Gene Therapy (1999) 6, 1995–2004
 1999 Stockton Press All rights reserved 0969-7128/99 $15.00
http://www.stockton-press.co.uk/gt
Comparison between cationic polymers and lipids in
mediating systemic gene delivery to the lungs
A Bragonzi1, A Boletta1, A Biffi1, A Muggia1, G Sersale2, SH Cheng3, C Bordignon1, BM Assael2
and M Conese1
1
Telethon Institute for Gene Therapy, San Raffaele Scientific Institute, Milano; 2Centro di Riferimento Fibrosi Cistica, Milano and
Department of Pediatrics, University of Milano, Italy; and 3Genzyme Corporation, Framingham, MA, USA
Airway inflammation frequently found in congenital and
acquired lung diseases may interfere with gene delivery by
direct administration through either instillation or aerosol.
Systemic delivery by the intravenous administration represents an alternative route of delivery that might bypass
this barrier. A nonviral approach for transfecting various
airway-derived cell lines in vitro showed that cationic polymers (PEI 22K and 25K) and lipids (DOTAP, GL-67/DOPE)
are able to transfect with high efficiency the reporter genes
firefly luciferase and E. coli lacZ. Notably, two properties
predicted that cationic vectors would be useful for a systemic gene delivery approach to the lung: (1) transfection
was not inhibited or increased when cells were incubated
with cationic lipids or polymers in the presence of serum;
and (2) cationic vectors protected plasmid DNA from
DNase degradation. A single injection of DNA complexed
to the cationic polymer PEI 22K into the tail vein of adult
mice efficiently transfected primarily the lungs and to a
lesser extent, heart, spleen, kidney and liver. The other
vectors mediated lower to undetectable levels of luciferase
expression in the lungs, with DOTAP ⬎ GL67/DOPE ⬎ PEI
25K ⬎ DOTMA/DOPE. A double injection protocol with a
15-min interval between the two doses of DOTAP/DNA
complexes was investigated and showed a relevant role of
the first injection in transfecting the lungs. A two log
increase in luciferase expression was obtained either when
the two doses were comprised of luciferase plasmid or
when an irrelevant plasmid was used in the first injection.
The double injection of luciferase/PEI 22K complexes
determined higher transgene levels than a single dose, but
a clear difference using an irrelevant plasmid as first dose
was not observed. Using lacZ as a reporter gene, it was
shown that only cells in the alveolar region, including type II
penumocytes, stained positively for the transgene product.
Keywords: cationic polymers; lipids; DNase; intravenous delivery; pneumocytes; gene therapy
Introduction
Gene therapy could be useful for congenital and acquired
lung diseases that are characterized by chronic airway
inflammation and alveolitis/fibrosis, as in cystic fibrosis
(CF),1 surfactant deficiency states,2 after ionizing
irradiation for cancer treatment,3 or when lung parenchyma destruction ensues, such as in alpha1-antitrypsin
deficiency.4 However, many factors can impede the gene
transfer process particularly in injured respiratory tracts
when a direct delivery is attempted. For example, in CF
the lung pathology is characterized by thick dehydrated
mucus production. In the other diseases, other barriers
such as alveolar accumulation of proteins and lipids or
uninhibited neutrophil elastase are present.
Cationic lipids have been shown capable of mediating
gene transfer to respiratory epithelial cells both in vitro5–7
and into the lungs of small animals. These in vivo studies
were conducted by administrating the DNA/lipid complexes via aerosol, intranasal or intratracheal injections
into airways of normal and uninjured lungs.8–11 Systemic
delivery by intravenous injection has also been shown to
be effective in transfecting lung tissue.12–18
Polycationic polymers, like polyethylenimine (PEI), are
a new class of nonviral vectors capable of transfecting in
vitro and in vivo a variety of cell types, including respiratory cells.19–23 Previous intravenous studies have not
directly compared cationic lipids and polymers in transfecting animal model airways.
We undertook the present study to extend our understanding of the feasibility of transfecting airway lung
cells by systemic delivery of DNA/lipid complexes
(lipoplexes) and to compare their efficiency with that of
DNA/polymer complexes (polyplexes). Our data show
that the cationic polymer PEI 22K is the optimal vector
with this route of administration and that it can transfect
lung alveolar cells following a single intravenous administration. We envisioned also a double injection protocol
aimed to increase gene transfer levels in vivo and to give
an insight into the mechanism of action of nonviral
vectors.
Results
Correspondence: M Conese, Telethon Institute for Gene Therapy, San
Raffaele Scientific Institute, Via Olgettina 58, Milano 20132, Italy
Received 8 March 1999; accepted 19 July 1999
In vitro optimization of cationic vectors
To test and compare the relative efficiencies of various
cationic lipids and cationic polymers in transfecting air-
Systemic gene transfer to the lungs
A Bragonzi et al
1996
Figure 1 Detection of luciferase expression following transfection of
human airway epithelial cell lines. Cells were plated on to 24-well plates
(40 000 cells per well), transfected 16 h later with the vectors complexed
to 2 ␮g of pCLuc, and incubated for an additional 24 h. Lipoplexes and
polyplexes were used at 7 N/P equivalents for PEI 22K and 25K, 2.4 N/P
for DOTAP, 6 N/P for DOGS, 5 N/P for DOTMA/DOPE and 0.5 N/P
for GL-67/DOPE. Relative light units (RLU) were normalized to the
protein content of cell lysates. The results are the mean (±s.d.) of three
experiments conducted in triplicate.
way-derived cell lines, we evaluated expression of the
reporter gene firefly luciferase. In A549, type II pneumocyte-derived epithelial cells, the cationic lipids DOTAP
and GL-67/DOPE exhibited higher transfection
efficiencies than the cationic lipids DOGS and
DOTMA/DOPE but approximately the same efficiency as
observed with the polymeric DNA-binding cations PEI
22K and 25K (Figure 1). In other airway cell lines tested,
for example, the human fetal tracheal 56FHT80- and the
human CF nasal epithelial cells CFNPE90-, PEI 22K
mediated higher luciferase levels than all the other vectors (Figure 1). Evaluation of E. coli ␤-galactosidase
expression in A549 cell line confirmed that DOTAP, GL67/DOPE, PEI 25K and PEI 22K showed a higher
efficiency
of
transfection
than
DOGS
and
DOTMA/DOPE (Table 1).
Table 1 Percentage of ␤-galactosidase-positive A549 cells upon
transfection with cationic lipids and polymers
Vector
PEI 22K (7)
PEI 25K (7)
DOTAP (2.4)
DOGS (6)
DOTMA/DOPE (5)
GL-67/DOPE (0.5)
% of transfected cells Protein concentration
(␮g/␮l)
25.6 ± 4.1
48.3 ± 7
49.3 ± 2.7
16 ± 1.6
9.5 ± 0.7
55 ± 1
0.31 ± 0.13
1.07 ± 0.4
0.47 ± 0.3
0.76 ± 0.44
0.51 ± 0.16
0.49 ± 0.05
A549 cells were seeded on to 24-well plates (40 000 cells per
well) and transfected with the cationic vectors complexed to
2 ␮g of pCMV␤. After 24 h, cells were stained with X-gal, and
10 microscopic fields were evaluated for ␤-galactosidase-positive cells. In parallel, protein concentration was evaluated by
Bradford’s assay. Control untransfected cells presented a protein concentration in the range of 0.8–1.0 ␮g/␮l. Lipoplexes and
polyplexes were used at the indicated N/P equivalents. Data
are shown as the mean ± s.d. of two experiments conducted
in duplicate.
It has been previously shown that transfection activity
is significantly decreased by the addition of serum to the
incubation medium.24 The effect of serum on gene transfer into A549 cells mediated by DOTAP, GL-67/DOPE,
PEI 25K or PEI 22K was tested. We found that the
efficiency of gene transfer mediated either by PEI 22K,
PEI 25K or DOTAP was not appreciably affected by the
presence of serum (Figure 2a, b and c). In serum-free conditions, luciferase activity mediated by GL-67/DOPE was
lower than in the presence of serum at all the N/P equivalents tested (Figure 2d). The resistance of these cationic
vectors to serum suggests that all four vectors could be
useful for systemic gene delivery.
Figure 2 also shows that optimization of the charge
ratio is an important parameter for transfection of airway
epithelial cells. For all four vectors, the complexes that
gave the maximal luciferase expression were slightly
positive (ie 7 N/P equivalents for PEI 22K and PEI 25K,
corresponding to 1.2 +/− charge ratio of the complexes;
2.4 N/P equivalents for DOTAP, corresponding to 2.4
+/− charge ratio; 0.5 N/P equivalents for GL-67/DOPE,
corresponding to 1.5 +/− charge ratio).
To evaluate the integrity of plasmid DNA in the presence of cationic lipids, its protection from DNase digestion was evaluated. Free DNA presented three forms
(relaxed, linear and supercoiled) and was completely
degraded when incubated in the presence of DNase I
(Figure 3a). DOTAP partially protected DNA from degradation, whereas, using DOGS as a carrier, no DNA forms
were observed. DOTAP was equally effective in protecting DNA at all the lipid/DNA N/P equivalents tested
(Figure 3b). GL-67/DOPE also protected DNA from
degradation at the 0.5–1 N/P equivalents but less at the
0.25 N/P equivalent (Figure 3c).
Relative efficiency of cationic vectors in mediating gene
transfer in vivo
We next compared the luciferase levels obtained in various organs of C57Bl/6 mice after intravenous injection of
pCLuc DNA complexed to cationic vectors. The peak levels of luciferase expression were found between 24 and
48 h after injection with the complexes (data not shown).
Using PEI 22K, the highest levels of luciferase expression
were found in the lungs, followed by the other organs,
including spleen, liver, kidney and heart (Figure 4a).
Among the other vectors, only DOTAP could determine
significant luciferase levels into the lungs, with low levels
obtained in this organ with GL-67/DOPE and PEI 25K.
Both DOTMA/DOPE and DOGS (not shown) gave negligible expression in all the organs. Injected free plasmid
DNA did not gain any gene reporter expression
(Figure 4a).
As for in vitro studies, the optimal ratio of the vectors
to DNA was also evaluated in vivo. Figure 4b shows that
the higher the amount of the cationic vector in the complexes the higher the luciferase levels obtained in the
lungs. PEI 22K and PEI 25K were most effective when 10–
15 N/P equivalents were used, while DOTAP and GL67/DOPE gave the highest luciferase expression at the
3.6–4.8 and 2–4 N/P equivalents, respectively. Even
though the transfection mediated by PEI 25K, DOTAP or
GL-67/DOPE increased when more cationic vector in the
complexes was used, luciferase levels were approximately from one to four logs below those obtained by
PEI 22K. The transfection mediated by PEI 22K or
Systemic gene transfer to the lungs
A Bragonzi et al
1997
Figure 2 Effect of serum and charge ratio on transfection efficiency. A549 cells plated on to 24-well plates were transfected with 2 ␮g per well of pCLuc
complexed to cationic vectors at the indicated vectors/DNA N/P equivalents either in the absence or in the presence of 10% FCS. We used from 2.5 to
15 N/P equivalents for PEI 22K (a), from 2.5 to 15 N/P equivalents for PEI 25K (b), from 0.4 to 4.8 N/P equivalents for DOTAP (c), and from 0.12
to 1 N/P equivalent for GL-67/DOPE (d). Evaluation of luciferase levels was carried out after 24 h from the addition of complexes.
DOTAP was dependent on the N/P equivalents also in
all the other organs tested (Figure 4c and d). Higher N/P
equivalents than those showed determined animal mortality and they were not considered further.
Double injections of cationic vector/DNA complexes
In an attempt to further increase the levels of transfection
in the lungs, we administered a second dose of complexes
16 h after the first one, but saw no increase in transfection
activity (not shown). The same experiment was performed with shorter intervals between the two injections.
When the injections were done with an interval of only
15 min with the pCLuc DNA complexed to DOTAP (at
2.4 N/P equivalents), the luciferase levels increased two
logs compared with that obtained using a single injection
(Figure 5a). Surprisingly, similar high levels of expression
were obtained when the first dose was comprised of an
irrelevant plasmid (pCMV␤n). However, when we
inverted the order of the injections (first the pCLuc and
secondly the pCMV␤n), the luciferase levels did not
increase (Figure 5a). Indeed, the single injection of
pCMV␤n/DOTAP complexes resulted in background
levels. Two injections of PEI 22K/pCLuc (7.5 N/P
equivalents) at a 15-min interval resulted in one log
increase of luciferase levels (Figure 5b). Differently than
to DOTAP, injection of polyplexes containing pCLuc
either before or after the irrelevant PEI 22K/pCMV␤n
complexes did not yield any significant difference with
respect to a single injection (Figure 5b), suggesting a different mechanism of transfection in vivo between cationic
lipids and polymers.
Localization of cationic lipid-transfected cells into the
lungs
At 24 h following intravenous injection of a plasmid
encoding ␤-galactosidase complexed to PEI 22K at 15 or
Systemic gene transfer to the lungs
A Bragonzi et al
1998
Figure 3 DNase sensitivity of DNA complexed to cationic vectors. (a) pCLuc (576 ng) was complexed to either DOTAP (at 2.4 N/P equivalents) or
DOGS (at 6 N/P equivalents) and incubated in the absence or presence of 8 U of DNase I. Free DNA was used as control. (b) pCLuc was complexed
to DOTAP at 1.2, 2.4 and 3.6 N/P equivalents and incubated in the absence or in the presence of DNase I. (c) pCLuc was complexed to GL-67/DOPE
at 0.25, 0.5 and 1 N/P equivalents and incubated in the absence or in the presence of DNase I.
Figure 4 Expression of luciferase in mice organs after systemic administration of cationic vector/DNA complexes. (a) C57Bl/6 mice were injected in
the tail vein with 50 ␮g of pCLuc complexed to cationic vectors at the following N/P equivalents: 15 for PEI 22K and PEI 25K, 5 for DOTMA/DOPE,
4.8 for DOTAP and 2 for GL-67/DOPE. After 24 h, mice were killed, organs were homogenated with lysis buffer (see Materials and methods), and
luciferase expression was evaluated. (b) Intravenous injections were performed with different N/P equivalents of the complexes as indicated and luciferase
expression was revealed only in the lungs. In (c) and (d) is shown the optimization of gene transfer to the mice organs using different N/P equivalents
for PEI 22K (c) or DOTAP (d). Each experiment was performed with three–seven animals.
Systemic gene transfer to the lungs
A Bragonzi et al
1999
Figure 5 Expression of luciferase after two injections of PEI 22K/DNA or DOTAP/DNA complexes. C57Bl/6 mice were injected in the tail vein with
the vector complexed to 50 ␮g plasmid (I) and a second dose was given within a 15-min interval (II). The single injection of the cationic vector/pCLuc
or cationic vector/pCMV␤n complexes was performed for comparison (I, pCLuc or I, pCMV␤n). pCLuc, a plasmid encoding firefly luciferase, and
pCMV␤n, encoding for a nuclear ␤-galactosidase, were used either in the first or in the second injection as indicated. DOTAP/DNA complexes (a) and
PEI 22K/DNA complexes (b) were used at 2.4 or 7.5 N/P equivalents, respectively. Luciferase levels were evaluated 24 h later (n = 3 per group).
DOTAP at 4.8 N/P equivalents, we evaluated the presence of transfected cells by X-gal staining of whole lungs.
Since the pCMV␤n expression vector contains a nuclear
localization signal, expression of ␤-galactosidase was
observed in cell nuclei of cells lining the alveoli for both
PEI 22K and DOTAP (Figure 6a and b). The higher cell
number in PEI 22K-transfected lungs confirmed the
luciferase data. Staining of bronchi and bronchioli was
not observed. The same pattern of staining was obtained
when GL-67/DOPE was used as vector and with double
injection protocols using PEI 22K or DOTAP (data not
shown). No stained nuclei were found in the lungs of
mice transfected with an irrelevant plasmid complex
either with PEI 22K or DOTAP. In the attempt to characterize the cellular type target of the cationic polymers and
lipids, immunohistochemistry on cryosections with an
antibody directed against surfactant protein A (SP-A)
was performed. Some of the transfected cells were positive for SP-A expression (Figure 6c, d and e), indicating
that these cells were type II pneumocytes. When we used
an anti-CD11b antibody to detect macrophages, no colocalization was found with nuclear ␤-galactosidase
positive cells (Figure 6f, g and h).
Discussion
In this study, we compared the relative ability of two different classes of DNA delivering agents, cationic lipids
and polycationic polymers, in transfecting lung tissue following a single intravenous injection. Although the cationic vectors tested were able to mediate gene delivery
to airway-derived cell lines in vitro with high efficiency,
only some of them showed significant activity in vivo (PEI
22K ⬎ DOTAP ⬎ GL-67/DOPE ⬎ PEI 25K).
Our data indicate that the behaviour of any vector
tested previously in vitro is only partially reproducible in
vivo for an intravenous administration. Indeed, optimization of the overall charge of the complexes revealed significant differences between in vitro and the in vivo system. The more positive the complexes injected in the tail
vein the more luciferase levels were obtained in the
lungs, whereas less positive complexes were optimal in
vitro. These data are in agreement with other studies in
which a larger excess of positive charges are required for
lung transfection through intravenous administration.25
On the other hand, neutral polyplexes were shown to be
optimal in transfecting the lungs through the intratracheal route (Ref. 22 and our data not shown) and
DNA/PEI complexes bearing ionic charge ratios closest
to neutrality gave the highest levels of transgene
expression in the mature mouse brain.26 This difference
between intravenous and the other routes of administration might be due to the presence of serum components in the first case. Although the presence of serum in
the transfection medium has been demonstrated to
inhibit gene transfer activity of cationic lipids,24 our data
show that both lipoplexes and polyplexes are resistant
to serum inhibition (Figure 2). DOTAP and GL-67/DOPE
could protect the plasmid DNA from DNase degradation,
a property not shared by DOGS (Figure 3). These properties could explain in part the cationic vector efficiency in
lung transfection after a single intravenous injection. PEI
transfection was shown to be resistant to serum but,
because it was not possible to separate DNA from PEI
with a simple organic phase extraction, due to its chemical nature, DNA protection assay could not be performed
with these polyplexes.
We obtained the higher levels of lung transfection
(3 × 107 RLU/mg protein) using a linear 22K form of PEI
than those achieved with branched counterpart PEI 25K.
Previous studies using scanning force microscopy have
shown that the vast majority of complexes were 20–40 nm
in diameter for both the linear PEI 22K and branched PEI
25K, although a small number of much larger complexes
resulted from condensation of DNA with PEI 25K.27 This
difference in size distribution might be important when
polyplexes have to interact with the endothelial luminal
membrane. Small size of the complexes has been considered a critical parameter for in vivo delivery to cross
barriers such as endothelial cell fenestrations and dif-
Systemic gene transfer to the lungs
A Bragonzi et al
2000
Figure 6 Identification of transfected lung cells in vivo following systemic administration of PEI 22K and DOTAP/DNA complexes. C57Bl/6 mice were
injected into the tail vein with 50 ␮g of pCMV␤n complexed to PEI 22K at 15 N/P equivalents (a, c, d, f and g) or DOTAP at 4.8 N/P equivalents
(b, e and h) and nuclear ␤-galactosidase expression was evaluated by X-gal staining (see Materials and methods) after 24 h. (a) and (b) Microphotographs
showing that the transfected cells are localized in the alveolar regions. Original magnification 20×. (c), (d) and (e) Co-localization of the nuclear ␤galactosidase staining (blue) with the detection of SP-A (brown) in a single alveolar cell. Original magnification 100×. Inset in (e) depicts an enlargement
of the cell showing both SP-A and ␤-galactosidase expression. (f), (g) and (h) Absence of co-localization between ␤-galactosidase-positive cells and
staining for CD11b (identifying macrophages). Original magnification 100×.
fusion into the tissues. It has been shown that the ionic
strength of the solution in which polyplexes are formed
influences their size. In fact, small polyplexes (40 nm)
were obtained at low ionic strength (5% glucose), while
large complexes (up to ⬎1000 nm) were formed in
physiological salt concentrations.28 Moreover, Goula et
al29 showed that increasing the N/P ratio of the
polyplexes decreased the size of particles formulated in
glucose, whereas this effect was not observed when the
complexes were formed in NaCl. Preliminary experiments using a physiological salt concentration for the
intravenous administrations of complexes resulted in
approximately 1000-fold less transgene levels in the lungs
than using complexes formed in 5% glucose (data not
shown), demonstrating that the small complexes are
more efficient in vivo for the intravenous injections than
the large complexes. To increase the activity of polyplexes in vivo, we tried different protocols, but we found
no improvement in lung transfection with either increasing amounts of DNA or by pre-condensing DNA to
poly-l-lysine (not shown).
Others have found that using linear PEI 22K at 4 equivalents and 125 ␮g DNA, it was possible to obtain high
levels of lung transfection (1 × 107 RLU/mg protein).23
Systemic gene transfer to the lungs
A Bragonzi et al
We achieved these same levels using PEI 22K at 15 equivalents with 50 ␮g DNA (Figure 4). Since it has been
shown that bacterial DNA can induce acute inflammation
in the lower respiratory tract after intratracheal instillation,30 our experimental conditions should favor less
toxicity of polyplexes. Infact, when Goula et al23 used
125 ␮g DNA with a N/P ratio of 6, those polyplexes were
toxic in 50% of the mice injected, while when we used the
same amount of the cationic polymer with 50 ␮g DNA at
N/P of 15, no toxicity was observed. This could also be
the reason why a high toxicity with PEI 25K was shown
even at low N/P ratios, making impossible for these
authors a direct comparison between the two forms of
PEI.23 On the other hand, it has been shown that PEI 25K
can activate the complement.31
DOTAP was the most efficient vector in vivo among the
cationic lipids, but nevertheless, it did not reach the levels
obtained with PEI 22K. Besides the different chemical
configuration, the lower activity of DOTAP might be
ascribed to the size of the particles which were shown to
be between 180 and 280 nm.32 DOTAP and GL-67/DOPE
are composed of a hydrophobic lipid anchor group capable of interacting with membranes, linked to a positively
charged headgroup that binds to and condenses the
DNA. GL-67 is a new cationic lipid that arose from structure–function studies and was shown to be more efficient
than other cationic lipid preparations when given intranasally.33 Therefore, it was surprising to observe that
DOTAP was more efficient than GL-67/DOPE after a single intravenous injection. Clearly, factors other than
chemical configuration must be important when the systemic route is used. Another two cationic lipids (DOGS
and DOTMA/DOPE) were not efficient in mediating
gene transfer either in vitro or in vivo. One possible explanation comes from the finding that DOGS can strongly
activate complement.31 In the same study, DOTAP and
DOTMA/DOPE were also shown to be weak activators
of complement. Again, the different efficiency of DOTAP
and DOTMA/DOPE must be found in other properties.
Repeated doses of DOTAP complexes (50 ␮g given
twice in a 16-h interval) did not increase luciferase
expression in the lung (not shown). In contrast, a similar
experiment carried out with a short time interval between
the two doses (15 min) determined luciferase levels that
were greater than two logs over those obtained with a
single administration. Since the double injection of
DOTAP/DNA complexes with 4.8 N/P equivalents was
toxic, it was possible to perform this protocol only with
2.4 N/P equivalents. Interestingly, the large increase in
luciferase
activity
was
obtained
with
either
DOTAP/pCMV␤n or DOTAP/pCLuc complexes used as
the first dose. These results are similar to those obtained
by pre-injecting cationic lipids and subsequently either
naked DNA34 or DNA/lipid complexes.35 A possible
explanation is that aggregates between DNA complexes
and serum proteins are trapped in lung capillaries and
thereby retain the DNA complexes injected later. It might
also be possible that the first injection saturates the
macrophage cells,36 allowing the second injection to overcome the cellular barriers to intravenous administration.
An increase of luciferase levels was reached with the
double injection of PEI 22K/pCLuc complexes in a 15min interval with respect to a single injection, but a clear
difference using an irrelevant plasmid as first dose was
not observed. Since this is the first time to our knowledge
that double injection protocols have been exploited with
polyplexes for an intravenous route, further studies are
needed to comprehend the mechanism underlying PEI
22K lung transfection in vivo.
In previous studies performed using the systemic
route, recombinant protein was localized either in the
alveolar region12,18 or in the endothelium of microvessels
present in the alveolar septi.16,17,25 Recently, DNA was
shown in type II alveolar epithelial cells,18 a finding confirmed by our protein data. Indeed, we have found transgene expression only in cells lining the alveoli. Cytochemical characterization indicates that some of those cells
are type II pneumocytes. Since alveolar macrophages
never stained positive for ␤-gal expression, these results
could be due to the lack either of macrophage transfection or of transgene expression by these cells.
Alveolar type II cells are potential targets for the gene
therapeutic treatment of congenital lung diseases such as
neonatal respiratory distress syndrome and congenital
pulmonary alveolar proteinosis, or ␣1-antitrypsin
deficiency and cystic fibrosis. In the case of CF, it is still
not known which cell type is responsible for the pathogenesis of the disease. However, it has been shown that
CFTR is expressed in alveolar type II precursors and in
differentiated type II pneumocytes.37 It would be interesting to see if CFTR transfection in CF-type II pneumocytes
in vivo resulted in any change of CF phenotype.
Materials and methods
Cell lines
A549 (human type II pneumocytes), 56FHT80- (human
fetal tracheal epithelium) and CFNPE90- (a CF nasal
epithelium) cells were grown in DMEM/HAM F12 (1:1)
containing 10% fetal calf serum (FCS), 2 mm l-glutamine,
100 U/ml penicillin and 100 ␮g/ml streptomycin. A549
cell line was from American Type Culture Collection
(ATCC, Rockville, MD, USA). 56FHT80- and CFNPE90cell lines were a kind gift of Dr D Gruenert (University
of California, San Francisco, CA, USA). Each cell line was
maintained in culture until the 10th-12th passage.
Plasmids
Plasmids pCLuc, carrying the P. pyralis luciferase coding
region under the control of the cytomegalovirus (CMV)
immediate–early enhancer/promoter region, and
pCMV␤n, coding for a nuclear form of E. coli lacZ gene
driven by the same promoter were kind gifts of Dr L
Monaco (HS Raffaele, Milano, Italy). Plasmid carrying the
cytosolic form of lacZ gene (pCMV␤) was purchased
from Clontech Laboratories (Palo Alto, CA, USA). Plasmid DNA preparation was performed by double CsCl
gradient purification.38 A260/A280 was determined spectrophotometrically and was between 1.8 and 2.0. Purity
of the plasmids was also checked by 1% agarose gel
electrophoresis.
Cationic vectors–DNA complexes formation
Linear 22-kDa polyethylenimine (PEI 22K) (Exgen 500;
Euromedex, Souffelweyersheim, France) was purchased
as 5.47 mm (0.4 mg/ml) aqueous stock solution. Branched
25-kDa PEI (PEI 25K) (Sigma-Aldrich, St Louis, MO,
USA) was used as 100 mm (4.5 g/l) aqueous stock solution. DOTAP (Boehringer Mannheim, Mannheim,
2001
Systemic gene transfer to the lungs
A Bragonzi et al
2002
Germany) was purchased as a 1 mg/ml aqueous solution. DOGS (Transfectam) was a kind gift of Dr J-P Behr
(Strasbourg, France) and was used as a 2 mm ethanolic
solution. DOTMA/DOPE (Lipofectin; Gibco BRL, Life
Technologies, Gaithersburg, MD, USA) was obtained as
a 1 mg/ml aqueous solution. GL-67 was formulated with
DOPE at a molar ratio of 1:233 and was reconstituted as
1 mm aqueous solution. For complex formation, the following amounts of each vector were used per 2 ␮g of
DNA: from 2.5 to 15 equivalents of PEI nitrogen per DNA
phosphate, that is from 0.15 to 0.6 ␮l of PEI 25K solution19
and from 2.7 to 16.4 ␮l of PEI 22K solution; for DOTAP,
from 1.2 to 4.8 equivalents of DOTAP per DNA phosphate, corresponding from one- to 12-fold weight excess,
that is from 6 to 24 ␮l of DOTAP solution; 6 equivalents
of DOGS protonated nitrogen per DNA phosphate, that
is 6 ␮l of DOGS solution;39 5 equivalents of
DOTMA/DOPE, corresponding to 20 ␮l of stock solution, that is a 10-fold weight excess; for GL-67/DOPE,
from 0.12 to 4 equivalents of GL-67/DOPE per DNA
phosphate, corresponding from 8/1 to 0.25/1 DNA/lipid
molar ratio, that is from 0.75 ␮l to 24 ␮l of GL-67/DOPE
solution. Complexes were formed by adding the solution
of the transfection reagent into the DNA solution and
waiting for 15 min at room temperature. For in vitro
transfections, DNA complexes with PEI 25K, PEI 22K,
DOGS or GL-67/DOPE were formed in 150 mm NaCl;
with DOTAP in 20 mm Hepes buffer, pH 7.4; with
DOTMA/DOPE in OPTI-MEM (Gibco BRL). For in vivo
experiments, all the complexes tested were formed in 5%
w/v glucose.
In vitro transfections
Cells were seeded on to 24-well plates at 40 000 per well
to obtain 60–70% confluency after an overnight incubation. Complexes prepared with 2 ␮g of DNA in 100 ␮l
of transfection solution were added to each well. All the
experiments with PEI 22K, PEI 25K, DOTAP and GL67/DOPE were performed in the presence of serum,
except where indicated. FCS (10%) was added to the
transfection medium after 3–6 h incubation of DOGS or
DOTMA/DOPE complexes with the cells. The plates
were centrifuged for 10 min at 410 g. Cells were incubated with the complexes for 24 h and then evaluated for
transgene expression.
Assay for luciferase activity in vitro
Twenty-four hours after transfection, cells were rinsed
twice in PBS and then lysed with 100 ␮l of cell lysis buffer
(25 mm Tris/HCl, 2 mm DTT, 2 mm EDTA, 10% glycerol,
1% Triton X-100, pH 8.0). Twenty microliters of cell lysate
were mixed with 70 ␮l of luciferase assay substrate
(Promega, Madison, WI, USA) at room temperature and
the light emission (integrated over 30 s) was quantified
with an LB 9501 luminometer (Berthold, Bad Willbad,
Germany). Relative light units (RLU) were normalized to
the protein content of each sample, determined according
to Bradford’s assay for protein concentration (Bio-RAD,
München, Germany).
Cytochemical assay for ␤-galactosidase activity in vitro
Individual cells expressing ␤-galactosidase were identified after incubation with 5-bromo-4-chloro-3-indolyl-␤d-galactopyranoside (X-gal). Briefly, the cells were fixed
with a solution of 0.2% glutaraldehyde, 2% formaldehyde
in PBS for 15 min at 4°C, washed twice with PBS, pH 7.5,
and then incubated with a solution containing 1 mg/ml
X-gal dissolved in 5 mm potassium ferricyanate, 5 mm
potassium ferricyanide, and 2 mm magnesium chloride in
PBS (pH 7.4) for 6–8 h at 37°C.
DNase protection assay
pCLuc (576 ng) complexed to the cationic vector was
incubated in a buffer containing 20 mm magnesium
chloride in the presence or absence of 10 U of DNase I
(Boehringer Mannheim). The reaction was carried out for
3 h at 37°C. DNA was isolated by using the phenol–
chloroform extraction and its integrity was examined by
fractionation through a 1% agarose gel and ethidium
bromide staining. Free DNA was used as a control.
Injection of DNA complexes into the tail vein
Complexes were formed by adding 500 ␮l of transfecting
reagent in 5% w/v glucose (at the appropariate ratio to
DNA as described above) into 100 ␮l of DNA solution
in glucose. For the highest DNA/DOTAP (4.8) and GL67/DOPE (4) N/P equivalents 600 ␮l of transfecting
reagent were used. C57Bl/6 mice of 4–6 weeks of age
were used. Individual mice in groups of three were
injected intravenously with 50 ␮g of DNA complexed
with the vector in a total volume of 600–700 ␮l. Twentyfour hours after i.v. injection, mice were killed and
organs were removed and assayed for gene expression.
Assay for luciferase activity in vivo
After reperfusion with 10 ml 150 mm NaCl, lung, heart,
spleen, liver and kidney were collected and washed with
cold saline. The organs were homogenized on ice in 2 ml
of lysis buffer as described above. The homogenates were
centrifuged at 15 800 g for 10 min at 4°C, and 70 ␮l of
the supernatant was analyzed for luciferase activity using
a single-well luminometer for 30 s. Luciferase activity
was measured employing a luciferase assay kit
(Promega). In parallel, purified luciferase (Boehringer
Mannheim) dissolved in the supernatant of homogenates
from the different organs of nontreated animals was used
as a standard. In these conditions, a conversion factor of
15160 RLU/pg luciferase was obtained for the lungs, 3106
RLU/pg for the liver, 7976 RLU/pg for the kidney, 10456
for the spleen, and 7058 for the heart.
In situ ␤-gal assay on lung cryosections
The lungs were reperfused through the heart with 10 ml
of normal saline solution. Lungs were removed by dissection, rinsed in normal saline solution and fixed for 60 min
at room temperature in 4% paraformaldehyde. The staining was performed in potassium phosphate buffer
100 mm (pH 7.4), 5 mm EGTA, 2 mm magnesium chloride, 0.01% Na deoxycholate, 0.02% Nonidet P-40,
0.5 mg/ml X-gal, 10 mm potassium ferricyanate, and
10 mm potassium ferricyanide overnight at 37°C. The
organs were cryoprotected in 10% sucrose and subsequently frozen samples were included in OCT (Miles,
Elkhart, IN, USA), cryosectioned into 8-␮m sections with
a cryostate, and mounted on glass slides. Slides were subsequently washed in normal saline solution, counterstained with Harris Hematoxylin (Sigma), and
coverslipped with aqueous mounting media (Electron
Microscopy Science, Washington, PA, USA). Blue cells
were detected by direct light microscopy.
Systemic gene transfer to the lungs
A Bragonzi et al
Immunohistochemical stainings
The rabbit polyclonal antibody raised against human surfactant protein A (SP-A) was kindly provided by Dr J
Alcorn (Southwestern Medical Center, Dallas, TX, USA)
and identifies type II penumocytes.40 The rat IgG2b
monoclonal antibody M1/70, a kind gift of Dr I Simon,
Harvard Medical School, Boston, MA, USA, was raised
against the unique ␣ subunit of mouse Mac-1 (CD11b)41
and identifies tissue macrophages. Immunohistochemistry was carried out using the avidin–biotin–peroxidase
detection system (ABC Vectastain Elite kit; Vector Laboratories, Burlingame, CA, USA). Briefly, incubation of
cryosections with the appropriate dilution of the primary
antibody (anti-SPA, 1:25; anti-CD11b, 1:50) was carried
out for 1 h at room temperature in PBS, 1% BSA, 2% normal goat serum. After washing in PBS, sections were
incubated with biotinylated secondary antibody (antirabbit IgG H+L and anti-rat IgG H+L from Vector
Laboratories) for 1 h at room temperature. Detection was
carried out by a chromogenic reaction using 3-amino-9ethyl-carbazole (Sigma) as substrate, according to the
manufacturer’s instructions. Slides were counterstained
with Harris Haematoxylin and mounted with aqueous
mounting media.
Acknowledgements
This study was supported by a grant for Cystic Fibrosis
from the Italian Ministero della Sanità (L 548/93) and the
Associazione Lombarda Fibrosi Cistica. We would like to
thank Dr D Gruenert for the 56FHT80- and CFNPE90cell lines, Dr L Monaco for the pCLuc and pCMV␤n plasmids, Dr J-P Behr for the DOGS cationic lipid, Dr J Alcorn
for the anti-SP-A polyclonal antibody, and Dr I Simon for
the anti-CD11b monoclonal antibody.
References
1 Grubb B et al. Inefficient gene transfer by adenovirus vector to
cystic fibrosis airway epithelia of mice and humans. Nature 1994;
371: 802–806.
2 Korst RJ, Bewig B, Crystal RG. In vitro and in vivo transfer and
expression of human surfactant SP-A- and SP-B-associated protein cDNAs mediated by replication-deficient, recombinant
adenoviral vectors. Hum Gene Ther 1995; 6: 277–287.
3 Epperly M et al. Prevention of late effects of irradiation lung
damage by manganese superoxide dismutase gene therapy.
Gene Therapy 1998; 5: 196–208.
4 Ferkol T et al. Transfer of the human alpha1-antitrypsin gene
into pulmonary macrophages in vivo. Am J Respir Cell Mol Biol
1998; 18: 591–601.
5 Fasbender AJ, Zabner J, Welsh MJ. Optimization of cationic
lipid-mediated gene transfer to airway epithelial. Am J Physiol
1995; 13: L45–L51.
6 McLachlan G et al. Evaluation in vitro and in vivo of cationic
liposome-expression construct complexes for cystic fibrosis gene
therapy. Gene Therapy 1995; 2: 614–622.
7 Biffi A et al. Restoration of bacterial killing activity of human
respiratory cystic fibrosis cells through cationic vector-mediated
CFTR gene transfer. Hum Gene Ther 1999; 10: 1923–1930.
8 Stribling R et al. Aerosol gene delivery in vivo. Proc Natl Acad
Sci USA 1992; 89: 11277–11281.
9 Alton EWFW et al. Non-invasive liposome-mediated gene delivery can correct the ion transport defect in cystic fibrosis mutant
mice. Nat Genet 1993; 5: 135–142.
10 Hyde LJ et al. Correction of the ion transport defect in cystic
fibrosis transgenic mice by gene therapy. Nature 1993; 362:
250–255.
11 Eastman SJ et al. Optimization of formulations and conditions
for the aerosol delivery of functional cationic lipid:DNA complexes. Hum Gene Ther 1997; 8: 313–322.
12 Zhu N et al. Systemic gene expression after intravenous DNA
delivery into adult mice. Science 1993; 261: 209–211.
13 Liu Y et al. Cationic liposome-mediated intravenous gene delivery. J Biol Chem 1995; 270: 24864–24870.
14 Thierry AR et al. Systemic gene therapy: biodistribution and
long-term expression of a transgene in mice. Proc Natl Acad Sci
USA 1995; 92: 9742–9746.
15 Thierry AR et al. Characterization of liposome-mediated gene
delivery: expression, stability and pharmacokinetics of plasmid
DNA. Gene Therapy 1997; 4: 226–237.
16 Li S, Huang L. In vivo gene transfer via intravenous administration of cationic lipid–protamine–DNA (LPD) complexes. Gene
Therapy 1997; 4: 891–900.
17 Liu Y et al. Factors influencing the efficiency of cationic liposome-mediated intravenous gene delivery. Nat Biotechnol 1997;
15: 167–173.
18 Griesenbach U et al. Comparison between intratracheal and
intravenous administration of liposome–DNA complexes for
cystic fibrosis lung gene therapy. Gene Therapy 1998; 5: 181–188.
19 Boussif O et al. A versatile vector for gene and oligonucleotide
transfer into cells in culture and in vivo: polyethylenimine. Proc
Natl Acad Sci USA 1995; 92: 7297–7301.
20 Boussif O, Zanta MA, Behr J-P. Optimized galenics improve in
vitro gene transfer with cationic molecules up to 1000-fold. Gene
Therapy 1996; 3: 1074–1080.
21 Boletta A et al. Nonviral gene delivery to the rat kidney with
polyethylenimine. Hum Gene Ther 1997; 8: 1243–1251.
22 Ferrari S et al. ExGen 500 is an efficient vector for gene delivery
to lung epithelial cells in vitro and in vivo. Gene Therapy 1997; 4:
1100–1106.
23 Goula D et al. Polyethylenimine-based intravenous delivery of
transgenes to mouse lung. Gene Therapy 1998; 5: 1291–1295.
24 Felgner PL et al. Lipofection: a highly efficient, lipid-mediated
DNA-transfection procedure. Proc Natl Acad Sci USA 1987; 84:
7413–7417.
25 Song YK et al. Characterization of cationic liposome-mediated
gene transfer in vivo by intravenous administration. Hum Gene
Ther 1997; 8: 1585–1594.
26 Abdallah B et al. A powerful nonviral vector for in vivo gene
transfer into the adult mammalian brain: polyethylenimine.
Hum Gene Ther 1996; 7: 1947–1954.
27 Dunlop DD et al. Nanoscopic structure of DNA condensed for
gene delivery. Nucleic Acids Res 1997; 25: 3095–3101.
28 Ogris M et al. The size of DNA/transferrin–PEI complexes is an
important factor for gene expression in cultured cells. Gene Therapy 1998; 5: 1425–1433.
29 Goula D et al. Size, diffusibility and transfection performance of
linear PEI/DNA complexes in the mouse central nervous system. Gene Therapy 1998; 5: 712–717.
30 Schwartz DA et al. CpG motifs in bacterial DNA cause inflammation in the lower respiratory tract. J Clin Invest 1997; 100:
68–73.
31 Plank C et al. Activation of the complement system by synthetic
DNA complexes: a potential barrier for intravenous gene delivery. Hum Gene Ther 1996; 7: 1437–1446.
32 Bragonzi A et al. Glomerular filtration is required for transfection of proximal tubular cells in the rat kidney following injection of DNA complexes in to the renal artery (submitted,
revised).
33 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.
34 Song YK, Liu F, Liu D. Enhanced gene expression in mouse lung
by prolonging the retention time of intravenously injected plasmid DNA. Gene Therapy 1998; 5: 1531–1537.
35 Song YK, Liu D. Free liposomes enhance the transfection activity
of DNA/lipid complexes in vivo by intravenous administration.
Biochim Biophys Acta 1998; 1372: 141–150.
2003
Systemic gene transfer to the lungs
A Bragonzi et al
2004
36 McCray PB et al. Alveolar macrophages inhibit retrovirusmediated gene transfer to airway epithelia. Hum Gene Ther 1997;
8: 1087–1093.
37 McCray P, Wohlford-Lenane C, Snyder J. Localisation of cystic
fibrosis transmembrane conductance regulator mRNA in human
fetal lung tissue by in situ hybridization. J Clin Invest 1992; 90:
619–625.
38 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press: Cold Spring
Harbor, 1989.
39 Behr J-P et al. Efficient gene transfer into mammalian primary
endocrine cells with lipopolyamine-coated DNA. Proc Natl Acad
Sci USA 1989; 86: 6982–6986.
40 Alcorn JL et al. Primary cell culture of human type II pneumocytes: maintenance of a differentiated phenotype and transfection with recombinant adenoviruses. Am J Respir Cell Mol Biol
1997; 17: 672–682.
41 Springer TA et al. Mac-1: a macrophage differentiation antigen
identified by monoclonal antibody. Eur J Immunol 1979; 9:
301–306.