In vivo insulin effect on ATPase activities in erythrocyte membrane from insulin-dependent diabetics.

Na+-K+-dependent ouabain-sensitive ATPase and Mg2+-ATPase have been assayed in the erythrocyte membranes of control subjects and in uncontrolled type I (insulin-dependent) diabetics. A decrease in Na+-K+-ATPase activity was observed in the patients that was significantly correlated with glycemia. The Mg2+-ATPase was increased moderately, and no correlation with glycemia was found. To study the in vivo effect of insulin, ATPase activities were measured in uncontrolled diabetics before and after a 24-h continuous insulin perfusion administered by means of an artificial pancreas. ATPase activities were corrected after normalization of glycemia. It therefore seems that glycemia and/or insulinemia are involved in the regulation of erythrocyte Na+-K+ ouabain-sensitive ATPase and to a lesser extent in that of Mg2+-dependent ATPase.

Abnormal taurocholate ileal transepithelial transport in atherosclerotic mini-pigs and effects of ACE inhibitors.

In atherosclerotic mini-pigs, we attempted to determine (i) whether high-fat atherogenic diet disturbs the taurocholate transepithelial transport and incorporation in the ileal epithelium mounted in Ussing chambers, and (ii) whether these processes are sensitive to angiotensin converting enzyme (ACE) inhibitors which slow the development of vascular atherosclerosis. In atherosclerotic mini-pigs, the mucosal to serosal transepithelial fluxes were markedly lower (72% inhibition) and free diffusion was more altered than active processes. Taurocholate incorporation into enterocyte (75% inhibition) paralleled the flux reduction. The transport disturbance observed here might be explained by changes in bile salt permeability in relation to alterations of the membrane properties. Taurocholate absorption was lowered by atherogenic diet, whereas bile salts were not trapped in the enterocyte, therefore atherosclerosis-induced alterations preferentially affected the passage through the brush-border. In the ACE inhibitor treated atherosclerotic mini-pigs, perindopril and enalapril similarly inhibited serum ACE activities. Perindopril further corrected taurocholate fluxes by 50% and fully restored taurocholate incorporation. Since enalapril did not restore the atherosclerosis-induced alterations, the involvement of intestinal ACE in bile acid recycling and of an ACE inhibitor class effect on these mechanisms both remain to be ascertained.

Human hepatic uptake of two vinca alkaloids: navelbine and vincristine.

A human liver plasma membrane model for the evaluation of the specific binding and transport processes of drugs presenting high hepatic clearance such as vinca alkaloids was developed. Uptake of the two structural antitumor analogs, navelbine (NVB) and vincristine (VCR), which exhibit wide variabilities in their respective pharmacokinetic parameters and antitumor spectra, was investigated. The high yield, the enzymatic profile and the retention of physiologic transport capacities, as demonstrated by taurocholate uptake, revealed that this membrane preparation was well suited for studies of hepatic drug transport systems. For both drugs two distinct processes were observed: mainly membrane binding and transport. NVB was found to bind to the membrane vesicles more intensively than VCR, but the transport processes were almost identical. However only NVB uptake seems to involve Na(+)-dependent processes. These significant differences may be related to the respective lipophilicity of the drugs. The more lipophilic molecule (NVB) presents the highest uptake, which is presumably at the origin of its greatest distribution volume in vivo.

Intestinal absorption of drugs: digitalis binding and transport by brush-border membrane vesicles from human duodenum.

The intestinal epithelial cell layer is the first major barrier to absorption encountered by xenobiotics. An understanding of the mechanism and sites of drug absorption and metabolism is thus a critical first step in developing orally active compounds. In this context human brush-border membrane vesicles obtained from multi-organ donor intestines have been purified. This model has been validated and used to investigate the duodenal absorption of drugs "in vitro", namely digitalis. It is well established that digitalis compounds present great variability in their respective "in vivo" bioavailability in human (60-90% for digoxin, 0% for ouabain). These particular characteristics prompted us to determine whether this membrane model constitutes a suitable tool in predicting the bioavailability or intestinal transport processes of these molecules. The uptake of [3H] digoxin and [3H] ouabain by membrane vesicles incubated in media of increasing osmolarities demonstrated that: i/two factors are involved in the uptake processes of digoxin: membrane binding and intravesicular transport (osmotic sensitive), ii/ for ouabain, no osmotic sensitivity was observed, indicating that no transport process occurred, but only membrane binding processes. These results are in complete agreement with the absolute bioavailability data reported for man "in vivo". This human brush-border model constitutes an interesting approach to the intestinal absorption phenomena which are known to be among the factors influencing the bioavailability of orally administered drugs.

[The effect of pentoxifylline and propentofylline on the membrane fluidity of red blood cells in uncontrolled insulin-dependent (type 1) diabetic patients]. / Etude de l'effet de la pentoxifylline et de la propentofylline sur la fluidité membranaire de globules rouges de sujets diabétiques insulino-dépendants (type 1) déséquilibrés.

Fluorescence polarization of red blood cell (RBC) membranes evaluated using DPH was measured in patients with uncontrolled insulin-dependent diabetes mellitus and in controls. The effect of in vitro addition of pentoxifylline and propentofylline (10(-5) and 10(-4) M) was studied. The fluorescence polarization parameter (P) was lower in the diabetic patients (p = 0.20 +/- 0.03, n = 8) as compared to the controls (p = 0.28 +/- 0.02), n = 8), reflecting an increase in probe mobility in the hydrophobic environment in the depth of the double lipid layer. In vitro addition of pentoxifylline had no effect on the fluorescence parameter of RBC membranes from controls, whereas both concentrations of pentoxifylline studied (p = 0.24 +/- 0.03) significantly increased the fluorescence parameter of RBC membranes from diabetics. Propentofylline had no effect. These findings suggest that the active site responsible for improved membrane fluidity of RBC from diabetics is located on the methyl radical present in the pentoxifylline molecule but not in the propentofylline molecule.

Ethynylestradiol alters lipid composition and phosphatidylethanolamine metabolism in red blood cells.

Treatment of female rats with ethinylestradiol at a dose of 60 micrograms/rat, daily for 21 days, produced marked changes in red blood cell lipids. Cholesterol was decreased by 22% and total phospholipids were increased by 13%, resulting in a 31% decrease in the cholesterol to phospholipid ratio. The mass distribution of phosphatidylcholine and phosphatidylethanolamine relative to total phospholipids was unchanged. Whereas control red cells incorporated preferentially fatty acids in phosphatidylcholine, ethinylestradiol stimulated their incorporation specifically in phosphatidylethanolamine, where increases occurred with palmitic acid (+75%), oleic acid (+68%) and arachidonic acid (+31%). Incorporation in phosphatidylcholine was unaffected with any of the 3 fatty acids. The stimulation of fatty acid incorporation in phosphatidylethanolamine is likely to reflect an estrogen-dependent increase in turnover rate of fatty acids in this phospholipid. Such alterations in lipid composition and fatty acid incorporation in red cell phospholipids may have significant effects on membrane function.

Correction by insulin added in vitro of abnormal membrane fluidity of the erythrocytes from type 1 (insulin-dependent) diabetic patients.

Filtrability of erythrocytes obtained from uncontrolled Type 1 (insulin-dependent) diabetic patients is abnormal, but is corrected by insulin added in vivo or in vitro. As erythrocyte filtrability depends on several determinants, we chose to study a membrane property of erythrocytes from diabetic subjects. Membrane fluidity was studied by fluorescence polarization using a lipophilic probe, the diphenyl-hexatriene and the Coulter Epics V together with a laser Spectra-physics 2000. Fluorescence polarization values obtained for 31 normal subjects (0.253 +/- 0.043 SD) and 31 uncontrolled Type 1 diabetic patients (0.231 +/- 0.043 SD) were significantly different (p less than 0.01). Insulin (2.5.10(-9) mol/l) added in vitro increased the fluorescence polarization values of red cell membranes from diabetic patients (without insulin, fluorescence polarization values = 0.210 +/- 0.032 SD; with insulin, fluorescence polarization values = 0.253 +/- 0.024 SD, p less than 0.001, n = 15), but had no effect on normal membranes (without insulin fluorescence polarization values = 0.255 +/- 0.037 SD, with insulin, fluorescence polarization values = 0.251 +/- 0.026 SD; n = 12). Given a relationship between the lipid bilayer and membrane cytoskeleton proteins, this insulin-correctable abnormality of erythrocyte membrane fluidity may be an important determinant of the rheological behaviour of erythrocytes from diabetic patients.

Correction by pentoxifylline of the abnormal fluorescence polarization of erythrocyte membranes from diabetic patients.

Erythrocytes from diabetic patients show abnormal rheology. Pentoxifylline, a methylxanthine, improves the abnormal deformability of diabetic erythrocytes, but its mechanism of action remains unclear. We have studied the effect of pentoxifylline on the lipid order of erythrocyte membranes from controls and patients with Type I diabetes. We studied the structural organization of membrane lipids in individual erythrocyte ghosts by fluorescence polarization using a cell sorter. Fluorescence polarization values (P) for 17 controls (P = 0.244) and 20 diabetic patients (P = 0.215) were significantly different. Pentoxifylline added in vitro had no effect on normal membranes, but significantly increased at 10(-5) mol X l-1 (P = 0.233), and normalized at 10(-4) mol X l-1 (P = 0.243), the P value of membrane ghosts from diabetics.