Vasoconstriction occurs frequently following coronary angioplasty and is implicated in the pathogenesis of abrupt closure and restenosis. Control of vasomotor tone is regulated in part directly by smooth muscle cells and indirectly through the endothelium. To study the mechanisms underlying vasoconstriction, the effect of angioplasty and endothelial denudation on endothelium-dependent and -independent relaxation was examined in 15 mongrel dogs. Percutaneous transluminal angioplasty and endothelial denudation of the right femoral artery were performed. Endothelial injury was assessed by adhesion of indium-111-labeled platelets. Endothelium-dependent and -independent relaxation were assessed using acetylcholine and nitroglycerin, respectively. Vessels precontracted with potassium chloride and exposed to acetylcholine showed impaired relaxation in both the angioplasty and denuded groups (angioplasty = 14 ± 5%, denuded = 0 ± 0%, normal = 73 ± 12%; P < 0.05 for both angioplasty and denuded compared to normal). Precontraction with phenylephrine yielded similar results (angioplasty = 16 ± 8%, denuded = 4 ± 2%, normal = 39 ± 10%; P < 0.05 only for denuded segment compared to normal). Segments precontracted with phenylephrine and exposed to nitroglycerin did not demonstrate impaired relaxation (angioplasty = 73 ± 9%, denuded = 68 ± 9%, normal = 71 ± 7%, P = ns). Mean indium-111 counts were similar in both the angioplasty and denuded segments (2820 ± 1481 and 2963 ± 1228 counts/ming/g, respectively) compared to a lower count in the normal segment (1514 ± 956 counts/min/g). Thus, angioplasty produces significant vascular injury and impairment of vasodilator function, comparable to that caused by endothelial denudation alone. This implies that vasoconstriction seen following coronary angioplasty may be due to endothelial injury and the resultant loss of control of vasomotor tone.

Thrombosis is a common sequela of total parenteral nutrition. We have recently demonstrated in vitro that hypertonic total parenteral nutrition solutions are potent inducers of a tissue factor monocyte procoagulant activity, the initiating cofactor of the extrinsic clotting cascade. We have further studied, in vitro, the effects of the component solutions of total parenteral nutrition on the induction and modulation of endothelial cell procoagulant activity. Cultured porcine aortic endothelial cells were incubated with (a) 200 microliters of dextrose solution (5%, 10%, 20%, 25%, and 50%), (b) 200 microliters of amino acid solution [full strength (N), one-fourth strength, and one-half strength], and (c) 200 microliters of 10% lipid emulsion. Cocultures of lipid emulsion and 20% dextrose, lipid emulsion and full-strength 10% amino acid solution (N-amino acid), and lipid emulsion and bacterial lipopolysaccharide also were studied. Cells were incubated for intervals of 3-108 h, washed and frozen, harvested, and assayed for endothelial cell procoagulant activity. Units of endothelial cell procoagulant activity were derived from a standard thromboplastin curve. Our results show that amino acid and hypertonic dextrose total parenteral nutrition solutions are able to strongly induce endothelial cell procoagulant activity expression in vitro. In contrast, lipid emulsion significantly inhibited the induction of endothelial cell procoagulant activity by 20% dextrose, N-amino acid, and lipopolysaccharide. These results provide further evidence for the role of the cellular pathways of coagulation in total parenteral nutrition-induced thrombosis. Furthermore, the inhibitory properties of lipid emulsion may be of practical advantage in reducing total parenteral nut induced thrombosis.

Phagocytes of the smooth dogfish (Mustelus canis) contain no endogenous peroxidase within their lysosomes and constitute models for cells genetically deficient in lysosomal enzymes such as myeloperoxidase. We have obtained uptake of over 50% of exogenous horseradish peroxidase, provided the enzyme is exhibited to cells after incorporation into liposomes coated with heat-aggregated (62 degrees, 10 min), isologous IgM. Trapping of horseradish peroxidase (EC 1.11.1.7) by liposomes was established by chromatographic resolution (Sephadex G-200; Sepharose 2B and 4B) of free enzyme from that associated with liposomes; liposome-associated horseradish peroxidase, together with trapped markers of the aqueous compartment (glucose, CrO4 equals), were excluded, and free enzyme and markers were retained. Enzyme and marker trapping was not electrostatic, varied with the molar ratio of charged membrane components, and was reversed by detergent lysis (Triton X-100) of liposomes. Uptake at 30 degrees of aggregated IgM-coated liposomes containing trapped horseradish peroxidase exceeded that of free enzyme of 100-fold, and was more efficient than uptake of horseradish peroxidase presented in uncoated liposomes or in liposomes coated with native IgM. After phagocytosis, peroxidase-rich liposomes were localized exclusively in lysosomes of the phagocytes by ultrastructural histochemistry; the enzyme displayed over 50% latency to osmotic lysis. This method may prove to be of general use in the provision of exogenous enzymes to phagocytic cells genetically deficient in lysosomal hydrolases.