Role of the kidneys in the metabolism of plasma mevalonate. Studies in humans and in rhesus monkeys.

Studies were carried out in humans and in rhesus monkeys to determine the role of the kidneys in the metabolism of circulating mevalonic acid (MVA). Following intravenous infusion of [14C]MVA and [3H]cholesterol, there was a rapid appearance of [14C]squalene in the kidneys that exhibited a significantly longer half-life than plasma or hepatic squalene. In man and in rhesus monkeys there was a rapid equilibration between newly synthesized cholesterol from MVA and exogenously administered cholesterol in all tissues except the kidneys, where the specific activity ratio of newly synthesized to exogenous cholesterol was significantly higher. Estimates of the quantitative metabolism of intravenously infused radiolabeled MVA in the monkey demonstrated that 23% was excreted in the urine, 67% metabolized to cholesterol (58% in nonrenal tissues and 9% in the kidneys), and 10% catabolized to CO2 and nonsteroid products. Measurements of MVA metabolism in anephric and uninephric patients demonstrate that, in the absence of renal uptake of MVA, exogenous and newly synthesized cholesterol achieve almost instantaneous equilibrium in the plasma; whereas in control subjects with normal renal function, this equilibration required at least 21 d for the two cholesterol decay curves to become parallel. These results suggest that the kidneys are solely responsible for the observed disequilibrium between newly synthesized and exogenous cholesterol; we suggest that this was due to the delayed release of newly synthesized cholesterol from the kidneys into the plasma compartment following intravenous infusion with radiolabeled MVA. The data document the importance of the kidneys in the metabolism of circulating MVA. However, calculation of the quantitative significance of this pathway in relation to whole body MVA metabolism indicates that renal metabolism of MVA accounts for approximately 0.1% of daily MVA turnover, and that alterations in this pathway due to any form of renal pathology would not result in significant changes in hepatic or whole body sterol synthesis rates. We urge caution in the use of radiolabeled MVA in long-term kinetic studies of sterol metabolism because our data show that the plasma compartment of MVA is not necessarily in isotopic equilibrium with tissue MVA.

Heterogeneity of cholesterol homeostasis in man. Response to changes in dietary fat quality and cholesterol quantity.

Studies were carried out to examine the effects of dietary fat and cholesterol on cholesterol homeostasis in man. 75 12-wk studies were carried out during intake of 35% of calories as either saturated or polyunsaturated fat, first low and then high in dietary cholesterol. Dietary fat and cholesterol intakes, plasma lipid and lipoprotein levels, cholesterol absorption and sterol synthesis in isolated blood mononuclear leukocytes were measured during each diet period. In 69% of the studies the subjects compensated for the increased cholesterol intake by decreasing cholesterol fractional absorption and/or endogenous cholesterol synthesis. When an increase in plasma cholesterol levels was observed there was a failure to suppress endogenous cholesterol synthesis. Plasma cholesterol levels were more sensitive to dietary fat quality than to cholesterol quantity. The results demonstrate that the responses to dietary cholesterol and fat are highly individualized and that most individuals have effective feedback control mechanisms.

Unexpected inhibition of cholesterol 7 alpha-hydroxylase by cholesterol in New Zealand white and Watanabe heritable hyperlipidemic rabbits.

We investigated the effect of cholesterol feeding on plasma cholesterol concentrations, hepatic activities and mRNA levels of HMG-CoA reductase and cholesterol 7 alpha-hydroxylase and hepatic LDL receptor function and mRNA levels in 23 New Zealand White (NZW) and 17 Watanabe heritable hyperlipidemic (WHHL) rabbits. Plasma cholesterol concentrations were 9.9 times greater in WHHL than NZW rabbits and rose significantly in both groups when cholesterol was fed. Baseline liver cholesterol levels were 50% higher but rose only 26% in WHHL as compared with 3.6-fold increase with the cholesterol diet in NZW rabbits. In both rabbit groups, hepatic total HMG-CoA reductase activity was similar and declined > 60% without changing enzyme mRNA levels after cholesterol was fed. In NZW rabbits, cholesterol feeding inhibited LDL receptor function but not mRNA levels. As expected, receptor-mediated LDL binding was reduced in WHHL rabbits. Hepatic cholesterol 7 alpha-hydroxylase activity and mRNA levels were 2.8 and 10.4 times greater in NZW than WHHL rabbits. Unexpectedly, cholesterol 7 alpha-hydroxylase activity was reduced 53% and mRNA levels were reduced 79% in NZW rabbits with 2% cholesterol feeding. These results demonstrate that WHHL as compared with NZW rabbits have markedly elevated plasma and higher liver cholesterol concentrations, less hepatic LDL receptor function, and very low hepatic cholesterol 7 alpha-hydroxylase activity and mRNA levels. Feeding cholesterol to NZW rabbits increased plasma and hepatic concentrations greatly, inhibited LDL receptor-mediated binding, and unexpectedly suppressed cholesterol 7 alpha-hydroxylase activity and mRNA to minimum levels similar to WHHL rabbits. Dietary cholesterol accumulates in the plasma of NZW rabbits, and WHHL rabbits are hypercholesterolemic because reduced LDL receptor function is combined with decreased catabolism of cholesterol to bile acids.

Plasma mevalonate as a measure of cholesterol synthesis in man.

Measurement of mevalonic acid (MVA) concentrations in plasma or 24-h urine samples is shown to be useful in studies of the regulation of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase and cholesterol synthesis. Plasma MVA concentrations, measured either at 7-9 a.m. after an overnight fast, or throughout the 24-h cycle, were compared with cholesterol synthesis rates that were measured by the sterol balance method: plasma MVA concentrations were directly related to the rate of whole body cholesterol synthesis (r = 0.972; p less than 0.001; n = 18) over a tenfold range of cholesterol synthesis rates. Moreover, hourly examination of MVA concentrations throughout the day demonstrated that interventions such as fasting or cholesterol feeding cause suppression of the postmidnight diurnal rise in plasma MVA concentrations, with little change in the base-line of the rhythm. Thus, the daily rise and fall of plasma MVA appears to reflect changes in tissues and organs, such as the liver and intestine, that are known to be most sensitive to regulation by fasting or by dietary cholesterol. The hypothesis that short-term regulation of HMG-CoA reductase in tissues is quickly reflected by corresponding variations in plasma MVA was tested by using a specific inhibitor of HMG-CoA reductase, mevinolin, to block MVA synthesis. Mevinolin caused a dose-dependent lowering of plasma MVA after a single dose; and in patients who received the drug twice a day for 4 wk, it decreased 24-h urinary MVA output. Significant lowering of plasma cholesterol was achieved through administration of mevinolin at doses that only moderately limit MVA production.

Rapid infusion of a phospholipid emulsion attenuates the effects of endotoxaemia in horses.

REASONS FOR PERFORMING STUDY: Endotoxaemia currently is associated with a poor prognosis in horses. The results of recent trials in other species indicate that phospholipid emulsions reduce the deleterious effects of endotoxin (LPS). However, in a previous study in horses, a 2 h infusion of emulsion caused an unacceptable degree of haemolysis. HYPOTHESIS: Rapid administration of a lower total dose of emulsion would reduce the effects of LPS and induce less haemolysis; the emulsion would reduce inflammatory effects of LPS in vitro. METHODS: Twelve healthy horses received an i.v. infusion either of saline or a phospholipid emulsion (100 mg/kg), followed immediately by E. coli 055:B5 LPS (30 ng/kg). Clinical parameters, haematological profiles, serum tumour necrosis factor (TNF) activity, serum lipid profiles, urine analyses and severity of haemolysis were monitored before and at selected times after LPS. Monocytes were also incubated in vitro with LPS in the presence or absence of emulsion, after which TNF and tissue factor activities were determined. RESULTS: Clinical signs of endotoxaemia were reduced in horses receiving the emulsion, including clinical score, heart rate, rectal temperature, serum TNF activity, and the characteristic leucopenic response to LPS, when compared to horses not receiving the emulsion. Three horses receiving the emulsion had none, 2 had mild and one had moderate haemolysis. There were no differences in urinalysis results and creatinine concentrations, either within the groups over time or between the groups. Serum concentrations of phosphatidylcholine, bile acids and triglycerides peaked immediately after the infusion; there were no significant changes in concentrations of nonesterified fatty acids or cholesterol. Incubation of equine monocytes with emulsion prevented LPS-induced TNF and tissue factor activities. CONCLUSIONS: Rapid administration of emulsion significantly reduced inflammatory effects of LPS in vivo and caused a clinically insignificant degree of haemolysis. The results of the in vitro studies indicate that emulsion prevents not only LPS-induced synthesis of cytokines, but also expression of membrane-associated mediators (i.e. tissue factor). POTENTIAL RELEVANCE: Rapid i.v. administration of emulsions containing phospholipids that bind endotoxin may provide a clinically useful method of treating endotoxaemia in horses.

Inhibition of liver prenyltransferase by alkyl phosphonates and phosphonophosphates.

n-Pentyl and n-decyl phosphonate and the corresponding phosphonophosphates were found to inhibit cholesterol synthesis from mevalonate in the 10000 X g supernatants of liver homogenates and the synthesis of farnesyl pyrophosphate from geranyl and isopentenyl pyrophosphate by purified liver prenyltransferase. Kinetic analysis of the inhibition of prenyltransferase showed that the phosphonates and the phosphonophosphates interacted with two forms, or two sites, of the enzyme. The order of increasing potency was C5-phosphonate less than C10-phosphonate less than C5-phosphonophosphate less than C10-phosphonophosphate. The phosphonophosphates were at least ten times stronger inhibitors than the phosphonates.

Removal of low-density lipoproteins in patients by extracorporeal immunoadsorption.

A new technique called LDL-pheresis was used in patients to lower low-density lipoprotein cholesterol levels. This procedure combines continuous extracorporeal plasma separation with immunoadsorption of low-density lipoprotein on columns containing monospecific antibody to human apolipoprotein B. Six patients underwent a total of 164 procedures without significant side effects or nonspecific protein depletion. Acutely, LDL-pheresis lowered plasma cholesterol levels by removing up to 82 percent of the circulating low-density lipoprotein. Weekly LDL-pheresis combined with a portacaval shunt in a patient with homozygous familial hypercholesterolemia resulted in normalization of plasma cholesterol levels and rapid regression of skin xanthomata. Three of four patients with atherosclerotic coronary artery disease have noted improvement in their angina. LDL-pheresis appears to be a promising new technique capable of safely and efficiently lowering plasma low-density lipoprotein cholesterol levels.

Extracorporeal LDL cholesterol removal: role of LDL-pheresis in combination with other hypolipidemic therapy to regress vascular disease.

The direct relationship between hypercholesterolemia and atherosclerosis has resulted in formal cholesterol-lowering recommendations for patients at increased risk. The incomplete response to therapy of some forms of hypercholesterolemia as well as not uncommon drug intolerance prompted the development of extracorporeal techniques to reduce serum cholesterol levels. Nonhuman primate data and an analysis of human cholesterol epidemiology and reduction trials were used to establish guidelines that would maximize the likelihood of stabilizing or regressing established coronary artery atherosclerosis. These goals are a total cholesterol (TC) level of less than or equal to 150 mg/dL (3.9 mmol/L) and a ratio of TC to high-density lipoprotein cholesterol (HDL) of less than 2.8. Selective, extracorporeal removal of LDL cholesterol (LDL-pheresis) was combined with diet and hypolipidemic drugs in a pilot study at The Rogosin Institute to achieve these lipid end-points. Technical aspects of LDL-pheresis, the background rationale for its use as part of a combined hypolipidemic therapy, the initial experience at The Rogosin Institute, and plans for future studies and applications are presented.

Interleukin-6 predicts hypoalbuminemia, hypocholesterolemia, and mortality in hemodialysis patients.

Low serum albumin and low serum cholesterol levels are among the most consistent predictors of mortality in patients with end-stage renal disease (ESRD) undergoing hemodialysis. Hypoalbuminemia is often interpreted as a marker of poor nutrition, but serum albumin and cholesterol levels can also be low as part of a cytokine-mediated acute-phase reaction to acute or chronic inflammation. Here we report the results from a 900-day prospective study designed to determine whether tumor necrosis factor-alfa (TNF-alpha) and interleukin-6 (IL-6) predict serum albumin and cholesterol levels and mortality in a group of 90 ambulatory, adult hemodialysis patients with no acute infection, hospitalization or surgery, and no known acquired immunodeficiency syndrome (AIDS), malignancy, or liver disease. Measurable levels of TNF-alpha and/or IL-6 were found in 89 of 90 patients. Significant relationships were found between TNF-alpha and IL-6 and the degree of hypoalbuminemia and dyslipoproteinemia. IL-6 was the strongest predictor of mortality in univariate and multivariate analysis, followed by age, albumin level, and body mass index (BMI). Although the cause of hypercytokinemia was not addressed in this study, the data support the view that hypoalbuminemia and hypocholesterolemia are negative acute-phase responses to inflammatory stimuli. These results suggest that efforts to identify the nature of the stimuli for cytokine production and to lower cytokine levels in hemodialysis patients might be effective in improving the survival of patients undergoing hemodialysis.

The correlation of cytokine levels with body weight after megestrol acetate treatment in geriatric patients.

BACKGROUND: Cachexia is associated with elevated levels of cytokines in cancer and human immunodeficiency virus patients. Studies in cancer and acquired immunodeficiency syndrome patients showed that treatment with megestrol acetate (MA) is associated with improvement in appetite and weight gain. Reduction in the levels of cytokines is associated with weight gain in laboratory animals with cancer. This study evaluates the correlation between changes in cytokine (or their receptor) levels and weight following MA treatment in geriatric weight-loss patients. METHODS: Veterans Administration Medical Center nursing home patients (N = 69) with a weight loss of > or =5% of usual body weight over the past 3 months or body weight 20% below their ideal body weight participated in a 12-week, randomized, double-blind, placebo-controlled trial, with an additional 13-week follow-up period. Patients were randomly assigned to receive a placebo or MA oral suspension of 800 mg/d for 12 weeks. Levels of the following cytokines (or their receptors) were measured at baseline and after 12 weeks of treatment: tumor necrosis factor soluble receptor (TNFR) subunits. TNFR-p55 and TNFR-p75: interleukin 6 (IL-6); and the soluble interleukin-2 receptor (sIL-2R). The subjects' weight and body composition were measured at the start of the study. Weight and mortality were followed up for another 13 weeks after discontinuing the MA study drug. RESULTS: Elevated levels of IL-6 in almost all geriatric cachexic patients, compared with normal volunteers (mean, <4.6 pg/ml). were noted at baseline. At 12 weeks after the study drug treatment, there was a decrease in cytokine levels (or their receptors) in the MA group (mean change in IL-6, 3.63+/-6.62 pg/ml; TNFR-p55, -0.06+/-0.11 ng/ml; TNFR-p75. -0.01+/-0.29 ng/ml; and sIL-2R, 0.08+/-0.07 ng/ml) and the placebo group (mean change in IL-6, -2.08+/-3.92 pg/ml; TNFR-p55, -0.02+/-0.08 ng/ml; TNFR-p75, -0.20+/-0.18 ng/ml; and sIL-2R, 0.02+/-0.03 ng/ml). Although the change in cytokine levels was not statistically significant between the two groups, significant negative correlation (p < .05) was found. For example, increased weight correlated with decreased sIL-2R levels (r = .36) and TNFR-p75 (r = -.31; fat-free mass (FFM) gain and reduction of sIL-2R (r = -.39), TNFR-p75 (r = -.30). There was a significant correlation between weight gain and reduction of TNFR-p75 (r = .54), TNFR-p55 (r- = .47), and sIL-2R (r = -.53); FFM gain and reduction of sIL-2R (r = -.59), TNFR-p75 (r = -.41), TNFR-p55 (r = -.42); and fat gain and reduction of TNFR-p75 (r = -.41) in the MA group (p < .05), but not in the placebo group. CONCLUSIONS: Although there was no significant change in cytokine levels between the two groups, the reduction in cytokine levels after MA treatment correlated with improvement in weight, fat mass, and FFM at 12 weeks.

Preoperative and intraoperative predictors of postoperative morbidity, poor graft function, and early rejection in 190 patients undergoing liver transplantation.

HYPOTHESIS: Preoperative and intraoperative variables predict in part adverse outcome after liver transplantation. DESIGN: Prospective, blinded, cohort study. SETTING: Tertiary care hospital. SUBJECTS: A total of 190 adult patients undergoing primary liver transplantation. MAIN OUTCOME MEASURE: Adverse outcome was prospectively defined as either in-hospital death or prolonged postoperative hospitalization (>14 days) associated with morbidity. Potential preoperative and intraoperative risk factors were collected. Associations were tested by univariate analysis followed by multivariate analysis in which preoperative factors were entered before intraoperative factors. RESULTS: Adverse outcome occurred in 44.7% of patients. Incidences of other complications were as follows: in-hospital mortality (8.4%), primary graft nonfunction (4.2%), poor early graft function (1.1%), and early rejection (31.2%). Univariate predictors of adverse outcome were United Network for Organ Sharing status (P =.003), Child-Turcotte-Pugh score (P =.02), POSSUM physiological score (P =.002), recipient age (P =.01), preoperative serum high-density lipoprotein cholesterol level (P =.03), preoperative serum creatinine level (P =.002), preoperative serum total IgG level (P =.004), duration in hospital preoperatively (P =.03), operative duration (P<.001), allogeneic erythrocyte transfusions (P<.001), total intraoperative fluids (P =.002), and use of inotropic agents (P =.01). In the final multivariate model, predictors of adverse outcome were United Network for Organ Sharing status (P =.03), recipient age (P =.002), and total intraoperative fluids (P =.04). Most patients who died or had a prolonged hospitalization exhibited dysfunction of more than 1 organ system, including pulmonary, renal, and infectious complications. CONCLUSIONS: Adverse outcome occurs frequently after liver transplantation, usually involves multiple organ systems, and is predicted in part by several preoperative and intraoperative factors.

Dextran-sulfate cellulose adsorption for lowering Lp[a].

Dextran sulfate selectively adsorbs lipoproteins that contain apolipoprotein-B100, including those associated with apolipoprotein-a, from human plasma with high affinity. Dextran sulfate immobilized on cellulose is incorporated into the Liposorber LA-15 system (Kaneka Corporation, Osaka, Japan). This system was evaluated in a nine-center controlled study of patients with familial hypercholesterolemia (FH) who had not responded adequately to diet and drug therapy. There were 54 patients with heterozygous FH: 45 randomized to treatment, 9 controls) and 10 patients with homozygous FH, all of whom received LDL apheresis. Removal of both LDL and Lp[a] was specific and highly efficient, > 90% of theory. Plasma LDL and Lp[a] concentrations were effectively lowered by repetitive LDL apheresis during the study and returned to baseline levels 3 to 4 weeks after the last apheresis treatment. The kinetics of the LDL rebound fit a simple one-pool kinetic model with a constant synthetic rate and constant fractional catabolic rate. The kinetics of the Lp[a] rebound are more complex and may require a model that includes a large extravascular Lp[a] pool in slow equilibrium with the plasma pool.

Quality of life and stimulation of weight gain after treatment with megestrol acetate: correlation between cytokine levels and nutritional status, appetite in geriatric patients with wasting syndrome.

BACKGROUND: The geriatric wasting syndrome (GWS) has been associated with proinflammatory cytokines, depression and progressive decline in quality of life (QOL). The objective of this study was to evaluate the correlation between the changes in cytokine levels and appetite, nutritional markers, and QOL in geriatric patients with GWS following a randomized clinical trial of megestrol acetate (MA) versus placebo. METHODS: This was a prospective, double-blind, placebo-controlled trial. We evaluated 69 predominantly male (3 females) nursing home residents with weight loss of > or =5% of their usual body weight over the past three months or body weight 20% below their ideal body weight. Patients were randomly assigned to receive either placebo or megestrol acetate (MA) oral suspension (O.S.) 800 mg/day for 12 weeks and were then followed for 13 weeks off treatment. Data on appetite, weight, nutritional status, QOL and cytokine levels were collected at baseline and week 12. The correlation between appetite, weight, nutritional status, sense of well being and cytokine level changes in response to MA treatment was examined at week 12. RESULTS: Appetite, sense of well being, and QOL assessed by an "enjoyment list" significantly improved in the MA arm. Rising prealbumin showed a negative correlation with decreasing IL-6 (r = -0.51), TNFR-p 55 (r = -0.49) and sIL-2R (r = -0.38). There was also an improvement in prealbumin and a decrease in IL-6 and TNFR-p55 in the MA-arm (p < 0.01). A correlation between a decrease in the IL-6 levels and improvement in depression (r = 0.50) was seen in the MA arm as well. Improvement in appetite positively correlated with increased enjoyment of life (r = -0.41), less depression (r = -0.34), improved sense of well being (r = 0.36), prealbumin gain (r = 0.30), and weight gain (r = 0.38) by 12 weeks. Also, improvement in appetite positively correlated with improvement in nutritional parameters such as prealbumin, albumin, fat free mass and weight in the MA arm. CONCLUSIONS: In a geriatric nursing home population with weight loss, reduction in cytokine levels after MA treatment correlates with improvement in appetite, prealbumin, albumin, and improvement in quality of life.

Immunochemical studies of extracellular glycoproteins (X-GPs) of goldfish brain.

Exoglycoproteins (X-GPs) are a family of soluble glycoproteins which are the most prominent constituent of the extracellular compartment of goldfish brain. On conventional two-dimensional polyacrylamide gels they typically display two primary molecular weight forms, averaging about 33 and 38 kDa, each appearing as a row of five to seven individual spots. When X-GP antibodies were applied by Western blotting, gels of goldfish brain extract prepared without a reducing agent showed, in addition to the primary molecular weight groups, at least one row of spots of slightly lower molecular weight and a major array of spots in the range of 45-60 kDa. The latter presumably represent dimers of the primary X-GP forms since they gave rise to the primary forms upon treatment with a reducing agent. However, on gradient gels prepared without detergents or reducing agents, X-GPs identified by immunostaining appeared only at 200 kDa and above, indicating that these proteins naturally occur in the form of large particles. Deglycosylation of the brain extract by N-glycosidase F reduced the molecular weight of each primary X-GP form by about 5 kDa, but did not abolish the microheterogeneity, which is at least partly due to minor differences in primary structure among the proteins in individual spots. Both rows of spots in the deglycosylated sample showed a coordinated shift toward the basic side of the gel, and a prominent new spot appeared on the basic end of the lower molecular weight group, which probably represents the fully deglycosylated form of the most abundant X-GP isoform.(ABSTRACT TRUNCATED AT 250 WORDS)

Plasmin catalyzes binding of lipoprotein (a) to immobilized fibrinogen and fibrin.

Lipoprotein (a) [Lp(a)] is a plasma component whose concentration is related to the development of atherosclerosis, although the underlying mechanisms are not known. Lp(a) contains a unique structure, apolipoprotein (a), that shares partial homology with plasminogen. We now report that plasmin catalyzes the binding of Lp(a) to both immobilized fibrinogen and fibrin in a manner analogous to our previously reported studies with plasminogen. Plasmin treatment of immobilized fibrinogen induces a 3.7-fold increase in Lp(a) binding. Low density lipoprotein, molecules similar to Lp(a) but lacking apolipoprotein (a), bind poorly to immobilized fibrinogen and binding is not increased by plasmin. Trypsin but not neutrophil elastase also increases the binding of Lp(a) to fibrinogen. Lp(a) also complexes to plasmin-fibrinogen digests, and binding increases in proportion to the time of plasmin-induced fibrinogen degradation. Lp(a) binding is lysine-binding site dependent as it is inhibited by epsilon-aminocaproic acid. Lp(a) inhibits the binding of plasminogen to plasmin-modified immobilized fibrinogen, indicating that both molecules compete for similar lysine-binding sites. These findings demonstrate an affinity between Lp(a) and protease-modified fibrinogen or fibrin and thereby provide a potential mechanism to explain the association between thrombosis, coronary atherosclerosis, and increased blood concentrations of Lp(a).

In vivo protection against endotoxin by plasma high density lipoprotein.

Overwhelming bacterial infection is accompanied by fever, hypotension, disseminated intravascular coagulation, and multiple organ failure leading to death in 30-80% of cases. These classical symptoms of septic shock are caused by potent cytokines that are produced in response to endotoxin released from Gram-negative bacteria. Treatments with antibodies and receptor antagonists to block endotoxin or cytokine mediators have given mixed results in clinical trials. High density lipoprotein (HDL) is a natural component of plasma that is known to neutralize endotoxin in vitro. We report here that raising the plasma HDL concentration protects mice against endotoxin in vivo. Transgenic mice with 2-fold-elevated plasma HDL levels had more endotoxin bound to HDL, lower plasma cytokine levels, and improved survival rates compared with low-HDL mice. Intravenous infusion of HDL also protected mice, but only when given as reconstituted HDL prepared from phospholipid and either HDL apoprotein or an 18-amino acid peptide synthesized to mimic the structure of apolipoprotein A-I of HDL. Intact plasma HDL was mildly toxic, and HDL apoprotein was ineffective. The effectiveness of the reconstituted peptide renders very unlikely any significant contribution to protection by trace proteins in apo-HDL. These data suggest a simple leaflet insertion model for binding and neutralization of lipopolysaccharide by phospholipid on the surface of HDL. Plasma HDL may normally act to protect against endotoxin; this protection may be augmented by administration of reconstituted HDL or reconstituted peptides.