In Vivo Gene Transfer Into Injured Carotid Arteries
Optimization and Evaluation of Acute Toxicity
Background Adenoviral vectors are very attractive agents for use in in vivo arterial gene transfer. In a previous study, we demonstrated high-efficiency adenovirus-mediated gene transfer into medial smooth muscle cells of balloon-injured rat carotid arteries. We now further characterize this system by investigating the reproducibility of recombinant gene expression, the presence of acute adenovirus-associated toxicity in the vessel wall, and the optimal virus concentration for transduction.
Methods and Results Balloon-injured rat carotid arteries were incubated with (1) an adenovirus expressing a β-galactosidase gene (Av1LacZ4), (2) a related adenovirus without the recombinant gene (Addl312), or (3) control solutions. Recombinant gene expression was determined 3 days after gene transfer by measurement of β-galactosidase activity in vessel extracts and by counting of smooth muscle cells in microscopic sections that were histochemically stained to detect recombinant β-galactosidase activity. Adenovirus-associated toxicity was assessed in microscopic cross sections by counting of total smooth muscle cell nuclei in the media (to identify cell loss) and characterization of medial cellular infiltrates with histochemical stains for specific inflammatory cells (neutrophils, lymphocytes, macrophages, and monocytes). Maximum recombinant gene expression after incubation with Av1LacZ4 was produced by virus concentrations ranging from 2×1010 to 5×1010 plaque-forming units (pfu)/mL. Surprisingly, use of a higher concentration of Av1LacZ4 virus (1×1011 pfu/mL) resulted in loss of recombinant gene expression. In addition, infusion of either Av1LacZ4 or Addl312 at 1×1011 pfu/mL resulted in statistically significant decreases in medial smooth muscle cell number (53% decrease, P<0.01 for Av1LacZ4; 39% decrease, P<.05 for Addl312) compared with vessels infused with control solution. This decrease in smooth muscle cell number was not present after the infusion of virus at lower concentrations. The number of neutrophils in vessel cross sections was significantly increased (fivefold; P<.05) after infusion of Av1LacZ4 at 1×1011 pfu/mL compared with vessels infused with control solution. Lymphocytes, macrophages, and monocytes were present only in low numbers in all vessel cross sections and were not increased consequent to adenovirus infusion.
Conclusions This model of focal in vivo adenovirus-mediated gene transfer into the media of injured arteries is highly reproducible and allows high-level recombinant gene expression over a fairly narrow range of virus concentrations. Acute adenovirus-associated tissue toxicity, as demonstrated by medial smooth muscle cell loss and neutrophilic infiltrates, places an upper limit on virus concentration and associated recombinant gene expression and suggests the presence of a “window” of virus concentration in which either therapeutic or biological effects of recombinant genes can be studied in the absence of associated acute toxicity.
Focal in vivo arterial gene transfer offers a method of studying the effects of specific genes in vessel wall biology and developing gene therapy protocols directed against localized vascular disorders such as neointimal proliferation or restenosis after angioplasty.1 2 3 4 A useful animal model of focal in vivo arterial gene transfer would allow reproducibly high levels of gene transfer and would be associated with very little tissue toxicity. In this manner, the effects of introduced genes might be studied with minimal interference from nonspecific effects of the gene transfer protocol. In addition, efforts at gene therapy might proceed in the absence of virus-related toxicity.
Adenoviral vectors are very useful tools for the transfer of genetic material into the arterial wall. Several groups have reported efficient in vivo adenovirus-mediated gene transfer into vascular endothelium,5 6 7 neointima,8 and media.9 Reported transduction efficiencies for cells in these vascular compartments range from 30% to 80% after infusion of adenovirus ranging in concentration from 108 to 1010 plaque-forming units (pfu)/mL. In none of these studies, however, was an attempt made to optimize adenovirus-mediated gene transfer while evaluating potential local toxicity. In two of these studies, only one concentration of virus appears to have been used.5 8 In a previous study, we used three concentrations of adenovirus, demonstrating a dose-responsive increase in recombinant gene expression between 108 and 1010 pfu/mL and noting possible toxicity at the highest concentration.9 Higher concentrations of adenovirus were not evaluated. As a result, the relations among higher virus concentration, higher levels of recombinant gene expression, and potential virus-related toxicity remain unexplored.
The potential for virus-related toxicity is suggested by data generated in nonvascular animal models of gene transfer. Acute hepatocellular damage resulted from the injection of high concentrations of adenoviral vectors into mouse bile ducts.10 In another study, intravenous injection into mice at multiplicities of infection (based on an estimate of hepatocyte number) of ≥700 was lethal to most of the animals.11 Animals that survived demonstrated patchy hepatic necrosis 2 days after inoculation. Acute inflammation has also been reported after the instillation of adenoviral vectors into lungs of both nonprimates and primates.12 13 In contrast, local adenovirus-related vascular toxicity has not been studied.
In the present study, we infused recombinant adenovirus into balloon-injured rat carotid arteries and investigated the relation among virus concentration, recombinant gene expression, and acute vessel toxicity. In addition, we determined the reproducibility of high-efficiency medial gene transfer. On the basis of our results, we have defined an optimal virus concentration for in vivo gene transfer in this particular animal model. The results suggest the presence of a relatively narrow window in which recombinant protein expression is maximized and acute toxic effects of adenovirus-mediated gene transfer are minimized.
The construction of the recombinant replication-defective adenovirus used in this study, Av1LacZ4, has been described in detail previously.9 This vector (provided by Dr Bruce Trapnell, Genetic Therapy Incorporated) was derived by homologous recombination between Addl327 (an adenovirus serotype 5 [Ad5] mutant with genomic deletions over much of the E1 and a small segment of the E3 coding regions14 ) and the pAv1LacZ4 plasmid (containing a nuclear targeted Escherichia coli β-galactosidase gene under the control of the Rous sarcoma virus long terminal repeat promoter). Exposure of a cell expressing this β-galactosidase gene to 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) results in a histochemical reaction that stains the nucleus blue. The replication-defective adenovirus strain Addl312, which does not contain a recombinant gene, is also an Ad5 derivative but contains a genomic deletion only in the E1a coding region.15 16 Virus stocks were prepared from lysates of infected 293 cells. The titer of cesium-banded virus stocks (1×1011 pfu/mL for both Av1LacZ4 and Addl312) was determined by plaque titration on 293 cells.
All animal procedures were approved by the Animal Care and Use Committee of the National Heart, Lung, and Blood Institute. Adult male Sprague-Dawley rats (Taconic Farms) weighing 350 to 450 g were used. Rats were anesthetized with intraperitoneal injections of ketamine (Fort Dodge Laboratories, Inc) and xylazine (Miles, Inc) in doses of 100 and 10 mg/kg, respectively. Animals were anticoagulated with intravenous injections of porcine heparin (100 USP U/kg, Abbott Laboratories) 3 minutes before isolation of the carotid segment for infusion. After surgery, animals were allowed water ad libitum but were not allowed food until 8 to 12 hours later.
In Vivo Gene Transfer Into Vessel Segments
In each animal, the left carotid system was surgically exposed, proximal and distal control was obtained, and an arteriotomy was made on the external carotid artery. The left common carotid artery was injured by passage of a 2F Fogarty balloon catheter (Baxter), as described previously.9 The catheter, filled with 0.2 mL air, was passed three times to ensure adequate injury. After balloon injury, a segment of common carotid artery, approximately 1 cm in length, was isolated as described.9 A 24-gauge polytetrafluoroethylene catheter was introduced through the external carotid arteriotomy, and the isolated vessel segment was flushed with M-199 medium (Biofluids).
Aliquots of adenovirus were thawed and used immediately whenever possible. Aliquots not used immediately were placed on wet ice and used within 2 hours of thawing. M-199 medium containing 1 mg/mL rat serum albumin (Sigma Chemical Company) was used as needed to dilute viral stocks. For each rat, 50 μL of adenovirus-containing or control solution was infused into the isolated carotid segment, also through an intra-arterial catheter. Intraluminal pressures, determined using a standard pressure transducer with saline infusion under similar conditions, ranged from 70 to 90 mm Hg (data not shown). After infusion, the solution was allowed to dwell in the vessel segment for 20 minutes, during which the carotid segment remained distended. After the incubation period, the solution was withdrawn, the external carotid artery was ligated, and blood flow was reestablished through the common and internal carotid arteries.
Fixation and Removal of Vessel Segments
Animals were killed 3 days after surgery by an overdose with pentobarbital. None of the animals had clinical evidence of wound infection at the time of death. For experiments involving histological examination of vessels, perfusion fixation was carried out in situ.17 After retrograde cannulation of the abdominal aorta, the arterial tree was cleared of blood by perfusion with normal saline at a pressure of 100 mm Hg. After clearing, certain vessels were perfusion-fixed in situ by infusion of either (1) 2% formaldehyde (Mallinckrodt) and 0.2% glutaraldehyde (Sigma Chemical) in phosphate-buffered saline (pH 7.2) for rats exposed to Av1LacZ4 (a fixative commonly used for X-Gal staining18 ) or (2) 2% paraformaldehyde (Fisher) with 0.1 mol/L cacodylic acid (Baker) (pH 7.4) for rats exposed to Addl312. Fixatives were perfused through the arterial cannula at 100 mm Hg for 5 minutes. All perfusion-fixed vessels were then excised and further immersed in the respective fixative for 2 hours. Other arteries were not perfusion-fixed but were removed and immediately frozen for later measurement of recombinant β-galactosidase activity in tissue extracts, as described previously.9
Evaluation of Recombinant β-Galactosidase Gene Expression
Recombinant β-galactosidase expression was evaluated both by histochemical staining and by measurement of β-galactosidase activity in tissue extracts. For histochemical staining, vessels were washed in phosphate-buffered saline and incubated in X-Gal for 2 hours, as described previously.9 All arteries were then divided into segments approximately 2 mm long, and these segments (three or four per vessel) were embedded in paraffin. Sections 5 μm thick were cut from the embedded segments and counterstained with either nuclear fast red (for Av1LacZ4-exposed vessels) or hematoxylin and eosin (for Addl312-exposed and control vessels). For Av1LacZ4-exposed vessels, both the total number of and the number of X-Gal–stained smooth muscle cells (as identified by spindle-shaped morphology and medial location) were counted in at least four randomly chosen cross sections per vessel. A mean total of the number of transduced cells per 5-μm vessel section was calculated for each artery. A mean percentage of transduced smooth muscle cells per vessel section was also calculated for each artery by dividing the mean number of transduced smooth muscle cells by the total mean number of smooth muscle cells.
Recombinant β-galactosidase activity in vessel extracts was measured after each vessel was minced into small pieces in lysis buffer (200 μL per vessel; 0.2% Triton X-100 and 100 mmol/L potassium phosphate [pH 7.8]). These fragments were homogenized in a Polytron (Brinkman), and aliquots of vessel lysate were assayed for β-galactosidase activity with the substrate 3-(4-methoxyspirol-[1,2-dioxetane-3,2′-tricyclo-[18.104.22.168,7]decan]-4-yl)phenyl-β-d-galactopyranoside (AMPGD, Galacto-Light, Tropix). Light emission was measured using a Monolight 2010 luminometer (Analytical Luminescence Laboratory) and calibrated with a standard curve generated using purified E coli β-galactosidase (specific activity, 300 U/mg; Boehringer Mannheim). The activity assay gave a linear response to levels of β-galactosidase standard ranging from 1.5 to 1500 μU, with light emission at 1.5 μU 6 to 12 times that of background. All samples were assayed in duplicate.
Evaluation of Smooth Muscle Cell Number in Vessel Sections
As part of an examination of potential toxicity of adenoviral vector-mediated gene transfer, we counted the total number of smooth muscle cells present in histological sections taken from control and virus-infused vessels. Vessel sections used for this analysis included those described above (taken from arteries exposed to Av1LacZ4 and also used for the determination of the percentage of transduced smooth muscle cells) as well as identically prepared sections taken from arteries exposed to Addl312 or to control solutions. The total medial smooth muscle cell nuclei were counted in three or four (depending on the number of segments into which the vessel had been divided for paraffin embedding) randomly chosen microscopic cross sections per vessel. These vessel cross sections were spaced a minimum of 50 μm apart to avoid counting the same cells twice. To confirm the identification of medial cells as smooth muscle cells, additional vessel cross sections were cut from the same paraffin blocks used to provide sections for cell counting and were processed for immunohistochemical detection of smooth muscle cell actin. Staining was performed using an antibody to smooth muscle cell actin (Sigma Chemical) at a dilution of 1:10 000. Bound antibody was detected with a secondary antibody conjugated to horseradish peroxidase.
Two groups of control vessels were used for the biochemical and histological analyses described above. The first group was infused with M-199 medium containing 1 mg/mL rat serum albumin, the medium used to dilute the virus stocks. This group served as the control for studies involving the evaluation of recombinant β-galactosidase gene expression, both in tissue extracts and in X-Gal–stained cross sections. An additional group of control vessels was used for studies involving analysis of potential medial smooth muscle cell loss after exposure to either Av1LacZ4 or Addl312. This second series of control vessels were exposed to 10 mmol/L Tris-HCl and 1 mmol/L MgCl2 (pH 7.5) containing 10% glycerol, the vehicle used in storing virus stocks. By controlling for both the virus storage buffer (10 mmol/L Tris-HCl and 1 mmol/L MgCl2, [pH 7.5] containing 10% glycerol) and diluent (M-199 medium containing 1 mg/mL rat serum albumin), we could attribute any differences in the number of smooth muscle cells between virus-exposed and control vessels to exposure to adenovirus.
Histochemical Identification of Inflammatory Medial Cells
Microscopic sections of vessels treated with either Av1LacZ4 or control solutions were processed for histochemical detection of neutrophils and immunohistochemical detection of lymphocytes, macrophages, and monocytes. Only those cells present in the media of vessel cross sections were counted. Neutrophils were identified by morphology and by histochemical staining using naphtol AS-D chloroacetate esterase (Leder stain),19 20 a protocol in which a chromogen turns neutrophilic myeloid cells and tissue mast cells red. Lymphocyte (T and B cells) staining was performed using a mixture of antibodies: (1) anti-rat CD3 (Pharmigen; CD3 is expressed on rat T lymphocytes21 ) and (2) anti-rat CD45RC (Pharmigen; CD45RC is expressed on pre–B lymphocytes, B cells, and a subset of T lymphocytes22 ). These antibodies were used at final dilutions of 1:500 for anti-rat CD3 and 1:1000 for anti-rat CD45RC. Specific monocyte and macrophage staining was performed using another antibody, ED1 (Harlan), which recognizes rat macrophages, monocytes, and dendritic cells.23 The ED1 antibody was used at a final dilution of 1:1600. Bound antibodies were detected with a secondary antibody conjugated to horseradish peroxidase. After peroxidase staining, sections were counterstained with hematoxylin. Sections of rat spleen were used as positive controls both for the antibodies and the Leder stain. The number of neutrophils (Leder stain positive), lymphocytes (CD3+ CD45RC-positive cells), and monocytes and macrophages (ED1-positive cells) were counted in three or four randomly selected cross sections per vessel, spaced at least 50 μm apart.
For each group of vessels, a mean number was calculated of the specific cell types per histological cross section for that group (ie, total number of smooth muscle cells, smooth muscle cells with blue nuclei, neutrophils, lymphocytes, or monocytes/macrophages per section). Group mean numbers of vessels receiving undiluted virus (1×1011 pfu/mL) and vessels receiving virus storage buffer control (10 mmol/L Tris-HCl and 1 mmol/L MgCl2 [pH 7.5] containing 10% glycerol) were compared using Student’s t test. Group mean numbers of vessels receiving virus at all concentrations lower than 1×1011 pfu/mL, which from dilution contained varying proportions of virus storage buffer and diluent (M-199 medium containing 1 mg/mL rat serum albumin), were compared with mean numbers of both control groups (control vessels infused with virus storage buffer and control vessels infused with diluent only) using one-way ANOVA.24
In Vivo Gene Transfer and Expression of β-Galactosidase
Thirty-two rat carotid arteries were injured and exposed to Av1LacZ4. Thirteen were used in β-galactosidase activity studies. Nineteen were stained with X-Gal and used for histological evaluation (see below). The arteries used for β-galactosidase activity studies were exposed to Av1LacZ4 at concentrations of 1×1011 pfu/mL (5×109 pfu per vessel segment; n=2), 5×1010 pfu/mL (2.5×109 pfu per vessel segment; n=3), 3×1010 pfu/mL (1.5×109 pfu per vessel segment; n=3), 2×1010 pfu/mL (1×109 pfu per vessel segment; n=2), and 1×1010 pfu/mL (5×108 pfu per vessel segment; n=3). Controls (n=2) received M-199 medium containing 1 mg/mL rat serum albumin. Three days after transduction, arteries were harvested and β-galactosidase activity was measured from tissue extracts (Fig 1⇓). Extracts of vessels exposed to 1×1010 pfu/mL had mean β-galactosidase activity of 5.6 μU/μg total vessel protein (range, 1.4 to 10.2 μU/μg). This value was only slightly higher than that of the control vessels (1.6 μU/μg total vessel protein; range, 1.2 to 2.1 μU/μg). Exposure of vessels to higher concentrations of Av1LacZ4 in the range of 2×1010 pfu/mL to 5×1010 pfu/mL resulted in an increase in β-galactosidase activity measured in tissue extracts. Mean β-galactosidase activity was 46.9 μU/μg total vessel protein at 2×1010 pfu/mL (range, 37.7 to 56.0 μU/μg), 47.4 μU/μg total vessel protein at 3×1010 pfu/mL (range, 38.1 to 64.1 μU/μg), and 31.3 μU/μg total vessel protein at 5×1010 pfu/mL (range, 12.0 to 62.3 μU/μg). Exposure to Av1LacZ4 at a concentration of 1×1011 pfu/mL resulted in a dramatic decrease in β-galactosidase activity in tissue extracts (1.0 μU/μg total vessel protein; range, 0.5 to 1.5 μU/μg), essentially equal to control values.
The 19 rat carotid arteries used for histological evaluation were exposed to Av1LacZ4 at concentrations of 1×1011 pfu/mL (5×109 pfu per vessel segment; n=5), 5×1010 pfu/mL (2.5×109 pfu per vessel segment; n=4), 3×1010 pfu/mL (1.5× 109 pfu per vessel segment; n=4), 2×1010 pfu/mL (1×109 pfu per vessel segment; n=3), and 1×1010 pfu/mL (5×108 pfu per vessel segment; n=3). Controls (n=4) received M-199 medium containing 1 mg/mL rat serum albumin. Three days after transduction, arteries were harvested and stained with X-Gal. Blue-stained cells were limited to the media of the arterial wall and were identified as smooth muscle cells by immunohistochemistry (data not shown). The number of transduced smooth muscle cells was determined by counting X-Gal–stained cells in cross sections (Fig 2⇓). Control vessels showed no X-Gal–stained cells; vessels exposed to 1×1010 pfu/mL Av1LacZ4 showed only a few stained cells. Exposure of vessels to concentrations of Av1LacZ4 in the range of 2×1010 pfu/mL to 5×1010 pfu/mL resulted in an increase in the number of X-Gal–stained smooth muscle cells per cross section. The mean number of X-Gal–stained smooth muscle cells at 2×1010 pfu/mL was 21 (range, 13 to 26 stained cells per section). At 3×1010 pfu/mL, 68 X-Gal–stained cells per section were present (22% of total medial smooth muscle cells; range, 45 to 91 stained cells per section); at 5×1010 pfu/mL, 28 X-Gal–stained cells per section were present (range, 0 to 55 stained cells per section). Exposure to Av1LacZ4 at a concentration of 1×1011 pfu/mL resulted in a dramatic loss of X-Gal staining in vessel cross sections; no X-Gal–stained cells were present in any section taken from five vessels infused with 1×1011 pfu/mL of Av1LacZ4. The results of this histological analysis therefore closely paralleled those obtained from the β-galactosidase activity studies (Fig 1⇑).
Morphology After Transduction With Av1LacZ4 and Addl312
Sections of balloon-injured rat carotid arteries infused with either Av1LacZ4 (n=19), Addl312 (n=22), or control solutions (n=10) were examined to determine the effect of virus concentration on vessel wall morphology. Av1LacZ4 was infused at concentrations ranging from 1×1010 to 1×1011 pfu/mL (5×108 to 5×109 pfu per vessel segment), as described above. Addl312 was infused at concentrations ranging from 1×109 to 1×1011 pfu/mL (5×107 to 5×109 pfu per vessel segment). All arteries were examined 3 days after injury and infusion. Microscopic examination of all vessels, including controls, revealed focal necrosis and intramural hemorrhage, as we have reported previously.9 This injury is most likely a result of the balloon denudation protocol. Areas of more significant medial injury (in which cellular nuclei were totally absent in large areas of histological sections) appeared more prominent in vessels exposed to undiluted stocks of both Av1LacZ4 and Addl312 (Fig 3A⇓ through 3C). The observation that medial smooth muscle cell loss appeared to be increased at higher virus concentrations was investigated quantitatively by counting smooth muscle cell nuclei in the stained tissue sections.
Compared with results obtained from vessels exposed to control solutions, the medial smooth muscle cell number was significantly decreased to 47% of control values (109±40 versus 231±63 cells per section; P<.01) consequent to exposure to Av1LacZ4 at 1×1011 pfu/mL (Fig 4A⇓). Results obtained with Addl312 were similar to those obtained with Av1LacZ4 (Fig 4B⇓). Again, smooth muscle cell number was significantly decreased to 61% of control values in vessels exposed to 1×1011 pfu/mL (141±72 versus 231±63 cells per section; P<.05). Infusion of either Av1LacZ4 or Addl312 at concentrations of less than 1×1011 pfu/mL did not result in significant changes in medial smooth muscle cell number compared with control solution–infused vessels (P=.21 and P=.25, respectively, by one-way ANOVA).
Identification of Inflammatory Cells in the Media of Arteries After Transduction With Av1LacZ4
Arteries were injured and exposed to Av1LacZ4 or control solutions, as described (see “Methods”). Cellular infiltrates in the media were noted in cross sections of both virus-exposed and control vessels. Infiltrates were not uniformly present, and their size varied greatly among vessels and among cross sections from the same vessel. The cells constituting the infiltrates were not medial smooth muscle cells as they failed to stain with a specific immunohistochemical stain against smooth muscle cell actin (data not shown). We further identified these cellular infiltrates by histochemical staining for the presence of specific inflammatory cell types.
Neutrophils in vessel cross sections were detected by Leder staining (see “Methods”; Fig 3D⇑). Approximately 5 to 10 neutrophils per histological section were present after exposure to control solutions (Fig 5⇓). Av1LacZ4 delivered at 1×1010 pfu/mL resulted in a similar number of neutrophils per histological section. Av1LacZ4 delivered over the range of 2×1010 to 1×1011 pfu/mL resulted in an increase in the number of neutrophils, peaking at 20 to 25 per histological section at the highest concentrations. Compared with results obtained from the control vessels exposed to 10 mmol/L Tris-HCl and 1 mmol/L MgCl2 (pH 7.5) containing 10% glycerol (virus storage buffer), the number of Leder stain–positive cells was significantly greater (26±17 versus 5.7±6.1 positive cells per section; P<.05) as a result of exposure to Av1LacZ4 at 1×1011 pfu/mL. Infusion of Av1LacZ4 at concentrations of less than 1×1011 pfu/mL did not result in statistically significant increases in the number of Leder stain–positive cells in histological sections compared with control vessels (P=.24 by one-way ANOVA). Nevertheless, at virus concentrations of 3×1010 and 5×1010 pfu/mL, there was a notable increase in Leder stain–positive cells per histological section: fourfold over control values (22±26 and 21±17, respectively).
T and B lymphocytes as well as macrophages and monocytes were detected within the vascular media by the use of specific immunohistochemical stains (see “Methods”). An average of fewer than two macrophages and monocytes were present per histological section in either vector-exposed or control vessels (maximum of four in any individual cross section; data not shown). T and B lymphocytes were absent from virtually all cross sections (maximum of one in any individual cross section; data not shown). The low-level staining for T and B lymphocytes as well as macrophages and monocytes in vessel sections was independent of Av1LacZ4 concentration and occurred in the presence of markedly positive staining in control sections of rat spleen (data not shown).
The present study confirms that adenoviral vectors can be used to achieve both high-efficiency gene transfer and high levels of recombinant gene expression in the media of balloon-injured rat carotid arteries. Three days after transduction with Av1LacZ4 at 3×1010 pfu/mL (the concentration of virus producing maximal recombinant gene expression), 22.0±8.4% of medial smooth muscle cells stained positively with X-Gal, and 47.5±14.4 μU of β-galactosidase activity per microgram of total vessel protein was detected in vessel extracts. These results are similar to those previously reported by our group (30% X-Gal positivity in the media; 100 μU β-galactosidase activity per microgram of total vessel protein9 ) and contrast with another report8 that suggested that the media of the injured rat carotid artery is amenable to only very low levels of in vivo adenovirus-mediated gene transfer and expression. At the optimal vector concentration of 3×1010 pfu/mL, the quantitative results of both X-Gal staining and β-galactosidase expression in the present study varied by no more than a factor of 2 between individual vessels with the most and least evidence of gene transfer. Therefore, this in vivo gene transfer system is both qualitatively and quantitatively reproducible.
The loss of recombinant β-galactosidase gene expression (as measured by both X-Gal staining and β-galactosidase activity) after exposure to undiluted Av1LacZ4 at 1×1011 pfu/mL was an unanticipated finding. This loss of expression contrasted sharply with expression levels obtained at only modestly lower concentrations of virus (2×1010 to 5×1010 pfu/mL) and was associated with a statistically significant loss of medial smooth muscle cell nuclei compared with controls. Exposure to Addl312 at 1×1011 pfu/mL also resulted in a significant decrease in medial smooth muscle cell number. Control experiments demonstrated that neither the loss of gene expression nor the loss of medial smooth muscle cells was caused by exposure to the virus storage buffer (10 mmol/L Tris-HCl and 1 mmol/L MgCl2 [pH 7.5] containing 10% glycerol). These data suggest that the decrease in β-galactosidase gene expression after exposure to Av1LacZ4 at 1×1011 pfu/mL, compared with exposure to concentrations of Av1LacZ4 ranging from 2×1010 to 5×1010 pfu/mL, is caused either by complete absence of transduction or by the death of any smooth muscle cells that are transduced at these high concentrations of virus. Because significant medial cell death also occurred after exposure to Addl312 at 1×1011 pfu/mL, toxicity is attributable to the adenovirus and not to expression of the recombinant β-galactosidase transgene.
Compared with virus delivered at 3×1010 pfu/mL, exposure of vessels to Av1LacZ4 at 1×1011 pfu/mL resulted in a nearly 50-fold decrease in recombinant β-galactosidase activity (and complete absence of X-Gal positivity in vessel sections) but only a 2-fold decrease in medial smooth muscle cell number. This apparent discrepancy arises from the fact that although transduced cells make up only a small proportion of the total number of medial cells (20% to 30% after exposure to Av1LacZ4 at 3×1010 pfu/mL in the present study), they harbor essentially 100% of the total vessel β-galactosidase activity (there is a low level of background endogenous β-galactosidase activity in the vessel extracts). Therefore, at a viral concentration of 3×1010 pfu/mL, loss of all potentially transduced cells would decrease cell number by only 20% to 30%, whereas gene expression would decrease nearly 100%, to background levels. With Av1LacZ4 at 1×1011 pfu/mL, loss of potentially transduced cells would be expected to be even greater, consistent with the 53% loss reported in the present study; loss of β-galactosidase expression would again be nearly 100%.
Acute toxic effects similar to those described have not been identified in several reports on adenovirus-mediated vascular gene transfer.5 6 7 8 25 A recent review of vascular gene transfer characterized adenovirus-associated vascular toxicity as “unknown.”4 Our findings that high virus concentrations per se are toxic to the arterial wall, however, are consistent with data generated in other organ systems suggesting that exposure to highly concentrated adenovirus does result in acute cytopathic effects in vivo.10 11 12 13 26 It appears from these data that a compromise must be achieved between attempts to maximize adenovirus-mediated gene expression and attempts to minimize acute toxicity.
In characterizing the cellular infiltrates present in the vessel media, we found the number of neutrophils in cross sections was significantly increased compared with control (fivefold; P<.05) after infusion of Av1LacZ4 at 1×1011 pfu/mL. Increases of similar magnitude, but falling short of statistical significance, were present after exposure to Av1LacZ4 at 3×1010 pfu/mL and 5×1010 pfu/mL and may indicate a lower level of cytopathology at these virus concentrations. Of particular interest was the demonstration that the cellular infiltrates in the vascular media were composed almost exclusively of neutrophils and were distinctly devoid of macrophages, monocytes, and lymphocytes. The timing of medial smooth muscle cell loss after transduction (3 days), the neutrophilic character of medial inflammatory infiltrates, and the near-complete absence of lymphocytes from the transduced vascular media all support the hypothesis that smooth muscle cell destruction results from the acute cytopathic effects of highly concentrated adenovirus and is not due to the development of cellular immunity to viral antigens.10 Additional evidence as to whether the immune system is involved in acute cell loss might be obtained from analysis of adenovirus-transduced vessels in immune-deficient animals.
In addition to acute vector-related toxicity, Yang et al10 reported delayed (1 to 2 weeks) cytotoxic effects after adenovirus-mediated hepatic gene transfer. This delayed toxicity was associated with a systemic cellular immune response to viral antigens expressed by the transduced cells, resulting in the disappearance of recombinant gene expression via the specific elimination of transduced cells and their replacement by untransduced hepatocytes.10 We have described a similar time course of disappearance of recombinant gene expression in adenoviral vascular gene transfer systems,9 27 and it is probable that a similar immune-mediated mechanism is operative. Efforts to circumvent the immune system to obtain prolonged recombinant gene expression after adenovirus-mediated gene transfer are under way at other laboratories and have resulted in significant progress.10 The goal of the present study, however, was not to obtain prolonged gene expression but rather to maximize short-term expression. Short-term recombinant gene expression may be adequate both for biological studies aimed at the elucidation of the role of specific genes in the arterial wall and for therapeutic studies in disease processes that are confined to relatively short time periods. Restenosis after angioplasty may be an example of a disease in which a short course of therapy will prove sufficient.28
In conclusion, the rat carotid injury model of focal in vivo adenovirus-mediated gene transfer is highly reproducible but can be associated with acute virus-mediated toxicity. A range of virus concentrations giving high-level gene expression with minimal acute toxicity can be identified and exploited for the Av1LacZ4 virus. It is probable that similar results will be obtained with adenoviruses expressing other genes, although the specific window in which biological or therapeutic effects can be studied in the absence of acute tissue pathology may depend on the specific gene that is expressed. The optimal concentration of each specific adenovirus will likely require empirical testing and will depend on the level of gene expression required to investigate the effect of interest, the inherent toxicity (if any) of the recombinant protein, and the virus-mediated toxicity described in the present study.
Andrew H. Schulick, MD, is generously funded by the Pharmacology Research Associates Training program of the National Institute of General Medical Sciences, National Institutes of Health. We thank Ms Eleonora Dorfman for help in the preparation of viral stocks.
- Received October 10, 1994.
- Revision received October 31, 1994.
- Accepted November 13, 1994.
- Copyright © 1995 by American Heart Association
Swain JL. Gene therapy: a new approach to the treatment of cardiovascular disease. Circulation. 1989;80:1495-1496.
Meidell RS, Gerard RD, Williams RS. The end of the beginning: gene transfer into the vessel wall. Circulation. 1992;85:1219.
Dichek DA. Interventional approaches to the introduction of genetic material into the vasculature. In: Topol EJ, ed. Textbook of Interventional Cardiology. Philadelphia, Pa: WB Saunders; 1994:989-1005.
Nabel EG, Pompili VJ, Plautz GE, Nabel GJ. Gene transfer and vascular disease. Cardiovasc Res. 1994;28:445-455.
Lemarchand P, Jaffe HA, Danel C, Cid MC, Kleinman HK, Stratford-Perricaudet LD, Perricaudet M, Pavirani A, Lecocq JP, Crystal RG. Adenovirus-mediated transfer of a recombinant human alpha-1 antitrypsin cDNA to human endothelial cells. Proc Natl Acad Sci U S A. 1992;89:6482-6486.
Rome JJ, Shayani V, Flugelman MY, Newman KD, Farb A, Virmani R, 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. 1994;14:148-161.
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.
Guzman RJ, Lemarchand P, Crystal RG, Epstein SE, Finkel T. Efficient and selective adenovirus-mediated gene transfer into vascular neointima. Circulation. 1993;88:2838-2848.
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.
Yang Y, Nunes FA, Berencsi K, Furth EE, Gonczol 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.
Ginsberg HA. The Adenoviruses. New York, NY: Plenum Press; 1984.
Shenk T, Williams J. Genetic analysis of adenoviruses. Curr Top Microbiol Immunol. 1984;111:1-38.
Maloney WC, McPherson K, Fliegelman L. Esterase activity in leukocytes demonstrated by the use of naphtol AS-D chloracetate substrate. J Histochem Cytochem. 1960;8:200-206.
Leder LD. Uber die selective fermenstoffchemische dartstellung von neutrophilen myeloischen zellen und gewebstrastzellen in paraffininschnitt. Klin Wochenschr. 1964;42:533.
De Jonge MWA, De Leij L, Mesander G, Deenen GJ, Opstelten D, Kroese FGM, van Imhoff GW, Nieuwenhuis P, The TH. LAMA tumor in the rat as an experimental model for pre-B-cell leukemia. Cancer Res. 1994;49:5682-5688.
Rosner B. Fundamentals of Biostatistics. Boston, Mass: Duxbury Press; 1982:410-435.
Chen S, Wilson JM, Muller DWM. Adenovirus-mediated gene transfer of soluble vascular cell adhesion molecule to porcine interposition vein grafts. Circulation. 1994;89:1922-1928.