Uptake of chemically modified low density lipoproteins in vivo is mediated by specific endothelial cells.

Acetoacetylated (AcAc) and acetylated (Ac) low density lipoproteins (LDL) are rapidly cleared from the plasma (t1/2 approximately equal to 1 min). Because macrophages, Kupffer cells, and to a lesser extent, endothelial cells metabolize these modified lipoproteins in vitro, it was of interest to determine whether endothelial cells or macrophages could be responsible for the in vivo uptake of these lipoproteins. As previously reported, the liver is the predominant site of the uptake of AcAc LDL; however, we have found that the spleen, bone marrow, adrenal, and ovary also participate in this rapid clearance. A histological examination of tissue sections, undertaken after the administration of AcAc LDL or Ac LDL (labeled with either 125I or a fluorescent probe) to rats, dogs, or guinea pigs, was used to identify the specific cells binding and internalizing these lipoproteins in vivo. With both techniques, the sinusoidal endothelial cells of the liver, spleen, bone marrow, and adrenal were labeled. Less labeling was noted in the ovarian endothelia. Uptake of AcAc LDL by endothelial cells of the liver, spleen, and bone marrow was confirmed by transmission electron microscopy. These data suggest uptake through coated pits. Uptake of AcAc LDL was not observed in the endothelia of arteries (including the coronaries and aorta), veins, or capillaries of the heart, testes, kidney, brain, adipose tissue, and duodenum. Kupffer cells accounted for a maximum of 14% of the 125I-labeled AcAc LDL taken up by the liver. Isolated sinusoidal endothelial cells from the rat liver displayed saturable, high affinity binding of AcAc LDL (Kd = 2.5 X 10(-9) M at 4 degrees C), and were shown to degrade AcAc LDL 10 times more effectively than aortic endothelial cells. These data indicate that specific sinusoidal endothelial cells, not the macrophages of the reticuloendothelial system, are primarily responsible for the removal of these modified lipoproteins from the circulation in vivo.

Expression of apolipoprotein E by cultured vascular smooth muscle cells is controlled by growth state.

Rat vascular smooth muscle cells (SMC) in culture synthesize and secrete a approximately 38,000-Mr protein doublet or triplet that, as previously described (Majack and Bornstein. 1984. J. Cell Biol. 99:1688-1695), rapidly and reversibly accumulates in the SMC culture medium upon addition of heparin. In the present study, we show that this approximately 38,000-Mr heparin-regulated protein is electrophoretically and immunologically identical to apolipoprotein E (apo-E), a major plasma apolipoprotein involved in cholesterol transport. In addition, we show that expression of apo-E by cultured SMC varies according to growth state: while proliferating SMC produced little apo-E and expressed low levels of apo-E mRNA, quiescent SMC produced significantly more apo-E (relative to other proteins) and expressed markedly increased levels of apo-E mRNA. Northern analysis of RNA extracted from aortic tissue revealed that fully differentiated, quiescent SMC contain significant quantities of apo-E mRNA. These data establish aortic SMC as a vascular source for apo-E and suggest new functional roles for this apolipoprotein, possibly unrelated to traditional concepts of lipid metabolism.

Isolation of a Golgi apparatus-rich fraction from rat liver. I. Method and morphology.

Golgi apparatus were released without fixatives from rat hepatocytes by gentle homogenization, concentrated by differential centrifugation, and purified by sucrose gradient centrifugation. Examination of sections of purified fractions by electron microscopy showed fields of morphologically intact units of Golgi apparatus consisting of stacks of parallel flattened cisternae, secretory vesicles, and small vesicular profiles. Negative staining of unfixed pellets revealed a complex network of anastomotic tubules continuous with platelike structures and secretory vesicles. These structures corresponded, respectively, to the small vesicular profiles and parallel flattened cisternae with attached secretory vesicles of sectioned material. Small fragments of granular endoplasmic reticulum were often closely associated with the peripheral tubules, suggesting sites of continuity in intact hepatocytes.

Apolipoprotein E: cholesterol transport protein with expanding role in cell biology.

Apolipoprotein E is a plasma protein that serves as a ligand for low density lipoprotein receptors and, through its interaction with these receptors, participates in the transport of cholesterol and other lipids among various cells of the body. A mutant form of apolipoprotein E that is defective in binding to low density lipoprotein receptors is associated with familial type III hyperlipoproteinemia, a genetic disorder characterized by elevated plasma cholesterol levels and accelerated coronary artery disease. Apolipoprotein E is synthesized in various organs, including liver, brain, spleen, and kidney, and is present in high concentrations in interstitial fluid, where it appears to participate in cholesterol redistribution from cells with excess cholesterol to those requiring cholesterol. Apolipo-protein E also appears to be involved in the repair response to tissue injury; for example, markedly increased amounts of apolipoprotein E are found at sites of peripheral nerve injury and regeneration. Other functions of apolipoprotein E, unrelated to lipid transport, are becoming known, including immunoregulation and modulation of cell growth and differentiation.

Lipoprotein uptake by neuronal growth cones in vitro.

Macrophages that rapidly enter injured peripheral nerve synthesize and secrete large quantities of apolipoprotein E. This protein may be involved in the redistribution of lipid, including cholesterol released during degeneration, to the regenerating axons. To test this postulate, apolipoprotein E-associated lipid particles released from segments of injured rat sciatic nerve and apolipoprotein E-containing lipoproteins from plasma were used to determine whether sprouting neurites, specifically their growth cones, possessed lipoprotein receptors. Pheochromocytoma (PC12) cells, which can be stimulated to produce neurites in vitro, were used as a model system. Apolipoprotein E-containing lipid particles and lipoproteins, which had been labeled with fluorescent dye, were internalized by the neurites and their growth cones; the unmetabolized dye appeared to be localized to the lysosomes. The rapid rate of accumulation in the growth cones precludes the possibility of orthograde transport of the fluorescent particles from the PC12 cell bodies. Thus, receptor-mediated lipoprotein uptake is performed by the apolipoprotein B,E(LDL) (low density lipoprotein) receptors, and in the regenerating peripheral nerve apolipoprotein E may deliver lipids to the neurites and their growth cones for membrane biosynthesis.

Three-dimensional structure of the LDL receptor-binding domain of human apolipoprotein E.

Human apolipoprotein E, a blood plasma protein, mediates the transport and uptake of cholesterol and lipid by way of its high affinity interaction with different cellular receptors, including the low-density lipoprotein (LDL) receptor. The three-dimensional structure of the LDL receptor-binding domain of apoE has been determined at 2.5 angstrom resolution by x-ray crystallography. The protein forms an unusually elongated (65 angstroms) four-helix bundle, with the helices apparently stabilized by a tightly packed hydrophobic core that includes leucine zipper-type interactions and by numerous salt bridges on the mostly charged surface. Basic amino acids important for LDL receptor binding are clustered into a surface patch on one long helix. This structure provides the basis for understanding the behavior of naturally occurring mutants that can lead to atherosclerosis.

Differential effects of apolipoproteins E3 and E4 on neuronal growth in vitro.

Apolipoprotein E4 (apoE4), one of the three common isoforms of apoE, has been implicated in Alzheimer's disease. The effects of apoE on neuronal growth were determined in cultures of dorsal root ganglion neurons. In the presence of beta-migrating very low density lipoproteins (beta-VLDL), apoE3 increased neurite outgrowth, whereas apoE4 decreased outgrowth. The effects of apoE3 or apoE4 in the presence of beta-VLDL were prevented by incubation with a monoclonal antibody to apoE or by reductive methylation of apoE, both of which block the ability of apoE to interact with lipoprotein receptors. The data suggest that receptor-mediated binding or internalization (or both) of apoE-enriched beta-VLDL leads to isoform-specific differences in interactions with cellular proteins that affect neurite outgrowth.

Identity of very low density lipoprotein apoproteins of plasma and liver Golgi apparatus.

In the rat, very low density lipoproteins isolated from hepatocyte Golgi apparatus, liver perfusates, and whole plasma appear identical in many respects. With specific immunochemical techniques and polyacrylamide-gel electrophoresis it can be demonstrated that the very low density lipoproteins from all three sources contain the same major lipoprotein apoproteins.

Altered metabolism (in vivo and in vitro) of plasma lipoproteins after selective chemical modification of lysine residues of the apoproteins.

Chemical modification of lysine residues by acetoacetylation of the apoproteins of iodinated canine and human low density lipoproteins (LDL) and canine high density lipoproteins (HDL) resulted in a marked acceleration in the rate of removal of these lipoproteins from the plasma after intravenous injection into dogs. Clearance of the lipoproteins from the plasma correlated with their rapid appearance in the liver. Acetoacetylated canine (125)I-LDL (30-60% of the lysine residues modified) were essentially completely removed from the plasma within an hour, and > 75% of the activity cleared within 5 min. Reversal of the acetoacetylation of the lysine residues of the LDL restored to these lipoproteins a rate of clearance essentially identical to that of control LDL. Identical results were obtained with modified human LDL injected into dogs. At 10 min, when congruent with 90% of the acetoacetylated human (125)I-LDL had been removed from the plasma, 90% of the total injected activity could be accounted for in the liver. Furthermore, it was possible to demonstrate an enhancement in uptake and degradation of acetoacetylated LDL by canine peritoneal macrophages in vitro. The mechanism(s) responsible for the enhanced removal of the LDL and HDL in vivo and in vitro remains to be determined. By contrast, however, acetoacetylation of canine (125)I-apoE HDL(c) did not accelerate their rate of removal from the plasma but, in fact, retarded their clearance. Control (native) apoE HDL(c) were removed from the plasma (64% within 20 min) and rapidly appeared in the liver (39% at 20 min). At the same time point, only 45% of the acetoacetylated apoE HDL(c) were cleared from the plasma and <10% appeared in the liver. Acetoacetylation of the apoE HDL(c) did not enhance their uptake or degradation by macrophages. The rapid clearance from the plasma of the native apoE HDL(c) in normal and hypercholesterolemic dogs suggests that the liver may be a normal site for the removal of the cholesteryl ester-rich apoE HDL(c). The retardation in removal after acetoacetylation of apoE HDL(c) indicates that the uptake process may be mediated by a lysine-dependent recognition system.

Identical structural and receptor binding defects in apolipoprotein E2 in hypo-, normo-, and hypercholesterolemic dysbetalipoproteinemia.

Apolipoprotein E (apoprotein E or apo-E) from type III hyperlipoproteinemic subjects with the E2/2 homozygous phenotype displays both structural and receptor binding heterogeneity. The apo-E from all subjects thus far studied, however, has been functionally defective, though to different degrees. Although nearly every type III hyperlipoproteinemic subject has the E2/2 phenotype, 95-99% of the people with this same phenotype do not display type III hyperlipoproteinemia, nor do they have elevated plasma cholesterol levels. Consequently, it became important to determine whether the apo-E2 from hypo- and normocholesterolemic individuals with the E2/2 phenotype is also functionally abnormal. To do this, apo-E2 was isolated from two hypo-, two normo- and two hypercholesterolemic homozygous E2/2 subjects. The apo-E2 was recombined with vesicles and tested for its ability to displace (125)I-low density lipoproteins (LDL) from apo-B,E (LDL) receptors on human fibroblasts. The apo-E2 from all six subjects was found to be severely defective in receptor binding (<2% of the binding activity of normal apo-E3). In all cases, the binding activity of the apo-E2 was increased 10- to 20-fold by treating the apoproteins with cysteamine, a reagent that converts cysteine residues to positively charged lysine analogues. The cysteine content of each apo-E was determined by monitoring the change in the isoelectric focusing position of the cysteamine-treated apo-E2. Using this method, it was found that the apo-E2 from each subject contained two cysteine residues per mole. A partial sequence analysis of the cysteine-containing regions of the apo-E from three of the six subjects indicated that the two cysteine residues were at residues 112 and 158 in the amino acid sequence. The cysteine at residue 158 has previously been implicated in the severe binding defect of the apo-E2 from a type III hyperlipoproteinemic subject. Since the apo-E2 of the hypo-, normo-, and hypercholesterolemic subjects in this study all displayed a severe functional abnormality, it is apparent that factors in addition to the defective receptor binding activity of the apo-E2 are necessary for the manifestation of type III hyperlipoproteinemia.

Regulation of hepatic lipoprotein receptors in the dog. Rapid regulation of apolipoprotein B,E receptors, but not of apolipoprotein E receptors, by intestinal lipoproteins and bile acids.

Two distinct lipoprotein receptors can be expressed in the dog liver. One is the apolipoprotein (apo-) B,E receptor. This receptor binds apo-B-containing low density lipoproteins (LDL), as well as apo-E-containing lipoproteins, such as the cholesterol-induced high density lipoproteins (HDL(c)). The second hepatic lipoprotein receptor is the apo-E receptor. It binds apo-E HDL(c) and chylomicron remnants, but not LDL. The present studies were undertaken to determine whether short-term (acute) regulation of the two receptors can occur in response to perturbations in hepatic cholesterol metabolism. The design used three groups of experimental animals: (a) immature dogs (with both hepatic apo-B,E and apo-E receptors expressed), (b) adult dogs (with predominantly the apo-E receptor expressed and little detectable apo-B,E receptor binding activity), and (c) dogs treated with the bile acid sequestrant cholestyramine or those that have undergone biliary diversion (with apo-E receptors and induced apo-B,E receptors). In the first series of experiments, changes in hepatic lipoprotein receptor expression were studied by delivering cholesterol to the liver via intestinal lymph lipoproteins. Dog lymph (5-11 mg of triglycerides/min per kg of body weight, 0.15-0.3 mg of cholesterol/min per kg) or saline were infused intravenously for 6-8 h into matched pairs of dogs. Serial liver biopsies were obtained at intervals of 1-2 h. A progressive loss of specific (calcium-dependent) binding of LDL was seen in hepatic membranes from both immature and cholestyramine-treated dogs. After 4-6 h of lymph infusion, almost no apo-B,E receptor binding could be detected. The decrease in binding of apo-E HDL(c) to the same membranes was much less pronounced, and could be explained by a loss of binding of HDL(c) to the apo-B,E receptor; there was little or no effect on apo-E receptor binding. In the second series of experiments, the effects of a diminished hepatic demand for cholesterol on lipoprotein receptor expression were studied by suppressing bile acid synthesis. The bile acid taurocholate (2-3 mumol/kg per min) was infused intravenously over a 6-h interval. This resulted in a progressive loss of LDL binding to liver membranes of immature or cholestyramine-treated dogs. The infusion of taurocholate for 6 h did not significantly alter the expression of the apo-E receptor binding activity, whereas apo-B,E receptor activity was rapidly down-regulated. Preparation of a bile fistula in adult dogs markedly induced the expression of the apo-B,E receptor. In this state, the binding activity of the apo-B,E receptor could be almost totally abolished by reinfusion of taurocholate for 6 h, without profoundly affecting apo-E receptor binding. Evidence from the analysis of plasma lipoprotein patterns and tissue culture reactivity suggested that changes in assayed hepatic lipoprotein receptor activity occurred in concert with changes in plasma lipoproteins.The results indicate that the two canine hepatic lipoprotein receptors differ in their metabolic regulation. The apo-B,E receptor responds rapidly to changes in hepatic requirements for cholesterol. The apo-E receptor appears to be more refractory to acute regulation. The rapidity of the changes in the activity of the apo-B,E receptor (within 2-4 h) suggests that the binding activity of this receptor may be regulated by factors independent of protein synthesis.

Cholesteryl ester accumulation in mouse peritoneal macrophages induced by beta-migrating very low density lipoproteins from patients with atypical dysbetalipoproteinemia.

The d < 1.006 lipoproteins of patients in a kindred with atypical dysbetalipoproteinemia induced marked cholesteryl ester accumulation in mouse peritoneal macrophages. The affected family members had severe hypercholesterolemia and hypertriglyceridemia, xanthomatosis, premature vascular disease, the apo-E3/3 phenotype, and a predominance of cholesterol-rich beta-very low density lipoproteins (beta-VLDL) in the d < 1.006 fraction. When incubated with mouse peritoneal macrophages, the d < 1.006 lipoproteins or beta-VLDL from the affected family members stimulated cholesteryl [(14)C]oleate synthesis 15- to 30-fold above that caused by normal, control d < 1.006 lipoproteins (VLDL). The ability of the beta-VLDL to stimulate macrophage cholesteryl ester accumulation was greatly reduced as a consequence of treatment with hypolipidemic agents, which specifically reduced the concentration of beta-VLDL. Two important differences were noted in a comparison of the beta-VLDL from these atypical dysbetalipoproteinemic subjects with that of classic E2/2 dysbetalipoproteinemics: (a) the beta-VLDL from the atypical subjects were severalfold more active in stimulating cholesteryl ester accumulation in macrophages, and (b) both the intestinal and hepatic beta-VLDL from the atypical subjects were active. The triglyceriderich, alpha(2)-migrating VLDL from the affected family members constituted <10% of the d < 1.006 fraction and were similar to normal VLDL in that they did not stimulate cholesteryl ester synthesis in the macrophages. Several lines of evidence indicate that the macrophage accumulation of cholesteryl esters was induced by a receptor-mediated uptake process and that the beta-VLDL were bound by a specific beta-VLDL receptor. First, the uptake and degradation of the lipoproteins and the induction of cholesteryl ester formation displayed qualities of high affinity, saturable kinetics. Second, the uptake and degradation process was inhibited when the lysyl residues of the beta-VLDL apoproteins were modified by reductive methylation. Third, the beta-VLDL from the affected subjects competed with diet-induced canine (125)I-beta-VLDL for the same cell surface receptors, but did not compete with chemically modified low density lipoproteins. Finally, the receptor-mediated uptake of these beta-VLDL resulted in lysosomal degradation of the lipoproteins, which could be prevented by incubating the cells with chloroquine. Normal, triglyceride-rich VLDL were also degraded when incubated with the macrophages, but they were not degraded by the same receptor-mediated process responsible for the degradation of the beta-VLDL of the patients. The degradation of the VLDL was not abolished by reductive methylation of the lipoproteins or by treatment of the cells with choloroquine. These studies demonstrate that the beta-VLDL from subjects with atypical dysbetalipoproteinemia are taken up by macrophages via the same receptor-mediated process responsible for the uptake of diet induced beta-VLDL. The accelerated vascular disease seen in these patients may be the result of high concentrations of beta-VLDL capable of binding to and delivering large quantities of cholesterol to macrophages and converting them into cells resembling the foam cells of atherosclerotic lesions.

A-IMilano apoprotein. Decreased high density lipoprotein cholesterol levels with significant lipoprotein modifications and without clinical atherosclerosis in an Italian family.

Significant hypertriglyceridemia with a very marked decrease of high density lipoproteins (HDL)-cholesterol levels (7-14 mg/dl) was detected in three members (father, son, and daughter) of an Italian family. The three affected individuals did not show any clinical signs of atherosclerosis, nor was the atherosclerotic disease significantly present in the family. Lipoprotein lipase and lecithin:cholesterol acyltransferase activites were normal or slightly reduced. Morphological and compositional studies of HDL in the subjects showed a significant enlargement of the lipoprotein particles (approximately 120 vs. approximately 94 A for control HDL) and a concomitant increase in the triglyceride content. Analytical isoelectric focusing of HDL apoproteins provided evidence for multiple isoproteins in the apoprotein(apo)-A-I range, with nine different bands being detected instead of the usual four bands observed in normal subjects. Two-dimensional immunoelectrophoresis against apo-A antiserum indicated a clear reduction of apo-A in the alpha electrophoretic region, with splitting of the protein "peak." The observation in otherwise clinically healthy subjects of hypertriglyceridemia, reduced HDL-cholesterol, and marked apoprotein abnormalities, without a significant incidence of atherosclerotic disease in the family suggests this is a new disease entity in the field of lipoprotein pathology, very probably related to an altered amino acid composition of the apo-A-I protein (see Weisgraber et al. 1980. J. Clin. Invest. 66: 901-907).

Two independent lipoprotein receptors on hepatic membranes of dog, swine, and man. Apo-B,E and apo-E receptors.

We have reported previously that canine livers possess two distinct lipoprotein receptors, an apoprotein (apo)-B,E receptor capable of binding the apo-B-containing low density lipoproteins (LDL) and the apo-E-containing cholesterol-induced high density lipoproteins (HDLc), and an apo-E receptor capable of binding apo-E HDLc but not LDL. Both the apo-B,E and apo-E receptors were found on the liver membranes obtained from immature growing dogs, but only the apo-E receptors were detected on th hepatic membranes of adult dogs. In this study, the expression of the apo-B,E receptors, as determined by canine LDL binding to the hepatic membranes, was found to be highly dependent on the age of the dog and decreased linearly with increasing age. Approximately 30 ng of LDL protein per milligram of membrane protein were bound via the apo-B,E receptors to the hepatic membranes of 7- to 8-wk-old immature dogs as compared with no detectable LDL binding in the hepatic membranes of adult dogs (greater than 1--1.5 yr of age). Results obtained by in vivo turnover studies of canine 125I-LDL correlated with the in vitro findings. In addition to a decrease in the expression of the hepatic apo-B,E receptors with age, these receptors were regulated, i.e., cholesterol feeding suppressed these receptors in immature dogs and prolonged fasting induced their expression in adult dogs. Previously, it was shown that the apo-B,E receptors were induced in adult livers following treatment with the hypocholesterolemic drug cholestyramine. In striking contrast, the apo-E receptors, as determined by apo-E HDLc binding, remained relatively constant for all ages of dogs studied (10--12 ng/mg). Moreover, the expression of the apo-E receptors was not strictly regulated by the metabolic perturbations that regulated the apo-B,E receptors. Similar results concerning the presence of apo-B,E and apo-E receptors were obtained in swine and in man. The hepatic membranes of adult swine bound only apo-E HDLc (apo-E receptors), whereas the membranes from fetal swine livers bound both LDL and apo-E HDLc (apo B,E and apo-E receptors). Furthermore, the membranes from adult human liver revealed the presence of the apo-E receptors as evidenced by the binding of 12--14 ng of HDLc protein per milligram of membrane protein and less than 1 ng of LDL protein per milligram. The membranes from the human liver also bound human chylomicron remnants and a subfraction of human HDL containing apo-E. These data suggest the importance of the E apoprotein and the apo-E receptors in mediating lipoprotein clearance, including chylomicron remnants, by the liver of adult dogs, swine, and man.

Fat feeding in humans induces lipoproteins of density less than 1.006 that are enriched in apolipoprotein [a] and that cause lipid accumulation in macrophages.

Formula diets containing lard or lard and egg yolks were fed to six normolipidemic volunteers to investigate subsequent changes in the composition of lipoproteins of d less than 1.006 g/ml and in their ability to bind and be taken up by receptors on mouse macrophages. Both formulas induced the formation of d less than 1.006 lipoproteins that were approximately 3.5-fold more active than fasting very low density lipoproteins (VLDL) in binding to the receptor for beta-VLDL on macrophages. Subfractionation of postprandial d less than 1.006 lipoproteins by agarose chromatography yielded two subfractions, fraction I (chylomicron remnants) and fraction II (hepatic VLDL remnants), which bound to receptors on macrophages. However, fraction I lipoproteins induced a 4.6-fold greater increase in macrophage triglyceride content than fraction II lipoproteins or fasting VLDL. Fraction I lipoproteins were enriched in apolipoproteins (apo) B48, E, and [a]. Fraction II lipoproteins lacked apo[a] but possessed apo B100 and apo E. The apo[a] was absent in normal fasting VLDL, but was present in the d less than 1.006 lipoproteins (beta-VLDL) of fasting individuals with type III hyperlipoproteinemia. The apo[a] from postprandial d less than 1.006 lipoproteins was larger than either of two apo[a] subspecies obtained from lipoprotein (a) [Lp(a)] isolated at d = 1.05-1.09. However, all three apo[a] subspecies were immunochemically identical and had similar amino acid compositions: all were enriched in proline and contained relatively little lysine, phenylalanine, isoleucine, or leucine. The association of apo[a] with dietary fat-induced fraction I lipoproteins suggests that the previously observed correlation between plasma Lp(a) concentrations and premature atherosclerosis may be mediated, in part, by the effect of apo[a] on chylomicron remnant metabolism.

Identification of a new structural variant of human apolipoprotein E, E2(Lys146 leads to Gln), in a type III hyperlipoproteinemic subject with the E3/2 phenotype.

A type III hyperlipoproteinemic subject having the apolipoprotein E (apo E) phenotype E3/2 was identified. From isoelectric focusing experiments in conjunction with cysteamine treatment (a method that measures cysteine content in apo E), the E2 isoform of this subject was determined to have only one cysteine residue, in contrast to all previously studied E2 apoproteins, which had two cysteines. This single cysteine was shown to be at residue 112, the same site at which it occurs in apo E3. From amino acid and sequence analyses, it was determined that this apo E2 differed from apo E3 by the occurrence of glutamine rather than lysine at residue 146. When phospholipid X protein recombinants of the subject's isolated E3 and E2 isoforms were tested for their ability to bind to the human fibroblast apo-B,E receptor, it was found that the E3 bound normally (compared with an apo E3 control) but that the E2 had defective binding (approximately 40% of normal). Although they contained E3 as well as E2, the beta-very low density lipoproteins (beta-VLDL) from this subject were very similar in character to the beta-VLDL from an E2/2 type III hyperlipoproteinemic subject; similar subfractions could be obtained from each subject and were shown to have a similar ability to stimulate cholesteryl ester accumulation in mouse peritoneal macrophages. The new apo E2 variant has also been detected in a second type III hyperlipoproteinemic subject.

A novel electrophoretic variant of human apolipoprotein E. Identification and characterization of apolipoprotein E1.

A new apolipoprotein E (apo E) phenotype has been demonstrated in a Finnish hypertriglyceridemic subject (R.M.). At the time of this study, R.M.'s plasma triglyceride and cholesterol levels were 1,021 and 230 mg/dl, respectively. The subject's apo E isoelectric focusing pattern was characterized by two major bands, one in the E3 position and the other in the E1 position. Normally the E1 position is occupied by sialylated derivatives of apo E4, E3, or E2. The E1 band of subject R.M. is not a sialylated form, however, because it was not affected by neuraminidase digestion. The identity of the E1 variant as a genetically determined structure was established by amino acid and partial sequence analyses, confirming that the variant is an example of a previously uncharacterized apo E phenotype, E3/1. Both cysteamine modification and amino acid analysis demonstrated that this variant contains two cysteine residues per mole. Sequence analysis of two cyanogen bromide fragments and one tryptic fragment of the apo E3/1 showed that it differs from E2(Arg158----Cys) at residue 127, where an aspartic acid residue is substituted for glycine. This single amino acid interchange is sufficient to account for the one-charge difference observed on isoelectric focusing gels between E2(Arg158----Cys) and the E1 variant. The variant has been designated E1 (Gly127----Asp, Arg158----Cys). When compared with apo E3, the E1 variant demonstrated reduced ability to compete with 125I-LDL for binding to LDL (apo B,E) receptors on cultured fibroblasts (approximately 4% of the amount of binding of apo E3). This defective binding is similar to that of E2-(Arg158----Cys). Therefore, the binding defect of the variant is probably due to the presence of cysteine at residue 158, rather than aspartic acid at residue 127. In contrast, the apo E3 isoform from this subject demonstrated normal binding activity, indicating that it has a normal structure. In family studies, the vertical transmission of the apo E1 variant has been established. It is not yet clear, however, if the hypertriglyceridemia observed in the proband is associated with the presence of the E1(Gly127----Asp, Arg158----Cys) variant.

Apolipoprotein E associated with astrocytic glia of the central nervous system and with nonmyelinating glia of the peripheral nervous system.

The plasma protein apolipoprotein (apo) E is an important determinant of lipid transport and metabolism in mammals. In the present study, immunocytochemistry has been used to identify apo E in specific cells of the central and peripheral nervous systems of the rat. Light microscopic examination revealed that all astrocytes, including specialized astrocytic cells (Bergmann glia of the cerebellum, tanycytes of the third ventricle, pituicytes of the neurohypophysis, and Müller cells of the retina), possessed significant concentrations of apo E. In all of the major subdivisions of the central nervous system, the perinuclear region of astrocytic cells, as well as their cell processes that end on basement membranes at either the pial surface or along blood vessels, were found to be rich in apo E. Extracellular apo E was present along many of these same surfaces. The impression that apo E is secreted by astrocytic cells was confirmed by electron microscopic immunocytochemical studies, which demonstrated the presence of apo E in the Golgi apparatus. Apo E was not present in neurons, oligodendroglia, microglia, ependymal cells, and choroidal cells. In the peripheral nervous system, apo E was present within the glia surrounding sensory and motor neurons; satellite cells of the dorsal root ganglia and superior cervical sympathetic ganglion as well as the enteric glia of the intestinal ganglia were reactive. Apo E was also present within the non-myelinating Schwann cells but not within the myelinating Schwann cells of peripheral nerves. These results suggest that apo E has an important, previously unsuspected role in the physiology of nervous tissue.

Macrophage-specific expression of human apolipoprotein E reduces atherosclerosis in hypercholesterolemic apolipoprotein E-null mice.

apoE deficiency causes hyperlipidemia and premature atherosclerosis. To determine if macrophage-specific expression of apoE would decrease the extent of atherosclerosis, we expressed human apoE in macrophages of apoE-null mice (apoE-/-) and assessed the effect on lipid accumulation in cells of the arterial wall. Macrophage-specific expression of human apoE in normal mice was obtained by use of the visna virus LTR. These animals were bred with apoE-/- mice to produce animals hemizygous for expression of human apoE in macrophages in the absence of murine apoE (apoE-/-,hTgE+/0). Low levels of human apoE mRNA were present in liver and spleen and high levels in lung and peritoneal macrophages. Human apoE was secreted by peritoneal macrophages and was detected in Kupffer cells of the liver. Human apoE in the plasma of apoE-/-,hTgE+/0 mice (n = 30) was inversely correlated (P < 0.005) with the plasma cholesterol concentration. After 15 wk on a normal chow diet, atherosclerosis was assessed in apoE-/-,hTgE+/0 animals and in apoE-/-,hTgE0/0 littermates matched for plasma cholesterol level (approximately 450 mg/dl) and lipoprotein profile. There was significantly less atherosclerosis in both the aortic sinus and in the proximal aorta (P < 0.0001) in the animals expressing the human apoE transgene. In apo-E-/-,hTgE+/0 animals, which had detectable atherosclerotic lesions, human apoE was detected in the secretory apparatus of macrophage-derived foam cells in the arterial wall. The data demonstrate that expression of apoE by macrophages is antiatherogenic even in the presence of high levels of atherogenic lipoproteins. The data suggest that apoE prevents atherosclerosis by promoting cholesterol efflux from cells of the arterial wall.

Intravenous infusion of apolipoprotein E accelerates clearance of plasma lipoproteins in rabbits.

Plasma cholesterol levels in cholesterol-fed rabbits were markedly reduced by the intravenous infusion or bolus injection of recombinant human apo E or rabbit plasma apo E. Administration of 6-70 mg of apo E resulted in an approximately 20-40% acute reduction in plasma cholesterol levels within 2-3 h. Plasma cholesterol levels remained reduced for 4-8 h after the administration of apo E. Furthermore, the intravenous injection of apo E reduced the plasma cholesterol levels in Watanabe heritable hyperlipidemic rabbits. The addition of apo E to [14C]cholesterol-labeled canine thoracic duct lymph or [14C]cholesterol-labeled chylomicrons resulted in accelerated plasma clearance of these diet-induced lipoproteins in normal rabbits, with the uptake occurring primarily in the liver. This study suggests that the amount or availability of apo E in the plasma of cholesterol-fed rabbits may be rate limiting for the normal clearance of diet-induced remnant lipoproteins.