Gene Therapy (1998) 5, 1180–1186
 1998 Stockton Press All rights reserved 0969-7128/98 $12.00
http://www.stockton-press.co.uk/gt
Enhanced in vitro and in vivo gene delivery using
cationic agent complexed retrovirus vectors
M Themis1 and SJ Forbes2, L Chan1, RG Cooper3, CJ Etheridge3, AD Miller3, HJF Hodgson2
and C Coutelle1
1
Division of Biomedical Sciences, 2Liver Group Laboratory, Imperial College School of Medicine; and 3Department of Chemistry,
Imperial College, London, UK
Retroviruses are, at present, the most efficient integrative
vectors available for gene delivery. However, these viruses
are still limited by relatively low titres. Although several protocols exist to improve virus titre most of them are timeconsuming and unable to provide sufficient virus for in vivo
applications. Virus titre can be enhanced by polybrene and
other cationic agents. By investigating a broad range of
cationic agents for their ability to enhance virus infectivity
we found that both ecotropic and amphotropic retrovirus
infection could be increased. In particular, the lipopolyamine dioctadecylamidoglycylspermine (DOGS) gave up to
one order of magnitude enhancement above polybrene-
mediated infection without cytotoxicity. To increase virus
infectivity further we combined the enhancing effect of
DOGS on virus infectivity with concentration of virus particles by ultrafiltration to reach titres of 1 × 109 IU/ml. The
in vivo transduction of regenerating rat liver, by an amphotropic retrovirus was increased approximately five-fold by
the addition of DOGS compared with virus alone. There
was no animal toxicity observed following the administration of DOGS. The improved transduction efficiency
seen both in vitro and in vivo following the co-administration of DOGS/virus complexes may be useful for future
gene therapy applications.
Keywords: gene therapy; retrovirus; liver; cationic agent
Introduction
Both viral and nonviral vectors are currently being
developed for gene therapy. The transfection of plasmid
DNA complexed with liposome carrier molecules into
cells is simple and of low cytotoxicity, but several factors
limit DNA delivery by this approach. These include comparatively poor transfection efficiency, nontargeted delivery and transient transgene expression. Alternatively,
viral vectors offer highly efficient DNA delivery and have
been engineered to target specific cellular receptors.
Adenoviral vectors have been shown to transduce cells
efficiently and can be purified to the high titres applicable
for successful in vivo gene therapy. However, adenovirus
vectors are limited by transient transgene expression and
intrinsic immunogenicity.1
Retroviral gene transfer offers more stable gene delivery as the retrovirus integrates into the host genome.
Recent advances in retroviral biology have further
increased the suitability of retroviral vectors for use in
clinical trials. These developments include safety features
such as self-inactivating vectors, which once integrated
leave the viral LTR promoter defective thereby avoiding
promoter/enhancer insertional mutagenesis.2 Retrovirus
infection is restricted to cells which express the appropriate receptor molecules for virus attachment and novel
Correspondence: M Themis, Division of Biomedical Sciences, Imperial
College School of Medicine at St Mary’s, Norfolk Place, London, W2
1PG, UK
The first two authors contributed equally to this paper
Received 28 November 1997; accepted 27 March 1998
virus pseudotypes have been engineered with envelopes
capable of recognising specific cell receptors.3 Retroviral
envelope pseudotyping is also employed to overcome the
relatively low viral titres which are a major barrier to in
vivo use. For example virions peudotyped with the envelope of vesicular stomatitis virus (VSV-G) can be concentrated by high speed centrifugation.4 More recently retroviruses resistant to complement inactivation by human
serum5 and virus producer cells with enhanced retrovirus
genome expression using an efficient cytomegalovirus
(CMV) promoter within the virus LTR have also enabled
increases in virus titre.6,7 Further methods used to
improve virus titre include ultrafiltration using low
speed centrifugation and hollow fibre filtration.8,9
Virus infectivity can be enhanced several-fold by the
use of positively charged molecules such as polybrene or
DEAE dextran. These compounds complex with the virus
envelope and target the cell wall thereby reducing the
repulsive negative charges, which are believed to inhibit
virus attachment to the target cell receptors.10 The use of
these agents is limited in vivo by both their toxicity and
lack of efficacy. Recently several authors have demonstrated in vitro that complexing virus particles with cationic amphiphiles can enhance virus infection above that
observed in the presence of polybrene.11–13 Such amphiphiles have been used for several years to mediate DNA
transfer to cells, albeit at a lower efficiency than viruses.
They include cationic liposomes which are lamellar vesicles and a variety of cationic compounds in either linear
or branched form with the ability to complex negatively
charged molecules.14,15
In this study we have screened a wide range of cationic
Retroviral gene delivery enhanced by cationic agents
M Themis et al
agents in search of the most effective and nontoxic compounds able to enhance retrovirus titre. We show that
virus titres of 1 × 109 IU/ml can be reached in vitro by
combining virus titre enhancement with cationic amphiphiles together with concentration of virus particles by
ultrafiltration. Furthermore, in an in vivo model of liver
transduction, we demonstrate that the addition of cationic amphiphiles to retroviral vectors before intraportal
injection increases the efficiency of hepatocyte transduction five-fold without any increase in toxicity.
Results
To determine the effects of cationic agents on retroviral
titre, several commercial and noncommercial cationic
liposomes, lipopolyamines and cationic polymers (listed
in Table 1) were investigated at a range of concentrations
from 0.025–30 ␮g/ml in combination with ecotropic and
amphotropic virus supernatants. The virosomes thus
generated by each combination were titred on NIH3T3
cells and the level of infection was compared with that
obtained with polybrene. In addition, the effects on
cellular morphology and survival of NIH 3T3 cells
were examined.
Enhancement of retroviral transduction by cationic
agents
Virosomes were generated with ecotropic viruses produced
by GPE+86LHL cells and amphotropic viruses from
PA317HyTk MLV cells. Each virus was used to infect
NIH3T3 cells for 24 h and the resulting hygromycin B resistant colonies were scored to determine the optimum virosome complex combination for highest virus titre.
Virus titres for each respective pseudotype without cationic agents were approximately five-fold lower than in
the presence of polybrene. Ecotropic viruses exhibited at
least a 20-fold greater level of transduction than their
amphotropic counterparts (data not shown). In the presence of polybrene, ecotropic and amphotropic virus
stocks were titred at 1 × 107 and 3 × 106 c.f.u./ml respectively. All cationic agents listed in Table 1 showed a variable and concentration-dependent ability to transduce
NIH3T3 cells.
Table 2 shows the concentration of each cationic agent
required for maximum virus titre in comparison to that
using virus alone. Many of the agents tested improved
transduction levels significantly above those observed
with virus only or with viruses in combination with
polybrene. Of the cationic polymers, the most effective
was PEI. Although cytotoxicity was observed with PEI at
5 ␮g/ml (Table 1), much lower concentrations were
required to enhance ecotropic and amphotropic virus
transduction by 17- and 19-fold, respectively (Table 2).
Several cationic liposomes also improved virus transduction substantially, especially LipofectAMINE liposomes
and those formulated from the noncommercial cationic
amphiphile N1-cholesteryloxycarbonyl-3,7-diazanonane1,9-diamine CDAN (see Materials and methods),16 which
stimulated ecotropic retroviral transduction 20- and 23fold, respectively. Intriguingly, some cationic liposomes
were able to enhance only one virus pseudotype and not
the other. For instance, DMRIE C liposomes and lipid 67
containing liposomes,17 were able to enhance infectivity
of ecotropic viruses marginally higher than polybrene,
whereas no increase was observed using amphotropic
virions. By contrast, Tfx50 liposomes were only able to
enhance amphotropic virions and not ecotropic particles.
Generally, a higher cationic agent concentration was
needed to reach maximum titre for amphotropic than for
Table 2 Concentration of cationic agent used for optimum
retrovirus transduction of NIH3T3 cells
Cationic agent
Table 1 Cationic agent concentration vs NIH3T3 cell survival
Maximum Concentration
concentration
of agent
of agent
producing
before
50% or less
cytotoxicity
cytoxicity
(␮g/ml)
(␮g/ml)
Cationic polymers
DEAE dextran
Protamine sulphate
PEI
Lipopolyamine
DOGS
Liposomes
Dimrie C
Tfx 50
LipofectAMINE
lipofectin
lipid 67:DOPE
CDAN (B198):DOPE
ACHx
(CJE52):DOPE
CTAp (B232):DOPE
% Cell
survival
(±s.d.)
15
30
30
0.8 ± 0.5
18 ± 3.8
19.9 ± 12.9
30
40
31 ± 4.8
1
5
1
1
1
5
1
15
10
15
40
10
5
10
15
40
18 ± 5.2
6 ± 2.9
18 ± 3.2
9 ± 2.8
17 ± 8.1
22 ± 4.1
18 ± 5.0
20 ± 8.2
5
5
0.5
Cationic polymers
Polybrene
DEAE dextran
Protamine
sulphate
PEI
Lipopolyamine
DOGS
Cationic liposomes
lipofectAMINE
ACHx
(CJE52):DOPE
CTAP
(B232):DOPE
CDAN
(B198):DOPE
DMRIE C
Lipid 67:DOPE
Tfx50
a
Positive
Amphotropic
Ecotropic
charge per
molecule Optimum Fold Optimum Fold
conc increasea conc increasea
(␮g/ml)
(␮g/ml)
Multiple
Multiple
Multiple
5
0.1
5
5
13
7
5
1.75
1
5
10
11
Multiple
0.25
19
1.25
17
4
5
24
1.75
36
4
2
0.5
5
11
14
0.7
1.75
20
13
4
5
18
1.75
14
3
0.35
21
0.25
23
1
3
2
N
N
0.35
N
N
6
0.75
0.25
N
7
8
N
Fold increase above that observed without amphiphile.
N, levels of infection below that observed with polybrene.
1181
Retroviral gene delivery enhanced by cationic agents
M Themis et al
1182
ecotropic viruses (Table 2). Of all the cationic agents
tested, lipopolyamine DOGS and cationic liposomes formulated from CDAN gave the best virus titre enhancement of 36- and 23-fold, respectively, for ecotropic virus,
and 24- and 21-fold, respectively, for amphotropic virus.
The concentration of DOGS required to reach these titres
for each virus pseudotype was higher than the liposome
concentration of CDAN (Table 2). However, DOGS produced the highest virus titres overall (Figure 1a and b).
We also used amphotropic virus/DOGS complexes on
the HepG2 liver cancer cell line to investigate whether a
liver tumour cell line could be more efficiently infected
using a concentration of 1.75 ␮g/ml of DOGS. This was
found to enhance the infection of the HepG2 cells by
amphotropic retrovirus five-fold over that observed with
a polybrene/virus combination. These data encouraged
us to investigate further possible cytotoxic effects of each
cationic agent.
Survival of NIH3T3 cells in the presence of virosome
complexes
To test the concentration of cationic agent which could
enhance virus titre without cytotoxicity, cells were
exposed to a range of cationic agent concentrations complexed with virus particles for 24 h (the length of time
chosen to infect cells by retroviral particles). They were
then examined morphologically and by a clonal assay for
cell survival. Cytotoxicity was determined by the
reduction in percentage survival of colony-forming units
(c.f.u./ml) in comparison with mock-treated control cells.
Table 1 shows the concentration of cationic agents which
reduced cell survival to 50% or less. Viruses alone
showed no cytotoxicity. In general, virosome complexes
formed with 1.75 ␮g/ml or less cationic agent typically
showed little or no cytotoxicity. Exceptions to this rule
were cationic liposomes formulated from cationic lipids
3-aza-N1-cholesteryloxycarbonylhexane-1,6-diamine ACHx
and 4-aza-N1-cholesteryloxycarbonyloctane-1,8-diamine
ACO,16 as well as polycationic polymer PEI. The last
reduced cloning efficiency to slightly below 90%. At
5 ␮g/ml several of the generated virosomes reduced cell
survival to approximately 50% or below (Table 1) (shown
in parallel by a fall in transduction levels). In contrast,
several cationic agents produced no cytotoxic effects at
this concentration; the least toxic being DOGS. CDAN
containing liposomes also promoted high virus titres
with both virus pseudotypes at concentrations well
below those inducing cytotoxicity.
Optimisation of virus titre by combining cationic agent
effects with virus concentration
To reach maximum virus titre virus particles were first
concentrated by ultrafiltration and then complexed with
DOGS at 1.75 ␮g/ml. Virus producer cells TELCeB/
MOF-1 and TELCeB/AL-7, generating ecotropic and
amphotropic virions, respectively, which encode the lacZ
gene were titred at 1 ± 107 IU/ml (in the presence of
polybrene). By complexing with DOGS these cells each
produced virus titres of 1 × 108 IU/ml (Figure 1a and b).
Ten-fold concentrated ecotropic virus stock (by
ultrafiltration) increased the virus titre, in the presence of
polybrene, to 1 × 108 IU/ml while a virus titre of nearly
1 × 109 IU/ml without observable cytotoxicity was
reached by complexing the concentrated virus with
DOGS. Application of this procedure with amphotropic
virus led to titres of approximately 3 × 108 IU/ml. These
data encouraged us to investigate the efficiency of virus
complexed with DOGS for in vivo gene delivery.
Figure 1 (a) and (b). The number of ␤-galactosidase-positive NIH3T3
cells represented as infectious units per ml (IU/ml) following infection
with ecotropic or amphotropic retrovirus supernatants.
Transduction of rat liver in vivo by retroviral vectors
Animals were transduced in vivo by portal vein injection
24 h following partial hepatectomy or sham hepatectomy,
all animals were killed at day 11. The amphotropic retroviral vector from TELCeB/AL-7 cells was used both with
and without the addition of cationic agents.
Animals which had received a sham hepatectomy
24 hrs before retroviral transduction showed a very low
level of hepatic ␤-galactosidase expression in their livers
following transfection with either the TELCeB/AL-7
virus alone, 0.04 ± 0.04% (n = 4) or the TELCeB/AL-7
virus and DOGS, 0.05 ± 0.03% (n = 4).
Animals which had received a partial hepatectomy
24 hrs before transduction had 5.2 ± 2.1% (n = 7) of their
hepatocytes expressing nuclear localised ␤-galatosidase.
Retroviral gene delivery enhanced by cationic agents
M Themis et al
The hepatocyte transduction efficiency was enhanced to
27.4 ± 5.8% (n = 7) by combining the TELCeB/AL-7 retroviral supernatant with DOGS before infusion and to
10.2 ± 2.8% (n = 8) by combining the TELCeB/AL-7 retroviral supernatant with CDAN containing liposomes
before use. Representative photomicrographs of X-galstained frozen sections of rat liver from day 11 are shown
in Figure 2. ␤-Galactosidase transgene expression was
also sought in the heart, lung, testes, skeletal muscle and
brain of all animals in each group but was not detected
in any animal.
Following administration of the retrovirus or
retroviral/cationic agent complexes to either normal animals or animals that had received a partial hepatectomy
24 h earlier, there was no mortality or signs of distress
observed. Histological sections of liver at the time of
death showed no evidence of either ongoing hepatitis or
recent hepatocyte necrosis.
Discussion
We have studied several novel retrovirus/cationic agent
combinations with the aim of enhancing both in vitro and
in vivo retrovirus infectivity. Cationic agent concentrations were optimised for enhancement of viral transduction and minimum cytotoxicity. DOGS complexed
with either ecotropic or amphotropic viruses gave the
largest enhancements in viral titre in comparison to all
the other cationic agents tested. Several other
virus/cationic agent combinations showed useful
improvements in transduction levels of the NIH3T3
indicator cell line.
Ideally the optimal concentration should allow complexing of all available viral particles while not causing
cytotoxicity. Interestingly, the cytotoxicity of each cationic agent did not prove to be the limiting factor for
transduction efficiency. The optimum concentration for
each cationic agent at which the maximum increase in
transduction was observed, was always less than the
cytotoxicity threshold. Therefore, the efficacy of each cationic agent appeared to depend directly upon its ability
to enhance virus–cell interaction rather than upon a
reduction in cell survival. DOGS enhanced virus titre by
almost one order of magnitude over polybrene. In
addition, DOGS showed only a partial fall in virus titre
enhancement over the optimum of 1.75 ␮g/ml and
5 ␮g/ml required for ecotropic and amphotropic virus,
respectively. DOGS appeared effective in promoting
efficient virus infection up to a cytotoxic threshold of
40 ␮g/ml which was significantly higher than the
majority of the other cationic agents tested as measured
by clonal assay and cellular morphology. Using
TELCeB/MOF-1 and TELCeB/AL-7 cells, to generate
ecotropic and amphotropic viruses respectively, encoding the lacZ vector, DOGS enabled titres to reach 1 × 108
IU/ml without concentration. DOGS was therefore considered an ideal candidate cationic agent for in vivo use.
Indeed, DOGS was able significantly to enhance infectivity in vivo without any animal mortality or observed
distress. In addition, no hepatocyte necrosis was
observed following histological analysis of liver sections.
The virosome complexes appeared to be relatively stable
in animal serum.
Possible mechanisms of retrovirus titre enhancement in
the presence of cationic agents include: more efficient cell
binding and entry of the virus, increased transport of the
reverse-transcribed viral DNA to the nucleus of the
infected cell or prolonged virus half-life. The transduction efficiencies exhibited by the virosome complexes
appeared to vary widely. For instance, Tfx50 liposomes
were unable to enhance ecotropic virus transduction
whilst enhancing amphotropic virus infection six-fold. By
contrast, most cationic liposome stimulation was more
pronounced on the ecotropic virus. Perhaps the ability of
the cationic agent to adapt to the surface conformation
of the virus envelope determines the ability of any given
cationic agent to enhance virus infection. That is to say,
the more efficiently the cationic agent may coat the virus
particle, the more effectively the repulsive forces between
virus particle and cell membrane are neutralised, thus
allowing the virus to enter the cells more easily. In support of this, DEAE dextran is known to be capable of
neutralising the repulsive forces between the virus particle and cell membrane through the formation of large
cationic polymer structures sandwiched between the
virus and the cells.10 Hence, DOGS appears to be structurally best suited for interaction with the envelopes of
both viral pseudotypes and therefore for the efficient
regulation of the repulsive forces between virus particles
and cell membranes.
Complexing retrovirus with nontoxic cationic agents is
a useful method to improve virus titres as exemplified
by the increase in infectivity of ecotropic virus by one
order of magnitude when combined with DOGS over the
infectivity observed when using polybrene. The combination of this approach with ultrafiltration of the virus
enhances virus titre further up to about 1 × 109 IU/ml.
Ultrafiltration is probably less harsh than ultracentrifugation. This combination could also improve infectivity in
vivo, however, the volumes of virus supernatant required
for in vivo application would require large-scale culture
and concentration procedures. DOGS, which was able to
improve the transduction of NIH3T3 cells by amphotropic retroviruses also improved the infectivity of the
liver tumour cell line HePG2 by five-fold over that
observed using polybrene. This suggested that virosome
complexes could be tested in vivo on the liver. The liver
was chosen as an in vivo target for nonconcentrated virosome complexes for several reasons. The liver is an
important target for gene therapy, it is the major metabolic organ and a site of metastatic cancer spread. The
model of intraportal injection of retrovirus 24 h following
partial hepatectomy is well characterised and repeatable.18,19 Using this model, if retroviral vectors are
injected intraportally they transduce approximately 5%
of hepatocytes.20 However, it is known that at this time
30–45% of the liver hepatocytes are in S phase of the cell
cycle21 and are therefore potential targets for retroviral
vectors. We therefore hypothesised that if the titre of the
retroviral vector could be increased then the number of
transduced hepatocytes could also be increased following
a single injection of vector. Polybrene is commonly used
to increase the titre of retroviral vectors in vitro but has
not been shown to be effective in vivo.22
We have demonstrated that the in vivo transduction
efficiency of retroviral vectors can be increased five-fold
by the simple addition of a lipopolyamine or cationic
liposomes to the retroviral supernatant before their use.
This process was well tolerated by the recipient animals
which had no excess of morbidity or mortality over ani-
1183
Retroviral gene delivery enhanced by cationic agents
M Themis et al
1184
a
b
c
Figure 2 Photomicrograph of frozen sections of rat liver which has been transducted at 24 h following partial hepatectomy with: (a) retrovirus alone,
(b) retrovirus+CDAN (B198), (c) retrovirus+DOGS. Animals were killed 10 days after transduction.
Retroviral gene delivery enhanced by cationic agents
M Themis et al
mals receiving the retroviral supernatant alone.
Importantly the addition of these cationic agents did not
broaden the tissue range beyond that seen with retroviruses alone.
Recent work has shown that the liver can be primed
to accept retroviral vectors without performing a partial
hepatectomy by administering growth factors23 or transducing the liver with the urokinase gene.24 These
methods are more clinically acceptable than partial hepatectomy, but because the liver remains intact, the number
of target cells is increased, consequently the titre of the
retroviral vector used when utilising these techniques
may have greater importance. Although the virus titres
obtained by these procedures are still somewhat lower
than those observed with adenovirus, this study has
shown that titres of integrative retrovirus vectors may be
increased safely to levels more closely suited for use in
gene therapy protocols with the advantage of no cytotoxic effects. In future experiments we aim to apply concentrated virus stocks from large-scale culture concentrated by hollow fibre filtration to in vivo delivery in
order to increase further the number of hepatocytes
expressing the ␤-galactosidase gene.
Materials and methods
Retroviruses
Retrovirus particles were generated by the packaging
cells lines GPE+86 LHL and PA317 tgLS[+]HyTK producing ecotropic and amphotropic virions, respectively. Each
encode the hygromycin phosphotransferase gene and
confer resistance to the phosphotransferase inhibitor hygromycin B (Calbiochem, CA, USA) on infected cells. The
cell lines TELCeB/AL-7 and TELCeB/MOF-1 were used
to produce amphotropic and ecotropic particles, respectively, and encode a LacZ gene which expresses nuclear
localising ␤-galactosidase. Each cell line was used to
infect NIH3T3 cells to determine virus titres using the
cationic polymer polybrene as the transduction-enhancing agent as previously described.25
Generation of high titre retrovirus stocks
High titres of amphotropic and ecotropic retrovirus
stocks were produced. Virus producer cells were grown
in DMEM supplemented with 10% foetal calf serum at
37°C. To prepare viruses, cells were grown to confluency
and refed with fresh serum containing medium overnight. The following day virus containing supernatants
were harvested and filtered through 0.8 ␮m filters for
determination of retrovirus titre and helper virus assay.
Helper virus assays were performed on NIH3T3 cells as
previously described, using serial dilutions of virus.25
Briefly, for virus titration 4 × 104 NIH3T3 cells were
seeded into each well of 12-well culture dishes and incubated overnight at 37°C. Cells were infected with retroviruses for 24 h followed by a 24-h period to allow for
transgene expression. Titres were analysed either by
using a ␤-galactosidase enzymatic assay or by seeding
cells at clonal density on to 60 mm dishes in hygromycin
B selection to score colonies resistant to this selective
agent. In the presence of polybrene (Sigma, Poole, UK) a
titre of 1 × 107 c.f.u./ml and 3 × 106 c.f.u./ml was established for GPE+86 LHL and PA317 tgLS[+]HyTK cells,
respectively, and 1 × 107 IU/ml for both TELCeB/AL-7
and TELCeB/MOF-1 cell lines.
Infection of NIH3T3 cells to determine amphiphile
effects on transduction
NIH3T3 cells were seeded into 12-well dishes at 5 × 104
cells per well and left overnight at 37°C. Virosome complexes were prepared by mixing an appropriate concentration of cationic agent with 20 ␮l of virus stock in a total
of 0.5 ml of serum-free DMEM. After 45-min incubation,
the volume of medium was made up to 3 ml with DMEM
containing FCS to bring the serum concentration to 10%.
The medium from each well was replaced with the
appropriate virosome complex and the cells were left for
24 h. The following day the virosome complexes were
removed from the wells and the cells were washed with
PBS and refed with fresh serum containing medium. The
cells were left for a further 24 h to allow transgene
expression. Each well was then washed in PBS and trypsinised. Using virus produced by GPE+86 LHL and
PA317 tgLS[+]HyTK cells, NIH3T3 cells were diluted
1/50 on to 60-mm dishes in hygromycin B (Calbiochem)
selection at 500 U/ml and left to grow for 7–10 days after
which time colonies were scored following methylene
blue staining. NIH3T3 cells infected by virus from
TELCeB/AL-7 and TELCeB/MOF-1 cell lines were examined using a ␤-galactosidase assay. High titre virus was
also tested on the HepG2 liver cancer cell line using the
amphotropic/cationic agent complexes which gave
optimum virus titres on NIH3T3 cells.
Cationic agents
The cationic agents used in this study are as follows:
(1) Cationic liposomes: LipofectAMINE, a combination of
2,3-dioleyloxy-N-[2-(sperminecarboxamido)-ethyl]-N,Ndimethyl-1-propanaminium trifluoroacetate (DOSPA)
and dioleoyl l-␣-phosphatidylethanolamine (DOPE), 3:1
(w:w) (GibcoBRL, New York, USA); Lipofectin, a combination of N-[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethyl
ammonium chloride (DOTMA) and DOPE, 1:1 (w:w)
(GibcoBRL) DMRIE C N-[1-(2,3-dimyristyloxy)propyl)
propyl]-N,N-dimethyl-N-(2-hydroxyethyl)
ammonium
bromide and DOPE, 1:1 (w:w) (GibcoBRL); Tfx-50, a combination of N,N,N′,N′-tetramethyl-N,N′-bis(2-hydroxyethyl)-2,3-di(oleoyloxy)-1,4-butane-diammonium iodide and DOPE (Promega, Southampton, UK); liposomes
containing Lipid 67 in combination with DOPE
(Genzyme, Framingham, MA, USA); liposomes containing N1-cholesteryloxycarbonyl-3,7- diazanonane-1,9diamine (CDAN, synonym B198), 3-aza-N1-cholesteryloxycarbonylhexane-1,6-diamine (ACHx, synonym CJE52),
4-aza-N1-cholesteryl-oxycarbonyloctane-1,8 (ACO, synonym B130) or N15-cholesteryloxycarbonyl-3,7,12-triazapentadecane-1,15-diamine (CTAP, synonym B232) in
combination with DOPE, 3:2 (mol:mol).16
(2) Lipopolyamine: DOGS containing dioctadecylamidoglycylspermine (Promega).
(3) Cationic polymers: Protamine sulphate (Sigma), DEAE
dextran (Sigma), polyethylenimine (PEI; Aldrich, Poole,
UK), polybrene (Sigma) 1,5-dimethyl-1,5-diazaundecamethylene polymethobromide.
In vivo transduction of rat liver with retrovirus
Male Wistar rats weighing 200–250 g were used. They were
housed in a temperature- and light-controlled room (12 h
light/ dark cycle) and were allowed water and food ad libitum.
1185
Retroviral gene delivery enhanced by cationic agents
M Themis et al
1186
For in vivo transduction, a modified method of
Rettinger et al18 was used: 24 h following partial hepatectomy the animal was anaesthetised with isoflurane
anaesthetic (Abbott Laboratories, Kent, UK), a mid-line
laparotomy was performed, the bowel was deflected to
the left and wrapped in normal saline soaked gauze. The
portal vein and hepatic artery were clamped using microaneurysm clips, the portal vein was cannulated with a
26-gauge needle and 2 ml of retroviral supernatant either
with or without the addition of cationic agent at the concentration used in vitro was injected over 5 min. Haemostasis was achieved by direct pressure to the portal vein
and topical thrombin (Sigma) application. The clips were
removed and the abdomen sutured, and animals were
allowed to recover in a warm environment. On day 11,
animals were anaesthetised using Hypnorm (Janssen,
High Wycombe, UK) and Hypnovel (Roche, Welwyn
Garden City, UK) and killed by exsanguination. Tissue
from the liver, testes, heart, lung, skeletal and brain was
fresh frozen in liquid nitrogen.
Histological processing and analysis
To analyse the tissue expression of the ␤-galactosidase
transgene, 10 ␮m frozen sections were cut on to glass
slides, fixed in a solution of 1.25% glutaraldehyde at 4°C,
washed with PBS, covered in X-gal (5-bromo-4-chloro3indolyl-␤-d-galactosidase; GibcoBRL) solution (pH 7.4)
and incubated overnight at 37°C. Sections were then
washed in PBS, counterstained lightly with eosin or carmalum stain, dehydrated and mounted in pertex
(Cellpath, Hemel Hempstead, UK). The proportion of
hepatocytes expressing nuclear localised ␤-galatosidase
was calculated by counting 2000 hepatocytes per section
from three separate areas. The specificity of staining was
confirmed by examining serial sections.
Analysis of in vitro and in vivo toxicity of the cationic
agents
In vitro analysis of the cytotoxic effects of retroviruses
complexed with cationic agents on NIH3T3 cells was performed both morphologically and by clonal assay.
Infected cells and control mock-infected cells were
seeded at clonal density on to 60-mm dishes at 100 cells
per dish. Colonies of cells either in the presence or
absence of selective agent were scored and compared
with unselected mock-treated cells.
Following
in
vivo
administration
of
the
retroviral/cationic agents to both normal animals and
animals which had received a partial hepatectomy 24 h
earlier, the animals were observed for signs of distress.
At the time of death (day 11) histological sections of the
liver were examined for either ongoing hepatitis or
evidence of recent hepatocyte necrosis.
Acknowledgements
Stuart Forbes is a Wellcome Trust Research Training Fellow. This work was supported by funds from the Wellcome Trust, MRC, Genzyme, the Sir Jules Thorne Charitable Trust and the Muller Bequest. We wish to thank
Yasuhiro Takeuchi for the ecotropic and amphotropic
retrovirus producer cells lines TELCeB6/MOF-1 and
TELCeB/delPMOSAF-7 and Robert Newbold for the
PA317 tgLS[+]HyTK producer cell line.
References
1 Yang Y, Li Q, Ertl HC, Wilson JM. Cellular and humoral
immune responses to viral antigens create barriers to lungdirected gene therapy with recombinant adenoviruses. J Virol
1995; 69: 2004–2015.
2 Olson P, Nelson S, Dornburg R . Improved self-inactivating
retroviral vectors derived from spleen necrosis virus. J Virol
1994; 68: 7060–7066.
3 Chu TH, Dornburg R. Retroviral vector particles displaying the
antigen-binding site of an antibody enable cell-type specific
gene transfer. J Virol 1995; 69: 2659–2663.
4 Burns JC et al. Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titre and
efficient gene transfer into mammalian and nonmammalian
cells. Proc Natl Acad Sci USA 1993; 90: 8033–8037.
5 Cosset FL et al. High titre packaging cells producing recombinant
retroviruses resistant to human serum. J Virol 1995; 69: 7430–7436.
6 Soneoka Y et al. A transient three-plasmid expression system for
the production of high titer retroviral vectors. Nucleic Acids Res
1995; 4: 628–633.
7 Paul RW et al. Increased viral titer through concentration of viral
harvests from retroviral packaging lines. Hum Gene Ther 1993;
4: 609–615.
8 Chu THT, Dornburg R. Toward highly efficient cell type-specific
transfer with retroviral vectors displaying single-chain antibodies. J Virol 1997; 71: 720–725.
9 Paul RW et al. Increased viral titre through concentration of viral
harvests from retroviral packaging lines. Hum Gene Ther 1993;
4: 609–615.
10 Kahn ML, Lees W, Dichek DA. Optimization of retroviral vector-mediated gene transfer into endothelial cells in vitro. Circ Res
1992; 71: 1508–1517.
11 Hodgson CP, Solaiman F. Virosomes: cationic liposomes enhance
retroviral transduction Nature Biotechnol 1996; 14: 339–343.
12 Kaneko Y, Tsukamoto A. Structural characteristics of cationic
liposomes with potent enhancing effect on retroviral transduction into human hepatoma cells. Cancer Lett 1996; 107: 211–215.
13 Singh D, Rigby PW. The use of histone as a facilitator to improve
the efficiency of retroviral gene transfer. Nucleic Acids Res 1996;
24: 3113–3114.
14 Gao X, Huang L. Cationic liposome-mediated gene transfer.
Gene Therapy 1995; 2: 710–722.
15 Smith JG, Walzem RL, German JB. Liposomes as agents of DNA
transfer. Biochem Biophys Acta 1993; 1154: 327–340.
16 Cooper RG et al. Polyamine analogues of 3␤-[N-(N′,N′dimethylaminoethane)carbamoyl]cholesterol
(DC-Chol)
as
agents for gene delivery. Chem Eur J 1998 (in press).
17 Lee ER et al. Detailed analysis of structures and formulations of
cationic lipids for efficient gene transfer to the lung. Hum Gene
Ther 1996; 7: 1701–1717.
18 Rettinger SD et al. In vivo hepatocyte transduction with retrovirus during in flow occlusion. J Surg Res 1993; 54: 418–425.
19 Ferry N et al. Retrovirus-mediated gene transfer into hepatocytes in vivo. Proc Natl Acad Sci USA. 1991; 88: 8377–8381.
20 Frederiks WM et al. Immunocytochemical determination of
ploidy class-dependent bromodeoxyuridine incorporation in rat
liver parenchymal cells after partial hepatectomy. Histochemistry
1990; 93: 627–630.
21 Hafenrichter DG et al. Optimization of liver-directed gene therapy. Transplant Proc 1994; 26: 3379–3780.
22 Pitt BR et al. Retrovirus-mediated gene transfer in lungs of living
fetal sheep. Gene Therapy 1995; 2: 344–350.
23 Bosch A et al. Proliferation induced by keratinocyte growth factor enhances in vivo retrovirus-mediated gene transfer to mouse
hepatocytes. J Clin Invest 1996; 98: 2683–2687.
24 Lieber A et al. Adenovirus-mediated urokinase gene transfer
induces liver regeneration and allows for efficient retroviral
transduction of hepatocytes in vivo. Proc Natl Acad Sci USA 1995;
92: 6210–6214.
25 Miller AD, Rosman GJ. Improved retroviral vectors for gene
transfer and expression. Biotechniques 1989; 7: 980–990.