Adventitial versus intimal liposome-mediated ex vivo transfection of canine saphenous vein grafts with endothelial nitric oxide synthase gene

Gene Therapy for the Extension of Vein Graft Patency: A Review

The mainstay of treatment for long-segment small-vessel chronic occlusive disease not amenable to endovascular intervention remains surgical bypass grafting using autologous vein. The procedure is largely successful and the immediate operative results almost always favorable. However, the lifespan of a given vein graft is highly variable, and less than 50% will remain primarily patent after 5 years. The slow process of graft malfunction is a result of the vein's chronic maladaptive response to the systemic arterial environment, its primary component being the uncontrolled proliferation of vascular smooth muscle cells (SMCs). It has recently been suggested that this response might be attenuated through pre-implantation genetic modification of the vein, so-called gene therapy for the extension of vein graft patency. Gene therapy seems particularly well suited for the prevention or postponement of vein graft failure since: (1) the stimulation of SMC proliferation appears to largely be an early and transient process, matching the kinetics of current gene transfer technology; (2) most veins are relatively normal and free of disease at the time of bypass allowing for effective gene transfer using a variety of systems; and (3) the target tissue is directly accessible during operation because manipulation and irrigation of the vein is part of the normal workflow of the surgical procedure. This review briefly summarizes the current knowledge of the incidence and basic mechanisms of vein graft failure, the vector systems and molecular targets that have been proposed as possible pre-treatments, the results of experimental genetic modification of vein grafts, and the few available clinical studies of gene therapy for vascular proliferative disorders.

Nanoparticle drug- and gene-eluting stents for the prevention and treatment of coronary restenosis

Percutaneous coronary intervention (PCI) has become the most common revascularization procedure for coronary artery disease. The use of stents has reduced the rate of restenosis by preventing elastic recoil and negative remodeling. However, in-stent restenosis remains one of the major drawbacks of this procedure. Drug-eluting stents (DESs) have proven to be effective in reducing the risk of late restenosis, but the use of currently marketed DESs presents safety concerns, including the non-specificity of therapeutics, incomplete endothelialization leading to late thrombosis, the need for long-term anti-platelet agents, and local hypersensitivity to polymer delivery matrices. In addition, the current DESs lack the capacity for adjustment of the drug dose and release kinetics appropriate to the disease status of the treated vessel. The development of efficacious therapeutic strategies to prevent and inhibit restenosis after PCI is critical for the treatment of coronary artery disease. The administration of drugs using biodegradable polymer nanoparticles as carriers has generated immense interest due to their excellent biocompatibility and ability to facilitate prolonged drug release. Despite the potential benefits of nanoparticles as smart drug delivery and diagnostic systems, much research is still required to evaluate potential toxicity issues related to the chemical properties of nanoparticle materials, as well as to their size and shape. This review describes the molecular mechanism of coronary restenosis, the use of DESs, and progress in nanoparticle drug- or gene-eluting stents for the prevention and treatment of coronary restenosis.

'No-touch' saphenous vein harvesting improves graft performance in patients undergoing coronary artery bypass surgery: a journey from bedside to bench

The saphenous vein is the most commonly used conduit in patients undergoing coronary artery bypass surgery yet its patency is inferior to the internal thoracic artery. Vascular damage inflicted to the vein when using conventional harvesting techniques affects its structure. Endothelial denudation is associated with early vein graft failure while damage of the outermost vessel layers has adverse long-term effects on graft performance. While many in vitro and in vivo experimental studies aimed at improving vein graft patency have been performed to date no significant 'bench to bedside' advances have been made. Among experimental strategies employed is the use of pharmacological agents, gene targeting and external stents. A 'no-touch' technique, where the saphenous vein is removed with minimal trauma and normal architecture preserved, produces a superior graft with long term patency comparable to the internal thoracic artery. Interestingly, many experimental studies are aimed at repairing or replacing those regions of the saphenous vein damaged when harvesting conventionally. 'No-touch' harvesting is superior in coronary artery bypass patients with long-term data published 5years ago. Here we describe a 'bedside to bench' situation where the mechanisms underlying the improved performance of 'no touch' saphenous vein grafts in patients have been studied in the laboratory.

The vasa vasorum and associated endothelial nitric oxide synthase is more important for saphenous vein than arterial bypass grafts

No-touch (NT) saphenous vein (SV) grafts are superior to SVs harvested by the conventional technique (CT), with a patency comparable with the internal thoracic artery (ITA). Preservation of the vasa vasorum is implicated in the success of NT harvesting. We compared the vasa vasorum and endothelial nitric oxide synthase (eNOS) in NT SV with ITA and radial artery (RA) grafts. Skeletonized SV (SSV) was also analyzed. The NT SV had a higher number and larger vasa vasorum compared with ITA (P = .0001) and RA (P = .0004) that correlated with eNOS protein. Activity of eNOS in SSV grafts was significantly lower than NT SV grafts (P = 004). Since a high proportion of the vasa vasorum are removed in SSV using the CT, we suggest that preservation of the vasa vasorum and eNOS-derived NO contributes to the high patency for NT as compared with SSV grafts.

Non-viral eNOS gene delivery and transfection with stents for the treatment of restenosis

BACKGROUND

In this study, we have examined local non-viral gene delivery, transfection, and therapeutic efficacy of endothelial nitric oxide synthase (eNOS) encoding plasmid DNA administered using coated stents in a rabbit iliac artery restenosis model.

METHODS

Lipopolyplexes (LPPs) with eNOS expressing plasmid DNA were immobilized on stainless steel stents using poly(D,L-lactide-co-glycolide) (PLGA) and type B gelatin coatings. The gene-eluting stents were implanted bilaterally in the denuded iliac arteries and eNOS transfection and therapeutic efficacy were examined 14 days after implantation.

RESULTS

The results show that non-viral lipopolyplex-coated stents can efficiently tranfect eNOS locally in the arterial lumen assessed by PCR and ELISA. Human eNOS ELISA levels were significantly raised 24 hours after transfection compared to controls (125 pg eNOS compared to <50 pg for all controls including naked DNA). Local eNOS production suppressed smooth muscle cell proliferation and promoted re-endothelialization of the artery showing a significant reduction in restenosis of 1.75 neointima/media ratio for stents with lipoplexes encoding eNOS compared with 2.3 neointima/media ratio for stents with lipoplexes encosing an empty vector.

CONCLUSIONS

These results support the hypothesis that a potent non-viral gene vector encoding for eNOS coated onto a stent can inhibit restenosis through inhibition of smooth muscle cell growth and promotion of a healthy endothelium.

Retaining perivascular tissue of human saphenous vein grafts protects against surgical and distension-induced damage and preserves endothelial nitric oxide synthase and nitric oxide synthase activity

OBJECTIVE

Conventional harvesting of saphenous vein used for coronary artery bypass surgery induces a vasospasm that is overcome by high-pressure distension. Saphenous vein harvested with its cushion of perivascular tissue by a "no touch" technique does not undergo vasospasm and distension is not required, leading to an improved graft patency. The aim of this study is to investigate the effect of surgical damage and high-pressure distension on endothelial integrity and endothelial nitric oxide synthase expression and activity in saphenous vein harvested with and without perivascular tissue.

METHODS

Saphenous veins from patients (n = 26) undergoing coronary artery bypass surgery were prepared with and without perivascular tissue. We analyzed the effect of 300 mm Hg distension on morphology and endothelial nitric oxide synthase/nitric oxide synthase activity using a combination of immunohistochemistry, Western blot analysis, reverse transcriptase polymerase chain reaction, and enzyme assay in distended (with and without perivascular tissue) compared with nondistended (with and without perivascular tissue) segments.

RESULTS

Distension induced substantial damage to the luminal endothelium (assessed by CD31 staining) and vessel wall. Endothelial nitric oxide synthase expression and activity were significantly reduced by high-pressure distension and removal of, or damage to, perivascular tissue. The effect of distension was significantly less for those with perivascular tissue than for those without perivascular tissue in most cases.

CONCLUSION

The success of the saphenous vein used as a bypass graft is affected by surgical trauma and distension. Veins removed with minimal damage exhibit increased patency rates. We show that retention of perivascular tissue on saphenous vein prepared for coronary artery bypass surgery by the "no touch" technique protects against distension-induced damage, preserves vessel morphology, and maintains endothelial nitric oxide synthase/nitric oxide synthase activity.

Tissue-engineered blood vessels with endothelial nitric oxide synthase activity

Nondegradable synthetic polymer vascular grafts used in cardiovascular surgery have shown serious shortcomings, including thrombosis, calcification, infection, and lack of growth potential. Tissue engineering of vascular grafts with autologous stem cells and biodegradable polymeric materials could solve these problems. The present study is aimed to develop a tissue-engineered vascular graft (TEVG) with functional endothelium using autologous bone marrow-derived cells (BMCs) and a hybrid biodegradable polymer scaffold. Hybrid biodegradable polymer scaffolds were fabricated from poly(lactide-co-epsilon-caprolactone) (PLCL) copolymer reinforced with poly(glycolic acid) (PGA) fibers. Canine bone marrow mononuclear cells were induced in vitro to differentiate into vascular smooth muscle cells and endothelial cells. TEVGs (internal diameter: 10 mm, length: 40 mm) were fabricated by seeding vascular cells differentiated from BMCs onto PGA/PLCL scaffolds and implanted into the abdominal aorta of bone marrow donor dogs (n = 7). Eight weeks after implantation of the TEVGs, the vascular grafts remained patent. Histological and immunohistochemical analyses of the vascular grafts retrieved at 8 weeks revealed the regeneration of endothelium and smooth muscle and the presence of collagen. Western blot analysis showed that endothelial nitric oxide synthase (eNOS) was expressed in TEVGs comparable to native abdominal aortas. This study demonstrates that vascular grafts with significant eNOS activity can be tissue-engineered with autologous BMCs and hybrid biodegradable polymer scaffolds.

Lower extremity vein graft failure: a translational approach

Abstract

Patients with the most severe manifestations of lower extremity arterial occlusive disease often require peripheral bypass surgery for limb salvage and preservation of function. Although good quality saphenous vein offers the most durable conduit for reconstruction, 5-year failure rates are 30–50% and have remained largely unchanged for the past two decades. The majority of these failures occur within the first year of implantation, which is regarded as the most biologically active time during which the vein graft adapts to the arterial environment. Although intimal hyperplasia is generally regarded as the primary culprit of vein graft failure, geometric remodeling of the healing vein graft has recently emerged as a potentially significant contributing factor. While hemodynamic forces, including an increase in shear stress and wall tension, are undoubtedly central to the magnitude and direction of vein graft remodeling, we have determined that these forces alone cannot account for the extent of variability noted in early remodeling patterns. Therefore, we hypothesize that circulating factors, such as mediators of inflammation, may modulate the vein graft response to mechanical forces. This article reviews the definition and diagnosis of vein graft failure and summarizes our current efforts to understand the mechanisms of normal and abnormal vein graft adaptation to the arterial environment.

Does periadventitial fat-derived nitric oxide play a role in improved saphenous vein graft patency in patients undergoing coronary artery bypass surgery?

BACKGROUND/AIMS

The saphenous vein is commonly used for coronary artery bypass surgery but its patency is poor. Vascular damage occurs during conventional surgery. However, patency improves when the graft is harvested with minimal surgical trauma, partly due to preservation of vascular endothelial nitric oxide synthase (eNOS) and tissue sources of nitric oxide (NO), a factor possessing both dilatory and anti-proliferative properties. Apart from these grafts exhibiting an intact luminal endothelium they are harvested complete with a surrounding cushion of tissue, much of which is fat.

METHODS

Immunostaining for eNOS was performed on vein graft sections and reverse-transcriptase polymerase chain reaction and Western blotting were used to identify eNOS mRNA and protein. NO synthase activity was measured using the citrulline assay.

RESULTS

Immunohistochemistry identified eNOS staining of vein graft segments, including dense staining of the cushion of perivascular fat and associated structures surrounding the vein. eNOS protein was confirmed in both the vein and surrounding fat by Western blot analysis. Using the citrulline assay, the perivascular fat and underlying vein possessed comparable NO synthase activity.

CONCLUSIONS

Our observations suggest that perivascular fat-derived NO plays a beneficial role in saphenous veins harvested atraumatically and used as grafts in patients undergoing coronary artery bypass surgery.

Evidence supporting changes in Nogo-B levels as a marker of neointimal expansion but not adaptive arterial remodeling

Both neointimal hyperplasia and inward remodeling contribute to restenosis and lumen loss. Nogo-B has been recently described as an inhibitor of vascular injury and neointimal hyperplasia. To determine whether Nogo-B expression may be a mediator of inward remodeling, we examine the localization of expression of Nogo-B in an in vivo model that examines both neointimal hyperplasia and inward remodeling. The rabbit carotid artery was subjected to balloon injury, outflow branch ligation to reduce flow, or both balloon injury and reduction in flow. In balloon injury-induced neointimal hyperplasia Nogo-B expression was reduced in the intima and media but stimulated in the adventitia. In low flow-induced inward remodeling medial Nogo-B expression was not reduced and adventitial Nogo-B expression was not stimulated. Low flow significantly augmented balloon injury-induced neointimal hyperplasia and was accompanied by reduced intimal and medial Nogo-B expression, and increased adventitial Nogo-B expression in both smooth muscle cells and macrophages. Low flow-induced inward remodeling is not associated with changes in medial Nogo-B expression and is distinct from injury-induced neointimal hyperplasia. Pharmacological strategies to inhibit neointimal hyperplasia and restenosis using normal flow models may only partially account for lumen loss and therefore may not accurately predict responses in patients with extensive outflow disease.

Cardiovascular gene delivery: The good road is awaiting

Atherosclerotic cardiovascular disease is a leading cause of death worldwide. Despite recent improvements in medical, operative, and endovascular treatments, the number of interventions performed annually continues to increase. Unfortunately, the durability of these interventions is limited acutely by thrombotic complications and later by myointimal hyperplasia followed by progression of atherosclerotic disease over time. Despite improving medical management of patients with atherosclerotic disease, these complications appear to be persisting. Cardiovascular gene therapy has the potential to make significant clinical inroads to limit these complications. This article will review the technical aspects of cardiovascular gene therapy; its application for promoting a functional endothelium, smooth muscle cell growth inhibition, therapeutic angiogenesis, tissue engineered vascular conduits, and discuss the current status of various applicable clinical trials.

Enhancement of in vivo endothelialization of tissue-engineered vascular grafts by granulocyte colony-stimulating factor

Successful reconstruction of large-diameter blood vessel in humans has been demonstrated using the tissue engineering technique, but improvement in patency of small-diameter bioartificial vascular graft remains a great challenge. This study reports that granulocyte colony-stimulating factor (G-CSF) can enhance in vivo endothelialization of tissue-engineered vascular grafts, which could be used to improve patency of small-diameter vascular graft. Vascular grafts were tissue engineered with decellularized canine abdominal aortas and canine autologous bone marrow-derived cells. Prior to cell seeding onto decellularized graft matrices, bone marrow-derived cells were induced to differentiate into endothelial cells and smooth muscle cells. The cell-seeded vascular grafts were implanted into the abdominal aortas of bone marrow donor dogs. Before and after graft implantation, G-CSF was administered subcutaneously to the dogs (n = 3). The grafts implanted into the dogs not receiving G-CSF were used as controls (n = 3). Eight weeks after implantation, grafts in both groups showed regeneration of vascular tissues including endothelium and smooth muscle. Importantly, endothelium formation was more extensive in the G-CSF-treated grafts than in the control grafts, as assessed with reverse transcription polymerase chain reaction, western blot, and immunohistochemistry. In addition, intimal hyperplasia was significantly reduced in the G-CSF-treated grafts compared to the control grafts. This study suggests that G-CSF administration could be applied to improve patency of small-diameter tissue-engineered vascular grafts.

Experimental study of tissue-type plasminogen activator gene to prevent vein grafts stenosis

The effects of in vivo local expression of recombined human tissue-type plasminogen activator (t-PA) gene on the thrombosis and neointima formation of vein grafts were explored. Jugular vein-to-artery bypass grafting was performed on 72 New Zealand white rabbits. The rabbits were divided into 3 groups according to the different processing methods: transfected t-PA gene group (n = 24), vector group (n = 24) and blank control group (n = 24). Samples of vein grafts were harvested at different time points after surgery. The expression of t-PA gene in vein graft was detected by RT-PCR and the synthesis of t-PA protein by Western-Blot assay. The t-PA activity was measured by chromogenic substrate assay. The Cr51 labeled platelets accumulation in vein grafts was counted. The histopathological changes were compared in intima hyperplasia index among the three groups after operation. The results showed that at the 2nd, 5th, 14th and 28th day after operation, RT-PCR and Western-blot confirmed the expression of t-PA mRNA and protein at the site of gene transfer. The t-PA activity detected on the 2nd, 5th, 14th and 28th day in experimental group was 370.63 +/- 59.44, 344.13 +/- 48.47, 252.87 +/- 51.80 and 161.75 +/- 68.94 U/g respectively, and disappeared on the 60th day and undetected in the control groups. The number of platelets accumulated in the vein grafts in gene group, vector group and blank control group was (85.04 +/- 21.58) 10(6), (225.87 +/- 85.13) 10(6) and (211.7 +/- 78.02) 10(6) respectively. The number of platelets accumulated in gene group was significantly fewer than that in the control groups. Morphometric analysis revealed that intimal hyperplasia was markedly reduced in the t-PA gene group as compared with that in the control groups. It was suggested that the local expression of t-PA gene in vein graft significantly inhibited the accumulation of platelets, thrombosis and concomitant intimal hyperplasia, by which stenosis of bypass graft could be prevented effectively.

Hypothesis: a potential role for the vasa vasorum in the maintenance of vein graft patency

Autologous saphenous vein is the most commonly used conduit for coronary artery bypass surgery with more than 50% grafts occluding within 10 years. In conventional preparation the vein undergoes considerable surgical trauma with damage to the outer layers during harvesting. Within these regions are situated the vasa vasorum and small vessels providing oxygen and nutrients to the vessel wall. Certain vasa vasorum terminate in the vessel lumen where it is suggested that they have a physiological role. Preservation of the vasa vasorum of saphenous veins used as bypass conduits may play an important role in the maintenance of graft patency.

Transduction of peptide analogs of the small heat shock-related protein HSP20 inhibits intimal hyperplasia

BACKGROUND

Human saphenous vein (HSV) is the autologous conduit of choice for peripheral vascular reconstructions. However, vasospasm can lead to early graft failure. The leading cause of delayed graft failure is intimal hyperplasia.

OBJECTIVE

To develop a proteomic approach to prevent vein-graft spasm and intimal hyperplasia.

METHODS

Biomimetic peptide analogs of the small heat shock-related protein HSP20, containing a protein transduction domain (PTD), a phosphorylated serine, and a sequence of HSP20 surrounding the phosphorylation site (PTD-pHSP20), or a scrambled sequence of the same amino acids surrounding the phosphorylation site (PTD-scHSP20) were synthesized. The peptides were used in muscle bath and organ culture experiments with human saphenous vein (HSV) segments. Cultured smooth muscle cell lines were used to determine the effect of the peptides on proliferation and migration.

RESULTS

In HSV rings precontracted with norepinephrine, PTD-pHSP20 but not PTD-scHSP20 led to relaxation. There was no significant difference in smooth muscle cell proliferation in cells treated with PTD-pHSP20 compared with PTD-scHSP20. Treatment with PTD-pHSP20 significantly inhibited cellular migration compared with PTD-scHSP20. Control, untreated, and PTD-scHSP20-treated saphenous veins had significant increases in intimal thickness after culture. This intimal thickening was completely inhibited by treatment with PTD-pHSP20.

CONCLUSIONS

Protein transduction of biologically active motifs of HSP20 can affect pathologic and physiologic responses of HSV and represents a novel proteomic-based therapeutic approach.

CLINICAL RELEVANCE

We have been a part of the genomics era and are now viewing the emergence of "proteomics." The genome is linear and relatively easy to examine; however the proteome is much more complex and dynamic. In essence, the purpose of gene therapy is to manipulate the genome to produce a particular protein. This manuscript describes a new proteomic approach in which the biologically active part of a protein is directly introduced into vascular cells. Peptides were synthesized which contained a total of 24 amino acids, 11 of which represent a protein transduction domain or "carrier" while the other 13 are the biologically active "cargo." These synthetic peptides prevent spasm (contraction) and intimal hyperplasia in segments of human saphenous vein treated ex vivo. Preclinical development is currently underway to develop these molecules as a proteomic-based vein harvest solution to enhance vein-graft patency.

Construction of a recombinant adenovirus carrying endothelial nitric oxide synthase gene and its expression and control in esophageal smooth muscle cells

OBJECTIVE

Nitric oxide (NO) is a major inhibitory neurotransmitter, and its deficiency plays an important role in the pathogenesis of motility disorders of the gastrointestinal tract. The present study was designed to generate a recombinant adenovirus containing the tetracycline (Tet)-regulated endothelial nitric oxide synthase (eNOS) gene and to detect the controllable expression of the gene in esophageal smooth muscle cells (ESMC).

METHODS

The construction of the recombinant adenovirus was completed in three steps: (1) a Tet-responsible expression cassette was made by cloning the full-length cDNA encoding eNOS into a pTRE-Shuttle Vector, which can be regulated by tetracycline or its analogs, such as doxycycline (Dox); (2) the expression cassette was transferred to Adeno-X viral DNA to form a recombinant adenoviral plasmid (pAd-eNOS) by means of an in vitro ligation reaction; and (3) the Ad-eNOS was packaged into infectious adenoviral particles (Adeno-X-TRE-eNOS) by transfecting human embryonic kidney (HEK) 293 cells. Cultured ESMC were coinfected by Adeno-X-TRE-eNOS and regulation virus (Adeno-X Tet-off virus), and the Dox-regulated eNOS expression was detected by RT-PCR and western blot.

RESULTS

The recombinant adenovirus (Adeno-X-TRE-eNOS) was generated successfully by an in vitro ligation reaction. The expression of the eNOS gene in the coinfected ESMC was confirmed by RT-PCR and western blot. Furthermore, the transcription could be precisely regulated in a dose-dependent manner in a series of concentrations of Dox, and it was completely turned off when the concentration reached 0.01 microg/mL.

CONCLUSIONS

A Tet- (or Dox-) regulated recombinant adenovirus carrying eNOS was successfully generated and controllable expression of eNOS in ESMC was achieved, which provides some material for conducting further gene therapy studies with eNOS.

Gastrointestinal smooth muscle cell as target for gene transfer of eNOS gene

AIM

To generate an adenoviral vector carrying endothelial nitric-oxide synthase (eNOS) gene in order to mediate the expression of eNOS gene in gastrointestinal smooth muscle cells (SMC) and assess the enzyme activity of eNOS.

METHODS

A recombinant adenovirus (Ad-eNOS) containing the bovine eNOS cDNA fragment was generated by homologous recombination in bacteria. The SMC of distal part of esophagus and gastric fundus of cat were isolated and cultured in vitro and infected with Ad-eNOS. The expression of eNOS gene was detected by Western blot and RT-PCR. The enzyme activity of NOS and the output of NO in SMC were measured by NOS and NO assay kit, furthermore, the different effects of given factors on the enzyme activity and the yield of NO were studied.

RESULTS

The Ad-eNOS can infect the cultured SMC efficiently (MOI = 50, infection rate = 74%). Western blot and RT-PCR confirmed the expression of eNOS in those infected cells. After the cells had been infected with Ad-eNOS, the basal activity of NOS significantly increased from 47±13 nkat/L to 93±13/L (P<0.05), and the level of NO in cell culture supernatants increased by 3 fold (45±13 vs 16±7 μmol/L). In the presence of L-arginine (NOS enzyme substrate), calcium, EGTA (calcium chelating agent), and L-NAME (NOS inhibitor), NOS activity was 94±8, 173±25, 29±6, 58±11 nkat/L and NO level was 48±14, 106±18, 6±2, 17±11 μmol/L, respectively.

CONCLUSION

The constructed recombinant adenovirus, Ad-eNOS, can efficiently mediate the expression of eNOS gene in cultured SMC of digestive tract. The activity of eNOS can be regulated by the concentration of calcium. L-arginine is Not the rate-limiting step for nitric oxide generation from endothelial nitric oxide synthase.

Prevention of restenosis by a herpes simplex virus mutant capable of controlled long-term expression in vascular tissue in vivo

Neointimal hyperplasia resulting from vascular smooth muscle cell (SMC) proliferation and luminal migration is the major cause of autologous vein graft failure following vascular coronary or peripheral bypass surgery. Strategies to attenuate SMC proliferation by the delivery of oligonucleotides or genes controlling cell division rely on the use of high concentrations of vectors, and require pre-emptive disruption of the endothelial cell layer. We report a genetically engineered herpes simplex virus (HSV-1) mutant that, in an in vivo rabbit model system, infects all vascular layers without prior injury to the endothelium; expresses a reporter gene driven by a viral promoter with high efficiency for at least 4 weeks; exhibits no systemic toxicity; can be eliminated at will by administration of the antiviral drug acyclovir; and significantly reduces SMC proliferation and restenosis in vein grafts in immunocompetent hosts.