Julian B. Marsh

Hepatic Secretion of Lipoproteins in the Rat and the Effect of Experimental Nephrosis

Livers from normal and nephrotic rats were perfused by the nonrecirculating technique. Nephrosis was studied on the 7th d after the injection of puromycin animonucleoside. Amino acid-labeled lipoproteins (d < 1.21) were isolated from the perfusion medium by agarose column chromatography or by sequential density ultracentrifugation. In both groups of animals, in addition to very low density lipoproteins and nascent high density lipoproteins, column chromatography revealed the presence of a peak of 2-3 x 10(6) daltons. This peak contained lipoproteins of densities corresponding to <1.006, 1.006 < d < 1.02, and 1.02 < d < 1.06, which indicated that rat liver secretes a heterogeneous mixture of triglyceride-rich lipoproteins. The amount of these lipoprotein density classes was measured and their lipid and apoprotein composition and their apoprotein specific activity were determined. In both groups of rats there was a progressive rise in phospholipid and decrease in triglyceride content as the isolation density increased from 1.006 and 1.06. The lipoproteins from the nephrotics had higher amounts of cholesterol. The livers from the nephrotic rats secreted two to three times as much lipoprotein as controls in all density classes in the first 20 min, but during the next 40 min only the 1.02 < d < 1.06 and nascent high density lipoproteins remained at this high level compared to controls. A larger total liver pool of apolipoproteins in nephrotic livers was inferred from their lower specific activities during the first 20 min. The apoprotein composition of liver perfusate lipoproteins from nephrotics differed from controls. There was a 40% decrease in the amount of low molecular weight apoproteins in all density classes, with corresponding increases in apo B and apo E in the triglyceride-rich fractions. The apo A-1 content of nascent HDL was increased from 16% in controls to 52% in nephrotics, with corresponding decreases in apo C and apo E. When these results were combined with specific activity measurements of the individual apoproteins and the net secretion rate of total protein in each lipoprotein class, it was possible to estimate the total amount of each apoprotein secreted and the total incorporation of labeled amino acids into each. The incorporation of label gave results similar to those obtained by direct measurement of the amounts of apoproteins. Apo E secretion was increased by a factor of 1.8, apo B by 2.8, and apo A-1 by 8.4, whereas the secretion of apo C was not significantly altered. We explain these results by postulating that the primary stimulus to hepatic plasma protein synthesis in response to proteinuria is general and that subsequent negative feedback regulation affects individual apolipoprotein synthesis rates. A corollary of this hypothesis is that the biosynthesis and secretion of an apoprotein may be regulated independently of the lipoprotein density class in which it is found.

Effects of eicosapentaenoic and docosahexaenoic acids on apoprotein B mRNA and secretion of very low density lipoprotein in HepG2 cells.

Oleic acid (18:1n-9, OA), docosahexaenoic acid (22:6n-3, DHA), or eicosapentaenoic acid (20:5n-3, EPA) was added to HepG2 cells at a concentration of 1 mM in a 5:1 or 2:1 molar complex with bovine serum albumin (BSA), and this was incubated for 3 hours. The incorporation of 3H-glycerol into cellular and medium triglyceride (TG), and the mass of TG were measured. The effects of these fatty acids on the secretion of very low density lipoprotein (VLDL) apolipoprotein B (apo B) were estimated from the incorporation of 3H-leucine into the medium apo B in comparison to cells incubated with fatty acid-poor albumin. The secretion of human albumin by the cells was also estimated by immunochemical precipitation of the labeled albumin. In addition, the intracellular levels of apo B messenger ribonucleic acid (mRNA) were measured by the dot-blot hybridization technique. Relative to control cells incubated with BSA, OA (complexed to BSA at a 5:1 molar ratio) stimulated TG synthesis and secretion sevenfold. Compared to OA, EPA was 24% less effective for both processes, whereas DHA inhibited only the secretion of TG (-43%). The secretion of VLDL apo B was not affected by OA, but was decreased 31% by EPA and 54% by DHA. When the molar ratio of fatty acid complexed to albumin was changed to 2:1, similar results were obtained with respect to TG production. The levels of apo B mRNA relative to actin mRNA were not significantly altered by any of the fatty acids.(ABSTRACT TRUNCATED AT 250 WORDS)

Catabolism of very low density lipoproteins in experimental nephrosis.

The effects of experimental nephrosis in rats, produced by puromycin aminonucleoside, include an elevation of plasma levels of all lipoprotein density classes and the appearance of high density lipoprotein (HDL) rich in apoprotein (apo) A-I and deficient in apo A-IV and apo E. The hyperlipoproteinemia is associated with an increase in hepatic synthesis of lipoproteins. The possible role of decreased very low density lipoprotein (VLDL were obtained from nonfasting animals by ultracentrifugation at d 1.006 and included chylomicrons) catabolism and its relationship to the apolipoprotein composition of nephrotic high density lipoproteins (1.063 less than d less than 1.210, or 1.072 less than d less than 1.210 [HDL]) was explored. When 125I-VLDL was injected, the faster plasma clearance of lower molecular weight apolipoprotein B (apo BL) compared with that of higher molecular weight apo BH which is seen in normal rats was not observed in nephrotic rats. Less labeled phospholipid, apo C, and apo E were transferred from VLDL to higher lipoprotein density classes. Heparin-releasable plasma lipoprotein lipase and hepatic lipase activities were decreased by 50% in nephrotic rats compared with pair-fed controls. Perfusion of livers with medium that contained heparin released 50% less lipase activity in nephrotic rats than in controls. When heparin was injected intravenously, significant decreases in plasma levels of triglycerides and significant increases in levels of free fatty acids were observed in both groups of animals. In the nephrotic rats, 86% of the free fatty acids were in the lipoprotein fractions, as compared with 16% in the controls. Heparin treatment did not restore to normal the decreased apo BL clearance in nephrotic rats but it produced an increased amount of apo A-IV and apo E in the plasma HDL. In vitro addition of partially pure lipoprotein lipase to whole serum from nephrotic rats significantly increased the content of apo E in HDL. We conclude that the abnormal apoprotein composition of HDL in experimental nephrosis is the result of altered entry of apolipoproteins from triglyceride-rich lipoproteins, probably because of decreased lipolysis.

Lipoproteins in experimental nephrosis: plasma levels and composition.

Experimental nephrosis was induced in rats by administration of puromycin aminonucleoside and the levels of plasma lipoproteins were examined 7 days later and compared to controls. As determined by density ultracentrifugation, VLDL, IDL, LDL, and HDL protein levels were increased by 8, 4, 5, and 5 times, respectively. These increases were accompanied by changes in lipid and apoprotein composition. The VLDL, IDL, and LDL fractions contained less triglyceride and more phospholipid and cholesterol, while HDL lipid composition was not altered. The apoprotein composition of VLDL and IDL were not measurably altered, but LDL contained less apoE. HDL had a markedly abnormal composition characterized by an almost complete absence of apoA-IV and apoE, increased apoA-1, and decreased apoC. While increased hepatic synthesis can account for much of the observed hyperlipoproteinemia in nephrosis, the changes in lipoprotein composition suggest impaired catabolism as a contributory factor.

Inhibition of apolipoprotein secretion and phosphatidate phosphohydrolase activity by eicosapentaenoic and docosahexaenoic acids in the perfused rat liver

Infusion of albumin-bound eicosapentaenoic acid (EPA) docosahexaenoic acid (DHA), or oleic acid (OA) in perfused rat livers was carried out for two hours at a rate that maintained the perfusate concentration at 1 mmol/L. When compared with fatty acid-poor albumin alone, triacylglycerol (TAG) output was not significantly increased with DHA or EPA, whereas OA infusion resulted in a twofold increase. Incorporation of labeled leucine into VLDL apo B-100, apo B-48, apo E, and apo Cs was decreased by 50% by DHA or EPA compared with OA. The total phosphatidate phosphohydrolase activity was decreased by 35% with DHA or EPA compared to oleic acid or albumin alone. In no case was there a significant change in the distribution of activity between the microsomal and cytosolic factions. Fatty acid infusion did not significantly change the liver TAG content. Total liver lipids, microsomal lipids, and lipids of secreted VLDL were enriched with the infused fatty acids. The degree of enrichment for secreted TAG averaged 24% for OA and 36% for DHA or EPA. The effects of DHA and EPA on PPH activity and on apo B secretion in feeding experiments with marine oils rich in these acids may relate to changes in the fatty acid composition of liver membranes.

Characterization of a bile salt-dependent cholesteryl ester hydrolase activity secreted from HepG2 cells

HepG2 cells and medium were assayed for cholesteryl ester hydrolase (CEH) activity in the presence and absence of sodium cholate. Although bile salt-dependent CEH activity was measured in the medium at 6 to 96 h (up to 4500 pmol/h per mg cell protein), there was very little activity detected in the corresponding cell homogenates (less than 70 pmol/h per mg cell protein). Activity in the medium was expressed only in the presence of trihydroxy bile salts and was maximal at 40 mM cholate and pH 7.5. Incubation of HepG2 cells with brefeldin A resulted in an 80 to 90% inhibition of secretion of the bile salt-dependent CEH activity, while only inhibiting total protein secretion by 42%. Bile salt-dependent CEH activity could also be detected in rat liver perfusates. Although there was measurable activity in all of 14 livers analyzed (47 +/- 10 and 53 +/- 17 nmol/h per g liver per h perfusion during two 5-min collections after 15 and 30 min of perfusion, respectively), it did not correlate with the activity found in corresponding liver homogenates, as only four livers had detectable bile salt-dependent CEH activity. These results provide evidence for the secretion of a bile salt-dependent CEH activity, from both a hepatic cell line and the intact liver, that has similar properties to the enzyme previously isolated from rat liver homogenates and rat pancreas.

THE EFFECT OF SERUM FROM INSULIN-RESISTANT CASES ON THE COMBINATION OF INSULIN WITH THE RAT DIAPHRAGM

Insulin resistance is currently believed to be associated with insulin-neutralizing substances in the blood. Lowell (1), employing the mouse convulsion assay for insulin, has shown that the injection of mixtures of insulin and serum from insulin-resistant diabetic patients into mice caused a significantly smaller incidence of convulsive symptoms than the injection of the same amounts of insulin alone. He points out, however, that the quantitative interpretation of these experiments is difficult because of the variations of the response of the animals to insulin and the nonspecific effects of the injection of serum. It would be desirable, therefore, to employ a method for the determination of insulin-inactivating properties of serum which does not involve the administration of insulin-serum mixtures to the intact animal. In the course of our recent work on the chemical combination of insulin with the rat diaphragm (2, 3) it became apparent that this technique would provide a method for the detection of insulin-inactivating substances in any solution. Accordingly, we have tested serum from normal, diabetic, and insulin-resistant subjects and found, in confirmation of the results of Lowell, that serum from insulin-resistant patients has the property of opposing the action of insulin.

Catabolism of the apolipoproteins of HDL in control and nephrotic rats

125I-labeled high density lipoprotein (HDL) from control rats, or from rats made nephrotic by puromycin aminonucleoside, was injected into control or nephrotic rats. At 5 and 20 h, the amount and distribution of label remaining in apolipoproteins HDL, A-I, E, A-IV and the C apolipoproteins was measured after ultracentrifugal isolation of HDL and SDS-polyacrylamide gel electrophoretic separation of each apolipoprotein. There were no significant differences in the removal rates of apolipoprotein HDL or of the individual apolipoproteins when the removal of HDL of controls was compared to HDL of nephrotics. HDL from nephrotic rats contains less than 10% of either the apolipoprotein A-IV or apolipoprotein E content of control HDL, indicating that neither apolipoprotein A-IV nor apolipoprotein E play a significant role in determining the catabolic fate of rat HDL. In severely nephrotic animals the apoliprotein C content of HDL was reduced to 50% of control values and the apolipoprotein A-I content of HDL rose to 87% of the total apolipoprotein. The individual apolipoproteins of HDL from either nephrotics or controls were catabolized at the same rates irrespective of the degrees of nephrosis or altered HDL apolipoprotein composition. The apparent fractional catabolic rates for apolipoprotein HDL and for each of the apolipoproteins, determined after 20 h, did not differ from one another, and all were reduced by half in the nephrotic rats compared to the normal controls. These results support the concept that HDL is catabolized as a particle mediated by apolipoprotein A-I recognition, and they reinforce earlier work indicating that increased synthesis is the dominant factor responsible for increased plasma HDL concentrations in experimental nephrosis.

The effect of fasting on the secretion of lipoproteins and two forms of apo B by perfused rat liver.

Abstract Rat livers from animals fed ad lib or fasted 24 hr were perfused by the single pass technique with 3H-amino acids and label incorporation into lipoprotein apoproteins was measured. The lipoproteins were separated into very low density lipoproteins (d < 1.006), low-density lipoproteins (1.006 < d < 1.06), and high-density lipoproteins (1.06 < d < 1.21) by ultracentrifugation and the apoproteins of each density fraction were analyzed by gel filtration column chromatography in the presence of sodium dodecyl sulfate. The apoprotein fractions included apo Bh (higher-molecular-weight apo B), apo B1 (lower-molecular-weight apo B), apo E, and apo C. Fasting decreased total hepatic apoprotein secretion from 75 to 53 μg·g−1·hr−1 and label incorporation into the apoproteins of very low density lipoproteins was reduced by 48% with the greatest reduction in both apo B fractions. There was a redistribution of apo Bh label into the low-density lipoproteins with fasting resulting in secretion of lipoproteins enriched in apo Bh. The high-density lipoprotein fraction contained 27 and 41% of the apo B1 label isolated from liver perfusates of fasted and fed animals, respectively. The nascent high-density lipoprotein was enriched in labeled apo B1 compared to low-density lipoproteins indicating that apo B in this density range represents a distinct metabolic fraction. With fasting the apo Bh and apo B1 apoproteins behave independently in terms of the distribution of label incorporation into nascent hepatic lipoproteins.

Catabolism of apoprotein A-1 of HDL in normal and nephrotic rats☆

Abstract High density lipoprotein was isolated from the plasma of control and puromycin aminonucleoside nephrotic rats and labeled with 125I. It was injected into control and nephrotic rats and the plasma analyzed after 3 min, 5 hr, and 20 hr. High density lipoprotein was isolated and the specific activity of apo A-I determined following apoprotein separation using SDS-polyacrylamide gel electrophoresis. The distribution of labeled apoproteins was determined in whole plasma by SDS-gel filtration column chromatography and the plasma concentration of apo A-I was calculated from its specific activity and the total plasma apo A-I radioactivity. A 3 min to 5 or 20 hr fractional catabolic rate was calculated. When multiplied by the plasma concentration of apo A-I, an estimate of the absolute catabolic rate was obtained. When injected into normal animals, the high density lipoprotein apo A-I had similar catabolic rates whether derived from control or nephrotic rat plasma, averaging 65 and 48 μg of apoprotein per ml of plasma per hr at 5 or 20 hr, respectively. Labeled HDL was injected into nephrotic rats of varying degrees of severity. The moderately nephrotic rats with plasma cholesterol levels averaging 177 mg/dl had apo A-I levels that were 3.6 times that of controls (1799 ± 195 μg/ml) and 2.4-fold increases in apo A-I catabolic rates (134 ± 28.9 μg/ml plasma/hr). The severely nephrotic rats with cholesterol concentrations averaging 396 mg/dl had apo A-I levels 7.1 times that of the controls (3533 ± 220 μg/ml) while the catabolic rate was 2.7 times the control rate (153 ± 19.5 μg/ml/hr), which was not a significant increase beyond that of the moderately nephrotic group. It was concluded that compositional differences of HDL resulting in an increased proportion of apo A-I, as in nephrotic rat plasma, do not affect apo A-I metabolism. The high levels of apo A-I in the plasma of nephrotic rats is due to increased hepatic synthesis that results in an expansion of the pool size and saturation of catabolic pathways. Small increases in apo A-I synthesis lead to large increases in the plasma concentration, an observation that may be important in the regulation of HDL levels that are known to be correlated with a decreased incidence of atherosclerosis.