Gene Therapy (2000) 7, 568–574
 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
www.nature.com/gt
VIRAL TRANSFER TECHNOLOGY
RESEARCH ARTICLE
Pseudotyped human lentiviral vector-mediated gene
transfer to airway epithelia in vivo
LG Johnson1, JC Olsen1, L Naldini2,3 and RC Boucher1
1
Cystic Fibrosis/Pulmonary Research and Treatment Center and Department of Medicine, The University of North Carolina at
Chapel Hill, Chapel Hill, NC; and 2Cell Genesys, Inc, Foster City, CA, USA
We used a replication defective human lentiviral (HIV) vector
encoding the lacZ cDNA and pseudotyped with the vesicular
stomatitis virus (VSV) glycoprotein (G) to evaluate the utility
of this vector system in airway epithelia. In initial studies,
apical application of vector to polarized well differentiated
human airway epithelial cell cultures produced minimal levels of transgene expression whereas basolateral application
of vector enhanced levels of transduction approximately 30fold. Direct in vivo delivery of HIV vectors to the nasal epithelium and tracheas of mice failed to mediate gene transfer,
but injury with sulfur dioxide (SO2) before vector delivery
enhanced gene transfer efficiency to the nasal epithelium of
both mice and rats. SO2 injury also enhanced HIV vectormediated gene transfer to the tracheas of rodents. These
data suggest that SO2 injury increases access of vector to
basal cells and/or the basolateral membrane of airway surface epithelial cells. Quantification of gene transfer efficiency
in murine tracheas demonstrated that transduction was more
efficient when vector was delivered on the day of exposure
(7.0%, n = 4) than when vector was delivered on the day
after SO2 exposure (1.7%, n = 4). Gene Therapy (2000) 7,
568–574.
Keywords: in vivo; airway epithelium; gene transfer; lentiviral vectors; HIV vectors
Introduction
The development of lentiviral gene transfer vectors has
been a recent advance in the field of gene transfer. Lentiviral vectors are attractive because they have the potential to provide long-term expression due to integration
into the host cell DNA while retaining the ability to transduce non-dividing cells.1–4 Studies in the brains of rats
have established the ability of lentiviral vectors to
accomplish these goals.1,2 Recent advances in human lentiviral (HIV) vector design have reduced safety concerns
regarding inadvertent production of replication competent lentivirus5,6 and the application of pseudotyped
vector technologies has led to production of vectors with
increased titers.1,7,8 Thus, HIV vectors appear to be promising gene transfer vehicles by virtue of their ability to:
(1) infect non-dividing cells; (2) be produced in high titer;
and (3) mediate persistence of transgene expression for
greater than 6 months.1,2
However, the airway epithelium remains a formidable
target for gene transfer vectors. Not only is the proliferative potential of the well differentiated surface epithelium
low,9 but also the apical membrane lacks entry mechanisms for a variety of gene transfer vectors including
Correspondence: LG Johnson, Cystic Fibrosis/Pulmonary Research and
Treatment Center, CB#7248, 7125 Thurston Bowles Building, The University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-7248,
USA
3
Current address: Laboratory for Gene Transfer and Therapy, IRCC, Institute for Cancer Research, University of Torino Medical School, SP 142
10060 Candiolo (Torino), Italy
Received 29 July 1999; accepted 26 November 1999
adenoviruses,10–12
liposomes,13,14
adeno-associated
viruses,15 and retroviruses.16,17 HIV vectors have not been
extensively studied in airways. In the only published
study to date, HIV vectors pseudotyped with the vesicular stomatitis virus (VSV) glycoprotein (G) efficiently
transduced undifferentiated airway epithelia, while failing to transduce the well differentiated pseudostratified
columnar epithelium of mature human bronchial xenografts.18
Given the potential advantages of HIV vectors, we
explored the utility of this vector system in airway epithelia. After establishing whether HIV vectors could
transduce non-dividing airway cells, we examined the
ability of these vectors to transduce polarized epithelia
in vitro via application of vector to the apical versus the
basolateral surfaces. Subsequently, we tested the ability
of HIV vectors to transduce the airways of rodents in vivo
using injury models to permit basolateral access of vector
to target cells.
Results and discussion
Transduction of non-dividing cells
As a first step, we investigated whether VSV-G pseudotyped HIV-based lentiviral vectors (Figure 1) could transduce non-dividing airway cells. Dividing and growtharrested (aphidicolin treated) CFT1 cells were infected
with an amphotropic enveloped vector [MuLV
(Ampho)], VSV-G pseudotyped MuLV vector [MuLV
(VSV-G)], or VSV-G pseudotyped HIV vector [HIV (VSVG)]. The MuLV (VSV-G) vector transduced 88% of cells
and the HIV (VSV-G) vectors transduced 97% of cells,
Airway lentiviral gene transfer
LG Johnson et al
Figure 1 Lentiviral vectors derived from human immunodeficiency virus
(HIV vectors). Vectors were produced by transient transfection into 293T
cells of vector, helper, and env plasmids. The plasmid backbones are not
shown. LacZ, E. coli gene coding for ␤-galactosidase; LTR, the U3, R and
U5 domains of the human immunodeficiency long terminal repeat (LTR)
region; SD, splice donor; SA, splice acceptor site; ⌿, packaging signal;
Ga, truncated gag gene which is also out of frame (X); RRE, Rev response
element; muPGK, murine phosphoglycerate kinase promoter; CMV,
human cytomegalovirus promoter.
whereas the lower titer MuLV (Ampho) vector transduced only 34% of cells (Figure 2). Only rare blue cells
were detected in aphidicolin-treated cells infected with
MuLV Ampho (0.2%) and MuLV (VSV-G) vectors (0.6%),
whereas 95% of aphidicolin-treated cells were transduced
with the HIV (VSV-G) lentiviral vector. Thus, all three
vectors efficiently transduced dividing cells (no
aphidicolin), whereas only the HIV-based lentiviral vector efficiently transduced growth-arrested cells (Figure 2).
Transduction of polarized well differentiated airway
epithelial cells
The localization of receptors for vector binding and
uptake is an important consideration for vectors delivered via the airway lumen. Previous studies have demonstrated a distinct polarity of viral receptor expression on
the apical versus the basolateral membranes of polarized
epithelia: (1) coxsackie-adenoviral receptors (CAR) are
expressed on basolateral domains to a much greater
extent than on apical domains of MDCK and human
airway epithelial cells;10–12 (2) amphotropic receptors
are similarly expressed predominantly on apical
Figure 2 HIV-based lentiviral transduction of non-dividing cells. Cells
were plated in the presence of either 0 or 15 ␮g/ml aphidicolin. The day
after plating, cells were infected at MOI 20, except for MuLV Ampho
(MOI 5) due to its lower titer. Cells were stained with X-gal 48 h after
infection. All vectors encode the E. coli ␤-galactosidase (lacZ) cDNA. HIV
(VSV-G), human lentiviral vector pseudotyped with vesicular stomatitis
virus glycoprotein; MuLV (VSV-G), murine leukemia virus derived retroviral vector pseudotyped with VSV-G; MuLV Ampho, amphotropic enveloped vector derived from murine leukemia virus.
domains;16,17 and (3) studies of wild-type vesicular
stomatitis virus infection of MDCK cells indicate that the
receptor for this virus is functionally located on the basolateral domain.19 To test the relevance of these observations to HIV vectors, we investigated the transduction
efficiency following apical versus basolateral application
of an HIV (VSV-G) vector (Figure 1) to polarized, well
differentiated airway epithelial cells. An important control for these experiments was the use of the transepithelial resistance as a measure of tight junctional integrity, since tight junctions are responsible for the physical
separation of the apical and basolateral domains. As
shown in Figure 3a, the transepithelial resistances of
these preparations were relatively high. Vector administration to the apical surface produced minimal levels of
transgene expression, whereas basolateral application of
vector to well differentiated human airway cells yielded
levels of transgene expression that were significantly (30fold) greater than levels achieved with apical application
of vector (basolateral 20.8 ± 3.6 mU/mg protein; apical
0.69 ± 0.20 mU/mg protein) as determined by ␤-galactosidase enzyme assay (Figure 3b; n = 3). In contrast, application of an MuLV (VSV-G) lacZ vector to either the apical or basolateral surfaces of well differentiated airway
epithelial cells failed to mediate transgene expression
(data not shown), consistent with low rates of airway epithelial cell proliferation (1.9%) in well differentiated cultures. These studies imply that uptake mechanisms for
VSV-G pseudotyped HIV vectors into quiescent, well differentiated polarized airway cells are predominantly
localized to the basolateral membrane.
569
Persistence in vitro
One attractive feature of retroviral and lentiviral vectors
is the potential for long-term transgene expression. Longterm transgene expression mediated by amphotropic and
VSV-G pseudotyped MuLV vectors has been reported in
undifferentiated airway cell lines with continuous passage on plastic.20,21 CFT1 cells, a human CF tracheal epithelial cell line, retrovirally transduced with wild-type
CFTR exhibited partial loss of functional expression
when cells in continuous passage were studied func-
Figure 3 HIV-based lentiviral transduction of polarized cells. (a) Transepithelial resistances of well differentiated human primary airway epithelial cells at the time of infection with HIV vector applied either to the
apical or basolateral surface. (b) Transduction of the same well differentiated primary human airway cells 72 h after infection with vector. Transduction (transgene expression) is expressed as milliunits (mU) ␤-gal per
milligram cellular total protein. ␤-Galactosidase activity was measured
using a chemiluminescence assay (Galacto-Star; Tropix, Bedford, MA,
USA). *Indicates significantly different from apical values by t test.
Although a significant difference is reported for the data in panel a, the
power or sensitivity of the performed t test at 0.576 is less than the desired
power of 0.8.
Gene Therapy
Airway lentiviral gene transfer
LG Johnson et al
570
tionally as polarized preparations at 6 months.22 Longterm expression of lentivirally transduced genes in airway cells has not been extensively studied. Primary airway cells grown on plastic possess many of the features
of pleuripotential cells in the airway including an undifferentiated phenotype, polyclonality, and the ability to
mature into a complex epithelium composed of multiple
differentiated epithelial cell phenotypes. Because injury
techniques typically increase vector access to both the
basolateral membrane of lumen-facing columnar cells
and the previously unexposed surfaces of airway basal
cells,21 we examined the ability of the lacZ transgene to
persist when undifferentiated cells were permitted to
develop into differentiated epithelial cell phenotypes.
Undifferentiated primary human airway cells grown
on plastic were transduced with VSV-G pseudotyped
HIV lacZ vectors containing either human cytomegalovirus (CMV) or murine phosphoglycerate kinase
(muPGK) internal promoters on culture day 1. Four days
later, an aliquot of cells was passaged on to TranswellCol inserts and permitted to grow into a well differentiated ciliated columnar epithelium over 30 days while
the remaining aliquot was stained in suspension with Xgal to assess initial levels of transduction. The percentages of cells expressing the transgene 4 days after transduction (day 5) and 30 days after transfer to TranswellCol inserts (day 35) are shown in Table 1. Primary human
airway cells transduced with an HIV (VSV-G) vector with
an internal murine PGK promoter exhibited a reduction
in the percentage of primary airway cells expressing lacZ
(24% of day 5 levels, n = 3 transwells) at 35 days. Persistence of lacZ transgene expression mediated by this vector
was less than persistence of transgene expression in airway cells mediated by an HIV vector with an internal
CMV promoter (55% of day 5 levels). In comparison, persistence of transgene expression was well maintained in
cells transduced with an MuLV (VSV-G) lacZ vector (LTR
promoter; 80% of day 5 levels). Histologic sections of
these cultures demonstrated that the lacZ transgene was
Table 1 Persistence of transgene expression in well differentiated cultures
Vector/Promoter
MuLV (VSV-G)
LTR promoter
HIV (VSV-G)
CMV promoter
HIV (VSV-G)
muPGK promoter
LacZ positive
cells on
day 5 (%)
LacZ positive
cells on
day 35 (%)
% Day 5
levels
60
48
80
69
38
55
14
3
24
Primary airway cells grown on plastic were infected with vector on
culture day 1, then disaggregated and split into two aliquots on
day 5. One aliquot was stained in suspension with X-gal and the
other plated on to two to three Transwell-Col inserts to permit the
cells to develop into a polarized, well differentiated epithelium over
30 days (day 35). The percentage of cells expressing the lacZ transgene at 35 days was determined by analysis of images captured
from histologic sections.
MuLV, murine leukemia virus; VSV-G; vesicular stomatitis virus
glycoprotein pseudotype; LTR, promoter elements of the MuLV
long terminal repeat; CMV, human cytomegalovirus; muPGK,
murine phosphoglycerate kinase.
Gene Therapy
expressed in basal and columnar cells after the 30 days
of growth on permeable substrates (data not shown).
Since similar transduction protocols, cell growth parameters, and retroviral envelope proteins were used, it
seems unlikely that differences in cell toxicity and growth
rates were responsible for the loss of transgene
expression. Rather, these data suggest that cis-acting
elements within the vector, ie promoters or other components of the vector backbone, contribute to the duration
of in vitro persistence.
In vivo infection of rodent nasal epithelium
No studies of direct in vivo airway HIV lentiviral gene
transfer have been published to date. We examined the
ability of luminal application of lentiviral vectors pseudotyped with VSV-G to transduce nasal airway epithelium
in vivo in the absence of injury. HIV (VSV-G) vectors
failed to transduce naive (unmodified or injury-free) well
differentiated airway epithelia in vivo following luminal
application of vector consistent with the findings in polarized cultures (data not shown). Based on previous
experience,21 we evaluated the effect of oxidant gas (SO2)induced modification of host airway on enhancement of
lentiviral gene transfer. Instillation of HIV (VSV-G) vector into the right nostril of SO2-exposed 3-week-old mice
(n = 6) and rats (n = 4) revealed patchy LacZ expression
(Figure 4) that was predominantly localized to the respiratory and transitional epithelium along the septum and
turbinates (Figure 4a–c). In general, expression was confined to the nostril to which vector was administered or
predominantly to the nostril receiving vector. Occasional
staining was also detected in the olfactory epithelium.
Expression was significantly greater in rats (66.5 ± 12.4
cells per perimeter) than in mice (6.6 ± 1.6 cells per perimeter, Figure 3d). No staining was detected in the nasal
epithelium of vehicle control rats (n = 3) or mice (n = 6). It
should be noted that both the vehicle control and vectortreated SO2-exposed rats had significant goblet cell
hyperplasia.
SO2 inhalational exposure has been shown to increase
tracheal retroviral gene transfer by direct injury manifested as denuding of surface epithelial cells and by
increased paracellular permeability of the airway epithelium in the less severely injured regions without frank
morphologic damage.21 Similar mechanisms are likely
operative in the nasal airways since SO2-induced effects
tend to be most pronounced in the more proximal airways.24
Intratracheal administration
The double tracheostomy technique was performed for
vector delivery to defined regions of the tracheas of anesthetized mice.11,21 Similar to the findings in the nasal epithelium, HIV vectors failed to transduce murine tracheal
epithelia in the absence of injury (data not shown). Mechanical injury, performed by gently pinching the trachea
between the cartilaginous rings with a small pair of forceps, followed by administration of an HIV (VSV-G) lacZ
vector resulted in patchy, inefficient gene transfer to the
tracheas of mice infected with vector, but not vehicle
control tracheas. This result was confirmed in histologic
sections (Figure 5).
The SO2 inhalational injury model was subsequently
employed to achieve more uniform tracheal injury and
to explore further the feasibility of in vivo HIV vector-
Airway lentiviral gene transfer
LG Johnson et al
a
b
c
d
Figure 4 HIV-based lentiviral transduction of rodent proximal nasal airway after SO2 inhalation. (a) Schematic of a coronal section demonstrating
the anatomy of the proximal nasal airway. The nasal epithelium is outlined
in bold black. Photomicrographs of X-gal stained histologic sections showing the septum of vehicle control (b) and vector-transduced rat nasal airways (c). Magnification (× 50). (d) Quantification of LacZ expression in
mouse and rat proximal nasal airway. Data are expressed as LacZ(+) cells
per perimeter transduced with vector. The epithelial circumference in one
nostril (outlined in bold black in panel a) is defined as a perimeter.23 Vector or vehicle was instilled intranasally 1–2 h after SO2 inhalation. X-gal
staining was performed 96 h following intranasal administration of vector
or vehicle. *Significantly different from vehicle control by t test or Mann–
Whitney rank sum test. †Significantly different from mouse by Mann–
Whitney rank sum test.
mediated gene transfer. In this model system, tracheal
airway cell proliferation is optimal 24 h after SO2 inhalation, whereas no significant airway cell proliferation
occurs in the first 12 h after SO2 inhalation as detected by
BrdU incorporation.21 Vector was instilled into the SO2exposed tracheas of anesthetized 3-week-old mice on the
same day of oxidant gas exposure (day 0) versus 24 h
later (day 1, Figure 6a). Quantification of gene transfer
efficiency (Figure 6b) demonstrated that transduction of
lacZ was more efficient when vector was instilled on the
day of injury when cells were not proliferating (7.0 ± 1.7%
cells transduced, n = 4) as compared with vector instillation on the day after injury when cell proliferation
peaks (1.7 ± 0.7% cells transduced, n = 4), or vehicle controls (0.1 ± 0.07%, n = 3). The efficiency of transduction
on the day of injury when cells were not proliferating
also falls within the 6–10% range predicted to correct the
CFTR-mediated chloride transport defect in cystic
fibrosis.25 Significant transduction of rat tracheas was also
feasible when vector (3.5 ± 1.3%, n = 3) was delivered on
the day of SO2 injury as compared with vehicle controls
(0.04 ± 0.02%).
These data differ from our experience with MuLV vectors21 where the maximal levels of transduction (6%)
were achieved when vector was delivered the day after
injury, consistent with the peak of airway proliferation
(24 h after injury). Presumably, increased paracellular
permeability and access to basolateral cell surfaces and
basal cells permissive for vector entry21 are the mechanisms responsible for the enhancement of HIV (VSV-G)
vector-mediated transduction of airway cells detected
after SO2 injury, whereas increased cell proliferation, perhaps complimented by increased access, is the dominant
mechanism responsible for enhancing MuLV (VSV-G)mediated gene transfer. The greater efficiency of transduction mediated by HIV VSV-G vectors in the first few
hours after SO2 inhalation, when rates of airway cell proliferation are minimal, suggests that HIV vectors can
transduce non-dividing airway cells in vivo.
The efficiency of transduction achieved with HIV vectors was remarkably similar to, rather than better than,
the efficiency of transduction reported for pseudotyped
MuLV vectors.21 Since both the MuLV and HIV vectors
use the vesicular stomatitis virus glycoprotein as the
ligand for vector entry, similar transduction efficiencies
might be the expected result if entry is limiting for gene
transfer. Alternatively, the use of the muPGK promoter
in the HIV vector, which is generally considered to be a
weak promoter, might have led to an underestimate of
HIV vector-mediated gene transfer efficiency since
expression levels might be lower with this promoter. We
attempted to use an HIV vector containing the stronger
CMV promoter, however, little or no gene transfer was
detected in naive (injury-free) or mechanically injured
murine tracheas infected with the CMV promoter construct (data not shown). Therefore, we focused all our
efforts on the use of the muPGK promoter in vivo.
We have not examined persistence of HIV vectormediated transgene expression in vivo beyond 5–6 days
since our initial goal was first to optimize efficiency.
Future studies will be necessary to fully evaluate the
potential for long-term expression from lentiviral vectors.
These studies will have to take into account that reporter
transgenes can induce an immune response limiting the
duration of transgene expression. Thus, it may be necessary to use species-specific transgenes or immunodeficient animals to assess in vivo persistence.
In summary, human lentiviral vectors pseudotyped
with VSV-G preferentially transduced polarized airway
epithelia from the basolateral surface similar to what has
been previously reported for other gene transfer vectors.10–17 As a result, a method (SO2 inhalation) that
increased access of vector delivered via the lumen to the
basal cells and the basolateral membrane was required
for successful gene transfer. If clinically effective and safe
methods can be developed to increase access of vector
to target cells, this approach may offer hope for efficient
lentivirus-mediated gene transfer.
571
Gene Therapy
Airway lentiviral gene transfer
LG Johnson et al
572
a
b
c
d
e
f
Figure 5 HIV-based lentiviral transduction of murine trachea after mechanical injury. Vector was instilled and confined to the mechanically injured
tracheas of anesthetized mice for 2 h at an estimated MOI of 30. Control tracheas are depicted in a and d, whereas lentiviral infected tracheas are
depicted in b, c, e, and f. Histologic sections were counterstained with nuclear fast red. Magnification (× 50).
Materials and methods
a
b
Figure 6 HIV-based lentiviral transduction of murine trachea after SO2
inhalation. Vector or vehicle was instilled into the tracheas of anesthetized
mice on the same day (day 0, n = 4) or 24 h (day 1, n = 4) after SO2
inhalational exposure. Tracheas were harvested and stained with X-gal 96
h after infection. (a) En face views of intact murine tracheas (upper row of
panels) and photomicrographs of histologic sections (lower row of panels,
magnification × 50) from control and test animals. (b) Quantitative analysis of gene transfer efficiency. Data were analyzed using a one way
ANOVA with a correction for multiple comparisons (Student Neuman
Keuls method). *Significantly different from control. †Significantly different from day 1.
Gene Therapy
Cell culture
Well differentiated cell cultures of primary normal (nonCF) human bronchial or nasal airway cells were prepared
as previously described.14,21,26 Cells were isolated from
surgical specimens by enzymatic digestion, plated on collagen-coated 10-cm plastic dishes at a density of 2 × 106
cells per dish, and maintained in hormone-supplemented
modified LHC9 medium.26 On culture day 5–6, cells were
harvested by trypsin digestion and plated at a density of
2 × 105 cells per 12 mm (0.4 ␮m pore size) Transwell-Col
insert (Costar, Cambridge, MA, USA) in a 50:50 mixture
of LHC Basal (Biofluids, Rockville, MD, USA) and
DMEM-H medium supplemented with growth factors,
retinoic acid, and bovine serum albumin as previously
described.14,21 When cultures attained confluence,
medium was gently aspirated from the apical surface creating an air–liquid interface. Cells were maintained in
culture for 2–3 weeks and used for experiments only after
they had developed a well differentiated phenotype
characterized by the presence of cilia on greater than 10%
of the cells upon visualization by phase contrast
microscopy.
X-gal histochemistry
For in vivo studies, animals were killed by CO2 asphyxia.
Excised tracheas were stained in X-gal solution for 6 h
and post-fixed in Omnifix II (Ancon Genetics, St Petersburg, FL, USA).21 To minimize background staining, the
pH of all solutions was adjusted to 8.0 with Tris buffer
(20 mm final concentration). The unique architecture of
the nasal epithelium required a modification of the X-gal
staining technique23 as follows: animal carcasses were
decapitated and the nasal cavity immediately fixed by
retrograde perfusion of 2% paraformaldehyde/0.2% glutaraldehyde into the nasal cavity via the tracheal remnant.
Following removal of soft tissue and the brain by sharp
dissection, heads were immersed in fixative for 2 h at
4°C punctuated by intermittent retrograde flushing of the
nasal cavity via the tracheal remnant. The nasal cavity
Airway lentiviral gene transfer
LG Johnson et al
was subsequently stained with X-gal by immersion combined with intermittent retrograde flushing for 6 h at
37°C. Heads were post-fixed in 10% neutral buffered formalin for 24 h, rinsed in tap water, and immersed in 70%
ethanol until decalcification (Formical; Decal, Congress,
NY, USA) was performed. Standard histological sections
were cut and counterstained with nuclear fast red.
␤-Galactosidase enzyme assay
Expression of the E. coli ␤-galactosidase (lacZ) transgene
in well differentiated primary airway epithelial cells was
measured using a chemiluminescent reporter assay kit
(Galacto-star, Tropix, Bedford, MA, USA). Samples were
prepared by excising the filter on which the cells are
attached from the plastic holder using a #11 blade scalpel
and immersing the filter containing the cells in 100 ␮l
lysis buffer. The filter in lysis buffer was freeze–thawed
six times in a 95% ethanol-dry ice bath and the remaining
solution pelleted at 3000 g in a microfuge. Supernatant
(2–20 ␮l) was mixed with 300 ␮l of reaction buffer containing the Galacto-Star substrate supplied by the manufacturer and incubated for 60–90 min. This reaction mixture was placed in a luminometer, measured for 5 s, and
the relative light units per sample recorded. A standard
curve was generated with serial dilutions of E. coli ␤galactosidase enzyme (Sigma Chemical, St Louis, MO,
USA). The amount of cellular total protein in the supernatant generated from each sample was measured using
the bicinchoninic acid (BCA) method (Pierce, Rockford,
IL, USA). Data are expressed as milliunits (mU or units
× 10−3) of active ␤-galactosidase enzyme. Typically, each
milligram of E. coli ␤-galactosidase contains 250–500 units
of active enzyme.
Quantitative analysis of gene transfer efficiency
To quantify the percentage of cells transduced in cultured
primary cells on plastic, cells were disaggregated, fixed
in 0.5% glutaraldehyde, and stained with X-gal in suspension. The percentage of cells transduced was calculated from hemacytometer counts of blue cells relative to
total cells in the sample.
For quantification of gene transfer efficiency in well
differentiated cultures, 10–15 images of randomly selected fields of histologic sections from each polarized culture were captured using a light microscope interfaced
to a cooled charge coupled device (CCD) camera
(Hamamatsu C5985, Hamamatsu, Japan). The percentage
of cells transduced from histologic sections of well differentiated cultures or from intact rodent tracheas was calculated by measuring the area of cells transduced relative
to the total area populated by cells dosed with vector
using the Metamorph image analysis system
(Metamorph; Universal Imaging, West Chester, PA,
USA).14,21
To determine the percentage of vector-transduced cells
in rodent tracheas, en face images of tracheas were captured with a CCD camera interfaced to a dissecting
microscope. The percentage of cells transduced was calculated by measuring the area of cells transduced relative
to the total area occupied by cells dosed with vector
using the Metamorph image analysis system. Once
images had been captured, intact tracheas were embedded in paraffin and five longitudinal sections (8 ␮m thick)
at 50–100 ␮m levels were taken, mounted on slides, then
counterstained with nuclear fast red. Histologic sections
from intact rodent tracheas were used to confirm localization of staining to airway epithelial cells, but were not
used to quantify transduction in sections due to the limitations of sampling error.
To quantify in vivo gene transfer efficiency in rodent
nasal epithelia, ⭓3 coronal histologic sections were taken
through the rodent head just caudad to the incisors.23 The
total number of LacZ-positive cells within the nasal epithelial circumference of a nostril, ie a perimeter (see Figure 4), was counted in each section by light microscopy.
573
Vector production
The VSV-G pseudotyped human lentiviral vector (HRmuPGKlacZ) used in this study is depicted in Figure 1.
An internal promoter, the murine phosphoglycerate kinase (muPGK) promoter was used to drive transcription
of the E. coli lacZ cDNA. This HIV (VSV-G) lacZ vector
was produced with the minimal packaging construct
which contains deletions of env, vif, vpu, nef, and vpr.5
Lentiviral vector preparations were purified and concentrated from the conditioned media of transfected cells by
ultracentrifugation as previously described.1,2 HIV (VSVG) vector stocks were titered in rat 208F fibroblasts and
generally exceeded 1 × 109 infectious units (IU)/ml.
The VSV-G pseudotyped murine leukemia virusderived lacZ vector HIT-LZ used in this study was produced by transient transfection as previously described.21
The amphotropic enveloped MuLV-derived lacZ vector
LNPOZ was produced from a clonal PA317 producer
cell line.27
Gene transfer
To assess polarity of infection, well differentiated primary airway cells with transepithelial resistances (Rt)
approximately 1000 Ohm-cm2 (measured using an epithelial ohmmeter, EVOM, World Precision Instruments,
Sarasota, FL, USA) were infected with HIV vector
applied to either the apical or basolateral surfaces (MOI
30) for 2 h at 37°C. For basolateral infections, transwells
were inverted with direct application of vector (100 ␮l)
to the basolateral surface of the transwell during the
period of vector incubation with cells.
In vivo intranasal instillation: Vector (20 ␮l) was administered drop by drop to the right nostril of anesthetized
mice or rats (age 3–4 weeks) using gel loading pipette
tips. The nasal epithelia of mice receiving vector or
vehicle were either injury-free or injured by oxidant gas
exposure 1–2 h immediately before vector delivery. All
SO2 treatments were whole body animal exposures in
stainless steel inhalation chambers for 3 h at 500 ppm as
previously described.21
Intratracheal instillation: Vector was delivered to the tracheas of anesthetized mice (age 3–4 weeks) or rats (age
3–4 weeks) using the double tracheostomy technique.11,21
Tracheas of anesthetized mice were surgically exposed
and a distal tracheotomy performed near the carina for
ventilation (breathing) and a proximal tracheotomy performed at the first cartilaginous tracheal ring for vector
instillation.11,21 The region between the two tracheostomies was filled with vector for a defined period of time
until subsequent removal by suctioning. The instilled
vector was limited to the region between the two tracheostomies by respiration through the distal tracheostomy,
Gene Therapy
Airway lentiviral gene transfer
LG Johnson et al
574
which was confirmed by visualization through a dissecting microscope. All intratracheal instillations of HIV
vector or vehicle (buffer) were performed as two sequential 10–15 ␮l aliquots for 1 h each (2 h total duration). This
time period was chosen as optimal because time periods
longer than this resulted in an unacceptably high mortality. The animals were killed 4–5 days after infection,
tracheas removed, and stained with X-gal for histochemical analysis.
Statistical analysis
Where possible, data are presented as the mean ± s.e. A
one way ANOVA with a correction for multiple comparisons (Student Neuman Keuls method), t tests, and Mann–
Whitney rank sum tests were used to determine statistical
significance (P ⬍ 0.05).
Acknowledgements
We thank Dr Daniel Costa of the US Environmental Protection Agency, Pulmonary Toxicology Branch, and his
assistant, Mr Todd Krantz, for their advice on the SO2
model and for performing the SO2 exposures. We are
grateful to the CF/Pulmonary and Research Center
Tissue Culture Core (Drs Scott Randell and James Yankaskas, Co-directors) for providing the primary human
airway cells. We thank Ms Jennifer Mewshaw, Ms Sarah
Mosier, Ms Hong Ni and Ms Miriam Kelly for technical
assistance and Ms Elizabeth Godwin for clerical assistance. This work was supported by HL58342 (LGJ).
References
1 Naldini L et al. In vivo gene delivery and stable transduction
of non-dividing cells by a lentiviral vector. Science 1996; 272:
263–267.
2 Naldini L et al. Efficient transfer, integration, and sustained
long-term expression of the transgene in adult rat brains injected
with a lentiviral vector. Proc Natl Acad Sci USA 1996; 93:
11383–11388.
3 Olsen JC. Gene transfer vectors derived from equine infectious
anemia virus. Gene Therapy 1998; 5: 1481–1487.
4 Ju Q et al. Transduction of non-dividing adult human pancreatic
beta cells by an integrating lentiviral vector. Diabetologia 1998;
41: 736–739
5 Zufferey R et al. Multiply attenuated lentiviral vector achieves
efficient gene delivery in vivo. Nature Biotech 1997; 15: 871–875.
6 Zufferey R et al. Self-inactivating lentivirus vector for safe and
efficient in vivo gene delivery. J Virol 1998; 72: 9873–9880.
7 Burns JC et al. Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titer and
efficient gene transfer into mammalian and nonmammalian
cells. Proc Natl Acad Sci USA 1993; 90: 8033–8037.
8 Yee J-K et al. A general method for the generation of high-titer,
pantropic retroviral vectors: highly efficient infection of primary
hepatocytes. Proc Natl Acad Sci USA 1994; 91: 9564–9568.
Gene Therapy
9 Leigh MW, Kylander JE, Yankaskas JR, Boucher RC. Cell proliferation in bronchial epithelium and submucosal glands of cystic fibrosis patients. Am J Resp Cell Mol Biol 1995; 12: 605–612.
10 Grubb BR et al. Inefficient gene transfer by adenovirus vector
to cystic fibrosis airway epithelia of mice and humans. Nature
1994; 371: 802–806.
11 Pickles RJ, Barker PM, Ye H, Boucher RC. Efficient adenovirusmediated gene transfer to basal but not columnar cells of cartilaginous airway epithelia. Hum Gene Ther 1996; 7: 921–931.
12 Pickles RJ et al. Limited entry of adenovirus vectors into well
differentiated airway epithelium is responsible for inefficient
gene transfer. J Virol 1998; 72: 6014–6023.
13 Chu Q et al. Binding and uptake of cation lipid:pDNA complexes by polarized airway cells. Hum Gene Ther 1999; 10: 25–36.
14 Matsui H, Johnson LG, Randell SH, Boucher RC. Loss of binding
and entry of liposome–DNA complexes decreases transfection
efficiency in differentiated airway epithelial cells. J Biol Chem
1997; 272: 1117–1126.
15 Duan DS et al. Polarity influences the efficiency of recombinant
adenoassociated virus infection in differentiated airway epithelia. Hum Gene Ther 1998; 9: 2761–2776.
16 Wang G et al. Influence of cell polarity on retrovirus-mediated
gene transfer to differentiated human airway epithelia. J Virol
1998; 72: 9818–9826.
17 Zsengeller ZK et al. Keratinocyte growth factor stimulates transduction of the respiratory epithelium by retroviral vectors. Hum
Gene Ther 1999; 10: 341–353.
18 Goldman MJ, Lee P-S, Yank J-S, Wilson JM. Lentiviral vectors
for gene therapy of cystic fibrosis. Hum Gene Ther 1997; 8:
2261–2268.
19 Fuller S, von Bonsdorff C-H, Simons K. Vesicular stomatitis
virus infects and matures only through the basolateral surface
of the polarized epithelial cell line, MDCK. Cell 1984; 38: 65–77.
20 Olsen JC et al. Retrovirus-mediated gene transfer to cystic
fibrosis airway epithelial cells: effect of selectable marker
sequences on long-term expression. Nucleic Acids Res 1993; 21:
663–669.
21 Johnson LG et al. Effect of host modification and age on airway
epithelial gene transfer mediated by a murine leukemia virusderived vector. J Virol 1998; 72: 8861–8872.
22 Olsen JC et al. Correction of the apical membrane chloride permeability defect in polarized cystic fibrosis airway epithelia following retroviral-mediated gene transfer. Hum Gene Ther 1992;
3: 253–266.
23 Parsons DW, Grubb B, Johnson LG, Boucher RC. Enhancement
of in vivo adenoviral gene transfer via modification of host barrier properties with a surface active agent. Hum Gene Ther 1998;
9: 2661–2672.
24 Hulbert WC et al. The response phase – the first six hours after
airway injury by SO2 inhalation: an in vivo and in vitro study.
Scanning Microsc 1989; 3: 369–378.
25 Johnson LG et al. Efficiency of gene transfer for restoration of
normal airway epithelial function in cystic fibrosis. Nat Genet
1992; 2: 21–25.
26 Gray TE et al. Mucociliary differentiation of serially passaged
normal human tracheobronchial epithelial cells. Am J Respir Cell
Mol Biol 1996; 141: 104–122.
27 Olsen JC, Sechelski J. Use of sodium butyrate to enhance production of retroviral vectors expressing CFTR cDNA. Hum Gene
Ther 1995; 6: 1195–1202.