 254 nm (a higher
energy) UV light source. Apparently, the presence of the “basic”
head group (contributed by the e-amino group of lysine) in Lys-
lipid, somehow, thermodynamically and/or kinetically impairs the
photocleavage reaction. The question arose whether or not the
presence of the basic head group of Lys in Lys-lipid would yield
relatively different types of spectral and kinetic profiles vis a vis
those observed with Glu- and Asp-lipids. To ascertain this, we
performed the photocleavage reaction of Lys-lipid exactly under
the experimental conditions of Glu-lipid and Asp-lipid, except for
using a 254 (instead of 365) nm UV light source. Fig. 3 (top left
and bottom left panels) shows the time dependent spectral data
during the course of the photocleavage of Lys-lipid. These spectral
data appear significantly different than those obtained during the
course of cleavage of Glu- (Fig. 2) and Asp-(ESI Figure S1†) lipids.
The spectral data of Fig. 3 reveal that even prior to the onset of the
cleavage process, the spectrum of Lys-lipid is significantly different
those of Glu- and Asp-lipids. Although the major absorption peak
of Lys-lipid at 245 nm is qualitatively similar to those observed
with Glu- and Asp-lipids, the broad, less-resolved shoulder peak
around 300 nm is unique for this lipid. A further difference is a
spectral feature that started to emerge during the time dependent
photocleavage reaction of Lys-lipid vis a vis Glu- and Asp-lipids.
To ascertain whether the spectral and/or kinetic profiles for
the photocleavage of Asp-lipid are similar or different than
those of Glu-lipid, we performed the photocleavage reaction of
the former lipid exactly under the experimental conditions of
Fig. 2. It was observed that the spectral changes during the
course of photocleavage of Asp-lipid were remarkably the same
as those observed with Glu-lipid. The only slight difference was
the magnitude of the rate constants, derived from the spectral
changes at 290, 304, and 320 nm, during the course of the
photocleavage of Asp-lipid. The values for the photocleavage of
Asp-lipid were determined to be 0.45 0.024, 0.40 0.026, and
0.43 0.030 min−1 at 290, 304, and 320 nm, respectively. Since
these kinetic parameters are not significantly different than those
derived from the photocleavage of Glu-lipid, we conclude that the
difference in the side chains of polar head groups (contributed
by Glu versus Asp amino acids) has practically no influence on
the spectral profiles and their associated rate constants during the
photocleavage reaction. To avoid repetition of similar spectral and
kinetic profiles, we provide the photocleavage data of Asp-lipid as
Electronic Supplementary Information (Fig. S1)† of this journal.
Unlike the marked similarities in the spectral and kinetic profiles
during the course of the photocleavage of Glu- and Asp-lipids,
Lys-lipid exhibited quite different behavior. We first noted that
1732 | Org. Biomol. Chem., 2006, 4, 1730–1740
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Fig. 3 Time dependent spectral changes upon photocleavage of Lys-lipid.
The difference spectra (i.e., spectra at different time intervals minus the zero
time spectrum of each panel) are shown just below the main spectra. The
time slices of the spectral changes at 285, 343 and 367 nm are shown on
the right, and the solid smooth lines are the best fit of the data according
to the single exponential rate equation with rate constants of 0.06
0.0011, 0.05 0.0010, and 0.08 0.0012 min−1 at 285, 343, and 367 nm,
respectively.
Fig. 2 Time dependent spectral changes upon photocleavage of Glu-lipid.
The difference spectra (i.e., spectra at different time intervals minus the zero
time spectrum) are shown just below the main spectra. The time slices of
the spectral changes at 290, 304 and 320 nm are shown on the right, and
the solid smooth lines are the best fit of the data according to the single
exponential rate equation with rate constants of 0.36 0.014, 0.32 0.009,
and 0.33 0.015 min−1 at 290, 304, and 320 nm, respectively.
In the case of Lys-lipid, the 290 nm band is slightly blue shifted
to 285 nm, but the 314 nm band is significantly red shifted to
343 nm, with emergence of another shoulder band at 367 nm. In
addition, neither the peak around 245 nm nor the negative peak
around 252 nm were evident in the difference spectra of Fig. 3.
To our further interest, unlike Glu- and Asp-lipids, the spectral
changes in the 280–400 nm region of Lys-lipid (during the course
of photocleavage) were significantly broad and asymmetrical. A
close inspection of the difference spectra (i.e., the spectra of Lys-
lipid at different times of photocleavage minus the uncleaved first
spectrum; bottom left panel of Fig. 3) led us to select the spectral
peaks at 285, 343, and 367 nm for undertaking kinetic analyses of
the photocleavage reaction. The right hand panels of Fig. 3 show
the time slices of spectral changes at the above wavelengths during
the course of photocleavage of Lys-lipid. As observed with Glu-
and Asp-lipids, the photocleavage data of Lys-lipid were also fitted
by the single exponential rate equation. The solid smooth lines
on the right hand panels of Fig. 3 are the nonlinear regression
analysis of the experimental data for the rate constants 0.06
0.0011, 0.05 0.0010, and 0.08 0.0012 min−1 at 285, 343, and
367 nm, respectively. Note that these values are about one order
of magnitude smaller than those observed for the photocleavage
of the Glu- and Asp-lipids. Hence, even upon exposure to a high-
energy source (i.e., irradiation with a 254 nm UV lamp), Lys-lipid is
resistant (compared to Glu- and Asp-lipids) to the photocleavage
reaction.
characterized the feasibility of photocleavage of acidic (viz., Glu
and Asp) and basic (Lys) amino acid head groups containing
o-nitrobenzyl conjugated lipids, we incorporated them into lipo-
somes. The liposomes were prepared in 25 mM HEPES buffer,
pH = 7.0, using 1,2-disteroylglycero-3-phosphocholine (DSPC)
as a major component (95 mol%) and individual photocleavable
lipids as minor components (5 mol%), encapsulating carboxyfluo-
rescein as the self-quenching23 reporter dye in the liposomal lumen.
The excess (unincorporated) dye was removed by gel filtration to
yield liposomes, which were only loaded with the reporter dye
(see the experimental section). A representative TEM picture of a
liposome formulated by using Glu-lipid is shown as Figure S2 in
the ESI.†
We determined the fluorescence excitation and emission maxima
of liposome-encapsulated carboxyfluorescein as being equal to 495
and 527 nm, respectively (data not shown). It is known that upon
release from liposomal lumen, the fluorescence emission intensity
of carboxyfluorescein increases,11,21 due to dilution induced de-
quenching of the excited state of the dye.23 Fig. 4A shows a
representative set of fluorescence emission spectra (kex = 495 nm)
for the release of carboxyfluorescein upon irradiating liposomes
containing the Glu-lipid at 365 nm for different time intervals. A
control experiment was performed in which the liposome solution
was not irradiated, but the fluorescence spectra were recorded at
corresponding time intervals (data not shown). From the spectral
data of Fig. 4A, we extracted the time dependent increase in the
fluorescence emission intensity (due to release of the liposome
encapsulated carboxyfluorescein to the exterior media) at 518 nm.
The ultimate goal of synthesizing these photocleavable lipids
was to utilize them for formulating liposomes for the triggered
release of their contents upon exposure to light. Hence, having
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1730–1740 | 1733
©

View Article Online
associated with the photocleavage of liposomal lipids, resulting in
the destabilization of lipid domains and leakiness of the liposomes.
This is followed by the slow release of the encapsulated dye during
the second (nearly a single exponential) phase. The experimental
data could be best fitted by a two step sequential model (see
Discussion), with fast (k1) and slow (k2) rate constants of 0.25
0.018 and 0.040 0.0024 min−1, respectively. Of these, the fast rate
constant (k1) corresponds to the single exponential rate constant
for the photocleavage of free Glu-lipid (0.32–0.36 min−1; Fig. 2).
The slow rate constant (k2) is ascribed to be the measure of
the release of the encapsulated (carboxyfluorescein) dye upon
destabilization of the liposome. This rate is about one order of
magnitude lower than the rate of the preceding photocleavage
step. The fluorescence signal obtained during these experiments is
due to the release of carboxyfluorescein. The photocleavage of the
lipids and the destabilization of liposomal domains do not yield
any fluorescence signal, and thus are manifested in the lag phase.
Fig. 5 Kinetic profiles for the release of carboxyfluorescein upon
irradiation of liposomes containing Lys-lipid. The increase in fluorescence
intensity (DF518/nm; kexc = 495 nm) as a function of time upon irradiation
of liposomes (open circles) and when they are kept in the dark (control
experiment; filled triangles) are shown. The inset shows an expansion of
the data at the initial time scale to show the initial lag phase. The solid
smooth line is the best fit of the data for a two-step release of the liposomal
content according to eqn. 2, with k1 and k2 values of 0.026 0.0037 and
0.025 0.0033 min−1, respectively.
Fig. 4 Spectral and kinetic profiles for the release of carboxyfluorescein
upon irradiation of Glu-lipid. A: Fluorescence emission spectra (kex
=
480 nm) as a function of irradiation time at 365 nm. B: The change in
fluorescence intensity (DF518/nm) as a function of time (open circle). The
control experiment (solid triangle) was performed without irradiating the
liposomes. The inset shows an expansion of the data at the initial time
scale to show the lag phase. The solid smooth line is the best fit of the data
for the photorelease of the liposomal content according to eqn. 2 with k1
and k2 values of 0.25 0.018 and 0.04 0.0024 min−1, respectively.
These along with the control (in which the liposomes were not
irradiated) data are shown as open circles and filled triangles,
respectively, in Fig. 4B. Note that the fluorescence emission
intensity of the reporter dye increases only when the liposomes
are irradiated at 365 nm, and not when they are kept in the dark.
Clearly, the irradiation of the carboxyfluorescein loaded liposomes
at 365 nm results in the release of the dye due to the photocleavage
of the resident Glu-lipid.
On examination of the time course for the release of carboxyflu-
orescein upon photocleavage of liposomes, it became evident
that the overall kinetic profile involved an initial lag phase. The
existence of such a lag phase is clear upon expanding the kinetic
data for the initial time scale (see the inset of Fig. 4B), and it is more
so in the case of photocleavage of Lys-lipid formulated liposomes
(see Fig. 5). As will be discussed subsequently, the lag phase is
We performed a control experiment to determine the extent
to which carboxyfluorescein is bleached during the course of
photocleavage of liposomes and their content release. This was
important since the amplitude of the fluorescence signal (due to
release of the dye from the liposomal lumen) would be affected by
the rate as well as the magnitude of photobleaching of the reporter
dye. However, when we performed such an experiment under the
experimental conditions of Fig. 4, we did not observe any sig-
nificant photobleaching of carboxyfluorescein (data not shown),
which could be envisaged to interfere in assessing the amplitude
of fluorescence changes during the course of the photocleavage of
Glu-lipid and unloading of the liposome encapsulated dye. Hence,
our overall mechanistic conclusion for the release of the liposomal
content via a “two-step” pathway is not affected by a potential
artifact due to photobleaching of the reporter dye.
1734 | Org. Biomol. Chem., 2006, 4, 1730–1740
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
Since the photocleavage of Lys-lipid was not as efficient as that
of Glu- and Asp-lipids (see Figs. 2 and 3), it was of interest to
investigate as to how the liposomes formulated by the former lipid
would photorelease their encapsulated contents. Toward this goal,
we formulated Lys-lipid containing liposomes and encapsulated
carboxyfluorescein as a reporter dye as described earlier. As
observed with Glu-lipid-containing liposomes, upon irradiation of
these liposomes (albeit at 254 nm, since Lys-lipid is not easily pho-
tocleaved at 365 nm), there was a time dependent increase in fluo-
rescence at 520 nm (kexc = 495 nm). Fig. 5 shows the kinetic profile
for the release of carboxyfluorescein as a function of the irradiation
time of the liposomes (open circles). We also performed a control
experiment in which liposomes were not irradiated (filled trian-
gles). Since the control experiment did not produce any increase in
fluorescence, it implied (also as observed in the case of Glu-lipid
formulated liposomes) that the unloading of carboxyfluorescein
was as a consequence of irradiation of the liposomes, which
resulted in the photocleavage of the liposome resident Lys-lipid.
An explicit feature of the kinetic data of Fig. 5 has been the pre-
ponderance of the “lag phase”, and the latter is more pronounced
upon magnifying the initial time scale data points (see the inset of
Fig. 5). This is presumably due to the fact that the rate of photo-
cleavage of Lys-lipid (k = 0.05–0.08 min−1) is considerably slower
than that of Glu- and Asp-lipids (k = 0.32 to 0.45 min−1). On
assumption that the mechanistic pathway for the photocleavage
of liposomes and their content release remains unaffected by the
nature of the photocleavable lipids, we analyzed the data of Fig. 5
by a two-step mechanism of eqn. 2. The solid smooth line is the best
fit of the data for the k1 and k2 values of 0.026 0.0037 and 0.025
0.0033 min−1, respectively. Note that, unlike the photocleavage of
Glu- and Asp-lipid formulated liposomes and their content release,
the magnitudes of k1 and k2 are nearly identical. However, it should
be pointed out that these values have been derived from the two
distinct kinetic features, i.e., lag and exponential phases, and thus
they are not due to apparent artifacts encountered in fitting the
kinetic data (comprised of either increasing or decreasing signals)
by single exponential versus biphasic rate equations.
are about one order of magnitude lower than those obtained with
Glu-lipid and its associated liposomes. On the other hand, the
magnitude of k2 is similar irrespective of whether the liposome
is formulated with Glu-lipid (k2 = 0.04 min−1) or Lys-lipid (k2 =
0.025 min−1). Given these observations, it is tempting to propose
that the photocleavage of the o-nitrobenzyl conjugated (acidic
as well as basic) lipids occur during the lag phase. Since the
photocleavage reaction is more facile with Glu-lipid than Lys-
lipid, the magnitude of k1 is about one order of magnitude higher
with liposomes formulated with Glu-lipid than with Lys-lipid.
But once the liposomes are photocleaved, the release of their
content (i.e., carboxyfluorescein as reporter dye) is not dependent
(as reflected in similar k2 values) on the type of amino acid head
group utilized to formulate the liposomes.
Discussion
The o-nitrobenzyl group-containing lipids exhibited a major
absorption peak (predominantly contributed by the aromatic ring)
around 250 nm, and a trailing shoulder peak at 300 nm. Although
it has been known that substitution at the o-nitrobenzyl ring alters
its spectral profile, there has been no precedence (to the best of
our knowledge) of the distal (side chain) groups of the amino
acids influencing the spectral profiles of the chromophore. This is
important since all amino acids differ only with respect to their
side chain groups, and those groups are far removed from the
aromatic ring to exhibit any inductive and/or resonance effects.
The question arises how acidic (Glu/Asp) and basic (Lys) amino
acid containing photocleavable lipids exhibit different spectral and
rate profiles during the course of the cleavage reaction. However,
before attempting to answer this question, let us consider a general
sequence of steps intrinsic to the photocleavage of o-nitrobenzyl
group-containing lipids (Scheme 2).
Based on the literature data,13 it is surmised that upon ab-
sorption of the UV light, anionic radical species are generated
on the excited nitro group (compound 8, Scheme 2), which
abstracts a benzylic proton and isomerizes to a metastable aci-
nitro intermediate (compound 9).14 The latter is rearranged to yield
the nitroso aminol derivative (compound 10). This is followed by
the hydrolysis of the aminol intermediate as the slowest step in the
overall photocleavage process.15 Since the absorption spectra
Irrespectively, it is noteworthy that the magnitude of k1
(0.026 min−1), derived from the photocleavage of Lys-lipid con-
taining liposomes (Fig. 5), is comparable to the rate constant for
the photocleavage of Lys-lipid (k = 0.05–0.08 min−1). These values
Scheme 2 The proposed sequence of steps during the photocleavage of o-nitrobenzyl conjugated lipids.
The Royal Society of Chemistry 2006 Org. Biomol. Chem., 2006, 4, 1730–1740 | 1735
This journal is
©

View Article Online
during the photocleavage of Asp/Glu-lipid are markedly different
than those of Lys-lipid, it follows that the side chain groups of
the amino acids are responsible for influencing the electronic
structures of either the corresponding precursors (compound 8) or
the intermediates (compounds 9 and 10), but not the final product
(compound 11), as the latter is already devoid of amino acids. Of
these, the electronic structural effect is most likely to be manifested
at the level of compound 10 or its subsequent derivative(s) (see
Scheme 3). At a neutral pH, the side chain groups of Asp/Glu and
Lyswouldpredominateas –COO− and–NH3+ species, respectively.
Of these, the –NH3+ group of Lys has the potential to interact with
either nitroso oxygen (panel A of Scheme 3) or the amide oxygen
(panel B of Scheme 3) of compound 10. Such interactions are
expected to facilitate the deprotonation (due to a decrease in the
pKa value) of the aminol carbon atom, resulting in the formation
of compounds 12 and 13, respectively (Scheme 3).
Based on the above discussion, it should not be envisaged
that the observed spectral changes Figs. 2 and 3 are primarily
dominated by the precursor “nitro” and the product “nitroso”
derivatives.14,15,25 Although other chromophoric intermediates are
likely to form during the photocleavage of o-nitrobenzyl conju-
gates, they are not detectable due to their shorter lifetimes.
Another interesting aspect of photocleavage is the fact that the
rate constant for the cleavage of Glu/Asp-lipid is about one order
of magnitude higher that of the Lys-lipid (Figs. 2 and 3). This
is besides the fact that the photocleavage of Lys-lipid requires
a higher energy (254 nm) UV source than Glu/Asp-lipid; the
latter lipids are cleaved even upon irradiation with a 365 nm UV
source. We think that the origin of both these features lays in
the preferential stabilization of the “precursor” ground state of
the Lys-lipid as compared to that of the Glu/Asp lipid. This can
+
occur due to the interaction of the –NH3 group of Lys with
Due to their conjugated nature, both compounds 12 and 13 can
yield the red-shifted absorption spectra, albeit they are expected
to be more pronounced with compound 13 than with compound
12. To differentiate which of the above compounds would be more
energetically stable, we performed the energy minimizations (ref.
24, employing the semiempirical force field PM3) of intermediates
12 and 13. This endeavour led to the suggestion that compound
13 was more stable than compound 12. Hence, the absorption
spectra during the course of the photocleavage of Lys-lipid is
likely to be dominated at least by the partial accumulation of
the intermediate 13. Therefore, we propose that, although the
photocleavage of both Asp/Glu and Lys-lipids predominantly
removes the corresponding amino acids (Scheme 2), a minor
fraction of the intermediate (particularly in the case of Lys-lipid) is
transformed into 13, and this process is responsible for red shifted
spectral profiles during the course of photocleavage of Lys-lipid.
the amide oxygen of the precursor (compound 8, Scheme 2),
essentially via the same mechanism involved in yielding the red-
shifted absorption spectra during the photocleavage of Lys-lipid
(Scheme 2). Hence, although the overall microscopic pathways for
the photocleavage of different amino acid containing lipids can
be the same, their spectral and kinetic profiles can be significantly
different. Therefore, detailed mechanistic studies for the overall
photocleavage reaction of o-nitrobenzyl group-containing amino-
lipid conjugates are essential for designing liposomes as efficient
drug delivery vehicles.
The release of the liposome-encapsulated carboxyfluorescein
upon irradiation can be envisaged to be the product of two
microscopic events: the first involving the photocleavage of
liposome resident o-nitrobenzyl conjugated lipids followed by the
destabilization of liposomes and release of the encapsulated dye
during the second step. These microscopic events are corroborated
Scheme 3 Alternative modes of binding of the side chain group of Lys of compound 10, and generation of conjugated reaction products.
1736 | Org. Biomol. Chem., 2006, 4, 1730–1740 This journal is The Royal Society of Chemistry 2006
©

View Article Online
by the emergence of a “lag” followed by an “exponential” kinetic
phase (Figs. 4 and 5) during the course of photocleavage and
“unloading” of the liposomal content. The prevalence of the “lag”
phase is more explicit on examining the kinetic data at shorter time
scales (insets of Figs. 4 and 5). Since the increase in the fluorescence
signal of carboxyfluorescein is due to the release of the dye from
photocleaved liposomes, and not due to the photocleavage event
per se, the latter process can be easily manifested in the “lag” phase
in the experimental conditions of Figs. 4 and 5. A cumulative
account of the experimental data leads us to propose a “two-step”
minimal model for the photocleavage of liposomes and unloading
of their contents (eqn. 1).
Conclusions
The experimental data presented herein lead to the following
conclusions: (1) Photocleavable lipids can be easily synthesized
by bridging a C18 chain fatty acid (non-polar group) and Glu,
Asp, and Lys as charged amino acids (as polar groups) with an o-
nitrobenzyl moiety. (2) Although all the charged amino acid head
groups can be utilized to formulate liposomes, the Glu/Asp-lipid-
harboring liposomes are more easily cleaved than Lys-lipid lipid-
harboring liposomes. (3) The photocleavage of liposomes and the
release of their contents proceed via a two-step mechanism, of
which the feasibility of the first (photocleavage) step is dependent
on whether the amino acid (forming the head group) is acidic or
basic in nature. The second (rate limiting) step is dictated by the
organization of the lipid domains following the separation of the
amino acid head groups. The insight gained from these studies will
serve as the prototype for formulating photocleavable liposomes
and using them as drug delivery vehicles for treating different types
of diseases.
k
k
2
1
L − F −−−→ L* − F −−−→ L* + F
(1)
Here, L and F represent liposomes and carboxyfluorescein,
respectively. The first step involves the photocleavage of the
o-nitrobenzyl conjugated lipids, incorporated in the liposomes,
leading to destabilization of the liposomes. The intermediary
species L*–F is representative of the photocleaved liposomes,
with fluorescence dye still entrapped in their lumen. The second
step is the release of the fluorescence dye with a concomitant
increase in the fluorescence emission intensity at 520 nm. For this
irreversible two step model, the time course of formation of free
carboxyfluorescein (F) can be given by equation 2.
Experimental
Commercial reagents were purchased from either Aldrich or Acros
Chemical Co. The protected amino acids were purchased from
Nova Biochem. Nitric acid (90%) was from Alfa Aesar. All
solvents used for reactions were analytical grade and were used
without further purification. Melting points were determined on
a micro melting point apparatus. H and 13C NMR spectra were
recorded using 300, 400 or 500 MHz spectrometers.
Solvents used for NMR were one of the following: CDCl3,
CD3OD or DMSO-d6 with TMS as the internal standard.
Elemental analyses were obtained from facilities at Desert
Analytics (Tucson, AZ). TLC was performed with Adsorbosil
plus IP, 20 × 20 cm plate, 0.25 mm (Altech Associates, Inc.).
Chromatography plates were visualized by either UV light
or in an iodine chamber. For drying water-wet compounds,
lyophilization (Freeze Dry system/Freezone 4.5; Labconco) was
used. Reactions were performed either under an N2 atmosphere
or using a guard tube. For extractive workups, the organic
layer was dried over anhydrous Na2SO4, and concentrated
in vacuo.
1
ꢀ
ꢁ
ꢂ
ꢃ
1
F = [L − F] 1+
{k2 exp(−k1t) − k1 exp(−k2t)}
(2)
k1 − k2
It should be mentioned that eqn. 2 has been explicitly derived
by Frost and Pearson,26 and has been utilized extensively for
analyzing the kinetic data of “two-step” irreversible reactions
such as that given by eqn. 1. The kinetic data of Figs. 4 and 5
could be easily analyzed by eqn. 2, yielding the k1 and k2 values
intrinsic to the photocleavage of differently formulated liposomes
and releasing their contents. Of derived kinetic parameters (from
the best fit of the experimental data), it is apparent that the k1 value
matches the single exponential rate constant for the photocleavage
of the corresponding lipid. For example, k1 is about one order of
magnitude higher for Glu-lipid (k1 = 0.25 min−1) than for Lys-lipid
(k1 = 0.026 min−1) since the rate of photocleavage of the former
lipid (k = 0.32–0.36 min−1; Fig. 2) is about one order of magnitude
faster than that of the latter lipid (k = 0.05–0.08 min−1; Fig. 3).
On the other hand, the magnitude of k2 is not affected by the type
of lipid utilized to formulate liposomes. For example, the k2 values
derived from photocleavage and unloading of encapsulated dye
are similar for Glu-lipid (0.04 min−1) and Lys-lipid (0.025 min−1).
Clearly, once the liposome resident lipids are cleaved, the release
of encapsulated dye is dominated by some common molecular
mechanism, and that mechanism is likely to be the organization of
the liposomal domains (after the photocleavage of the polar head
group). We surmise that the organization of the lipid domains
serves as the rate-limiting step for unloading its contents to exterior
media. We are currently in the process of testing this hypothesis by
formulating different types of liposomes, and we will report these
findings subsequently.
3-Nitro-4-(trifluoroacetylaminomethyl)benzoic acid (2)
Trifluoroacetic anhydride (5.9 mL, 41.3 mmol) was added in small
portions to solid 4-(aminomethyl) benzoic acid (2.5 g, 16.5 mmol)
◦
at 4 C. Upon completion of addition, the reaction mixture was
◦
homogeneous. Stirring was continued at 25 C for 2 h, and then
ice–water was added to precipitate the product. The white solid
was collected by filtration, washed with water and dried. Yield:
◦
1
3.63 g (88%), mp: 199–203 C; H NMR dH (300 MHz; CDCl3;
293 K; Me4Si) 7.91 (d, J 7.8, 2H, ArH), 7.37 (d, J 7.8, 2H, ArH),
4.44 (s, 2H, ArCH2).
The above compound (3.63 g, 14.7 mmol) was added over 1 h
to 90% nitric acid (20 mL) at −10 ◦C. The mixture was stirred
further for 1.5 h at 0 ◦C and then poured onto ice to precipitate the
product. The precipitated solid was filtered, washed with plenty of
water to neutral pH, and lyophilized to provide an off-white solid
◦
1
(3.95 g, 92%), mp: 210–211 C; H NMR dH (300 MHz; CDCl3;
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1730–1740 | 1737
©

View Article Online
298 K; Me4Si) 8.61 (d, J 1.6, 1H, ArH), 8.17 (dd, J 1.6, 8.1, 1H,
ArH), 7.51 (d, J 8.1 Hz, 1H, ArH), 4.74 (s, 2H, ArCH2).
(Na2SO4) and solvent was removed by rotary evaporation. The
crude product was purified by silica gel chromatography (eluant:
CHCl3, Rf = 0.2) to yield compound 5 as a white solid (0.480 g,
◦
1
99%), mp: 90–92 C; H NMR dH (300 MHz; CDCl3; 298 K,
Me4Si) 8.41 (d, J 1.6, 1H, ArH), 7.98 (dd, J 1.6, 8.0, 1H, ArH),
7.70 (d, J 8.0, 1H, ArH), 4.82–4.70 (m, 2H, ArCH2), 4.51–4.44
(m, 1H, COCH), 3.47 (q, J 7, 2H, CONHCH2), 2.96–2.88 (m,
1H, CH2CO2But), 2.62–2.58 (m, 1H, CH2CO2But), 1.66–1.58 (m,
2H, CONHCH2CH2), 1.45 (s, 9H, ester t-butyl CH3), 1.42 (s, 9H,
carbamate t-butyl CH3), 1.40–1.20 (m, 30H, stearyl CH2), 0.88 (t,
J 7.0, 3H, stearyl CH3).
4-(Boc-aminomethyl)-3-nitrobenzoic acid (3)
According to a general procedure,27 a solution of compound 2
(0.68 g, 2.3 mmol) and K2CO3 (0.81 g, 5.8 mmol) in MeOH–
◦
H2O (1 : 1, 16 mL) was maintained at 25 C for 10 h. The dark
yellow solution was concentrated under reduced pressure, and the
resultant solid was dissolved in dioxane–H2O (1 : 1, 10 mL). Di-
tert-butyl dicarbonate (0.77 g, 3.5 mmol) was added, and after
2.5 h, the reaction mixture was concentrated in vacuo. Ether and
water were added, and the aqueous phase was washed with ether,
the pH was adjusted to 3.0 with 10% aqueous citric acid, and
extracted with ethyl acetate. The combined ethyl acetate phases
were washed with brine, dried over Na2SO4, and concentrated
in vacuo to give the title product as a yellow solid (0.67 g, 97%),
4-t-Butoxycarbonylamino-4-(2-nitro-4-
octadecylcarbamoylbenzylcarbamoyl)butyric acid t-butyl ester (6)
Compound 4 (0.3 g, 0.67 mmol), Boc-Glu(OtBu)–OH (0.2 g,
0.67 mmol), HOBT (0.09 g, 0.67 mmol) and HBTU (0.25 g,
0.67 mmol) were taken up in DMF (15 mL) and N-
methylmorpholine (0.15 mL, 1.34 mmol) was added. The reaction
mixture was stirred at room temperature overnight. The work-up
procedure was the same as described for compound 5. The crude
product was then purified by silica gel chromatography (eluant:
CHCl3, Rf = 0.3) to provide the glutamic acid derivative 6 as a
yellow solid. Yield: 0.34 g (70%); 1H NMR dH (300 MHz; CDCl3;
298 K, Me4Si) 8.42 (d, J 1.6, 1H, ArH), 8.98 (dd, J 1.6, 8.0, 1H,
ArH), 7.71 (d, J 8.0, 1H, ArH), 4.75 (d, J 6, 2H, ArCH2), 4.15–4.08
(m, 1H, COCH), 3.47 (q, J 7, 2H, NHCH2), 2.43–2.37 (m, 1H,
CH2CH2CO2But), 2.31–2.25 (m, 1H, CH2CH2CO2But), 2.11–2.02
(m, 1H, CH2CH2CO2But), 1.95–1.87 (m, 1H, CH2CH2CO2But),
1.65–1.6 (m, 2H, NHCH2CH2), 1.45 (s, 9H, ester t-butyl CH3),
1.42 (s, 9H, carbamate t-butyl CH3), 1.40–1.20 (m, 30H, stearyl
CH2), 0.88 (t, J 7.0, 3H, stearyl CH3).
◦
1
mp 124–126 C; H NMR dH (300 MHz; CDCl3; 298 K; Me4Si)
8.74 (d, J 1.6, 1H, ArH), 8.30 (dd, J 1.6, 8.0, 1H, ArH), 7.77 (d, J
8.0, 1H, ArH), 4.65 (s, 2H, ArCH2), 1.44 (s, 9H, t-butyl CH3).
N-Stearyl-4-(aminomethyl)-3-nitrobenzamide (4)
Compound 3 (0.8 g, 2.7 mmol) was dissolved in CHCl3 (20 mL)
and stearic acid (0.71 g, 2.7 mmol), HOBT (0.364 g, 2.7 mmol),
HBTU (1.024 g, 2.7 mmol) and Et3N (0.75 mL, 5.4 mmol) were
added to the solution. The mixture was stirred at room tempera-
ture for 10 h. The reaction mixture was then washed with water; the
organic phase dried and solvent was removed in vacuo. The residue
was purified by column chromatography (eluant: 5% methanol in
chloroform, Rf = 0.3) to obtain the pure product as a yellow solid
◦
1
(1.19 g, 81%), mp: 84–86 C; H NMR dH (300 MHz; CDCl3;
298 K, Me4Si) 8.41 (d, J 1.8, 1H, ArH), 8.00 (dd, J 1.8, 8.1, 1H,
ArH), 7.68 (d, J 8.1, 1H, ArH), 6.41 (br s, 1H, CONH), 5.30 (br s,
2H, NH2), 4.59 (s, 2H, ArCH2), 3.45 (q, J 6.9, 2H, CONHCH2),
1.65–1.57 (m, 2H, CONHCH2CH2), 1.42 (s, 9H, t-butyl CH3),
1.24–1.33 (m, 30H, stearyl CH2), 0.87 (t, J 6.9, 3H, stearyl CH3).
To the above compound (1.16 g, 2.12 mmol), was added, 4 M
HCl in dioxane (8 mL) and the reaction mixture was stirred at
room temperature for 3 h. The solvent was then removed under
vacuum and water added to the residue. The insoluble white
solid was filtered, washed with plenty of water and dried to give
compound 4 (0.89 g, 94%) as a yellow solid. The compound was
carried on to the next step without further purification. 1H NMR
dH (300 MHz; CDCl3; 298 K, Me4Si) 8.54 (d, J 1.8, 1H, ArH),
8.03 (dd, J 1.8, 7.5, 1H, ArH), 7.75 (d, J 7.5, 1H, ArH), 4.31 (s,
2H, ArCH2), 3.31 (q, J 6.9, 2H, CONHCH2), 1.53–1.49 (m, 2H,
CONHCH2CH2), 1.40–1.12 (m, 30H, stearyl CH2), 0.78 (t, J 7,
3H, stearyl CH3).
[5-t-Butoxycarbonylamino-1-(2-nitro-4-
ocatdecylcarbamoylbenzylcarbamoyl)pentyl]carbamic
acid t-butyl ester (7)
Compound 4 (0.3 g, 0.67 mmol), Boc-Lys(Boc)–OH (0.22 g,
0.67 mmol), HOBT (0.09 g, 0.67 mmol) and HBTU (0.25 g,
0.67 mmol) were taken up in DMF (15 mL) and N-
methylmorpholine (0.15 mL, 1.34 mmol) was added. The work-up
procedure was the same as described for compound 5. The crude
product was purified by silica gel chromatography (eluant: 2%
MeOH in CHCl3, Rf = 0.2) to yield the lysine derivative 7 as a
1
white foamy solid. Yield: 0.45 g (87%); H NMR dH (500 MHz;
CDCl3; 298 K, Me4Si) 8.38 (d, J 1.6, 1H, ArH), 7.96 (dd, J 1.6, 8.0,
1H, ArH), 7.62 (d, J 8.0, 1H, ArH), 4.76–4.64 (m, 2H, ArCH2),
4.66–4.63 (m, 1H, COCH), 3.43 (q, J 7, 2H, NHCH2), 3.02–2.95
(m, 2H, CH2NHBoc), 1.82–1.72 (m, 2H, lysine CH2) 1.66–1.58
(m, 2H, NHCH2CH2), 1.43 (s, 9H, t-butyl CH3), 1.40 (s, 9H, t-
butyl CH3), 1.39–1.35 (m, 4H, lysine CH2), 1.35–1.15 (m, 30H,
stearyl CH2), 0.87 (t, J 7.2, 3H, stearyl CH3).
3-t-Butoxycarbonylamino-N-(2-nitro-4-
octadecylcarbamoylbenzyl)succinamic acid t-butyl ester (5)
Compound 4 (0.3 g, 0.67 mmol), Boc-Asp(OtBu)–OH dicyclo-
hexylamine salt (0.32 g, 0.67 mmol), HOBT (0.091 g, 0.67 mmol)
and HBTU (0.25 g, 0.67 mmol) were taken up in DMF (15 mL)
and N-methylmorpholine (0.15 mL, 1.34 mmol) was added. The
reaction mixture was stirred at room temperature overnight. The
solvent was removed in vacuo. Water was added to the residue
and extracted with ethyl acetate. The ethyl acetate layer was dried
3-Amino-N-(2-nitro-4-octadecylcarbamoyl-benzyl)-succinamic
acid (Asp-lipid)
To the Boc-Asp(OtBu) derivative 5 (0.40 g, 0.56 mmol), was
added 4 mL of trifluoroacetic acid and a drop of anisole. The
reaction mixture was stirred at room temperature for two hours.
It was then slowly added to water and aqueous NaOH solution
1738 | Org. Biomol. Chem., 2006, 4, 1730–1740
This journal is
The Royal Society of Chemistry 2006
©

View Article Online
(1 M) was slowly added to neutralize the TFA. The precipitate
was collected by filtration, washed with plenty of water and
dried to give Asp-lipid as a off-white solid (0.27 g, 85%); mp:
95 mol%) were dissolved in 5 mL of anhydrous chloroform and
500 lL of anhydrous methanol in a 25 mL clean, oven-dried
round bottomed flask. The organic solvents were then removed
in a rotary evaporator under reduced pressure maintaining the
bath temperature at 40 ◦C until a thin and uniform lipid film was
formed on the walls of the round-bottomed flask. The flask was
left on the rotary evaporator for an additional 15 minutes and then
allowed to dry in vacuo for 20 hours.
In another clean dry glass vial, 56 mg (150 lmoles) of 6-
carboxyfluorescein was taken up in 3 mL of HEPES buffer
(25 mM, pH = 7.0). The dye was dissolved by first bath-sonicating
(to reduce the particle size of the solid granules of the dye) to
form a dark brown transparent solution. The thin dry lipid film
was then hydrated with the dye solution (3 mL) by rotating slowly
in the rotary evaporator bath at 60 ◦C for 1 hour. The result-
ing suspension was then subjected to probe sonication (power:
50 W) at 60 ◦C for 1 hour with constant nitrogen bubbling, to get
a clear dark red liposome solution. The total lipid concentration
was 9 mM. The osmolarity of the liposome solution was measured
with a standard micro osmometer.
Sephadex G-50 resin (particle size 50–150 lm) was mixed with
an excess of water to form a gel and the gel was hydrated overnight
at 40 ◦C in the water bath of a regular rotary evaporator. A
chromatography column was packed with the gel after cooling to
room temperature and equilibrated with 200 mL of water whose
osmolarity was made equal to that of the liposome solution by
the addition of solid sodium chloride. The liposome solution
was then loaded on top of the column and slowly eluted. The
liposomes came out first as a yellow non-fluorescent solution and
were collected.
◦
1
154–157 C; H NMR dH (400 MHz; d6-DMSO; 298 K, Me4Si,
without exchangeable protons) 8.43 (d, J 1.6, 1H, ArH), 8.10
(dd, J 1.6, 8.0, 1H, ArH), 7.62 (d, J 8.0, 1H, ArH), 4.66–
4.56 (m, 2H, ArCH2), 3.96–3.92 (m, 1H, COCH), 3.24 (q, J 7,
2H, NHCH2), 2.75–2.58 (m, 2H, CH2CO2H), 1.60–1.40 (m, 2H,
NHCH2CH2), 1.35–1.17 (m, 30H, stearyl CH2), 0.82 (t, J 7.0, 3H,
stearyl CH3); 13C NMR dC (100 MHz; d6-DMSO; 298 K) 176.05,
171.99, 64.42, 148.52, 136.77, 135.50, 132.54, 130.66, 123.81,
50.71, 37.19, 31.87, 30.38, 29.58–29.23, 27.10, 22.64, 14.45. Anal.
Calcd. for C30H50N4O6.3CF3COONa.4H2O: C, 41.46; H, 5.61; N,
5.37. Found: C, 41.25; H, 5.92; N, 5.43%.
4-Amino-4-(2-nitro-4-
octadecylcarbamoylbenzylcarbamoyl)butyric acid (Glu-Lipid)
To the Boc-Glu(OtBu) derivative 6 (0.26 g, 0.36 mmol), was added
4 mL of trifluoroacetic acid and a drop of anisole. The reaction
mixture was stirred at room temperature for 2 h. The work-up
procedure was the same as described for Asp-lipid. Glu-lipid was
isolated as a white solid (0.2 g, 99%); mp: 150–152 ◦C; 1H NMR
dH (400 MHz; d6-DMSO; 298 K, Me4Si, exchangeable protons not
reported) 8.49 (d, J 1.6, 1H, ArH), 8.14 (dd, J 1.6, 8.0, 1H, ArH),
7.64 (d, J 8.0, 1H, ArH), 4.70–4.62 (m, 2H, ArCH2), 3.92–3.84
(m, 1H, COCH), 3.24 (q, J 7, 2H, NHCH2), 2.36–2.22 (m, 2H,
CH2CH2CO2H), 1.99–1.90 (m, 2H, CH2CH2CO2H), 1.58–1.40
(m, 2H, NHCH2CH2), 1.35–1.10 (m, 30H, stearyl CH2), 0.82 (t, J
7.0, 3H, stearyl CH3); 13C NMR dC (100 MHz; d6-DMSO; 298 K)
176.02, 169.43, 165.68, 147.94, 135.73, 135.38, 132.15, 130.50,
124.08, 52.69, 40.97, 40.47, 31.96, 30.38, 29.73–29.38, 27.13, 26.72,
22.70, 14.03. Anal. Calcd. for C31H52N4O6.CF3COONa.H2O: C,
55.61; H, 7.85; N, 7.86. Found: C, 55.27; H, 8.07; N, 8.02%.
Uncorking of liposomes and their content release
The fluorescence emission spectrum of the dye-encapsulated
liposomes was recorded with excitation at 580 nm. The quartz
cuvet was then placed under a UV lamp (100 W lamp for the
365 nm irradiation for liposomes incorporating Asp-lipid and
Glu-lipid; 10 W lamp for 254 nm irradiation for liposomes
incorporating Lys-lipid; distance from the lamp: 5 cm; area of
illumination: 1 cm × 1 cm). Every 5 minutes, the cuvet was
transferred to the fluorimeter and the emission spectrum was
recorded. The intensity of the emission maximum (518 nm) was
plotted as a function of time to generate the release curves for the
dye-encapsulated liposomes.
4-[(2,6-Diaminohexanoylamino)methyl]-3-nitro-N-octadecyl
benzamide (Lys-lipid)
To the di-Boc-lysine derivative 7 (0.38 g, 0.49 mmol), were added
4 mL of trifluoroacetic acid and a drop of anisole. The reaction
mixture was stirred at room temperature for 2 h. The work-up
procedure was the same as described for Asp-lipid. Lys-lipid
was obta◦ined as a light yellow solid (0.28 g, quantitative); mp:
1
130–134 C; H NMR dH (400 MHz; d6-DMSO; 298 K, Me4Si,
exchangeable protons not reported) 8.35 (d, J 1.6, 1H, ArH), 7.89
(dd, J 1.6, 8.0, 1H, ArH), 7.48 (d, J 8.0, 1H, ArH), 4.58–4.48 (m,
3H, ArCH2 and COCH), 3.25 (q, 7.2, 2H, NHCH2), 3.21–3.18 (m,
2H, lysine CH2NH2), 2.50 (t, J 6.8, 2H, lysine CH2), 1.62–1.54 (m,
2H, NHCH2CH2), 1.50–1.42 (m, 2H, lysine CH2), 1.41–1.15 (m,
2H, lysine CH2), 1.15–1.07 (m, 30H, stearyl CH2), 0.73 (t, J 7.2,
3H, stearyl CH3); 13C NMR dC (100 MHz; d6-DMSO; 298 K)
176.63, 164.37, 148.39, 137.97, 135.02, 132.51, 130.30, 123.85,
55.48, 42.01, 35.60, 31.94, 29.66–29.35, 27.11, 23.35, 22.74, 14.58.
Anal. Calcd. for C32H57N5O4.2CF3COOH. H2O: C, 52.61; H, 7.48;
N, 8.52. Found: C, 52.37; H, 7.08; N, 8.37%.
Acknowledgements
This research was supported by the National Institutes of Health
grants 1R15 DK56681-01A1 to D. K. S. and 1P20 RR 15566-01
to S. M.
References
1 V. P. Torchilin, Nat. Rev. Cancer, 2005, 4, 145–160.
2 D. Ppahadjopous, in Liposomes: Rational Design, ed. A. S. Janoff,
Marcel Dekker, New York, 1999, pp. 1–12; T. Ishida, H. Harashima
and H. Kiwada, Biosci. Rep., 2002, 22, 197–224.
Preparation of carboxyfluorescein-encapsulated liposomes
3 R. J. Banerjee, J. Biomater. Appl., 2001, 16, 3–21.
4 A. S. L. Derycke and P. A. M. de Witte, Adv. Drug. Delivery Rev., 2004,
56, 17–30.
The photocleavable lipid (0.45 lmoles, 5 mol%) and solid 1,2-
distearoyl-sn-glycero-3-phosphocholine (6.716 mg, 8.55 lmoles,
This journal is
The Royal Society of Chemistry 2006
Org. Biomol. Chem., 2006, 4, 1730–1740 | 1739
©

View Article Online
5 H. M. Lee and J. Chmielewski, J. Pept. Res., 2005, 65, 355–363; T. L.
Andresen, S. S. Jensen and K. Jorgensen, Prog. Lipid Res., 2005, 44,
68–97.
6 S. L. Huang and R. C. MacDonald, Biochim. Biophys. Acta, 2004,
1665, 134–141; N. Karoonuthaisiri, K. Titiyevskiy and J. L. Thomas,
Colloids Surf., B, 2003, 27, 365–375.
7 S. C. Davis and F. C. Szoka, Bioconjugate Chem., 1998, 9, 783–792.
8 H. D. Han, T. W. Kim, B. C. Shin and H. S. Choi, Macromol. Res.,
2005, 13, 54–61; K. Yoshino, A. Kadowaki, T. Takagishi and K. Kono,
Bioconjugate Chem., 2004, 15, 1102–1109.
9 Z. Li, Y. Wan and A. G. Kutateladze, Langmuir, 2003, 19, 6381–6391;
Y. Wan, J. K. Angleson and A. G. Kutateladze, J. Am. Chem. Soc., 2002,
124, 5610–5611; T. Spratt, B. Bondurant and D. F. O’Brien, Biochim.
Biophys. Acta, 2003, 1611, 35–43. For photo-polymerization induced
liposomal leakage, see: A. Mueller and D. F. O’Brien, Chem. Rev., 2002,
102, 727–757.
O. Wichmann and C. Schultz, Tetrahedron Lett., 2003, 44, 1153–1155;
C. G. Bochet, J. Chem. Soc., Perkin Trans. 1, 2002, 125–142.
15 T. S. Seo, X. Bai, H. Ruparel, Z. Li, N. J. Turro and J. Ju, Proc. Natl.
Acad. Sci. U. S. A., 2004, 101, 5488–5493; X. Bai, S. Kim, Z. Li, N. J.
Turro and J. Ju, Nucleic Acids Res., 2004, 32, 535–541; Z. Li, X. Bai, H.
Ruparel, S. Kim, N. J. Turro and J. Ju, Proc. Natl. Acad. Sci. U. S. A.,
2003, 100, 414–419.
16 Y. V. Il’ichev, M. A. Schwoerer and J. Wirz, J. Am. Chem. Soc., 2004, 126,
4581–4595; B. V. Zemelman, N. Nesnas, G. A. Lee and G. Miesenbock,
Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 1352–1357; J. E. T. Corrie, A.
Barth, V. R. N. Munasinghe, D. R. Trentham and M. C. Hutter, J. Am.
Chem. Soc., 2003, 125, 8546–8554.
17 A. Dussy, C. Meyer, E. Quennet, T. A. Bickle, B. Giese and A. Marx,
ChemBioChem., 2002, 3, 54–60.
18 L. Wang, J. E. T. Corrie and J. F. Wootton, J. Org. Chem., 2002, 67,
3474–3478.
10 T. L. Anderson, J. Davidsen, M. Begtrup, O. G. Mouritsen and K.
Jorgensen, J. Med. Chem., 2004, 47, 1694–1703; X. Guo and F. C.
Szoka, Acc. Chem. Res., 2003, 36, 335–341; J. Davidsen, K. Jorgensen,
T. L. Andersen and O. G. Mouritsen, Biochim. Biophys. Acta, 2003,
1609, 95–101.
11 N. R. Sarkar, T. Rosendahl, A. B. Krueger, A. L. Banerjee, K. Benton,
S. Mallik and D. K. Srivastava, Chem. Commun., 2005, 999–1001.
12 M. Egeblad and Z. Werb, Nat. Rev. Cancer, 2002, 2, 161–174.
13 A. Blanc and C. G. Bochet, J. Am. Chem. Soc., 2004, 126, 7174–7175;
M. C. Pirrung, W. H. Pieper, K. P. Kalippan and M. R. Dhananjeyan,
Proc. Natl. Acad. Sci. U. S. A., 2003, 100, 12548–12553; A. Blanc and
C. G. Bochet, J. Org. Chem., 2003, 68, 1138–1141; K. Schaper, S. A. M.
Mobarekeh and C. Grewer, Eur. J. Org. Chem., 2002, 1037–1046; R.
Wieboldt, D. Ramesh, E. Jabri, P. A. Karplus, B. K. Carpenter and
G. P. Hess, J. Org. Chem., 2002, 67, 8827–8831.
19 M. Wilcox, R. W. Viola, K. W. Johnson, A. P. Billington, B. K.
Carpenter, J. A. McCray, A. P. Guzikowski and G. P. Hess, J. Org.
Chem., 1990, 55, 1585–1589.
20 Z. Zhang, H. Hatta, T. Ito and S. I. Nishimoto, Org. Biomol. Chem.,
2005, 3, 592–596; M. Nomura, S. Shuto and A. Matsuda, Bioorg. Med.
Chem., 2003, 11, 2453–2461; M. P. Hay, G. J. Atwell, W. R. Wilson,
S. M. Pullen and W. A. Denny, J. Med. Chem., 2003, 46, 2456–2466.
21 For preliminary results, see: B. Chandra, S. Mallik and D. K. Srivastava,
Chem. Commun., 2005, 3021–3023.
22 T. Nagasaki, A. Taniguchi and S. Tamagaki, Bioconjugate Chem., 2003,
14, 513–516.
23 H. Komatsu and P. L. G. Chong, Biochemistry, 1998, 37, 107–
115.
24 Spartan 04 for Windows, Wavefunction Inc., Irvine, CA, USA, 2004.
25 H. Feuer, in The Chemistry of the Nitro and Nitroso Groups. Part 1,
Interscience Publishers, Wiley and Sons, NY, 1969.
14 Z. Guan, J. T. Roland, J. Z. Bai, S. X. Ma, T. M. McIntire and
M. Nguyen, J. Am. Chem. Soc., 2004, 126, 2058–2065; Y. Luo and
M. S. Shoichet, Biomacromolecules, 2004, 5, 2315–2323; C. P. Salerno
and H. J. Cleaves, Synth. Commun., 2004, 34, 2379–2386; C. Dinkel,
26 A. A. Frost and R. G. Pearson, 1961, Kinetics and Mechanism, Wiley,
New York.
27 H. Newman, J. Org. Chem., 1965, 30, 1287–1288.
1740 | Org. Biomol. Chem., 2006, 4, 1730–1740
This journal is
The Royal Society of Chemistry 2006
©