Efficient encapsulation of plasmid DNA in pH-sensitive PMPC-PDPA polymersomes: Study of the effect of PDPA block length on copolymer-DNA binding affinity

Abstract

We report the self‐assembly of a series of amphiphilic diblock copolymers comprising a biocompatible, hydrophilic block, poly(2‐methacryloyloxyethyl phosphorylcholine) (PMPC) and a pH‐sensitive block, poly(2‐(diisopropylamino)ethyl methacrylate) (PDPA), into a dispersion of colloidally stable, nanometer‐sized polymersomes at physiological pH and salt concentration. The pH‐sensitivity of the PDPA block affords the electrostatic interaction of these block copolymers with nucleic acids at endocytic pH, as a result of the protonation of its tertiary amine groups at pH values below its pK a . Herein we investigate the effect of PDPA block length on the binding affinity of the block copolymer to plasmid DNA.
magnified image

Full text

Efficient Encapsulation of Plasmid DNA in pH-
Sensitive PMPC–PDPA Polymersomes: Study of
the Effect of PDPA Block Length on Copolymer–
DNA Binding Affinity
a
Hannah Lomas, Jianzhong Du, Irene Canton, Jeppe Madsen,
Nicholas Warren, Steven P. Armes, Andrew L. Lewis, Giuseppe Battaglia*
1
Introduction
Polymersome
[1]
is the common term used for a polymeric
vesicle. It comprises amphiphilic block copolymer macro-
molecules that can self-assemble in water forming highly
organised membrane structures, which bend to form
spherical, enclosed structures known as vesicles. Polymer-
somes have a hydrophilic corona, which attracts water
Full Paper
G. Battaglia, H. Lomas, I. Canton, J. Madsen, N. Warren
Department of Biomedical Science, The Krebs Institute, University
of Sheffield, Western Bank, Sheffield, S10 2TN, UK
E-mail: g.battaglia@sheffield.ac.uk
H. Lomas
Biomaterials and Tissue Engineering Group, Department of
Engineering Materials, University of Sheffield, The Kroto Research
Institute, Broad Lane, Sheffield, S3 7HQ, UK
a
:Supporting information for this article is available at the bottom
of the article’s abstract page, which can be accessed from the
journal’s homepage at http://www.mbs-journal.de, or from the
author.
J. Du, J. Madsen, N. Warren, S. P. Armes
Department of Chemistry, University of Sheffield, Dainton
Building, Brook Hill, Sheffield, S3 7HF, UK
J. Du
Current address: School of Materials Science and Engineering,
Tongji University, 1239 Siping Road, Shanghai, 200092, China
A. L. Lewis
Biocompatibles UK Ltd, Chapman House, Farnham Business Park,
Weydon Lane, Farnham, Surrey, GU9 8QL, UK
We report the self-assembly of a series of amphiphilic diblock copolymers comprising a
biocompatible, hydrophilic block, poly(2-methacryloyloxyethyl phosphorylcholine) (PMPC)
and a pH-sensitive block, poly(2-(diisopropylamino)ethyl methacrylate) (PDPA), into a dis-
persion of colloidally stable, nanometer-sized polymer-
somes at physiological pH and salt concentration. The
pH-sensitivity of the PDPA block affords the electro-
static interaction of these block copolymers with
nucleic acids at endocytic pH, as a result of the proto-
nation of its tertiary amine groups at pH values below
its pK
a
. Herein we investigate the effect of PDPA block
length on the binding affinity of the block copolymer to
plasmid DNA.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083 513

molecules, and an aqueous core. The aqueous core of
polymersomes can afford the entrapment of polar com-
pounds, making the use of polymersomes for gene therapy
and drug delivery applications a more promising alter-
native to the currently available viral and non-viral vectors
that have been more widely studied over the past two
decades. There are numerous examples in the literature of
hydrophilic pharmaceutical agents, such as anti-cancer
drugs
[2]
and nucleic acids
[3]
being incorporated within a
polymersome core. Polymersomes are also capable of
incorporating hydrophobic molecules,
[4]
which can be
entrapped within their hydrophobic membrane core, and
amphiphilic molecules,
[4a,5]
which can be aligned with
the amphiphilic copolymer chains comprising the mem-
brane.
The synthetic amphiphilic block copolymer membrane
displays similar properties to its natural counterpart, with
its elastic, supramolecular fluid-like nature.
[5b]
Due to the
higher molecular weight of block copolymer macromole-
cules compared to the average molecular weight of a
phospholipid, the subsequently longer hydrocarbon chains
of block copolymers become entangled at the core of the
membrane bilayer,
[5a]
resulting in interdigitation of
the chains, and a smaller membrane thickness than would
be expected from the copolymer molecular weight. This in
turn affords block copolymer membranes, which are
typically 10–20 nm in thickness, lower permeability
compared to phospholipid membranes, which have a
thickness of 4–5 nm.
[6]
The non-ergodic nature of polymeric
amphiphilic aggregates
[7]
suggests that the kinetics of their
dissocation is slow, and this further enhances polymer-
somes’ retention of encapsulated cargo.
[5a,8]
In the blood
circulation, it is important that the drug delivery carrier is
relatively impermeable as there will be greater retention of
the payload and an increased probability of the therapeutic
agent reaching its target site. The interdigitation of the
hydrocarbon chains in copolymer membranes also imparts
increased mechanical stability,
[1,5a]
the evidence for which
has been obtained both experimentally by micropipette
aspiration techniques
[9]
and theoretically using coarse-
grain molecular dynamics.
[10]
Polymersomes thus have
many advantages over the more widely studied lipid
vesicles (or liposomes
[11]
). Liposomal systems have proven
to be ‘leaky’,
[12]
and have shown a lack of stability for use in
practical applications,
[13]
due to the limit on the thickness of
the hydrophobic core of the liposomal membrane,
[9]
and
hence a reduced thickness of the lipid vesicle ‘walls’
compared to the walls of polymersomes.
[14]
The fact that polymersomes are generated from syn-
thetic macromolecules allows the design of a vector to suit
the purpose. Design parameters can include the nature of
the polymers used to impart a certain surface chemistry to
polymersomes, for example the presence of PEG at the
corona to afford polymersomes stealthiness in vivo;
[15]
targeting ligands can be introduced to the polymersome
corona for the binding of specific cell surface receptors;
[16]
a
moiety can be incorporated which is responsive to a
particular stimulus,
[17]
for example, variation in
temperature,
[18]
pH,
[2a,5b,19]
hydrogen peroxide concentra-
tion
[20]
and light.
[21]
The possibilities are endless, and it is
due to this array of opportunities regarding the design of
block copolymer systems that can self-assemble into
nanometer-sized, robust, kinetically stable polymersomes,
that such systems have been studied extensively in recent
years for their potential as intracellular delivery vehicles for
various pharmaceutical agents.
In the case of DNA delivery, this requires entry into the
cell nucleus, in order for the therapeutic protein encoded
within the DNA to be expressed. Cellular internalisation of
naked DNA lacks efficiency due to its size and, in particular,
its negative charge.
[22]
The anionic phosphate backbone is
electrostatically repelled by the negatively charged plasma
membrane. The use of vectors to assist with intracellular
DNA delivery therefore dramatically improves the delivery
efficiency, providing protection and stability to DNA both in
the extracellular milieu and as it is trafficked intracellularly.
Viruses are currently the most efficient gene transport
vectors. They can survive the extracellular environment,
target and then penetrate the cell surface, and are replicated
once inside the cell. However, the gene expression from
viral vectors is rarely long term,
[23]
and they have been
found to interact unfavourably with the human immune
system,
[24]
leading to reduced circulation times,
[23]
diffi-
culties in using the same vector more than once or in
patients previously exposed to the virus
[23]
and even the
death of a patient participating in a clinical trial.
[24]
Thus,
despite their efficacy, there has been much concern
regarding the toxicity of viruses.
[22]
Consequently, recent
research work has focussed on the design of non-viral gene
delivery systems. The majority of this work has concen-
trated on vectors that contain a cationic component (lipid or
polymer),
[25]
to exploit the negative charge of the phos-
phate groups within nucleic acids, permitting electrostatic
interaction between vector and nucleic acid, and the
formation of a stable complex. Research has shown that
the presence of a positive charge density at the delivery
vector exterior can promote cellular uptake by endocytosis,
due to interaction between the cationic component and
negatively charged receptors present within the cell plasma
membrane, such as the proteoglycan heparin sulphate.
[26]
However, it is hypothesised that an ionic interaction can
induce cytotoxicity,
[27]
due to aggregation of the complexes
when their positive charge is screened as the cationic
groups within the complex interact with negatively
charged receptors in the plasma membrane, leading to
accumulation of the complexes at the membrane.
[28]
This is
thought to cause damage to the plasma membrane, for
example, by the appearance of pores within, which may be
H. Lomas et al.
514
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

a major contributory factor to the induction of cell
apoptosis.
[29]
A second major drawback of using a net
positively charged DNA-vector complex for in vivo applica-
tions is the propensity to attract serum proteins, which are
generally negatively charged.
[30]
Interaction with serum
proteins may result in the nucleic acid being displaced from
the gene delivery vector, leading to a short circulation time
in the bloodstream.
[31]
Indeed, positively charged species
have the potential to bind to any cells that they come into
contact with, for example blood cells,
[30]
due to their
binding affinity for the proteoglycans present within the
plasma membrane of most mammalian cells. The optimal
vector should therefore be designed to (a) incorporate a
‘stealth’ component and (b) contain cell-targeting moieties.
For the ‘stealth’ component, a domain with hydrophilic,
non-toxic and non-immunogenic properties should be
used, for example PEG,
[32]
pHPMA,
[33]
and oligosaccharides,
such as cyclodextrin.
[34]
Coating a polyplex/lipoplex with
one of these three materials confers resistance to complex
aggregation under physiological salt conditions due to a
reduced surface charge (reduced z-potential in the case of
colloidal aggregates), and a lower binding affinity for serum
constituents.
[35]
Covalent attachment of PEG
[36]
or
pHPMA
[33a]
to the surface of a DNA–vector complex can
confer steric stabilisation to a complex, due to their non-
ionic, hydrophilic nature, thereby shielding a polyplex/
lipoplex from nuclease degradation.
Herein we present an amphiphilic diblock copolymer
consisting of a pH-responsive polyelectrolyte poly(2-
(diisopropylamino)ethyl methacrylate) (PDPA) and a bio-
compatible, hydrophilic polymer poly(2-methacryloyloxy-
ethyl phosphorylcholine) (PMPC), that has previously been
investigated as a potential gene delivery vector.
[3a,3b]
This
diblock copolymer self-assembles into colloidally stable
polymersomes in aqueous media at physiological pH.
[2a]
The pH-sensitivity imparted by the PDPA block stimulates
dissolution of the polymersomes if the pH is reduced below
the PDPA pK
a
, which is approximately 6.4,
[2a,3a]
depending
on the concentration of ions in the media
[37]
(Figure 1), due
to the protonation of its tertiary amine groups. This affords
electrostatic binding to plasmid DNA phosphate groups at
weakly acidic pH.
[3a]
At neutral pH, previous work has
shown that this interaction is weakened,
[3a]
and instead
plasmid DNA (comprising 3800 base pairs) can be
encapsulated within PMPC
25
-PDPA
70
polymersomes with
a relatively high encapsulation efficiency,
[3a,3b]
and with-
out compromising the colloidal stability or shelf-life of the
polymersomes.
[3b]
Moreover, these polymersomes are
capable of transfecting human dermal fibroblast (HDF)
cells, which are primary cells, and Chinese hamster ovary
(CHO) cells (an animal cell line), with a high transfection
efficiency compared to the use of more common non-viral
methods of transfection, including the use of Lipofectamine
and calcium phosphate,
[3a]
largely due to the lower levels of
cytotoxicity imparted by the PMPC
25
-PDPA
70
polymer-
somes.
A selection of PMPC–PDPA block copolymers with
varying PDPA block lengths have been synthesised and
characterised within the Armes research group, and probed
for their potential to interact with and encapsulate
plasmid DNA. These alternative copolymers include
PMPC
25
-PDPA
90
, PMPC
25
-PDPA
120
and PMPC
25
-PDPA
160
,
the latter two of which have been shown in previous work
to self-assemble into colloidally stable polymersomes at
neutral pH.
[2a]
We present here an investigation into
whether a longer PDPA block length relative to the PMPC
block can afford a higher binding affinity for plasmid DNA,
due to the greater number of free amine groups on the PDPA
block, which are weakly cationic in mildly acidic conditions.
Such conditions are analogous to the pH range a gene
delivery vector would be exposed to once internalised by
cells via an endocytic pathway.
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 1. Chemical structure of the PMPC–PDPA block copolymer, where mand nare the mean degrees of polymerisation. Below the PDPA
pK
a
of 6.4, the tertiary amine groups on the PDPA chains are protonated, rendering the overall block copolymer hydrophilic, and allowing its
molecular dissolution in water, as unimers. Above pH 6.4, the PDPA tertiary amine groups become deprotonated, rendering this block now
hydrophobic, and the overall block copolymer amphiphilic, thereby triggering its self-assembly into membranes in an aqueous
environment, as illustrated by the cartoon.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 515

Firstly, transmission electron microscopy and dynamic
light scattering analyses of solutions of each of the PMPC–
PDPA copolymers were performed to confirm their ability to
yield polymersomes at physiological pH and salt concen-
tration. Polymersome shelf-life after storage at 4 8C was also
explored. z-potential measurements were performed on
copolymer–plasmid DNA solutions as a function of pH.
Further, ethidium bromide displacement assays and TEM
analyses at pH 6 were undertaken to explore the extent of
copolymer–DNA complexation at a [DPA]/[phosphate]
molar ratio of 1:1. The % DNA release from the polymer-
somes was probed by dropping the pH to endocytic
conditions and assessing the proportion re-encapsulated
within the polymersomes as they formed at neutral pH
conditions. Finally, the loading efficiency within PMPC–
PDPA polymersomes was investigated, using a new method
to calculate the % loading efficiency, which uses the highly
sensitive fluorescent probe Quant-iT PicoGreen dsDNA
reagent.
Experimental Part
Materials
For the synthesis of the PMPC–PDPA block copolymers, 2-
(methacryloyloxy)ethyl phosphorylcholine (MPC; >99%) was
kindly donated by Biocompatibles UK Ltd. 2-(Diisopropylami-
no)ethyl methacrylate (DPA) was purchased from Scientific
Polymer Products (USA). Copper(I) bromide (CuBr; 99.999%), 2,2-
bipyridine (bpy), methanol and isopropanol were purchased from
Sigma–Aldrich (UK) and were used as received. The silica used for
removal of the ATRP copper catalyst was column chromatography
grade silica gel 60 (0.063–0.200 mm) purchased from E. Merck
(Darmstadt, Germany). 2-(N-Morpholino)ethyl 2-bromo-2-methyl-
propanoate (ME-Br) initiator was synthesised according to a
previously reported procedure.
[38]
The pEGFP DNA construct used to prepare the GFP-encoding
plasmid DNA, and the Qiagen columns, were purchased from
CLONTECH Laboratories (Palo Alto, CA) and the strain JM109 used
for propagation in Escherichia coli was purchased from Promega
UK. Luria–Bertani (LB) media, kanamycin solution, Sfi I, Tris–EDTA
buffer solution and ethidium bromide solution were purchased
from Sigma–Aldrich (UK). The pmaxGFP plasmid (3.5 kbp) used in
some of the DNA encapsulation experiments was purchased from
Amaxa Biosystems.
Chloroform and ethanol were purchased from Fisher Scientific
(UK), and phosphate buffered saline (PBS) tablets (Dulbecco A) from
Oxoid Ltd. PBS solution made up by dissolving 1 tablet per 100 mL
distilled water, typically contained the following concentration of
ions in solution: Sodium chloride (8.0 gL
1
), potassium chloride
(0.2 gL
1
), disodium hydrogen phosphate (1.15gL
1
) and potas-
sium dihydrogen phosphate (0.2 gL
1
), resulting in a final
concentration of ions in solution of 150 mM. Sepharose 4B, sodium
hydroxide, hydrochloric acid and 40,6-diamidino-2-phenylindole
(DAPI) were bought from Sigma–Aldrich (UK). Quant-iT PicoGreen
dsDNA reagent was purchased from Invitrogen (UK).
Methods
PMPC–PDPA Block Copolymer Synthesis
PMPC–PDPA copolymers were synthesised by the Armes research
group using an atom transfer radical polymerisation (ATRP)
procedure, as reported in ref.
[2a]
Briefly, a Schlenk flask with a
magnetic stir bar and a rubber septum was charged with Cu(I)Br
(25.6 mg, 0.178 mmol) and MPC (1.32 g, 4.46 mmol). ME-Br initiator
(50.0 mg, 0.178 mmol) and bpy ligand (55.8 mg, 0.358 mmol) were
dissolved in methanol (2 mL), and this solution was deoxygenated
by bubblingN
2
for 30 min before beinginjected into the flask using a
syringe. The [MPC]: [ME-Br]: [CuBr]:[bpy] relative molar ratios were
25:1:1:2. The reaction was carried out under a nitrogen atmosphere
at 20 8C. After 65 min, a mixture of deoxygenated DPA (6.09g,
28.6 mmol) and methanol (7mL) was injected into the flask. After
48 h, the reaction solution was diluted by addition of isopropanol
(about 200mL)and then passed through a silica column to remove
the catalyst.To increase the purityof the block copolymers for future
cell delivery experiments, solutions of the block copolymer were
dialysedfor approximately 7 days againstwater, with daily changes
of the water (1L per 20g of block copolymer).
Characterisation
The Mnand Mw=Mnvalues of the four PMPC–PDPA diblock
copolymers were assessed by gel permeation chromatography
(GPC). The GPC set-up comprised a Polymer Labs PLgel 5 mm Mixed
‘C’ column operating at 40 8C in combination with a refractive index
detector. The eluent was a 3:1 chloroform: methanol mixture at a
flow rate of 1.0 mLmin
1
and calibration was carried out using five
near-monodisperse poly(methyl methacrylate) standards.The data
were processed by Cirrus GPC offline GPC/SEC software.
Please note that copolymers 1 and 2 were characterized using an
optimized GPC protocol. The data for copolymers 3 and 4 were taken
from ref
[2a]
. However, the method for performing GPC analysis on
PMPC-PDPA block copolymers has since been optimized, and very
recent GPC analysis of copolymers 3 and 4 gave the following
results: Copolymer 3: Mn, GPC 31 300, M
w
/M
n
1.39; copolymer 4:
M
n
, GPC 48 800, M
w
/M
n
2.68. It may be the case that the presence of
water had an influence on the results, and further the polymers
may require further purification before performing GPC analysis;
however these results imply that copolymer 4 is polydisperse, and
also that copolymer 3 is of significantly lower molecular weight
than that targeted. This new data should be kept in mind when
interpreting the results presented herein.
1
H NMR spectra were recorded using a Bruker DRX250 (250 MHz)
spectrometer at ambient temperature using DCl/CD
3
OD as the
solvent mixture to characterise the PMPC
25
–PDPA
70
and PMPC
25
–
PDPA
90
block copolymers and DCl/D
2
O to characterise the PMPC
25
–
PDPA
120
and PMPC
25
–PDPA
160
block copolymers.
The Mnand Mw=Mnvalues for each of the four PMPC–PDPA
copolymers are displayed in Table 1.
Plasmid DNA Preparation
pEGFP DNA construct (3.8 kilo base pairs) was used in DNA
encapsulation and characterisation studies, and the strain JM109
for propagation in E. coli. Selection for the transformed bacteria was
H. Lomas et al.
516
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

maintained by growing the culture at 37 8CinLBmediawith
100 mgmL
1
kanamycin. Plasmid DNA was isolated from E. coli JM109
using Qiagen columns, following the procedure provided by the
manufacturer. DNA purity was assessed by running a 1% agarose gel
stained with ethidium bromide and calculating the ratio of optical
absorbances at 260 and 280 nm, using a UV spectrophotometer (Perkin
Elmer Lambda 25). The optical absorbance at 260 nm and the molar
extinction coefficient were used to calculate the concentration of
nucleic acid present in the final solution.
Preparation of PMPC–PDPA Polymersomes
The technique of film rehydration was used to prepare nanometer-
sized PMPC–PDPA polymersomes, which were then further
processed by sonication, to narrow the polymersome size
distribution. In a typical experiment, PMPC–PDPA block copolymer
(20 mg) was added to a cylindrical glass vial and dissolved in a
solution of 2:1 chloroform:methanol, at a concentration of
2.25 mgmL
1
. For samples comprising the PMPC
25
–PDPA
160
copolymer, this solution was then filtered using filter paper
(Whatman). The solvent was evaporated overnight under vacuum,
resulting in a copolymeric film deposited on the walls of the vial.
The film was rehydrated by the addition of phosphate buffer saline
(150 mM) at pH 2 to form a solution of the copolymer. The pH was
increased by the addition of 1 MNaOH to neutral pH to form the
polymersomes, as observed by the increase in the turbidity of the
sample. The following were the approximate maximum concen-
trations found to be suitable for preparing kinetically stable
polymersomes at neutral pH for each of the copolymers: PMPC
25
–
PDPA
70
5mgmL
1
; PMPC
25
–PDPA
90
4mgmL
1
; PMPC
25
–PDPA
120
3mgmL
1
; PMPC
25
–PDPA
160
2mgmL
1
. The solution of polymer-
somes was sonicated for 5–10 min using a sonicator (Sonicor
Instrument Corporation), and then immediately purified by
preparative GPC, using a size exclusion column containing
Sepharose 4B and PBS eluent at pH 7.3. Fractions containing pure
polymersomes were then stored at 4 8C until use.
Preparation of PMPC–PDPA Polymersomes
Encapsulating Plasmid DNA
The pH of the PMPC–PDPA polymersome solution was reduced
to pH 6.0 via the dropwise addition of 0.1 MHCl. For this particular
experiment, a commercially available pmaxGFP plasmid (3.5 kbp)
(Amaxa Biosystems) 0.5mgmL
1
stock solution was used. The
stock solution of pDNA was added dropwise to the aqueous PMPC–
PDPA copolymer solution at a concentration that corresponded to a
[DPA]/[phosphate] molar ratio of circa 100:1. The solution was
vortexed after the addition of each drop of DNA to the copolymer
solution, checking that the pH did not rise above pH 6.3. After all of
the DNA had been added, the pH was raised to pH 7.3, and the
resulting polymersome dispersion was further processed via
sonication. Polymersomes encapsulating DNA were purified by
preparative GPC, as described above.
DNA Release: Estimation of % Plasmid DNA released
from PMPC
25
–PDPA
70
Copolymer upon Increasing
the pH from pH 6.0 to pH 7
Plasmid DNA (3.8 kbp) was added to 1 mL of a 5 mgmL
1
PMPC
25
-
PDPA
70
copolymer aqueous solution at pH 6.0, up to a concentration
that corresponded to a [DPA]/[phosphate] molar ratio of circa 100:1.
The solution was briefly vortexed, checked for the absence of
aggregation or DNA precipitation visible to the naked eye, and
the pH was increased to pH 7.3 by the dropwise addition of 1 M
NaOH. The resulting dispersion of polymersomes was manually
extruded 21 times using a LiposoFast-Basic through a polycarbo-
nate membrane of defined pore size (200 nm), using gas-tight glass
syringes, and sonicated for 10min. Polymersomes encapsulating
plasmid DNA were purified by preparative GPC, as described above.
The DNA encapsulation efficiency was determined by the addition
of 1 mgmL
1
40,6-diamidino-2-phenylindole (DAPI) to aliquots of
both the pure and impure polymersome solutions, followed by
analysis using a fluorimeter (Cary Eclipse Varian). Fluorescence
emission spectra were recorded using an excitation wavelength of
350 nm for DAPI detection. In the presence of double-stranded DNA,
the broad fluorescence emission peak corresponding to DAPI at
approximately 460 nm increases and shifts to a slightly lower
wavelength.
[3b]
Calibration curves recorded using 1 mgmL
1
DAPI
at known DNA and polymersome concentrations were used to
calculate the % DNA encapsulation efficiency, using the method
described by Lomas et al., 2008.
[3b]
This was typically found to be as
high as 50% under these optimal conditions for DNA encapsulation.
The pH of the GPC-purified polymersome solution was then
dropped to pH 6.0 via the addition of 1MHCl, kept at pH 6.0 for
5 min, and then increased to pH 7.3 by the dropwise addition of 1 M
NaOH. To calculate the % of DNA re-encapsulated inside the
polymersomes, the free DNA was separated off by preparative GPC.
Analysis of the Plasmid DNA Encapsulation
Efficiency within PMPC–PDPA Polymersomes
The encapsulation efficiency of plasmid DNA within PMPC–PDPA
polymersomes was determined by two different methods. The first
of these methods involves the addition of the fluorescent probe
DAPI, as described above and in Lomas et al., 2008.
[3b]
Recently, a
more sensitive technique has been implemented to efficiently
calculate plasmid DNA concentrations and hence measure the %
DNA loading efficiency. This technique uses Quant-iT PicoGreen
dsDNA reagent (‘PicoGreen’), which is a fluorescent probe with an
excitation wavelength of 502 nm and an emission wavelength of
523 nm when bound to nucleic acid. The PicoGreen reagent was
prepared approximately 1 h before use by diluting it 1 in 200 Tris–
EDTA (TE) buffer (pH 7.4). It was stored at 4 8C until use. PMPC–PDPA
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Table 1. Copolymer compositions calculated using
1
H NMR in DCl/
CD
3
OD for copolymers 1and 2and DCl/D
2
O at pH 2 for copo-
lymers 3and 4. GPC analyses were conducted in a 3:1 chloroform/
methanol mixture in the presence of 5 mM LiBr using poly(methyl
methacrylate) standards.
Copolymer Composition Mn, GPC Mw=Mn
1 PMPC
25
–PDPA
70
22 000 1.35
2 PMPC
25
–PDPA
90
37 800 1.18
3 PMPC
25
–PDPA
120
55 000 1.25
4 PMPC
25
–PDPA
160
74 000 1.23
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 517

polymersome solutions containing different concentrations of
plasmid DNA were diluted 1 in 5 with PBS buffer, and then
100 mL of each diluted sample was mixed with 100 mLPicoGreen
reagent in TE buffer. After gentle mixing and 5 min incubation at
roomtemperature,each samplewas transferredin 2 90 mLaliquots
(to perform the measurements in duplicate) to a 96 well plate. The
plate was analysed using a fluorescence-based microplate reader
(FL
X
800, Biotek Instruments, Inc.) at an excitation wavelength of
485 nm andan emission wavelength of 528nm for detectionof DNA
boundto PicoGreen reagent. Themicroplate reader was setup to take
readings at the top of each well and the sensitivity was set at 100. A
calibration graph wasobtained which was found to be linear up to a
DNA concentration of 0.5 mgmL
1
(which corresponded to a
concentration of 5 mgmL
1
before dilution of the polymersomes
with PBSand PicoGreen reagent). The calibration graph showed very
little dependency on the concentration/type of PMPC–PDPA
polymersomes used, presumably due to the relatively low concen-
tration of polymersomes used. This calibration graph was used to
measure concentrations of plasmid DNA encapsulated in PMPC
25
–
PDPA
70
,PMPC
25
–PDPA
90
and PMPC
25
–PDPA
120
polymersomes.
Transmission Electron Microscopy (TEM) Sample
Preparation, Imaging and Analysis
Copper grids were prior to the sample preparation covered with a
thin coat (10–20 nm) of carbon film, using a Carbon Coater. These
grids were typically glow discharged for 30–40 s, to increase their
hydrophilicity, and therefore their affinity for attachment of the
polymer- /DNA-based samples. The grids were submerged into a
solution of the sample of interest for 60 s, blotted dry, and finally
submerged into a solution of a staining agent for 20 s, before being
blotted dry, and briefly dried using a small vacuum pump. This
technique stains the samples by positive staining. A suitable
staining agent was chosen depending on both the nature of the
sample, and the required pH. An aqueous uranyl acetate solution
(1 w/v %, made up using deionised water) was used to stain samples
which required visualisation at pH values below the copolymerpK
a
of 6.4. Uranyl acetate was also used to stain samples of plasmid
DNA. Phosphotungstic acid solution (0.75 w/v %, made up using
deionised water) was used to stain samples of PMPC–PDPA
polymersomes at neutral pH, since it is the more suitable staining
agent for pH-sensitive samples, especially within this pH range.
Imaging was performed using either a FEI Tecnai Spirit TEM
operating voltage max 120 kV equipped with a Gatan 1k MS600CW
CCD camera, or a Philips CM100 instrument operating at 100 kV,
also equipped with a Gatan 1k CCD camera.
Images obtained from the TEM were analysed digitally using
DigitalMicrograph Demo software (Gatan). The images taken of
plasmid DNA, both alone and in the presence of PMPC–PDPA
copolymers, were processed using the fast Fourier transformation
(FFT) technique. A region of interest in the ‘real space’ micrograph
was converted into a Fourier space image using the ‘reduced FFT’
function. A mask was then applied to the Fourier space image to
filter out any background noise and the image was inverted back to
reveal a clearer picture of the ‘real’ image. This procedure was also
carried out on an area of background to ensure the introduction of
artefacts was avoided. These ‘cleaner’ images were then used to
measure the thickness of the strands of plasmid DNA and deduce
whether the copolymer was bound to the DNA.
Dynamic Light Scattering (DLS)
Dynamic light scattering (DLS) measurements were performed on a
Brookhaven Instruments 200SM laser light scattering goniometer
using a He-Ne 125 mW 633 nm laser. Aqueous polymersome
solutions were diluted, if necessary, with filtered PBS to a
concentration of 1 mgmL
1
and placed into glassvials. Singlescans
of five minutes exposure were performed and particle sizes were
estimated usingthe CONTIN method of dataanalysis at angles of 458,
708,908and1208,unless otherwisestated.Foranalysisof thecolloidal
stabilityofthe PMPC–PDPApolymersomeswithtime,thecorrelation
functions recorded at a 908angle to the incident laser beam were
analysed. Measurements were recorded on day 2 (within 24 h of
polymersome preparation), day 3 (within 48 h of polymersome
preparation), day 6, day 9, day 14, day 21 and finally day 50.
z-Potential Analysis
PMPC–PDPA polymersomes were prepared in the presence of
plasmid DNA to form solutions at approximately pH 7.3 at a [DPA]/
[phosphate] (mol/mol) ratio of 100:1. The polymersomes were
diluted in 150 mMPBS to give a final concentration of 0.01 w/w %
(0.1 mgmL
1
copolymer in PBS). A solution of the plasmid DNA
alone (1 mgmL
1
) was also prepared. The z-potentials of all
solutions were subsequently measured as a function of the
solution pH using a Malvern ZetaSizer Nano ZS.
Ethidium Bromide Displacement Assay to Measure
the Binding Strength of PMPC–PDPA Copolymers to
Plasmid DNA as a Function of pH
Ethidium bromide (2 mg) was dissolved in PBS (1 mL), the pH of
which had been adjusted to pH 7, and the solution was transferred
to a quartz fluorescence cuvette. The fluorescence of the solution
was measured at an ex l¼212 nm and em l¼590 nm using a Cary
Eclipse Varian fluorimeter. pEGFP DNA construct (10 mg) was then
added to the solution and the fluorescence was measured after
stirring the sample for half an hour. 5 mL aliquots of PMPC–PDPA
copolymer solution were titrated into the ethidium bromide–DNA
solution, thereby obtaining [DPA]/[phosphate] (mol/mol) ratios of
0.2, 0.4, 0.6, 0.8, 1, 2, 3, 4, 5 and 10:1. [DPA] refers to the concentration
of tertiary amine groups on the copolymer DPA chains, and
[phosphate]refers to the concentration ofphosphateson the plasmid
DNA. Each solution was mixed gently before the fluorescence was
measured. The above was repeated under conditions of pH 6.8, 6.6,
6.4, 6.2 and 6.0. The relative fluorescence, I/I
0
, was calculated
as follows, adapting a previously reported assay:
[39]
I=I0¼½Iobs:IEtBr þcopolymer=½IDNAþEtBrIEtBrþcopolymer:
(1)
Results and Discussion
TEM and DLS Analysis of Polymersomes Formed by
PMPC–PDPA Copolymers at Physiological pH
Transmission electron microscopy (TEM) and dynamic light
scattering (DLS) analyses were performed on PMPC–PDPA
H. Lomas et al.
518
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

copolymers at pH 7.3 in 150 mMPBS to assess the
morphology of structures generated under these condi-
tions. The results displayed in Figure 2 show that at this pH
the copolymers self-assemble into colloidally stable,
nanometer-sized polymersomes. The typical particle size
distribution ranges from 60–400 nm, with the general trend
of a longer hydrophobic PDPA block resulting in a slightly
larger polymersome diameter, following treatment of the
polymersomes by sonication for 10–20 min. These results
are in agreement with those obtained in previously
published work on the PMPC
25
–PDPA
70
,
[3b]
PMPC
25
–
PDPA
120[2a]
and PMPC
25
–PDPA
160[2a]
copolymers. Due to
the removal of water during sample preparation for TEM
analysis, the collapsed polymersomes can be observed
clustering together in the micrographs, appearing as
‘deflated balloons’. The removal of water is also the reason
why the observed particle sizes are slightly smaller in the
TEM images compared with the DLS data. The application of
(ultrasound) energy to the system results in a narrower, less
polydisperse particle size distribution. Extrusion techni-
ques can also be used to exert control over polymersome
size, by passing the samples through a polycarbonate
membrane containing pores of a specific diameter.
Figure 2(e) shows the particle size distribution for each of
the PMPC–PDPA polymersomes, collected when the detec-
tor was placed at 908to the incident laser beam. Data were
also collected at angles of 458,708and 1208to the incident
radiation, and the subsequent particle size distributions
acquired from the correlation functions using the CONTIN
method of data analysis. By plotting the reciprocal of the
relaxation time, t, of the data peak which corresponds to the
most abundant particle diameter, versus the square of
the scattering vector, q, an indication of the sphericity of the
solute particles can be gained from the linearity of the
graph: A linear graph is a satisfactory sign that the solution
contains a monodisperse distribution of spherical colloidal
particles. In the Supporting Information figure S1, 1/t
versus q
2
was plotted for each of the PMPC–PDPA
polymersome dispersions from particle size distributions
recorded 28 days after their initial preparation. These
polymersomes were prepared by film rehydration in
150 mMPBS at low pH (2.5), increasing the pH to 7.3
and sonicating the resulting polymersome dispersion for
10 min. Samples were then purified by preparative GPC and
stored at 4 8C. The figure shows that an approximately
linear relationship is obtained for each of the samples,
providing evidence for a dispersion of spherical colloidal
particles with a relatively narrow size distribution, i.e., a
monodisperse sample of polymersomes.
PMPC–PDPA polymersome shelf-life was tested over a
50 day period, following a method used previously by our
group.
[3b]
The DLS correlation functions were recorded at
different time points, with very little variation observed
for PMPC
25
–PDPA
70
(Figure 3a) and PMPC
25
–PDPA
160
(Figure 3d) suggesting these polymersomes had main-
tained their integrity as locally isolated, kinetically trapped
colloidal particles when stored at 4 8C for up to 50 days.
However, PMPC
25
–PDPA
90
(Figure 3b) and PMPC
25
–PDPA
120
(Figure 3c) polymersome dispersions showed some varia-
tion in the recorded correlation functions. A shift of the
correlation function to a longer relaxation time was
observed at a few of the time points, indicating a smaller
particle diffusion coefficient and therefore an increase in
the average particle size. This suggests that the polymer-
somes had begun to aggregate together to form higher order
phases such as interconnected vesicles or hexagonally
packed vesicles.
[40]
This phenomenon occurs as a function
of copolymer concentration.
[41]
Once a certain critical
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 2. (a)–(d) TEM micrographs and (e) DLS particle size distri-
butions, recorded at pH 7.3 and in 150 mMPBS, of polymersomes
formed from PMPC–PDPA block copolymers with varying PDPA
block length. Both sets of data show that vesicles are the pre-
dominant colloidal aggregate present at this pH for all four
PMPC–PDPA copolymers investigated. The TEM samples were
positively stained using a 0.75% w/v solution of phosphotungstic
acid prior to analysis. The DLS data were collected at 908to the
incident laser beam and the particle size distribution acquired
using the CONTIN method of data analysis.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 519

concentration is reached (which varies depending on the
molecular weight of the copolymer) the polymersomes are
no longer stable as locally isolated entities and sponta-
neously assemble into higher order phases. It may be the
case that the concentrations used to prepare the PMPC
25
–
PDPA
90
and PMPC
25
–PDPA
120
polymersomes (4 mgmL
1
and 3 mgmL
1
, respectively) are too high to impart
colloidal stability to the polymersomes, whilst for the
PMPC
25
–PDPA
160
copolymer, a lower concentration of
2mgmL
1
was applied. Increasing the copolymer
molecular weight reduces the maximum copolymer con-
centration that can be used to form colloidally stable
polymersomes, thus the above concentrations were
selected initially.
Interactions of PMPC–PDPA Copolymers with Plasmid
DNA at Varying pH
From previous experiments carried out on the PMPC
25
–
PDPA
70
copolymer by the Battaglia group,
[3a]
it is known
that this copolymer has a high binding affinity for plasmid
DNA (3.8 kbp) below its pK
a
, forming a complex with DNA
via the electrostatic binding of individual copolymer chains
to the DNA phosphate backbone. Changing the pH has a
significant effect on the strength of this interaction and
thus the morphology of the copolymer self-assembled
structure in the presence of pDNA. Above the copolymer
pK
a
, the tertiary amine groups on the PDPA chains become
deprotonated, rendering the PDPA block uncharged and
causing it to lose its high binding affinity for DNA. The
increase in pH triggers the self-assembly of the amphiphilic
block copolymer into polymersomes, and a large proportion
of the DNA becomes entrapped within the polymersomes’
aqueous lumen (up to 55% encapsulation efficiencies have
consistently been obtained in recent experimental work
carried out by our group, upon the application of extra
energy via sonication – see Supporting Information figure
S2). This is a physical encapsulation rather than a binding
phenomenon. At physiological pH the copolymer yields
colloidally stable vesicles, whilst at pH 6, copolymer–DNA
complexes are generated.
[3a,3b]
From this evidence, it was predicted that a greater
number of positively charged amine groups per mole of
PMPC–PDPA copolymer as a result of increasing the mole
H. Lomas et al.
Figure 3. DLS correlation functions recorded at various time intervals after the initial preparation of (a) PMPC
25
–PDPA
70
, (b) PMPC
25
–PDPA
90
,
(c) PMPC
25
–PDPA
120
and (d) PMPC
25
–PDPA
160
polymersomes. In graphs (a) and (d), there is very little change to the correlation function
indicating that these copolymers form a dispersion of colloidally stable polymersomes that can be stored for up to 50 days at 4 8C. However,
in graphs (b) and (c) some deviations to the observed correlation function were obtained. The mean polymersome diameter was typically
200–300 nm for each of the block copolymers.
520
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

fraction of the PDPA block would result in a greater
copolymer binding affinity for plasmid DNA. It was
hypothesised that copolymer–DNA complexes with
improved stability at weakly acidic pH would initiate a
higher DNA encapsulation efficiency within polymersomes
upon increasing the pH to 7.3. If the loading efficiency of a
certain drug/nucleic acid can be optimised, this reduces the
concentration of potentially toxic delivery vehicle needed
to transport the required dosage of the particular cargo to its
target site. Greater delivery of DNA into the cytoplasm of
the target cells would in turn result in a higher probability
that a few of the DNA molecules manage to avoid
degradation by the nucleases present in the cytosol and
enter the nucleus where they can exert their therapeutic
effect. There are numerous examples in the literature of
non-viral vectors being used to deliver DNA intracellularly.
However, what many of these current systems lack is a high
transfection efficiency, usually due to the toxicity of the
vector (e.g., Lipofectamine and calcium phosphate).
[3a]
PMPC–PDPA polymersomes are already proven to be non-
toxic and non-immunogenic to HDF cells.
[3b]
Part of the
work described here is based on improving the pDNA
loading efficiency within these polymersomes, in order to
optimise the transfection efficiency.
Firstly, to prove that an increasing PDPA block length
results in a greater number of protonated amine groups per
mole of PMPC–PDPA copolymer below its pK
a
,z-potential
measurements were performed on aqueous copolymer–
DNA solutions as a function of pH (Figure 4). A [DPA]/
[phosphate] ratio of 100:1 mol/mol was used and measure-
ments were taken within the pH range 5.5–7.3. This ratio
was chosen since it was used in previous pDNA encapsula-
tion and transfection studies on the PMPC
25
–PDPA
70
copolymer.
Above the PDPA pK
a
of 6.4, a zero z-potential was
obtained for all the copolymer solutions investigated in the
presence of DNA: The zwitterionic PMPC block is net neutral
and therefore does not contribute to the overall z-potential.
The PDPA block is deprotonated within this pH range, and
consequently does not contribute either. At the 100:1
[DPA]/[phosphate] molar ratio used the concentration of
phosphate ions is too low compared with the DPA
concentration to contribute to the overall z-potential.
Below the PDPA pK
a
, a positive z-potential was acquired
for each of the copolymer solutions, due to the protonation
of the PDPA tertiary amine groups. The greater the length of
the PDPA block relative to the PMPC block, the greater the
number of moles of free amine groups per mole of
copolymer, and subsequently the more positive the z-
potential. Thus below pH 6.4, the z-potential of the
copolymer–DNA solution containing PMPC
25
–PDPA
120
copolymer was higher than that of PMPC
25
–PDPA
90
copolymer, which in turn was higher than for PMPC
25
–
PDPA
70
(Figure 4).
The next step was to carry out ethidium bromide
displacement assays to determine the binding strength
of the copolymers to plasmid DNA as a function of pH and
PDPA block length, using the method previously reported
for the PMPC
25
–PDPA
70
copolymer.
[3a]
For each of the
PMPC
25
–PDPA
70
, PMPC
25
–PDPA
90
, PMPC
25
–PDPA
120
and
PMPC
25
–PDPA
160
copolymers, some pH-dependence
regarding the binding strength of the copolymer to DNA
can be observed, with the general trend of a lower pH, i.e.,
below the copolymer pK
a
of 6.4, resulting in a greater
copolymer–DNA bond strength. As previously stated, this is
due to the protonation of the PDPA amine groups below its
pK
a
and therefore their higher propensity for forming
electrostatic interactions with the anionic DNA phosphates
compared to when they are deprotonated. In general, up to a
[DPA]/[phosphate] ratio of 1:1, a large decrease in the
relative fluorescence emission intensity, corresponding to
increased ethidium bromide displacement as a result of
copolymer–DNA binding, was observed. This was followed
by a more gradual decline in emission intensity, until a
plateau value was attained, at a [DPA]/[phosphate] ratio of
5:1 (data not shown).
However, in the experiments performed on each of the
alternative PMPC–PDPA copolymers, the degree of observed
pH-dependence was not as significant as in the case of the
PMPC
25
–PDPA
70
copolymer.
[3a]
It appeared that some of
the ethidium bromide bound to DNA was sequestered at
the centre of the DNA coils, preventing its complete
displacement. This was particularly the case for copolymers
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 4. z-potential analysis of aqueous PMPC–PDPA copolymer
solutions prepared in the presence of plasmid DNA (3.8 kbp) at a
[DPA]/[phosphate] ratio of 100:1mol/mol, and a solution of the
plasmid DNA alone as a control. Above the pK
a
of the PDPA block
(approximately 6.4), a zero z-potential was obtained for all the
copolymer solutions in the presence of DNA. Below the PDPA pK
a
,
a positive z-potential was acquired for each of the copolymer
solutions, which increased with increasing PDPA block length.
N.B. Analysis of the PMPC
25
–PDPA
160
copolymer was not carried
out.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 521

with increasing PDPA block length. This is illustrated more
clearly by plotting the % ethidium bromide displacement as
a function of pH at a [DPA]/[phosphate] molar ratio of 1:1
(Figure 5). At 50% displacement, a plateau appeared to be
reached in the case of the PMPC
25
–PDPA
90
, PMPC
25
–
PDPA
120
and PMPC
25
–PDPA
160
copolymers, and lowering
the pH further did not result in greater displacement,
regardless of the copolymer used. For example, upon the
addition of PMPC
25
–PDPA
120
at [DPA]/[phosphate] 1:1,
50% displacement was obtained at pH 6.6. However,
performing experiments at pH values lower than this
did not raise the % displacement above 50%, even though
z-potential data showed that decreasing the pH below
6.6 resulted in protonation of this copolymer, until a plateau
level of ionisation was reached at pH 6.1 (Figure 4).
N.B. There is no obvious indication as to why the pK
a
of
the PMPC
25
–PDPA
120
copolymer should appear higher in
Figure 4 than that of the PMPC
25
–PDPA
90
copolymer (in
contrast to the zeta potential data shown in Figure 3, where
the pK
a
for each of the copolymers analysed appeared to be
around 6.3 under the equal experimental conditions that
were applied). Further, in future work, this assay should be
performed at pH 5.5; these conditions would ensure the
analysis is being performed below the block copolymer pK
a
.
The interactions between pDNA and the PMPC–PDPA
copolymers at pH 6 have been explored by TEM
microscopy of positively stained samples using 1% w/v
aqueous uranyl acetate. This was undertaken at a [DPA]/
[phosphate] 1:1 mol/mol ratio for each of the copolymers.
The strands of pDNA were clearly visualised, both in
samples containing solely DNA and in the presence of the
PMPC–PDPA copolymers (Figure 6).
Using fast FFT techniques, the images were transformed
in order to reduce the background noise and gain a clearer
picture of the DNA strands in order for their width to be
measured and ascertain whether complexation between
the individual copolymer chains (unimers) and the DNA
could be observed (Figure 6). The thickness of double
stranded DNA is approximately 4 nm but it was predicted
that with the copolymer bound to the phosphate groups on
both bases within a DNA base pair, this value could be
doubled to approximately 8 nm. 100 DNA strand thickness
measurements were made per sample, over two digital
micrographs each. From these measurements, the mean
and median average values for each sample were
calculated, and are presented in Figure 6 f. From this
preliminary analysis, the ds-pDNA strands appeared to be,
on average, thicker in the presence of the PMPC–PDPA
copolymers at [DPA]/[phosphate] 1:1 mol/mol compared to
the sample containing only DNA. Any ‘nucleation centres’
(i.e., the dark centres which indicate overlapping of the DNA
strands and copolymer chains) observed were also included
in the measurements; these were more frequent in the
presence of the copolymers. In the presence of the PMPC
25
–
PDPA
70
copolymer, the mean DNA strand thickness was
virtually doubled from 4.25 nm for the DNA alone to
8.15 nm upon the addition of the copolymer, suggesting
complexation between DNA and this copolymer. When
PMPC–PDPA copolymers with longer PDPA block lengths
were added to the DNA solution, the increase in the mean
strand thickness was not as great, consistent with the
theory that increasing the PDPA block length decreases the
pK
a
of the copolymer (see below), therefore at pH 6, there is
not such a great extent of tertiary amine group protonation,
thereby reducing DNA binding affinity.
A further important observation from the current data is
that the DNA strands appeared to be more coiled in the
presence of PMPC
25
–PDPA
70
copolymer (Figure 6b) com-
pared to in the absence of copolymer (Figure 6a). A selection
of ‘long rod’ and ‘loose ring’ structures was observed in the
presence of the PMPC
25
–PDPA
70
copolymer, which are very
similar in appearance to structures observed under TEM
analysis of PMPC
30
–PDMA
20
/DNA complexes (DMA ¼2-
(dimethylamino)ethyl methacrylate) in work conducted by
Chim et al.
[42]
The ‘spaghetti-like’ structures observed in the
samples of DNA mixed 1:1 [DPA]/[phosphate] with
PMPC
25
–PDPA
90
(Figure 6c), PMPC
25
–PDPA
120
(Figure 6d),
and PMPC
25
–PDPA
160
(Figure 6e) block copolymers, are
comparable to those observed by Chim et al. on mixing
PMPC
30
–PDMA
10
copolymer with plasmid DNA.
[42]
In this
H. Lomas et al.
Figure 5. % Ethidium bromide displacement by PMPC–PDPA copo-
lymers of varying PDPA block length upon their titration into a
solution of plasmid DNA which was initially bound to ethidium
bromide. The data shown were collected at a [DPA]/[phosphate]
ratio of 1:1, within a pH range of 6.0–7.0. Data points have only
been included for experiments that have been performed in
duplicate for the PMPC
25
–PDPA
90
, PMPC
25
–PDPA
120
and
PMPC
25
–PDPA
160
copolymers. For the PMPC
25
–PDPA
70
copolymer,
the data shown were obtained from previously reported
results.
[3a]
N.B. The error bars shown represent the standard error
of the mean, and a DNA concentration of 10 mgmL
1
was applied
in all cases.
522
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

latter study, the effect of increasing the length of the
cationic (PDMA) block relative to the PMPC block on the
extent of plasmid DNA binding and subsequent condensa-
tion was investigated. The authors suggest that the
‘spaghetti-like’ structures obtained for
the PMPC
30
–PDMA
10
copolymer are
indicative of ‘incomplete DNA conden-
sation’, and as the relative length of the
cationic PDMA block was increased, the
‘linear structures became progressively
thicker and shorter’, indicative of ‘suc-
cessive folding and twisting of DNA,
hence more efficient DNA condensa-
tion’.
[42]
This is the opposite of the results
obtained here in this preliminary study
on the PMPC–PDPA copolymers. For
longer PDPA blocks, i.e., for the
PMPC
25
–PDPA
90
, PMPC
25
–PDPA
120
and
PMPC
25
–PDPA
160
copolymers, the
degree of DNA condensation appeared
to be lower than for PMPC
25
–PDPA
70
.A
possible explanation for this is that due
to the high concentrations of DNA used
in this study (1.5 mgmL
1
) it was
necessary to work with very small
volumes, the pH of which was difficult
to measure accurately. In any case, by
working at pH 6, the samples were
prepared very close to the PDPA pK
a
.
As the PDPA pK
a
is approached, the
degree of protonation of the PDPA chains
is reduced, which in turn reduces the
binding affinity of these chains for DNA.
A further important point is that it is
described in the literature on research
work performed on PMPC–PDPA copo-
lymers by Zhao and coworkers that a
larger PDPA block results in a lower
copolymer pK
a
,
[43]
with the exact value
of the pK
a
depending on the salt
concentration.
[37]
At high ionic strength,
the charged copolymer structures are
stabilised, moving the position of the
equilibrium to favour the production of
charged structures.
[37]
This facilitates
the formation of charged copolymer
species at a higher pK
a
, at high salt
concentrations. If the ionic strength of
the media is lowered, the pK
a
of the
PMPC–PDPA copolymer is also lowered.
Modification of the copolymer pK
a
with
changing PDPA block length would
clearly explain the apparent lack of
significant copolymer–DNA complexa-
tion observed in the TEM micrographs for the copolymers
with a longer PDPA block (Figure 6d and e), and would also
provide an explanation as to why the degree of copolymer–
DNA complexation at a [DPA]/[phosphate] 1:1 mol/mol
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 6. TEM digital micrographs, transformed using fast FFT to reduce background
noise, for samples of plasmid DNA prepared at pH 6 (a) alone (diluted in PBS), and in the
presence of the following copolymers: (b) PMPC
25
–PDPA
70
, (c) PMPC
25
–PDPA
90
, (d)
PMPC
25
–PDPA
120
, and (e) PMPC
25
–PDPA
160
. A concentration of 1.5 mgmL
1
was used
with respect to the pDNA, the [DPA]/[phosphate] molar ratio was 1:1, and (f) Mean and
median ds-pDNA strand thicknesses measured from TEM digital micrographs of samples
prepared at pH 6 in the presence of PMPC–PDPA copolymers. Two digital micrographs
were analysed per sample, and 50 measurements were taken per micrograph. The error
bars represent the standard deviation.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 523

ratio was incomplete at pH 6.0 in the analysis by ethidium
bromide displacement assays (Figure 5). The % ethidium
bromide displacement was especially low for the PMPC–
PDPA copolymer with the longest PDPA block length
(PMPC
25
-PDPA
160
), whereby even at pH 6.2, the % ethidium
bromide displacement was just 25%.
In the experiment described here a DNA concentration of
1.5 mgmL
1
was used, but in the literature, typically
concentrations of 10–50 mgmL
1
are applied,
[42,44]
i.e., two
orders of magnitude lower. Moreover, raising the DNA
concentration (at constant monomer:nucleotide molar
ratio) has been found to affect the structure of the complex
formed. In research conducted by Chim et al.,
[42]
a higher
proportion of more compact/‘highly condensed’ complexes
(e.g., ‘partial rods’, ‘toroids’ and ‘rectangular block’-like
structures) was observed with increasing DNA concentra-
tion. Furthermore, working at as high a DNA concentration
as 1.5 mgmL
1
is not an accurate reflection of the in vitro /
in vivo situation. This concentration was initially chosen to
aid visualisation of the complexes, but for future work a
lower concentration would be sufficient.
The effect of the PMPC block on DNA condensation
should also be considered. The ability of the PMPC block to
impart steric stability to copolymer–DNA complexes,
thereby enhancing their colloidal stability, has been
noted.
[39,42]
Despite this, various studies have shown that
the PMPC block can impede DNA condensation:
[39,42]
The
PMPC block itself occupies a lot of physical space compared
to other steric stabilisers such as PEG. Besides this, due to its
highly hydrophilic nature, each MPC residue can attract a
minimum of 12 water molecules,
[45]
whilst PEG has just
three water molecules associated with each monomer
unit.
[46]
Compared with PEG monomers, MPC therefore
covers more spatial volume and hence provides a greater
steric barrier, which may be the cause of impeded DNA
condensation. It is therefore obvious that a number of
different factors need to be considered when tuning a DNA
delivery vector for the required degree of DNA condensa-
tion. Too long a PMPC block may hinder DNA condensation
by creating a steric barrier around copolymer–DNA
complexes, whilst the degree of DNA condensation is
typically raised by increasing the relative block length of
the cationic polyelectrolyte comprising the copolymer.
Whilst it is necessary to take advantage of the ability of a
copolymer to efficiently condense DNA in order to protect
the DNA from intracellular enzymatic degradation, the
strength of the copolymer–DNA bond must not be so high
that the DNA cannot be released from the vector.
One should also consider the fact that TEM sample
preparation requires dehydration of the samples, and since
the PMPC block is highly hydrated in its native state, this
dehydration may result in alteration of the morphology of
formed DNA–copolymer complexes. A supplementary
technique such as liquid AFM, where the structures can
be observed closer to their native conditions, would be
beneficial. Imaging by cryo-TEM would also be useful,
since unlike TEM, the use of staining agents to visualise
specimens is not necessary, and moreover, samples are
immediately vitrified using a cryogen such as liquid
ethane upon their loading onto TEM grids, and can therefore
be imaged in their native, hydrated states.
DNA Release from its Vector: Estimation of % Plasmid
DNA Released from PMPC
25
-PDPA
70
Copolymer upon
Increasing the pH from 6.0 (i.e., Endocytic pH) to pH
7.3 (i.e., Physiological pH)
It was investigated whether pDNA initially encapsulated
within the aqueous cores of PMPC
25
–PDPA
70
polymersomes
could be re-encapsulated inside the polymersomes after the
pH had been dropped to pH 6.0, kept at pH 6.0 for 5 min, and
then increased again to pH 7.3 (see schematic representa-
tion in Figure 7a). The purpose of this was to attempt to
reconstruct the conditions the DNA-loaded polymersomes
would experience once taken up by cells. A timescale of
5 min was chosen for this initial investigation, since
previous work performed on PMPC
25
–PDPA
70
polymer-
somes encapsulating an amphiphilic fluorescent dye
showed that these polymersomes appeared to be able to
escape the endocytic pathway within this timescale due to
localisation of the dye with the cell cytosol.
[3b]
Additional
studies have provided proof that PMPC–PDPA polymer-
somes are internalised by cells by an endocytic pathway,
[5c]
meaning that once taken up, they become contained within
an internal cell organelle known as an early endosome. The
early endosomal lumen is kept at pH 5.5–6.5 via the influx of
hydrogen ions. This drop in pH (from neutral outside the
cell) would trigger each polymersome to dissociate into
thousands of single copolymer chains. The PDPA block of
the copolymer can bind to DNA within this pH range, as
discussed above. The increase in particle number inside the
endosome triggered by the drop in pH in turn results in an
osmotic pressure build-up within the endosome, which is
thought to cause its eventual lysis and subsequent release
of its contents into the cell cytosol.
[3a]
Once the copolymer–
DNA complexes are freed into the cytosol it is presumed
that the majority of the bound DNA will be released as the
DPA chains become deprotonated at the pH of 7.3 inside
the cell cytosol. The deprotonation of the DPA chains
will render this block hydrophobic, thereby making the
overall PMPC–PDPA block copolymer amphiphilic and
triggering its self-assembly into polymersomes. This self-
assembly is a rapid and spontaneous process, resulting
in the formation of the most thermodynamically
stable structure under the given conditions. As the
copolymer macromolecules re-assemble into polymer-
somes (and presumably also become inserted into the
membranes of the internal cell organelles), the majority of
H. Lomas et al.
524
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

the DNA bound to the copolymer at low pH is
presumed to be freed.
The aim of this experiment was to predict the %
of DNA that would be released from the
copolymer as the endosomal contents are
released into the cell cytosol, and thus foretell
the % that would be re-captured inside the
polymersomes. It is obviously crucial that as
much DNA is released as possible, so that there is a
higher probability of some of the DNA molecules
escaping degradation by the nucleases present in
the cytosol and entering the nucleus, where
transcription and translation into encoded pro-
tein can take place.
Following the method described in previous
work published by our group,
[3b]
the addition of
DAPI to polymersome solutions was used to
calculate % DNA encapsulation efficiencies. The
fluorescence emission intensity corresponding to
DAPI bound to DNA was measured (see Figure 7a
for a schematic representation of this, and
Supporting Information Figure S4 for the
recorded fluorescence emission spectra). The %
plasmid DNA re-captured inside the polymer-
somes following reduction in the solution pH to
6.0 for 5 min was typically found to be less than
10%, meaning that the majority of DNA is
released from copolymer–DNA complexes and
not re-encapsulated if the polymersomes are not
further processed by extrusion and/or sonication
(see Figure 7b).
Encapsulation of Plasmid DNA within
PMPC–PDPA Polymersomes
As stated above, our group has previously
reported
[3b]
a method which uses the probe DAPI
to detect DNA using fluorescence techniques.
DAPI can form a strong bond to DNA as a result of
its aromatic, planar structure and abundance of
amino groups. Furthermore, it can permeate the
polymersome membrane. However, one of the
major drawbacks of this technique is the fact that
DAPI was found to display fluorescence proper-
ties even in the absence of DNA, although the
fluorescence intensity is lower.
Recently, a more sensitive technique has been
implemented by our group to efficiently deter-
mine pDNA concentration and hence measure its
encapsulation efficiency within PMPC–PDPA
polymersomes. This technique uses Quant-iT
PicoGreen dsDNA reagent (‘PicoGreen’), which
is a fluorescent probe that can detect as low as
picomolar concentrations of double stranded
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 7. (a). Schematic representation of the experiment to assess % DNA
release. The initial concentration of pDNA loaded within PMPC–PDPA poly-
mersomes was determined by the addition of DAPI to an aliquot of the sample,
and use of an appropriate calibration graph
[3b]
. The upper spectrum in the
figure above was recorded upon the addition of DAPI to an aliquot of the initial
sample of DNA-loaded polymersomes. The pH of the polymersome samples
was then lowered to pH 6.0, causing the polymersomes to dissociate, as
illustrated in the cartoon, and generate, in some cases, copolymer–DNA
complexes. The pH was maintained at pH 6 for 5 min, and then increased
to pH 7.3. As illustrated in the figure, some of the DNA present was re-
encapsulated within the polymersomes as they formed. Purification of these
polymersomes by preparative GPC removed any unencapsulated DNA, and
allowed determination of the % encapsulation efficiency, by measuring the
fluorescence emission spectrum upon the addition of DAPI. The % DNA
released could subsequently be calculated. N.B. Please note that the spectra
for samples before and after the pH drop and increase were recorded at
different sample concentrations (i.e., different dilution factors were applied).
(b). % DNA initially encapsulated within PMPC
25
–PDPA
70
polymersomes at
[DPA]/[phosphate] 100:1 mol/mol and % of the DNA originally added that is
still encapsulated within the polymersomes following a pH reduction to 6 and
increase back to pH 7.3.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 525

nucleic acid. PMPC–PDPA polymersome solutions contain-
ing different concentrations of pDNA were mixed with
PicoGreen reagent and analysed using an excitation
wavelength of 485 nm and an emission wavelength of
528 nm for detection of pDNA bound to PicoGreen reagent.
A calibration graph was acquired which was found to be
linear up to a DNA concentration of 0.5 mgmL
1
(which
corresponded to a concentration of 5 mgmL
1
before
dilution of the polymersomes with PBS and PicoGreen
reagent). The calibration graph showed very little depen-
dence on the concentration or type of PMPC–PDPA
polymersomes used (see Supporting Information Figure
S5 a), presumably due to the dilution of the polymersomes
prior to analysis, and therefore the low concentration of
copolymer in the analysed samples. The mean calibration
graph (Figure 8a) can therefore be used to measure
concentrations of DNA encapsulated in PMPC
25
–PDPA
70
,
PMPC
25
–PDPA
90
, PMPC
25
–PDPA
120
and PMPC
25
–PDPA
160
polymersomes, up to a DNA concentration of 0.5 mgmL
1
.
The advantages of this method are (i) the small volume of
polymersomes required (10 mL before dilution) due to the
sensitivity of the technique, (ii) a very low concentration of
pDNA can be detected and (iii) the technique can be carried
out within a relatively short time period (less than 1 h)
without the necessity to dissociate the polymersomes prior
to taking the measurements.
Following the protocol used to gain optimal plasmid DNA
loading efficiency within PMPC
25
–PDPA
70
polymer-
somes,
[3b]
DNA encapsulation experiments were conducted
using each of the alternative PMPC–PDPA polymersomes.
Plasmid DNA (commercial pmaxGFP, 3.5 kbp) was added
dropwise to the clear copolymer solutions at pH 6.0, and
the pH was increased to pH 7.3 to generate DNA-loaded
polymersomes. Following purification by preparative GPC,
the calibration graph displayed in Figure 8a was used to
calculate DNA concentrations, and after taking into account
the dilution factors, the concentration of DNA present could
be calculated (see Supporting Information Figure S6 a–c).
The % encapsulation efficiencies obtained for each of the
PMPC–PDPA polymersomes (PMPC
25
–PDPA
70
, PMPC
25
–
PDPA
90
and PMPC
25
–PDPA
120
) are presented in Figure 8b.
This was found to be as high as 64% in the case of the
PMPC
25
–PDPA
70
polymersomes and 63% for the PMPC
25
–
PDPA
90
polymersomes. For the PMPC
25
–PDPA
120
polymer-
somes, the encapsulation efficiency was slightly lower
(virtually 50%).
The above data indicate that these PMPC–PDPA block
copolymers with longer PDPA block lengths show potential
as gene delivery vectors, with polymersome sonication
resulting in a relatively high % encapsulation efficiency of
plasmid DNA (50% or above). The procedure for preparing
DNA-loaded PMPC–PDPA polymersomes has thus been
optimised since previous publications
[3a,3b]
by this group,
and loading efficiencies of 50–65% are consistently
obtained within the PMPC
25
–PDPA
70
polymersomes during
routine experiments.
DLS and TEM analyses of the PMPC–PDPA polymersomes
encapsulating pDNA were carried out on GPC-purified
samples at pH 7.3 (Figure 9 i and ii). The polymersomes
investigated were (as stated above) (a) PMPC
25
–PDPA
70
,
(b) PMPC
25
–PDPA
90
and (c) PMPC
25
–PDPA
120
. Initial analy-
sis of the DLS data by measuring the diameter by intensity
revealed that for all three of the PMPC–PDPA copolymer
samples, in the DLS profiles measured at 908to the incident
radiation beam, particles with a hydrodynamic diameter
greater than 1 mm were detected (data not shown); however
H. Lomas et al.
Figure 8. (a) Calibration graph for the calculation of pDNA concentration within PMPC–PDPA polymersomes by the addition of Quant-iT
PicoGreen dsDNA reagent. The figure displays the mean calibration graph, taking into account the data points collected for known DNA
concentrations in the presence of each of the PMPC–PDPA polymersomes, (n¼4 for each data point). As shown in the figure, up to a pDNA
concentration of 0.5 mgmL
1
, there is a linear relationship between the fluorescence emission intensity at 528 nm and the pDNA
concentration. The equation of this graph can thus be used to measure the % pDNA encapsulation efficiency within each of the
polymersomes. (b) % pDNA loading efficiencies observed for PMPC
25
–PDPA
70
, PMPC
25
–PDPA
90
, and PMPC
25
–PDPA
120
polymersomes, using
the equation of graph (a) to calculate the concentration of DNA encapsulated.
526
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

vesicles were clearly the main morphology present. It is
essential that samples of polymersomes are free from
aggregation and self-assembly into higher order lyotropic
phases when treating cells, as any polymersome aggrega-
tion could induce cell cytotoxicity. Furthermore, the rate of
particle uptake by endocytic pathways is related to particle
size.
[5c]
In order to analyse the data acquired and establish the
effect of adding DNA to a solution of the PMPC–PDPA
copolymer, particle size distributions were obtained by
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Figure 9. (i) DLS and (ii) TEM analysis of GPC-purified samples of pDNA-loaded PMPC–PDPA polymersomes, for (a) PMPC
25
–PDPA
70
,
(b) PMPC
25
–PDPA
90
, and (c) PMPC
25
–PDPA
120
copolymers. The DLS particle size distributions were measured at an angle of 908to the
incident laser beam, and were determined using the CONTIN method of analysis, using the observed range of diameters by number (rather
than diameter by intensity) to analyse the data. These particle size distributions were each compared to a typical sample of the
polymersomes prepared in the absence of pDNA, to ensure that the presence of DNA did not adversely affect the self-assembly of the
copolymer into colloidally stable polymersomes. Samples for TEM analysis were positively stained using 0.75% w/v phosphotungstic acid.
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 527

measuring the CONTIN particle diameter distribution by
number rather than by intensity. Since the particle diameter
by intensity is a measure of the cumulative intensity, larger
particles appear to be relatively more abundant compared
with smaller particles, therefore the presence of micro-
metre-sized aggregates may mask the presence of some of
the nanometre-sized aggregates, or at least reduce the
intensity of these peaks. The particle diameter by number
was utilised to obtain the particle size distributions given in
Figure 9 (i) (recorded at 908to the incident laser beam) and
Supporting Information figures S7 (i), (ii) and (iv) (recorded
at angles of 458,708and 1208to the incident laser beam,
respectively).
Once the larger aggregates had been removed by
analysing the DLS data in this way, the effect of the DNA
on the ability of the copolymer to self-assemble into
nanometer-sized polymersomes could be assessed more
qualitatively. The data below indicate that compared to a
sample containing the empty polymersomes, the presence
of the DNA does not have a detrimental effect on the
copolymer self-assembling properties, and a distribution of
polymersome sizes similar to that obtained for the
unloaded polymersomes was obtained.
TEM analysis of each of the samples (Figure 9 ii) revealed
a larger particle size distribution than that obtained by DLS,
since the aggregates were removed from the DLS analysis to
prevent these from masking the peaks corresponding to the
polymersomes (due to the reasons described above).
Despite this presence of aggregates, the micrographs show
that vesicles are the main morphology present.
Conclusion
Herein we have demonstrated the ability of PMPC
m
–PDPA
n
copolymers (where ‘m’ and ‘n’ are the mean degrees of
polymerisation; m¼25 and n¼70–160) to generate poly-
mersomes at physiological pH and salt concentration via
the technique of copolymer film rehydration at acidic pH,
followed by sonication of the polymersomes upon increas-
ing the pH to 7.3. From the data obtained it can be deduced
that the colloidal stability of these polymersomes is
dependent on both the copolymer concentration and
molecular size. Studies on various polymersome-forming
block copolymers reported in the literature have shown
that as the block copolymer concentration surpasses a
critical minimum value, which varies depending on the
copolymer molecular weight, interactions between the
isotropic phases (i.e., the polymersomes) occur.
[41]
To keep
the free energy of the system at a minimum, these
interactions result in the formation of higher order
mesophases,
[41,47]
known as lyotropic phases. The poly-
mersomes pack to form interconnected or hexagonally
packed vesicles, bicontinuous phases and finally lamellae
(stacked membranes), with increasing copolymer concen-
tration.
[41,48]
Studies on PEO-PBO block copolymers have
confirmed that it is not only the copolymer concentration
that determines the type of self-assembly that is formed;
the molecular weight of the amphiphile also plays a
significant role.
[40,48c]
Dispersions of locally isolated poly-
mersomes were generated by low molecular weight block
copolymers even at relatively high concentrations, whereas
those of a higher molecular weight initially produced
‘peculiar vesicular gel clusters’,
[48c]
before ultimately
breaking up to yield dispersed vesicles. This information
should therefore be taken into consideration when select-
ing a suitable concentration for a particular amphiphile
molecular size.
Despite z-potential analysis confirming that an increas-
ing molar fraction of the PDPA block results in a greater
number of protonated amine groups below the copolymer
pK
a
of 6.4, it is inconclusive from the evidence gathered so
far that this results in a greater PMPC–PDPA binding affinity
for plasmid DNA. Ethidium bromide displacement assays
which were carried out for each of the copolymers at
varying pH and [DPA]/[phosphate] molar ratio did not yield
the results that were expected, since both the extent of
ethidium bromide displacement and the copolymer pK
a
are
dependent on the salt concentration, which was sometimes
difficult to control precisely when preparing samples at
different pH conditions.
An improved technique to assess % DNA encapsulation
efficiency within PMPC–PDPA polymersomes has been
developed, using Quant-iT PicoGreen dsDNA reagent. In this
initial investigation, relatively high % pDNA encapsulation
efficiencies within PMPC
25
–PDPA
70
, PMPC
25
–PDPA
90
, and
PMPC
25
–PDPA
120
polymersomes were obtained, at circa
50% or above. DLS and TEM analysis of these pDNA-loaded
samples confirmed the self-assembly of these copolymers
into polymersomes even in the presence of DNA, although
some larger aggregates were also observed. This initial
encapsulation data are very promising since in previous
work
[3a]
conducted by the Battaglia group a high transfec-
tion efficiency was obtained using PMPC
25
–PDPA
70
poly-
mersomes to transfect HDF cells, compared with the use of
Lipofectamine or calcium phosphate. Furthermore, the high
transfection efficiency obtained was not compromised by
observed cytotoxicity. In future work, the cytotoxicity and
transfection efficiency of pDNA-loaded PMPC–PDPA poly-
mersomes should be tested on HDF cells as a function of
PDPA block length.
The potential ability of PMPC–PDPA polymersomes to
release their DNA payload within the physiological pH
conditions of the cell cytosol was investigated by mimick-
ing the conditions of endocytic release in terms of the pH
changes. Reducing the pH from physiological (i.e., pH 7.3) to
endocytic (pH 6) and then raising the pH back to
cytosolic pH (pH 7.3) caused approximately 90% of initially
H. Lomas et al.
528
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083

encapsulated plasmid DNA to be released from its
polymersome vector. This is essential for therapeutic
applications, since the DNA needs to ideally be dissociated
from its vector in order to penetrate the nuclear membrane.
Many of the gene delivery vectors in the literature require
the use of endosomolytic agents in order to escape the
endocytic pathway, and/or degradable linkages in order to
release the nucleic acid from the vector. With the PMPC–
PDPA polymersomes, such requirements are not necessary,
due to the pH-responsiveness of the PDPA block. This aids
release from the endocytic pathway via the influx of H
þ
ions
into the endocytic vesicle triggering the dissociation of each
PMPC–PDPA polymersome into thousands of copolymer
unimers. This increase in particle number within the
endocytic vesicle is thought to cause the osmotic pressure
within it to build up, leading to its rupture and release of the
contents into the cell cytosol. Some of the pDNA molecules
may already be free following polymersome dissolution,
and some may have become bound to the PDPA chains
within the copolymer as they became protonated under the
weakly acidic pH conditions. Either way, release from the
endosome would cause the PDPA block to lose its charge,
and it is hypothesised that this would trigger spontaneous
release of the majority of bound DNA.
Acknowledgements: Professor Anthony J. Ryan is gratefully
acknowledged for access to dynamic light scattering facilities.
This work was supported by funding from the EPSRC, Biocompa-
tibles UK Ltd, and the Department of Engineering Materials,
University of Sheffield. SPA is the recipient of a 5-year Royal
Society/Wolfson Research Merit Award. HL is funded by
the EPSRC.
Received: February 21, 2010; DOI: 10.1002/mabi.201000083
Keywords: diblock copolymers; encapsulation; pH-sensitive;
plasmid DNA; vesicles
[1] B. M. Discher, Y.-Y. Won, D. S. Ege, J. C.-M. Lee, F. S. Bates, D. E.
Discher, D. A. Hammer, Science 1999,284,1143.
[2] [2a] J. Du, Y. Tang, A. L. Lewis, S. P. Armes, J. Am. Chem. Soc.
2005,127,17982; [2b] C. Zheng, L. Qiu, K. Zhu, Polymer 2009,
50,1173; [2c] F. Ahmed, R. I. Pakunlu, A. Brannan, F. Bates,
T. Minko, D. E. Discher, J. Controlled Release 2006,116,150;
[2d] F. Ahmed, R. I. Pakunlu, G. Srinivas, A. Brannan, F. Bates,
M. L. Klein, T. Minko, D. E. Discher, Mol. Pharm. 2006,3,340.
[3] [3a] H. Lomas, I. Canton, S. MacNeil, J. Du, S. P. Armes, A. J.
Ryan, A. L. Lewis, G. Battaglia, Adv. Mater. 2007,19,4238;
[3b] H. Lomas, M. Massignani, A. K. Abdullah, I. Canton,
C. LoPresti, S. MacNeil, J. Du, A. Blanazs, J. Madsen, S. P. Armes,
A. L. Lewis, G. Battaglia, Faraday Discuss. 2008,139,143;
[3c] Y. Kim, M. Tewari, J. D. Pajerowski, S. Cai, S. Sen,
J. Williams, S. Sirsi, G. Lutz, D. E. Discher, J. Controlled Release
2009,134,132.
[4] [4a] G. Battaglia, A. J. Ryan, Angew. Chem., Int. Ed. 2006,45,
2052; [4b] P. P. Ghoroghchian, P. R. Frail, K. Susumu,
D. Blessington, A. K. Brannan, F. S. Bates, B. Chance, D. A.
Hammer, M. J. Therien, Proc. Natl. Acad. Sci. USA 2005,102,
2922; [4c] P. P. Ghoroghchian, J. J. Lin, A. K. Brannan, P. R. Frail,
F. S. Bates, M. J. Therien, D. A. Hammer, Soft Matter 2006,2,
973; [4d] W. Mueller, K. Koynov, K. Fischer, S. Hartmann,
S. Pierrat, T. Basche, M. Maskos, Macromolecules 2009,42,357.
[5] [5a] G. Battaglia, A. J. Ryan, J. Am. Chem. Soc. 2005,127,8757;
[5b] A. Blanazs, M. Massignani, G. Battaglia, S. P. Armes, A. J.
Ryan, Adv. Funct. Mater. 2009,19,2906; [5c] M. Massignani,
C. LoPresti, A. Blanazs, J. Madsen, S. P. Armes, A. L. Lewis,
G. Battaglia, Small 2009,5,2424.
[6] G. Cevc, D. Marsh, Phospholipid Bilayers: Physical Principles
and Models, Vol. 5, John Wiley & Sons, New York 1987.
[7] S. Jain, F. S. Bates, Macromolecules 2004,37,1511.
[8] G. Battaglia, A. J. Ryan, S. Tomas, Langmuir 2006,22,4910.
[9] H. Bermudez, A. K. Brannan, D. A. Hammer, F. S. Bates, D. E.
Discher, Macromolecules 2002,35,8203.
[10] G. Srinivas, D. E. Discher, M. L. Klein, Nat. Mater. 2004,3,638.
[11] A. D. Bangham, M. M. Standish, J. C. Watkins, J. Mol. Biol. 1967,
13,238.
[12] F. Ahmed, D. E. Discher, J. Controlled Release 2004,96,37.
[13] D. E. Discher, A. Eisenberg, Science 2002,297,967.
[14] R. A. L. Jones, Nat. Mater. 2004,3,209.
[15] P. J. Photos, L. Bacakova, B. Discher, F. S. Bates, D. E. Discher,
J. Controlled Release 2003,90,323.
[16] P. Broz, S. M. Benito, C. Saw, P. Burger, H. Heider, M. Pfisterer,
S. Marsch, W. Meier, P. Hunziker, J. Controlled Release 2005,
102,475.
[17] [17a] F. Meng, Z. Zhong, J. Feijen, Biomacromolecules 2009,10,
197; [17b] O. Onaca, R. Enea, D. W. Hughes, W. Meier, Macro-
mol. Biosci. 2009,9,129; [17c] M.-H. Li, P. Keller, Soft Matter
2009,5,927.
[18] [18a] S. Y. Kim, H. J. Kim, K. E. Lee, S. S. Han, Y. S. Sohn, B. Jeong,
Macromolecules 2007,40,5519; [18b] A. Rank, S. Hauschild,
S. Forster, R. Schubert, Langmuir 2009,25,1337; [18c] S. Qin,
Y. Geng, D. E. Discher, S. Yang, Adv. Mater. 2006,18,2905;
[18d] J. S. Lee, W. Zhou, D. Zhang, C. Otto, J. Feijen, J. Controlled
Release 2008,132,e28.
[19] [19a] H. Kukula, H. Schlaad, M. Antonietti, S. Fo¨rster, J. Am.
Chem. Soc. 2002,124,1658; [19b] C. Fernyhough, A. J. Ryan,
G. Battaglia, Soft Matter 2009,5,1674.
[20] A. Napoli, M. Valentini, N. Tirelli, M. Mu¨ ller, J. A. Hubbell, Nat.
Mater. 2004,3,183.
[21] G. R. Robbins, M. Jimbo, J. Swift, M. J. Therien, D. A. Hammer,
I. J. Dmochowski, J. Am. Chem. Soc. 2009,131,3872.
[22] S. D. Patil, D. G. Rhodes, D. J. Burgess, AAPS J. 2005,7,61.
[23] C. E. Thomas, A. Ehrhardt, M. A. Kay, Nat. Rev. Genet. 2003,4,
346.
[24] T. Hollon, Nat. Med. 2000,6,6.
[25] I. A. Khalil, K. Kogure, H. Akita, H. Harashima, Pharmacol. Rev.
2006,58,32.
[26] [26a] P. L. Felgner, T. R. Gadek, M. Holm, R. Roman, H. W. Chan,
M. Wenz, J. P. Northrop, G. M. Ringold, M. Danielsen, Proc.
Natl. Acad. Sci. USA 1987,84,7413; [26b] P. L. Felgner, G. M.
Ringold, Nature 1989,337,387; [26c] V. Vijayanathan,
T. Thomas, T. J. Thomas, Biochemistry 2002,41,14085.
[27] D. M. Morgan, V. L. Larvin, J. D. Pearson, J. Cell Sci. 1989,94,
553.
[28] D. Fischer, T. Bieber, Y. Li, H. P. Elsasser, T. Kissel, Pharm. Res.
1999,16,1273.
[29] [29a] A. Mecke, I. J. Majoros, A. K. Patri, J. R. Baker, Jr, M. M.
Banaszak Holl, B. G. Orr, Langmuir 2005,21,10348;
Efficient Encapsulation of Plasmid DNA in pH-Sensitive ...
Macromol. Biosci. 2010,10,513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.mbs-journal.de 529

[29b] S. Hong, P. R. Leroueil, E. K. Janus, J. L. Peters, M.-M.
Kober, M. T. Islam, B. G. Orr, J. R. Baker, Jr, M. M. Banaszak Holl,
Bioconjugate Chem. 2006,17,728.
[30] M. Burke, Chem. Britain 2003,39,36.
[31] C. W. Poulton, L. W. Seymour, Adv. Drug Delivery Rev. 2001,46,
187.
[32] J. G. Fewell, M. Matar, G. Slobodkin, S.-O. Han, J. Rice,
B. Hovanes, D. H. Lewis, K. Anwer, J. Controlled Release
2005,109,288.
[33] [33a] D. Oupicky, K. A. Howard, C. Konak, P. R. Dash,
K. Ulbrich, L. W. Seymour, Bioconjugate Chem. 2000,11,
492; [33b] R. Duncan, K. Ulbrich, Macromol. Chem., Macromol.
Sym. 1993,70/71,157.
[34] S. Hwang Pun, M. E. Davis, Bioconjugate Chem. 2002,13,630.
[35] [35a] D. L. Elbert, J. A. Hubbell, Annu. Rev. Mater. Sci. 1996,26,
294; [35b] M. Ogris, S. Brunner, S. Schuller, R. Kircheis,
E. Wagner, Gene Ther. 1999,6,595.
[36] D. W. Bartlett, M. E. Davis, Bioconjugate Chem. 2007,18,456.
[37] C. Giacomelli, L. Le Men, R. Borsali, J. Lai-Kee-Him, A. Brisson,
S. P. Armes, A. L. Lewis, Biomacromolecules 2006,7,817.
[38] K. L. Robinson, J. V. M. Weaver, S. P. Armes, E. D. Marti, F. C.
Meldrum, J. Mater. Chem. 2002,12,890.
[39] J. K. W. Lam, Y. Ma, S. P. Armes, A. L. Lewis, T. Baldwin,
S. Stolnik, J. Controlled Release 2004,100,293.
[40] G. Battaglia, A. J. Ryan, Macromolecules 2006,39,
798.
[41] T. Smart, H. Lomas, M. Massignani, M. V. Flores-Merino,
L. RuizPerez, G. Battaglia, Nano Today 2008,3,38.
[42] Y. T. A. Chim, J. K. W. Lam, Y. Ma, S. P. Armes, A. L. Lewis, C. J.
Roberts, S. Stolnik, S. J. B. Tendler, M. C. Davies, Langmuir
2005,21,3591.
[43] X. Zhao, Z. Zhang, F. Pan, T. A. Waigh, J. R. Lu, Langmuir 2008,
24,6881.
[44] L. Wan, Y. You, Y. Zou, D. Oupicky, G. Mao, J. Phys. Chem. B.
2009,113,13735.
[45] [45a] K. Ishihara, H. Nomura, T. Mihara, K. Kurita, Y. Iwasaki,
N. Nakabayashi, J. Biomed. Mater. Res. 1998,39,323;
[45b] K. Konno, K. Kurita, Y. Iwasaki, N. Nakabayaschi,
K. Ishihara, Biomaterials 2001,22,1883.
[46] T. P. Smart, O. O. Mykhaylyk, A. J. Ryan, G. Battaglia, Soft
Matter 2009,5,3607.
[47] T. P. Smart, C. Fernyhough, A. J. Ryan, G. Battaglia, Macromol.
Rapid Commun. 2008,29,1855.
[48] [48a] A. Blanazs, S. P. Armes, A. J. Ryan, Macromol. Rapid
Commun. 2009,30,267; [48b] G. Battaglia, A. J. Ryan, Nat.
Mater. 2005,4,869; [48c] C. LoPresti, H. Lomas, M. Massignani,
T. Smart, G. Battaglia, J. Mater. Chem. 2009,19,
3576.
H. Lomas et al.
530
Macromol. Biosci. 2010,10, 513–530
ß2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim DOI: 10.1002/mabi.201000083