Physiol Genomics
2: 117–127, 2000.
Adeno-associated virus vector transduction of vascular
smooth muscle cells in vivo
MATTHIAS RICHTER,1 AKIKO IWATA,2 JOHN NYHUIS,2 YOSHIO NITTA,2
A. DUSTY MILLER,3 CHRISTINE L. HALBERT,3 AND MARGARET D. ALLEN2
2Division of Cardiothoracic Surgery, Department of Surgery, University of Washington;
the 3Fred Hutchinson Cancer Research Center, Seattle, Washington;
and the 1Humboldt Universität zu Berlin, Berlin, Germany
viral vectors; gene therapy; vascular biology
GENE THERAPY COULD PROVIDE new treatment options for
genetic and acquired diseases but is currently limited
by vector technology. An ideal vector would have high
transduction efficiency, transfer genes into quiescent as
well as dividing cells, and provide long-term gene
expression. Additionally, the ideal vector should not
create inflammatory or immune responses nor direct
cytopathic effects, and it should not be pathogenic for
humans.
Adeno-associated virus (AAV) vectors provide relatively high transduction efficiency and are able to
transduce nondividing as well as dividing cells (1, 10,
Received 21 May 1999; accepted in final form 20 February 2000.
Article published online before print. See web site for date of
publication (http://physiolgenomics.physiology.org).
15, 16, 18–21, 36, 37, 40). The structure of their
single-stranded DNA can be reduced to two terminal
repeats, providing for packaging and integration (34),
while leaving most of the 4.5-kb DNA to accommodate a
promoter to enhance gene expression and a reporter or
functional gene. Finally, AAV appears to be nonpathogenic for humans.
Given these advantages, AAV vectors would seem to
be suitable for vascular transduction, given the quiescent state and slow turnover of vascular cells and the
desirability of long-term gene expression for the treatment of such vascular diseases as atherosclerosis,
posttransplant arteriopathy, and fibromuscular hyperplasia. Since the objective in treating arterial pathology
is to provide local delivery of gene product, our aim in
this study is to maximize the prevalence of gene
expression in the vascular wall rather than to produce
systemic levels of secreted gene products.
AAV vectors have been previously studied in airway,
heart, liver, brain, and skeletal and cardiac muscle (1,
10, 15, 16, 18–21, 36, 37, 40), but fewer investigations
have examined the potential for in vivo AAV vector
transduction of the vascular wall (3, 13, 23, 24, 30).
Controversy exists as to whether intraluminal delivery
results in transduction of just endothelial cells in the
vaso vasorum (23) or whether these small vectors can
traverse the internal elastic lamina (IEL) and transduce medial smooth muscle cells (3, 13, 30). Compared
with previous studies, these experiments utilize the
high viral titers now available, a different promoter,
varying intraluminal pressures, a surgical technique,
and quantitative assessments of the prevalence of gene
expression over the entire arterial luminal surface on
whole vessel preparations as well as on cross sections.
In addition, host immune reaction was specifically
sought by immunocytochemical identification of infiltrating leukocytes. The result of optimizing several
variables in this model is that we have arrived at a
method of achieving reporter gene expression in approximately one third or more of the subintimal vascular
smooth muscle cells (VSMCs) in a rabbit carotid artery
model with an intact endothelium. These techniques
would be suitable for clinical applications in either
intraoperative intra-arterial delivery or catheter-based
percutaneous interventions for a number of vascular
pathologies where alteration of smooth muscle cell
targets would be advantageous.
1094-8341/00 $5.00 Copyright r 2000 the American Physiological Society
117
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
Richter, Matthias, Akiko Iwata, John Nyhuis, Yoshio
Nitta, A. Dusty Miller, Christine L. Halbert, and Margaret D. Allen. Adeno-associated virus vector transduction of
vascular smooth muscle cells in vivo. Physiol Genomics 2:
117–127, 2000.—Adeno-associated virus (AAV) vectors might
offer solutions for restenosis and angiogenesis by transducing
nondividing cells and providing long-term gene expression.
We investigated the feasibility of vascular cell transduction
by AAV vectors in an in vivo rabbit carotid artery model. Time
course of gene expression, inflammatory reaction to the
vector, and effects of varying viral titer, exposure time, and
intraluminal pressures on gene expression were examined.
Recombinant AAV vectors with an Rous sarcoma virus promoter and alkaline phosphatase reporter gene were injected
intraluminally into transiently isolated carotid segments.
Following transduction, gene expression increased significantly over 14 days and then remained stable to 28 days, the
last time point examined. Medial vascular smooth muscle
cells were the main cell type transduced even with an intact
endothelial layer. Increasing the viral titer and intraluminal
pressure both enhanced transduction efficiency to achieve a
mean of 34 ⫾ 7% of the subintimal layer of smooth muscle
cells expressing gene product. A mild inflammatory reaction,
composed of T cells with only rare macrophages, with minimal intimal thickening was demonstrated in 40% of transduced vessels; inflammatory cells were not detected in shamoperated control arteries. These findings demonstrate that
AAV is a promising vector for intravascular applications in
coronary and peripheral vascular diseases.
118
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
MATERIALS AND METHODS
Fig. 1. Operative technique for adeno-associated virus (AAV) vector
transduction.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
Animals. New Zealand White and inbred Stauffland
rabbits (R and R Rabbitry, Stanwood, WA), ranging in
weight from 2.2 to 3.6 kg, were used for all experiments.
All experiments were performed in accordance with the
‘‘Principles of Laboratory Animal Care’’ formulated by
the National Society for Medical Research, and the
‘‘Guide for the Care and Use of Laboratory Animals’’
[DHHS Publication No. (NIH) 85–23, Revised 1985,
Office of Science and Health Reports, Bethesda, MD
20892].
Operative technique for in vivo AAV vector transduction of rabbit carotid arteries. Rabbits were sedated
with a mixture of ketamine (35 mg/kg), xylazine (5
mg/kg), and atropine (0.032 mg/kg) given intramuscularly. After endotracheal intubation, animals were mechanically ventilated with halothane in 100% oxygen to
maintain general anesthesia. Each rabbit was heparinized systemically with 300 U/kg of heparin sodium
prior to carotid artery manipulation. The left carotid
artery was surgically exposed, and vascular spasm was
averted with topical lidocaine. A 3.0-cm segment of the
carotid was isolated between atraumatic vascular
clamps. A 25-gauge butterfly needle was introduced
into the proximal end of the isolated segment. The
distal clamp was opened briefly, and blood was flushed
out of the vessel by injecting Ringer solution through
the butterfly needle; then the distal clamp was again
closed. A second needle for later vector instillation was
then inserted into the distal end of the isolated segment
and fixed in place with a pursestring to ensure that no
leakage could occur. The Ringer solution was withdrawn up into the proximal needle to empty the vascular lumen, and a third clamp was placed just distal to
the ‘‘Ringer needle’’ tip so that the site of needle entry
would be excluded from the surface area exposed to the
vector. Into this isolated and now-empty vascular segment, the AAV vector was injected with a volume
between 35 and 80 µl, distending the artery to decrease
endothelial infolding and to enhance contact of the AAV
vector with all portions of the luminal endothelium
(Fig. 1). The length of the artery exposed to the vector
was between 5 and 17 mm.
For intraluminal pressure measurements, a pressure
line was connected to the proximal (Ringer) needle with
a three-way stopcock. The proximal (Ringer) and distal
(vector) needle insertion sites were surrounded with
pursestring sutures to provide a closed system in which
the intraluminal pressure could be measured. As before, Ringer solution was withdrawn into the proximal
needle after the lumen was flushed free of blood; then,
the proximal stopcock was opened to the pressuremonitoring line. Now the vessel was filled with vector
from the distal needle, and the intraluminal pressure
continuously recorded. Both injection sites were repaired with single adventitial sutures of 7–0 prolene
after the injection needles were withdrawn. The clamps
were removed after a 30- or 60-min exposure time, and
blood flow was reestablished. The cervical incision was
closed in layers, and the rabbits were returned to their
cages for recovery.
Sham operations were performed in eight animals,
using exactly the same operative technique, but injecting Ringer solution instead of AAV vector. These served
as controls for the degree of inflammatory reaction that
might be incurred by surgical manipulation alone,
unrelated to gene transduction and viral vector exposure. The contralateral carotid arteries in all rabbits
were also harvested and examined as normal, untreated control arteries.
In the interests of both vector and animal conservation, data from 19 of the 58 animals in various treatment groups were utilized in more than one experiment. The data from the 11 animals killed at 14 days in
the time course experiment were used as a basis of
comparison for the effects of higher viral titer. Data
from the eight animals transduced with higher viral
titers in the titer experiments were used as controls for
the effects of exposure time, but this time the titer was
kept constant.
AAV vectors. The AAV vector CWRAP (11) contains
the cDNA encoding human placental alkaline phosphatase (AP) driven by a Rous sarcoma virus (RSV) promoter and enhancer sequence (15). AAV vector stocks
were generated as previously described (16). Briefly,
cells were infected with adenovirus 5, and then vector
plasmid (4 µg) and the AAV packaging plasmid pMTrepCMVcap (2) were cotransfected by using the calcium
phosphate transfection method. Vector stocks were
purified by cesium chloride centrifugation and stored at
⫺80°C. Vector titers were determined by using HT1080 cells as targets for transduction and were equal to
5 ⫻ 108 or 2 ⫻ 109 AP-positive (AP⫹) focus-forming
units (FFU) per milliliter (ml). The ratio of genomecontaining vector particles to FFU for the CWRAP
vector under these production conditions ranges from
100 to 700, and for the vector stock with a titer of 5 ⫻
108 AP⫹ FFU/ml, the ratio was 100. Vector stocks were
tested for the presence of infectious adenovirus by
plaque assay (16), and none were detected [⬍100
plaque-forming units (PFU) per ml]. Vector stocks were
also assayed for contamination by replication-competent AAV by infectious center assays, and contained no
detectable replication-competent AAV (⬍100 infectious
units/ml).
Histochemical AP staining of carotid arteries. Animals were killed at 3, 7, 10, 14, and 28 days following
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
from one part of the vessel to another, we felt that the
best overview of the prevalence of gene product expression was provided by the en face preparations, with
data substantiated in selected experiments by analysis
of the cross sections.
Immunocytochemistry. To determine cell type specificity on the cross sections, antibody to ␣-actin (BoehringerMannheim, Germany) was used to identify smooth
muscle cells; antibody to CD-31, PECAM-1 (DAKO,
Carpinteria, CA), was used to identify endothelial cells.
Immunocytochemistry utilizing antibodies specifically
directed against rabbit T lymphocytes, MAb Tib 188
(American Type Culture Collection, Rockville, MD),
and rabbit macrophages, Ram-11 (DAKO), was used to
determine the cell-type specificity of leukocytes infiltrating the intima, media, and adventitial compartments of
the vascular wall. MAb Tib 188 is an L11/135 hybridoma that recognizes a cell surface determinant on all
rabbit thymocytes and peripheral blood T cells, but not
rabbit B cells (17). Splenocytes marked by this antibody
include all that respond to T cell mitogens, both resting
and activated, and allogenic stimuli (17). Ram-11,
developed against a cytoplasmic component of rabbit
alveolar macrophages, binds broadly to all tissue macrophages (39).
Briefly, histological cross sections of the arteries were
embedded in paraffin. These cross sections included en
face specimens that had previously been stained for AP
activity. The specimens were then deparaffinized and
hydrated through xylene and alcohol series. Endogenous peroxidase was blocked with 1.5% H2O2 in methanol for 10 min. Nonspecific binding was controlled for
by treating with a blocking solution for 30 min (1%
BSA/PBS, 1% horse serum, and 0.1% Triton X-100).
The tissue was then incubated with primary antibody
diluted in BSA/PBS at room temperature for an hour
(MAb Tib at 1:200, MAb Ram-11 at 1:400, MAb to
␣-actin at 1:1,000, and MAB to CD31 at 1:100 dilutions). After washing with PBS, sections were incubated with biotinylated horse anti-mouse IgG (Vector
Laboratories, Burlingame, CA) diluted in PBS (1:100)
for 30 min at room temperature, then washed again
and treated with the avidin-biotin peroxidase technique (ABC Elite Kit, Vector Laboratories) for 30 min.
The reaction product was developed using diaminobenzidine (DAB) substrate (Vector Laboratories) for 4 to 10
min, resulting in the deposition of a brown reaction
product, which visualizes bound antibodies. Nuclei
were counterstained with hematoxylin.
On the cross sections from 42 arteries (35 transduced
and 7 sham operated), the presence or absence of
intimal lesions and the extent of T lymphocyte and
macrophage infiltration were assessed semiquantitatively. By and large, these specimens represented those
in the high-titer and high-pressure groups, the ‘‘at risk’’
groups for lesion development, and their controls. In 23
specimens in which lesions were identified, including
both transduced and untransduced specimens, the
intimal lesions were further characterized by assessing
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
vector instillation. Rabbits were anesthetized (as described above) and systemically heparinized. Both the
transduced and the contralateral native carotid arteries were surgically exposed, isolated proximally and
distally, then ligated and excised. The rabbits were
then killed with concentrated pentobarbital given intravenously at 120 mg/kg.
Each carotid specimen was divided into three parts:
the central part was fixed in zinc fixative (6) or 4%
neutral buffered formalin (NBF)(pH 7.2) for histological cross sectioning. The remaining vessel ends were
opened longitudinally, pinned open to expose the endothelial surface, and fixed in NBF to produce two en face
specimens, or whole tissue mounts of the entire vessel
wall. En face tissue specimens were warmed up in a
microwave for 40 s to flatten the fixed specimen, then
washed in PBS and 0.1% Triton X-100 for 10 min. To
heat-inactivate endogenous AP, specimens were placed
in a water bath at 68.5°C for 1 h. To assess AP
enzymatic activity, specimens were then placed in AP
substrate solution (Zymed Laboratories, San Francisco,
CA) for 4 h at 37°C. The contralateral carotid artery
was similarly processed.
Quantification of the prevalence of gene product expression. By viewing the arterial luminal surface on the en
face preparations, the prevalence of gene expression
was quantified as the percentage of the arterial surface
expressing the transduced human placental AP enzyme, using the Optimas computerized image analysis
program (Optimas, Bothell, WA). After enzymatic visualization of alkaline phophatase, the margins of the
transduced area were traced and the stained area was
measured. This AP-stained area was then calculated as
a percentage of the entire luminal surface that had
been exposed to the vector to determine the prevalence
of transduction by area. The smooth muscle cells that
were transduced were found predominantly just underneath the luminal endothelium and could be visualized
on these en face preparations and quantified as the
percent of gene expression in the subendothelial, subintimal layer.
Cross sections were used to examine which cell types
were transduced and, on these cross sections, the
number of transduced cells were also counted manually, and the data were compared with the quantitative
results from the en face preparations. At least two cross
sections were evaluated for gene expression in 29 of the
50 transduced arteries in which gene expression was
calculated by the en face method. The entire circumference of each arterial cross section was examined in
8–12 low-powered microscopic fields. In each field, the
number of transduced cells in the media was compared
with the total number of medial cell nuclei. The sum of
the percentage of transduced cells for all fields was
calculated for each vessel cross section, and the results
of two to three cross sections were averaged for each
carotid artery. In the intima, the number of transduced
cells, although rare, was also counted manually. Because of the considerable variation in gene expression
119
120
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
scopic fields. In these specimens, intimal areas were
also quantified by computerized image analysis.
Statistical analysis. Data were compared between
groups using Student’s t-test. Differences were considered to be statistically significant at a P ⬍ 0.05.
Standard errors of the mean follow the mean values.
RESULTS
the percentage of the luminal circumference involved
with lesion development and by noting the number of
cell layers present between the endothelium and the
IEL.
Furthermore, in 11 arteries exposed to high titers of
vector and 6 sham-operated arteries, the absolute
number of T lymphocytes in the media were manually
counted, averaging results from at least two cross
sections per artery. As above, each arterial cross section
comprised a summation of 8–12 low-powered micro-
Fig. 3. Transduction of vascular smooth muscle cells (VSMCs). This example of a rabbit carotid artery cross section
using Nomarski technique shows VSMC expression of AP at 14 days after AAV vector transduction. It has been
stained histochemically for alkaline phosphatase (AP, dark blue-black), and then nuclei were lightly counterstained
with methyl green; white arrowheads point to the stained VSMCs in the media expressing AP protein. Overall, 55%
of medial VSMCs were expressing gene product on this specimen. The viral titer used was 2 ⫻ 109 AP FFU/ml. The
intact endothelial layer is marked with solid black arrowheads (original magnification, ⫻200). The en face
preparation of this artery is shown in Fig. 5B.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
Fig. 2. Time course of gene expression. Area of gene expression,
visualized by histochemical staining, was measured as a percentage
of total subluminal surface at days 3, 7, 10, 14 and 28 after AAV vector
transduction on en face specimens (see MATERIALS AND METHODS). The
viral titer used was 5 ⫻ 108 AP FFU/ml (with an actual volume
between 35 and 80 µl). The solid circles show the absolute values of
transduced areas, and the horizontal lines are mean values of all
transduced specimens at each time point. Open circles are the values
from untransduced contralateral carotid control specimens from the
same rabbits. We found a significant difference in the percentage area
transduced between carotid specimens at days 10 and 14, whereas
the differences were not statistically significant between 14 and 28
days. AP, alkaline phosphatase. FFU, focus-forming units. NS, not
significant.
Time course of expression within 28 days. First, the
time course of gene expression after AAV vector transduction was investigated (Fig. 2). We examined AP gene
expression at 3, 7, 10, 14, and 28 days, using between
35 and 80 µl of vector stock with a viral titer of 5 ⫻ 108
AP FFU/ml, in 25 rabbits. Minimal gene expression
was evident as early as days 3 and 7 after treatment
(mean 0.1% and 0.9 ⫾ 0.4% of the luminal surface area
with AP expression, respectively). The prevalence of
gene expression increased significantly between day 10
and day 14 (from a mean of 5 ⫾ 2% at day 10 to a mean
of 16 ⫾ 3% at day 14) and decreased somewhat, but
without statistical significance, by day 28.
No AP activity was seen in the sham-operated controls nor in the contralateral normal arteries, indicating that the histological techniques effectively eliminated any endogenous cross-reacting AP.
Cell types transduced. In these en face preparations
of the entire luminal surface, gene expression was not
uniform: areas of high enzyme activity were seen as
well as areas with little or no expression. Cross sections
revealed that, despite an intact endothelial layer,
VSMCs in the media were the main cell type transduced as evidenced by AP histochemical staining (Fig.
3). Specifically, gene expression was most commonly
seen in the layer of smooth muscle cells immediately
below the IEL. Immunocytochemistry using antibodies
to smooth muscle cells, T lymphocytes, and macrophages confirmed that the medial cells with histochemical staining for AP were smooth muscle cells and not T
lymphocytes or macrophages. AP staining was occasionally detected in rare endothelial cells in the intima.
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
Correlation between VSMC gene expression on en face
preps and cross sections. In 29 of the 50 arteries
examined, gene expression was also quantified on at
least two cross sections for each artery. On the cross
sections, gene expression was calculated as the percentage of medial cells with positive AP staining divided by
the total number of cells in the media, defined by the
number of nuclei. Across all specimens examined, gene
expression on the en face preparations, representing
the surface area transduced, correlated well with gene
expression on the cross sections, measured as a percent
of cells transduced (R2 ⫽ 0.69). However, in the specimens from the groups transduced with high titers of
vector and those transduced at high pressures, the
correlation between transduced areas on the en face
preps and gene expression on the cross sections was
even more striking (R2 ⫽ 0.82). The maximum transduced area on the en face preparations was 58%, which
compared well with 55% of medial smooth muscle cells
transduced on the cross sections of the same specimen.
Effect of viral titer. The percentage of the subluminal
area expressing gene product on the en face preparations varied with the viral titer (Fig. 4). A comparison
was made between the AP-stained subluminal area
using a viral titer of 5 ⫻ 108 AP FFU/ml (n ⫽ 11) vs. a
viral titer of 2 ⫻ 109 AP FFU/ml (n ⫽ 8) at 14 days
following transduction (volumes for both groups were
between 35 and 80 µl of vector stock). A significantly
higher subluminal area of gene expression was seen at
the higher titer (P ⬍ 0.05). The mean percentage area
expressing gene product on the en face preparations
was 34 ⫾ 7% for the higher titer vs. 15 ⫾ 3% for the
vessels transduced with the lower viral titer. Among
the specimens transduced with the higher viral titer,
the highest percentage of intra-arterial area expressing
gene product on the en face preparations was 58%,
demonstrating that widespread transduction of the
subintimal VSMC layer is feasible (Fig. 5).
Similarly, on the cross sections, the highest percentage of total medial smooth muscle cells expressing gene
product in any specimen was 55% (Fig. 3). When gene
product expression was calculated across the 8–12
fields comprising an entire arterial circumference and
averaged over the arteries comprising each group, the
average percentage of medial smooth muscle cells
expressing gene product was 26 ⫾ 6% for the specimens
transduced with the higher-titer vector compared with
Fig. 5. En face specimens after transduction with different viral titers. En face preparations of the luminal surfaces
of two carotid arteries at 14 days following transduction: one was exposed to a low viral titer (5 ⫻ 108 AP FFU/ml)
(A), and the other was exposed to a high viral titer (2 ⫻ 109 AP FFU/ml) (B). A greater prevalence of gene expression
(58%) is seen in the specimen exposed to the high viral titer (B) than in the specimen exposed to the lower viral titer
(A) (22%). Smooth muscle cells expressing gene product are visualized through the endothelium and internal elastic
lamina (IEL) on these whole tissue mounts, which accounts for the slightly hazy margins. Arrows demonstrate the
clamp position, where no AP expression is seen. White asterisks show the transduced areas (original magnification,
⫻40).
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
Fig. 4. Effect of viral titer on gene expression. Prevalence of gene
expression in specimens transduced with low viral titer (5 ⫻ 108 AP
FFU/ml) vs. high viral titer (2 ⫻ 109 AP FFU/ml) was compared at
day 14 after gene delivery (volume between 35 and 80 µl). En face
specimens were histochemically stained, and gene expression was
measured as the transduced area. Solid circles show the transduced
areas as a percentage of total subluminal surface area, and horizontal lines are mean values of all transduced specimens at each time
point. Open circles are the values for the untransduced contralateral
carotid artery control specimens. We found that increasing the viral
titer significantly increased the prevalence of gene expression, an
effect seen both on these en face specimens as well as on the cross
sections.
121
122
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
Fig. 6. Effect of exposure time on gene expression. Prevalence of gene
expression by percentage area transduced in specimens with exposure times to the virus of 30 min compared with 60 min, using a viral
titer of 2 ⫻ 109 AP FFU/ml (volume between 35 and 80 µl). Solid
circles show the absolute values of transduced areas, and horizontal
lines are mean values of all transduced specimens at each time point.
Open circles are values for untransduced contralateral carotid control specimens from the same rabbits. Although a longer exposure
time did increase the prevalence of gene expression, this did not
reach statistical significance (NS) with the sample size employed.
Fig. 7. Effect of intraluminal pressure on gene expression. Percentage surface area transduced in specimens subjected to low pressure
(peaks between 30 and 40 mmHg) vs. high intraluminal pressure
(peaks between 67 and 110 mmHg). Solid circles show values of
transduced areas, lines are mean values of all specimens in each
group, and open circles represent the control specimens. In this
model, higher intraluminal pressures resulted in a higher prevalence
of gene expression.
On the cross sections (n ⫽ 4 per group), the average
percentage of medial cells expressing gene product was
25 ⫾ 5% for the high-pressure group vs. 6 ⫾ 3% for the
low-pressure group, also a significant difference (P ⬍
0.05). Gene expression ranged from 14% to 38% in the
high-pressure group compared with 0 to 11% in the
low-pressure group.
Lesion development. Small but identifiable intimal
lesions, consisting of one or more cell layers beneath an
intact endothelium, were seen in 15 of 35 transduced
carotids examined (Fig. 8). Immunocytochemical analysis of the cell types involved revealed that the intimal
lesions were composed of ␣-actin-positive VSMCs. Similar mild degrees of intimal thickening were also seen in
6 of 7 sham-operated and in 2 of 18 native unoperated
vessels. Intimal lesions were not circumferential in any
of the specimens, and, overall, the observed lesions
would be classified as mild degrees of intimal thickening. In the 23 arteries with lesions, the lesions encompassed an average of 35% of the luminal circumference
in both the transduced and sham-operated vessels.
Similarly, in most of the involved arteries, the lesions
were only 1–2 cell layers deep. Only three of the
transduced vessels and two of the sham-operated arteries had lesions that were 2–4 cell layers in depth.
Although the majority of observations were made at 14
days following transduction, equivalent lesions were
seen in both a 10-day specimen (without gene expression) and a 28-day specimen (with gene expression).
Intimal areas were calculated by computerized image analysis for 11 ‘‘at risk’’ arteries transduced with
high-titer vectors and for 6 sham-operated control
arteries, and no significant difference was found between the transduced arteries and the sham controls.
When specimens were separated out by exposure to
high viral titers and/or elevated intraluminal pres-
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
2 ⫾ 1% for those at the lower titer, a very significant
difference (P ⫽ 0.002). Among the high-titer specimens,
gene expression on the cross sections ranged from 0 to
55%, with 7 of 10 high-titer specimens evidencing gene
expression of over 19%.
Effect of exposure time. Because gene expression
could also vary with the time that the vessel lumen was
exposed to the viral vector, we compared exposure
times of 30 min (n ⫽ 7) and 60 min (n ⫽ 8), with a
volume between 35 and 80 µl of vector stock and a viral
titer at 2 ⫻ 109 AP FFU/ml for both groups. A doubling
of the exposure time increased the area of gene expression (Fig. 6); however, this did not reach statistical
significance with the numbers of carotids we examined
(mean area expressing gene product was 21 ⫾ 4% for
30-min exposures compared with a mean area of 34 ⫾
7% for 60-min exposures).
Effect of intraluminal pressure. In this in vivo system, it was not possible to keep intraluminal pressure
absolutely constant over an hour’s time, so animals
were divided into two treatment groups (n ⫽ 5 per
group) where the vector was introduced under ‘‘high’’ or
‘‘low’’ intraluminal pressure, using a constant viral titer
of 5 ⫻ 108 FFU/ml. In the ‘‘high-pressure’’ group,
systolic peak pressures ranged for different animals
between 67 and 110 mmHg and, in the ‘‘low-pressure’’
group, pressures were maintained with peaks between
30 and 40 mmHg. The percentage of the vascular wall
surface expressing gene product was higher in arteries
transduced under high (mean area expressing gene
product was 28 ⫾ 2%) vs. low peak systolic pressures
(mean area of 12 ⫾ 6%) (P ⬍ 0.05) (Fig. 7).
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
123
sures, lesion development could not be tied to viral titer
(1/9 low-titer and 2/8 high-titer specimens had lesions).
In the transduced arteries, lesions were seen in arteries
both with and without gene expression. However, in
this sample, there were more lesions among the transduced specimens in the high-pressure group (4/5 specimens) than in the low-pressure group (2/5 specimens).
In two sham-operated arteries also exposed to high
intraluminal pressures, one of the two developed an
intimal lesion.
Inflammatory/immune reactivity. What differentiated the lesions in the transduced vessels from those in
the sham-operated arteries was the presence of T
lymphocytes in the transduced vessels (Fig. 8). Of 35
transduced vessels examined by immunocytochemistry, T lymphocytes were detected in the intima in 8, in
the media in 14 (or 40%), and in the adventitia in 24. In
contrast, only rare T lymphocytes were seen in the
adventitia and surrounding fibrinous tissue of shamoperated controls and the unoperated contralateral
carotid arteries, and no T lymphocytes were seen in
either the intima or media of these control specimens (7
sham-operated arteries and 15 native carotids were
examined).
Although adventitial infiltration of transduced arteries could be extensive, this was usually seen in vessels
in which a lesser degree of T lymphocytes were also
present in the media. Intimal T cells were much less
common and virtually always coexisted with T lymphocyte infiltration of the media.
In a sample of ‘‘at risk’’ arteries, we counted the
absolute number of T lymphocytes in the media, as
being the most accurate representation of vessel wall
infiltration. When the mean number of T lymphocytes
in the arterial media was compared between 11 arteries
exposed to high-titer vector (8 for 60 min, 3 for 30 min)
and 6 sham-operated arteries, the infiltration of T
lymphocytes was 100-fold higher in the transduced
vessels than in the sham controls (25 ⫾ 14 vs. 0.25 ⫾
0.2 T cells/total medial area, P ⬍ 0.05).
T lymphocyte infiltration of the media appeared to
correlate with gene expression, but, with this sample
size, clear differences related to intraluminal pressure,
viral titer, exposure time, or lesion development could
not be determined. Among the examined specimens, T
lymphocyte infiltration was seen exclusively in those
specimens with gene expression. Three of five specimens exposed to high intraluminal transduction pressures exhibited T cell infiltration compared with one of
four specimens in the low-pressure group. T lymphocyte infiltration was found in 5/8 specimens transduced
with high titers of virus and 4/9 low-titer specimens.
Among the high-titer specimens, T lymphocytes were
present in 5/7 specimens with 60-min exposure times
and 3/5 specimens with 30-min exposure times. Intimal
lesions were seen both in specimens with and without T
lymphocyte infiltration.
Macrophages were also examined by immunocytochemistry in 31 transduced and 3 sham-operated control specimens. Compared with T lymphocytes, macrophages were much less commonly detected, being
present in the media in only 7 (23%) of transduced
specimens. Intimal macrophages were present as more
than a single cell in only two transduced specimens.
Like the T lymphocytes, macrophages were not present
in the sham-operated controls.
Also like the T lymphocytes, macrophages were
found only in specimens with overt gene expression.
However, among the examined specimens with gene
expression, only 25% had medial macrophages. Also,
although five of the seven specimens with medial
macrophages also had intimal lesions, among all examined specimens with intimal lesions, only 31% had
macrophages present in the media.
Taken together, intimal lesion formation was found
with equal frequency and extent among transduced
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
Fig. 8. Inflammatory reaction to viral transduction at 14 days. Immunocytochemistry of two representative cross
sections of rabbit carotid arteries at 14 days after operation using antibodies to T lymphocytes [MAb Tib 188, T
lymphocytes stained brown (DAB)] and nuclear counterstaining with hematoxylin. A: specimen shows a
sham-operated carotid with a small intimal lesion but with no T lymphocytes in the media or intima and only rare T
lymphocytes in the peri-adventitial area (marked with the white arrowhead). B: in comparison, another specimen
shows an example of a transduced carotid artery, with a similar degree of overlying intimal lesion. In the transduced
specimen, T lymphocytes (white arrowheads) are evident in the media and in the peri-adventitia area, and a single
T lymphocyte is seen in the intima. The IEL, dividing the intima from the media, is marked with solid black arrows,
and the endothelial cell layer with solid black arrowheads, in both pictures (original magnification, ⫻400).
124
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
arteries and sham-operated controls, whereas T lymphocyte infiltration and, to a much lesser extent, macrophage infiltration were unique to the transduced arteries. Leukocyte infiltration appeared to relate to gene
expression.
DISCUSSION
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
This study demonstrates that, with optimization of
viral titer, exposure time, and intraluminal pressure,
AAV vectors can transduce approximately one-third of
the subendothelial layer of VSMCs in this rabbit carotid artery model of intra-arterial gene transduction,
with a maximum transduction rate of 58%. One of the
most important findings of this study is that, despite an
intact endothelial layer, VSMCs were preferentially
transduced by this AAV vector. This preferential transduction could be an advantage therapeutically, since
VSMC transduction could be an important gene therapy
target in the treatment of postangioplasty restenosis,
atherosclerosis, and posttransplant arteriopathy. Importantly, those smooth muscle cells that immediately
underlie the IEL are the ones most likely to express
adhesion molecules (8, 28) and synthetic phenotypes
(12, 22, 29) in several pathological processes and therefore, could be the preferred target cells for therapy (4,
25). Previous in vitro work has shown that VSMCs can
be transduced by AAV more readily than endothelial
cells (23), and that, in cultured VSMCs, gene expression can be sustained for at least 8 wk (25). It may be
that there is an inherent difference in relevant receptor
density on VSMCs as opposed to endothelial cells or
that the conversion rate from single to double-stranded
DNA may differ between cell types.
In vivo studies to date, however, have given conflicting results. Although skeletal and cardiac muscle is
readily transduced by AAV vector in vivo (10, 26, 37,
40), it has not been clear that the same is true for
vascular smooth muscle. Using the same AAV vector as
in our studies, Halbert et al. (16) did report VSMC gene
expression in murine pulmonary blood vessels after
AAV vector delivery by nasal aspiration, supporting a
potential predilection for VSMC transduction. However, among previous studies utilizing direct AAV vector transduction of arteries, only two (3, 13, 30) of four
research teams (3, 13, 23, 24, 30) detected gene expression in the media of the arterial wall. The Bahou
laboratory (3, 13) found evidence of transferred DNA by
in situ PCR in up to 90% of both endothelial and
VSMCs following intra-arterial catheter-based AAV
transduction of rat carotid arteries. Interestingly, using
higher intraluminal pressures and lower viral titers
than in our studies, they found a pattern of transduction similar to our findings, with preferential transduction of the layer of VSMCs just under the IEL in a
vessel with an intact endothelium. The low levels of
recombinant protein expression, which could not be
detected histochemically, may well be related to the
considerably lower viral titers (2–5 ⫻ 105 IU/ml) and
total viral dose utilized for these studies. Rolling et al.
(30) also found transduction of both medial VSMCs and
adventitial cells of rat carotid arteries utilizing a
cytomegalovirus (CMV) promoter, a viral titer and dose
near the range of our low-titer group, and a 20-min
exposure time. However, in their study, as in Lynch et
al. (23), balloon-injured carotids demonstrated significantly greater transduction than uninjured arteries,
whereas in our work, we see effective gene transfer
through an intact endothelium.
The preferential VSMC transduction we found is in
contrast to the findings of Lynch et al. (23), who found
that endothelial cells in the adventitial vaso vasorum
were preferentially transduced in hypercholesterolemic primates. In our rabbit carotid model, microvessels surrounding the adventitia are dissected off prior
to transduction, thus limiting the opportunities for
transduction of adventitial vessels, which may explain
some of the differences between our findings and those
of Lynch et al. (23); vaso vasorum may also be more
prominent in primates. Maeda et al. (24) also found
transduction of only endothelial and adventitial cells
when they exposed cross sections of rat aorta to AAV
vectors. The viral titer used by Lynch et al. (23) is
probably equivalent to our low range, and that used by
Maeda et al. (24) is probably equivalent to our high
range. However, both of these laboratories were examining gene expression at very early time points, 1–3
days posttransduction. In our studies, we saw only
minimal gene expression at 3 days, and our cross
sections to evaluate the cell types expressing gene
product were performed at 14 days, which may explain
some of the variance in findings.
The choice of promoter is also known to influence cell
type-specific and organ-specific expression. A CMV
promoter, known to provide efficient gene expression in
arterial endothelial cells, was used by both Lynch et al.
(23) and Maeda et al. (24). At much lower viral titers,
the Bahou laboratory (3) found functional recombinant
protein could be detected only with a collagen ␣(I)
promoter and not with a CMV promoter, although
similarly efficient in vivo transgene delivery into both
endothelial cells and VSMCs was demonstrated with
both promoters (3). The RSV promoter in the construct
we used may have facilitated gene expression in VSMCs
vs. endothelial cells. We did observe limited gene
expression in endothelial cells in the intima in some of
our carotid specimens, so we know that in vivo AAV
vector transduction of endothelial cells is possible with
the vector construct we used.
Using this same AAV vector construct, RSV promoter, and human AP reporter gene in vitro, we found
surprisingly efficient transduction of nonconfluent rabbit vena caval endothelial cells (14) with an actual viral
dose of 5 ⫻ 103 AP FFU/ml (data not shown). However,
our in vitro results were not predictive of the low
incidence of endothelial cell transduction in vivo. One
reason for this difference between in vitro and in vivo
gene expression might be the proliferating state of the
cultured cells compared with the quiescent state of
endothelial cells in vivo, given that AAV vectors may
preferentially transduce cells in S phase (32). Alternatively, it seems likely that the endothelial cells in the
artery walls in the in vivo experiments may have been,
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
In these studies, we first detected appreciable gene
expression histochemically on day 10, and expression
increased significantly by day 14. Similarly, Rolling et
al. (3) found that gene expression in AAV-transduced
rat carotid arteries accelerated between days 10 and
20, persisted for 30 days, and, in one animal, was still
detectable at 6 mo (30). Although we also show here
that gene expression can be sustained for at least 4 wk,
we did not examine later time points, and, certainly,
gene expression might have been demonstrable for a
much longer period of time in these vessels. In murine
studies, gene expression in transduced skeletal muscle
has been reported for up to 1.5 years after AAV vector
gene delivery (40).
An unexpected finding was an apparent mild inflammatory response to AAV vector transduction with small
intimal lesions in some arteries. AAV vectors are generated by excluding viral regulatory and structural genes
to reduce host cellular inflammatory response. However, in 14 of 35 transduced specimens examined, we
observed mild T lymphocyte infiltration into at least
two of the three layers of the vessel wall. This inflammatory cell migration into the media was clearly different
in the transduced vessels compared with sham-operated carotid arteries. This inflammatory response might
be triggered by the AAV vector capsid, since Halbert et
al. (16) have found that a humoral immune response
can be generated to AAV vector proteins, and, therefore,
a cellular response might also be possible. If so, other
AAV vector serotypes (33) may be found to be less
immunogenic. Alternatively, this could be a reaction to
the expressed protein (27), in our case the human
placental AP gene product, although, by and large, AP
is not known to engender an inflammatory reaction.
Brockstedt et al. (7) have recently reported that
high-titer intravenous, but not intramuscular, AAV
transduction can induce a cytotoxic T lymphocyte response as well as a humoral response against the
foreign protein. Their intravenous delivery model is
likely to be more immunogenic than ours, because the
intravenous vector-DNA constructs could traffic through
the spleen and lymphoid organs, whereas, in our experiments, the vector is isolated in an intra-arterial section.
However, even plasmid DNA encoding an endogenous
protein has been shown capable of inducing a host
immune response (35).
Finally, vascular endothelial injury and repair due to
a needle tip abrasion or surgical dissection of the
adventitia (5) might explain the small intimal lesions
in sham-operated vessels, but this does not explain the
greater prevalence of T cell infiltration in the transduced vessels. Certainly, among experimental animals,
rabbits are more prone to vasculopathy following any
stimulus than other experimental animals; the question is whether such sensitive vascular reactivity in the
rabbit would actually mimic expected vascular responses in humans. Further investigations are needed
to confirm whether these preliminary findings represent a cellular inflammatory response to the viral
vector or to the foreign protein.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
in fact, transduced, but that gene expression might
have been at a lower level than what we could detect
histologically.
AAV belongs to the parvovirus family and, with a
particle size of 20 nm, is one of the smallest viruses.
However, its small size, although generally a limiting
factor for gene packaging, might also enable AAV
vectors to gain access to cells and portions of the vessel
wall that are not accessible to larger vectors. It has
been proposed (31) that, during intravascular gene
delivery, anatomical barriers, in particular the IEL,
may prevent larger vectors from penetrating into the
media. Rome et al. (31) have even suggested that the
media of the vessel wall is virtually inaccessible to
vectors with a size around 70 to 100 nm such as
adenoviruses or the larger liposome vectors. In contrast, smaller particles such as low-density lipoproteins
with a diameter of 22–24 nm, could pass, although to a
limited degree, into the media beyond the IEL (38). In
our studies, the AAV vector, smaller than even the
above-mentioned low-density lipoproteins, might reach
the tunica media simply due to its small size. Similarly,
in the lung (16), the AAV vector must have been able to
circumvent physical barriers such as the basement
membrane of the airway epithelium to gain access to
VSMCs. Dzau et al. (9), using HVJ-liposome vectors,
have also demonstrated that factors other than size
alone may affect vector penetration of tissue (9).
Although we did not create a demonstrable arterial
injury to the endothelium or disrupt the IEL, it is also
possible that, in our surgical model, simply occluding
the blood flow through the vessel for 30–60 min might
result in some ischemia of the vascular wall itself,
which could, theoretically, alter the permeability of the
endothelial layer and facilitate penetration of the vector through endothelial gap junctions to deeper layers
of the vessel wall. In comparison, other investigators (3,
13, 23, 30) have introduced the AAV vectors through
intraluminal catheters without subjecting the vessel
wall to this mild degree of ischemia. Differences in
endothelial cell permeability may thus also account for
some differences in results.
Our findings show that AAV vectors can transduce
VSMCs in the media in vivo with an apparent transduction efficiency of up to 58%. This was achieved with
intraluminal pressures up to 110 mmHg and without
prior injury of the vessel and, thus an intact endothelial
layer. Although Nabel et al. (27) have proposed that
pressures in the range of 150–200 mmHg might be
needed for effective DNA-liposome transfection, we
chose pressures around 100 mmHg to avoid barotrauma to the artery. In our comparisons between
arteries transduced with pressures between 67 and 110
mmHg and those transduced at lower pressures, we
found an increase in the prevalence of gene expression
at the higher pressures. In both groups, however,
VSMCs were clearly the main cell type transduced as
detected by the AP histochemical staining: the higher
pressures did not increase the number of endothelial
cells expressing gene product.
125
126
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
We acknowledge the contributions of Edward Sien in the review
and preparation of this report.
Portions of this project were supported by National Institute of
Diabetes and Digestive and Kidney Diseases Grant DK-47754 and by
Cystic Fibrosis Foundation Grant CFFR565.
Address for reprint requests and other correspondence: M. D.
Allen, Division of Cardiothoracic Surgery, Mailstop 356310, 1959 NE
Pacific St., Seattle, WA 98195 (E-mail: [email protected]).
12.
13.
14.
15.
16.
17.
18.
REFERENCES
1. Alexander IE, Russell DW, Spence AM, and Miller AD.
Effects of gamma irradiation on the transduction of dividing and
nondividing cells in brain and muscle of rats by adeno-associated
virus vectors. Hum Gene Ther 7: 841–850, 1996.
2. Allen JM, Debelak DJ, Reynolds TC, and Miller AD. Identification and elimination of replication-competent adeno- associated virus (AAV) that can arise by nonhomologous recombination
during AAV vector production. J Virol 71: 6816–6822, 1997.
3. Arnold TE, Gnatenko D, and Bahou WF. In vivo gene transfer
into rat arterial walls with novel adeno-associated virus vectors.
J Vasc Surg 25: 347–355, 1997.
4. Bai H, Morishita R, Kida I, Yamakawa T, Zhang W, Aoki M,
Matsushita H, Noda A, Nagai R, Kaneda Y, Higaki J,
Ogihara T, Sawa Y, and Matsuda H. Inhibition of intimal
hyperplasia after vein grafting by in vivo transfer of human
senescent cell-derived inhibitor-1 gene. Gene Ther 5: 761–769,
1998.
5. Barker SG, Tilling LC, Miller GC, Beesley JE, Fleetwood
G, Stavri GT, Baskerville PA, and Martin JF. The adventitia
and atherogenesis: removal initiates intimal proliferation in the
rabbit which regresses on generation of a ‘‘neoadventitia’’. Atherosclerosis 105: 131–144, 1994.
6. Beckstead JH. A simple technique for preservation of fixationsensitive antigens in paraffin-embedded tissues. J Histochem
Cytochem 42: 1127–1134, 1994.
7. Brockstedt DG, Podsakoff GM, Fong L, Kurtzman G,
Mueller-Ruchholtz W, and Engelman EG. Induction of immunity to antigens expressed by recombinant adeno-associated
virus depends on the route of administration. Clin Immunol 92:
67–75, 1999.
8. Duplaa C, Couffinhal T, Dufourcq P, Llanas B, Moreau C,
and Bonnet J. The integrin very late antigen-4 is expressed in
human smooth muscle cell. Involvement of alpha 4 and vascular
cell adhesion molecule-1 during smooth muscle cell differentiation. Circ Res 80: 159–69, 1997.
9. Dzau VJ, Mann MJ, Morishita R, and Kaneda Y. Fusigenic
viral liposomes for gene therapy in cardiovascular diseases. Proc
Natl Acad Sci USA 93: 11421–11425, 1996.
10. Fisher KJ, Jooss K, Alston J, Yang Y, Haecker SE, High K,
Pathak R, Raper SE, and Wilson JM. Recombinant adenoassociated virus for muscle directed gene therapy. Nat Med 3:
306–312, 1997.
11. Fisher-Adams G, Wong KK Jr, Podsakoff G, Forman SJ,
and Chatterjee S. Integration of adeno-associated virus vectors
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
in CD34⫹ human hematopoietic progenitor cells after transduction. Blood 88: 492–504, 1996.
Giuriato L, Chiavegato A, Pauletto P., and Sartore S.
Correlation between the presence of an immature smooth muscle
cell population in tunica media and the development of atherosclerotic lesion: a study on different-sized rabbit arteries from
cholesterol-fed and Watanabe heritable hyperlipidemic rabbits.
Atherosclerosis 116: 77–92, 1995.
Gnatenko D, Arnold TE, Zolotukhin S, Nuovo GJ, Muzyczka N, and Bahou WF. Characterization of recombinant adenoassociated virus-2 as a vehicle for gene delivery and expression
into vascular cells. J Investig Med 45: 87–98, 1997.
Goldman CK and Gillespie GY. Physiologic responses of
REVC, a continuous rabbit endothelial vascular cell. J Vasc Res
32: 31–40, 1995.
Halbert CL, Standaert TA, Aitken ML, Alexander IE, Russell DW, and Miller AD. Transduction by adeno-associated
virus vectors in the rabbit airway: efficiency, persistence, and
readministration. J Virol 71: 5932–5941, 1997.
Halbert CL, Standaert TA, Wilson CB, and Miller AD.
Successful readministration of adeno-associated virus vectors to
the mouse lung requires transient immunosuppression during
the initial exposure. J Virol 72: 9795–9805, 1998.
Jackson S, Chused TM, Wilkinson JM, Liserson WM, and
Kindt TJ. Differentiation antigens identify subpopulations of
rabbit T and B lymphocytes. J Exp Med 157: 34–46, 1983.
Kaplitt MG, Leone P, Samulski RJ, Xiao X, Pfaff DW,
O’Malley KL, and During MJ. Long-term gene expression and
phenotypic correction using adeno- associated virus vectors in
the mammalian brain. Nat Genet 8: 148–154, 1994.
Kaplitt MG, Xiao X, Samulski RJ, Li J, Ojamaa K, Klein IL,
Makimura H, Kaplitt MJ, Strumpf RK, and Diethrich EB.
Long-term gene transfer in porcine myocardium after coronary
infusion of an adeno-associated virus vector. Ann Thorac Surg
62: 1669–1676, 1996.
Kessler PD, Podsakoff GM, Chen X, McQuiston SA, Colosi
PC, Matelis LA, Kurtzman GJ, and Byrne BJ. Gene delivery
to skeletal muscle results in sustained expression and systemic
delivery of a therapeutic protein. Proc Natl Acad Sci USA 93:
14082–14087, 1996.
Koeberl DD, Alexander IE, Halbert CL, Russell DW, and
Miller AD. Persistent expression of human clotting factor IX
from mouse liver after intravenous injection of adeno-associated
virus vectors. Proc Natl Acad Sci USA 94: 1426–1431, 1997.
Kuwahara M, Jacobsson J, Kuwahara M, Kagan E, Ramwell PW, and Foegh ML. Coronary artery ultrastructural
changes in cardiac transplant atherosclerosis in the rabbit.
Transplantation 52: 759–765, 1991.
Lynch CM, Hara PS, Leonard JC, Williams JK, Dean RH,
and Geary RL. Adeno-associated virus vectors for vascular gene
delivery. Circ Res 80: 497–505, 1997.
Maeda Y, Ikeda U, Ogasawara Y, Urabe M, Takizawa T,
Saito T, Colosi P, Kurtzman G, Shimada K, and Ozawa K.
Gene transfer into vascular cells using adeno-associated virus
(AAV) vectors. Cardiovasc Res 35: 514–521, 1997.
Mohuczy D, Gelband CH, and Phillips MI. Antisense inhibition of AT1 receptor in vascular smooth muscle cells using
adeno-associated virus-based vector. Hypertension 33: 354–359,
1999.
Monahan PE, Samulski RJ, Tazelaar J, Xiao X, Nichols TC,
Bellinger DA, Read MS, and Walsh CE. Direct intramuscular
injection with recombinant AAV vectors results in sustained
expression in a dog model of hemophilia. Gene Ther 5: 40–49,
1998.
Nabel EG, Plautz G, and Nabel GJ. Transduction of a foreign
histocompatibility gene into the arterial wall induces vasculitis.
Proc Natl Acad Sci USA 89: 5157–5161, 1992.
O’Brien KD, McDonald TO, Chait A, Allen MD, and Alpers
CE. Neovascular expression of E-selectin, intercellular adhesion
molecule-1, and vascular cell adhesion molecule-1 in human
atherosclerosis and their relation to intimal leukocyte content.
Circulation 93:672–682, 1996.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
In summary, this AAV vector transduced vascular
tissue and, in particular, VSMCs with an AP reporter
gene for up to 28 days. The prevalence of gene expression, which varied with viral titer, exposure time, and
intraluminal pressure, could be as high as 58% by area.
These studies of rabbit carotid artery AAV vector
transduction also demonstrated that a mild cellular
inflammatory response could be elicited, which appeared to occur only in the specimens with gene product
expression.
This study shows that AAV vectors have the potential
for becoming very promising vectors for clinical applications in cardiovascular disease because of their relatively high transduction efficiency, potential for longterm gene expression, and selective targeting of VSMCs.
AAV VECTOR TRANSDUCTION OF VASCULAR SMOOTH MUSCLE CELLS
35. Shull RM, Lu X, McEntee MF, Bright RM, Pepper KA, and
Kohn DB. Myoblast gene therapy in canine mucopolysaccharidosis. I. Abrogation by an immune response to alpha-L-iduronidase.
Hum Gene Ther 7: 1595–1603, 1996.
36. Snyder RO, Miao CH, Patijn GA, Spratt SK, Danos O, Nagy
D, Gown AM, Winther B, Meuse L, Cohen LK, Thompson
AR, and Kay MA. Persistent and therapeutic concentrations of
human factor IX in mice after hepatic gene transfer of recombinant AAV vectors. Nat Genet 16: 270–276, 1997.
37. Svensson EC, Marshall DJ, Woodard K, Lin H, Jiang F,
Chu L, and Leiden JM. Efficient and stable transduction of
cardiomyocytes after intramyocardial injection or intracoronary
perfusion with recombinant adeno-associated virus vectors. Circulation 99: 201–205, 1999.
38. Tompkins RG. Quantitative analysis of blood vessel permeability of squirrel monkeys. Am J Physiol Heart Circ Physiol 260:
H1194–H1204, 1991.
39. Tsukada T, Rosenfeld M, Ross R, and Gown AM. Immunocytochemical analysis of cellular components in atherosclerotic
lesions. Use of monoclonal antibodies with the Watanabe and
fat-fed rabbit. Arteriosclerosis 6: 601, 1986.
40. Xiao X, Li J, and Samulski RJ. Efficient long-term gene
transfer into muscle tissue of immunocompetent mice by adenoassociated virus vector. J Virol 70: 8098–8108, 1996.
http://physiolgenomics.physiology.org
Downloaded from http://physiolgenomics.physiology.org/ by 10.220.33.2 on June 18, 2017
29. Okamoto E, Suzuki T, Aikawa M, Imataka K, Fujii J,
Kuro-o M, Nakahara K, Hasegawa A, Yazaki Y, and Nagai
R. Diversity of the synthetic-state smooth muscle cells proliferating in mechanically and hemodynamically injured rabbit arteries. Lab Invest 74: 120–127, 1996.
30. Rolling F, Nong Z, Pisvin S, and Collen D. Adeno-associated
virus-mediated gene transfer into rat carotid arteries. Gene Ther
4: 757–761, 1997.
31. Rome JJ, Shayani V, Flugelman MY, Newman KD, Farb A,
Virmani R, and Dichek DA. Anatomic barriers influence the
distribution of in vivo gene transfer into the arterial wall.
Modeling with microscopic tracer particles and verification with
a recombinant adenoviral vector. Arterioscler Thromb 14: 148–
161, 1994.
32. Russell DW, Miller AD, and Alexander IE. Adeno-associated
virus vectors preferentially transduce cells in S phase. Proc Natl
Acad Sci USA 91: 8915–8919, 1994.
33. Rutledge EA, Halbert CL, and Russell DW. Infectious clones
and vectors derived from adeno-associated virus (AAV) serotypes
other than AAV type 2. J Virol 72: 309–319, 1998.
34. Samulski RJ, Chang LS, and Shenk T. Helper-free stocks of
recombinant adeno-associated viruses: normal integration does
not require viral gene expression. J Virol 63: 3822–3828, 1989.
127