Using genetically modified mice to study apolipoprotein B

The B apolipoproteins, apo-B48 and apo-B100, are key proteins in mammalian lipoprotein metabolism and are components of all classes of lipoproteins considered to be atherogenic. Our laboratory has generated an array of genetically modified mice for studying apo-B biology. Using gene targeting in mouse embryonic stem cells, we have generated apo-B-deficient mice. Heterozygotes had low plasma levels of apo-B and cholesterol; homozygotes died early in embryonic development, most likely because the absence of lipoprotein secretion by the yolk sac interfered with the delivery of lipid nutrients to the developing embryo. We have also generated human apo-B transgenic mice with an 80-kb genomic DNA fragment spanning the entire human apo-B gene; those mice had markedly increased plasma levels of low density lipoprotein cholesterol and exhibited increased susceptibility to atherosclerosis. The human apo-B transgenic mice have also yielded insights regarding the regulation of apo-B expression in different tissues. Although the 80-kb transgene contained nearly 20 kb of 5' and 3' flanking sequences and was expressed at high levels in the liver, no transgene expression was detectable in the intestine. Subsequent transgenic mouse studies have demonstrated that the expression of the apo-B gene in the intestine is controlled by DNA sequences that are very distant from the structural gene. Transgenic mice have also proved useful for studying apo-B structure/function relationships. By expressing mutant forms of human apo B in transgenic mice, we have examined the structural features of the apo-B molecule that are required for lipoprotein (a) formation. We have demonstrated that the carboxyl terminal cystine residue of apo-B100, cysteine-4326, is required for apo-B100's disulfide linkage with apo(a) to form lipoprotein (a). Finally, we have used gene targeting techniques to generate mice that synthesize exclusively apo-B48 (apo B48-only mice) and mice that synthesize exclusively apo-B100 (apo-B100 only mice): These mice have helped to clarify the unique metabolic roles of the two apo-B proteins.     

Transgenic mice expressing high plasma concentrations of human apolipoprotein B100 and lipoprotein(a)

The B apolipoproteins, apo-B48 and apo-B100, are key structural proteins in those classes of lipoproteins considered to be atherogenic [e.g., chylomicron remnants, beta-VLDL, LDL, oxidized LDL, and Lp(a)]. Here we describe the development of transgenic mice expressing high levels of human apo-B48 and apo-B100. A 79.5-kb human genomic DNA fragment containing the entire human apo-B gene was isolated from a P1 bacteriophage library and microinjected into fertilized mouse eggs. 16 transgenic founders expressing human apo-B were generated, and the animals with the highest expression had plasma apo-B100 levels nearly as high as those of normolipidemic humans (approximately 50 mg/dl). The human apo-B100 in transgenic mouse plasma was present largely in lipoproteins of the LDL class as shown by agarose gel electrophoresis, chromatography on a Superose 6 column, and density gradient ultracentrifugation. When the human apo-B transgenic founders were crossed with transgenic mice expressing human apo(a), the offspring that expressed both transgenes had high plasma levels of human Lp(a). Both the human apo-B and Lp(a) transgenic mice will be valuable resources for studying apo-B metabolism and the role of apo-B and Lp(a) in atherosclerosis.     

Molecular analysis of the aspartylglucosaminidase gene in Japanese patients with aspartylglucosaminuria

Molecular analysis of the aspartylglucosaminidase (AGA) gene was performed on two Japanese siblings with aspartylglucosaminuria (AGU). They were the first Japanese patients diagnosed as AGU and up to now, no other patients with AGU have been discovered in Japan. AGA cDNA from one patient contained 7-nt (TCTCCAG) insertion between exons 3 and 4, which was identical with the 3'-terminal sequence of intron 3. Sequencing of amplified genomic DNA from both patients showed the same single base transition (A-->G) at the 5'-side adjacent to the insertion sequence. This mutation newly created a consensus AG dinucleotide at the 3'-splice site and caused alternative splicing of AGA pre-mRNA by leaving the 3'-terminal 7-nt of intron 3. BsmAI restriction site analysis of amplified genomic DNA indicated that they were homozygotes of this mutation. We conclude that splicing defect of intron 3 caused AGA deficiency in them.     

XbaI polymorphism of the apolipoprotein B gene and plasma lipid and lipoprotein response to dietary fat and cholesterol: a clinical trial

A dietary trial was carried out on a group of offspring whose parents were hospitalized for an acute myocardial infarction. The XbaI Restriction Fragment Length Polymorphism (RFLP) was used to examine the genetic contribution of variation at this apo B locus to the response of lipids and lipoproteins to dietary manipulations. Twenty participants were homozygotes for the 8.0 kb fragment (X1X1), two were homozygotes for the 5.0 kb fragment (X2X2), and 15 were heterozygotes (X1X2). Subjects were randomized to a 5-week crossover study. Half began on a low SFA--cholesterol (LSC) diet for 5 weeks and, after a washout period of 4 weeks, they were placed on a high SFA--cholesterol (HSC) diet for a second period of 5 weeks. This order was reversed in the second group of participants. Significant changes in total cholesterol, LDL-C, and apo B were observed when subjects were moved from the LSC to the HSF diet. The corresponding average change induced by the dietary manipulations in X1X1 subjects compared with subjects with X2 allele were: 18.1 +/- 17.6 mg/dl and 9.5 +/- 19.6 mg/dl for total cholesterol and 15.8 +/- 15.3 mg/dl and 4.8 +/- 20.9 mg/dl for LDL-C, respectively. Our observation indicated that variation at the apo B XbaI locus may interact with baseline levels to determine individual dietary response in LDL-C level. However, the differences between the genotypic classes were not statistically significant, suggesting that the apo B XbaI locus is not a major determinant of interindividual differences in lipid and lipoprotein response to diet in this population.     

Using genetically engineered mice to understand apolipoprotein-B deficiency syndromes in humans

Several human diseases are characterized by defects in the synthesis and secretion of the apolipoprotein (apo) B-containing lipoproteins. Familial hypobetalipoproteinemia is caused by mutations in the apo-B gene and is characterized by abnormally low plasma concentrations of apo-B and low-density lipoprotein (LDL) cholesterol. Another apo-B deficiency syndrome, abetalipoproteinemia, is caused by mutations in the gene for microsomal triglyceride transfer protein (MTP). MTP is a microsomal protein that is thought to transfer lipids to the apo-B protein as it is translated, allowing it to attain the proper conformation for lipoprotein assembly. A third apo-B deficiency syndrome, Anderson's disease (or chylomicron retention disease), is characterized by the inability to secrete apo-B-containing chylomicrons from the intestine but an apparently normal capacity to secrete lipoproteins from the liver. To more fully understand these human apo-B deficiency syndromes, our laboratory has generated and characterized gene-targeted mouse models. This review summarizes what has been learned from these animal models.     

Identification of the low density lipoprotein receptor-binding site in apolipoprotein B100 and the modulation of its binding activity by the carboxyl terminus in familial defective apo-B100

Familial defective apolipoprotein B100 (FDB) is caused by a mutation of apo-B100 (R3500Q) that disrupts the receptor binding of low density lipoproteins (LDL), which leads to hypercholesterolemia and premature atherosclerosis. In this study, mutant forms of human apo-B were expressed in transgenic mice, and the resulting human recombinant LDL were purified and tested for their receptor-binding activity. Site-directed mutagenesis and other evidence indicated that Site B (amino acids 3,359-3,369) binds to the LDL receptor and that arginine-3,500 is not directly involved in receptor binding. The carboxyl-terminal 20% of apo-B100 is necessary for the R3500Q mutation to disrupt receptor binding, since removal of the carboxyl terminus in FDB LDL results in normal receptor-binding activity. Similarly, removal of the carboxyl terminus of apo-B100 on receptor-inactive VLDL dramatically increases apo-B-mediated receptor-binding activity. We propose that the carboxyl terminus normally functions to inhibit the interaction of apo-B100 VLDL with the LDL receptor, but after the conversion of triglyceride-rich VLDL to smaller cholesterol-rich LDL, arginine-3,500 interacts with the carboxyl terminus, permitting normal interaction between LDL and its receptor. Moreover, the loss of arginine at this site destabilizes this interaction, resulting in receptor-binding defective LDL.     

Construction of a DBA/2.Fv-2r congenic strain. Apparent lethality of the homozygous Fv-2r genotype: brief communication

From the (C57BL/6 X DBA/2)F1 cross, 13 serial backcrosses to the DBA/2 parental mouse strain were bred with selection by progeny testing in each generation for the Fv-2s/Fv-2r heterozygous genotype. Intercrossing heterozygotes of the 13th backcross generation produced no Fv-2r/Fv-2r homozygotes. Homozygosity for the Fv-2r allele thus appeared to be lethal on a DBA/2 background and in the absence of protector gene(s) of the C57BL strain.     

A sequence-specific RNA-binding protein complements apobec-1 To edit apolipoprotein B mRNA

The editing of apolipoprotein B (apo-B) mRNA involves the site-specific deamination of cytidine to uracil. The specificity of editing is conferred by an 11-nucleotide mooring sequence located downstream from the editing site. Apobec-1, the catalytic subunit of the editing enzyme, requires additional proteins to edit apo-B mRNA in vitro, but the function of these additional factors, known as complementing activity, is not known. Using RNA affinity chromatography, we show that the complementing activity binds to a 280-nucleotide apo-B RNA in the absence of apobec-1. The activity did not bind to the antisense strand or to an RNA with three mutations in the mooring sequence. The eluate from the wild-type RNA column contained a 65-kDa protein that UV cross-linked to apo-B mRNA but not to the triple-mutant RNA. This protein was not detected in the eluates from the mutant or the antisense RNA columns. Introduction of the mooring sequence into luciferase RNA induced cross-linking of the 65-kDa protein. A 65-kDa protein that interacted with apobec-1 was also detected by far-Western analysis in the eluate from the wild-type RNA column but not from the mutant RNA column. For purification, proteins were precleared on the mutant RNA column prior to chromatography on the wild-type RNA column. Silver staining of the affinity-purified fraction detected a single prominent protein of 65 kDa. Our results suggest that the complementing activity may function as the RNA-binding subunit of the holoenzyme.     

DNA polymorphisms in the ACE gene, serum ACE activity and the risk of nephropathy in insulin-dependent diabetes mellitus

BACKGROUND: To determine the relationship between DNA polymorphisms in the angiotensin I converting enzyme (ACE) gene, serum ACE activity and the risk of diabetic nephropathy. METHODS: A case-control study was carried out in a population of Jewish insulin-dependent diabetes mellitus (IDDM) patients. Cases (77 IDDM patients with diabetic nephropathy) and controls (89 IDDM patients with normoalbuminuria) were genotyped with PCR protocols for detecting two DNA polymorphisms in the ACE gene: one in intron 7 detected with the restriction enzyme PstI and the other in intron 16 identified as an insertion/deletion (I/D). RESULTS: The risk of nephropathy was increased only in patients homozygous for the allele with the PstI site. These homozygotes had a nephropathy risk that was 2.3 times (95% C.I.: 1.2-4.5) that of the other genotypes. Furthermore, these individuals did not have elevated serum ACE activity. CONCLUSIONS: The results of this study are evidence that the risk of diabetic nephropathy in IDDM is influenced by genetic variability at the ACE locus, but the responsible variant is not the I/D polymorphism in intron 16. Our findings require further studies in other populations.     

Complete thyroxine-binding globulin (TBG) deficiency produced by a mutation in acceptor splice site causing frameshift and early termination of translation (TBG-Kankakee)

Fourteen T4-binding globulin (TBG) variants have been identified at the gene level. They are all located in the coding region of the gene and 6 produce complete deficiency of TBG (TBG-CD). We now describe the first mutation in a noncoding region producing TBG-CD. The proband was treated for over 20 yr with L-T4 because of fatigue associated with a low concentration of serum total T4. Fifteen family members were studied showing low total T4 inherited as an X chromosome-linked trait, and affected males had undetectable TBG in serum. Sequencing of the entire coding region and promoter of the TBG gene revealed no abnormality. However, an A to G transition was found in the acceptor splice junction of intron II that produced a new HaeIII restriction site cosegregating with the TBG-CD phenotype. Sequencing exon 1 to exon 3 of TBG complementary DNA reverse transcribed from messenger RNA of skin fibroblasts from an affected male, confirmed a shift in the ag acceptor splice site. This results in the insertion of a G in exon 2 and causes a frameshift and a premature stop at codon 195. This early termination of translation predicts a truncated TBG lacking 201 amino acids.     

Regulation of the assembly and secretion of very low density lipoproteins by the liver

Recent studies have suggested that there are three sites at which VLDL secretion by the liver may be controlled: (i) Newly synthesised apo-B either remains associated with the RER membrane and is degraded by the ubiquitin/proteasome system, or is translocated into the lumen and incorporated into lipid poor VLDL precursors; (ii) the lumenal apo-B is either degraded or moves on, and (iii) acquires the remaining VLDL lipids in the SER/cis-Golgi. Newly synthesised apo-B, at the cytosolic side of the RER, is stabilised and protected from degradation by the chaperone protein, hsp-70. Triacylglycerol, cholesterol ester and phospholipids have all been implicated in the translocation of apo-B and microsomal triglyceride protein plays a major role. If translocation does not occur then the apo-B is degraded. Dietary fish-oils, but not sunflower oil, inhibit movement of apo-B containing precursors from the RER and their assembly with lipids and target lumenal apo-B to degradation. This effect is reversed by inhibition of lumenal proteolysis, but not by inhibition of cytosolic proteolysis. Therefore lumenal degradation of apo-B and secretion appear to be in balance, so that if assembly of VLDL precursors is slowed, then degradation becomes predominant. If however, degradation is inhibited then VLDL assembly can proceed. These observations suggest that movement of VLDL precursors from the RER lumen to the second stage of assembly may be a further regulated step.     

Video-assisted thoracoscopy in single-stage resection of a para-aortic posterior mediastinal dumbbell tumor

The incidence of cystic fibrosis (CF) has previously been calculated from epidemiological surveys and from neonatal screening. With the cloning of the CF gene it has become possible to derive incidence figures from heterozygote frequencies, provided that the distribution of mutant alleles among healthy carriers is the same as among affected people. We have estimated the allele frequencies for four CF mutations, AF508, G551D, G542X and R117H, in 14360 unselected women undergoing antenatal heterozygote screening. The proportion of R117H, an allele of known mild effect, was much greater for heterozygotes than for homozygotes. The incidence of CF was therefore calculated from the heterozygote frequencies of AF508, G551D and G542X in a larger cohort of 27 161 successively screened women. The point estimate for the incidence of CF in the Scottish population was 1 in 1984, with 95% confidence intervals of 1 in 1692 to 1 in 2336.     

Hereditary defect in biosynthesis of aldosterone: aldosterone synthase deficiency 1964-1997

We studied two of the three patients with a hereditary defect in the biosynthesis of aldosterone originally described by Visser and Cost in 1964. All three presented as newborns with salt-losing syndrome and failure to thrive. The original biochemical studies showed a defect in the 18-hydroxylation of corticosterone. According to the nomenclature proposed by Ulick, this defect would be termed corticosterone methyl oxidase deficiency type I. We measured plasma steroids in the untreated adult patients and performed molecular genetic studies. Aldosterone and 18-OH-corticosterone were decreased, whereas corticosterone and 11-deoxycorticosterone were elevated, thus confirming the diagnosis of corticosterone methyl oxidase deficiency type I. Cortisol and its precursors were in the normal range. Genetic defects in the gene CYP11B2 encoding aldosterone synthase (P450c11Aldo) have been described in a few cases. We identified a homozygous single base exchange (G to T) in codon 255 (GAG) causing a premature stop codon E255X (TAG). This mutation destroys a Aoc II restriction site. Digestion of a PCR fragment containing exon 4 of CYP11B2 (261 bp) with this restriction enzyme revealed in the two patients homozygous for the E255X mutation only a 261-bp fragment, whereas the heterozygous parents had three fragments (261 bp from the mutant allele and 194 and 67 bp from the wild-type allele). The mutant enzyme had lost the five terminal exons containing the heme binding site, and thus there was a loss of function enzyme. We conclude that the biochemical phenotype of these prismatic cases of congenital hypoaldosteronism can be explained by the patients genotype.     

Cholesteryl ester transfer protein gene polymorphism is a determinant of HDL cholesterol and of the lipoprotein response to a lipid-lowering diet in type 1 diabetes

The TaqIB cholesteryl ester transfer protein (CETP) gene polymorphism (B1B2) is a determinant of HDL cholesterol in nondiabetic populations. Remarkably, this gene effect appears to be modified by environmental factors. We evaluated the effect of this polymorphism on HDL cholesterol levels and on the lipoprotein response to a linoleic acid-enriched, low-cholesterol diet in patients with type 1 diabetes. In 44 consecutive type 1 diabetic patients (35 men), CETP polymorphism, apolipoprotein (apo) E genotype, serum lipoproteins, serum CETP activity (measured with an exogenous substrate assay, n = 30), clinical variables, and a diet history were documented. The 1-year response to diet was assessed in 14 type 1 diabetic patients, including 6 B1B1 and 6 B1B2 individuals. HDL cholesterol was higher in 10 B2B2 than in 14 B1B1 homozygotes (1.63 +/- 0.38 vs. 1.24 +/- 0.23 mmol/l, P < 0.01). HDL cholesterol, adjusted for triglycerides and smoking, was 0.19 mmol/l higher for each B2 allele present. CETP activity levels were not significantly different between CETP genotypes. Multiple regression analysis showed that VLDL + LDL cholesterol was associated with dietary polyunsaturated:saturated fatty acids ratio (P < 0.02) and total fat intake (P < 0.05) in the B1B1 homozygotes only and tended to be related to the presence of the apo E4 allele (P < 0.10). In response to diet, VLDL + LDL cholesterol fell (P < 0.05) and HDL cholesterol remained unchanged in 6 B1B1 homozygotes. In contrast, VLDL + LDL cholesterol was unaltered and HDL cholesterol decreased (P < 0.05) in 6 B1B2 heterozygotes (P < 0.05 for difference in change in VLDL + LDL/HDL cholesterol ratio). This difference in response was unrelated to the apo E genotype. Thus, the TaqIB CETP gene polymorphism is a strong determinant of HDL cholesterol in type 1 diabetes. This gene effect is unlikely to be explained by a major influence on the serum level of CETP activity, as an indirect measure of CETP mass. Our preliminary data suggest that this polymorphism may be a marker of the lipoprotein response to dietary intervention.