Search

Abstract: Endothelial cells adhere to one another through junctional structures formed by intercellular adhesion molecules. These intercellular proteins regulate barrier function in response to the hemodynamic shear rate and enable the selective passage of solutes and fluids across the endothelium. After endovascular device implantation, the endothelial barrier is compromised and becomes discontinuous, which increases permeability, allowing transmigration of leukocytes and lipoproteins and leading to the accumulation of lipid-laden foamy macrophages in the subendothelial space. Drug-coated bioresorbable vascular scaffold implants have been associated with unexpected thrombotic complications, which were not predicted in animals because of dissimilarities in endothelial regeneration and realignment between animals and humans. The development of a microengineered, microfluidics-based system of patterned channels lined with human endothelial and smooth muscle cells perfused with blood allows for the evaluation of endothelial function and barrier integrity. This review highlights the translational potential of vasculature-on-chip, which recreates the microphysiological milieu to evaluate the impact of drug-eluting bioresorbable vascular scaffolds on endothelial barrier integrity and to characterize polymer biodegradation behavior and drug release kinetic profiles over time., Competing Interests: The author reports no conflicts of interest., (Copyright © 2021 Wolters Kluwer Health, Inc. All rights reserved.)

The drug-eluting stent platform has a limited surface area, and a polymer carrier matrix is coated to enable sufficient loading of drugs. The development of a suitable polymer has been challenging because it must exhibit biocompatibility with the intravascular milieu. The use of biodegradable polymers seems to be attractive because it enables drug release as it degrades and is eventually eliminated from the body leaving the permanent metallic stent polymer-free. The aim of this study was to investigate the biocompatibility of biodegradable polymers using the human monocyte cell line. Cultured monocytes differentiated into functional macrophages (THP-1) were incubated with various polymers including poly-L-lactide (PLA), polycaprolactone (PCL), or poly-D, L-lactide-co-glycolide (PLGA) for up to 5 days. Exposure of cells to the polymers resulted in macrophage-polymer adhesion and induced marked pro-oxidant species as measured by calcein AM uptake assay and flow cytometric analysis of 2',7'-dichlorofluorescin fluorescence, respectively. Real-time reverse-transcription polymerase chain reaction and Western blot analysis of expression of nicotinamide-adenine dinucleotide phosphate (NADPH) oxidases revealed enhanced expression of NADPH oxidase subunits in response to PLA and PLGA compared with that of PCL. Flow cytometric analysis of fluorescein isothiocyanate-Annexin V and propium iodide-stained PLA and PGLA polymer-exposed THP-1 cells showed early and late apoptotic changes. Similarly, exposure to the PLA and PGLA polymers, but not to the PCL polymer, resulted in enhanced staining for cleaved poly(ADP-ribose) polymerase-1, a protein fragment produced by caspase cleavage. These results indicate that biodegradable polymers are associated with cell adhesion, NADPH oxidase-induced generation of reactive oxygen species and excess apoptosis.

To overcome the limitations of balloon expandible metal stent-induced neointimal smooth muscle cell proliferation, drug-coated stent devices have been developed. Drug eluting stents release high concentrations of antiproliferative agents, such as paclitaxel, to reduce neointimal hyperplasia. The proinflammatory cytokine, tumor necrosis factor-alpha (TNF-alpha), is known to cause severe endothelial dysfunction and accelerate atherosclerotic lesion progression. The interaction of TNF-alpha and paclitaxel on the release of prothrombotic molecules was examined in endothelial cells. Treatment of endothelial cells with paclitaxel had no direct effect on tissue factor (TF) expression, but TNF-alpha increased TF. Cotreatment of paclitaxel with TNF-alpha markedly augmented the release of TF. TNF-alpha induced release of plasminogen activator inhibitor but no synergism occurred with paclitaxel. Treatment of endothelial cells with paclitaxel and TNF-alpha reduced expression of thrombomodulin and protein C receptor. Tissue factor pathway inhibitor expression was reduced by prolonged treatment with either paclitaxel or TNF-alpha. The adhesion molecule, CD62 E, was induced by TNF-alpha; however, CD31, CD62 P, and CD106 were not affected by paclitaxel and TNF-alpha. Apoptosis was not observed with cotreatment of endothelial cells with paclitaxel and TNF-alpha. CD59-positive microparticles were released in response to TNF-alpha, but the release was not augmented by paclitaxel. Paclitaxel and TNF-alpha increased the nitrotyrosination of proteins. These findings indicate that paclitaxel enhances TNF-alpha-induced release of TF, and downregulated thrombomodulin, increased protein nitration, which may subsequently favor prothrombotic intimal surface.

In an effort to overcome the limitations of balloon-expandible intravascular metal stent-induced neointimal formation, drug-coated stent devices have been developed. The stent platform allows the local delivery of drugs to an injury site, thereby reducing the amount of drug exposure to the systemic circulation and other organs. The drug carrier matrix allows the release of the drug in a diffusion-controlled manner over an extended time period after the stent implant. The drugs are chosen such that the complex cascade of events that occurs after stent implantation that leads to smooth muscle cell proliferation and migration towards the intima are inhibited. The success of an antirestenotic drug therapy from a drug-coated stent is dependent, at least partially, on the extent of drug elution from the stent, the duration and rate of release, and accumulation of drug in the arterial wall in such a way that it covers the initiation and progression of vessel wall remodeling. The local vascular drug concentrations achieved are directly correlated with the biological effects and local vascular toxicity, and there is therefore a challenge in finding an optimum dose of drug to be delivered to tissues (ie, one that has the desired therapeutic effect without local adverse effects). There is increased focus on optimization of various factors that affect drug release from the stent system, including the physicochemical properties of the drugs, carrier vehicle formulation, and profile of elution kinetics. This review highlights the various factors involved in drug release kinetics, local vascular toxicity, carrier vehicle matrix, tissue deposition, and distribution through the arterial wall from stent-based drug delivery systems.

The direct effects of the cholesterol-lowering agents, 3-hydroxy-3-methylglutaryl-coenzyme A (HMG CoA) reductase inhibitors, on vascular smooth muscle responsiveness were examined by incubation of isolated aorta from normocholesterolemic rats with simvastatin, atorvastatin, or pravastatin. The smooth muscle contractions caused by phenylephrine were progressively inhibited with increasing concentrations of simvastatin. Similarly, atorvastatin at the higher concentration caused decreased responses to phenylephrine. In contrast, incubation with pravastatin had no significant effect at all concentrations studied. In Ca2+-free buffer, the transient contraction caused by phenylephrine, which results from intracellular release of Ca2+, also was inhibited by simvastatin and atorvastatin but not by pravastatin. In cultured rat aortic smooth muscle cells loaded with fura-2, increases in intracellular free-Ca2+ concentration ([Ca2+]i) induced by angiotensin II were markedly inhibited in cells incubated with simvastatin and atorvastatin but not pravastatin. The inhibitory effects of simvastatin and atorvastatin were reversed by mevalonate. These findings demonstrate that inhibition of HMG CoA reductase by using simvastatin and atorvastatin, but not pravastatin, has effects on vascular smooth muscle cell responsiveness that involve alteration of Ca2+ homeostasis through a mevalonate-dependent pathway.

We examined the effects of diabetes on eicosanoid metabolism and endothelium-dependent relaxation in isolated aorta from alloxan-induced diabetic rabbits and that from normal rabbits incubated in increased concentrations (44 mM) of glucose in vitro for 6 h. Immunoreactive 15-hydroxyeicosatetraenoic acid (HETE) was assayed in the incubation media of isolated aortic segments. Basal and acetylcholine (ACh)-stimulated release of 15-HETE was significantly greater in aorta of diabetic animals as compared with those of normal rabbits. Incubation of aortic segments from normal rabbits in increased concentrations of glucose caused a significant increase in basal and ACh-stimulated release of 15-HETE; and the release was significantly greater in aortic segments with endothelium than in segments without endothelium. Basal and ACh-stimulated release of 15-HETE was inhibited by indomethacin, a cyclooxygenase inhibitor. 15-HETE caused contractions of aortic rings that were inhibited by the prostaglandin H2 (PGH2) thromboxane A2 (TXA2) receptor blocker SQ-29548, but not by the TXA2 synthase inhibitor carbethoxyhexyl imidazole or indomethacin. Treatment of aortic rings with subthreshold concentrations of 15-HETE impaired ACh-induced relaxation; this was prevented by treatment with SQ-29548. Thus, abnormal release of endothelium-derived 15-HETE may play a role in endothelial cell dysfunction and increased vasoconstriction in diabetes by a mechanism that involves interaction with PGH2/TXA2 receptors.

A possible relationship between increased aldose reductase activity and abnormal endothelium-dependent relaxation was examined in aorta from alloxan-induced diabetic rabbits. Isolated aorta of diabetic rabbits, contracted submaximally with phenylephrine, showed significantly decreased endothelium-dependent relaxations induced by acetylcholine or adenosine diphosphate compared to those from normal rabbits. Basal and acetylcholine-stimulated levels of cyclic GMP and the relaxations in response to an endothelium-independent vasodilator, sodium nitroprusside, were not significantly different between diabetic and normal rabbits, indicating that nitric oxide release and action on the vascular smooth muscle were unchanged. The release of thromboxane A2 from diabetic vessels was increased, as previously demonstrated. Treatment with an aldose reductase inhibitor, zopolrestat, normalized the elevated red blood cell sorbitol levels in diabetic rabbits. Zopolrestat also restored the abnormal acetylcholine- and adenosine diphosphate-induced relaxations of the aorta. The aldose reductase inhibitor had no effect on the levels of cyclic GMP or on the increased release of thromboxane A2 in diabetic aorta. These findings suggest that increased activity of the aldose reductase pathway in hyperglycemia is responsible for the abnormal endothelium-dependent relaxation in diabetic blood vessels. Significant alterations in endothelial production of neither nitric oxide nor vasoconstrictor prostanoids could be directly implicated in the improvement caused by the drug, suggesting another mechanism of action.

Transmural electrical stimulation was used to elicit frequency-dependent adrenergic neurogenic contractions in isolated carotid arteries from cholesterol-fed and control rabbits. In rings with endothelium, responses to adrenergic nerve stimulation were significantly greater in arteries from cholesterol-fed as compared with those from control rabbits. Responses to adrenergic nerve stimulation of rings without endothelium were not different between the two groups. Methylene blue, a guanylate cyclase inhibitor, increased contractions of rings with endothelium and abolished the difference between the responses of arteries from cholesterol-fed and control rabbits. Methylene blue had no significant effect on arteries without endothelium. The overflow of endogenous norepinephrine (NE) caused by transmural electrical stimulation was not different between segments of arteries from cholesterol-fed and control rabbits. In control rabbits, exogenously applied NE contracted arteries with endothelium less than arteries without endothelium, whereas in cholesterol-fed rabbits the contractions caused by NE were not different between arteries with and without endothelium. Acetylcholine-induced relaxations were not different between rings with endothelium from cholesterol-fed and control rabbits. These results suggest that hypercholesterolemia selectively impairs the inhibitory influence of the endothelium on adrenergic contractions.