An enhanced autogene-based dual-promoter cytoplasmic expression system yields increased gene expression.

Abstract

The relatively low levels of transfection that can be achieved by current gene-delivery systems have limited the therapeutic utility of gene transfer. This is especially true for nonviral gene-delivery systems, where the levels of gene expression achieved are usually below the levels achieved by viral gene transfer systems. One strategy for increasing gene expression is to design a cytoplasmic expression system that does not require nuclear delivery for gene expression to occur. This can be achieved through the use of an autocatalytic cytoplasmic expression system using phage RNA polymerases. Here we describe cytoplasmic expression systems that yield increased levels of gene expression following in vitro transfection. We demonstrate direct evidence for an exponential, autocatalytic increase in gene expression using autogenes, as well as levels of reporter gene expression that are 20-fold higher than standard CMV-based nuclear expression systems. The development of a high-efficiency plasmid-based expression system could significantly improve the gene expression properties of nonviral gene-delivery systems, thereby increasing their clinical utility.

Full text

Page 1

RESEARCH ARTICLE
An enhanced autogene-based dual-promoter
cytoplasmic expression system yields increased
gene expression
J Finn1, ACH Lee2, I MacLachlan2and P Cullis1,3
1Department of Biochemistry and Molecular Biology, University of British Columbia, Vancouver, BC, Canada;2Protiva Biotherapeutics
Inc, Burnaby, BC, Canada; and3Inex Pharmaceuticals Corp., Burnaby, BC, Canada
The relatively low levels of transfection that can be achieved
by current gene-delivery systems have limited the therapeu-
tic utility of gene transfer. This is especially true for nonviral
gene-delivery systems, where the levels of gene expression
achieved are usually below the levels achieved by viral
gene transfer systems. One strategy for increasing gene
expression is to design a cytoplasmic expression system that
does not require nuclear delivery for gene expression to
occur. This can be achieved through the use of an
autocatalytic cytoplasmic expression system using phage
RNA polymerases. Here we describe cytoplasmic expression
systems that yield increased levels of gene expression
following in vitro transfection. We demonstrate direct
evidence for an exponential, autocatalytic increase in gene
expression using autogenes, as well as levels of reporter
gene expression that are 20-fold higher than standard CMV-
based nuclear expression systems. The development of a
high-efficiency plasmid-based expression system could
significantly improve the gene expression properties of
nonviral gene-delivery systems, thereby increasing their
clinical utility.
Gene Therapy (2004) 11, 276–283. doi:10.1038/sj.gt.3302172
Keywords: cytoplasmic expression; gene expression system; T7 RNA polymerase; bicistronic; plasmid design; autogene
Introduction
A major limitation of gene-delivery systems is the
relatively low level of gene expression in transfected
tissues. One strategy to increase levels of gene expression
following transfection employing a nonviral vector
involves improving the plasmid design. The incorpora-
tion of a cytoplasmic expression system represents one
such approach.1–3Cytoplasmic expression systems by-
pass the requirement for nuclear delivery of plasmid
DNA, a major obstacle in present day gene therapy.4,5–7
In addition, they take advantage of the large number of
plasmids found in the cytoplasm of the cell following
transfection with nonviral vectors.8Cytoplasmic expres-
sion systems can be designed to utilize the unique
properties of the bacteriophage RNA polymerases
(RNAPs).Phage RNAPs
(B100 kDa), single subunit proteins capable of synthe-
sizing RNA from DNA templates. They require no
additional cofactors and have demonstrated efficient
cytoplasmic transcriptional activity.9,10These features
make phage RNAPs attractive candidates for the devel-
opment of autocatalytic cytoplasmic expression systems
using autogenes. Phage RNAP autogenes consist of an
RNAP gene, driven by its own cognate promoter.3In
order to evade the requirement for exogenous RNAP to
initiate the expression system, a nuclear promoter can
aremoderatelysized
be added upstream of the RNAP promoter.11Although
the first round of RNAP expression must occur via
the nuclear promoter, the resulting RNAP in the
cytoplasm drives the cytoplasmic expression system,
producing RNA from plasmid DNA template in the
cytoplasm.
RNA produced in the cytoplasm lacks the 50cap that
stabilizes nuclear transcripts and assists in ribosomal
recruitment.12,13
Viral Internal Ribosome Entry Site
(IRES) elements are sequences that have been shown to
enhance the recruitment of the cytoplasmic translational
machinery in the absence of 50capping.14Early dual-
promoter cytoplasmic expression systems did not con-
tain IRES elements and, as a result, the vast majority of
the mRNA produced was not translated.11Although
an autogene based on the T7 bacteriophage RNAP
that contained an EMCV IRES has been previously
described,15it did not contain a eukaryotic promoter and
required the cotransfection of RNAP protein or mRNA,
thereby limiting its utility.
In this study, we have characterized a novel autogene
construct containing both the EMCV IRES element (to
allow cap-independent translation) and the CMV pro-
moter (to bypass the need for additional exogenous
RNAP protein or mRNA during transfection). In addi-
tion, we have combined the autogene and reporter gene
cassettes on a single plasmid, whereas other systems
have utilized separate plasmids. We also describe a cell-
free system for the evaluation of RNAP-based trans-
fection reagents and use it to demonstrate the potential
of the autogene system to elicit high levels of gene
expression.
Received 02 March 2003; accepted 03 September 2003
Correspondence: Dr J Finn, Department of Biochemistry and Molecular
Biology, University of British Columbia, 2146 Health Sciences Mall,
Vancouver, BC, Canada V6T 1Z3
Gene Therapy (2004) 11, 276–283
& 2004 Nature Publishing Group All rights reserved 0969-7128/04 $25.00
www.nature.com/gt

Page 2

Materials and methods
Plasmids and primers
Plasmid R023 consists (Figure 1) of a basic autogene
cassette driven by a CMV promoter and intron. The
autogene cassette was derived from the plasmid T7-G1, a
gift of Dr Jon Wolff (Waisman Center, WI, USA). T7-G1
contains the basic autogene cassette, consisting of the T7
promoter, EMCV IRES, and T7 RNAP gene. The nuclear
localization sequence was removed from the T7 RNAP
via PCR prior to subcloning into R023. L059 consists of a
pTRI-Amp (Ambion) backbone with EMCV IRES, Photi-
nus pyralis luciferase and beta-globin poly-adenylation
site derived from EMC-Luc (Jon Wolff). L053 consists of
the CMV promoter (with intron) from NGVL3 and the
Photinus pyralis luciferase gene. L069 and L070 consist of
L053 containing one or two irrelevant 2.5 kb spacer
fragments, respectively. R037 consists of R023 without
the T7 and T3 promoters. R011 is a bicistronic plasmid
consisting of R023 with a downstream luciferase reporter
gene cassette from L059 (bicistronic). PT7-Luc (Promega)
consists of the Photinus pyralis luciferase gene driven by a
T7 RNAP promoter. RPA-RNAP consists of a 350 bp
KpnI–AflII T7 RNAP fragment blunted and ligated into
the SmaI site of pTRI-Amp in reverse orientation. RPA-
Luc consists of a 250 bp Xcm1–BsrG 1 luciferase fragment
blunted and ligated into the Sma1 site of pTRI-Amp in
the reverse orientation.
The NVSC1 primer sequence is 50-TCCTGCAGCCC
GGGGGATCCTCTAG-30.
Transcription and translation assay
A 25 ml reaction was set up using a Promega (Wisconsin)
Coupled In Vitro Transcription and Translation kit, as
per the manufacturer’s instructions. A measure of
250 ng of PT7-Luc was added to all reactions. Then,
250 ng of either R023 (containing a T7 RNAP gene driven
by the T7, SP6, and T3 promoters) or R037 (containing a
T7 RNAP gene driven only by the SP6 promoter) was
added to the reactions. SP6 RNAP (0.5 U; Promega)
was added and each reaction was incubated at 301C.
At the time points indicated, 2 ml of reaction mixture
was removed and assayed for luciferase expression,
as described below. All reactions were performed in
triplicate.
Transfections
Lipoplexes were formed by mixing plasmid DNA with
large unilamellar vesicles (LUVs) composed of equimo-
lar amounts of DOPE:DODAC (50:50) on ice and
incubated for 20 min prior to use. All transfections were
performed at a cationic lipid to plasmid DNA charge
ratio of 3:1. Lipoplexes were diluted with serum-contain-
ing media before addition directly to the cell media. BHK
cells were plated at 25 000 cells per well in 24-well plates.
Neuro2A cells were plated at 30 000 cells per well in 24-
well plates. The total mass of plasmid added was
identical in all transfections. Equimolar transfections
using plasmids of different sizes were achieved through
the addition of an empty vector (pBlueScript) to normal-
ize the total mass of DNA in each transfection. All
transfections were performed in triplicate. Data are
presented as mean values7standard error.
Luciferase and BCA assays
Cells were washed twice with 1 ml PBS, followed by the
addition of 0.2 ml lysis buffer (PBS with 0.1% Triton
X-100) before being stored at ?701C. Cells were thawed
and 5–20 ml of sample was assayed in duplicate on a 96-
well plate. Samples were assayed using a Berthhold
Centro LB960 Microplate Luminometer and Luciferase
Assay System (Promega). Standard luciferase assays
were performed and transfection data are reported as
mass quantities of luciferase protein, using a standard
curve obtained from serial 10-fold dilutions of a 20 mg/ml
Photinus pyralis luciferase standard (Promega). Cell-free
luciferase assays are reported in RLUs. Total protein was
quantified using a Pierce BCA assay kit, as per the
manufacturer’s instructions.
Immunofluorescence
BHK cells were plated on glass coverslips in six-well
plates (150 000 cells/well) and transfected with 1.5 mg of
plasmid DNA. At 24 h post-transfection, cells were
washed once with 2 ml PBS-IF (10 mM sodium phos-
phate, 140 mM sodium chloride, pH 7.4) prior to fixation
for 10 min with 2 ml 2% paraformaldehyde. Cells were
subjected to three 30 s washes before permeabilization
with 0.25% Triton X-100 in PBS-IF for 5 min. After
washing three times for 30 s with PBS-IF, cells were
incubated with blocking buffer (10% BSA in PBS-IF) for
1 h, shaking gently at room temperature. Cells were
washed three times for 10 min with PBS-IF, followed by
addition of a primary antibody solution consisting of a
1:1000 dilution of goat anti-T7 RNAP antibody (a gift
from Dr Paul Fisher at the Department of Pharmacolo-
gical Sciences, State University of New York at Stoney
Brook) or 1:1000 dilution of mouse antiluciferase mono-
clonal antibody (Abcam) in 2% BSA in PBS-IF. Cells
were incubated with primary antibody solution for 2 h
while shaking at room temperature. Cells were washed
three times for 10 min in PBS-IF, followed by the addition
Figure 1 Plasmid diagrams of major constructs used. R023 is the basic
autogene construct, containing the T7 RNAP gene driven by the T7, T3,
and SP6 promoters (PTRI). L059 is the luciferase reporter gene cassette.
R011 is the bicistronic autogene construct (R023þL059). L053 is the
CMV-driven nuclear expression construct.
Enhanced cytoplasmic expression system
J Finn et al
277
Gene Therapy

Page 3

of secondary antibody (rabbit anti-goat IgG, FITC
labelled (QED Bioscience Inc) or rabbit anti-mouse Texas
Red labelled (Abcam), 1:200 dilution in 2% BSA-PBS-IF)
and incubation for 2 h while shaking at room tempera-
ture. Cells were washed four times for 10 min with PBS-
IF before being mounted and photographed on a Zeiss
Axiovert S100 fluorescence microscope. The percentage
of cells transfected was determined by counting trans-
fected and nontransfected cells under the microscope.
Data indicate the average of six separate counts from
three different experiments.
RNase protection assay
RNAP and luciferase probes were prepared from
EcoR1-digested RPA-RNAP or RPA-Luc plasmid, respec-
tively. GAPDH probe was purchased from Pharmingen.
Probes were labelled following the manufacturer’s
protocol using
(NEN).
BHK cells were plated on six-well plates (150 000
cells/well) and transfected with 1.5 mg of R011 or L053 in
triplicate. After 24 h, cells were treated with 20 mg/ml
Actinomycin D. At 0, 2, 4, 6 or 8 h after actinomycin D
treatments, cells were washed once with PBS and
recovered by trypsinization. Cells from triplicate wells
were pooled before harvesting total RNA (RNeasy
miniprep kit, Qiagen). A measure of 10, 5 or 2.5 mg of
total RNA was subjected to RNase protection analysis
using the RiboQuant RPA system (Pharmingen), accord-
ing to the manufacturer’s protocol. All values shown are
the average7standard deviation of two independent
experiments. Data were collected using a Typhoon
Phosphoimager (Amersham Biosciences) and analysis
was performed using ImageQuant software (Amersham
Biosciences).
32P-aUTP (3000 Ci/mmol, 10 mCi/ml)
Primer extension
Primer extension analysis using
NVSC1 and 100 mg of RNA isolated from R011-trans-
fected BHK cells (24 h post transfection) was performed
using a Primer Extension System (Promega). The ladder
was prepared by end-labelling FX174 HinfI DNA
markers with
7standard deviation of two independent experiments.
Data were collected as described for RNase Protection
assay above.
32P-labelled primer
32P. All values shown are the average
Results
Autocatalytic gene expression results in an
exponential time-dependent increase in gene
expression
A hallmark of an autocatalytic, self-amplifying system is
an exponential, time-dependent increase in the product
being amplified. This exponential relationship would be
limited only by the amount of substrate available (ie
charged tRNA, GTP, etc), and would continue as long as
the template plasmid is in excess. In order to verify the
autocatalytic nature of the autogene, a cell-free transcrip-
tion and translation assay was performed. R023 plasmid
DNA (consisting of T7 RNAP driven by both SP6 and T7
RNAP promoters) was incubated with a PT7-Luc
reporter gene plasmid (consisting of luciferase driven
by only the T7-promoter) in the presence of rabbit
reticulocyte lysate and SP6 RNAP. SP6 RNAP transcribes
T7 RNAP RNA from the R023 plasmid, leading to the
production of T7 RNAP protein that is then able to drive
the expression of both the T7 RNAP gene from R023 in
an autocatalytic fashion, as well as expression of the
luciferase gene from PT7-Luc. Figure 2 shows a dramatic
increase in luciferase expression over time, indicating an
exponential, autocatalytic increase in T7 RNAP protein.
This increase is not observed when a control plasmid
(R037, consisting of T7 RNAP driven only by the SP6
promoter) lacking the T7 promoter needed for auto-
catalytic amplification is used. The reason for the lack of
expression from R037 is that, without the autocatalytic
amplification, the amount of T7 RNAP produced is not
enough to give rise to detectable levels of luciferase
expression.
Figure 2 (a) Schematic diagram of the transcription and translation assay.
SP6 RNAP binds to the SP6 promoter (PSP6) on R023 (T7 RNAP driven
by SP6 and T7 promoters) (1) transcribing T7 RNAP mRNA, which is (2)
translated into the T7 RNAP protein. The T7 RNAP protein then binds
the T7 promoter (PT7) on R023 (3), resulting in more T7 RNAP protein
(2) and initiating the autocatalytic cycle and an exponential increase in T7
RNAP production. T7 RNAP also transcribes luciferase mRNA from PT7-
Luc (4), resulting in an increase in luciferase expression proportional to the
amount of T7 RNAP present. In the control reaction (below), the lack of
PT7in R037 (T7 RNAP gene driven by only SP6 promoter) prevents any
autocatalytic production of T7 RNAP (3). (b) An in vitro coupled
transcription and translation (Promega) assay. A measure of 250 ng of
PT7-Luc was combined with 250 ng of either R023 or R037 in a total
reaction volume of 15 ml, and 0.5 U of SP6 RNAP (Promega) was added
and incubated at 301C. The aliquots (2 ml) were removed at the time points
indicated and subjected to luciferase analysis, as described in Materials and
methods. After an initial lag phase, the R023 reaction resulted in an
exponential increase in luciferase expression, verifying the autocatalytic
nature of the system.
Enhanced cytoplasmic expression system
J Finn et al
278
Gene Therapy

Page 4

A bicistronic construct results in higher levels of gene
expression than a dual-plasmid transfection
Previously published work on cytoplasmic expression
systems employed an autogene cassette and a reporter
gene cassette on separate plasmids. It was of interest to
compare the expression resulting from a dual-plasmid
transfection system with a single-plasmid bicistronic
system in which the autogene and reporter gene were on
one large plasmid. When equimolar amounts of auto-
gene and reporter gene constructs were used to transfect
BHK cells, it was found that the bicistronic construct
yielded two- to four-fold higher levels of gene expression
than the analogous dual plasmid transfection (Figure 3).
This result was unexpected because previous results
suggest that transfection (delivery to the nucleus
and subsequent expression) would be more efficient for
the smaller autogene plasmid than the larger bicistronic
construct.16For the dual-plasmid transfection, this would
result in a greater number of cells expressing RNAP
via the CMV promoter in the nucleus, and accordingly
greater levels of luciferase via the RNAP promoter in
the cytoplasm. In order to understand this phenomenon,
a series of luciferase plasmids of increasing size
were prepared to determine the effect of plasmid size
on transfection efficiency in BHK cells. It was found
that L053 (5.8 kb), L069 (8.3 kb), and L070 (10.8 kb)
yieldedsimilar levelsof
transfected in equimolar amounts (Figure 4). This
suggests that, for the system described here, larger
plasmids are not at a disadvantage compared to
the smaller plasmids. In addition, immunofluorescence
studies using anti-T7 RNAP and antiluciferase anti-
bodies showed that the same percentage of cells are
being transfected with the bicistronic construct as with
the dual-plasmid transfection (data not shown), further
supporting this finding.
gene expressionwhen
Cytoplasmic expression system results in a 20-fold
increase in gene expression per cell, compared
to a nuclear expression system
In order to compare the relative efficiency of nuclear
versus cytoplasmic expression, BHK cells were trans-
fected with equimolar amounts of a CMV-Luciferase
(L053) and a bicistronic autogene plasmid containing
both the autogene cassette, as well as the luciferase
reporter gene cassette (R011). As shown in Figure 5, the
autogene system yielded a 20-fold increase in luciferase
expression when compared with the CMV-mediated
nuclear expression system.
It was of interest to determine whether the increase in
luciferase expression was the result of greater levels of
luciferase production in each transfected cell or due to an
increase in the total number of cells being transfected.
The number of cells transfected with the autogene
system was experimentally determined and compared
with the number of cells transfected with the standard
nuclear expression system. Transfected cells were quan-
tified using immunofluorescence with both anti-T7-
RNAP and antiluciferase antibodies, and BHK cells
transfected with either the autogene or nuclear expres-
sion construct. As seen in Figure 6, the autogene and
nuclear expression constructs both result in transfection
of approximately the same number of cells (autogene
11.473.5%, nuclear 10.772.9%). The increase in reporter
gene expression from the bicistronic autogene construct
can therefore be attributed to an increase in the level of
gene expression in transfected cells, as opposed to an
increase in the number of cells being transfected.
It was important to make sure that the 20-fold increase
in expression was not limited to BHK cells. As can be
seen in Figure 7, a similar autogene-mediated improve-
ment in gene expression was also observed in the Neuro-
2a murine neuroblastoma cell line.
The system described here is initially dependent on
the nuclear transcription of T7 RNAP. It was of interest to
determine the proportion of nuclear transcripts derived
Figure 3 Comparison of bicistronic construct versus a dual plasmid
transfection. BHK cells were transfected with 1 mg/well of plasmid.
Equimolar amounts of plasmids were added, and the total mass of DNA
per transfection was kept equal by adding an unrelated plasmid
(pBlueScript). Transfections and luciferase assays were performed as
described in Materials and methods. Error bars indicate the standard error.
Transfection with the bicistronic autogene construct (R011) resulted in
expression levels that were two- to four-fold higher than the dual-plasmid
transfection (autogene and reporter gene on separate plasmids). There is no
luciferase expression in the absence of the autogene cassette.
Figure 4 Plasmid size does not effect the transfection of BHK cells. BHK
cells were transfected with a total of 1 mg/well. Equimolar amounts of
plasmid were added, and the total mass of DNA per transfection was
normalized by adding an unrelated plasmid (pBlueScript). Error bars
indicate the standard error. The size of plasmid, ranging from 5.8 to
10.8 kb, does not have an effect on transfection in BHK cells.
Enhanced cytoplasmic expression system
J Finn et al
279
Gene Therapy

Page 5

from the CMV promoter versus cytoplasmic transcripts
derived from the T7 promoter. As the two promoters
have different transcription start sites, the two transcripts
will have different length 50-untranslated regions (UTRs).
Therefore, a primer extension assay was performed
using a primer that binds downstream of the two
promoters, 90 bp downstream from PT7 and 300 bp
downstream of the PCMV. Figure 8 shows that a much
higher proportion of mRNA is transcribed from the T7
promoter than from the CMV promoter (B57711 fold).
This is consistent with previous work that found that the
large majority of transcripts in the cell were transcribed
by the T7 RNAP in the cytoplasm.11This further
demonstrates that only a catalytic amount of RNAP
needs to be expressed in the nucleus for large amounts of
cytoplasmic mRNA to be produced.
Cytoplasmic mRNA transcripts have a shorter half-life
than nuclear transcripts
It was of interest to determine the half-life of the
cytoplasmic transcripts, as the lack of 50cap structure
on the cytoplasmic transcripts would be expected to
result in a decrease in mRNA stability.13,17–19An RNase
Figure 8 Primer extension assay performed on BHK cells transfected with
the bicistronic autogene construct (R011). (a) The transcripts initiated at
the nuclear CMV promoter were predicted to have a longer 50UTR,
resulting in larger fragments, B300 bp in size, while transcripts initiated
at the T7 promoter were predicted to have a shorter 50UTR, B90 bp in
size. (b) Transfection of BHK cells with the bicistronic autogene construct
(R011) yielded approximately 57711-fold more short cytoplasmic
transcripts as long nuclear transcripts, indicating that the majority of
transcription is initiated by the T7 RNAP in the cytoplasm.
Figure 6 Immunofluorescence of BHK cells transfected with cytoplasmic
or nuclear expression constructs. BHK cells were transfected with
equimolar amounts of R011 (autogene) or L053 (nuclear) plasmids. Cells
were fixed 24 h post-transfection and subjected to immunofluorescence
using anti-T7 RNAP antibodies (with a FITC conjugated secondary) and
antiluciferase (with a Texas Red-conjugated secondary). A similar number
of cells were transfected when using either the autogene or nuclear
expression plasmids. In the case of the autogene transfection, every cell that
is expressing RNAP is also expressing luciferase.
Figure 7 Increased autogene expression is also seen in Neuro-2a cells.
Neuro-2a cells were transfected with a total of 2 mg/well. Equimolar
amounts of plasmids were added, and the total mass of DNA per
transfection was kept equal by adding an unrelated plasmid (pBlueScript).
Error bars indicate the standard error. Transfection with the autogene
(R011) yielded a 20-fold increase in expression over the standard nuclear
expression plasmid (L053).
Figure 5 A comparison of autogene and nuclear expression. BHK cells
were transfected with a total of 1 mg/well. Equimolar amounts of plasmids
were added, and the total mass of DNA per transfection was kept equal by
adding an unrelated plasmid (pBlueScript). Error bars indicate the
standard error. Transfection with the autogene (R011) yielded a 20-fold
increase in expression over the standard nuclear expression plasmid
(L053).
Enhanced cytoplasmic expression system
J Finn et al
280
Gene Therapy