Page 1 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
1
Ligand-modified human serum albumin
2
nanoparticles for enhanced gene delivery
3
Jennifer Look†#, Nadine Wilhelm‡#, Hagen von Briesen‡, Nadja Noske§, Christine Günther§,
4
Klaus Langer†, Erwin Gorjup‡*
5 6 7 8 9 10 11 12 13 14 15 16
†
Institute of Pharmaceutical Technology and Biopharmacy, University of Muenster, Corrensstr. 48, Muenster 48149, Germany ‡
Fraunhofer Institute for Biomedical Engineering, Ensheimer Straße 48, 66386 St. Ingbert, Germany §
apceth GmbH & Co. KG, Max-Lebsche-Platz 30, 81377 Munich, Germany
#
Both authors contributed equally to this work.
*
To whom correspondence should be addressed: Erwin Gorjup, Fraunhofer Institute for Biomedical Engineering, Ensheimer Straße, 48, 66386 St. Ingbert, Germany, Tel: +49 (0) 6894/980-274, Fax: +49 (0) 6894/980-185,
[email protected] ACS Paragon Plus Environment
1
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 40
17
ABSTRACT
18
The development of non-viral gene delivery systems is a great challenge to enable safe gene
19
therapy. In this study, ligand-modified nanoparticles based on human serum albumin (HSA)
20
were developed and optimized for an efficient gene therapy. Different glutaraldehyde
21
crosslinking degrees were investigated to optimize the HSA nanoparticles for gene delivery. The
22
peptide sequence arginine-glycine-aspartate (RGD) as well as the HIV-1 transactivator of
23
transduction sequence (Tat) are well known as promising targeting ligands. Plasmid DNA loaded
24
HSA nanoparticles were covalently modified on their surface with these different ligands. The
25
transfection potential of the obtained plasmid DNA loaded RGD- and Tat-modified nanoparticles
26
was investigated in vitro and optimal incubation conditions for these preparations were studied.
27
It turned out, that Tat-modified HSA nanoparticles with the lowest crosslinking degree of 20%
28
showed the highest transfection potential. Taken together, ligand-functionalized HSA
29
nanoparticles represent promising tools for efficient and safe gene therapy.
30 31
KEYWORDS
32
albumin, nanoparticle, modification, gene delivery, human mesenchymal stem cell
33
ACS Paragon Plus Environment
2
Page 3 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
34
INTRODUCTION
35
Gene therapy is a high-potential therapeutic strategy for the treatment of diseases which are
36
based on a genetic defect. Currently, most clinical trials in gene therapy target cancer. These
37
multifactorial diseases can be cured by gene therapy through delivery of e.g. suicide genes which
38
destroy malignant cells by enzymatic function 1. Genetic modified mesenchymal stem cells
39
(MSC) are suitable to function as a delivery agent for suicide genes in vitro and in vivo in cancer
40
treatment 2-4. In most studies the MSC were genetically modified ex vivo which reduces side
41
effects for the patients. Unfortunately, efficient genetic modification of MSC is current
42
exclusively achieved by viral vectors 5.
43
Virus-based gene delivery is extremely potent and ensures long-term expression of genes 6, 7.
44
However, application in gene therapy is limited due to serious drawbacks like carcinogenicity,
45
immunogenicity, inflammation or high-cost production 8-13. Consequently, non-viral options for
46
gene transfer were developed such as lipoplexes, polyplexes based on poly-L-lysine (PLL) or
47
poly amidoamine (PAMAM), and new nanomaterials such as quantum dots or silica
48
nanoparticles 14-19. Beside the advantages of a low immune response or low cost production in
49
large quantities, most of the non-viral vectors were not applied for gene therapy due to low
50
transfection efficiency or high cytotoxicity in vitro 20, 21. However, human serum albumin based
51
nanoparticles (HSA-NP) are known to be nontoxic, non-immunogenic and biodegradable 22-24. In
52
addition, clinical studies proved the promising use of Abraxane™, a HSA-based nanoparticulate
53
drug, for breast cancer therapy 25. Nevertheless, unmodified HSA nanoparticles are inefficient in
54
gene delivery; mostly due to their negative surface charge, which on the one hand impedes
55
binding of the negatively charged plasmid DNA and on the other hand impedes the cellular
56
uptake of the vectors 26. To overcome this hurdle, Fischer et al. developed cationized HSA
ACS Paragon Plus Environment
3
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 40
57
nanoparticles 27. However, these NP were still not able to sufficiently transfect cells without an
58
additional endosomolytic agent. Combination of HSA with polyethylenimine (PEI), known for
59
its positive effect on endosomal release, led to an increased gene expression in HEK293 cells 28.
60
In contrast, PEI-HSA combined nanoparticles interact with cells nonspecifically, which limits
61
their application in vivo. Therefore, modification of HSA nanoparticles with specific ligands for
62
an efficient uptake seems to be a suitable strategy for non-viral gene delivery. Studies have
63
reported that modification of nanoparticles with the arginine-glycine-aspartate (RGD)-containing
64
peptide led to an efficient transfer across the cell membrane of integrin-positive cells such as
65
B16F10 or HUVEC 29. Moreover, Gojgini et al. showed a RGD-mediated gene delivery in
66
mouse MSC with hyaluronic acid hydrogels in a scaffold for local gene therapy 30. Beside
67
specific ligands, cell penetrating peptides (CPP) like the HIV-1 transactivator of transduction
68
sequence (Tat) facilitate cellular uptake of a large variety of cargos 31-33. Suk et al. demonstrated
69
that modification of PEI/DNA complexes with Tat peptides enhanced gene transfection
70
efficiency up to 14-fold in neuronal cells 34. Nevertheless, most of the studies included cytotoxic
71
nanoparticle formulations such as dendrimers or PEI or showed poor transfection efficiency.
72
The aim of this study was to modify HSA nanoparticles with different ligands like RGD or Tat in
73
order to achieve a non-viral and biocompatible ex vivo gene delivery system for the gene therapy
74
with human mesenchymal stem cells (hMSC). Biocompatibility of the gene delivery system is
75
especially important for a future genetic modification of stem cells due to the fact that
76
mesenchymal stem cells are known to differentiate spontaneously in vitro due to stress situations
77
35, 36
78
In the present study, the efficiency of modified HSA-NP for non-viral ex vivo gene delivery was
79
investigated. HSA-nanoparticles were prepared using an ethanol desolvation method and
.
ACS Paragon Plus Environment
4
Page 5 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
80
characterized the RGD- and Tat-modified delivery systems with regard to particle size, surface
81
charge (zeta potential), ligand binding, and plasmid release. The transfection potential of the
82
different modifications was assessed in varying incubation media and was successfully even in
83
absence of an endosomolytic agent in HEK293T cells. Results were promising and indicate that
84
the study needs to follow up with an optimization of the formulation for non-viral ex vivo gene
85
delivery in the hard-to-transfect hMSC.
86
ACS Paragon Plus Environment
5
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 6 of 40
87
EXPERIMENTAL SECTION
88
Materials
89
Human serum albumin (fraction V) and glutaraldehyde 25% solution were obtained from Sigma
90
(Steinheim, Germany). The succinimidyl ester of methoxy poly (ethylene glycol) hexanoic acid
91
(mPEG5000-SHA) and the crosslinker NHS-PEG5000-Maleinimide (NHS-PEG5000-Mal) were
92
purchased from JenKem Technology (Plano, USA). The peptides RGD and RAD were obtained
93
from Peptides International (Louisville, USA) and HIV Tat 48-60 Cys peptide from Innovagen
94
AB (Lund, Sweden). All chemicals were of analytical grade and used as received.
95
Preparation of surface-modified plasmid-loaded HSA nanoparticles
96
HSA nanoparticles were prepared by a desolvation technique as described previously
97
(Steinhauser et al., 2009). In principle, 1 ml human serum albumin solution (20 mg/ml, pH 6.0)
98
was incubated with 100 µg of the respective plasmid (pEGFP-N1, pcMV-Luc) for 15 min under
99
constant stirring (550 rpm) at room temperature (RT). Nanoparticle preparation was performed
100
by dropwise addition of 2.7 ml ethanol 96% (v/v) at a rate of 1 ml/min under stirring (550 rpm).
101
After the desolvation process the nanoparticles were stabilized by crosslinking with
102
glutaraldehyde. Therefore volumes of 2.36 µl and 11.80 µl of a glutaraldehyde solution 8%
103
(m/v) were added, which corresponds to a theoretical calculated crosslinking degrees of 20% and
104
100%, respectively. It was assumed that a glutaraldehyde concentration of 100% enables a
105
theoretic crosslinking of 60 primary amino groups present in one HSA molecule but does not
106
necessarily lead to a quantitative HSA crosslinking. The crosslinking process was performed for
107
at least 12 h under constant stirring (220 rpm) of the NP suspension at RT. Particles were
108
purified in two cycles by centrifugation (14,000 g, 8 min) and redispersion of the pellet in
ACS Paragon Plus Environment
6
Page 7 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
109
phosphate buffer (pH 8.0). For redispersion a Thermomixer Comfort (Eppendorf AG, Hamburg,
110
Germany) and sonication were used.
111 112
Surface modification with RGD and RAD peptides
113
First step of surface modification with RGD and RAD peptide was the activation of
114
nanoparticles with the heterobifunctional crosslinker NHS-PEG5000-Mal. Therefore, NHS-
115
PEG5000-Mal was dissolved in phosphate buffer (pH 8.0) and added in a 11fold molar excess
116
(0.88 mg per mg NP) to the nanoparticle suspension. After incubation for 1 h in a Thermomixer
117
(600 rpm) at RT the nanoparticles were purified in two cycles by centrifugation at 14,000 g for 8
118
min and redispersion in phosphate buffer (pH 8.0). In the second coupling reaction step an
119
equimolar amount of the deacetylated RGD and RAD peptide was added to the nanoparticles
120
(11.08 µg RGD, 11.29 µg RAD per mg NP), respectively. For deacetylation of the peptides an
121
aliquot of 100 µl of deacetylation solution (0.5 M hydroxylamine, 25 mM EDTA in PBS, pH
122
7.2-7.5) was added to the peptide solution in PBS. After an incubation of at least 12 hours in the
123
Thermomixer (600 rpm, RT) the nanoparticles were purified by centrifugation (14,000 g, 8 min)
124
and redispersion in water. The supernatant after centrifugation was collected for quantification of
125
free peptide.
126
Surface modification with TAT peptide
127
For NP surface modification with Tat peptide the first step of coupling reaction was performed
128
as described above. In brief, the crosslinker NHS-PEG5000-Mal was added in a 11fold molar
129
excess to the purified plasmid-loaded nanoparticles. After incubation and purification the Tat
130
peptide was given to the activated NP in a 1:8 molar ratio (3.5 µg per mg NP). After incubation
131
in the Thermomixer (600 rpm, RT) for at least 12 hours the nanoparticle suspension was purified
ACS Paragon Plus Environment
7
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 8 of 40
132
once and the supernatant collected for quantification of the uncoupled peptide amount. In
133
addition, particles were PEG-modified as a negative control. Therefore, plasmid-loaded
134
nanoparticles in phosphate buffer (pH 8.0) were mixed with a 11fold molar excess of
135
mPEG5000-SHA and purified two times after an incubation of 1 h in the Thermomixer (600
136
rpm, RT).
137
Particle characterization
138
The resulting particle yield after purification was determined gravimetrically. Therefore an
139
aliquot (20.0 µl) of the respective nanoparticle sample was put in micro weighing dishes (VWR
140
International GmbH, Darmstadt, Germany) and dried for 2 h at 80°C. The content of the
141
nanoparticles was calculated from the difference of the empty and the nanoparticle-filled dish.
142
Determination of particle diameter and polydispersity index was performed by photon
143
correlation spectroscopy (PCS) with a Zetasizer Nano ZS (Malvern Instruments GmbH,
144
Herrenberg, Germany). The measurements were carried out at 22°C and a scattering angle of
145
173°. Particle diameter was calculated from the intensity of the scattered light (Z-average).
146
Investigations on the nanoparticle surface charge were measured by laser Doppler
147
microelectrophoresis using the zetapotential mode of the same instrument. Before measurement,
148
the samples were diluted with purified water to a concentration of 0.05 mg/ml.
149
Digestion of plasmid-loaded nanoparticles and agarose gel electrophoresis
150
An aliquot of 2 mg NP with maximum 10 µg plasmid was digested with proteinase K solution
151
followed by plasmid extraction via a silica-based spin column. For these digestion and extraction
152
steps DNeasy® Blood & Tissue Kit (QIAGEN GmbH, Hilden, Germany) was used. Agarose gel
153
(0.6%) was prepared with broad range agarose (Carl Roth GmbH & Co. KG, Karlsruhe,
154
Deutschland) in TAE buffer. The marker, control plasmid and the extracted plasmids from the
ACS Paragon Plus Environment
8
Page 9 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
155
different nanoparticle batches were loaded onto the gel together with 2 µl of the nucleic acid
156
stain SafeWhite (NBS Biologicals Ltd., Cambridgeshire, UK), respectively. Electrophoresis was
157
performed at a constant voltage of 80 V for 1 h in TAE buffer. Bands corresponding to the
158
plasmid were detected under UV light and photographed.
159
Entrapment efficiency of plasmid-loaded HSA nanoparticles
160
The amount of plasmid incorporated into the HSA nanoparticles was measured indirectly in the
161
supernatant after nanoparticle purification. The content of free plasmid DNA was determined
162
using Quant-iTTM PicoGreen® dsDNA Assay Kit (Life Technologies, Eugene, USA) according
163
to the operating instruction. The fluorescence was measured by a microplate reader SynergyTM
164
Mx (BioTek Instruments GmbH, Bad Friedrichshall, Germany) at excitation and emission
165
wavelengths of 480 nm and 520 nm, respectively. The incorporated amount of plasmid in the
166
nanoparticles was calculated by the difference between the used amount per mg NP and the
167
detected free amount of plasmid per mg NP in the supernatants of the purification process.
168
Storage of HSA nanoparticles
169
To analyze the storage stability of the HSA nanoparticles, the presence of free plasmid DNA was
170
analyzed with unmodified 20% and 100% crosslinked nanoparticles under four different
171
conditions over 28 days. The following storage conditions were investigated: purified water at
172
8°C, cell medium DMEM with 10% fetal bovine serum (FBS), cell medium DMEM with 10%
173
heat inactivated FBS and cell medium OptiMEM®, all at 37°C, respectively. The heat
174
inactivation of FBS was performed in a water bath at 56°C for 30 min. Typically, lyophilized
175
nanoparticles were resuspended in the respective medium at a concentration of 2.5 mg/ml and
176
aliquoted in separate microcentrifuge tubes for each time point and then stored at 8°C and 37°C,
177
respectively. At predetermined time intervals the tubes were withdrawn and centrifuged at
ACS Paragon Plus Environment
9
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 10 of 40
178
20,238 g for 30 min. The supernatants were collected and the amount of released DNA was
179
measured fluorimetrically with Quant-iTTM PicoGreen® dsDNA Assay Kit in a microplate
180
reader as described above.
181
Peptide quantification by HPLC analysis
182
The amount of RGD, RAD, and Tat peptide bound to nanoparticle surface was calculated as the
183
difference between the total amount of the initial peptide added and the amount of peptide
184
measured in the supernatant obtained during the purification steps. The peptide amount was
185
determined by a C18-RP-HPLC method on a Gemini® 5 µm NX C18 110 Å column (250 x 4.6
186
mm) with a gradient elution program. As mobile phase 0.1% trifluoro acetic acid in purified
187
water (A) and acetonitrile (B) were used. For the RGD and RAD peptide the column was
188
equilibrated with 90% A and separation was performed at a flow rate of 1.0 ml/min with a linear
189
gradient (eluent A : eluent B) with the following steps: 0 min (90:10), 8 min (50:50), 10 min
190
(90:10), and 13 min (90:10). For the Tat peptide the column was also equilibrated with 90% A
191
and the separation was performed with the steps: 0 min (90:10), 10 min (70:30), 12 min (90:10),
192
and 15 min (90:10). In both cases the injection volume was 20.0 µl. The detection was performed
193
using a diode array detector by measuring the absorbance at 220 nm.
194
Cell culture
195
All cells were cultured in a humidified atmosphere at 5% CO2 and 37°C. Medium was changed
196
twice a week and cells were subcultured at a maximum confluence of 80-90%. Human epithelial
197
kidney (HEK) 293T cells were cultured with culture medium (DMEM, 10% fetal bovine serum
198
(FBS) and 100 U/ml penicillin and 100 µg/ml streptomycin), unless stated otherwise. Human
199
mesenchymal stem cells (hMSC) were isolated from bone marrow of the caput femoris and
200
cultured in α-MEM supplemented with 15% FBS, 100 U/ml penicillin and 100 µg/ml
ACS Paragon Plus Environment
10
Page 11 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
201
streptomycin. Isolated hMSC were characterized by the expression of the surface marker CD29,
202
CD44, CD73, CD90, CD105, CD106 and HLA-ABC and absence of CD34, CD45, CD133 and
203
HLA-DR, as well as by their adipogenic and osteogenic differentiation capacity.
204
Flow cytometry analysis
205
HEK 293T cells and human mesenchymal stem cells were seeded 24 h before nanoparticle
206
incubation at 7.5x104 or 1.5x104 cells/cm2, respectively. Cells were treated with 50 µg
207
nanoparticles per cm2 growth area in fresh culture medium. After incubation for 24 h, cells were
208
washed with PBS and harvested. After fixation by 10 g/l PFA and 8.5 g/l NaCl in PBS, pH 7.4
209
for 30 min cells were analyzed by flow cytometry with 10,000 cells per sample, using
210
FACSCalibur and CellQuest Pro software (Becton Dickinson, Heidelberg, Germany).
211
Nanoparticles could be detected via their autofluorescence at 488/520 nm.
212
CLSM analysis
213
Cells were grown on culture slides (Becton Dickinson, Heidelberg, Germany) and incubated with
214
HSA nanoparticles as described above for binding analysis. After 48 h of incubation, cells were
215
washed with PBS and cytoplasm was stained with CellTracker™ Blue CMAC dye (Life
216
Technologies, Darmstadt, Germany) according to the manufacturer´s instructions. Cells were
217
fixed with ice-cold 70% ethanol for 5 min and covered with VECTASHIELD HardSet Mounting
218
medium (Vector Laboratories, Burlingame, CA, USA). Samples were stored at 4°C until analysis
219
with a TCS SP8 confocal microscope (Leica microsystem, Heidelberg, Germany).
220
Gene expression analysis
221
Cells were seeded 24 h before transfection experiments as described before. Culture medium was
222
replaced with fresh medium, fresh medium supplemented with 0.1 mM chloroquine or
ACS Paragon Plus Environment
11
Molecular Pharmaceutics
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 12 of 40
223
OptiMEM® (Life Technologies, Darmstadt, Germany) and added with 50 µg/cm2 nanoparticles.
224
Incubation medium was changed after 24 h and cells were cultured in culture medium for further
225
48 h. Expression of the reporter gene eGFP was analyzed with a fluorescence microscope (IX71,
226
Olympus, Hamburg, Germany). Luciferase activity was quantified 72 h post-transfection using
227
luciferase assay system (Promega GmbH, Mannheim, Germany) following manufacturer´s
228
protocol. Luciferase activity was measured in relative light units (RLU) using Tecan Infinite 200
229
microplate reader (Tecan, Mainz, Germany).
230
ACS Paragon Plus Environment
12
Page 13 of 40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
231
RESULTS
232
Preparation of peptide-modified plasmid-loaded HSA nanoparticles
233
Plasmid-loaded nanoparticles based on human serum albumin (HSA) were prepared by a well-
234
established desolvation method (Fig. 1A). After incubation of the plasmid with the HSA
235
solution, the protein precipitates with the plasmid DNA due to ethanol addition. Particles were
236
stabilized by crosslinking with two different amounts of glutaraldehyde which led to crosslinking
237
degrees of 20% and 100%, respectively. After purification RGD and Tat peptides, as well as
238
their negative controls RAD and PEG (Fig. 1C-E), were attached to the particle surface.
239
Therefore, in the first reaction step a bifunctional PEG-based crosslinker was used for NP
240
activation. In the second step the thiol-reactive maleinimide part of the crosslinker was reacted
241
with the thiol group containing peptides (Fig. 1B). RGD- and RAD-modified nanoparticles with
242
20% crosslinking degree were obtained in a diameter range of about 250 nm (Table 1). The sizes
243
of RGD- and RAD-modified 100% crosslinked particles were in the range of 190 nm and were
244
significantly smaller (p
