Gene Therapy (2000) 7, 1753–1760
 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
www.nature.com/gt
NONVIRAL TRANSFER TECHNOLOGY
RESEARCH ARTICLE
Biodistribution and transgene expression with nonviral
cationic vector/DNA complexes in the lungs
A Bragonzi1, G Dina2, A Villa2, G Calori3, A Biffi1, C Bordignon1, BM Assael4 and M Conese1
1
Institute for Experimental Treatment of Cystic Fibrosis, San Raffaele Scientific Institute, Milano; 2MIA, DIBIT, San Raffaele
Scientific Institute and University of Milano-Bicocca, Milano; 3Epidemiology Unit, San Raffaele Scientific Institute; and
4
Department of Pediatrics, University of Milano, Italy
Biodistribution of nonviral cationic vector/DNA complexes
was studied after systemic or intratracheal administration to
the lungs and correlated with transgene expression. Intravenous injection in C57Bl/6 mice gave maximal and significant luciferase expression in the lungs with the cationic polymer PEI 22K/DNA complexes at the highest ratios of
positive/negative charges versus DNA alone. While
DOTAP/DNA complexes with high charge ratio determined
lower but still significant luciferase activity versus uncomplexed DNA, GL-67A and PEI 25K mediated negligible
luciferase expression. Labelled PEI 22K and DOTAP complexes were evenly distributed in the alveolar region, where
GFP expression was revealed, while PEI 25K and GL-67A
complexes were not detected, suggesting a different interaction of these complexes with the plasma membrane of endothelial cells. Following an intratracheal injection, the highest
and significant levels of transfection were obtained with
slightly positive PEI complexes as compared with DNA
alone, whereas cationic lipid-based vectors, DOTAP and
GL-67A, gave not significant luciferase activity. Both types
of polyplexes gave similar levels of lung luciferase
expression by targeting different airway cell populations. PEI
25K complexes determined high levels of GFP in the bronchial cells, confirming confocal data on fluorescent complexes internalization. PEI 22K complexes gave mainly high
GFP signal in the distal tract of the bronchial tree, where
tagged complexes were recovered. Fluorescent lipid complexes were found in aggregates in the lumen of bronchi totally (DOTAP) or partially (GL-67A) co-localizing with surfactant protein A. Results indicated that cationic polymers could
overcome the surfactant barrier which inhibited airway cell
transfection mediated by cationic lipids. Gene Therapy
(2000) 7, 1753–1760.
Keywords: gene transfer; cationic polymers; lipids; intravenous; intratracheal; barriers
Introduction
Gene delivery to the airway epithelium is a major goal
for lung gene therapy, but success depends on overcoming the biological and physicochemical barriers which the
vector encounters in the tracheo-bronchial tree. This concept has been highlighted when demonstration of limited
efficient gene transfer was obtained in human airways,
as in the case of cystic fibrosis (CF) patients with both
nonviral and viral vectors.1–4 Considering only nonviral
DNA delivering agents, several previous works have achieved gene transfer with cationic lipids (lipofection) and
polymers (polyfection) to the lungs of laboratory animals
by either direct instillation in the nose and trachea5–7 or
via the aerosol administration.8,9 Systemic intravenous
infusion of cationic vector/DNA complexes has been
attempted and shown to transfect preferentially the lung
tissue.7,10–13 The most important determinants for cationic
lipid-mediated gene transfer through intravenous administration are the cationic lipid structure, the charge ratio
(+/−) between the cationic vector and DNA,14–16 and the
Correspondence: M Conese, Institute for Experimental Treatment of Cystic Fibrosis, San Raffaele Scientific Institute, Via Olgettina 58, Milano
20132, Italy
Received 28 February 2000; accepted 22 June 2000
size of liposomes.11 Charge ratio between cationic polymers (ie PEI) and DNA has also been shown to be determinant in obtaining efficient lung transfection.12 A formulation suitable for intravenous injection may not work
when administered in trachea. Few studies have
addressed this hypothesis and showed that in case of
liposome DODAC/DOPE the same ratio of lipid to DNA
was optimal for both intravenous and intratracheal
administrations.7 On the other hand, neutral complexes
formed between linear PEI and DNA gave higher levels
than positive complexes when injected in the trachea of
newborn rabbits.6 The observation that tracheal insufflation of plasmid DNA resulted in transfection of rat
lungs to the same extent as injection of plasmid–cationic
liposome complex17 has thus brought us to investigate
the role of surfactant in cationic lipid-mediated gene
transfer to the lungs. In vitro studies have demonstrated
that natural and synthetic preparations of surfactant
markedly inhibited gene transfer mediated by cationic
liposome,18 but not by PEI and dendrimers.19 However,
the role of surfactant has not been studied in vivo.
While previous studies have addressed the physicochemical properties of lipoplexes and polyplexes, the
tissue factors that regulate the transfection efficiency of
nonviral vectors are still largely unknown. Indeed, the
efficacy of gene therapy would also depend on the cells
Barriers to gene transfer to the lungs
A Bragonzi et al
1754
targeted by the nonviral vectors. The respiratory cell type
to be transfected could be different in diseases which
involve either bronchial/bronchiolar cells, such as CF,2
or the alveolar cells, like neonatal respiratory distress
syndrome.20 We have thus undertaken this study to
explore the biodistribution of cationic vector/DNA complexes in the lungs and the cell types which are targeted
directly comparing two ways of administration: intravenous and intratracheal.
Our results show that the same cationic vector/DNA
complexes behaved differently depending on the route of
administration. Following a single intravenous injection,
the main constraint to gene transfer was represented by
the affinity of the complexes for the plasma membrane
of endothelial cells, while for the intratracheal delivery
the interaction with the surfactant played a major role.
Moreover, immunohistochemistry studies revealed that
the cationic vectors had different cellular targets when
administered through the trachea.
Results
Luciferase expression in the lungs
Transfection efficiency of gene delivery to the airway epithelium was studied comparing two different classes of
cationic vectors after systemic (intravenous) or local
(intratracheal) administration. We have previously
shown that a single injection of DNA complexed to the
cationic polymer polyethylenimine (PEI) 22K into the tail
vein of adult mice efficiently transfected primarily the
lung and, to a lesser extent, heart, spleen, kidney, and
liver.13 As shown in Figure 1a, the other vectors seemed
to mediate lower levels of luciferase expression in the
lungs, with DOTAP ⬎ PEI 25K ⬎ GL-67A. The most positive complexes gave the highest luciferase levels in the
lungs and the maximal activity was obtained with PEI
22K at 15 equivalents of amino-to-phosphate (N/P)
groups (corresponding to 2.4 +/− charge ratio). No
luciferase expression was obtained with DNA alone or
with complexes containing a control plasmid coding for
␤-galactosidase. Statistical analysis revealed that PEI 22K
(15 N/P) and DOTAP (3.6 and 4.8 N/P) mediated luciferase levels which were significantly higher than those achieved by DNA alone. No significant difference in luciferase expression was observed with PEI 25K/DNA or GL67A/DNA complexes versus DNA alone for all N/P equivalents.
In the case of intratracheal injections, the same complexes behaved in a different manner. The cationic lipid
DOTAP mediated from low to undetectable levels of
expression with the most positive complexes (4.8 N/P
equivalents, 4.8 +/− charge ratio) (Figure 1b). The other
cationic lipid-based vector GL-67A seemed to give higher
levels of lung luciferase than DOTAP with the less positive complexes (0.25–0.5 N/P, corresponding to 0.75–1.5
+/− charge ratio), but nevertheless showed a decrease as
DOTAP as the amount of the vectorial component
increased. PEI-mediated transfection determined significantly higher levels and maximal luciferase activity with
slightly positive complexes (5–10 N/P equivalents, 0.8–
1.7 +/− charge ratios) (Figure 1b). Lung luciferase levels
in mice transfected with PEI 22K and PEI 25K were the
highest and the most reproducible in all the animals.
Uncomplexed DNA gave detectable expression only in
four mice out of eight, indicating a poor reproducibility
in lung transfection. Statistical analysis showed that PEI
22K (10 N/P) determined significant higher luciferase
levels than uncomplexed DNA. When the conventional
cut-off of 102 RLU/mg protein was considered, also PEI
25K (5 N/P) resulted in significantly higher luciferase
expression than DNA alone. On the contrary, both cationic lipid-based vectors DOTAP and GL-67A did not
reach significance versus DNA alone with all the N/P
ratios, although in the case of GL-67A (0.25 N/P) the
mean luciferase level was similar to that of PEI 22K (10
N/P).
Figure 1 Effect of DNA/vectors charge ratio on luciferase gene expression in mice lungs after intravenous and intratracheal injections. C57Bl/6 mice
were injected in the tail vein (a) or instilled in the trachea (b) with 50 ␮g of pCLuc complexed to different equivalents of cationic vectors as detailed
(N/P). DNA alone was also injected as experimental control. The luciferase expression was evaluated in the lungs after 24 h from transfection. Bars
represent means +/− s.d. (n = 3–9); squares represent individual mice measurements. Medians are shown as dashes inside the bars. Kruskal–Wallis test
for comparisons among groups, each group comprising of different N/P equivalents and DNA alone: (a) PEI 22K, P = 0.01; PEI 25K, P = 0.02; DOTAP,
P ⬍ 0.001; GL-67A, P = 0.2. (b) PEI 22K, P = 0.03; PEI 25K, P = 0.005; DOTAP, P = 0.6; GL-67A, P = 0.4. (*) Only significance (P ⬍ 0.05) versus
DNA alone is reported. (°) Chi-square test: statistically significant at P ⬍ 0.05 versus DNA alone.
Gene Therapy
Barriers to gene transfer to the lungs
A Bragonzi et al
These results showed that the same complexes transfected the lungs with different efficiency when injected
intravenously or in the trachea, suggesting that different
barriers to gene transfer are involved in the two routes
of administration.
Biodistribution of the complexes
Cationic vector/DNA complexes were fluorescently
tagged by inclusion of 0.4 N/P equivalents of FITC-polyL-lysine (pLys) to follow their biodistribution after injection. The transfection efficiency of labeled complexes was
not affected either in vitro or in vivo (data not shown). In
order to assure the co-localization between FITC-pLys
and DNA, the plasmid was covalently labeled by
incorporating dUTP-Texas Red with a nick translation
system. A549 cells were incubated with double labeled
complexes and the transfection was stopped at 1 h of
incubation. The signal corresponding to both Texas Redplasmid DNA and FITC-pLys was observed in the cells
and co-localized completely either in DOTAP (Figure 2a
and b) or PEI 22K complexes (Figure 2d and e). These
results showed that the FITC-pLys was present in the
complexes as long as 1 h after the transfection.
FITC-labeled complexes were injected in the tail vein
of the C57Bl/6 mice, the animals were killed after 1, 10
and 30 min from the injection, the lungs were cryosectioned and observed at confocal microscopy. After 10
min, the FITC-complexes were evenly distributed in the
alveolar region either with DOTAP/DNA (3.6 N/P)
(Figure 3a) or PEI 22K/DNA (15 N/P) (Figure 3c) complexes. No substantial difference was seen at the different
time-points indicating that the complexes reached the
lungs in large amount in a time as short as 1 min and
remained in the organ for at least 30 min. Examination
of lung cryosections of nontransfected mice or injected
with DNA/FITC-pLys revealed only background autofluorescence (not shown). The reperfusion of the lungs
before their removal showed no difference in the distribution pattern suggesting the presence of the complexes
in one of the cellular compartments of the alveoli. Intravenous injections were also performed with GL-67A (2
a
d
b
e
N/P) (Figure 3b) or PEI 25K/DNA complexes (10 N/P)
(Figure 3d) containing FITC-pLys. No fluoresceinated
complexes were observed in the lung cryosections at the
same time-points considered before. Biodistribution of
fluoresceinated complexes is in accordance with luciferase levels, suggesting that after an intravenous injection
the first barrier to be overcome is the interaction of complexes with endothelial plasma membrane.
Intratracheal injection of labeled complexes revealed a
different signal pattern for the various cationic vectors in
epithelial bronchial cells (Figure 4a, c, e, and g) and in
the alveolar region (Figure 4b, d, f, and h) in animals
killed after 30 min. PEI 25K complexes (10 N/P) were
mainly localized in the epithelial bronchial cells (Figure
4g), with very few complexes in the alveolar region
(Figure 4h). Lower amounts of PEI 22K complexes (10
N/P) were detected inside bronchial cells (compare Figure 4e and 4g). More DNA/FITC-pLys/PEI 22K complexes were recovered at the level of the alveolar region
(Figure 4f). In contrast with PEI complexes, DOTAP (2.4
N/P) and GL-67A (0.5 N/P) showed aggregates in the
lumen of bronchi (Figure 4a and c), with more GL-67A
than DOTAP complexes detected in the alveolar region
(compare Figure 4b and d). Immunofluorescence with an
antibody directed against human surfactant protein A
(TRITC signal, not shown) revealed a complete or a partial co-localization (yellow signal) between lipid complexes and surfactant in the bronchial lumen, for DOTAP
and GL-67A respectively (Figure 4a and c, insets). At
variance with what was seen after the intravenous
injections, we observed patched distribution of labeled
complexes in the bronchial tree, independently of the
vector used.
Confocal data reveal a potential role for pulmonary
surfactant as a barrier to transfection mediated by cationic lipids but not by cationic polymers, as confirmed
by luciferase data.
a
b
c
d
1755
c
f
Figure 2 Confocal microscopy of A549 cells transfected with doubly-labeled complexes. A549 cells were transfected for 1 h with doubly-labeled
DOTAP (a–c) or PEI (d–f) complexes containing FITC-pLys (a and d)
and Texas-Red labeled DNA (b and e). Transmission images of cells with
superimposed FITC-pLys fluorescence are shown in (c) and (f). Co-localization between Texas-Red DNA fluorescence and FITC-pLys was complete.
Bar: 12 ␮m.
Figure 3 Biodistribution of fluorescent complexes in the lungs after intravenous administration. Plasmid DNA was partially pre-condensed with
0.4 N/P equivalents of FITC-pLys followed by complete condensation with
DOTAP (a), GL-67A (b), PEI 22K (c) or PEI 25K (d). Complexes were
injected in the tail vein in C57Bl/6 mice and the lungs were removed 10
min after transfection. Cryosections of the lungs at this time-point showed
DOTAP labeled complexes equally distributed in all the alveolar areas (a)
like PEI 22K complexes (c), while GL-67A (b) and PEI 25K (d) complexes
were not detected. Bar: 10 ␮m.
Gene Therapy
Barriers to gene transfer to the lungs
A Bragonzi et al
1756
single GFP-positive cells both at the bronchial and respiratory bronchiolar level (Figure 5e). No GFP-expressing
cells were detected after the intratracheal administration
of DOTAP complexes (not shown).
Discussion
a
b
c
d
e
f
g
h
Figure 4 Biodistribution of fluorescent complexes in the lungs after an
intratracheal injection. Complexes were prepared as for an intravenous
administration and injected into the mice tracheas. Lungs were removed
30 min after transfection. Cryosection of the lungs showed DOTAP (a)
and GL-67A (c) complexes aggregated in the bronchial lumen, with few
DOTAP (b) and many GL-67A (d) complexes in the alveolar region.
Immunofluorescence against human-surfactant protein A (TRITC, not
shown) revealed a co-localization with FITC-complexes (a and c, insets).
In contrast, PEI 22K complexes lined the apical part of bronchial epithelial
cells (e) and were also found in the alveoli (f). PEI 25K complexes distributed mostly in the cytosol of bronchial epithelial cells (g), with very few
polyplexes in the alveolar region (h). Bar: 15 ␮m.
Localization of transgene expressing cells
Transgene expression in mouse lung was detected
performing immunohistochemistry with an antibody
directed against green fluorescent protein (GFP). Intravenous injections of GFP-coding plasmid complexed with
PEI 22K gave positive signal in the alveolar region comprising endothelial cells and pneumocytes (Figure 5a),
confirming the previous observations with ␤-galactosidase transgene.13 After an intratracheal injection, different cell targets were obtained with PEI 25K, PEI 22K, and
GL-67A. PEI 25K complexes determined GFP positivity
in bronchial cells (Figure 5c), with lower signal in the
alveolar region (not shown). In contrast, PEI 22K complexes gave mainly high GFP signal in the distal tract of
the bronchial tree (alveolar ducts) (Figure 5d), whereas
few positive airway cells in the bronchi were detected.
GL-67A transfection resulted in equally distributed few
Gene Therapy
Intravenous and intratracheal delivery of DNA complexes have been envisioned as two strategies to obtain
gene transfer to the lungs. A direct comparison between
these two routes of administration was assessed considering lipid and polymeric classes of nonviral vectors. Data
showed that complexes with the same charge ratio
behaved differently according to the route of administration. In the systemic delivery, the more positive the
complexes the higher the lung luciferase levels. The
requirement of high +/− charge ratio for efficient intravenous gene transfer has also been demonstrated for
other cationic vectors such as DOTMA/DOPE and
DOLCE/DOPE21 and LPD (lipid-protamine-DNA).16 It is
possible that the excess amount of positive charges
required for an optimal transfection through this route is
essential in neutralizing negatively charged blood
components, that otherwise could inhibit the activity of
the DNA/cationic vector complexes at low charge
ratio.14,22 Moreover, the role of cationic vectors is thought
to protect plasmid DNA from DNase degradation in the
presence of serum.13,23 Indeed, naked DNA gave negligible gene expression when administered intravenously,
as shown by us in this report and by others.17,21
The efficiency of gene expression differed greatly
among the vectors we have tested (Ref. 13 and data
shown herein). Chemical structure of the vectors and cationic vector/DNA ratio might determine different sizes
of complexes that could be relevant for an efficient gene
transfer through intravenous administration. The size of
the DOTAP/DNA complexes was shown to be 181–280
nm, GL-67A complexes were 400–600 nm (SH Cheng,
personal communication), while 60–80 nm were obtained
for PEI 25K and 22K complexes.24,25 The effect of particle
size determined five-fold higher levels of expression for
complexes prepared using 400- or 800-nm extruded liposomes as compared with those prepared using 100-nm
extruded liposomes.21 As shown for MLV and
LUV/DNA complexes,7 lipid complexes might become
trapped in capillaries of lung when injected intravenously, due to their large size. On the other hand,
highly compacted PEI/DNA complexes might circulate
faster than lipoplexes and consequently the contact time
between polyplexes and cell surface may be reduced
resulting in decreased transfection efficiency. Our data on
the biodistribution in the lungs of the different types of
complexes demonstrated that factors other than size
might have significant importance. In fact, polyplexes
which ranged in the same size were differently deposited
in the lung 10 min after the administration: PEI 22K complexes were evenly distributed, whereas no labeled PEI
25K/DNA complexes were detected justifying different
levels of expression. Moreover, the large GL-67A complexes were not retained in the lungs, as shown by the
absence of fluorescence at 10 min. Overall, these data
suggest that high affinity of complexes for the endothelial
luminal plasma membrane may be essential for an
efficient lung gene transfer through the systemic route.
After an intravenous injection, cationic vector/DNA
Barriers to gene transfer to the lungs
A Bragonzi et al
1757
a
b
c
d
e
f
Figure 5 Immunohistochemical detection of GFP transgene expressing cells. Intravenous injection: a positive signal in alveoli of mice injected with PEI
22K complexes (a). Intratracheal injection: bronchial epithelial positive cells in mice injected with PEI 25K complexes (c) and distal tract of bronchoalveolar tree injected with PEI 22K complexes (d). Single bronchial epithelial cells and respiratory bronchiolar cells (e) in mice injected with GL-67A
complexes. Immunohistochemistry on untransfected mice (negative controls) showed no staining (b and f). Pictures are representative of the lung sections
obtained from three animals. Bar: 10 ␮m.
complexes gave GFP expression in the alveolar region,
confirming our previous results with X-gal staining for
␤-galactosidase.13 Although DOTAP and PEI 22K labeled
complexes were biodistributed in a similar fashion, the
transgene levels in the lungs were higher for PEI 22K. It
may be possible that highly compacted PEI 22K complexes favorably cross the endothelium, as was shown for
pulmonary endothelium in mice26 and endothelial glomerular barrier in the rat kidney.25 Moreover, greater diffusibility of these complexes related to their small size
and stability was also demonstrated in the mouse central
nervous system.27
Transfection through intratracheal administration
involved different factors for successful gene delivery
into the lungs as demonstrated by injecting the same
DNA complexes used for systemic administration. The
highest levels of transfection were obtained using PEIs
with low ratio of positive charges. As opposed to systemic delivery, intratracheal-injected complexes do not
have to interact with the negatively charged blood
components. Thus, less positive complexes are needed for
high levels of lung transfection, as in the case of intraventricular injection in the mouse central nervous system.27
Confocal microscopy studies revealed that lipoplexes, but
not polyplexes, aggregated in the lumen of bronchi and
co-localized either totally (DOTAP) or partially (GL-67A)
with the surfactant protein A. GL-67A was chosen since
it was shown to be a more efficient lipid formulation than
other commercial lipids, such as DOTAP, for mouse and
human lung transfection.9,28 It includes a neutral lipid,
DOPE, and a polyethylene glycol-containing lipid,
DMPE-PEG5000, which stabilizes the bilayer configuration. It could be possible that these adjunctive components might have influenced the interaction with surfactant. These data confirm for the first time in vivo, the
previous in vitro findings of the inhibition of cationic
lipid-mediated gene transfer by natural and synthetic
surfactant preparations.18,19 The lack of inhibition by surfactant of cationic polymer-mediated gene transfer is
similar to what has been seen using adenoviral vectors.
Indeed, it has been shown that surfactant facilitates adenovirally mediated gene transfer in the peripheral lung
and enhances transgene expression.29
Both types of polyplexes gave similar levels of lung
luciferase expression by targeting different airway populations. PEI 25K gave high levels of GFP in the bronchi,
confirming confocal data on complex internalization by
this pulmonary area, while GFP expression and FITC signal were correspondingly low in the alveoli. PEI 22K
complexes were found on the apical surface of bronchial
cells and gave low levels of GFP in the bronchi, whereas
more polyplexes and corresponding GFP expression
were revealed in the distal tract of the broncho-alveolar
tree. Therefore, PEI 22K and 25K complexes could be useful for obtaining therapeutic gene transfer in different
lung diseases, depending on the cellular population to be
Gene Therapy
Barriers to gene transfer to the lungs
A Bragonzi et al
1758
targeted.30,31 GL-67A transfection resulted in a low GFP
expression in both the bronchial and the distal pulmonary districts. This distribution may reflect some complexes which escaped the surfactant barrier. However,
the degree of cellular correction in the CF epithelium is
in the order of 5%,32 indicating that GL-67A could be of
therapeutic relevance in this disease.
A number of studies have demonstrated that intratracheal administration of naked DNA results in equal or
better in vivo gene transfer to the airway as compared
with the administration of plasmid–cationic liposome
complexes in the mouse,33 rat,17 and man.34 Our data
show that intratracheal injection of DNA resulted in gene
expression in mouse airways in the absence of any
enhancing agent; however, naked DNA gave positive
lung transfection in half of the treated mice, demonstrating low reproducibility in gene transfer. On the other
hand, DNA/cationic polymer complexes gave high, significant and reproducible lung transfection and therefore
constitute a better choice for directly targeting the
airways than does naked DNA.
In summary, our results suggest that the interaction of
complexes with plasma membrane of endothelial cells
(likely due to affinity for a cellular receptor) could play
a relevant role in the efficiency of cationic vectormediated gene delivery to the lungs after systemic
administration. Moreover, the intratracheal injections
show that polyplexes could overcome the surfactant barrier which inhibit lung transfection by cationic lipids.
These data would indicate that polyplexes, and in particular PEI 25K, might be more suitable vectors among
those used so far in preclinical and clinical studies for a
gene therapy approach to lung monogenic diseases, like
CF, in which bronchial cells must be targeted.2
Materials and methods
Plasmids
Plasmid pCLuc carrying Photinus pyralis luciferase coding
region under the control of the cytomegalovirus (CMV)
immediate–early enhancer/promoter region was a kind
gift of Lucia Monaco (HS Raffaele, Milano, Italy). pEGFP-C3 carrying the coding region for an enhanced version of Aequorea victoria GFP was purchased from Clontech (Palo Alto, CA, USA). Plasmid DNA was purified
by the double CsCl-ethidium bromide gradient centrifugation method.35 Purity of the plasmids was determined
by absorbance at 260 and 280 nm and by 1% agarose
gel electrophoresis.
Labeled DNA was prepared by incorporating dUTPTexas Red (ChromaTide Texas Red-12-dUTP, Molecular
Probe, Eugene, OR, USA). Fluorescent nucleotides were
covalently linked using the Nick Translation System
(Gibco-BRL, Life Technologies, San Giuliano, Milan,
Italy). Separation between labeled DNA from non-incorporated nucleotides was performed with Sephadex G-50
Fine column (Amersham Pharmacia Biotech, Uppsala,
Sweden).
Preparation of cationic vector/DNA complexes
The complex formation of DNA with linear PEI 22K
(Exgen 500, Euromedex, Souffelweyersheim, France),
branched PEI 25K (Sigma-Aldrich, St Louis, MO, USA),
and DOTAP (Roche Molecular Biochemicals, Mannheim,
Gene Therapy
Germany) was carried out as previously described.13 GL67/DOPE/DMPE/PEG5000 (GL-67A; kindly provided
by SH Cheng, Genzyme, Framingham, MA, USA) was
formulated at 1:2:0.05 molar ratio9 and reconstituted as 2
mm aqueous solution. From 0.25 to 4 N/P equivalents of
GL-67A per DNA phosphate were used, that is from 0.75
to 12 ␮l per 2 ␮g DNA. 16.1 kDa FITC-conjugated polyl-lysine hydrobromide (FITC-pLys) (Sigma-Aldrich) was
prepared as 1 mm stock solution in water. Labeled complexes containing FITC-pLys were prepared by pre-incubation of 2.7 ␮l of 1 mm FITC-pLys per microgram DNA,
corresponding to 0.4 N/P equivalents, for 15 min at room
temperature, followed by addition of cationic vectors.
In vitro transfections
A549 (human type II pneumocytes) cell line was from
ATCC and was cultured in DMEM/HAMF12 (1/1), 10%
fetal calf serum (Hyclone, Logan, UT, USA), 2 mm l-glutamine, 100 U/ml penicillin, 100 ␮g/ml streptomycin
(Gibco BRL). Cells were seeded in 24-well plates on to
glass coverslips at 50% confluence. One hour after transfection with fluoresceinated complexes, cells were fixed
in 4% paraformaldehyde in PBS for 10 min at room temperature and mounted with aqueous mounting media
(Electron Microscopy Science, Washington, PA, USA).
Samples were analyzed with a confocal microscope
(BioRad MRC 1024; Hercules, CA, USA) equipped with
an argon-krypton laser.
In vivo transfections
C57Bl/6 mice of 4–6 weeks of age were injected either
through the tail vein or in the trachea with cationic
vector/DNA complexes in 5% w/v glucose. 50 ␮g of
pCLuc and 75 ␮g of p-EGFP-C3 were complexed with
the vector in a range of 400–700 ␮l for the tail vein and
in a range of 300–650 ␮l for the intratracheal injection.
Plasmid DNA or DNA/vectors complexes were delivered to the airways by either direct intratracheal injection or using a 24G catheter (Terumo, Leuven, Belgium).
No difference of luciferase expression was observed
between the two procedures. In both cases the trachea
was exposed with a local incision, slowly injected and
the incision was then sutured. Twenty-four hours after
injection, mice were killed and the lungs were removed
and assayed for luciferase or GFP expression.
Evaluation of transgene expression
Quantification of luciferase activity in the mice lungs has
already been described.13 Briefly, transfected lungs were
homogenized on ice in 2 ml of lysis buffer (25 mm
Tris/HCl, 2 mm DTT, 2 mm EGTA, 10% glycerol, 1% Triton X-100, pH 8.0) and centrifuged at 15 800 g for 10 min
at 4°C. Seventy microliters of the supernatant was analyzed for luciferase activity and in parallel, purified
luciferase (Roche), dissolved in the lung homogenates of
nontreated animals, was used as a standard. In these conditions, a conversion factor of 15 160 relative light units
(RLU)/pg luciferase was obtained for the lungs.
GFP expression was evaluated on cryosections by
immunohistochemistry with an anti-GFP polyclonal antibody (Molecular Probes) and using the avidin-biotin-peroxidase detection system (ABC Vectastain Elite kit, Vector Laboratories, Burlingame, CA, USA). Incubation of
cryosections with primary antibody (1:800) was carried
out in PBS/1% BSA/2% normal goat serum for 1 h at
Barriers to gene transfer to the lungs
A Bragonzi et al
room temperature. After extensive washing, lung
cryosections were incubated with biotinylated secondary
antibody (1:400) (anti-rabbit IgG H+L, from Vector
Laboratories). Slides were counterstained with Harris
Hematoxylin (Sigma-Aldrich). Immunohistochemistry
was performed on lung sections taken randomly from all
the lobes of three animals per group and showed a
similar transgene distribution.
Biodistribution of labeled complexes
C57Bl/6 mice transfected with fluoresceinated complexes
were killed after 1–30 min. Lungs were fixed in 4% paraformaldehyde in PBS at 4°C, cryoprotected in 10%
sucrose in PBS at 4°C and frozen in liquid nitrogen.
Organs were included in OCT (Miles, Elkhart, IN, USA),
cryosectioned in 5-␮m thick sections, coverslipped with
aqueous mounting media, and analyzed with a
confocal microscope.
6
7
8
9
10
11
12
Immunofluorescence staining
Lung cryosectioned slides were stained using a rabbit
polyclonal antibody raised against human surfactant protein-A (kindly provided by John Alcorn, Southwestern
Medical Center, Dallas, TX, USA) for 30 min at 37°C in
PBS 1% BSA. Sections were incubated with TRITC-conjugated secondary antibody for 30 min at 37°C, mounted
with aqueous mounting media, and analyzed at
confocal microscope.
13
14
15
16
Statistical analysis
As the variables did not show a Gaussian distribution,
non-parametric tests were used. The comparisons of cationic vectors/DNA at all N/P equivalent and DNA alone
were performed by using Kruskal–Wallis analysis and for
post hoc comparisons the Dunn procedure was used. A
cut-off of 102 RLU/mg protein for luciferase expression
was used to compare the proportion of positive mice
between cationic vectors/DNA N/P equivalents and
DNA alone by means of the Chi-square test. SAS Statistical Package was used for statistical analysis.36
Acknowledgements
This study was supported by a grant for cystic fibrosis
from the Italian Ministero della Sanità (L548/93) and the
Associazione Lombarda Fibrosi Cistica. G Dina was supported by Giovanni Armenise-Harvard Foundation. We
would like to thank SH Cheng for providing us with the
GL-67A vector.
17
18
19
20
21
22
23
24
References
1 Geddes D, Alton EWFW. Cystic fibrosis clinical trials. In: Kabanov AV, Felgner PL, Seymour LW (eds). Self-Assembling Complexes for Gene Delivery. From Laboratory to Clinical Trial. John
Wiley: Chichester, 1998, pp 403–420.
2 Boucher RC. Status of gene therapy for cystic fibrosis lung disease. J Clin Invest 1999; 103: 441–445.
3 Alton EWFW et al. Cationic lipid-mediated CFTR gene transfer
to the lungs and nose of patients with cystic fibrosis: a doubleblind placebo-controlled trial. Lancet 1999; 353: 947–954.
4 Zuckerman JB et al. A phase I study of adenovirus-mediated
transfer of the human cystic fibrosis transmembrane conductance regulator gene to a lung segment of individuals with cystic
fibrosis. Hum Gene Ther 1999; 10: 2973–2985.
5 Oudrhiri N et al. Gene transfer by guanidium-cholesterol cat-
25
26
27
28
ionic lipids into airway epithelial cells in vitro and in vivo. Proc
Natl Acad Sci USA 1997; 94: 1651–1656.
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.
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.
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.
Eastman SJ et al. A concentrated and stable aerosol formulation
of cationic lipid:DNA complexes giving high-level gene
expression in mouse lung. Hum Gene Ther 1997; 8: 765–773.
Templeton NS et al. Improved DNA:liposome complexes for
increased systemic delivery and gene expression. Nat Biotechnol
1997; 15: 647–652.
Liu Y et al. Factors influencing the efficiency of cationic liposome-mediated intravenous gene delivery. Nat Biotechnol 1997;
15: 167–173.
Goula D et al. Polyethylenimine-based intravenous delivery of
transgenes to mouse lung. Gene Therapy 1998; 5: 1291–1295.
Bragonzi A et al. Comparison between cationic polymers and
lipids in mediating systemic gene delivery to the lungs. Gene
Therapy 1999; 6: 1995–2004.
Song YK, Liu F, Chu S, Liu D. Characterization of cationic liposome-mediated gene transfer in vivo by intravenous administration. Hum Gene Ther 1997; 8: 1585–1594.
Liu F, Qi H, Huang L, Liu D. Factors controlling the efficiency
of cationic lipid-mediated transfection in vivo via intravenous
administration. Gene Therapy 1997; 4: 517–523.
Li S, Huang L. In vivo gene transfer via intravenous administration of cationic lipid-protamine-DNA (LPD) complexes. Gene
Therapy 1997; 4: 891–900.
Tsan M-F, White JE, Sheppard B. Lung-specific direct in vivo
gene transfer with recombinant plasmid DNA. Am J Physiol
1995; 268: L1052–L1056.
Tsan M-F, Tsan GL, White JE. Surfactant inhibits cationic liposome-mediated gene transfer. Hum Gene Ther 1997; 8: 817–825.
Ernst N et al. Interaction of liposomal and polycationic transfection complexes with pulmonary surfactant. J Gen Med 1999; 1:
331–340.
Nogee LM et al. A mutation in the surfactant protein B gene
responsible for fatal neonatal respiratory disease in multiple kindreds. J Clin Invest 1994; 93: 1860–1863.
Mahato RI et al. Biodistribution and gene expression of
lipid/plasmid complexes after systemic administration. Hum
Gene Ther 1998; 9: 2083–2099.
Yang J-P, Huang L. Overcoming the inhibitory effect of serum
on lipofection by increasing the charge ratio of cationic liposome
to DNA. Gene Therapy 1997; 4: 950–960.
Li S et al. Dynamic changes in the characteristics of cationic lipidic vectors after exposure to mouse serum: implications for
intravenous lipofection. Gene Therapy 1999; 6: 585–594.
Dunlop DD, Maggi A, Soria MR, Monaco L. Nanoscopic structure of DNA condensed for gene delivery. Nucleic Acids Res
1997; 25: 3095–3101.
Foglieni C et al. Glomerular filtration is required for transfection
of proximal tubular cells in the rat kidney following injection
of DNA complexes into the renal artery. Gene Therapy 2000; 7:
279–285.
Goula D et al. Rapid crossing of the pulmonary endothelial barrier by polyethylenimine/DNA complexes. Gene Therapy 2000;
7: 499–504.
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.
Chadwick SL et al. Safety of a single aerosol administration of
escalating doses of the cationic lipid GL-67/DOPE/DMPEPEG5000 formulation to the lungs of normal volunteers. Gene
Therapy 1997; 4: 937–942.
1759
Gene Therapy
Barriers to gene transfer to the lungs
A Bragonzi et al
1760
29 Jobe AH et al. Surfactant enhances adenovirus-mediated gene
expression in rabbit lungs. Gene Therapy 1996; 3: 775–779.
30 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.
31 Thurston G et al. Cationic liposomes target angiogenic endothelial cells in tumors and chronic inflammation in mice. J Clin
Invest 1998; 101: 1401–1413.
32 Johnson L et al. Efficiency of gene transfer for restoration of normal airway epithelial function in cystic fibrosis. Nat Genet 1992;
2: 21–25.
Gene Therapy
33 Meyer KB et al. Intratracheal gene delivery to the mouse airway:
characterization of plasmid DNA expression and pharmacokinetics. Gene Therapy 1995; 2: 450–460.
34 Zabner J et al. Comparison of DNA–lipid complexes and DNA
alone for gene transfer to cystic fibrosis airway epithelia in vivo.
J Clin Invest 1997; 100: 1529–1537.
35 Sambrook J, Fritsch EF, Maniatis T. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press: New York,
1989.
36 Neave HR, Worthington PL. Distribution-Free Tests. Rontledge:
London, 1988.