A poly(ethylene) glycolylated peptide for ocular delivery compacts DNA into nanoparticles for gene delivery to post-mitotic tissues in vivo.

Abstract

We have previously shown that a novel synthetic peptide for ocular delivery (POD) can efficiently compact DNA and deliver it to cells in vitro. This observation prompted us to develop use of POD as a nonviral vector in vivo.
POD peptide was modified using poly(ethylene) glycol (PEG-POD) and used to compact DNA into nanoparticles that were then analysed using electron microscopy, dynamic light scattering, and fluorescent labeling. Transfection efficiency and localization were determined 48 h post-injection into the subretinal space of the mouse eye using luciferase and LacZ, respectively. Efficiency of ocular transfection was compared to two other PEGylated peptides: PEG-TAT and PEG-CK30.
PEG-POD can compact DNA and form discrete nanoparticles of approximately 136 nm that can penetrate and transduce the retinal pigment epithelium (RPE) in vivo. PEG-POD significantly increased expression of plasmid DNA by 215-fold, PEG-TAT by 56.52-fold, and PEG-CK30 by 24.73-fold relative to DNA injected alone. In all cases beta-galactosidase was observed primarily in the RPE layer after subretinal injection. Electrophysiological analyses of PEG-POD transduced retina indicates an absence of PEG-POD-mediated toxicity. PEG-POD can protect plasmid DNA from DNaseI digestion, resulting in significant transfection of the lung after intravenous injection in mice.
PEG-POD was found to significantly increase gene delivery relative to both DNA alone and other pegylated peptides. These findings highlight the use of pegylated peptides, and specifically PEG-POD, as novel gene delivery vectors.

Full text

Page 1

THE JOURNAL OF GENE MEDICINE
J Gene Med 2010; 12: 86–96.
Published online 24 November 2009 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/jgm.1415
RESEARCH ARTICLE
A poly(ethylene) glycolylated peptide for ocular
delivery compacts DNA into nanoparticles for gene
delivery to post-mitotic tissues in vivo
Sarah Parker Read
Siobhan M. Cashman
Rajendra Kumar-Singh*
Department of Ophthalmology, Tufts
University School of Medicine, Boston,
MA, USA
*Correspondence to:
Rajendra Kumar-Singh, Department
of Ophthalmology, Tufts University
School of Medicine, 136 Harrison
Avenue, Boston, Massachusetts
02111, USA. E-mail:
rajendra.kumar-singh@tufts.edu
Received: 22 June 2009
Revised: 8 October 2009
Accepted: 13 October 2009
Abstract
Background
ocular delivery (POD) can efficiently compact DNA and deliver it to cells
in vitro. This observation prompted us to develop use of POD as a nonviral
vector in vivo.
We have previously shown that a novel synthetic peptide for
Methods
POD) and used to compact DNA into nanoparticles that were then analysed
using electron microscopy, dynamic light scattering, and fluorescent labeling.
Transfection efficiency and localization were determined 48 h post-injection
into the subretinal space of the mouse eye using luciferase and LacZ,
respectively. Efficiency of ocular transfection was compared to two other
PEGylated peptides: PEG-TAT and PEG-CK30.
POD peptide was modified using poly(ethylene) glycol (PEG-
Results
approximately 136 nm that can penetrate and transduce the retinal pigment
epithelium (RPE) in vivo. PEG-POD significantly increased expression of
plasmid DNA by 215-fold, PEG-TAT by 56.52-fold, and PEG-CK30 by 24.73-
fold relative to DNA injected alone. In all cases β-galactosidase was observed
primarily in the RPE layer after subretinal injection. Electrophysiological
analyses of PEG-POD transduced retina indicates an absence of PEG-POD-
mediated toxicity. PEG-POD can protect plasmid DNA from DNaseI digestion,
resulting in significant transfection of the lung after intravenous injection in
mice.
PEG-POD can compact DNA and form discrete nanoparticles of
Conclusions
relative to both DNA alone and other pegylated peptides. These findings
highlight the use of pegylated peptides, and specifically PEG-POD, as novel
gene delivery vectors. Copyright  2009 John Wiley & Sons, Ltd.
PEG-POD was found to significantly increase gene delivery
Keywords
cell penetrating peptide; gene delivery; nonviral; POD
Introduction
The development of nonviral gene transfer technologies for the treatment of
genetic diseases has seen significant progress over the last 10 years. How-
ever, efficient gene transfer to post-mitotic cells or tissues in vivo, such as the
retina or brain,remains a significant barrier for the widespread use of nonviral
vectorsintheclinic.Retinitispigmentosa(RP)andage-related maculardegen-
eration (AMD) collectively describe a common group of disorders that lead to
Copyright  2009 John Wiley & Sons, Ltd.

Page 2

PEG-POD is a novel gene therapy vector
87
blindness. In RP and AMD, the retinal photoreceptor cells
or the retinal pigment epithelium (RPE) are among the
first group of cells to degenerate. Hence, the protection of
these cells is a target application of nonviralgene delivery.
Recently, we described a novel cell penetrating
peptide for ocular delivery (POD) that can bind the
plasma membrane of human embryonic retinoblasts
in culture and enter the cytoplasm [1]. The design
ofPOD[CGGG(ARKKAAKA)4]
glycosaminoglycan-binding domain of acidic and basic
fibroblast growth factor. When covalently linked to
small dye molecules such as lissamine (0.6 kDa), POD
retained the ability to bind and traverse the plasma
membrane. POD was demonstrated to enter retinal
cells in vivo, including RPE, photoreceptors and ganglion
cells. Although this property of POD may potentially
be useful for the delivery of small molecule drugs to
the retina, the ability to deliver substantially larger
molecules such as DNA (approximately 4.3 MDa) to
post-mitotic cell nuclei in vivo remains a challenge. To
achieve transgene expression in post-mitotic cells in vivo,
POD/DNA complexes would need to overcome the barrier
of the plasma membrane, escape endosomes, avoid
degradation in the cytoplasm and cross the intact nuclear
membrane pore [2]. Although viruses or recombinant
viral gene transfer vectors are very efficient at overcoming
each of these barriers, nonviral vectors have thus far
enjoyed limited success, preventing their widespread use
in gene transfer to post-mitotic tissues in vivo.
Because POD is a positively-charged peptide, we
wished to examine whether POD may compact and
protect DNA from endonucleases at the same time as
retaining its cell-penetrating properties. The addition
of poly(ethylene) glycol (PEG) has been previously
demonstrated to stabilize protein–nucleic acid complexes
[3]. Therefore, we postulated whether DNA compacted
with pegylated POD may form discrete nanoparticles
suitable as a formulation for gene delivery to post-mitotic
tissue in vivo.
wasbasedonthe
Materials and methods
Plasmids, peptide and cell lines
pCAGLuc was cloned by first placing the Luciferase
cDNA from pGL3-Control (Promega, Madison, WI, USA)
into pBluescript II KS (Stratagene, La Jolla, CA, USA)
using HindIII and XbaI. An XhoI/NotI fragment of this
plasmid was then inserted into pCAGEN (kindly provided
by C. Cepko, Harvard Medical School, Boston, MA). A
SmaI/NotIfragmentofpCMVb(Clontech,MountainView,
CA, USA) was cloned into EcoRV/NotI-digested pCAGEN
to generate pCAGLacZ. c-POD, CGGG(ARKKAAKA)4,
was generated at Tufts University Peptide Synthesis
Core Facility (Boston, MA, USA) and purified by
high-performance liquid chromatography (HPLC). TAT
(CGGGGYGRKKRRQRRR) was synthesized and purified
by HPLC by Sigma Genosys (Woodlands, TX, USA).
CK30 (CK30) was synthesized and purified by HPLC by
PolyPeptide Laboratories (Torrance, CA, USA). Human
embryonicretinoblasts (HER)
grown in Dulbecco’s modified Eagle’s medium (DMEM)
supplemented with 10% fetal bovine serum (FBS)
(Invitrogen, Carlsbad, CA, USA).
911 cells[4] were
In vivo delivery of POD/DNA complexes
All experiments involving animals were in accordance
with the Statement for the Use of Animals in Ophthalmic
and Vision Research, set out by the Association for
Research in Vision and Ophthalmology. C57BL/6J and
BALB/cJ mice were purchased from Jackson Laboratories
and maintained under 12:12 h dark/light cycles in
accordance with federal, state, and local regulations. Mice
were anesthetized by intraperitoneal injection of Xylazine
(Xyla-Ject, Phoenix Pharmaceutical, Inc., St Joseph, MO,
USA; 10 mg/ml)/Ketamine HCl (Ketaset, Fort Dodge
AnimalHealth,FortDodge,IA,USA;1 mg/ml).Subretinal
injections were performed with a 32-G needle (Becton
Dickinson, Franklin Lakes, NJ, USA) and a 5 µl glass
syringe (Hamilton, Reno, NV, USA) by a trans-scleral
trans-choroidal approach. Animals were sacrificed by CO2
inhalation followed by cervical dislocation.
c-POD/DNA compaction: in vitro
and in vivo delivery
pCAGLuc was compacted using c-POD as previously
described [1]. Briefly, 1.8 nmol c-POD and 2 µg (0.466
pmol) of DNA were each suspended in 15 µl of 100 mM
sodium phosphate buffer. The DNA and c-POD solutions
were mixed and incubated for 30 min prior to addition
to the media of HER cells. Under serum starvation
conditions, the cells were incubated with serum free
DMEM for 24 h prior to transfection. Under nonserum
starved conditions, the cells were maintained in 10% FBS
until just prior to transfection, at which time the media
was changed to 2% FBS. After 48 h, cells were prepared
for luciferase assay according to the manufacturer’s
protocol (Promega). Live cells were visualized with
light and fluorescent microscopy using an Olympus IX51
with differential interference contrast and appropriate
fluorescent filters (Olympus, Tokyo, Japan). Images
were captured using a Retiga 2000R FAST camera and
QCapture Pro 5.0 (QImaging, British Columbia, Canada).
For use in explants and in vivo injection, c-POD/DNA
complexes were prepared as above using 1.8 nmol c-POD
to compact either 0.233 or 0.466 pmol DNA. Explants
were prepared from 6–8-week-old C57BL/6J wild-type
mice by enucleation and removal of the anterior chamber.
The retina and RPE/sclera were separated, flattened and
placed in 500 µl of 5% CO2equilibrated DMEM with 10%
FBSand100U/mlPen-Strep.Compacted pCAGLuc(2 µg)
or the plasmid alone was added to explants and incubated
for 48 h. For injections in vivo, 0.2 µg of compacted DNA
Copyright  2009 John Wiley & Sons, Ltd.
J Gene Med 2010; 12: 86–96.
DOI: 10.1002/jgm

Page 3

88
S. Parker Read et al.
or naked DNA was injected into the subretinal space
as described above. The eyes were harvested 48 h later
and the anterior chamber removed. In both explants and
in vivo injections, tissue was homogenized using a VWR
PowerMax AHS 200 homogenizer (VWR, West Chester,
PA, USA) in homogenization buffer (50 mM Tris HCl, pH
8.0, 150 mM NaCl) and assayed for luciferase expression
according to the manufacturer’s protocol (Promega) using
a Glomax 20/20 luminometer over a 10-s integration
(Promega). Protein concentration was measured using a
Quick Start Bradford Protein Assay (Bio-Rad, Hercules,
CA, USA) and comparing to bovine serum albumin (Bio-
Rad) standards in corresponding buffer.
Preparation and characterization
of pegylated c-POD, TAT and CK30
nanoparticles
Peptides were resuspended in 0.1M sodium phosphate
(pH 7.2), 5 mM ethylenediaminetetraacetic acid to
form a 20 mg/ml solution. An equimolar amount of
methoxy-PEG-maleimide −10 kDa (Nektar Transforming
Therapeutics, San Carlos, CA, USA) was resuspended in
the same volume of dimethylsulfoxide. PEG was added
dropwise to the peptide over approximately 10 min,
vortexing between drops. The solution was shaken
overnight at room temperature, and dialysed using a
Bio-Gel P6 column (Bio-Rad) into 0.1% trifluoroacetic
acid for POD and TAT and 50 mM ammonium acetate for
CK30 [5]. The pegylated POD (PEG-POD) was quantified
by Coomassie stain of a Tris-HCl gel (Bio-Rad) using
a c-POD standard curve. Both CK30 and TAT are not
within the detectable limits of Coomassie stain, and so
protein concentration was performed using BCA Protein
Assay Kit (Thermo Fisher Scientific, Waltham, MA, USA)
using the un-pegylated peptide to create a standard curve.
DNA was compacted by diluting the plasmid in water to
a final concentration of 0.2 µg/µl and added dropwise
to PEG-POD or PEG-TAT with a final ratio 1.8 nmol
peptide: 2 µg of DNA (0.466 pmol pCAGLuc or 0.367
pmol pCAGLacZ), vortexing to mix. The nanoparticles
were dialysed three times in 5% dextrose using Biomax
10K centrifugal filter (Ultrafree, Millipore Corp., Billerica,
MA, USA) and stored at 4◦C. PEG-CK30 particles were
compacted following a previously described protocol
[5,6], but briefly, 0.9 ml of DNA (200 µg/ml) was added
to 0.1 ml of 7.1 mg/ml PEG-CK30 in 0.1 ml aliquots
over approximately 2 min at 25◦C. The compacted DNA
was dialysed overnight into 0.9% filtered NaCl at 4◦C
(Tube-O-Dialyser, G-Biosciences, St Louis, MO, USA)
and concentrated by centrifugation (Ultrafree, Millipore
Corp.; nominal molecular weight limit 100 kDa). Plasmid
compaction for all peptides was verified by reduced
mobility through a 1% agarose gel (Invitrogen), which
was relieved after 15 min of incubation with 0.25%
Trypsin (Invitrogen) at 37◦C. PEG-POD particles were
analysed by incubation on glow-discharged Formvar
copper grids (Electron Microscopy Sciences, Hatfield, PA,
USA), stained with 0.4% uranyl acetate, and visualized
using CM10 Transmission Electron Microscope (FEI,
Hillsboro, OR, USA) using Digital Micrograph software
(Gatan, Pleasanton, CA, USA). PEG-POD particles and
control samples were diluted to 0.2 mg/ml and analysed
at 25◦C using a BI-200SM research goniometer and
laser light scattering system (Brookhaven Instruments
Corporation, Holtsville, NY, USA) with 532 nm laser light.
Data was collected for 5 min for each sample and the
mean particle diameter determined by quadratic fit using
software supplied by the manufacturer.
Localization of compacted DNA in vitro
Rhodamine labeled DNA (pGeneGrip; Gene Therapy
Systems, San Diego, CA, USA) was compacted with either
c-PODorPEG-POD asdescribed.Compactionswereadded
to HER cells grown to approximately 80% confluency
on Lab Tek-II chamber slides (NUNC, Thermo Fisher
Scientific, Rochester, NY, USA) in serum free DMEM to a
final concentration of 0.66 µg/ml and incubated at 37◦C,
5% CO2for 24 h (for each condition, n = 4). Cells were
fixed with 4% formalin for 15 min at room temperature.
Imaging was performed as above using an Olympus BX51
microscope (for each condition, n = 4).
PEG-peptide/DNA nanoparticle
transfections in vivo
Subretinal injections were performed as above and
luciferase activity assayed as for c-POD compactions.
All injections were performed in a 2 µL total volume.
For tail vein injections, compacted or naked DNA was
injected in a total volume of 200 µl with 0.01% fast
green (Thermo Fisher Scientific) using a 25-G needle
(BectonDickinson).Forty-eighthoursafterinjection,mice
were sacrificed using CO2 and perfused with 800 µl
of PBS through the left ventricle. The organs were
harvested, placed in 500 µl of Luciferase Cell Culture
Lysis Reagent (Promega) with 20 µl/ml of a protease
inhibitor cocktail (Sigma, St Louis, MO, USA) and
homogenized for 20 s. Homogenates were centrifuged at
4◦C for 10 min at 12000 g and assayed for luciferase
activity as per manufacturer’s instructions (Promega)
and protein concentration as mentioned above. Standard
curve values for the luminometer were determined
using QuantiLum Recombinant Luciferase (Promega).
Conversion from relative light units (RLU) to pg was
calculated as: luciferase (pg) = (3.696 × 10−5× RLU)
−0.0815 (R2= 0.9999). For β-galactosidase expression,
1.2 µg of either compacted or naked pCAGLacZ was
injected into the subretinal space of BALB/cJ mice. After
48 h,eyeswereenucleated,fixedin0.25%glutaraldehyde
for 30 min, and washed three times in PBS for 30 min.
The eyes were incubated in X-Gal solution (FisherBiotech,
Fisher Scientific, Fair Lawn, NJ, USA) for 16–18 h and
then rinsed in phosphate buffer (pH 7.4) for 45 min.
Copyright  2009 John Wiley & Sons, Ltd.
J Gene Med 2010; 12: 86–96.
DOI: 10.1002/jgm

Page 4

PEG-POD is a novel gene therapy vector
89
The eyes were fixed for 24 h in 4% PFA, dehydrated,
embedded, and 18-µm sections were collected. Whole
eye bright-field images were taken using a Nikon C-FMC
microscope (Nikon, Tokyo, Japan) prior to sectioning. For
all conditions, n = 4.
Serum stability assay
Seven hundred nanograms of PEG-POD compacted or
naked pCAGLuc was treated with 2.5 U or 0.25 U DNaseI
(Sigma) at 37◦C for 15 min, after which 10 µg of pronase
(Sigma) was added and incubated for a further 10 min at
37◦C. The samples were then run on a 1% agarose gel to
check for DNA degradation.
Electroretinography
Mice were injected into the subretinal space with 1 µl of
PEG-POD∼Luc nanoparticles (700 ng) or 1 µl of 5% dex-
trose buffer and analysed by electroretinography 48 h
later. After overnight dark-adaptation, the mice were
anesthetized as described above, pupils dilated with 1%
Tropicamide (Akorn, Inc., Lake Forest, IL, USA), and sco-
topic electroretinograms recorded at two different flash
intensities (−10 and 1 dB) using contact lens electrodes
and the UTAS system with BigShot ganzfeld (LKC Tech-
nologies, Inc., Gaithersburg, MD, USA). Five to ten flashes
were averaged for the high-low intensities.
Statistical analysis
All data analyses were performed using Prism Software 4
(GraphPad Software, Inc., San Diego, CA, USA). During
analysis of the data, there was shown to be a highly
significant difference (F-test, p < 0.0001) between the
variances of the data sets that was mean dependent.
This precludes analysis by Student’s t-test, which assumes
equal variances between samples [7]. The data (RLU/mg)
were logarithmically transformed to create a Gaussian
distribution, eliminating the variability between data set
variance, and allowing use of the parametric t-test [7]. All
statistical analyses were done using log-transformed data
sets. Although all data points are shown, only those above
an uninjected mean ±3 SD are used in the statistical
analysis and to calculate the means [5].
Results
Pegylated POD compacts DNA
to form discrete nanoparticles
We functionalized the free sulfhydryl bond on the
cysteine (c) of c-POD with a 10 kDa PEG-maleimide
group. Although the molecular weight of c-POD after
synthesis was determined by mass spectroscopy to be
the predicted 3.5 kDa, we observed a unique band
corresponding to approximately 17.5 kDa on a Coomassie
stained gel, suggesting that c-POD forms pentamers in
solution (Figure 1a). Functionalization of this 17.5-kDa
product with 10-kDa PEG generated a gel shift to the
expected 27.5-kDa protein (Figure 1a). A minor product
was also detected at 37 kDa, suggesting that some c-POD
pentamers may be functionalized by two 10-kDa PEG
molecules (Figure 1a). Similar to c-POD [1], pegylated
POD (PEG-POD) compacted DNA and prevented its
migration in an agarose gel (Figure 1b). DNA from such
complexes could be readily released by incubation of
the PEG-POD/DNA complex with trypsin (Figure 1b).
The PEG-POD/DNA complexes were suspended in 5%
dextrose buffer for all experiments.
Closer examination of particle size by dynamic light
scattering (DLS) revealed a size range of 136 ± 27.2 nm.
By contrast, the controls (i.e. buffer or uncompacted
DNA) were determined to be 1.6 ± 0 nm and 1221 ±
753.4 nm, respectively (Figure 1c). The large standard
deviation for DNA is not surprising given that DLS is
primarily accurate for spherical particles and may be
detecting rotational as well as translational motion of the
plasmid DNA. Visualization of PEG-POD complexed with
pCAGLuc plasmid DNA (PEG-POD∼Luc) by transmission
electron microscopy revealed relatively homogeneous
100–175 nm spherical nanoparticles (Figure 1d). In
conclusion, PEG-POD compacts DNA and forms discrete
nanoparticles in solution.
Relative to c-POD, PEG-POD
nanoparticles do not aggregate
and are resistant to DNAseI
One key difference between actively dividing and
quiescent cells in the context of gene delivery is that,
during mitosis, the nuclear membrane disintegrates
and allows easier access of large molecules such as
DNA into the nucleus. The intact nuclear membrane
in post-mitotic cells in vivo is considered to be a
significant barrier to nonviral gene transfer [8]. We
compacted rhodamine-labeled plasmid DNA using c-POD
and found that, even at the level of light-microscopy, large
c-POD/DNA aggregates with a size in the range
10–100 µm were readily visible when incubated with
HER 911 cells in culture (Figure 2a). Unlike c-POD, PEG-
POD did not form large aggregates when complexed with
DNA and incubated with HER cells (Figure 2a).
Serum nuclease activity significantly decreases the half-
life of plasmid DNA. DNA injected intravenously into
mice is degraded within 10 min [9], making protection
from degradation an important property for genes to
be successfully delivered in vivo by intravenous injection.
DNaseI is an endonuclease found in human serum and
many tissues in the body [10] and it has been shown
that serum nuclease activity is primarily derived from
DNaseI [11]. Hence, DNaseI was utilized to test the
potential stability of pegylated POD/DNA nanoparticles
Copyright  2009 John Wiley & Sons, Ltd.
J Gene Med 2010; 12: 86–96.
DOI: 10.1002/jgm

Page 5

90
S. Parker Read et al.
Figure 1. PEG-POD and c-POD can both compact DNA but PEG-POD compacts DNA to form small, discrete nanoparticles. (a) cPOD is
an approximately 3.5 kDa peptide that migrates at approximately 17.5 kDa as a pentamer. The addition of a 10 kDa PEG increased
the molecular weight to primarily a 27.5 kDa protein, with higher species also seen. (b) Both c-POD and pegylated c-POD (PEG-POD)
are able to compact DNA and retard movement during electrophoresis. Compaction is relieved upon degradation of the peptide with
trypsin. (c) DLS indicated a mean average particle radius of 136 ± 27.2 nm for PEG-POD∼Luc nanoparticles, 1.6 ± 0 nm for buffer
alone, and 1221 ± 753.4 nm for the pCAGLuc plasmid alone. DLS shown as mean ± SD. (d) Transmission electron microscopy
revealed that PEG-POD compacted DNA particles were smaller and more spherical and homogeneous in size than uncompacted
plasmid DNA, which formed large irregular aggregates, or c-POD compacted DNA. Size bar, top = 1 µm, bottom = 200 nm
Figure 2. PEG-POD compacted DNA nanoparticles have reduced aggregation and increased protection against DNaseI.
(a) Rhodamine-labeled DNA was complexed with c-POD or PEG-POD and added to HER cells in culture and observed after
24 h. c-POD/DNA complexes form large aggregates in cell culture while PEG-POD/Rhodamine labeled DNA nanoparticles showed
a diffuse punctate pattern across cells. Scale bar = 50 µm. (b) In the absence of pronase, plasmid migration was inhibited by the
compaction of the PEG-POD∼Luc nanoparticles. Nanoparticle DNA was released following a 10-min pronase incubation. Treatment
with a low (0.25 U) or high (2.5 U) dose of DNaseI led to partial and complete digestion of pCAGLuc, respectively. However,
PEG-POD nanoparticles remained largely intact under both conditions
Copyright  2009 John Wiley & Sons, Ltd.
J Gene Med 2010; 12: 86–96.
DOI: 10.1002/jgm