Quantification of vascular graft seeding by use of computer-assisted image analysis and genetically modified endothelial cells

Vascular Grafts and the Endothelium

This article discusses the importance of the endothelium for successful vascular grafts derived from both native arteries and synthetic materials. It also discusses the fundamental strategies to endothelialize synthetic grafts in animal experiments and in the clinic, as well as the use of endothelial progenitor cells (EPCs), bone marrow-derived cells, and mesothelium as endothelial substitutes.

Engineering blood vessels by gene and cell therapy

Cardiovascular-related syndromes are the leading cause of morbidity and mortality worldwide. Arterial narrowing and blockage due to atherosclerosis cause reduced blood flow to the brain, heart and legs. Bypass surgery to improve blood flow to the heart and legs in these patients is performed in hundreds of thousands of patients every year. Autologous grafts, such as the internal thoracic artery and saphenous vein, are used in most patients, but in a significant number of patients such grafts are not available and synthetic grafts are used. Synthetic grafts have higher failure rates than autologous grafts due to thrombosis and scar formation within graft lumen. Cell and gene therapy combined with tissue engineering hold a great promise to provide grafts that will be biocompatible and durable. This review describes the field of vascular grafts in the context of tissue engineering using cell and gene therapies.

Cellular engineering of vascular bypass grafts: role of chemical coatings for enhancing endothelial cell attachment

Surgical treatment of vascular disease has become common. The use of synthetic materials is limited to grafts larger than 5-6mm, because of the frequency of occlusion observed with small-diameter prosthetics. An alternative would be a hybrid or tissue-engineered graft with the surface coated with a monolayer of the patient's own cells. Currently, to be effective, high-density seeding regimens have to be undertaken. This is because endothelial cells (ECs) are washed off the graft lumen once exposed to physiological blood flow. EC attachment has been shown to be significantly improved by pre-coating with substances known to attach ECs selectively. The review examines the various types of coating and bonding technology used to date to enhance endothelial cell attachment onto the surface of prosthetic vascular bypass grafts.

Endothelial cell delivery for cardiovascular therapy

Obstructive atherosclerotic vascular disease stands as one of the greatest public health threats in the world. While a number of therapies have been developed to combat vascular disease, endothelial cell delivery has emerged as a distinct therapeutic modality. In this article, we will review the anatomy of the normal blood vessel and the biology of the intact endothelium, focusing upon its centrality in vascular biology and control over the components of the vascular response to injury so as to understand better the motivation for a cell-based form of therapy. Our discussion of cell delivery for cardiovascular therapy will be divided into surgical and interventional approaches. We will briefly recount the development of artificial grafts for surgical vascular bypass before turning our attention towards endothelial cell seeded vascular grafts, in which endothelial cells effectively provide local delivery of endogenous endothelial secretory products to maintain prosthetic integrity after surgical implantation. New techniques in tissue and genetic engineering of vascular grafts and whole blood vessels will be presented. Methods for percutaneous interventions will be examined as well. We will evaluate results of endoluminal endothelial cell seeding for treatment of restenosis and gene therapy approaches to enhance endogenous re-endothelialization. Finally, we will examine some innovations in endothelial cell delivery that may lead to the development of endothelial cell implants as a novel therapy for controlling proliferative vascular arteriopathy.

Gene therapy for vascular disease

Atherosclerosis is a degenerative process characterized by endothelial cell dysfunction, inflammatory cell adhesion and infiltration, and the accumulation of cellular and matrix elements leading to the formation of fibrocellular plaques. In the end stages, advanced occlusive plaques limit blood flow and oxygen delivery resulting in the well-known ischemic syndromes of the coronary, skeletal muscle, mesenteric, and cerebrovascular circulation. Moreover, sudden critical ischemic events may be precipitated by plaque disturbance, rupture, hemorrhage, and/or thrombosis. Traditional pharmacologic approaches have been effective in reducing serum cholesterol and controlling thrombosis but, in the main, have had little impact on the treatment of advanced lesions. The purpose of this review is to examine the current status of gene therapy for vascular proliferation, aberrant endothelial function, thrombosis, peripheral ischemia, and modification of the blood/biomaterial interface.

Healing characteristics of intraarterial stent grafts in an injured artery model

Previous investigations reveal in the absence of endothelial cell (EC) injury, intraarterial polytetrafluoroethylene (PTFE) stent graft (SG) exhibit greater EC repaving than PTFE interposition grafts (CG). The investigation evaluated EC repaving of SG compared to CG after balloon injury. Twenty adult dogs underwent SG (n = 10) or CG (n = 10) placement in the infrarenal aorta after balloon injury with harvest at 1 and 6 weeks. Endothelial repaving, intima-to-media height ratios (IMHR), and inflammatory stains were performed. Endothelial repaving was greater in 6-week SG compared to CG (51% +/- 5.0 versus 10% +/- 5.0, p < or = 0.05). IMHR was less in 6-week SG compared to CG at the proximal (1.22 +/- 0.16 versus 1.82 +/- 0.16, p < or = 0.05) and distal anastomosis (0.81 +/- 0.25 versus 1.33 +/- 0.25, p < or = 0.05). Smooth muscle cell (SMC) alpha-actin was greater in 1-week SG compared to CG at the distal anastomosis (121.5 +/- 7.2 versus 94.0 +/- 7.2, p < or = 0.05). Proliferating cell nuclear antigen (PCNA) was less in 6-week SG compared to CG at the proximal (5.6 +/- 1.4 versus 9.4 +/- 1.1, p < or = 0.05) and distal anastomosis (3.8 +/- 0.6 versus 11.2 +/- 1.1, p < or = 0.05). Macrophage CD-44 was less in 6-week SG compared to CG at the proximal (10.4 +/- 1.6 versus 32.6 +/- 3.6, p < or = 0.05) and distal anastomosis (8.6 +/- 0.9 versus 35.6 +/- 3.6, p < or = 0.05). Intraarterial SG placed after balloon injury exhibited significantly greater endothelialization and less intimal hyperplasia when compared to CG.

Retroviral-mediated transduction of endothelial cells with the lac Z gene impairs cellular proliferation in vitro and graft endothelialization in vivo

PURPOSE

An endothelialized lumen within a synthetic graft that expresses recombinant proteins with anticoagulant or antiproliferative activity has the potential to improve graft function. However, preliminary data suggest that genetic modification of endothelial cells (ECs) impairs their proliferation. The purpose of this investigation was to test the hypothesis that retroviral transduction of cultured ECs with the lac Z gene encoding for beta-galactosidase would decrease EC proliferation in vitro and graft endothelialization in vivo.

METHODS

In vitro studies compared canine EC proliferation over a 14-day period among early-passage ECs (two and three) and late-passage ECs (six and nine) transduced with the BAG vector (containing the lac Z gene and the neomycin resistance gene), ECs transduced with the neomycin resistance gene only, the nontransduced ECs. In vivo canine studies assessed endothelialization of expanded polytetrafluoroethylene thoracoabdominal grafts seeded with autologous lac Z-transduced ECs (n = 7) or nontransduced ECs (n = 3) compared with that of nonseeded grafts (n = 3). Histochemical staining and DNA polymerase chain reaction was used 6 weeks after implantation to detect the presence of the lac Z gene in the grafts' cellular linings and perigraft tissues. Endothelialization was assessed by light microscopy and electron microscopy.

RESULTS

Proliferation of late-passage lac Z-transduced ECs in vitro was significantly decreased compared with neomycin resistance-transduced ECs or nontransduced ECs. Among early-passage ECs smaller but significant decreases in proliferation were noted among lac Z-transduced cells compared with nontransduced cells. Six of seven expanded polytetrafluoroethylene grafts seeded with transduced ECs showed lac Z gene expression. Lac Z gene expression was not found on grafts seeded with nontransduced ECs or nonseeded grafts. The endothelialized luminal surface area was significantly less in grafts seeded with lac Z-transduced ECs compared with grafts seeded with nontransduced ECs.

CONCLUSIONS

Retroviral-mediated transduction of canine ECs with the lac Z gene encoding for beta-galactosidase impairs EC proliferation in vitro and the ability of transduced ECs to form a confluent EC monolayer on the luminal surface of synthetic grafts in vivo.

Seeding of vascular grafts with genetically modified endothelial cells. Secretion of recombinant TPA results in decreased seeded cell retention in vitro and in vivo

BACKGROUND

Seeding of small-diameter vascular grafts with endothelial cells (ECs) genetically engineered to secrete fibrinolytic or antithrombotic proteins offers the potential to improve graft patency rates.

METHODS AND RESULTS

Sheep venous ECs were transduced with a retroviral vector encoding human tissue plasminogen activator (TPA). The ECs were seeded onto 4-mm-ID synthetic (Dacron) grafts. Retention of the seeded ECs was measured 2 hours after placement of the seeded grafts both in vitro in a nonpulsatile flow system and in vivo (in sheep) as femoral and carotid interposition grafts. On exposure to flow in vitro, ECs transduced with TPA were retained at a significantly lower rate (median, 67%) than either untransduced ECs (81%) or ECs transduced with a control retroviral vector producing beta-galactosidase (beta-Gal) (80%) (P < .05 for TPA versus either control). On implantation in vivo, ECs transduced with TPA were retained at a very low rate (median, 0%), significantly less than the retention of ECs transduced with the beta-Gal vector (32%; P < .00001). Decreased in vivo retention of ECs transduced with TPA correlated modestly with increased in vitro cellular passage level (r2 = .48; P < .0001) but not with in vivo blood flow rate (P = .45). Addition of the protease inhibitor aprotinin to the cell culture and graft perfusion media resulted in a significant (P < .05) increase in in vitro retention of ECs transduced with TPA.

CONCLUSIONS

Increased TPA expression significantly decreases seeded EC adherence in vitro and in vivo. Gene therapy strategies for decreasing graft thrombosis may require expression of antithrombotic molecules that lack proteolytic activity.

Fluid shear induced endothelial cell detachment from glass--influence of adhesion time and shear stress

In this study, human umbilical vein and human saphenous vein endothelial cells were seeded on glass and exposed to fluid shear in a parallel-plate flow chamber. Cell retention, morphology and migration were studied as a function of shear stress and of adhesion time prior to exposure to shear. Three-hour and 24-h adhesion times gave rise to comparable cell retention values after 2 h of flow for both cell types. Cell retention decreased from 85 to 20% as shear stress increased from 88 to 264 dynes cm-2 (8.8 to 26 Pa). Mean spreading areas decreased after the onset of flow, but subsequently stabilized to plateau values, which were smaller at higher shear stresses. Shape factors increased faster to higher values as cells were exposed to higher shear stresses, without any obvious preference in orientation of the cells with respect to the direction of flow. Migration was unidirectional with flow and linear with time. Migration was faster for cells which had adhered for 24 h than for cells which had adhered for 3 h and was accompanied by the presence of fibrillar structures left behind on the surface upstream of migrating cells. It is concluded that after 3 h adhesion to glass, cells have adhered with an adhesion strength that does not substantially increase during the next 21 h. However, during this time changes in cell-substratum interactions seem to occur judging by the differences in, e.g., migration rates.

Effect of pulsatile shear stress on endothelial attachment to native vascular surfaces

An in vitro model of vascular damage was used to investigate the ability of seeded endothelial cells to resist shear stresses generated in a perfusion circuit. At perfusion rates of 100 ml/min the maximum shear stress reached 16.5 dyn/cm2. At this level the rate of cell detachment from the damaged vascular surface was 88 per cent per h for the first 20 min of flow but gradually decreased to 5 per cent per h after 90 min. These findings suggest that endothelial cells may be retained on damaged vascular surfaces in conditions that approximate to arterial flow.

Quantification of in vitro endothelial cell adhesion to vascular graft material

This study tests the hypothesis that denucleating vascular graft material and binding cell adhesion molecules increases endothelial attachment. Removal of gas nuclei (denucleation) increases the available surface area of biomaterials for modification and/or cell adhesion, while adhesion molecules provide specific attachment sites. Microvascular endothelial cells (MVEC) were isolated from fat, fluorescently labeled, and allowed to settle onto expanded polytetrafluoroethylene (ePTFE) vascular patches. Patch treatments included fibronectin alone (F), gas denucleation followed by fibronectin (D/F), denucleation followed by the surfactant tridodecylmethylammonium chloride (TDMAC) (D/T), denucleation followed by TDMAC followed by fibronectin (D/T/F), or denucleation followed by TDMAC followed by a synthetic polymer with numerous arginine-glycine-aspartic acid sequences (D/T/R). After 1 h of incubation, the 45 mm2 patch area covered with microvascular endothelial cells was assessed using computer-aided fluorescence microscopy. Initial graft coverage with D/T (26.2 +/- 2.4 mm2) and D/T/F (25.9 +/- 2.1 mm2) was better than with F (16.8 +/- 2.5 mm2) (P < .05). Patches were then exposed to a detachment stress and coverage was again measured. Following stress, coverage was greater with D/T (20.7 +/- 3.4 mm2) and D/T/F (20.7 +/- 2.0 mm2) than with D/T/R (8.4 +/- 1.8 mm2) or F (3.6 +/- 0.9 mm2) (P < .001). Percent retention of cells following stress was better with D/T and D/T/F than with D/T/R, D/F, or F (P < .0001). Scanning electron micrographs were consistent with the qualitative findings. The results indicate that TDMAC alone or with fibronectin increases adhesion of human microvascular endothelial cells to denucleated ePTFE.

Low level in vivo gene transfer into the arterial wall through a perforated balloon catheter

BACKGROUND

Gene transfer into the arterial wall may provide a novel therapeutic strategy for the treatment of coronary artery restenosis. Previously described methods for gene transfer into the arterial wall require total vessel occlusion for 30 minutes. We sought to develop a protocol for gene transfer within a more clinically relevant time frame.

METHODS AND RESULTS

We used a perforated balloon (Wolinsky) catheter to inject retroviral vector-containing virions into rabbit aortas in vivo. The virions were injected within 1 minute. Aortas were removed 5-14 days after injection and analyzed for evidence of gene transfer. In initial studies, nine rabbits were injected with a vector expressing the beta-galactosidase gene, and nine rabbits were injected with either non-beta-galactosidase-containing vectors or with a vehicle control. Histochemical staining of aortic tissues revealed blue (positive) cells in eight of nine experimental rabbits and six of nine controls. Because of the lack of specificity of the beta-galactosidase detection system, we adopted a polymerase chain reaction-based protocol in which oligonucleotide primers were used to amplify specific vector-related sequences from aortic tissue extracts. The polymerase chain reaction protocol, calibrated with standards containing known numbers of transduced cells, revealed low amounts of vector-related sequences in six of 12 vector-injected rabbits and in one of 13 controls (p less than 0.03). Comparison with standards indicated that fewer than 100 transduced cells were present in a 2-cm length of the injected aortic tissue.

CONCLUSIONS

Although in vivo gene transfer through an infusion balloon catheter can be accomplished within 1 minute, the therapeutic use of this protocol is limited by the small number of cells that are transduced.