A General Approach to Enzyme-Responsive Liposomes

Article abstract of DOI:10.1002/chem.202000529

Liposomes are effective nanocarriers due to their ability to deliver encapsulated drugs to diseased cells. Nevertheless, liposome delivery would be improved by enhancing the ability to control the release of contents at the target site. While various stimuli have been explored for triggering liposome release, enzymes provide excellent targets due to their common overexpression in diseased cells. We present a general approach to enzyme-responsive liposomes exploiting targets that are commonly aberrant in disease, including esterases, phosphatases, and β-galactosidases. Responsive lipids correlating with each enzyme family were designed and synthesized bearing an enzyme substrate moiety attached via a self-immolating linker to a non-bilayer lipid scaffold, such that enzymatic hydrolysis triggers lipid decomposition to disrupt membrane integrity and release contents. Liposome dye leakage assays demonstrated that each enzyme-responsive liposome yielded significant content release upon enzymatic treatment compared to minimal release in controls. Results also showed that fine-tuning liposome composition was critical for controlling release. DLS analysis showed particle size increases in the cases of esterase- and β-galactosidase-responsive lipids, supporting alterations to membrane properties. These results showcase an effective modular strategy that can be tailored to target different enzymes, providing a promising new avenue for advancing liposomal drug delivery.

Full text of DOI:10.1002/chem.202000529

Accepted Article
Accepted Article
Title: A General Approach to Enzyme-Responsive Liposomes
Authors: Michael D Best and Jinchao Lou
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202000529
Link to VoR: https://doi.org/10.1002/chem.202000529
01/2020

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
A General Approach to Enzyme-Responsive Liposomes
Jinchao Lou [a] and Michael D. Best *[a]
Dedicated to Professor Eric V. Anslyn on the occasion of his 60th birthday
[a]
J. Lou, Prof. M. D. Best
Department of Chemistry
University of Tennessee
1420 Circle Drive, Knoxville, TN, 37996 (USA)
E-mail: mdbest@utk.edu
Supporting information for this article is given via a link at the end of the document.
Abstract: Liposomes are effective nanocarriers due to their ability to
deliver encapsulated drugs to diseased cells. Nevertheless, liposome
delivery would be improved by enhancing the ability to control the
release of contents at the target site. While various stimuli have been
explored for triggering liposome release, enzymes provide excellent
targets due to their common overexpression in diseased cells. We
ideal for maximizing activity.[8] Therefore, controlling the timing
and location of cargo release is a key point in advancing liposomal
drug delivery.
In modern liposome research, different strategies for
achieving control over drug release have been reported, which
can be divided into two main categories: passive release (by
internal stimuli) and active release (by external stimuli).[8-9] For
passive release, variations in biological conditions between
healthy and diseased cells have been used to differentiate release,
such as pH profiles,[10] redox environment,[11] metabolites[12] and
enzyme expression.[13] For active release, external stimuli
including light,[14] heat,[15] and ultrasound[16] have been
investigated for disrupting liposomes and cause encapsulated
cargo release. Though a variety of triggered release approaches
have been pursued, many challenges remain to be overcome in
order to develop clinically applicable controlled release strategies.
Active release protocols commonly suffer from challenges
associated with selective delivery of stimuli to diseased cells. For
example, while light-initiated liposomal release has been
extensively studied, this typically entails irradiation with UV light,
for which disadvantages include limited tissue penetration and
damage to healthy tissue due to photodecomposition of
biomolecules.[17] Considering passive release, minimal variations
between diseased and healthy cells often offers only a narrow
window for differentiation. For example, for pH-triggered release,
reported pH differences between normal cells and cancer cells
are quite small (6.5-6.9 for cancer cells and 7.2-7.4 for normal
tissue),[18] which makes it difficult to develop systems that respond
to these specific pH variations.
Among the potential stimuli for triggered release, the
targeting of enzymes is particularly promising due to the
significant overexpression of enzyme abundance commonly
associated with diseased cells. Despite this potential, the
development of enzyme-responsive liposomes has received less
attention than other strategies, and has been limited to a small
group of enzyme targets.[19] Most previously reported enzyme-
responsive systems have targeted matrix metalloproteinase
(MMP)[13a,13b], cholinesterase[13c] or phospholipase enzymes.[20]
The latter approach has exploited the fact that phospholipase
enzymes directly modify lipids that compose the liposome. As an
example, the Andresen group[20a] developed a class of prodrugs
incorporated within lipid scaffolds, for which drug release is driven
by lipid hydrolysis catalyzed by secretory phospholipase A2
present
a general approach to enzyme-responsive liposomes
exploiting targets that are commonly aberrant in disease, including
esterases, phosphatases and β-galactosidases. Responsive lipids
correlating with each enzyme family were designed and synthesized
bearing an enzyme substrate moiety attached via a self-immolating
linker to a non-bilayer lipid scaffold, such that enzymatic hydrolysis
triggers lipid decomposition to disrupt membrane integrity and release
contents. Liposome dye leakage assays demonstrated that each
enzyme-responsive lipid yielded significant content release upon
enzymatic treatment compared to minimal release in controls. Results
also showed that fine-tuning liposome composition was critical for
controlling release. DLS analysis showed particle size increases in
the cases of esterase- and β-galactosidase-responsive lipids,
supporting alterations to membrane properties. These results
showcase an effective modular strategy that can be tailored to target
different enzymes, providing a promising new avenue for advancing
liposomal drug delivery.
Introduction
Liposomes are spherical structures formed by lipid self-
assembly that have proven to be effective for encapsulating and
delivering a wide variety of drugs with different properties in order
to enhance therapeutic attributes.[1] Comparing to free drugs,
liposomal delivery systems exhibit advantages including
protection against degradation, reduction of drug toxicity,[2] and
side effects[3] as well as optimization of pharmacokinetic
properties[4] and therapeutic indices of drugs.[5] As a result,
various designer liposome platforms have been intensely studied
in recent years in order to further enhance delivery
characteristics.[6] Indeed, there are approximately fifteen
liposomal drugs currently approved by FDA that are commercially
available, including the Doxil/Caelyx (Johnson & Johnson)
formulation for the delivery of the anti-cancer drug doxorubicin,
and AmBisome (Gilead) encapsulating amphotericin B to treat
fungal infections.[7] Drug release from clinically used formulations
currently relies upon liposome carrier breakdown, which is not
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
A.
(sPLA2). These prior studies support the hypothesis that
enzymatic reactions can be harnessed to drive liposome release.
However, there are many other enzymes that are significantly
overexpressed in different types of cancerous and otherwise
diseased cells that provide exciting targets.
Enzyme
hydrolysis
Liposome
destabilization
and release
In this work, we report a general, modular strategy for
developing liposomal drug delivery platforms that is effective for
targeting a range of enzymes that exhibit dramatic upregulation in
diseases such as cancer, including esterases, phosphatases, and
β-galactosidases. Esterase concentrations have been reported to
be enhanced by two to three orders of magnitude in cancer
cells,[21] and these enzymes have been implicated in other
diseases such as neuronal,[22] liver[23] and Alzheimer’s
diseases.[24] Overabundance of multiple phosphatase enzymes,
such as alkaline phosphatase, has also been correlated with
cancer[25] as well as liver,[26] cardiovascular[27] and kidney
diseases.[28] β-Galactosidases are also overexpressed in cancer,
and thus have been targeted for cancer imaging.[29] Glycosidase
enzymes in general are involved in numerous diseases including
diabetes, cancer, viral infections such as HIV, and lysosomal
storage disorders.[30] These particular enzyme targets have been
the subject of few studies. In a rare example, the ganglioside GM1
has been reported for triggered release using β-galactosidase
based on the modulation of membrane properties upon truncation
of the carbohydrate head group.[31] While this is an elegant
strategy, it relies upon a complex glycolipid structure for release.
In addition, the Szoka group developed a cholesterol-based
phosphate lipid analog, which was able to stabilize the DOPE
bilayer.[13d] Addition of alkaline phosphatase and hydrolysis of the
phosphate head group disrupted the bilayer and caused
encapsulated cargo release.
key
= Enzyme-responsive Lipid
= Hydrophilic cargo
= Hydrophobic cargo
= Trigger/Enzyme Substrate
= Self-immolating Linker
= Non-bilayer Forming Lipid
= Bulk lipid
B. Structures of Enzyme-Responsive Lipids
Galactose
substrate/trigger
Ester or phosphate
substrate/trigger
R
OH
O
O
O
H
HO
OH
O
O
N
N
P
H
O
O
O
O
O
O
O
OH
O
O
15
15
SIL = Quinone methide (QM)
O
O
SIL = Trimethyl lock (TML):
Very fast kinetics for
lactonization
7
7
GRL
7
7
O
R =
R =
(ester trigger)
ERL
PRL
or
O
(phosphate
trigger)
P
OH
O
Scheme 1. Design of enzyme-responsive liposomes. A. Cartoon depicting
hypothetical model for liposome release. Enzyme-responsive lipids contain three
major regions including enzyme substrate head group, SIL and a non-bilayer lipid
scaffold. After hydrolysis of the substrate by the appropriate enzyme, SIL
decomposition will cause the release of a non-bilayer lipid to disrupt the
membrane integrity and release of encapsulated cargo. B. Structures of enzyme-
responsive lipids. Esterase- and phosphatase-responsive lipids (ERL and PRL,
respectively) include TML as the SIL and DOPE as the lipid scaffold. β-
Galactosidase-responsive lipid GRL instead bears a QM-generating SIL and an
aminodialkylglycerol analogue as the lipid scaffold.
Results and Discussion
that will destabilize the membrane and trigger release of liposome
contents. The lipids dioleoylphosphatidylethanolamine (DOPE)
and aminodialkylglycerol were employed for this purpose, with the
latter selected to avoid synthetic challenges associated with
introduction of the sugar head group of GRL. It has been well-
studied that due to the small head-to-tail volume ratio, lipids such
as DOPE typically prefer to form hexagonal phase (HII) lipid
assemblies in aqueous solution at physiological conditions.[36]
However, increasing the head-to-tail volume ratio through N-
acylation with a bulky group results in the self-assembly of
resulting lipids into stable membrane bilayers.[37] Therefore, the
release of DOPE has previously been harnessed for liposome
release strategies. For example, the Smith group reported
photocleavable liposomes using a light-responsive lipid analogue
that generates DOPE upon UV light irradiation resulting in release
of the entrapped dye calcein.[14a] In an example more closely
related to this work, McCarley and co-workers[11a] developed
redox-responsive liposomes by coupling DOPE to a quinone
redox switch using TML as a SIL. Addition of Na2S2O4 initiated the
self-immolating process of the quinone head group and the
release of DOPE, which ultimately perturbed the membrane and
released the loaded dye calcein. Putting all of these groups
together, our systems are designed such that enzymatic removal
of trigger groups will stimulate decomposition of the SIL to
produce non-bilayer lipids that disrupt the membrane and release
encapsulated cargo.
Herein, we report a versatile stimuli-responsive lipid design
by exchanging appended substrate moieties to target multiple
enzymes that are commonly overexpressed in cancer cells. The
design strategy is shown in Scheme 1. Each lipid analog contains
three functional regions. First, a variable substrate moiety
corresponding to each target enzyme is included that acts as the
trigger for liposome release. This includes an ester that can be
cleaved by an esterase (esterase-responsive lipid (ERL)), a
phosphate group that is hydrolyzed by
(phosphatase-responsive lipid (PRL)) or a β-galactose moiety for
enzymatic cleavage by β-galactosidase (galactosidase-
a
phosphatase
a
responsive lipid (GRL)). Secondly, a self-immolating linker (SIL)
is present in between the trigger head group and the lipid scaffold
to be released, a strategy that has been used in stimuli-
responsive systems including sensors[32] and nanomedicine.[33]
Upon trigger removal, the SIL is designed to quickly undergo a
disassembly reaction that results in release of an appended
leaving group. Specifically, the SILs for our responsive lipids
include the trimethyl lock (TML, o-hydroxydihydrocinnamic
acid)[34] for ERL and PRL or the quinone methide (QM)-
generating 4-hydroxylmethylphenol group for GRL,[35] both of
which are well known for fast kinetics of decomposition upon
trigger removal.
Finally, these responsive lipids are designed to form stable
liposome membranes that are perturbed upon enzymatic removal
of the trigger, in this case through the release of a non-bilayer lipid
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
A. Synthetic Route for Esterase-Responsive Lipid ERL
release. Prior to incorporating this compound into liposomes, we
set out to evaluate esterase hydrolysis of the free lipid using a
TLC assay in a manner similar to a previous report.[13e] For this
experiment, 2 µg of ERL was dissolved in 50 µL of TBS buffer and
was then incubated with 0.45 U of commercially available porcine
liver esterase at room temperature for three hours. After this, the
reaction mixture was spotted on a silica gel TLC plate, which was
run with 20% MeOH-chloroform as eluant and visualized with
potassium permanganate stain.[41] As is shown in Figure S1, this
plate indicated complete disappearance of the ERL spot in the
product coupled with the appearance of a new spot attributed to
DOPE. This result indicates that ERL acts as an appropriate
substrate and provided evidence that the esterase hydrolyzed
only the TML head group while leaving the DOPE fatty acid ester
chains intact. In addition, the catalytic activities of phosphatase
and β-galactosidase enzymes were confirmed via colorimetric
assays using p-nitrophenylphosphate and p-nitrophenylgalactose,
respectively (Figure S2).
With evidence that ERL acts as a substrate for esterase
enzyme, we next moved on to evaluate liposome triggered
release properties using fluorescence-based dye leakage assays.
These assays can be used to evaluate the release of either
hydrophobic or hydrophilic dyes encapsulated within the
membrane bilayer or aqueous core, respectively, as judged by
changes in dye emission properties. The first assay we pursued
utilized Nile red (NR), which is a hydrophobic dye that can mimic
common non-polar drugs. NR is widely used in cell biology
studies[42] since it only fluoresces in aqueous media when it is
solubilized through encapsulation within membrane bilayers.
Therefore, following triggered release into aqueous media, NR
will no longer fluoresce due to precipitation,[43] and the resulting
decrease in fluorescence intensity can be used to track release.
We incorporated responsive lipids at varying percentages into
liposomes primarily composed of bulk lipids including
phosphatidylcholine (PC) and DOPE. Other lipid additives such
as 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP) and
phosphatidic acid (PA) were also included to probe their effects
on liposome membrane stability before and after enzyme
treatment. Unilamellar liposomes were prepared using standard
thin-film hydration methods including lipid film preparation,
hydration, freeze-thaw cycles and extrusion through 200 nm
polycarbonate membranes. Dynamic light scattering (DLS)
experiments were performed to verify the successful formation of
the stable liposome vesicles, as will be later discussed.
O
O
O
O
O
O
DOPE, EDCI, HOBt, DIEA
O
O
O
P
N
H
O
O
DMF
OH
OH
O
H
92%
O
1
ERL
1)
B. Synthetic Route for Phosphatase-Responsive Lipid PRL
O
BnO
N
P
O
P
O
TBDMSCl
DMAP
THF
OH
OH
BnO
OBn
O
OH
OBn
tetrazole
2) m-CPBA
O
conc. H2SO4
LiAlH4
THF
O
Si
O
OH
reflux
O
Si
93%
82%
94%
95%
2
3
4
5
O
OR
O
P
P
O
KF
O
BnO
OBn
OR
O
O
P
OH
Jones Reagent
DOPE, HOBt, EDCI
OH
O
O
N
H
O
O
acetone
DMF
O
O
H
87%
77%
O
6
1. TMS-I, DCM
R = Bn (7)
2. MeOH/H2
O
R = H (PRL)
60%
C. Synthetic Route for Galatosidase-Responsive Lipid GRL
O
O
O
O
O
O
O
O
O
O
HO
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
Br
O
O
HBr, AcOH
NaBH4
O
O
O
O
O
O
O
O
O
O
Ag2O, MeCN
CHCl3/i-PrOH
DCM
O
Quant.
77%
96%
O
O
8
9
10
O
H
HO
O
O
O
OH
O
O
O
O
H2
N
HO
OH
O
O
O
O
O
O
O
O
O
O
O
O
O
NO2
O
O
O
15
13
15
OH
O
NaOMe
MeOH
90%
Cl
O
O
O
O
O
Pyridine, EtOAc
TEA, DMF
O
N
H
93%
40%
O
N
H
O
O
O
O
O
O
11
12
GRL
15
15
15
15
NO2
Scheme 2. Synthetic routes developed to access ERL (A) PRL (B) and GRL (C).
The synthetic routes used to access the three enzyme-
responsive lipids are shown in Scheme 2. The synthesis of ERL
benefitted from commercially available esterified TML-carboxylic
acid 1, which allowed for convenient access to this product
through a one-step amide-bond coupling reaction with DOPE
(Scheme 2A). To access PRL, compounds 2-4 were synthesized
from 3,5-dimethylphenol as previously reported[38] through a
Michael addition with methyl 3-methylbut-2-enoate, spontaneous
lactonization to 2, reductive ring opening to 3, and silyl protection
to 4. Phosphoramidite chemistry was next performed to produce
the phosphodiester of 5, followed by Jones oxidation and
simultaneous silyl deprotection to afford 6.[38a] Finally, an amide
coupling reaction with DOPE yielding 7 was followed by benzyl
deprotection to produce PRL. For GRL, the design and synthetic
route were different, as shown in Scheme 2C. As discussed
above, to overcome synthetic issues associated with the use of
acetyl protecting groups, 4-hydroxy benzyl alcohol was instead
used as the SIL and dialkylglycerol as the non-bilayer forming lipid
scaffold. Compounds 8-11 were synthesized from β-D-galactose
pentaacetate as previously reported by Toth et al[39] through HBr
treatment to form ꢀ-D-galactopyranosyl bromide (8), glycosylation
with 4-hydroxyl benzaldehyde to produce 9, aldehyde reduction
to 10, and conversion of the resulting alcohol to p-nitrophenyl
carbonate 11. Dialkyl aminoglycerol lipid 13 was synthesized from
solketal in five steps as we previously reported[40] through tosylate
introduction, acetonide deprotection, substitution of tosylate with
azide, Williamson ether synthesis to introduce lipid alkyl chains,
and reduction of azide to the primary amine 13 (not shown).
Amine 13 and carbonate 11 were combined to produce the
carbamate moiety of protected lipid 12, followed by deprotection
of the acetyl groups on galactose access GRL.
During the course of these studies, we systematically varied
lipid composition to analyze liposomes containing a wide range of
percentages of different mixtures of PC, DOPE, DOTAP and PA.
The results of these studies are summarized in Table S1. Many
of these liposome formulations were found to not release contents
or to not be sufficiently stable before enzyme treatment, indicating
that the stability of the liposome before and after enzymatic
release has to be carefully fine-tuned. Our initial success in
esterase-responsive liposome release was achieved using
liposomes composed of PC, DOPE, and DOTAP. DOTAP is a
cationic lipid that is widely used in transfection,[44] and has been
reported to render membranes more fusogenic, which is why we
began introducing this compound into liposomes in an effort to
induce cargo release.[13e] After screening a variety of different lipid
mixtures, liposomes composed of 30% ERL, 50% DOPE, 10%
PC and 10% DOTAP treated with 0.45 U of porcine liver esterase
and incubated in a 30 °C water bath in between measurements
Following the successful synthesis of all three lipids, we next
tested their enzyme-responsive properties. Considering the
accessibility of ERL compared to the other two lipids, we initiated
studies with this compound to determine conditions for liposomal
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
Figure 1. Release of NR over time from different formulations of enzyme-responsive liposomes. A. Data for ERL liposomes composed of
30% ERL, 50% DOPE and varying percentage of PC or DOTAP addition. B. Data for ERL-containing liposomes containing PA. A 30/10/50/10
ratio of ERL/PC/DOPE/PA yielded ~60% release over time upon esterase treatment. C. Data for PRL-containing liposomes. A 30/15/45/10
ratio of PRL/PC/DOPE/PA gave ~50% release overtime upon phosphatase treatment. D. Data for GRL-containing liposomes. A 30/10/50/10
ratio of GRL/PC/DOPE/PA resulted in ~60% release overtime upon galactosidase treatment. Control experiments involving treatment with
buffer or heat-denatured enzyme, along with removal of responsive lipid, did not show significant fluorescent decrease in all cases. Error bars
indicate standard errors from at least three studies.
showed an ~30% decrease in fluorescent intensity upon esterase
treatment within 40 minutes (Figure 1A and Table S1, Entry 1).
We have previously found that heating of liposomes can be
necessary to achieve effects such as liposome fusion.[45] Different
control sets were also tested using the exact same liposome
formulation by treating with only buffer or with heat-denatured
enzyme, leading to diminished fluorescence changes
(background release ranging from ~8-12%, Figure 1A). Similar
results were also observed for liposomes lacking ERL in which
this lipid was replaced by PC. To drive home the sensitive nature
of liposome composition, we will note that liposomes containing
30% ERL, 50% DOPE, 15% PC and 5% DOTAP showed almost
the same fluorescent intensity decrease (~10%) as the control
sets, indicating that trading only 5% of DOTAP for PC was
sufficient to stabilize the membrane and shut down release of
encapsulated dye. On the other hand, liposomes containing 30%
ERL, 50% DOPE, 5% PC and 15% DOTAP only showed ~15%
release. In this way, the liposome composition needed to be
carefully fine-tuned in order to optimize release properties.
PA). NR release curves for these experiments are shown in
Figure 1B. Upon treatment with esterase and heating at 30 °C in
between measurements, the fluorescence intensity gradually
decreased by ~60%, although this process took a significantly
longer time (approximately five hours) to run to completion.
Control experiments were also done as before, either treating the
exact same liposome solution with buffer or heat denatured
esterase or by treating liposomes in which ERL was replaced by
PC with esterase. All of these, again, showed only minimal
background release. We will also note that the percent change in
fluorescence may not indicate the total percentage of dye release,
for example since the released hydrophobic dye may be re-
encapsulated into other lipid assembly structures after being
initially released. These results indicate that ERL enables
effective enzyme-responsive liposome properties for the release
of hydrophobic cargo.
To pursue our goal of developing a general strategy that could
be adapted to target different enzymes, we next evaluated
triggered release conditions driven by PRL and GRL. For the
former, we again tested different formulations (Table S2) and
arrived at an effective composition of 30% PRL, 15% PC, 45%
DOPE and 10% PA. After adding 0.35 U alkaline phosphatase
from Escherichia coli and incubating at 30 °C in between
measurements, a fluorescence intensity decrease of ~50% was
observed after about thirteen hours, as is shown Figure 1C.
Interestingly, this release curve appeared to exhibit an induction
The sensitive nature of the DOTAP liposomes led us to next
explore another lipid additive, PA. PA is an ionic lipid that is known
to exaggerate the non-bilayer properties of DOPE within
membranes.[46] After evaluating different lipid percentages (Table
S1, Entries 15-17), we identified that the formulation in which
DOTAP was simply replaced by PA showed significantly
enhanced activity (30% ERL, 10% PC, 50% DOPE and 10% egg
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
Figure 2. DLS results from enzyme-responsive liposomes used in NR release experiments. Data for ERL-containing liposomes (A) and GRL-containing liposomes (C) yielded significant increases in
average particle sizes upon esterase or β-galactosidase treatment, respectively. For PRL-containing liposomes (B), minimal changes were observed under any conditions. Control experiments in all
three cases did not show any significant change. Error bars indicate standard errors from at least three studies.
period of approximately 5 hours before a steeper drop in
fluorescence intensity. This result was only observed with PRL,
and since this is the only anionic enzyme-responsive lipid tested,
it is possible that the altered charge of the membrane affects
enzymatic modification and/or release properties. Also, as
mentioned above, a key point we encountered with each of these
responsive systems is that initial and final stability of the
membrane must be carefully tuned to maximize release. In this
case, this was done by dropping the percentage of DOPE from
50% for ERL to 45% for PRL, which was perhaps needed to
compensate for the introduction of the negatively charged PRL
lipid. GRL exhibited release properties that were quite similar to
ERL, despite being composed of entirely different structural
attributes (altered enzymatic substrate, SIL and lipid scaffold that
is released). Again, different formulations were screened (Table
S3), and a 30/10/50/10 ratio of GRL/PC/DOPE/PA resulted in an
~60% decrease in fluorescent intensity within nine hours upon
addition of β-galactosidase from Escherichia coli (Figure 1D). For
both PRL and GRL, multiple control experiments were again
carried out. Liposomes containing responsive lipids didn’t show
significant release upon treatment with either buffer or heat-
denatured enzyme addition. Similar results were again observed
for liposomes in which responsive lipids were replaced by PC and
treated with enzyme. These dye release results showcase that
these enzyme-responsive liposome systems are effective for
controlling the release of encapsulated contents in a manner
driven by enzyme treatment. These also indicate our design
indeed provides a general strategy for targeting different enzymes
by simply exchanging the enzymatic substrate trigger
incorporated into the head groups of enzyme-responsive lipid
structures. The stabilities of liposomes formed by incorporating
each of these three responsive lipids as well as PA in the
formulation were separately tested via tracking encapsulated Nile
red fluorescence over time (Figure S3). All of them were stable for
at least three days when being kept at 4 °C, which is common of
stabilized liposomal nanoparticles.
expected for particles prepared via extrusion by passing through
membranes of 200 nm. This indicates that these formulations
containing responsive lipids form stable liposomes. As can be
seen in Figure 2A as well as Figure S4, incubation of ERL
containing liposomes with esterase resulted in a dramatic
increase in average particle size. In the corresponding control
experiments, liposomes containing ERL but instead treated with
TBS buffer or liposomes without ERL subjected to esterase only
showed minimal change. These results are in line with previous
stimuli-responsive liposomes we have developed,[40b,47] and
indicate that liposomes underwent structural changes upon
esterase treatment. Possible explanations include that lipids
adopt different assembly properties after enzyme modification
such as the inverted hexagonal phase, which is favored by DOPE
at physiological conditions.[36] Considering the complicated lipid
composition, another reasonable explanation would be that after
enzyme treatment, the membrane was destabilized, ultimately
leading to fusion and concomitant release of encapsulated cargo.
Liposomes containing GRL yielded comparable increases in
average particle sizes after adding β-galactosidase and null
results in controls (Figure 2C and Figure S5), suggesting that this
system underwent alterations to lipid self-assembly properties
that were similar to ERL liposomes. However, somewhat
surprisingly, PRL liposomes did not yield any significant size
changes in any of these experiments, suggesting that PRL
containing liposomes did not undergo significant changes in lipid
assembly properties (Figure 2B and Figure S6). While PRL
liposomes did exhibit NR release, the response curve was unique
from the others since it showed an induction period and it
ultimately yielded less of
a decrease in percentage of
fluorescence. Again, the most significant difference in this
compound is the presence of the charged phosphate group, which
may impact either liposome membrane packing or the interaction
of phosphatase enzyme with the liposomes. Nevertheless, we
have previously observed similar results in which a light-triggered
release liposomal platform yielded significant NR released but no
size changes were detected via DLS.[14d]
A benefit of the unique structure of liposomes is that they are
quite versatile in being able to encapsulate both hydrophobic
cargo within the membrane bilayer but also hydrophilic contents
within the aqueous core. The ability of such liposomal platforms
to deliver hydrophilic molecules is also important for polar
therapeutics such as siRNA. However, this is usually more
challenging as this requires polar contents to escape through the
hydrophobic bilayer region to ultimately achieve release. As a
In order to understand potential structural changes to self-
assembled lipid structures caused by the introduction of
enzymatic stimuli, we next performed DLS studies using each
system. For these experiments, liposome formulations that
yielded optimal NR release results in the prior studies were
analyzed by DLS before and after enzyme treatment, with results
shown in Figure 2. In all cases, the original liposome samples
prior to enzyme treatment showed uniform vesicle sizes with
average diameters ranging from 150 nm – 200 nm, which is
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
lipid scaffold. These novel lipids were incorporated into liposomes,
and after careful fine-tuning of membrane stability based on
composition, each compound was found to be successful for
achieving triggered release of the hydrophobic dye NR upon
treatment with the appropriate enzyme. These results showcase
that this approach provides a general strategy for enzyme-
responsive liposomes in which different enzymes can be targeted
by modifying the substrate trigger displayed at the lipid headgroup.
However, despite the generality of this approach, there were
differences observed in release properties based on structural
nuances. In particular, PRL was found to exhibit NR release
curves (less release and induction period) and DLS results (no
observed change) that were different from ERL and GRL, despite
the fact that the structures of ERL and PRL only differ by one
functional group, and indeed they are expected to produce the
same lipid product upon release. This exception indicates that
response properties can vary based on changes to the structures,
in this case most likely due to the charge of the resulting enzyme-
responsive liposomes. Finally, ERL-containing liposomes also
showed ability to release hydrophilic cargo. Enzymes provide
exciting targets for controlling release of encapsulated drugs from
nanocarriers using stimuli-responsive materials since these are
commonly overexpressed in many diseases. However, the
existence of numerous families of enzymes that catalyze wide-
ranging reactions by which varying substrates are modified
provides a grand challenge for exploring different potential
therapeutic targets. The modular strategy reported herein
provides an efficient approach to overcoming this barrier as an
initial step for advancing enzyme-responsive liposomal
therapeutics.
Figure 3. SRB release data over time upon esterase addition to liposomes containing 30%
ERL, 50% DOPE, 10% PC and 10% PA or control experiments of buffer treatment or
liposomes lacking ERL. Only liposomes containing ERL with esterase treatment showed
significant fluorescence increases. Error bars indicate standard errors from at least three
studies.
result, significant membrane disruption is needed to induce
hydrophilic cargo release. In our studies, we selected
Sulforhodamine B (SRB) to evaluate polar dye release because it
is a water soluble fluorescent dye commonly used in release[10b]
and cytotoxicity assays.[48] The concentration of SRB used for
encapsulation was carefully chosen to make sure it is high enough
for fluorescence to be self-quenched due to collisional effects
when entrapped within liposomes. Therefore, when the dye is
released from liposomes, it will be diluted leading to
a
fluorescence turn-on effect. When preparing these liposomes,
similar methods were used that were modified by the addition of
purification through a size exclusion column (SEC), which is
needed after extrusion to remove unencapsulated dye. The
successful preparation of liposomes was again determined by
DLS.
Experimental Section
To assess the hydrophilic cargo release properties of
esterase-responsive liposomes, SRB was encapsulated within
liposomes containing 30% ERL, 10% PC, 50% DOPE and 10%
PA as well as control liposomes containing 30% DPPC, 10% egg
PC, 50% DOPE and 10% PA. After treating with esterase and
incubating in a 30 °C water bath for 17 hours when not performing
measurements, the detergent triton X-100 was added into the
General experimental
Reagents and solvents were generally purchased from Acros, Sigma-
Aldrich, or Fisher Scientific and used without further purification. PC (L-α-
Phosphatidylcholine, mixed isomers from chicken eggs), PA (L-α-
phosphatidic acid sodium salt from chicken eggs), DOPE (1,2-dioleoyl-sn-
glycero-3-phosphoethanolamine), DPPC (1,2-dipalmitoyl-sn-glycero-3-
phosphocholine) and DOTAP (1,2-dioleoyl-3-trimethylammonium-propane,
chloride salt) were purchased from Avanti Polar Lipids, Inc (Alabaster, AL).
Esterase from porcine liver was purchased from Sigma-Aldrich (lyophilized
powder, ≥15 units/mg solid, SKU: E3019). β-Galactosidase from
Escherichia coli was purchased from Abnova (in 1.6 M ammonium sulfate,
Catalog #: P5270). Alkaline phosphatase from Escherichia coli was
purchased from Sigma-Aldrich (in 2.5 M ammonium sulfate, SKU: P4252).
β-Galactosidase substrate, 4-nitrophenyl β-D-galactopyranoside was
purchased from Carbosynth. Phosphatase substrate, 4-nitrophenyl
phosphate disodium salt hexahydrate (pNPP) was purchased from Sigma-
Aldrich (5 mg tablet, SKU S0942). Compounds 4,[38] 11,[39] and 13[40] were
synthesized as previously reported with matching characterization data.
Dry solvents were obtained from a Pure Solv MD-7 solvent purification
system purchased from Innovative Technology, Inc (Newburyport, MA).
Column chromatography was performed using 230-400 mesh silica gel
purchased from Sorbent Technologies. NMR spectra were obtained using
Varian 300 MHz, 500 MHz or 600 MHz spectrometers. Mass spectra were
obtained with JEOL DART-AccuTOF and Waters Synapt G2-Si mass
spectrometers (Milford, MA). Liposome extruder and polycarbonate
membranes were obtained from Avestin (Ottawa, Canada). Ultrapure
water was purified via a Millipore water system (≥ 18 MW·cm triple water
purification system). Small quantities (< 5 mg) were weighed on an
OHRUS analytical-grade mass balance. Fluorescence studies were
liposome solution as
a measure of the total release of
encapsulated contents. The normalized results are reported for
each point as a percentage of fluorescent increase compared to
Triton X-100 treatment. As is shown in Figure 3, liposomes
containing ERL exhibited a significant increase in fluorescence
compared to control experiments in which liposomes containing
ERL were instead treated with buffer or liposomes without ERL
were subjected to enzyme, which both showed minimal
fluorescence change. These results showcase that enzyme-
responsive systems are also capable of releasing hydrophilic
contents.
Conclusion
In conclusion, we have designed and synthesized three
enzyme-responsive lipids targeting unique enzyme families that
are commonly aberrant in disease, including esterase,
phosphatase and β-galactosidase. These lipids shared similar
design strategies including an enzyme substrate/trigger at the
head group, a self-immolating linker and a non-bilayer forming
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
performed using a Cary Eclipse Fluorescence Spectrophotometer from
Agilent Technologies. Plots were generated using Origin Pro 2018. All
error bars in plots show the standard errors of at least three experimental
replicates.
was filtered through a column packed with florisil and the filtrate was
concentrated under reduced pressure. The crude was loaded onto silica
gel and subjected to column chromatography. Gradient elution from DCM
to 5% MeOH-DCM containing 1 drop of acetic acid was used to obtain 6
as a yellow oil. (0.31 g, 0.64 mmol, 87% yield). Rf=0.4 (5% MeOH-
DCM)/0.625 (50% EtOAc-hexane). 1H NMR (300 MHz, Chloroform-d) δ
7.31 (t, J = 1.8 Hz, 10H), 7.05 (d, J = 1.9 Hz, 1H), 6.72 – 6.68 (m, 1H), 5.12
(s, 2H), 5.09 (s, 2H), 2.89 (s, 2H), 2.50 (s, 3H), 2.13 (s, 3H), 1.59 (s, 6H).
13C NMR (126 MHz, cdcl3) δ 150.13, 150.08, 138.72, 136.68, 135.53,
135.48, 132.08, 132.02, 131.56, 128.77, 128.71, 128.18, 128.17, 119.13,
119.11, 70.17, 70.12, 47.50, 39.39, 31.79, 29.84, 25.69, 20.43. 31P NMR
(202 MHz, cdcl3) δ -6.77. HRMS-DART: [M+H]+ calcd for C27H32O6P,
483.1936, found 483.1644
Synthesis
ERL. DOPE (100 mg, 0.1344 mmol), Compound 1 (42.63 mg, 0.1613
mmol) and hydroxybenzotriazole (HOBt, 24.7 mg, 0.1613 mmol) were
added into a small vial under N2. Then, 1 mL dry DMF was added into the
vial and the reaction mixture was cooled to 0 °C. After being stirred for 5
min, 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDCI, 30.92 mg,
0.1613 mmol) and N,N-Diisopropylethylamine (DIEA, 55 μL, 0.336 mmol)
were added. The reaction was then stirred at rt for 5 h before being
quenched by pouring into 100 mL 1 M HCl. The aqueous layer was
extracted three times with 25 mL chloroform. The combined organic layer
was washed with 20 mL water five times and once with 20 mL brine. After
being dried over Na2SO4, filtered and concentrated, the crude was
subjected to column chromatography using gradient elution from 100%
chloroform to 20% methanol-chloroform, which yielded ERL as a colorless
oil (123 mg, 0.124 mmol, 92 % yield). Rf=0.48 (10% methanol-chloroform).
1H NMR (300 MHz, Chloroform-d) δ 6.80 (s, 1H), 6.55 (d, J = 2.0 Hz, 1H),
5.37 – 5.29 (m, 4H), 5.20 (s, 1H), 4.36 (d, J = 11.8 Hz, 1H), 4.11 (dd, J =
12.2, 7.2 Hz, 1H), 3.89 (s, 2H), 3.68 (s, 2H), 3.33 (d, J = 10.6 Hz, 2H), 2.52
(d, J = 25.0 Hz, 5H), 2.33 – 2.23 (m, 8H), 2.21 (s, 3H), 2.00 (q, J = 6.6 Hz,
8H), 1.56 (s, 10H), 1.37 – 1.20 (m, 46H), 0.91 – 0.84 (m, 6H). 13C NMR
(75 MHz, cdcl3) δ 173.70, 173.32, 171.29, 149.50, 138.23, 136.49, 133.45,
132.50, 129.89, 129.51, 129.49, 123.11, 70.24, 63.62, 62.48, 39.69, 39.39,
34.06, 33.91, 31.76, 31.28, 29.60, 29.37, 29.17, 29.01, 27.06, 27.03, 24.75,
22.52, 13.84. 31P NMR (121 MHz, cdcl3) δ 1.46. ESI-MS: [M-H]- calcd for
C56H95NO11P, 988.6643, found 988.6609.
(2R)-3-(((2-(3-(2-((bis(benzyloxy)phosphoryl)oxy)-4,6-
dimethylphenyl)-3-
methylbutanamido)ethoxy)(hydroxy)phosphoryl)oxy)propane-1,2-
diyl dioleate (7)
DOPE (29.9 mg, 0.04 mmol) and HOBt (7.35 mg, 0.048 mmol) were
dissolved in 0.8 mL of dry DMF containing 6 (23.16 mg, 0.048 mmol) in a
small vial. The reaction was cooled down to 0 °C and stirred for 5 min
followed by addition of EDCI (9.2 mg, 0.048 mmol) and DIEA (17 μL, 0.1
mmol). The mixture was stirred overnight before being quenched by
pouring into 100 mL 1 M HCl. The aqueous layer was extracted three times
with 20 mL chloroform. The combined organic layer was washed five times
with 50 mL water, once with 50 mL brine, dried with Na2SO4, filtered and
concentrated under reduced pressure. Column chromatography using
gradient elution from chloroform to 10% MeOH-chloroform was needed to
obtain 7 as a yellow oil. (37.22 mg, 0.03 mmol, 77% yield). Rf=0.18 (10%
MeOH-chloroform). 1H NMR (300 MHz, Chloroform-d) δ 7.27 (s, 11H),
6.84 (s, 1H), 6.65 (s, 1H), 5.38 – 5.29 (m, 5H), 5.14 (d, J = 8.4 Hz, 6H),
4.29 (d, J = 12.1 Hz, 1H), 4.03 (s, 1H), 3.88 (s, 2H), 3.77 – 3.59 (m, 2H),
3.29 (s, 2H), 2.70 (s, 2H), 2.47 (s, 4H), 2.26 – 2.10 (m, 5H), 2.09 – 1.92
(m, 14H), 1.27 (h, J = 9.3, 8.3 Hz, 57H), 0.86 (d, J = 6.9 Hz, 6H). 31P NMR
(121 MHz, cdcl3) δ -3.98, -9.02.
Dibenzyl (2-(4-((tert-butyldimethylsilyl)oxy)-2-methylbutan-2-yl)-3,5-
dimethylphenyl) phosphate (5)
Compound 4 (0.1 g, 0.31 mmol) was dissolved in 4 mL dry DCM under
nitrogen followed by addition of tetrazole (2.1 mL, 0.93 mmol, 0.45 M in
MeCN). The reaction mixture was cooled down to 0 °C and then dibenzyl
N,N-diisopropylphosphoramidite (0.155 mL, 0.456 mmol) was added. After
the reaction was allowed to warm up to rt and further stirred for 1.5 h, it
was brought back to 0 °C again. m-CPBA (0.281 g, 0.93 mmol, 57% purity)
was added and the reaction was stirred for another 1.5 h. After completion,
the reaction was quenched by adding 100 mL saturated NaHCO3 and the
aqueous phase was extracted three times with 25 mL CHCl3. The
combined organic layer was then washed with 50 mL water, 50 mL brine
and dried over Na2SO4. After being filtered and concentrated under
reduced pressure, the crude was purified through column chromatography
using gradient elution from hexane to 20% EtOAc-hexane. Compound 5
was obtained as a yellow oil. (0.1714 g, 0.3 mmol, 95% yield). Rf=0.24
(10% EtOAc-hexane). 1H NMR (300 MHz, Chloroform-d) δ 7.43 – 7.26 (m,
10H), 7.11 (d, J = 2.1 Hz, 1H), 6.75 – 6.69 (m, 1H), 5.12 (d, J = 8.2 Hz,
4H), 3.49 (dd, J = 7.9, 6.9 Hz, 2H), 2.50 (s, 3H), 2.18 (s, 3H), 2.13 – 2.05
(m, 2H), 1.53 (d, J = 0.8 Hz, 6H), 0.84 (d, J = 0.8 Hz, 9H), -0.04 (d, J = 0.8
Hz, 6H). 13C NMR (75 MHz, cdcl3) δ 150.44, 150.35, 138.66, 136.21,
135.82, 135.73, 132.86, 132.75, 131.09, 128.62, 128.59, 128.04, 118.95,
118.93, 69.80, 69.73, 61.11, 45.91, 39.63, 32.24, 26.05, 25.69, 20.42,
18.31, -5.22. 31P NMR (121 MHz, cdcl3) δ -6.98. HRMS-DART: [M+H]+
calcd for C33H48O5PSi: 583.3009, found: 583.2745.
PRL
Compound 7 (34.4 mg, 0.0285 mmol) was dissolved in 1 mL dry DCM in a
small vial at 0 °C. TMS-I (10.2 μL, 0.0712 mmol) was then added. The
reaction was stirred for 40 min followed by removal of the solvent under
reduced pressure. After further drying under high vacuum for 30 min, 2 mL
of MeOH/H2O (95/5, v/v) was added into the vial and the reaction was
stirred for another one hour. Upon removal of the solvent, the crude was
subjected to column chromatography. Gradient elution from chloroform to
30% methanol-chloroform and final elution with MeOH/chloroform/H2O
(25/65/4, v/v) was performed to purify PRL as a yellow oil. (17.6 mg,
0.0171 mmol, 60% yield). 1H NMR (300 MHz, 20% CD3OD-CDCl3) δ 6.42
(m, 2H), 5.21 – 5.17 (m, 4H), 5.01 (s, 1H), 4.19 (d, J = 11.5 Hz, 2H), 3.72
(d, J = 33.7 Hz, 6H), 2.29 (s, 4H), 2.12 (t, J = 15.6 Hz, 8H), 1.86 (d, J = 6.2
Hz, 8H), 1.44 (s, 4H), 1.13 (t, J = 6.0 Hz, 46H), 0.75 – 0.69 (m, 6H). 31P
NMR (121 MHz, 20% CD3OD-CDCl3) δ 0.63, -4.55. ESI-MS: [M-H]- calcd
for C54H94NO13P2, 1026.6200, found 1026.6173. [M-2H]2- calcd for
C54H93NO13P2, 512.8061, found 512.8077. [M+I]- calcd for C54H95NO13P2I,
1154.5323, found 1154.5304.
(2R,3S,4S,5R,6S)-2-(acetoxymethyl)-6-(4-((((2,3-
bis(hexadecyloxy)propyl)carbamoyl)oxy)methyl)phenoxy)tetrahydro
-2H-pyran-3,4,5-triyl triacetate (12)
3-(2-((bis(benzyloxy)phosphoryl)oxy)-4,6-dimethylphenyl)-3-
methylbutanoic acid (6)
In a 50 mL RBF, 11 (172 mg, 0.278 mmol), and 13 (125 mg, 0.23 mmol)
were added under nitrogen and the flask was cooled down to 0 °C in an
ice bath. 2 mL DMF containing triethylamine (96.2 μL, 0.69 mmol) was
then added into the flask. After overnight stirring, the reaction was
quenched by pouring into 100 mL water. The aqueous layer was extracted
three times with 25 mL chloroform. After washing the combined organic
layer with 50 mL water five times, and 50 mL brine once, the resulting
Compound 5 (0.43 g, 0.74 mmol) was dissolved with 4 mL acetone in a 50
mL RBF followed by addition of KF (47.34 mg, 0.815 mmol) and stirred
briefly before being cooled down to 0 °C. Jones reagent (~1 mL, containing
1 mL H2O, 0.23 mL concentrated H2SO4 and 0.27 g CrO3) was then added
dropwise to obtain a clear orange solution. The reaction was allowed to
warm up to rt and further stirred for 3 h. After this time, the reaction mixture
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
solution was dried over Na2SO4, filtered and concentrated under vacuum,
the crude was purified by column chromatography using gradient elution
from 10% EtOAc-hexane to 40% EtOAc-hexane to yield 12 as a white solid.
(174 mg, 0.17 mmol, 74% yield). Rf=0.7 (50% EtOAc-hexane). 1H NMR
(300 MHz, Chloroform-d) δ 7.34 – 7.28 (m, 2H), 7.04 – 6.92 (m, 2H), 5.55
– 5.41 (m, 2H), 5.16 – 4.98 (m, 5H), 4.28 – 3.99 (m, 3H), 3.60 – 3.36 (m,
8H), 3.29 – 3.17 (m, 1H), 2.18 (s, 3H), 2.06 (s, 6H), 2.01 (s, 3H), 1.54 (t, J
= 7.3 Hz, 4H), 1.25 (s, 56H), 0.93 – 0.84 (m, 6H). 13C NMR (75 MHz, cdcl3)
δ 170.46, 170.37, 170.25, 169.49, 156.95, 156.56, 131.81, 129.94, 117.06,
99.81, 71.96, 71.29, 71.20, 70.96, 70.41, 68.76, 67.02, 66.97, 66.27, 61.49,
42.60, 32.07, 30.18, 29.85, 29.81, 29.77, 29.63, 29.51, 26.24, 22.84, 20.87,
20.82, 20.74, 14.27. HRMS-DART: [M+H]+ cacld for C57H98O14N:
1020.6987, found: 1020.6404.
concentration was 350 U/mL. The solution was stored in a 4 °C fridge. β-
Galactosidase was also used directly after purchasing. According to the
LOT bioactivity analysis, the unit concentration was 550 U/mL. The
solution was stored in 4 °C fridge. Denatured enzyme solutions were
prepared by heating the above-mentioned stock solutions in a 60 °C water
bath for one hour and then slowly cooling back down to room temperature.
Nile red release studies with enzyme addition
A 100 µL aliquot of the prepared 2 mM liposome solution was added into
a sub-micro quartz cuvette. Enzyme was added directly into the cuvette
(esterase: 10 µL enzyme stock (0.45 U), β-galactosidase: 1 µL enzyme
stock (0.55 U), alkaline phosphatase: 1 µL enzyme stock (0.35 U)). The
cuvettes were heated in a 30 °C water bath in between measurements.
Buffer control sets were also done by switching enzyme solution into either
TBS or ammonia sulfate buffer based on the buffers used for the different
enzymes used in study. Fluorescence intensity was then measured over
time (excitation wavelength=552 nm). For GRL and ERL, spectrometer
settings were: excitation slit=5 nm, emission slit=5 nm. For PRL, these
were: excitation slit=5 nm, emission slit=10 nm. When processing the data,
fluorescence intensities at 635 nm were selected. Experiments were run
at least 3 times, each with different batches of liposomes, and averaged
data were reported with error bars showing standard error.
GRL
Compound 12 (20 mg, 0.0196 mmol) was dissolved in 2 mL MeOH in a
large 4 dr vial at 0 °C. NaOMe (27 μL, 0.1176 mmol, 4.375 M in MeOH)
was next added slowly. The reaction mixture immediately turned yellow
and was stirred overnight before being neutralized with 1 N HCl. The
solvent was evaporated, and the crude was purified via a column
chromatography using gradient elution from 100% chloroform to 30%
MeOH-chloroform. GRL was obtained as a white solid. (15 mg, 0.0176
mmol, 90% yield). Rf=0.2 (10% MeOH-chloroform). 1H NMR (600 MHz,
20% CD3OD-Chloroform-d) δ 7.12 (d, J = 7.9 Hz, 2H), 6.92 – 6.86 (m, 2H),
4.92 – 4.83 (m, 2H), 4.70 (dd, J = 7.7, 1.9 Hz, 1H), 3.80 (t, J = 2.6 Hz, 1H),
3.70 – 3.59 (m, 3H), 3.49 (t, J = 6.1 Hz, 1H), 3.46 – 3.38 (m, 2H), 3.38 –
3.20 (m, 7H), 3.05 (dd, J = 14.3, 6.1 Hz, 1H), 1.39 (q, J = 6.6 Hz, 4H), 1.09
(d, J = 2.0 Hz, 58H), 0.71 (td, J = 7.0, 2.0 Hz, 6H). 13C NMR (126 MHz,
20% CD3OD-Chloroform-d) δ 157.21, 156.88, 130.45, 129.43, 116.56,
101.18, 77.36, 76.97, 75.00, 73.33, 71.69, 70.88, 70.23, 68.53, 66.22,
61.07, 49.24, 49.13, 49.07, 48.96, 48.90, 48.78, 48.73, 48.69, 48.61, 48.56,
48.44, 48.27, 48.10, 41.93, 31.75, 29.79, 29.52, 29.50, 29.48, 29.46, 29.39,
29.30, 29.18, 25.90, 25.87, 22.49, 13.79. ESI-MS: [M+H]+ cacld for
C49H90O10N: 852.6565, found 852.6561, [M+NH4]+ cacld for C49H93O10N2:
869.6830, found: 869.6847. [M+Na]+ cacld for C49H89O10NNa:874.6384,
found: 874.6409.
Preparation of liposomes for sulforhodamine B (SRB) studies
The same lipid stocks prepared for NR study were used. SRB sodium salt
was dissolved with 1×TBS (pH=8, containing 25 mM Tris/Tris HCl, 0.10 M
NaCl, 0.0027 M KCl) to produce a 20 mM solution. Proper volumes of each
stock solution were pipetted into a small vial to obtain the desired
percentage of each lipid. The solvents were evaporated under nitrogen
stream and the resulting lipid film was further dried under vacuum for at
least one hour. After that, the lipid film was hydrated with 20 mM SRB stock
solution in a 60 °C water bath for 4 sets of 15 min with vortexing after each
set, followed by ten freeze-thaw cycles in a dry ice-acetone bath and a
60 °C water bath. Then, the solutions were extruded through a 200 nm
polycarbonate membrane for 15 passes with a LiposoFast extruder
(Avestin, Inc.). Finally, the unencapsulated dye was removed with a size
exclusion column (SEC). Sephadex G-50 used for SEC was pre-
equilibrated with 1×TBS (pH=8, containing 25 mM Tris/Tris HCl, 0.13 M
NaCl, 0.0027 M KCl) for at least one hour prior to use. A micro-column was
used for separation and gravity elution was sufficient for separation.
Fractions were collected every ~1mL, and the first fraction showing pink
color was usually the liposome solution, which was further checked by
adding Triton X-100 under a UV lamp, after which an increase in
fluorescent intensity was observed denoting the release of encapsulated
dye.
Preparation of liposomes for Nile red (NR) release studies
Enzyme-responsive lipids, bulk lipids and NR stock solutions were
prepared in either chloroform or MeOH/chloroform mix and stored in -20 °C
freezer. Proper volumes of each stock solution were pipetted into a small
vial to obtain the desired percentage of each lipid. NR was added as 5%
of the total lipid content. The solvents were then evaporated under nitrogen
stream and the resulting lipid film was further dried under vacuum for at
least one hour. After that, the appropriate buffer or water was added into
the vial. For ERL and GRL containing liposomes, 1×TBS (pH=8,
containing 25 mM Tris/Tris HCl, 0.13 M NaCl, 0.0027 M KCl) was used.
For PRL containing liposomes, 1×TBS with cation activators (pH=8,
containing 25 mM Tris/Tris HCl, 0.13 M NaCl, 0.0027 M KCl, 0.25 mM
MgCl2 and 0.25 mM ZnCl2) was used. The concentration of the total lipid
content used in this release assay was 2 mM. The film was hydrated in a
60 °C water bath for 4 sets of 15 min with vortexing after each set, followed
by ten freeze-thaw cycles using a dry ice-acetone bath and 60 °C water
bath. Finally, the solutions were extruded through a 200 nm polycarbonate
membrane for 15 passes with a LiposoFast extruder (Avestin, Inc.). The
resulting liposomes were stored at 4 °C and were used up within two days.
SRB release assays with esterase addition
A 100 µL aliquot of the prepared liposome solution was added into a sub-
micro quartz cuvette. Esterase stock solution (10 µL, 3 mg/mL, 0.45 U)
was added directly into the cuvette to a final concentration of 0.27 mg/mL.
The cuvette was heated in a 30 °C water bath in between measurements.
Buffer control sets were also performed by switching enzyme solution into
TBS buffer and performing the same readings. Fluorescence intensity was
then measured over time (excitation wavelength=550 nm, excitation
slit=10 nm, emission slit=2.5 nm). After completion, 1 µL of an aqueous
20% Triton X-100 solution was added to trigger complete release. When
processing the data, fluorescence intensities at 590 nm were selected and
fluorescence increases were reported as a percentage of the fluorescence
after triton X-100 treatment for each sample. Experiments were run at least
3 times each with different batches of liposomes, and averaged data were
reported with error bars showing standard error.
Enzyme stock preparation
The esterase stock solution was prepared in 1×TBS (pH=8, containing 25
mM Tris/Tris HCl, 0.13 M NaCl, 0.0027 M KCl) at a concentration of 3
mg/mL. According to LOT bioactivity analysis and unit definition provided
by Sigma-Aldrich, the unit concentration was ~45 U/mL. The stock
solutions were divided into small aliquots in Eppendorf tubes and stored in
a -20 °C freezer until their use. Alkaline phosphatase was used directly
after purchasing. According to the LOT bioactivity analysis, the unit
DLS analysis of particle sizes before and after the enzyme treatment
8
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
Ballesteros, J. Alvarez, S. Cerdán, J.-P. Galons, R. J. Gillies, Magn.
Reson. Med. 1999, 41, 743-750.
DLS measurements were carried out with a Malvern Zetasizer Nano ZS
instrument equipped with a 4.0 mW laser operating at ꢀ=633 nm. Samples
were prepared by diluting the liposome solutions before or after adding
enzyme by 10x with the proper matching buffer. As an example, a 5 µL of
the liposome solution was added into 45 µL of proper buffer in a micro
cuvette for measurement. All samples were determined at a scattering
angle of 173° at 25 °C. The reported data were the average of three tests
with error bars showing standard error.
[19] a) P. Meers, Adv. Drug Deliv. Rev. 2001, 53, 265-272; b) F. Fouladi, K.
J. Steffen, S. Mallik, Bioconjugate Chem. 2017, 28, 857-868; c) Z.
Huang, F. C. Szoka Jr, in Liposome technology, CRC Press, 2018, pp.
187-218; d) T. L. Andresen, D. H. Thompson, T. Kaasgaard, Mol.
Membr. Biol. 2010, 27, 353-363.
[20] a) L. Linderoth, G. H. Peters, R. Madsen, T. L. Andresen, Angew.
Chem. Int. Ed. 2009, 48, 1823-1826; b) C. Foged, H. M. Nielsen, S.
Frokjaer, Int. J. Pharm. 2007, 331, 160-166; c) S. Thamphiwatana, W.
Gao, D. Pornpattananangkul, Q. Zhang, V. Fu, J. Li, J. Li, M. Obonyo,
L. Zhang, J. Mater. Chem. B 2014, 2, 8201-8207.
[21] a) R. B. Cohen, M. M. Nachlas, A. M. Seligman, Cancer Res. 1951, 11,
709; b) H. Dong, L. Pang, H. Cong, Y. Shen, B. Yu, Drug Deliv. 2019,
26, 416-432; c) H. Yano, S. Kayukawa, S. Iida, C. Nakagawa, T. Oguri,
T. Sanda, J. Ding, F. Mori, A. Ito, M. Ri, A. Inagaki, S. Kusumoto, T.
Ishida, H. Komatsu, H. Inagaki, A. Suzuki, R. Ueda, Cancer Sci. 2008,
99, 2309-2314.
[22] S. Rainier, M. Bui, E. Mark, D. Thornas, D. Tokarz, L. Ming, C. Delaney,
R. J. Richardson, J. W. Albers, N. Matsunami, J. Stevens, H. Coon, M.
Leppert, J. K. Fink, Am. J. Hum. Gen. 2008, 82, 780-785.
[23] H. G. Kunkel, S. M. Ward, J. Exp. Med. 1947, 86, 325-337.
[24] H. Kaduszkiewicz, T. Zimmermann, H. P. Beck-Bornholdt, H. van den
Bussche, Brit. Med. J. 2005, 331, 321-323.
Acknowledgements
This material is based upon work supported by the National
Science Foundation under Grant No. DMR-1807689.
Keywords: enzyme • drug delivery • lipids • liposomes • modular
strategy
[25] a) A. Ostman, C. Hellberg, F. D. Bohmer, Nat. Rev. Cancer 2006, 6,
307-320; b) S. Zhao, D. Sedwick, Z. Wang, Oncogene 2015, 34, 3885-
3894; c) M. Kim, M. Baek, D. J. Kim, Curr. Pharm. Des. 2017, 23, 4226-
4246.
[26] a) M. J. Kendall, R. Cockel, J. Becker, C. F. Hawkins, Ann. Rheum. Dis.
1970, 29, 537-&; b) K. Ooi, K. Shiraki, Y. Morishita, T. Nobori, J. Clin.
Lab. Anal. 2007, 21, 133-139.
[1]
[2]
a) G. Gregoriadis, B. E. Ryman, Biochem. J. 1971, 124, 58P; b) G.
Gregoriadis, FEBS Lett. 1973, 36, 292-296.
C.-C. Chang, D.-Z. Liu, S.-Y. Lin, H.-J. Liang, W.-C. Hou, W.-J. Huang,
C.-H. Chang, F.-M. Ho, Y.-C. Liang, Food Chem. Toxicol. 2008, 46,
3116-3121.
[27] J. W. Li, C. Xu, Y. Fan, Y. Wang, Y. B. Xiao, Plos One 2014, 9.
[28] M. Haarhaus, V. Brandenburg, K. Kalantar-Zadeh, P. Stenvinkel, P.
Magnusson, Nat. Rev. Nephrol. 2017, 13, 429-442.
[3]
[4]
H.-I. Chang, M.-K. Yeh, Int. J. Nanomed. 2012, 7, 49-60.
T. M. Allen, C. B. Hansen, D. E. L. de Menezes, Adv. Drug Deliv. Rev.
1995, 16, 267-284.
[29] a) D. Asanuma, M. Sakabe, M. Kamiya, K. Yamamoto, J. Hiratake, M.
Ogawa, N. Kosaka, P. L. Choyke, T. Nagano, H. Kobayashi, Y. Urano,
Nat. Commun. 2015, 6, 7; b) H. B. Bosmann, T. C. Hall, Proc. Natl.
Acad. Sci. U.S.A. 1974, 71, 1833-1837; c) S. K. Chatterjee, M.
Bhattacharya, J. J. Barlow, Cancer Res. 1979, 39, 1943-1951; d) J. J.
Zhang, P. H. Cheng, K. Y. Pu, Bioconjugate Chem. 2019, .30, 2089-
2101.
[30] T. Kajimoto, M. Node, Curr. Top. Med. Chem. 2009, 9, 13-33.
[31] P. Pinnaduwage, L. Huang, Biochim. Biophys. Acta, Biomembr. 1988,
939, 375-382.
[5]
[6]
A. A. Gabizon, Cancer Res. 1992, 52, 891-896.
a) B. S. Pattni, V. V. Chupin, V. P. Torchilin, Chem. Rev. 2015, 115,
10938-10966; b) A. Bunker, A. Magarkar, T. Viitala, Biochim. Biophys.
Acta, Biomembr. 2016, 1858, 2334-2352; c) C. Zylberberg, S.
Matosevic, Drug Deliv. 2016, 23, 3319-3329; d) T. M. Allen, P. R. Cullis,
Adv. Drug Deliv. Rev. 2013, 65, 36-48.
a) F. Muggia, J. Hainsworth, S. Jeffers, S. Groshen, M. Tan, F. Greco,
in Proc Am Soc Clin Oncol, Vol. 15, 1996, p. 287; b) R. N. Davidson, L.
D. Martino, L. Gradoni, R. Giacchino, R. Russo, G. B. Gaeta, R.
Pempinello, S. Scott, F. Raimondi, A. Cascio, T. Prestileo, L. Caldeira,
R. J. Wilkinson, A. D. M. Bryceson, QJM 1994, 87, 75-81.
S. Bibi, E. Lattmann, A. R. Mohammed, Y. Perrie, J. Microencapsul.
2012, 29, 262-276.
[7]
[32] J. Yan, S. Lee, A. Zhang, J. Yoon, Chem. Soc. Rev. 2018, 47, 6900-
6916.
[33] M. Gisbert-Garzarán, M. Manzano, M. Vallet-Regí, Chem. Eng. J. 2018,
340, 24-31.
[8]
[9]
S. Alam, S. I. Mattern-Schain, M. D. Best, in Comprehensive
Supramolecular Chemistry II, Vol. 5 (Ed.: J. L. Atwood), Elsevier,
Oxford, 2017, pp. 329-364.
[34] a) L. D. Lavis, T.-Y. Chao, R. T. Raines, ChemBioChem 2006, 7, 1151-
1154; b) B. Wang, S. Gangwar, G. M. Pauletti, T. J. Siahaan, R. T.
Borchardt, J. Org. Chem. 1997, 62, 1363-1367; c) M. N. Levine, R. T.
Raines, Chem. Sci. 2012, 3, 2412-2420; d) R. B. Greenwald, Y. H.
Choe, C. D. Conover, K. Shum, D. Wu, M. Royzen, J. Med. Chem.
2000, 43, 475-487; e) S. S. Chandran, K. A. Dickson, R. T. Raines, J.
Am. Chem. Soc. 2005, 127, 1652-1653.
[35] a) P. L. Carl, P. K. Chakravarty, J. A. Katzenellenbogen, J. Med. Chem.
1981, 24, 479-480; b) S. Gnaim, D. Shabat, Acc. Chem. Res. 2014, 47,
2970-2984.
[10] a) A. A. Kale, V. P. Torchilin, Bioconjugate Chem. 2007, 18, 363-370; b)
W. Viricel, A. Mbarek, J. Leblond, Angew. Chem. Int. Ed. Engl. 2015,
54, 12743-12747.
[11] a) W. Ong, Y. Yang, A. C. Cruciano, R. L. McCarley, J. Am. Chem. Soc.
2008, 130, 14739-14744; b) N. L. Abbott, C. M. Jewell, M. E. Hays, Y.
Kondo, D. M. Lynn, J. Am. Chem. Soc. 2005, 127, 11576-11577.
[12] J. Lou, X. Zhang, M. D. Best, Chem. Eur. J. 2019, 25, 20-25.
[13] a) A. I. Elegbede, J. Banerjee, A. J. Hanson, S. Tobwala, B. Ganguli, R.
Wang, X. Lu, D. K. Srivastava, S. Mallik, J. Am. Chem. Soc. 2008, 130,
10633-10642; b) L. Zhu, P. Kate, V. P. Torchilin, ACS Nano 2012, 6,
3491-3498; c) F. Menger, D. Johnston Jr, J. Am. Chem. Soc. 1991,
113, 5467-5468; d) S. Davis, F. Szoka, Bioconjugate Chem. 1998, 9,
783-792; e) C. C. Pak, S. Ali, A. S. Janoff, P. Meers, Biochim. Biophys.
Acta, Biomembr. 1998, 1372, 13-27; f) C. C. Pak, R. K. Erukulla, P. L.
Ahl, A. S. Janoff, P. Meers, Biochim. Biophys. Acta, Biomembr. 1999,
1419, 111-126.
[14] a) Z.-Y. Zhang, B. D. Smith, Bioconjugate Chem. 1999, 10, 1150-1152;
b) T. Furuta, S. S.-H. Wang, J. L. Dantzker, T. M. Dore, W. J. Bybee, E.
M. Callaway, W. Denk, R. Y. Tsien, Proc. Natl. Acad. Sci. 1999, 96,
1193-1200; c) K. A. Carter, S. Shao, M. I. Hoopes, D. Luo, B. Ahsan, V.
M. Grigoryants, W. Song, H. Huang, G. Zhang, R. K. Pandey, J. Geng,
B. A. Pfeifer, C. P. Scholes, J. Ortega, M. Karttunen, J. F. Lovell, Nat.
Commun. 2014, 5, 3546; d) A. M. Bayer, S. Alam, S. I. Mattern-Schain,
M. D. Best, Chem. Eur. J. 2014, 20, 3350-3357.
[15] a) M. Yatvin, J. Weinstein, W. Dennis, R. Blumenthal, Science 1978,
202, 1290-1293; b) L. Li, T. L. M. ten Hagen, D. Schipper, T. M.
Wijnberg, G. C. van Rhoon, A. M. M. Eggermont, L. H. Lindner, G. A.
Koning, J. Controlled Release 2010, 143, 274-279.
[16] a) X. Liang, J. Gao, L. Jiang, J. Luo, L. Jing, X. Li, Y. Jin, Z. Dai, ACS
Nano 2015, 9, 1280-1293; b) S. Ibsen, M. Benchimol, D. Simberg, C.
Schutt, J. Steiner, S. Esener, J. Controlled Release 2011, 155, 358-
366.
[17] Y. Matsumura, H. N. Ananthaswamy, Toxicol. Appl. Pharmacol. 2004,
195, 298-308.
[18] a) R. J. Gillies, Z. Liu, Z. Bhujwalla, Am. J. Physiol. Cell Physiol. 1994,
267, C195-C203; b) R. Van Sluis, Z. M. Bhujwalla, N. Raghunand, P.
[36] D. C. Litzinger, L. Huang, Biochim. Biophys. Acta, Biomembr. 1992,
1113, 201-227.
[37] D. P. Siegel, R. M. Epand, Biophys. J. 1997, 73, 3089-3111.
[38] a) W. Zhou, C. Andrews, J. Liu, J. W. Shultz, M. P. Valley, J. J. Cali, E.
M. Hawkins, D. H. Klaubert, R. F. Bulleit, K. V. Wood, ChemBioChem
2008, 9, 714-718; b) K. L. Amsberry, R. T. Borchardt, J. Org. Chem.
1990, 55, 5867-5877.
[39] T. Chauvin, P. Durand, M. Bernier, H. Meudal, B. T. Doan, F. Noury, B.
Badet, J. C. Beloeil, E. Toth, Angew. Chem. Int. Ed. Engl. 2008, 47,
4370-4372.
[40] a) E. A. Losey, M. D. Smith, M. Meng, M. D. Best, Bioconjugate Chem.
2009, 20, 376-383; b) X. Zhang, D. S. Alves, J. Lou, S. D. Hill, F. N.
Barrera, M. D. Best, Chem Commun (Camb) 2018, 54, 6169-6172.
[41] L. Feng, K. Manabe, J. C. Shope, S. Widmer, D. B. DeWald, G. D.
Prestwich, Cell Chem. Biol. 2002, 9, 795-803.
[42] P. Greenspan, E. P. Mayer, S. D. Fowler, J. Cell. Biol. 1985, 100, 965-
973.
[43] X. Liang, X. Yue, Z. Dai, J.-i. Kikuchi, Chem. Commun. 2011, 47, 4751-
4753.
[44] D. Simberg, S. Weisman, Y. Talmon, Y. Barenholz, Crit. Rev. Ther.
Drug 2004, 21, 62.
[45] S. A. Whitehead, C. D. McNitt, S. I. Mattern-Schain, A. J. Carr, S. Alam,
V. V. Popik, M. D. Best, Bioconjugate Chem. 2017, 28, 923-932.
[46] E. Tanguy, Q. Wang, H. Moine, N. Vitale, Front. Cell. Neurosci. 2019,
13.
[47] J. Lou, A. J. Carr, A. J. Watson, S. I. Mattern-Schain, M. D. Best,
Chem. Eur. J. 2018, 24, 3599-3607.
[48] V. Vichai, K. Kirtikara, Nat. Protoc. 2006, 1, 1112.
9
This article is protected by copyright. All rights reserved.

10.1002/chem.202000529
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
A modular approach is reported for the development of enzyme-responsive liposomes. These exploit synthetic lipid switches
containing variable enzyme substrates that, when removed, yield decomposition of a self-immolating linker producing a non-bilayer
lipid that perturbs the membrane and triggers release of contents. This approach enables the targeting of a range of enzymes that
are overexpressed in diseased cells for drug delivery applications.
Institute and/or researcher Twitter usernames: @chemistryUTK
10
This article is protected by copyright. All rights reserved.