Prevention of Arterial Thrombosis by Adenovirus-Mediated Transfer of Cyclooxygenase Gene
Background Prostacyclin is an important vasoprotective molecule. It inhibits platelet aggregation, monocyte interaction with endothelium, and smooth muscle cell lipid accumulation. Vascular cyclooxygenase-1 (COX-1) is the rate-limiting step in prostacyclin synthesis. The objective of this study was to determine whether adenovirus-mediated transfer of COX-1 could restore COX-1 activity, augment prostacyclin synthesis, and prevent thrombus formation in a porcine carotid angioplasty model.
Methods and Results Human COX-1 cDNA driven by a cytomegalovirus promoter was constructed into a replication-defective adenovirus 5 vector by homologous recombination. Recombinant adenovirus without a foreign gene (Ad-RR) and buffer were included as controls. Recombinant Ad-LacZ was used for marking the transfected cells in vivo. In the in vitro experiments, cultured human endothelial cells (ECs) and porcine arterial smooth muscle cells (SMCs) were incubated with Ad-COX-1 for 2 hours and 6-keto-PGF1α level and the transgene expression were determined 72 hours after infection. In the in vivo experiments, recombinant adenoviruses were directly instilled into angioplasty-injured porcine carotid arteries for 30 minutes. Cyclic flow changes were monitored for 10 days and thrombus formation was examined histologically thereafter. Transgene expression and prostaglandin I2 (PGI2) synthesis by the injured arteries were determined. Cultured ECs infected with Ad-COX-1 produced a fivefold to eightfold increase in PGI2, and the transgene expression in cultured porcine SMCs was demonstrated by Northern analysis. Direct administration of Ad-COX-1 at a dose of 3×1010 pfu completely inhibited carotid cyclic flow changes and thrombus formation accompanied by a fourfold increase in PGI2 synthesis by the injured arteries 10 days after infection, whereas Ad-COX-1 at a lower dose, 5×109 pfu, had no antithrombotic effects when compared with Ad-RR vector and buffer controls.
Conclusions Adenovirus-mediated transfer of COX-1 to angioplasty-injured carotid arteries was efficacious in augmenting PGI2 synthesis and was associated with an inhibition of thrombosis when a relatively high titer of adenovirus was instilled.
Prostacyclin is a key vasoprotective molecule that is synthesized in vascular cells and acts as an autacoid in inhibiting platelet aggregation and maintaining SMC relaxation.1 2 3 Its biosynthesis is catalyzed in succession by phospholipase A2, which liberates arachidonic acid (AA) from the sn-2 position of membrane phospholipids; cyclooxygenase (also known as prostaglandin H synthase), which converts the liberated AA into PGG2 and then PGH2; and finally prostacyclin synthase, which transforms PGH2 into prostacyclin.4 Injury to the endothelium may cause a reduction in prostacyclin synthesis with an attendant risk of thrombosis. Attempts to treat thrombosis by systemic prostacyclin infusion have been unsatisfactory because of unacceptable complications.5 6 Experimental data have indicated that the constitutive COX-1 or PGH synthase-1 is the key rate-limiting step in controlling the extent of prostacyclin synthesis.7 8 9 Overexpression of COX-1 in a cultured human endothelial cell line by retrovirus-mediated transfer of COX-1 cDNA is accompanied by a sustained increase in prostacyclin synthesis.10 The in vitro gene transfer results suggest that local production of prostacyclin to the vascular injury sites may be augmented by transfer of COX-1. However, retroviral vectors are inefficient for direct gene transfer to the damaged vessel wall.11 By contrast, replication-defective recombinant adenoviruses are efficient vectors for direct in vivo arterial gene transfer.12 In the present study, we constructed a replication-defective recombinant Ad5 viral vector containing a human COX-1 cDNA (Ad-COX-1) insert driven by cytomegalovirus promoter and showed its ability to augment COX-1 expression and PGI2 synthesis in cultured vascular cells and injured porcine carotid arteries. Direct instillation of Ad-COX-1 into angioplasty-injured porcine carotid arteries significantly reduced CF changes and was associated with the inhibition of carotid arterial thrombosis. The antithrombotic effect is achieved only at a relatively high titer of Ad-COX-1.
Preparation of Recombinant Adenovirus
Construction and Preparation of Replication-Defective Recombinant Adenovirus
Ad5/CMV-COX-1 (Ad-COX-1), Ad5/CMV-LacZ (Ad-LacZ), and Ad5 vector control (Ad-RR) were essentially as previously reported.13 14 High-titer stocks of recombinant viruses were prepared by a modification of the method described previously.14 Confluent monolayers of transformed human embryonic kidney (293) cells were infected by exposure to primary recombinant viruses for 1 hour at a multiplicity of infection of 0.01 and incubated in Dulbecco’s modified Eagle’s medium containing 4.5 g/L glucose supplemented with 2% fetal bovine serum until extensive cytopathic effects were observed. Three to 5 days after infection, 293 cells were lysed in isotonic saline buffer (mmol/L: Tris HCl 10, NaCl 137, KCl 4, MgCl2 1, pH 7.4) and 0.1% Nonidet P-40. The lysate was centrifuged at 12 000g, and virus was precipitated from the supernatant by addition of 0.5 vol of 20% polyethylene glycol/2.5 mol/L NaCl. The precipitated virus was resuspended and purified by cesium chloride density centrifugation and sepharose CL4B chromatography. The cells were stored at −80°C in phosphate-buffered saline containing 0.1 mg/mL bovine serum albumin until use. Titers of purified virus preparations were determined by plaque assay on monolayers of 293 cells. Virus preparations were not screened for the presence of wild-type virus, which had not been reported at the time virus was grown for the experiments reported here.
Infection of Cultured Vascular Cells With Ad-COX-1
Cultured human ECs, EA.hy 926,15 or porcine aortic SMCs16 were incubated with fresh medium containing Ad-COX-1 (60 pfu/cell) or vehicle for 2 hours. The cells were washed and incubated in fresh medium for 48 hours. Prostacyclin content in the medium produced by the infected, cultured ECs was measured as 6-keto-PGF1α by a highly sensitive RIA.17 The mRNA expression in the infected porcine SMCs was determined by Northern blot analysis by using a 1.3-kb human COX-1 riboprobe,17 which is highly selective for COX-1 mRNA.7 To control for differences in the efficiency of total RNA extraction or RNA transfer to the membranes, porcine glyceraldehyde 3-phosphate dehydrogenase probed with an [α-32P]-labeled human cDNA was included as an internal control.
Porcine Carotid Crush-Injury and Angioplasty Model
Because of its simplicity, a standardized, crush-injury model18 was initially used to establish whether enhancement of PGI2 synthesis by adenovirus-mediated transfer of COX-1 cDNA in vivo was feasible. Because a single carotid cutdown is used to allow for the creation of crush injury and virus delivery, surgery is rapid and relatively low cost, not requiring x-ray and angioplasty equipment.
After the feasibility of enhancing PGI2 synthesis by COX-1 cDNA transfer was established, an animal model bearing similarity to events taking place during angioplasty in humans was used to study the effects of enhanced PGI2 synthesis on thrombus formation. Pig carotid arteries were injured by an angioplasty procedure modified from that of Steele et al.19 In addition to complete endothelial denudation in all balloon-damaged arteries, deep arterial injury (tear extending beyond the internal elastic membrane) is observed in 50% of damaged arteries in this model.19 In preliminary experiments, a constrictor was applied within the first hour of angioplasty to the center of the injury, often leading to irreversible thrombotic occlusion despite administration of heparin to ACTs of 500 to 800 seconds. However, by 2 hours from injury, cyclic thrombus deposition in the injured artery spontaneously decreased and a constrictor could be applied without promoting irreversible occlusion in the majority of vessels. By 12 hours from injury, only rare cyclic flow variations were observed. To prevent acute thrombosis during surgery in the study, an intravenous bolus of heparin (200 U/kg body wt) was given 5 minutes before angioplasty; this prolonged the ACT to threefold to fivefold higher than the basal ACT. ACT prolongation was comparable in all experimental groups. Bilateral carotid arteries were exposed and 5 mL 1% lidocaine was applied. A 5F-polyethylene balloon catheter was introduced into either common carotid artery via the right femoral artery, and once the catheter was in place under fluoroscopy, the balloon was inflated for 30 seconds ×5 with a 60-second interval between each inflation. Both ends of the injured carotid segments were temporarily ligated, and Ad-COX-1, Ad-RR, or buffer was instilled into the lumen with a 22-gauge Teflon catheter for 30 minutes. Five minutes before removing the instillate and allowing reflow, we administered a second bolus of heparin (100 U/kg). Carotid flow was monitored for 2 hours before application of a constrictor, and thereafter the pigs were allowed to recover. Heparin 50 U · kg−1 · h−1 was given for the first 24 hours to all pigs.
Carotid flow was continuously recorded for 10 days according to a procedure previously described.20 Despite a high degree of heparin anticoagulant activity evidenced by prolonged ACT, severe CF reductions were noted during the first 2 hours in all animals, with frequent zero-flow requiring massage of the artery to dislodge thrombi. The flow became stabilized thereafter. There were no differences in the initial flow reductions among various experimental groups of animals. The flow rate at 24 hours after surgery was used as the postangioplasty baseline for assessment of carotid flow changes. Heparin was discontinued after the initial 24 hours. A “zero” flow was defined as a complete stop of flow for at least 24 hours. When zero-flow was present at the time of euthanasia (one Doppler reading was taken at this time in all pigs), the presence of an occlusive thrombus was confirmed by gross and histological examination in 100% of cases. All the animals that developed zero-flow had persistent zero-flow for the entire period of study except for one control animal in which the flow returned on day 9. This was thought to be due to spontaneous fibrinolysis. The severity of CF variations is arbitrarily classified according to the extent of CF reductions. A reduction of CF >75% was considered severe; 25% to 74%, moderate; and <25%, mild. The frequency of CF changes during the entire monitoring period was compared between Ad-COX-1 and the two control groups. At the end of 10 days, 10 000 U heparin were administered prior to euthanasia with a barbiturate overdose. Animal experiments were approved by the Animal Care Committee.
Common carotid arteries were removed en bloc and postfixed in 10% buffered formalin for ≈72 hours. Sections were paraffin embedded by using standard laboratory procedures and stained with hematoxylin-eosin or Movat’s pentachrome stain. Paraffin sections were evaluated by light microscopy as to the extent of thrombosis and neointimal proliferation. The morphological evaluation was performed in a blinded fashion. For β-galactosidase staining, Ad-LacZ (6×109 pfu/mL) was instilled into the angioplasty-injured artery according to identical procedures as described above. The animals were killed on day 10, and the carotid arterial tissues were prepared and stained with X-Gal by procedures previously reported.21
Prostacyclin Formation in Porcine Carotid Arterial Segments
The injured porcine carotid arterial segments treated with Ad-COX-1, Ad-RR, or buffer were dissected on day 10 after the animals were killed. The arterial segments were cut into rings of approximately equal size. The arterial rings were gently washed and incubated with fresh PBS, pH 7.4, containing 20 μmol/L arachidonic acid at 37°C for 30 minutes. The medium was removed and the content of the stable metabolite of prostacyclin, 6-keto-PGF1α, was measured by a specific and sensitive RIA.
The effect of Ad-COX-1, Ad-RR, and mock saline on carotid blood flow and thrombus deposition after injury was studied in 48 animals. In addition, 3 animals received recombinant adenovirus carrying the LacZ gene (Ad-LacZ) to monitor the duration of gene expression independent of PGI2 expression. Isolated, balloon-injured carotid segments were incubated for 30 minutes with 3.2×1010 pfu of Ad-RR, 5×109 pfu of Ad-COX-1, 3×1010 pfu of Ad-COX-1, or buffer alone. Each viral dose was suspended in about 0.5 to 0.7 mL buffer. This volume was found to completely fill the isolated arterial segments under an instillation pressure of 300 to 500 mm Hg. Two animals assigned to Ad-COX-1 and one control pig were excluded from analysis because of irreversible thrombotic occlusion within the first 6 hours after surgery. In addition, one control and one Ad-COX-1–treated pig tore off the wires connected to the probes on day 7 and 8, respectively, and were excluded from analysis because of lack of an interpretable flow signal and the presence of uncontrolled, added trauma to the artery. CFVs and histological thrombus at death were analyzed in 8 pigs receiving 3×1010 pfu of Ad-COX, 17 receiving 5×109 pfu Ad-COX-1, 5 receiving 3.2×1010 pfu Ad-RR, and 13 pigs treated with buffer as additional control.
Whole-blood aggregation was performed in a Chronolog Lumi-aggregometer (model 560VS) in all pigs. Aggregation was stimulated with type I collagen (2, 5, and 10 μg/mL), the thromboxane A2 analog U46619 (0.5, 1, and 2 μg/mL), thrombin (2.5 and 5 U/mL), and ADP (0.05, 0.1, and 0.2 μmol/L). Except for blood used for aggregation with thrombin and ADP (collected in 1/9 vol of 3.8% citrate), blood was drawn into heparin (final concentration 4 U per milliliter of blood). Platelet-rich plasma was prepared for aggregation with ADP. Platelet counts in control pigs were 324 800±110 366 at baseline and 275 800±143 000 at 10 days. The relative platelet counts were 384 666±95 980 and 363 670±115 590 in COX-1 cDNA–transduced pigs. Hematocrits (baseline and 10-day values) were 28.98±3.36 and 29.68±2.44 in control pigs and 25.02±2.89 and 28.48±2.47 in gene-treated pigs.
ANOVA was used to determine the presence of significant (P<.05) differences in end points between the study groups. Multiple-comparison testing with the Student-Newman-Keuls test was then carried out to isolate the groups producing significant differences.
Augmentation of Prostacyclin Production in Cultured Human Endothelial Cells by Ad-COX-1 Infection
We initially determined whether Ad-COX-1 infection of cultured EA hy 926 human EC line increased PGI2 synthesis. EA hy 926 is a hybrid cell line that possesses morphological and biochemical characteristics of a human EC.15 We have previously shown that retrovirus-mediated transfer of COX-1 into this cell line leads to a sustained overexpression of COX-1.10 Monolayers of confluent EA cells in T-25 flasks were treated with Ad-COX-1 (60 pfu/cell) or Ad-LacZ at 37°C for 2 hours. For comparison, EA cells were incubated in cultured medium alone at 37°C for 2 hours. The cells were washed and stimulated with arachidonate (10 μmol/L), ionophore A23187 (10 μmol/L), or buffer alone at 37°C for 10 minutes. 6-Keto-PGF1α contents in the medium were measured by RIA. The results are shown in Table 1⇓. The baseline levels of 6-keto-PGF1α produced by Ad-COX-1– and Ad-LacZ–transfected cells were not significantly different from those of untransfected cells (Table 1⇓). Arachidonate and ionophore treatments of Ad-COX-1 cells were accompanied by about a fivefold and eightfold increase in 6-keto-PGF1α levels, respectively. By contrast, the 6-keto-PGF1α levels produced by Ad-LacZ cells in response to arachidonate or ionophore were not different from those of untransfected native cells. These results indicate overexpression of COX-1 activity in Ad-COX-1 cells. This effect is not due to a nonspecific adenoviral effect on COX-2 induction, as the Ad-LacZ transfected cells produced essentially identical baseline and stimulated levels of 6-keto-PGF1α as the native untransfected cells.
Expression of COX-1 Transgene in Cultured Porcine SMCs
After denudation by angioplasty, the SMCs are the main cell type to be exposed to the recombinant adenoviruses. To determine whether vascular SMCs took up the recombinant Ad-COX-1, cultured porcine arterial SMCs were treated with Ad-COX-1 (60 pfu/cell) for 2 hours. COX-1 mRNA was determined by Northern blot analysis, using a 1.3-kb riboprobe for COX-1. As shown in Fig 1⇓, uninfected cells exhibited a single 2.7-kb COX-1 band, whereas in addition to this inherent COX-1 mRNA band, the Ad-COX-1–infected SMCs expressed a 3.2-kb band, suggestive of transcription of the 430-bp SV40 polyadenylation sequence present in the adenoviral vector downstream from the COX-1 cDNA.
Effect of Ad-COX-1 Transfer on Thrombus Formation
In preliminary experiments, we used a simple porcine carotid injury model to determine whether instillation of Ad-COX-1 for 30 minutes would elicit an increased PGI2 synthesis. Porcine carotid arteries were injured by modified Spencer-Welles forceps according to a procedure described by Butler et al.18 Ad-COX-1 (1×1010 pfu suspended in 0.5 mL buffer [titer 2×1010 pfu/mL]) or buffer alone was instilled in the injured arterial segments for 30 minutes before to reflow. The animals were euthanitized at 72 hours after infection, and carotid arteries were isolated and dissected into rings. PGI2 synthesis by Ad-COX-1–infected arteries was threefold higher than that of buffer controls.
The effect of direct adenovirus-mediated transfer of COX-1 on thrombus formation was evaluated in a porcine carotid angioplasty model.19 Carotid arterial CFVs that occurred during the 10-day period and thrombus formation on the damaged arterial wall at the end of the 10-day period were compared between pigs treated with 5×109 pfu and 3×1010 pfu Ad-COX-1, pigs treated with buffer alone, and pigs treated with an adenoviral vector without foreign genes (Ad-RR, 3.2×1010 pfu). Five pigs were excluded from analysis (see “Methods”). The incidence of CFVs and histological thrombus at death in the 43 pigs is presented in Table 2⇓. All pigs developing zero-flow during the time of flow monitoring had occlusive carotid thrombi at death. Neither the ACTs during the first 24 hours nor the response of platelets to agonists at baseline and 10 days, including the thromboxane A2 analogue U46619 and thrombin, were significantly different between groups (data not shown).
Without exception, pigs that had received 5×109 pfu Ad-COX-1 developed frequent CFVs, which was not statistically different from the buffer or the Ad-RR control groups. In contrast, no CF reduction or thrombus deposition was found in the eight pigs that had been treated with 3×1010 pfu Ad-COX-1, which retained normal flow throughout the study period from day 2 to 10, whereas only one buffer control pig retained normal flow, and 54% of the buffer control and 40% of the Ad-RR control animals developed zero-flow (Table 2⇑). Thirty percent of pigs receiving 5×109 pfu of Ad-COX-1 developed zero-flow. A representative set of tracings comparing CFVs between a buffer control and a high-titer Ad-COX-1 pig is shown in Fig 2⇓.
Histological examinations of the injured carotid arteries at 10 days after infection confirmed the presence of occlusive thrombi in animals exhibiting zero-flows (Fig 3⇓). Despite a similar extent of vascular injury comparable to control and low-titer Ad-COX-1 groups, thrombi were not detected in any of the eight pigs infected with 3×1010 pfu of Ad-COX-1 (Fig 3⇓). Expression of β-galactosidase in Ad-LacZ–infected carotid arteries was evaluated by staining the arterial wall in situ with X-Gal. Positive staining was noted in small numbers of vascular cells scattered throughout the injured arterial segment 10 days after Ad-LacZ infection (Fig 3⇓).
Prostacyclin Synthesis in Arterial Segments Infected With Recombinant Adenovirus
The level of prostacyclin production by angioplasty-injured arterial rings on day 10 was fourfold to fivefold higher in pigs receiving 3×1010 pfu Ad-COX-1 (137±74 ng/mL) than in those receiving the buffer (32±4.8 ng/mL) or Ad-RR (24±2.7 ng/mL) control. The prostacyclin levels produced in the group infected with 5×109 pfu Ad-COX-1 (35±5 ng/ml) were not different from those of the control groups.
It has recently been shown in the retrovirus-mediated COX-1 overexpressed human endothelial cell line that, despite a short COX-1 half-life of about 10 minutes due to COX-1 autoinactivation during catalysis, not all the enzymes are autoinactivated; only about 30% of the overexpressed COX-1 is inactivated during each catalysis.7 PGI2 accounts for the overwhelming majority of cyclooxygenase products generated by endothelial and vascular SMCs, 95% and 73%, respectively.22 Most of the rest is accounted for by PGE2, PGF2α, and PGD2, while thromboxane A2 is produced in only very small amounts by endothelium and vascular SMCs.22 In agreement with these studies, we found that after retrovirus-mediated transfer of COX-1 into endothelial cells, overexpression of PGI2 accounted for most of the prostanoids elaborated by endothelial cells.7 Thus, overexpression of COX-1 results in a sustained production of a high level of PGI2. These findings lend credence to COX-1 gene transfer. As replication-defective retroviral vectors are inefficient for direct arterial gene transfer, we tested the efficacy of replication-defective Ad-COX-1 transfer in cultured cells and a porcine angioplasty model. Infection of cultured ECs with Ad-COX-1 for a short time period of 2 hours leads to a severalfold increase in PGI2 synthesis at 72 hours after infection. Transgene expression is also evident in porcine SMCs under a similar experimental protocol. Furthermore, direct instillation of Ad-COX-1 into a carotid arterial segment injured with mechanical clamps for 30 minutes increased PGI2 production by about threefold. These results are consistent with several recent reports on direct arterial gene transfer by adenoviral vectors and support the notion that adenoviral vectors are feasible for transferring COX-1 cDNA into damaged arterial wall for augmenting PGI2 synthesis.
Our results indicate that Ad-COX-1 transfer into the angioplasty-injured porcine carotid arterial wall immediately after injury is effective in protecting the injured arteries from developing CF changes and thrombus formation. Thrombus deposition immediately after injury was prevented by administration of high-dose heparin (total of 300 U/kg). In addition, application of the constrictor to the injured artery was delayed by 2 hours after completion of angioplasty and incubation of the injured arteries with virus or control buffer. Preliminary experiments indicated that earlier addition of stenosis to the injury caused irreversible occlusion in the majority of animals. Between 2 and 12 hours after injury, spontaneous thrombus formation in this model gradually abates. Only 3 of 45 pigs developed irreversible carotid occlusion between 2 and 72 hours after angioplasty. In a similar injury model in the dog, the early nadir in the number of CFVs was observed to coincide with degranulation of platelets and their nearly complete refractoriness to several agonists, including the thromboxane analog U46619 (Willerson JT, 1995, unpublished observations). Recurrent thrombus deposition, mirrored by CFVs, returned about 3 days after angioplasty in Ad-RR– and saline-treated carotid arteries. Clearly, the antithrombotic effects of Ad-COX-1 gene transfer in our study did not protect against thrombus deposition immediately after balloon injury, which required high-dose heparin for its prevention. On the other hand, chronic recurrent thrombus deposition plays an important role in the acute and chronic progression of atherothrombotic disease,23 24 25 and the ability to continuously monitor arterial flow after injury has revealed the important role played by recurrent thrombus deposition in the development of chronic stenosis after experimental injury.20 CFVs also were observed in humans after angioplasty and were abolished by platelet antagonists.26 27 In our model, CFVs tended to fade spontaneously by day 10, in accordance with the reported time course of endothelial healing.19 Thus, the duration of CFVs in this model coincides with that of foreign gene expression by recombinant first-generation adenoviruses, making these vectors suitable for testing of antithrombotic gene therapy.
The antithrombotic effect was conferred by a relatively high viral dose of Ad-COX-1 (3×1010 pfu). Lowering the dose to 5×109 pfu resulted in a disappearance of the protective properties against CF changes and thrombus formation. PGI2 synthesis by carotid arteries in the high-titer Ad-COX-1 group was increased over the control groups by fourfold at 10 days after infection, whereas PGI2 synthesis in the low-titer Ad-COX-1 group was not different from controls. Although thromboxane A2 levels were not measured in this study, these findings suggest that the platelet agonist thromboxane A2 was not a major product after COX-1 gene transfer to vascular cells or injured arteries in vivo.
Induction of COX-2 is unlikely to explain our results in vitro and in vivo. First, infection of porcine SMCs in vitro resulted in a 3.2-kb message hybridizing with the COX-1 riboprobe. We speculate, but do not prove, that the 3.2-kb band represents additional transcription of the 430-bp SV40 polyadenylation sequence present in the adenoviral vector downstream from the 2.7-kb COX-1 cDNA. Since there is about 60% homology between the COX-1 and COX-2 messages, we would have expected a hybridization of the riboprobe to the 4.3-kb COX-2 message, had COX-2 been induced.28 No 4.3-kb message was detected in the SMCs. Second, although the inducible COX-2 has been shown to be expressed in balloon-catheter–injured arteries and may contribute to the control level of PGI2 synthesis,29 this level of PGI2 synthesis appears to be inadequate for protecting against thrombosis. As the transgene expression in adenovirus-mediated transfer tends to dissipate 2 to 3 weeks after infection,12 PGI2 production may be much higher during the immediate periods after infection than those on day 10. Sustained elevations of PGI2 production in the high-titer Ad-COX-1 group during the 10-day period are likely to be responsible for the antithrombotic effects. Third, no elevation in PGI2 synthesis in arterial rings stimulated ex vivo or thrombus protection in vivo was noted after infection of the artery with Ad-RR at a dose identical to that of Ad-COX-1.
Taken together, these data suggest that direct adenovirus-mediated transfer of a physiologically relevant gene, COX-1, at a sufficiently high titer is capable of restoring COX-1 expression and PGI2 synthesis and is associated with preventing angioplasty-induced arterial thrombosis in a pig model. Beneficial effects on pulmonary arterial pressure have been shown by liposome-mediated transfer of COX-1.30 If our observations are confirmed in similar and other experimental models in the future, gene transfer of COX-1 may be considered as a possible, future therapeutic modality for human arterial thrombotic disorders. However, there are a number of problems that must be solved before COX-1 gene therapy may be applied to human diseases. For example, adenovirus-mediated gene transfer may be less efficient when applied to atherosclerotic arteries.31 In most human arterial thrombotic diseases, thrombus is formed on atherosclerotic lesions. Hence, a high titer of adenoviruses or other more efficient vectors may be required for transferring the COX-1 gene to achieve an efficacious antithrombotic effect. Unfortunately, adenoviruses at high titers are associated with inflammation due to immune reactions, which have been most prominent after adenoviral gene transfer to liver14 32 and lung.33 We were unable to show that Ad-COX-1 at 6×1010 pfu/mL (a dose of 3×1010 pfu) caused inflammation beyond that associated with the angioplasty in our model by day 10. Only scant, mostly adventitial, mononuclear infiltrates were observed in both virus- and saline control–treated arteries, possibly related to the severe medial balloon injury or external manipulation of the artery during surgery. In agreement with earlier reports,34 35 delivery of recombinant adenovirus to the artery at the doses used in this study seems to be associated with minimal or no inflammation.
Selected Abbreviations and Acronyms
|ACT||=||activated clotting time|
|SMC(s)||=||smooth muscle cell(s)|
Effects of Ad-LacZ
Although one batch of high-titer Ad-LacZ, later found to be contaminated with endotoxin, had shown protection against thrombosis, not associated with an increase in PGI2 synthesis, subsequent experiments with a second, endotoxin-free batch of Ad-LacZ at the original titer (6×1010 pfu/mL) and at a 50% higher and a 50% lower titer (9×1010 and 3×1010 pfu/mL), respectively, failed to show vasoprotection. In these experiments, Ad-LacZ treatment was associated with CFV and thrombus incidence not different from treatment with Ad-RR and buffer control. Further, in vitro experiments with Ad-LacZ failed to show an increase in PGI2 synthesis or expression of inducible nitric oxide synthase. The issue of “vasoprotection” seen with the endotoxin-contaminated batch of high-titer Ad-LacZ initially used was not pursued further.
This work was supported by NIH grants HL-35387, HL-50179, HL-17669, and NS-23327; American Heart Association Grant-in-Aid 92008550; and Texas Advanced Technology program 00366020. We thank C.J. Edgell for providing EA.hy926 cells.
Guest editor for this article was Valentin Fuster, MD, Mt Sinai Medical Center, New York, NY.
- Received October 16, 1995.
- Accepted November 5, 1995.
- Copyright © 1996 by American Heart Association
Vane JR. Prostaglandins and the cardiovascular system. Br Heart J. 1983;49:405-409.
Pomerantz K, Hajjar D. Eicosanoids in regulation of arterial smooth muscle cell phenotype, proliferative capacity, and cholesterol metabolism. Arteriosclerosis. 1989;9:413-429.
Wu KK, Kulmacz RJ, Wang L-H, Loose-Mitchell DS, Tsai A-L. Molecular biology of prostacyclin biosynthesis. In: Rubanyi G, Vane JR, eds. Prostacyclin: New Perspectives for Basic Research and Novel Therapeutic Indications. Amsterdam, Netherlands: Elsevier Science Publishers BV; 1992:11-23.
Sanduja SK, Tsai A-L, Matijevic-Aleksic N, Wu KK. Kinetics of prostacyclin synthesis in a PGHS-1 overexpressed endothelial cell. Am J Physiol. 1994;267:C1459-C1466.
Kent RS, Diedrich SL, Whorton R. Regulation of vascular prostaglandin synthesis by metabolites of arachidonic acid in perfused rabbit aorta. J Clin Invest. 1983;72:455-465.
McIntire TM, Zimmerman GA, Satoh K, Prescott SM. Cultured endothelial cells synthesize both PAF and PGI2 in response to histamine, bradykinin and ATP. J Clin Invest. 1985;76:271-280.
Xu X-M, Ohashi K, Sanduja SK, Ruan K-H, Wang L-H, Wu KK. Enhanced prostacyclin synthesis in endothelial cells by retrovirus-mediated transfer of prostaglandin H synthase cDNA. J Clin Invest. 1993;91:1843-1849.
Flugelman MY, Jaklitsch MT, Newman KD, Casscells SW, Brattauer GL, Dichek DA. Low level in vivo gene transfer into the arterial wall through a perforated balloon catheter. Circulation. 1992;85:1110-1117.
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.
Gomez-Foix AM, Coats WS, Baque S, Alam T, Gerard RD, Newgard CB. Adenovirus-mediated transfer of the muscle glycogen phosphorylase gene into hepatocytes confers altered regulation of glycogen metabolism. J Biol Chem. 1992;267:25129-25134.
Herz J, Gerard RD. Adenovirus-mediated transfer of low density lipoprotein receptor gene acutely accelerates cholesterol clearance in normal mice. Proc Natl Acad Sci U S A. 1993;980:2812-2816.
Edgell CJ, McDonald CC, Graham JB. Permanent cell line expressing human factor VIII-related antigen established by hybridization. Proc Natl Acad Sci U S A. 1983;80:3734-3738.
Ross R. The smooth muscle cell, II: growth of smooth muscle in culture and formation of elastic fibers. J Cell Biol. 1971;50:172-186.
Wu KK, Sanduja R, Ferhanoglu B, Loose-Mitchell DS. Aspirin inhibits interleukin 1-induced prostaglandin H synthase expression in cultured endothelial cells. Proc Natl Acad Sci U S A. 1991;88:2384-2387.
Steele PM, Chesebro JH, Stanson AW, Holmes DR, Dewanjee MK, Badimon K, Fuster V. Balloon angioplasty: natural history of the pathophysiological response to injury in a pig model. Circ Res. 1985;57:105-112.
Willerson JT, Yao GK, McNatt J, Benedict CR, Anderson HV, Golino P, Murphree SS, Buja LM. Frequency and severity of cyclic flow alternations and platelet aggregation predict the severity of neointimal proliferation following experimental coronary stenosis and endothelial injury. Proc Natl Acad Sci U S A. 1991;88:10624-10628.
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.
Chen L, Chester MR, Redwood S, Huang J, Leatham E, Kaski JC. Angiographic stenosis progression and coronary events in patients with ‘stabilized unstable angina.’ Circulation. 1995;91:2319-2324.
Duguid JB. Mural thrombosis in arteries. Br Med Bull. 1955;11:36-38.
Falk E. Unstable angina with fatal outcome: dynamic coronary thrombosis leading to infarction and/or sudden death. Circulation. 1985;71:699-708.
Anderson HV, Kirkeeide RL, Krishnaswami A, Weigelt LA, Revana M, Weisman HF, Willerson JT. Cyclic flow variations after coronary angioplasty in humans: clinical and angiographic characteristics and elimination with 7E3 monoclonal antiplatelet antibody. J Am Coll Cardiol. 1994;23:1031-1037.
Jones DA, Carlton DP, McIntyre TM, Zimmerman GA, Prescott SM. Molecular cloning of human prostaglandin endoperoxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem. 1993;268:9049-9054.
Pritchard KA Jr, O’Banion MK, Miano JM, Vlasic N, Bhatia UG, Young DA, Stemerman MB. Induction of cyclooxygenase-2 in rat vascular smooth muscle cells in vitro and in vivo. J Biol Chem. 1994;269:8504-8509.
Conary JT, Parker RE, Christman BW, Faulks RD, King GA, Meyrick BO, Brigham KL. Protection of rabbit lungs from endotoxin injury by in vivo hyperexpression of the prostaglandin G/H synthase gene. J Clin Invest. 1994;93:1834-1840.
Feldman LJ, Steg PG, Zheng LP, Chen D, Kearney M, McGarr SE, Barry JJ, Dedieu J-F, Perricaudet M, Isner JM. Low-efficiency of percutaneous adenovirus-mediated arterial gene transfer in the atherosclerotic rabbit. J Clin Invest. 1995;95:2662-2671.
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.
Lemarchand P, Jones M, Yamada I, Crystal RG. In vivo gene transfer and expression in normal uninjured blood vessels using replication-deficient recombinant adenovirus vectors. Circ Res. 1993;72:1132-1138.
Schulick AH, Newman KD, Virmani R, Dichek DA. In vivo gene transfer into injured carotid arteries: optimization and evaluation of acute toxicity. Circulation. 1995;91:2407-2414.