1. The ability of blood to flow through capillaries with diameters smaller than that of erythrocytes and leucocytes has been explained hypothetically on the basis of electrostatic repulsion between vessel walls and circulating cells, but in the absence of in vivo evidence. Following our discovery of extraordinarily high densities of sialic acids on endothelial surfaces, we have now devised experiments to determine to what extent microcirculatory blood flow might depend on these negative charges. 2. Hindlimbs of rats anaesthetized with ether and pentobarbitone were perfused through the femoral artery at constant volume with continuous recording of inflow resistance: first with carotid arterial blood containing continually infused acetylcholine at concentrations which minimized peripheral resistance without affecting heart rate; then for 5 min with buffered saline containing acetylcholine at the same concentration plus neuraminidase (from Vibrio cholerae, at 0.1 mg/ml) to remove endothelial sialic acids (for control experiments without neuraminidase); and then again with acetylcholine-infused carotid blood. 3. As a consequence of the saline perfusion, there was a rapid initial rise and fall of inflow resistance. The subsequent changes in resistance were calculated as the difference between its lowest value in the first 10 min and the value after 60 min. Controls and neuraminidase-perfused groups were compared using the t test for unpaired measurements. 4. In four control experiments inflow resistance decreased by 10-40% (14 +/- 10%, mean +/- S.E.M.). In three experiments with neuraminidase, inflow resistance increased by 24-38% (29 +/- 4%, mean +/- S.E.M.). The difference between the saline and the neuraminidase-perfused groups was significant (P less than 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)

Sialic acids present on luminal surfaces of vascular endothelium were determined by perfusing neuraminidase free of proteolytic activity through carotid arteries, iliac arteries and jugular veins of anaesthetized rabbits and guinea-pigs and through human umbilical veins. Total sialic acids released in I h from arteries and veins, determined fluorimetrically, were 24-51 X 10(6) molecules/micron 2 endothelial surface; this was more, by up to two orders of magnitude, than sialic acids releasable by neuraminidase from other types of cells, i.e. from 0.15 X 10(6) for human erythrocytes to 15 X 10(6) for human platelets. Sialic acids therefore provide extraordinarily high negative charge densities on vascular surfaces exposed to the flowing blood. As all circulating cells are also negatively charged, strong electrostatic repulsion must exist between them and normal vessel walls. These observations can therefore account for the general property of non-adherence of circulating cells in normal blood vessels of which the so-called 'non-thrombogenicity', meaning the non-adherence of platelets to normal vascular endothelium, is one example. It is suggested that a major biological function of these extraordinarily high negative charge concentrations is the mutual repulsion between endothelial surfaces and blood cells which promotes their unimpeded circulation.

Erythrocyte deformability was demonstrated to be influenced by platelet activation. Deformability of erythrocytes suspended in autologous platelet poor plasma (PPP), obtained from platelet rich plasma (PRP), was significantly reduced when PRP had previously been incubated with a platelet activating substance (arachidonic acid, adrenaline or ADP). The possibility of a direct influence of the activating substance on erythrocyte deformability was examined and malondialdehyde formation was determined as an indicator of platelet activation. Erythrocyte deformability was not impaired when endoperoxide formation in platelets was blocked by an inhibitor of cyclooxigenase (acetylsalicylic acid). Plasma viscosity was not influenced by platelet activation as demonstrated by filtration and viscosimetry. Recent studies showed that prostacyclin (PGI2) increases erythrocyte deformability (1). The antagonistic action between prostacyclin released by vessel walls and products of platelet metabolism being well known, we discuss possible mechanisms of this effect and pathophysiological relevance of our results.

Capillary occlusions are characteristic features of the early diabetic retinopathy and are presumed to initiate neovascularization. Activated leukocytes can cause microvascular occlusions and cell damage by release of cytotoxic products. To explore the role of leukocytes in capillary occlusion, nonperfusion, and neovascularization of diabetic retinopathy, a rat model was used, in which a diabetic state was induced by alloxan. Retina flat preparations were differentially stained for monocytes and granulocytes. Capillary occlusion, nonperfusion, and neovascularization were assessed microscopically in the center, midperiphery, and periphery of the retina. In contrast to control retinas, 2- to 9-month diabetic rats showed many capillary occlusions by leukocytes, especially monocytes, endothelial cell damage, extravascular macrophage accumulation, and tissue damage. The percentage of activated monocytes and granulocytes in the circulating blood of diabetic rats was greatly increased, and areas of capillary 'loss' and neovascularization in the retina coincided with sites of extravascular leukocytes. The authors' results suggest a potential role of monocytes and macrophages in the pathogenesis of diabetic retinopathy.

Atherosclerotic lesions contain oxidized LDL (OxLDL), immunoglobulins, and immune-competent cells. Low levels of circulating autoantibodies against malondialdehyde (MDA)-modified lysine, an epitope of OxLDL, occur in several species, and immune complexes between such autoantibodies and OxLDL are present in lesions. To study the potential role of autoantibodies against OxLDL in the atherogenic process, we prospectively hyperimmunized LDL receptor-deficient rabbits with homologous MDA-LDL and determined the effects of this intervention on the development of atherosclerosis. Immunization with MDA-LDL generated high titers of antibodies with similar specificity as naturally occurring autoantibodies. Immunized animals showed a significant reduction in the extent of atherosclerotic lesions in the aortic tree after 6.5 months, compared with "saline"-immunized controls (48% vs. 68%, P < 0.005). Immunization with keyhole limpet hemocyanin produced no change in lesion formation. Although the mechanisms by which immunization led to a protective effect are unknown, these results suggest an important role for the immune system in modulating the atherogenic process and may indicate a novel approach for inhibiting the progression of atherosclerosis.

Murine strains susceptible to atherosclerosis provide valuable models to study factors involved in atherogenesis. In some murine models, limited hypercholesterolemia can be achieved and lesions develop primarily in the aortic origin, in the vicinity of the aortic valve. In other models such as LDL receptor-deficient and apoE-deficient mice, diet-induced or spontaneous hypercholesterolemia and atherogenesis are much greater. To determine whether lesion formation in the aortic origin, where particular pathogenic conditions may exist, correlates with lesion formation throughout the entire aorta, we measured the extent of atherosclerosis in both areas in 8 apoE- and 11 LDL receptor-deficient mice fed cholesterol-rich diets for 3-6 months, as well as in 9 C57BL/6 mice fed an atherogenic diet for a year, using two different morphometric methods. Both apoE-deficient and LDL receptor-deficient mice developed extensive lesions throughout the aorta, and in these models a significant correlation was observed between the extent of lesions in the entire aorta (measured as percent of surface area) and that at the aortic origin (measured as averaged lesion area per cross-section) (r = 0.77, P < 0.0001). In contrast, the plasma cholesterol levels achieved in C57BL/6 mice were much lower, and atherosclerotic lesions were found almost exclusively in the aortic origin. These results demonstrate that in murine models developing extensive aortic lesions, both morphometric methods provide valid and complementary information on the degree and distribution of atherosclerosis, and suggest that under severe atherogenic conditions lesion formation throughout the aorta is determined by the same pathological factors, in each model. Comparison of the extent of atherosclerosis in the entire aorta between genders also showed that male LDL receptor-deficient mice had significantly more lesions than females (29.2 vs. 14.8%, P < 0.005, n = 16). A similar trend was also seen in apoE-deficient mice.

Oxidation of LDL plays all important role in atherogenesis. The lag-time in the formation of conjugated dienes provides a sensitive measure of the resistance of plasma LDL to oxidation and is widely assumed to be an indicator of atherogenic risk. To test this assumption, we investigated whether different antioxidants yielding similar lag-times result in similar reduction of atherosclerosis. A 6-months intervention study was carried out in three groups of 10 LDL receptor-deficient rabbits each. Because previous studies indicated that antioxidants that reduce atherosclerosis resulted in very long lag-times, the first group was treated with an antioxidant combination containing 1,000 IU vitamin E, 0.05% probucol analogue (BM15.0639), and 0.025% probucol. The lag-times achieved throughout the intervention period by this combination (952 +/- 39 min) were matched in a second group (829 +/- 46 min) treated with a variable dose of probucol (0.0575-0.11%, average 0.091%). A third, untreated group served as a control. Plasma cholesterol levels of all groups were matched. Even though both treatments yielded similar antioxidant protection of plasma LDL, 0.091% probucol reduced aortic atherosclerosis by 51.7% compared to the untreated group (P < 0.005), whereas the antioxidant combination failed to reduce lesion formation. Thus, the lag-time is clearly not correlated with the antiatherogenic efficacy of different antioxidants. However, a weak correlation was found within the group treated with probucol only. Our results suggest that the degree of antioxidant protection of plasma LDL may not be a good indicator of the atherogenic risk, in general, and that probucol reduces atherogenesis by mechanisms not shared by all antioxidants.

Oxidised low-density lipoprotein (Ox-LDL) has been associated with arterial foam-cell formation, and autoantibodies to Ox-LDL are present in human serum. Lipid peroxidation is enhanced in pre-eclampsia. We assessed whether the titre of IgG autoantibody to an epitope of Ox-LDL, malondialdehyde-conjugated low-density lipoprotein (MDA-LDL), was increased in the sera of pre-eclamptic patients. 16 such patients had significantly higher mean titres of autoantibodies to MDA-LDL than healthy pregnant women (p = 0.028). In a multiple regression model, pre-eclamptic patients still had a significantly higher mean titre (p = 0.048). Enhanced lipid peroxidation may be involved in the foam-cell formation of decidua and in the pathogenesis of pre-eclampsia.

The engineering of mice that express a human apoB transgene has resulted in animals with high levels of human-like LDL particles and through crosses with human apo(a) transgenics, high levels of human-like lipoprotein (a) (Lp[a]) particles. In this study, these animals have been used to compare the atherogenic properties of apo(a), LDL, and Lp(a). The presence of the high expressing apoB (apoBH) transgene was associated with a 2.5-fold increase in VLDL-LDL cholesterol (primarily in the LDL fraction) and a 15-fold increase in proximal lesions compared with non-transgenic mice (P < or = 0.0001), while the presence of the low expressing human apoB (apoBL) transgene was not associated with major changes in lipoprotein profiles or increases in aortic lesion size. Examination of aortas of apoBH mice demonstrated lesions along the entire length of the aorta and immunochemical analysis of the lesions revealed features characteristically seen in human lesions including the presence of oxidized lipoproteins, macrophages, and immunoglobulins. Unlike animals with the apoBL transgene, animals with the apo(a) transgene had significant increases in proximal aortic fatty streak lesions compared to nontransgenic control animals (threefold; P < 0.02), while animals with both transgenes, the apo(a)/apo BL double transgenics, had lesions 2.5 times greater than animals expressing the apo(a) transgene alone and eightfold (P < 0.0006) greater than nontransgenic animals. These murine studies demonstrate that marked increases in apoB and LDL resulted in atherosclerotic lesions extending down the aorta which resemble human lesions immunochemically and suggest that apo(a) associated with apoB and lipid may result in a more pro-atherogenic state than when apo(a) is free in plasma.