Search

A new kindred with asymptomatic hypobetalipoproteinemia is reported. The proband, age 67, differs from previously described cases in several respects: (a) unusually low levels of low density lipoprotein (LDL) cholesterol (4-8 mg/dl); (b) normal triglyceride levels; (c) low levels of high density lipoprotein; (d) mild fat malabsorption; and (e) a defect in chylomicron clearance. On a high-carbohydrate diet his plasma triglyceride levels, instead of rising, actually fell. Turnover of triglycerides in very low density lipoproteins (VLDL) was low (2.8 mg/kg per h). Fractional catabolic rate of LDL protein was just above the normal range (0.655/d) but net turnover was <10% of normal (0.65 mg/kg per d). The half-life of his chylomicrons was 29 min, five times the normal value. Postheparin lipoprotein lipase activity was normal and apolipoprotein C-II, the activator protein for lipoprotein lipase, was present and functional. Apolipoprotein C-III(1), however, was not detected in the VLDL fraction, a finding previously reported in patients with abetalipoproteinemia. Fecal excretion of cholesterol was almost twice normal; total sterol balance was increased by congruent with40%. The unusual features in the proband that distinguish him from previously described cases and from his affected first-degree relatives suggested that, in addition to the basic gene defect affecting LDL metabolism, he might have a second abnormality affecting clearance of chylomicrons and VLDL. The ratio of apolipoprotein E(3) to E(2) in his VLDL fraction was 0.93, just below the lower limit of normal, suggesting heterozygosity for E(3) deficiency. Whether or not this contributes to his hypertriglyceridemia remains to be established.

The effect of insulin on hepatic triglyceride synthesis and secretion is controversial. Previously, we have described a cell culture system of adult rat hepatocytes that synthesize and secrete very low density lipoprotein (VLDL) triglycerides with small and irreproducible effects of insulin on triglyceride metabolism. To study the primary effects of insulin on hepatic triglyceride metabolism a method was developed utilizing fibronectin-coated culture dishes that allowed adhesion, spreading, and maintenance of hepatocytes for 2-3 d in the absence of serum and insulin. This culture system allowed mass measurements of both cellular and secreted VLDL triglycerides for long time periods after the addition of physiological concentrations of insulin to hormone-free culture medium. In the absence of insulin and after an initial 4 h in culture, the medium was replenished and triglyceride mass was measured at the end of 18-h incubations. VLDL triglyceride accumulated in the culture medium at a linear rate over this time-course with increasing accumulation as the medium glucose concentration was raised from 2.5 to 25 mM glucose (1.77+/-0.24 to 3.09+/-0.76 mug triglyceride/mg cell protein per h). There was no apparent significant lipolysis or hepatocellular reuptake of secreted VLDL triglycerides. In the absence of insulin cellular triglyceride levels were unchanged between 3 and 24 h in culture while insulin (50-500 muU/ml) significantly increased cellular triglyceride content at all glucose concentrations tested (0-25 mM). The addition of insulin to the culture medium progressively reduced the rate of VLDL triglyceride secretion accompanied by an increase in cellular triglyceride at insulin concentrations > 50 muU/ml. Most or all of the observed increase in cell triglyceride content could in all experiments be accounted for by the insulin-induced inhibition of VLDL secretion. Incorporation of [2-(3)H]glycerol into cellular and VLDL triglycerides as a function of insulin concentration was also measured. Glycerol incorporation data at 20-22 h after plating of the cells closely paralleled the insulin-induced changes in cellular and VLDL triglyceride as determined by mass analysis. The observed effects of insulin occurred at concentrations close to the physiological range and suggest that the direct hepatic effect is to suppress VLDL secretion although the net effect in vivo will clearly reflect many additional accompanying changes.

A partially purified hormone-sensitive triglyceride lipase of human adipose tissue was found to be activated twofold by the addition of cyclic 3',5'-AMP, ATP, and magnesium ions. Lipase activities against diolein and monoolein were not affected. Addition of protein kinase inhibitor at zero time completely inhibited activation, and this inhibition was prevented by prior addition of an excess of exogenous protein kinase (from rabbit skeletal muscle). Addition of protein kinase inhibitor during the activation step blocked the activation process without a time lag, suggesting that protein kinase operates directly on hormone-sensitive lipase. Further purification yielded a fraction free of protein kinase, and lipase activation in this fraction depended absolutely on addition of exogenous kinase. Incubation of human fat with epinephrine or isoproterenol stimulated lipolysis and caused conversion of nonactivated hormone-sensitive lipase to its activated form, as indicated by a decrease in the activation subsequently obtainable in fractions prepared from such hormone-treated tissues. These findings strongly suggest that the stimulation of lipolysis by hormonal treatment is the consequence of the activation of hormone-sensitive triglyceride lipase by cyclic 3',5'-AMP-dependent protein kinase.

High density lipoprotein (HDL) inhibited the binding (trypsin-releasable radioactivity), internalization (cell-associated radioactivity after trypsinization), and degradation (TCA-soluble non-iodide radioactivity) of (125)I-low density lipoprotein ((125)I-LDL) by cultured normal human fibroblasts. At HDL:LDL molar ratios of 25:1 (protein ratios about 5:1), these parameters were reduced by about 25%. Unlabeled LDL was about 25 times more effective in reducing (125)I-LDL binding, implying that if HDL and LDL bind at common sites the affinity of HDL for these sites is very low or that the interaction is on some other basis. The fractional reduction in (125)I-LDL binding at a given HDL: (125)I-LDL ratio was independent of (125)I-LDL concentration and occurred equally with fibroblasts from a subject with homozygous familial hypercholesterolemia. Reciprocally, the binding, internalization, and degradation of (125)I-HDL were reduced by LDL. Preincubation of fibroblasts with HDL (or LDL) reduced the subsequent binding of (125)I-LDL (or (125)I-HDL) during a second incubation. In other studies HDL reduced the net increase in cell cholesterol content induced by incubation with LDL. HDL alone had no net effect on cell cholesterol content. These findings suggest that HDL reduces both the high affinity and the low affinity binding of LDL to human fibroblasts and that this in turn reduces the internalization and degradation of LDL. The effect of HDL on the LDL-induced changes in cell cholesterol content could be in part on this basis and in part on the basis of an HDL-stimulated release of cholesterol from the cells. These effects of HDL in vitro may be relevant to the negative correlations reported from in vivo studies between plasma HDL concentration and both body cholesterol pool size and the prevalence of clinically manifest atherosclerosis but further studies will be needed to establish this.

Measurements of transport of triglycerides (TG) in very low density lipoproteins (VLDL) were carried out in 59 patients by injection of radioactive glycerol, determinations of specific activities of VLDL-TG for 48 h thereafter, and treatment of the data by multicompartmental analysis. The patients were divided into three groups: normal weight (89-120% ideal weight), mildly obese (120-135% ideal weight), and markedly obese (135% ideal weight). They had varying levels of VLDL-TG ranging from normal to markedly elevated. In many subjects, there was a positive correlation between concentrations and transport of VLDL indicating that overproduction of VLDL-TG contributed to hypertriglyceridemia. In others, and particularly in several markedly obese subjects, transport rates were greatly increased without significant hypertriglyceridemia, suggesting that they had enhanced capacity to clear TG. In all groups, however, there were patients whose degree of hypertriglyceridemia seemed out of proportion to their transport rates. This finding and the fact that many patients have increased secretion of VLDL-TG without elevated plasma TG suggests that both overproduction of VLDL-TG and insufficient enhancement of clearance contributed to the development of hypertriglyceridemia.The data showed a poor correlation between transport rates determined by our multicompartment analysis and single-exponential analysis used previously by other investigators (r = 0.46); this comparison was not improved by segregating patients according to their degree of obesity. Although two conversion pathways (fast and slow synthetic paths) were required to fit the data, there was no correlation between transport rates and the ratio of the two pathways. Also, despite the known pathway of conversion of VLDL to low density lipoprotein, no correlation was found between VLDL-TG transport rates and estimated low density lipoprotein-cholesterol concentrations.

Three different multicompartmental models of free fatty acid (FFA) and very low density lipoprotein triglyceride fatty acid (VLDL-TGFA) transport in man are formulated from plasma FFA and VLDL-TGFA tracee and tracer data collected over a 24 hr interval after the injection of palmitate-(14)C. All modeling and data fitting were performed on a digital computer using the SAAM program. Structural differences in the three models relate to the position of the slowly turning over compartment required to generate the late portion of the plasma VLDL-TGFA tracer data. The positions of this slow compartment are along the hepatic pathway from FFA to VLDL-TGFA (model A) or in the distribution system of VLDL-TGFA (model B) or in the distribution system of FFA (model C). Although all three models are equally consistent with our experimental data and are supported by observations of others, each reveals inconsistency with some data obtained from the literature. Consequently, a combination model of FFA-TGFA transport, incorporating properties of models A, B, and C would be more consistent with all available data. Experiments that would help to determine the quantitative significance of each of the slow compartments in the combination model are suggested. Several other models suggesting recycling of plasma VLDL-TGFA through the plasma FFA pool, kinetic heterogencity of the plasma VLDL-TGFA pool, and contamination of plasma VLDL-TGFA radioactivity with low density lipoprotein (LDL) TGFA radioactivity were tested. The first model does not explain the late portion of the plasma VLDL-TGFA tracer data. The second and third models, while consistent with our tracee and tracer data, have steady-state implications with respect to the extent of kinetic heterogeneity and size of the LDL-TGFA contaminant that make them unlikely. Assumptions underlying other investigator's models of FFA and TGFA transport in man are reviewed within the logical framework of our models. Quantitative differences among the various models are shown by evaluating all of the models with respect to a common set of plasma FFA and VLDL-TGFA data.

We have developed a method for the rapid infusion into plasma of large amounts of long-chain free fatty acids (FFA). Unanesthetized dogs were connected by a peripheral artery to a closed, continuousflow centrifuge from which cells and plasma emerged in separate lines. Sodium oleate was infused directly into the plasma line before cells and plasma were recombined and returned to the animal through a peripheral vein.The centrifugation procedure itself produced only small changes in circulating levels of glucose, FFA, and electrolytes. Plasma flow rates as high as 100 ml/min could be maintained, and centrifugations of 12 hr were accomplished without complications. During centrifugation, sodium oleate was infused at rates up to 80 muEq/kg per min for 2.5 hr; the maximum molar ratio of FFA to albumin without hemolysis was 10:1. Plasma FFA levels rose rapidly after infusions were started and reached constant elevated levels within 15-20 min. Oleate infusion at 10-50 muEq/kg per min produced a rise in plasma FFA proportional to the infusion rate. The maximum increment in plasma FFA above control values was 1.66 muEq/ml. When infusions ended, plasma FFA declined rapidly to control levels. Oleate infusion at rates below 30 muEq/kg per min did not reduce levels of other plasma FFA. Infusion at high rates was accompanied by a marked fall in blood glucose. This method permits adminsitration of long-chain fatty acids in sufficient quantities to study their individual metabolic effects, and provides a new way to supply lipid calories parenterally.

The rate of oxidation of phytanic acid-U-(14)C to (14)CO(2) in three patients with Refsum's disease was less than 5% of that found in normal volunteers. In contrast, the rate of oxidation of alpha-hydroxyphytanic acid-U-(14)C and of pristanic acid-U-(14)C to (14)CO(2), studied in two patients, while somewhat less than that in normal controls, was not grossly impaired. These studies support the conclusion that the defect in phytanic acid oxidation in Refsum's disease is located in the first step of phytanic acid degradation, that is, in the alpha oxidation step leading to formation of alpha-hydroxyphytanic acid. The initial rate of disappearance of plasma free fatty acid radioactivity after intravenous injection of phytanic acid-U-(14)C (t(1/2) = 5.9 min) was slower than that seen with pristanic acid-U-(14)C (t(1/2) = 2.7 min) or palmitic acid-1-(14)C (t(1/2) = 2.5 min). There were no differences between patients and normal controls in these initial rates of free fatty acid disappearance for any of the three substrates tested. There was no detectable lipid radioactivity found in the plasma 7 days after the injection of palmitic acid-1-(14)C or pristanic acid-U-(14)C in either patients or controls. After injection of phytanic acid-U-(14)C, however, the two patients showed only a very slow decline in plasma lipid radioactivity (estimated t(1/2) = 35 days), in contrast to the normals who had no detectable radioactivity after 2 days. Incorporation of radioactivity from phytanic acid-U-(14)C into the major lipid ester classes of plasma was studied in one of the patients; triglycerides accounted for by far the largest fraction of the total present between 1 and 4 hr.

Transport of plasma-free fatty acids (FFA) and of fatty acids in triglycerides of plasma very low density lipoproteins (VLDL-TGFA) was studied in two normal subjects, five patients with type IV hyperlipoproteinemia, and two patients with type I hyperlipoproteinemia. After intravenous pulse-labeling with albumin-bound 1-palmitate-(14)C, specific radioactivity of plasma FFA and VLDL-TGFA were determined at intervals up to 24 hr. The results were analyzed using several different multicompartmental models each compatible with the experimental data. Fractional transport of VLDL-TGFA was distinctly lower (no overlap) in the type IV patients than in the control subjects, both on a usual balanced diet (40% of calories from carbohydrate) and on a high-carbohydrate diet (80% of calories). However, net or total transport of VLDL-TGFA in the type IV patients was not clearly distinguishable from that in the control subjects, there being considerable overlap on either diet. The results suggest that in this group of type IV patients the underlying defect leading to the increased pool size of VLDL-TGFA is not overproduction but a relative defect in mechanisms for removal of VLDL-TGFA. Since some of these type IV patients had only a moderate degree of hypertriglyceridemia at the time they were studied, and since it is not established that patients with the type IV phenotype constitute a biochemically homogeneous population, the present results should not be generalized. Four studies were done (in two control subjects and two type IV patients) in which the kinetic parameters in the same individual were determined on the balanced diet and on the high-carbohydrate diet. All subjects showed an increase in VLDL-TGFA pool size. Using two of the models for analysis, all showed an increase in net transport of VLDL-TGFA; using the third model, three of the four studies showed an increase in VLDL-TGFA transport. The results are compatible with the interpretation that the carbohydrate-induced increase in VLDL-TGFA, both in controls and type IV patients, is at least in part due to an increased rate of production of VLDL-TGFA. The magnitude of the increase was approximately the same in controls and patients. Thus, metabolic adjustment to a high-carbohydrate regimen in these type IV patients may not be basically different from that in normal controls; the higher levels of VLDL-TGFA reached may simply be another reflection of a defective removal mechanism. An alternative interpretation, compatible with the data, would involve both a carbohydrate-induced increase in fractional rate of release of VLDL-TGFA from liver to plasma and a decrease in fractional removal of VLDL-TGFA from plasma without increase in net production rate. The simpler hypothesis of a single primary effect on net VLDL-TGFA production from FFA seems more likely.

Plasma transport of free fatty acids (FFA) and triglyceride fatty acids (TGFA) was studied in seven subjects with normal lipid metabolism, one case of total lipodystrophy, and one case of familial hyperlipemia (Type V). Studies were carried out after intravenous injection of radioactive FFA, of lipoproteins previously labeled in vitro in the triglyceride moiety, or both. Computer techniques were used to evaluate a series of multicompartmental models, and a general model is proposed that yields optimum fitting of experimental data for both FFA and TGFA. The results show that as much as 20-30% of FFA leaving the plasma compartment in normal subjects is transported to an exchanging extravascular pool and quickly reenters the plasma pool as FFA. The rate of irreversible delivery of FFA from plasma to tissues averaged 358 muEq/min in normals. The lipodystrophy patient, despite the virtual absence of adipose tissue (confirmed at autopsy), had a plasma FFA concentration and a total FFA transport, both more than twice normal. Total TGFA transport ranged from 25 to 81 muEq/min in four normal controls. The rate constant for TGFA turnover in the patient with Type V hyperlipemia was so small that total transport could not be quantified from the data available; the TGFA half-life was over 500 min. In two normal subjects given injections of autologous lipoproteins labeled in vitro with triolein-(14)C and simultaneously given oleic acid-(3)H, it was shown that the time course for the disappearance of the TGFA in the in vitro labeled samples conformed almost exactly to that of the physiologically labeled lipoprotein TGFA synthesized from injected FFA (as evidenced by the simultaneous fitting of both sets of data using the same multicompartmental model and the same rate constants). Radioactivity appeared in the plasma FFA fraction at a significant rate after injection of plasma labeled in vitro with TGFA. It was estimated that as much as 50% of the total TGFA transported underwent rapid and rather direct conversion to FFA in the two normal subjects studied this way. The kinetic data suggest that such conversion of TGFA to FFA was not preceded by any extensive dilution, such as would result from complete mixing with tissue triglyceride stores.

The acute elevation of plasma free fatty acid (FFA) levels by direct infusion of sodium oleate into the plasma of conscious dogs was accompanied by the rapid onset of a 2- to 12-fold increase in plasma immunoreactive insulin, and, subsequently, a marked fall in plasma glucose, even in dogs receiving intravenous glucose throughout the infusion. The magnitude of both the insulin and glucose responses correlated with the mean FFA level during infusion. A large increase in plasma insulin and fall in glucose also occurred when glycerol was infused with oleate in order to simulate endogenous lipolysis more closely. Insulin levels in pancreaticoduodenal vein blood rose markedly during oleate infusion, while plasma ketone levels rose only slightly. In contrast to the effects of oleate infusion, elevation of plasma FFA to correspondingly high levels by triolein ingestion and intravenous heparin produced only small increases in plasma insulin, which did not correlate well with the FFA level reached, and small increases in plasma glucose.The results indicate that under certain conditions elevated FFA levels may be a potent stimulus of insulin secretion. This response is modified under other conditions such as during chylomicron removal under the influence of heparin. This effect may play a role in the regulation of lipolysis and ketone formation, but determination of the exact mechanism of FFA stimulation of the pancreas and its physiological significance will require further investigation.

Studies utilizing mevalonic acid-2-(14)C and D(2)O as precursors failed to provide evidence for an appreciable rate of endogenous biosynthesis of phytanic acid in a patient with Refsum's disease. Orally administered tracer doses of phytol-U-(14)C were well absorbed both by seven normal control subjects (61 to 94%) and by two patients with Refsum's disease (74 and 80%). The fraction of the absorbed dose converted to (14)CO(2) in 12 hours was 3.5 and 5.8% in Refsum's disease patients and averaged 20.9% in seven control subjects. Labeled phytanic acid was demonstrated in the plasma of both control subjects and patients given phytol-U-(14)C, establishing phytol in the diet as a potential precursor of phytanic acid. This labeled phytanic acid had disappeared almost completely from the plasma of the seven control subjects by 24 to 48 hours, whereas it persisted at high concentrations in the plasma of the two patients for many days. We conclude that the phytanic acid accumulating in Refsum's disease is primarily of exogenous origin and that patients with Refsum's disease have a relative block in the degradation of phytanic acid and possibly other similar branched-chain compounds. This may relate to a deficiency in mechanisms for release of phytanic acid from stored ester forms or, more probably, to reactions essential to oxidative degradation of the carbon skeleton.

A continuous-flow centrifuge was used to infuse sodium salts of oleic, linoleic, lauric, or palmitic acid into the pancreatic artery of anesthetized dogs. In these regional perfusion studies there was no increase in FFA levels in the general circulation. Elevation of pancreatic FFA levels produced an immediate increase in pancreatic venous immunoreactive insulin (IRI). After 10 min of FFA infusion. IRI levels declined somewhat from the initial peak response but soon rose again to high levels which were then sustained until the infusion was terminated. All four long-chain FFA tested produced a similar biphasic IRI response. Clearcut increases in IRI were associated with absolute FFA levels (measured in pancreaticoduodenal venous plasma) as low as 0.6-0.8 mueq/ml and with increments over basal levels of as little as 0.4-0.5 mueq/ml. At higher levels of FFA, absolute IRI levels in the pancreatic venous effluent exceeded 1,000 muU/ml in some experiments and 5- to 10-fold increases over basal values were observed. These studies indicate that long-chain FFA, in physiological concentrations, can markedly stimulate insulin secretion by a direct effect on the pancreas. The results lend support to the concept of insulin as a hormone that is importantly involved in regulating the metabolism of all three principal classes of metabolic substrates and whose release is in turn regulated by all of them. The relative importance and precise nature of its physiologic role in the regulation of lipolysis, lipid deposition, and ketone body formation remains to be established.

Refsum's disease (heredopathia atactica polyneuritiformis, HAP) is an inherited neurological disorder associated with storage of the branched-chain fatty acid, phytanic acid (3,7,11,15-tetramethylhexadecanoic acid). Cultured fibroblasts derived from skin biopsies of HAP patients did not contain elevated levels of phytanate, yet showed rates of phytanate-C-(14)C oxidation less than 3% of those seen in cells from control subjects. Cells of control subjects converted phytanate to alpha-hydroxyphytanate, to pristanate (the [n-1] homologue of phytanate) and to 4,8,12-trimethyltridecanoate, compounds previously identified as intermediates on the major pathway for phytanate metabolism in animals, providing the first direct evidence that this same oxidative pathway is operative in human cells. None of these breakdown products could be found after incubation of phytanate with HAP cells. Labeled alpha-hydroxyphytanate and labeled pristanate were oxidized at normal rates by HAP cells. Oxidation of the latter proceeded at normal rates both when added to the medium at very low tracer levels and at levels 100 times greater. Phytanate was incorporated into and released from lipid esters at normal rates by HAP cells. Elevated levels of free phytanate in the medium were no more toxic to HAP cells than to control cells over the 48- to 72-hr exposures involved in these studies, as evidenced by morphologic criteria and by ability to oxidize labeled palmitate. These findings are consistent with the hypothesis that the cells from HAP patients are deficient in a single enzyme involved in the alpha-hydroxylation of phytanate, while the enzymes involved in later steps are present at normal or near-normal levels.