© 1999 American Heart Association, Inc.
UltraRapid Communication |
A Mouse Model of Arterial Gene Transfer
Antigen-Specific Immunity Is a Minor Determinant of the Early Loss of Adenovirus-Mediated Transgene Expression
From the Gladstone Institute of Cardiovascular Disease (G.V., R.A., R.Q., C.A., D.A.D.), Daiichi Research Center (G.V., D.A.D.), and Department of Medicine (R.A., R.Q., D.A.D.), University of California, San Francisco, Calif.
Correspondence to David A. Dichek, MD, Gladstone Institute of Cardiovascular Disease, PO Box 419100, San Francisco, CA 94141-9100. E-mail ddichek{at}gladstone.ucsf.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abstract—We developed a murine model of arterial gene transfer and used it to test the role of antigen-specific immunity in the loss of adenovirus-mediated transgene expression. Adenoviral vectors encoding either ß-galactosidase (ß-gal) or green fluorescent protein were infused to the lumen of normal common carotids of CD-1 and C57BL/6 mice and atherosclerotic carotids of Apoe-/- mice. At 3 days after gene transfer, significant reporter gene expression was detected in all strains. Transgene expression was transient, with expression undetectable at 14 days. Next, a ß-gal–expressing vector was infused into carotids of ROSA26 mice (transgenic for, and therefore tolerant of, ß-gal) and RAG-2-/- mice (deficient in recombinase-activating gene [RAG]-2 and therefore lacking in antigen-specific immunity). ß-Gal expression was again high at 3 days but declined substantially (>90%) by 14 days. In vivo labeling with bromodeoxyuridine revealed that carotid endothelial proliferation was increased dramatically by the gene-transfer procedure alone, likely leading to the loss of episomal adenoviral DNA. Gene transfer to normal and atherosclerotic mouse carotids can be accomplished; however, elimination of antigen-specific immune responses does not prevent the early loss of adenovirus-mediated transgene expression. Efforts to prolong adenovirus-mediated transgene expression in the artery wall must be redirected. These efforts will likely include strategies to avoid the consequences of increased cell turnover. Nevertheless, despite the brevity of expression, this mouse model of gene transfer to normal and severely atherosclerotic arteries will likely be useful for investigating the genetic basis of vascular disease and for developing gene therapies. The full text of this article is available at http://www.circresaha.org.
Key Words: adenovirus • ß-galactosidase • gene therapy • ROSA26 • RAG-2-/-
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Arterial gene transfer with adenoviral vectors is a useful technique for investigations in vascular biology and potentially for human gene therapy.1 2 3 The utility of adenoviral vectors as experimental and therapeutic tools is limited, in part, by brevity of transgene expression. Transgene expression declines precipitously within 2 weeks after adenoviral arterial gene transfer,4 5 6 thereby confining use of adenoviral vectors to investigations and therapies for which transient expression is adequate. Prolongation of adenovirus-mediated transgene expression could substantially increase its utility. The development of methods to prolong adenovirus-mediated transgene expression is most rationally guided by an understanding of the mechanisms by which expression is lost.
Substantial data, generated primarily in mouse models of hepatic, pulmonary, and skeletal muscle gene transfer, implicate antigen-specific immune responses to either adenoviral proteins or foreign, immunogenic transgenes in the loss of adenovirus-mediated transgene expression.7 8 9 10 11 12 Much of the data implicating antigen-specific immune responses to adenoviral proteins were produced in either athymic "nude" mice or in mice with targeted deletions of components of the immune system ("knockout" mice).7 8 9 Data supporting a role for foreign, immunogenic transgenes (such as Escherichia coli ß-galactosidase [ß-gal]) in the loss of adenovirus-mediated expression were also produced in mice.10 11 13 14 In contrast to the substantial data produced in liver, lung, and skeletal muscle models, a relatively small amount of data supports a role for the immune system in loss of transgene expression after arterial gene transfer.15 Technical difficulties involved in performing gene transfer in mouse arteries have prevented the use of informative mouse strains to test more definitively the role of immunity in the loss of transgene expression after arterial gene transfer.
We sought to test the role of the antigen-specific immune system in the loss of gene expression after arterial gene transfer and to determine, simultaneously, whether mouse arteries were a suitable substrate for informative vascular gene transfer studies. We developed a technique for gene delivery to the mouse common carotid artery and applied it successfully to both normal and atherosclerotic arteries. We then tested whether transgene persistence is prolonged in transgenic ROSA26 mice that are tolerant of the ß-gal reporter gene16 or in mice in which the recombinase-activating-2 gene (RAG-2-/-) is deleted, which therefore lack antigen-specific immunity.17 Surprisingly, neither transgene tolerance nor absence of antigen-specific immunity insured prolonged, stable transgene expression. Our data are most consistent with the notion that transgene expression is lost primarily due to increased cell proliferation that is triggered by manipulation of the artery at the time of gene delivery.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adenoviral Vectors
We used four replication-defective adenoviral vectors: AdRSVnlacZ and AdCMVnlacZ (both express nuclear-targeted ß-gal), AdCMV-EGFP (expresses green fluorescent protein), and AdNull (does not contain a transgene). Concentrated viral stocks were prepared, stored, and titered as described.4 As a control infusate, virus storage buffer (10 mmol/L Tris-HCl, pH 7.4, 1 mmol/L MgCl2, and 10% glycerol) was diluted with M-199 (Life Technologies) in the same proportions as was done for the virus stocks.
Gene Transfer Into Mouse Carotid Arteries
All animal procedures were approved by the Animal Care and Use
Committee of the University of California, San Francisco. Mice used
included male CD1, C57BL/6, and ROSA26 mice16
(C57BL/6x129 background; Jackson Laboratories, Bar Harbor,
Maine); RAG-2-/- mice17
(C57BL/6x129 background; Taconic Farms, Germantown, NY); and
Apoe-/- mice (C57BL/6 background;
Berlex Biosciences, Richmond, Calif). Anesthesia, carotid
surgery, ß-gal activity assay, and histochemical detection of ß-gal
activity were performed according to procedures similar to those we
have described previously for rats.4 Within each
experimental group, background (endogenous) ß-gal
activity was defined as the highest activity measured in an extract
from a vessel transduced with AdNull. As expected, background ß-gal
activity was high in carotid arteries of ROSA26 mice (see below).
Histology and Immunohistochemistry
Vessel harvest, embedding, frozen sectioning, hematoxylin and
eosin (H&E) staining, and immunohistochemistry were performed
essentially as described previously.6 T lymphocytes
were detected by incubation of sections with a rat monoclonal antibody
to mouse CD5 (Ly-1; 53-7.3; Pharmingen) at a final dilution of
1:100.
Injection of Mice With Bromodeoxyuridine and 5-Azacytidine
Bromodeoxyuridine (BrdUrd; Sigma) injection and
immunohistochemical detection were performed essentially as described
previously.5 Injection with 5-azacytidine (5-Aza), an
agent that can demethylate genes and induce transcription both in vitro
and in vivo, was performed as described.18
Statistical Analysis
Group results are presented as median (range).
Comparisons between groups were made using the Mann-Whitney U test (for
two groups) and Kruskal-Wallis ANOVA (for three groups).
An expanded Materials and Methods section is available online at http://www.circresaha.org.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Adenoviral Gene Transfer to Normal Mouse Carotid Arteries
Carotid arteries of outbred CD1 mice were transduced with an adenoviral vector expressing ß-gal (AdRSVnlacZ; 108 to 1011 plaque-forming units [pfu]/mL) or with a vector that does not express a transgene (AdNull; 1x1010 pfu/mL) and were harvested 3 days later. Arteries infused with AdRSVnlacZ at 108 pfu/mL did not contain detectable vector-driven ß-gal activity (Figure 1A
). ß-gal expression was detected
above background levels in arteries infused with higher concentrations
of AdRSVnlacZ: 109 pfu/mL, 0.4 µU/µg (0 to
1.7); 1010 pfu/mL, 1.8 µU/µg (0.4 to 2.3);
2x1010 pfu/mL, 1.5 µU/µg (0.3 to 11.3);
5x1010 pfu/mL, 1.4 µU/µg (0.5 to 1.9); and
1011 pfu/mL, 0.5 µU/µg (0.3 to 0.8). ß-Gal
activity in arteries infused with AdNull at
1x1010 pfu/mL was 0.2 µU/µg (0 to 0.4;
P<0.01 versus AdRSVnlacZ infused at this concentration).
Because maximal ß-gal activity was obtained at 1 to
5x1010 pfu/mL, further studies were carried out
with infusion of AdRSVnlacZ in this range.
|
Many informative transgenes and knockout mutations are maintained on
the inbred C57BL/6 genetic background. To determine whether
arterial gene delivery was also successful in this strain,
carotid arteries of C57BL/6 mice were transduced with
AdRSVnlacZ at 1x1010,
2x1010, and 5x1010
pfu/mL. ß-Gal activity in these arteries was 2.7 µU/µg (0.5 to
8.2), 1.5 µU/µg (0.3 to 31.0), and 0.53 µU/µg (0.5 to 2.8),
respectively (Figure 1B
). Arteries infused with AdNull at
1x1010 pfu/mL had ß-gal activity of 0.24
µU/µg (0.2 to 1.0; P<0.05 versus AdRSVnlacZ at
1x1010 pfu/mL).
To identify the cell type that expressed ß-gal, we first examined
X-Gal–stained sections of arteries harvested 3 days after infusion of
AdRSVnlacZ in CD1 mice. Blue nuclei were present almost exclusively
in endothelial cells (Figure 2A). Approximately 10% of luminal
endothelial cells had blue nuclei. No blue nuclei were
seen in control arteries infused with AdNull and stained with X-Gal
(not shown). Similarly, in CD-1 arteries infused with AdCMV-EGFP,
10% of luminal endothelial cells demonstrated
bright green fluorescence (Figure 2B). No such
fluorescence was seen in arteries infused with AdNull.
|
Adenoviral Gene Transfer to Atherosclerotic Carotid Arteries of
Apoe-/- Mice
We tested whether advanced atherosclerotic lesions in the carotid
arteries of Apoe-/- mice were also
susceptible to adenovirus-mediated gene transfer. ß-Gal activity in
Apoe-/- carotids transduced with
AdCMVnLacZ (6x109 pfu/mL) was 490 µU/µg (450
to 2400; n=3) compared with 40 µU/µg (40 to 41; n=2) in
Apoe-/- carotids transduced with
AdNull. The relatively high levels in both controls and experimentals
likely result from a high background from tissue macrophages
(which can express endogenous ß-gal19 )
and from use of the powerful AdCMVnlacZ vector (see below).
X-gal–stained sections (Figure 2C) revealed ß-gal expression
in 13% of endothelial cells (12% to 15%; n=3) and
9% of adventitial cells (7% to 15%; n=3). No nuclear-localized X-gal
staining was detected in arteries transduced with AdNull.
ß-Gal Is Expressed Transiently After Arterial
Gene Transfer
In addition to the arteries harvested at 3 days (reported above),
carotid arteries in CD-1 mice were transduced with AdRSVnlacZ at
2x1010 pfu/mL and excised at 7 or 14 days after
gene transfer. ß-Gal activity in arterial extracts was
unchanged between 3 days (1.5 µU/µg [0.3 to 11.3]) and 7 days
(1.4 µU/µg [0.81 to 1.8]) but declined to an essentially
undetectable level at 14 days (Figure 3A). Similar results were obtained in
C57BL/6 mice (Figure 3B): 1.8 µU/µg (0 to 54) at 7 days and
0.15 µU/µg (0.14 to 0.24) at 14 days.
|
Adenovirus Infusion in Mouse Carotid Arteries Has Minimal Effects
Beyond Those Caused by Vehicle Infusion Alone
Mouse carotid surgery is technically demanding. The utility of the
mouse carotid model for biological investigations might be limited if
surgery and adenovirus infusion per se had significant effects on the
structure of the carotid artery wall. For example, if substantial
neointimal formation, vessel damage, or inflammation was
produced by adenovirus infusion, then it could be challenging to
discern specific effects of overexpressed transgenes.6 We
performed two independent experiments to investigate this issue. In the
first experiment, frozen sections of arteries of CD1 mice infused with
either AdRSVnlacZ, AdNull, or vehicle were stained with H&E and with
specific antibodies to CD5 (a T-cell antigen) at 3 days (two
arteries per group) and 10 days (three arteries per group). At 3 days,
H&E–stained sections from arteries in all of the groups appeared
similar: occasional medial necrosis was present (evidenced by
absence of smooth muscle cell nuclei), and scant inflammatory cell
infiltrates were seen in the adventitia. Immunohistochemical staining,
however, detected virtually no CD5+ cells in
these infiltrates. At 10 days, H&E–stained sections were again
indistinguishable among the three experimental groups. In all arteries,
there were areas of hypercellularity in both the media and adventitia.
CD5+ cells were primarily restricted to
the adventitia. There was a median of three (2 to 32)
CD5+ cells per section in arteries infused with
AdNull (n=3), three (2 to 3) CD5+ cells per
section in arteries infused with AdRSVnlacZ, and one (0 to 11)
CD5+ cell per section in arteries infused with
buffer. The number of CD5+ cells per section did
not differ significantly between the three groups
(P=0.63).
In a second series of experiments, sections of perfusion-fixed arteries
of C57BL/6 mice infused with either AdNull or vehicle were examined at
3 days (two arteries per group) and at 14 days (two arteries per
group). At both time points, the histologic appearance of arteries from
the AdNull and vehicle groups was similar. At 3 days, sections of
arteries from both groups showed occasional endothelial
denudation and smooth muscle cell loss. Apparent inflammatory cell
infiltrates were seen primarily in the adventitia. At 14 days, sections
of vessels from both groups demonstrated increased medial cellularity
when compared with either uninstrumented arteries or to arteries
harvested 3 days after gene transfer (Figure 4A through 4C). The adventitia of the
14-day arteries was also thickened compared with 3-day arteries and to
unmanipulated arteries, and there were focal areas of inflammation
(Figure 4A). There was only minimal evidence of
neointimal growth: occasionally, one cell layer was
present between the internal elastic lamina and the
endothelium (Figure 4B). Thus, the surgical
procedure itself can result in moderate tissue damage, inflammation,
and vascular remodeling that is manifested by increased medial
cellularity and adventitial thickening. (These changes were less
prominent in arteries manipulated by more experienced operators [data
not shown].) However, on the basis of histological and
immunohistochemical analyses performed on multiple sections
from a total of 12 adenovirus-infused arteries in two independent
series of experiments, we did not detect an increase in inflammation or
vessel damage that could be attributed to the infusion of
adenovirus.
|
ß-Gal Expression Is Not Stable After Arterial Gene
Transfer in Transgene-Tolerant or Immunodeficient Mice
As an initial application of the mouse carotid model, we
investigated the role of antigenicity of the transgene product (in
this case, E coli ß-gal) in the rapid loss of transgene
expression. We performed carotid gene transfer in ROSA26 mice, which
are transgenic for (and therefore immunologically tolerant of) ß-gal.
We did not use the AdRSVnlacZ vector for these experiments because
preliminary experiments in untransduced carotid arteries of ROSA-26
mice revealed endogenous ß-gal expression at a level
above that produced by AdRSVnlacZ infusion in either CD-1 or C57BL/6
mice (data not shown). Therefore, we constructed a new vector
(containing a different promoter and an intron) that expressed higher
levels of ß-gal: AdCMVnlacZ (see Materials and Methods). Left carotid
arteries in six ROSA26 mice were transduced with AdCMVnlacZ. Four
additional arteries were transduced with AdNull. Three AdCMVnlacZ
arteries and two AdNull arteries were harvested at both 3 and 14 days
after gene transfer (Figure 5A). At 3
days, ß-gal activity in the AdCMVnlacZ arteries was 279 µU/µg
(275 to 353); ß-gal activity in the AdNull arteries was 37.5
µU/µg (21.8 to 53.2; P=0.057 versus AdCMVnlacZ). At 14
days, ß-gal activity in AdCMVnlacZ arteries had fallen by nearly
90%, to 29 µU/µg (25 to 42; P<0.05 versus 3 days).
AdNull arteries harvested at 14 days contained 13 µU/µg (7 to 20;
P=0.057 versus AdCMVnlacZ). Notably, the
endogenous ß-gal levels in the carotids of ROSA26 mice
(as measured in extracts from AdNull-transduced arteries) were
substantially lower at 14 days than at 3 days after gene transfer (13
versus 37.5 µU/µg). This difference might be due to release
of endogenous ß-gal from cells injured at the time of
gene transfer. Because of this substantial difference, we set the
"baseline" ß-gal levels differently at 3 and at 14 days
(Figure 5A). This drop in baseline ß-gal expression suggested
that transgene expression may persist in the ROSA26 mice at 14
days. Nevertheless, there was a large, statistically significant drop
in the level of transgene expression between days 3 and 14.
|
We next tested whether the absence of mature T and B cells (which would
eliminate antigen-specific immune responses to both the transgene and
adenoviral antigens) would permit persistent transgene expression in
carotid arteries. We infused AdRSVnlacZ or AdNull to the carotid
arteries of RAG-2-/- mice (Figure 5B).
At 3 days after gene transfer, ß-gal activity was 14.8 µU/µg (5.9
to 33) in AdRSVnlacZ arteries but was undetectable in AdNull arteries
(P<0.01). At 14 days, ß-gal activity in AdRSVnlacZ
arteries decreased by nearly 90%, to 1.5 µU/µg (0.2 to 1.9;
P<0.01 versus AdRSVnlacZ at 3 days). ß-Gal activity in
AdNull arteries was 0.5 µU/µg (0.4 to 0.7; P<0.01
versus AdRSVnlacZ at 14 days). Thus, in the absence of antigen-specific
immunity, transgene expression persists at 14 days, but at a
significantly lower level.
We considered whether the apparent persistence of low levels of ß-gal
activity in the ROSA-26 and RAG-2-/- mice at 14
days might be due solely to the higher initial levels of ß-gal
activity in these mice (compare 3-day values in Figure 5
[ROSA-26 and RAG-2-/-]) to those in Figure 3 [CD1 and C57BL/6]). To test this hypothesis, we infused the
high-expressing AdCMVnlacZ vector to the carotids of six C57BL/6 mice
(Figure 6). Three days after gene
transfer, ß-gal activity in these arteries was 480 µU/µg (360 to
780). This activity is similar to that found at 3 days in both
Apoe-/- mice and ROSA-26 mice
infused with AdCMVnlacZ (Figure 5A) and far higher than the
3-day values in CD1, C57BL/6, and RAG-2-/-
carotids infused with AdRSVnlacZ (Figures 3 and 5B).
Despite these high initial levels, the ß-gal activity in C57BL/6
carotids transduced with AdCMVnlacZ and harvested at 14 days declined
by more than 97% (11 µU/µg [7 to 12]; P<0.05 versus
3 days) and did not differ from the ß-gal activity in C57BL/6
carotids transduced in parallel with AdNull: (8.5 µU/µg 6.8 to 24;
P=0.66 versus AdCMVnlacZ at 14 days). Thus, the low-level
persistence of ß-gal expression found in
RAG-2-/- mice (and possibly in ROSA26 mice) is
not due solely to the presence of high initial levels of ß-gal
expression.
|
Treatment With 5-Aza Does Not Reactivate Gene
Expression
Carotid arteries of 11 C57BL/6 mice were infused with AdCMVnLacZ
at 1.5x109 pfu/mL. Four arteries were harvested
3 days after gene transfer. The remaining seven arteries were harvested
at 14 days. Four days before harvest, four of the mice were injected
with the demethylating agent 5-Aza. ß-Gal expression in
arterial extracts was high at 3 days, 16 µU/µg (7.6 to
51), but fell to background levels in all 14-day arteries, regardless
of 5-Aza treatment (0.05 µU/µg [0.020 to 0.071] without 5-Aza
versus 0.038 [0.013 to 0.067] with 5-Aza; P=0.4).
Increased Endothelial Turnover in Surgically
Manipulated Arteries
We previously described a transient increase in
endothelial cell proliferation after infusion of either
buffer or adenoviral vector into rat carotid arteries.5 20
Because adenoviral vectors persist as episomes, they could gradually be
lost in replicating cells. To investigate whether increased cell
proliferation induced by the surgical procedure or virus infusion might
contribute to the loss of transgene expression, we infused vehicle or
AdNull into carotid arteries of C57BL/6 mice. Four days later, the mice
were injected with BrdUrd to label proliferating cells, and their
arteries were harvested. As controls, BrdUrd was also infused in two
unoperated mice. As expected,21 there were no
proliferating endothelial cells in the carotid arteries
of the unoperated mice. In contrast, endothelial
proliferation was 16% (3.0% to 30%; n=7) in vehicle-infused arteries
and 19% (14% to 29%; n=4) in arteries infused with AdNull
(P=0.32). Thus, endothelial cell
proliferation is increased substantially by the infusion procedure
alone. This increased proliferation may account for the loss of
transgene expression.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The present study had two principal aims: to determine whether the mouse carotid artery is a suitable substrate for informative gene transfer experiments and to test whether avoidance of antigen-specific immune responses would permit prolonged adenovirus-mediated transgene expression in the artery wall. Our major findings were as follows: (1) adenoviral gene transfer can be accomplished reproducibly in both normal and atherosclerotic mouse arteries; (2) transgene expression declines to background levels by 2 weeks; (3) infusion of adenovirus does not cause vascular injury beyond that caused by the surgical procedure alone; (4) loss of transgene expression is not due solely to antigen-specific immune responses; and (5) the surgical procedure results in increased endothelial turnover, which accompanies the loss of transgene expression.
Despite the technical difficulties involved in isolating, cannulating,
and infusing a mouse carotid artery, gene delivery in the mouse carotid
promises to be a powerful experimental tool. Transgene expression is
reproducible and robust (Figures 5 and 6) with levels of
ß-gal expression equivalent to those we have reported in rat and
rabbit arteries.4 5 6 Because we have obtained biologically
interesting phenotypes after adenoviral gene transfer into rats
and rabbits,20 22 23 it is likely that this expression
level is adequate to manipulate the arterial
phenotype in mice. Indeed, in a small series of experiments in
which mouse carotids were transduced with a vector expressing
transforming-growth factor-ß1
(TGF-ß1), we reproduced the phenotype
of cartilaginous metaplasia that we first reported in
rats20 (R. Agah, unpublished data, 1998). The
availability of genetically engineered mice (both knockout and
transgenic lines) may allow us to identify the molecular mechanisms
through which TGF-ß1 generates this dramatic
phenotype. In addition, gene transfer experiments in the
carotid arteries of Apoe-/- mice
might delineate the local effects of gene products on plaque
progression, regression, or rupture. Again, this mouse model may be
particularly useful because advanced atherosclerotic plaques such as
those in carotids of Apoe-/- mice
are either difficult or impossible to produce in rats, rabbits, and
pigs.
Probably as a result of its size, the mouse carotid artery is
occasionally injured during gene transfer surgery (although this injury
is minimized as an operator gains experience). Importantly, despite
this injury, we observed almost no neointimal
formation in transduced arteries. The absence of neointimal
formation indicates that the mouse carotid model will also be useful
for gene transfer experiments that investigate the genetic basis of
neointimal formation.24 25 26 Such experiments
can be carried out productively, despite the brevity of transgene
expression. Similar investigations in other species have yielded
insights, despite only short-term (
2 weeks) transgene
expression.2 20 27 28
We applied the mouse model to test whether arterial wall transgene expression could be prolonged by avoiding antigen-specific immune responses. Numerous reports, all performed in other organ systems, describe prolongation of adenovirus-mediated transgene expression by strategies that defeat immune surveillance, either by immunosuppression of the host9 29 30 31 32 or by engineering of the vector to decrease production or presentation of foreign antigens.12 33 34 35 The strongest evidence for a role of the immune system in loss of transgene expression comes from studies in liver and skeletal muscle showing prolongation of expression both in immune-deficient mice and in mice that are tolerant of an adenovirus-encoded transgene.7 9 10 Surprisingly, in both the ROSA26 and the RAG-2-/- mice, we detected a steep (>90%) drop in transgene expression between 3 and 14 days. Thus, although in other organ systems immune responses to adenoviral or transgene antigens are both necessary and sufficient to eliminate adenoviral vector–mediated transgene expression,10 36 37 these responses do not play a major role in the near-complete loss of transgene expression after adenoviral gene delivery to mouse carotids.
If the antigen-specific immune system is not the major factor determining loss of arterial wall transgene expression, then what is? Immune system components that function independently of the generation of antigen-specific immunity (such as macrophages) can eliminate adenoviral vectors from the livers of both immunocompetent and immunodeficient mice.38 However, macrophage-mediated elimination of the adenoviral genome occurs primarily in the first 24 hours, too early to explain the loss of expression in our study. We also considered that promoter shutdown, for example due to methylation,39 might play a role in extinction of transgene expression. However, promoter shutdown is an unlikely explanation for two reasons. First, the decline in transgene expression followed a nearly identical time course in mice transduced either with AdRSVnlacZ or AdCMVnlacZ. It is unlikely that the two promoters in these vectors, which are both capable of mediating long-term transgene expression in other settings,15 40 would be shut down so promptly in this model and with precisely the same kinetics. Second, treatment with the demethylating agent 5-Aza, which leads to recovery of in vivo expression from silenced murine retroviral vectors,18 41 did not increase transgene expression.
Our data are most consistent with the notion that transgene expression declines because of cell proliferation combined with vector and cell loss. This is not a novel mechanism for loss of transgene expression. Indeed, the slow decline in transgene expression after adenoviral gene transfer to mouse livers has been attributed, in part, to the gradual replacement of transduced hepatocytes with proliferating, untransduced cells.8 14 42 If low rates of cell proliferation are responsible for the gradual loss of hepatic transgene expression in these studies, then in highly proliferative tissues, such as the carotid endothelium on day 4 after infusion, transgene expression would decline rapidly. Elevated endothelial proliferation is also present after gene transfer to rat5 20 and rabbit arteries (D. Dichek, unpublished data, 1999). Thus, this feature is not a specific limitation of a mouse arterial gene transfer model.
Several plausible mechanisms might be responsible for loss of adenovirus-mediated transgene expression in highly proliferative tissues such as the endothelium of a recently operated artery. First, adenoviral DNA would be diluted as cells (and their chromosomal DNA) replicate whereas the episomal adenoviral DNA does not. Second, the proliferative cells likely include progeny of untransduced cells from areas of the artery adjacent to the site of vector infusion. As these cells proliferate and migrate to cover areas exposed by cell loss, they may replace transduced cells. Third, episomal DNA may be less tightly controlled and therefore less stable than chromosomal DNA during mitosis, leading to loss of vector DNA as transduced cells divide. Fourth and perhaps most importantly, endothelial proliferative rates of 16% per day (see above) in an artery that is not growing rapidly or developing thickened endothelium (neither of which was noted at 14 days) must be accompanied by cell loss. Because endothelial cells spread, proliferate, and migrate rapidly in response to adjacent cell loss,43 it is far more difficult to document cell loss than cell proliferation. Nonetheless, whenever a transduced cell is lost, the adenoviral DNA in the cell is lost as well, and this DNA is not replaced. Confirmation of any of these mechanisms in vivo represents a substantial experimental challenge.
If arterial transgene expression is lost due to increased cell proliferation, will the duration of expression after adenovirus-mediated vascular gene transfer also be limited in other settings? Data from our laboratory and other laboratories suggest that this is likely. In previous studies performed in rat carotids, we detected increased proliferation of both untransduced and transduced cells after infusion of adenoviral vectors.5 20 Proliferation was increased for both endothelial and smooth muscle cells and was unrelated to the presence of a specific transgene. Other groups have reported that vascular cell loss and proliferation are increased simply by surgical manipulation of an artery, in the absence of infusion.44 45 Although these results were all obtained in surgical models, it is probable that catheter-based46 or ex vivo gene transfer techniques47 will also increase vascular cell proliferation as a consequence of trauma to the normally quiescent artery wall. Taken together, these data predict that strategies that circumvent the immune response to adenovirus are unlikely to result in prolonged, stable transgene expression in the artery wall. This prediction appears to be borne out by a recent preliminary report in which infusion of a second generation adenoviral vector in rabbit carotid arteries caused less inflammation but yielded only minimal prolongation of transgene expression.48 Studies in our own rabbit carotid model have yielded similar results (D. Dichek, unpublished data, 1999). It is possible that our previous report of prolonged gene expression in carotid arteries of rats treated with cyclosporin A may reflect a cell-autonomous action of cyclosporin A (ie, decreased apoptosis) that is independent of any effect on the immune system.49 Prolongation of adenovirus-mediated transgene expression in the artery wall will likely require both avoidance of antigen-specific immune responses and circumvention of the consequences of cell proliferation and loss. Use of vectors that achieve chromosomal integration (such as retrovirus or adeno-associated virus) or enhancement of transduced cell survival (for example by more refined surgical techniques or transfer of genes that enhance cell survival) are two means by which this goal might be accomplished.
In summary, experiments performed in a mouse model suggest that nonimmune mechanisms are largely responsible for the early loss of adenovirus-mediated transgene expression. These experiments also set the stage for informative, short-term gene transfer experiments in the arteries of knockout, transgenic, and atherosclerotic animals.
|
Acknowledgments |
|---|
G.V. was supported by the Swiss National Science Foundation, the Swiss Biomedical Foundation, and the Swiss Association against Hypertension. R.A. was supported by a National Research Service Award from the National Institutes of Health. (T32HL07731). This work was supported in part by grant HL60504 and HL61860 (to D. D.). D.D. is an Established Investigator of the American Heart Association. We thank Ruth Linnemann, Dale Newland, Ricky Quan, and Ilsa Sauerwald for technical assistance; John Carroll and Neile Shea for help with graphics; Gary Howard and Stephen Ordway for editorial advice; Dr R. Sahli (University of Lausanne, Switzerland) for AdCMV-EGFP; and Dr Gabor Rubanyi (Berlex Biosciences) for providing the Apoe-/- mice.
Received September 22, 1999; accepted September 23, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. Nabel EG. Gene therapy for cardiovascular disease. Circulation. 1995;91:541–548.
2. Nabel EG, Nabel GJ. Complex models for the study of gene function in cardiovascular biology. Annu Rev Physiol. 1994;56:741–761.[Medline] [Order article via Infotrieve]
3. Dichek DA. Interventional approaches to the introduction of genetic material into the cardiovascular system. In: Topol EJ, ed. Textbook of Interventional Cardiology. 3rd ed. Philadelphia, Pa: WB Saunders; 1999:703–724.
4.
Lee SW, Trapnell BC, Rade JJ, Virmani R, Dichek DA. In
vivo adenoviral vector–mediated gene transfer into balloon-injured rat
carotid arteries. Circ Res. 1993;73:797–807.
5.
Schulick AH, Dong G, Newman KD, Virmani R, Dichek DA.
Endothelium-specific in vivo gene transfer. Circ
Res. 1995;77:475–485.
6. Newman KD, Dunn PF, Owens JW, Schulick AH, Virmani R, Sukhova G, Libby P, Dichek DA. Adenovirus-mediated gene transfer into normal rabbit arteries results in prolonged vascular cell activation, inflammation, and neointimal hyperplasia. J Clin Invest. 1995;96:2955–2965.
7. Yang Y, Ertl HCJ, Wilson JM. MHC class I–restricted cytotoxic T lymphocytes to viral antigens destroy hepatocytes in mice infected with E1-deleted recombinant adenoviruses. Immunity. 1994;1:433–442.[Medline] [Order article via Infotrieve]
8.
Yang Y, Nunes FA, Berencsi K, Furth EE,
Gönczöl E, Wilson JM. Cellular immunity to viral antigens
limits E1-deleted adenoviruses for gene therapy. Proc Natl Acad
Sci U S A. 1994;91:4407–4411.
9.
Dai Y, Schwarz EM, Gu D, Zhang W-W, Sarvetnick N,
Verma IM. Cellular and humoral immune responses to adenoviral vectors
containing factor IX gene: tolerization of factor IX and vector
antigens allows for long-term expression. Proc Natl Acad Sci
U S A. 1995;92:1401–1405.
10. Tripathy SK, Black HB, Goldwasser E, Leiden JM. Immune responses to transgene-encoded proteins limit the stability of gene expression after injection of replication-defective adenovirus vectors. Nat Med. 1996;2:545–550.[Medline] [Order article via Infotrieve]
11.
Yang Y, Su Q, Wilson JM. Role of viral antigens in
destructive cellular immune responses to adenovirus vector-transduced
cells in mouse lungs. J Virol. 1996;70:7209–7212.
12. Yang Y, Nunes FA, Berencsi K, Gönczöl E, Engelhardt JF, Wilson JM. Inactivation of E2a in recombinant adenoviruses improves the prospect for gene therapy in cystic fibrosis. Nat Genet. 1994;7:362–369.[Medline] [Order article via Infotrieve]
13. Wadsworth SC, Zhou H, Smith AE, Kaplan JM. Adenovirus vector-infected cells can escape adenovirus antigen-specific cytotoxic T-lymphocyte killing in vivo. J Virol. 1997;71:5189–5196.[Abstract]
14. Michou AI, Santoro L, Christ M, Julliard V, Pavirani A, Mehtali M. Adenovirus-mediated gene transfer: influence of transgene, mouse strain and type of immune response on persistence of transgene expression. Gene Ther. 1997;4:473–482.[Medline] [Order article via Infotrieve]
15. Schulick AH, Vassalli G, Dunn PF, Dong G, Rade JJ, Zamarron C, Dichek DA. Established immunity precludes adenovirus-mediated gene transfer in rat carotid arteries. Potential for immunosuppression and vector engineering to overcome barriers of immunity. J Clin Invest. 1997;99:209–219.[Medline] [Order article via Infotrieve]
16.
Zambrowicz BP, Imamoto A, Fiering S, Herzenberg LA,
Kerr WG, Soriano P. Disruption of overlapping transcripts in the
ROSA ßgeo 26 gene trap strain leads to widespread expression of
ß-galactosidase in mouse embryos and hematopoietic cells. Proc
Natl Acad Sci U S A. 1997;94:3789–3794.
17. Shinkai Y, Rathbun G, Lam K-P, Oltz EM, Stewart V, Mendelsohn M, Charron J, Datta M, Young F, Stall AM, Alt FW. RAG-2-deficient mice lack mature lymphocytes owing to inability to initiate V(D)J rearrangement. Cell. 1992;68:855–867.[Medline] [Order article via Infotrieve]
18.
Jaenisch R, Schnieke A, Harbers K. Treatment of mice
with 5-azacytidine efficiently activates silent retroviral
genomes in different tissues. Proc Natl Acad Sci U S A. 1985;82:1451–1455.
19. Weiss DJ, Liggitt D, Clark JG. In situ histochemical detection of ß-galactosidase activity in lung: assessment of X-Gal reagent in distinguishing lacZ gene expression and endogenous ß-galactosidase activity. Hum Gene Ther. 1997;8:1545–1554.[Medline] [Order article via Infotrieve]
20.
Schulick AH, Taylor AJ, Zuo W, Qiu C-B, Dong G,
Woodward RN, Agah R, Roberts AB, Virmani R, Dichek DA. Overexpression
of transforming growth factor ß1 in
arterial endothelium causes hyperplasia,
apoptosis, and cartilaginous metaplasia. Proc Natl Acad
Sci U S A. 1998;95:6983–6988.
21.
Schwartz SM, Benditt EP. Aortic
endothelial cell replication. Effects of age and
hypertension in the rat. Circ Res. 1977;41:248–255.
22. Rade JJ, Schulick AH, Dichek DA. Adenovirus-mediated delivery of hirudin to injured rat arteries reduces neointima formation. Circulation. 1995;92:I-501. Abstract.
23. Schneider DB, Vassalli G, Linnemann R, Driscoll RM, Sassani AB, DeYoung MB, Virmani R, Dichek DA. An atherogenic role for Fas ligand. Circulation. 1998;98:I-320. Abstract.
24.
Nabel EG, Shum L, Pompili VJ, Yang Z-Y, San H, Shu HB,
Liptay S, Gold L, Gordon D, Derynck R, Nabel GJ. Direct transfer of
transforming growth factor ß1 gene into
arteries stimulates fibrocellular hyperplasia. Proc Natl Acad Sci
U S A. 1993;90:10759–10763.
25. Nabel EG, Yang Z, Liptay S, San H, Gordon D, Haudenschild CC, Nabel GJ. Recombinant platelet-derived growth factor B gene expression in porcine arteries induces intimal hyperplasia in vivo. J Clin Invest. 1993;91:1822–1829.
26. Nabel EG, Yang Z-Y, Plautz G, Forough R, Zhan X, Haudenschild CC, Maciag T, Nabel GJ. Recombinant fibroblast growth factor-1 promotes intimal hyperplasia and angiogenesis in arteries in vivo. Nature. 1993;362:844–846.[Medline] [Order article via Infotrieve]
27.
Chang MW, Barr E, Seltzer J, Jiang Y-Q, Nabel GJ, Nabel
EG, Parmacek MS, Leiden JM. Cytostatic gene therapy for vascular
proliferative disorders with a constitutively active form of the
retinoblastoma gene product. Science. 1995;267:518–522.
28.
Guzman RJ, Hirschowitz EA, Brody SL, Crystal RG,
Epstein SE, Finkel T. In vivo suppression of injury-induced vascular
smooth muscle cell accumulation using adenovirus-mediated transfer of
the herpes simplex virus thymidine kinase gene. Proc Natl Acad
Sci U S A. 1994;91:10732–10736.
29. Lee MG, Abina MA, Haddada H, Perricaudet M. The constitutive expression of the immunomodulatory gp19 k protein in E1-, E3- adenoviral vectors strongly reduces the host cytotoxic T cell response against the vector. Gene Ther. 1995;2:256–262.[Medline] [Order article via Infotrieve]
30. Kay MA, Holterman A-X, Meuse L, Gown A, Ochs HD, Linsley PS, Wilson CB. Long-term hepatic adenovirus-mediated gene expression in mice following CTLA4Ig administration. Nat Genet. 1995;11:191–197.[Medline] [Order article via Infotrieve]
31. Wilson C, Kay MA. Immunomodulation to enhance gene therapy. Nat Med. 1995;1:887–889.[Medline] [Order article via Infotrieve]
32. Fang B, Eisensmith RC, Wang H, Kay MA, Cross RE, Landen CN, Gordon G, Bellinger DA, Read MS, Hu PC, Brinkhous KM, Woo SLC. Gene therapy for hemophilia B: Host immunosuppression prolongs the therapeutic effect of adenovirus-mediated factor IX expression. Hum Gene Ther. 1995;6:1039–1044.[Medline] [Order article via Infotrieve]
33.
Engelhardt JF, Ye X, Doranz B, Wilson JM. Ablation of
E2A in recombinant adenoviruses improves transgene
persistence and decreases inflammatory response in mouse liver.
Proc Natl Acad Sci U S A. 1994;91:6196–6200.
34.
Ilan Y, Droguett G, Chowdhury NR, Li Y, Sengupta K,
Thummala NR, Davidson A, Chowdhury JR, Horwitz MS. Insertion of the
adenoviral E3 region into a recombinant viral vector prevents antiviral
humoral and cellular immune responses and permits long-term gene
expression. Proc Natl Acad Sci U S A. 1997;94:2587–2592.
35.
Kochanek S, Clemens PR, Mitani K, Chen H-H, Chan S,
Caskey CT. A new adenoviral vector: replacement of all viral coding
sequences with 28 kb of DNA independently expressing both full-length
dystrophin and ß-galactosidase. Proc Natl Acad Sci
U S A. 1996;93:5731–5736.
36. Yang Y, Li Q, Ertl HCJ, Wilson JM. Cellular and humoral immune responses to viral antigens create barriers to lung-directed gene therapy with recombinant adenoviruses. J Virol. 1995;69:2004–2015.[Abstract]
37. Barr D, Tubb J, Ferguson D, Scaria A, Lieber A, Wilson C, Perkins J, Kay MA. Strain related variations in adenovirally mediated transgene expression from mouse hepatocytes in vivo: comparisons between immunocompetent and immunodeficient inbred strains. Gene Ther. 1995;2:151–155.[Medline] [Order article via Infotrieve]
38. Worgall S, Wolff G, Falck-Pedersen E, Crystal RG. Innate immune mechanisms dominate elimination of adenoviral vectors following in vivo administration. Hum Gene Ther. 1997;8:37–44.[Medline] [Order article via Infotrieve]
39. Groudine M, Eisenman R, Weintraub H. Chromatin structure of endogenous retroviral genes and activation by an inhibitor of DNA methylation. Nature. 1981;292:311–317.[Medline] [Order article via Infotrieve]
40. Jooss K, Yang Y, Wilson JM. Cyclophosphamide diminishes inflammation and prolongs transgene expression following delivery of adenoviral vectors to mouse liver and lung. Hum Gene Ther. 1996;7:1555–1566.[Medline] [Order article via Infotrieve]
41. Di Ianni M, Terenzi A, Perruccio K, Ciurnelli R, Lucheroni F, Benedetti R, Martelli MF, Tabilio A. 5-Azacytidine prevents transgene methylation in vivo. Gene Ther. 1999;6:703–707.[Medline] [Order article via Infotrieve]
42.
Kay MA, Landen CH, Rothenberg SR, Taylor LA, Leland F,
Wiehle S, Fang B, Bellinger D, Finegold M, Thompson AR, Read M,
Brinkhous KM, Woo SLC. In vivo hepatic gene therapy: complete albeit
transient correction of factor IX deficiency in hemophilia B dogs.
Proc Natl Acad Sci U S A. 1994;91:2353–2357.
43. Khachigian LM, Lindner V, Williams AJ, Collins T. Egr-1–induced endothelial gene expression: a common theme in vascular injury. Science. 1996;271:1427–1431.[Abstract]
44.
Kockx MM, De Meyer GRY, Andries LJ, Bult H, Jacob WA,
Herman AG. The endothelium during cuff-induced
neointima formation in the rabbit carotid artery.
Arterioscler Thromb. 1993;13:1874–1884.
45.
Kockx MM, De Meyer GRY, Jacob WA, Bult H, Herman AG.
Triphasic sequence of neointimal formation in the cuffed
carotid artery of the rabbit. Arterioscler Thromb. 1992;12:1447–1457.
46.
Willard JE, Landau C, Glamann DB, Burns D, Jessen ME,
Pirwitz MJ, Gerard RD, Meidell RS. Genetic modification of the vessel
wall. Comparison of surgical and catheter-based techniques for delivery
of recombinant adenovirus. Circulation. 1994;89:2190–2197.
47.
Chen S-J, Wilson JM, Muller DWM. Adenovirus-mediated
gene transfer of soluble vascular cell adhesion molecule to porcine
interposition vein grafts. Circulation. 1994;89:1922–1928.
48. Qian H, Wang Q, Channon K, Finer M, McArthur J, George S. An adenovirus with E1 and E4 deletions shows improved performance for vascular gene therapy. Circulation. 1998;98:I-669. Abstract.
49. Hojo M, Morimoto T, Maluccio M, Asano T, Morimoto K, Lagman M, Shimbo T, Suthanthiran M. Cyclosporine induces cancer progression by a cell-autonomous mechanism. Nature. 1999;397:530–534.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  J. Wang, X. Xian, W. Huang, L. Chen, L. Wu, Y. Zhu, J. Fan, C. Ross, M. R. Hayden, and G. Liu Expression of LPL in Endothelial-Intact Artery Results in Lipid Deposition and Vascular Cell Adhesion Molecule-1 Upregulation in Both LPL and ApoE-Deficient Mice Arterioscler Thromb Vasc Biol, January 1, 2007; 27(1): 197 - 203. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Otsuka, R. Agah, A. D. Frutkin, T. N. Wight, and D. A. Dichek Transforming Growth Factor Beta 1 Induces Neointima Formation Through Plasminogen Activator Inhibitor-1-Dependent Pathways Arterioscler Thromb Vasc Biol, April 1, 2006; 26(4): 737 - 743. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Wen, S. Graf, P. G. Massey, and D. A. Dichek Improved Vascular Gene Transfer With a Helper-Dependent Adenoviral Vector Circulation, September 14, 2004; 110(11): 1484 - 1491. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Wen, R. M. Driscoll, D. B. Schneider, and D. A. Dichek Inclusion of the E3 Region in an Adenoviral Vector Decreases Inflammation and Neointima Formation After Arterial Gene Transfer Arterioscler Thromb Vasc Biol, November 1, 2001; 21(11): 1777 - 1782. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||