Homozygous Familial Hypobetalipoproteinemia
Increased LDL Catabolism in Hypobetalipoproteinemia Due to a Truncated Apolipoprotein B Species, Apo B-87Padova
Mutations on the apolipoprotein (apo) B gene that interfere with the full-length translation of the apoB molecule are associated with familial hypobetalipoproteinemia (FHBL), a disease characterized by the reduction of plasma apoB and LDL cholesterol. In this report, we describe an FHBL kindred carrying a unique truncated apoB form, apoB-87Padova. Sequence analysis of amplified genomic DNA identified a single G deletion at nucleotide 12 032, which shifts the translation reading frame and causes a termination at amino acid 3978. Two homozygous subjects and seven heterozygous relatives were studied. Although homozygous individuals had only trace amounts of LDL, they were virtually free from the symptoms typical of homozygous FHBL subjects. We investigated the in vivo turnover of radiolabeled normal apoB-100 LDL and apoB-87 LDL in one homozygous patient and two normal control subjects. ApoB-87 LDL showed a similar metabolism in all three subjects, with a fractional catabolic rate more than double that of normal LDL. The rate of entry of apoB-87 in the LDL compartment was also markedly decreased compared with normal apoB-100. The increased in vivo catabolism of apoB-87 LDL was paralleled in vitro by a 2.5-fold increased ability of these particles to inhibit the uptake and degradation of normal apoB-100 LDL by normal human cultured fibroblasts. These results indicate that apoB-87 LDL has an enhanced ability to interact with the LDL receptor; the increased apoB catabolism contributes to the hypobetalipoproteinemia and may explain the mild expression of the disease in the two homozygous individuals.
Presented in part at the 62nd Scientific Sessions of the American Heart Association, New Orleans, La, November 13-16, 1989, and at the 9th International Symposium on Atherosclerosis, Rosemont, Ill, October 6-11, 1991, and published in abstract form in Circulation. 1989;80(suppl II):II-276.
- Received May 30, 1995.
- Revision received February 26, 1996.
Human FHBL is a genetic disorder inherited as an autosomal-dominant trait characterized by decreased levels of LDL cholesterol and plasma apoB. In homozygous individuals, cholesterol is found at very low levels in plasma and is mainly associated with HDL, whereas plasma apoB is absent or present only in trace amounts. In the past, the “classic” clinical picture of the homozygous patient1 included fat malabsorption, clotting abnormalities due to vitamin K deficiency, red cell acanthocytosis, and after the first decade of life, severe neurological disorders caused by vitamin E malabsorption.2 Heterozygous individuals have usually been described as free from symptoms, even if their LDL cholesterol and apoB concentrations are half the normal value or less.
The analysis of the apoB gene in a number of kindreds with FHBL revealed a variety of mutations that interfere with the full-length translation of the apoB molecule.3 In general, it has been observed that longer truncated apoB species are associated with higher LDL cholesterol levels and milder symptoms.4 This phenomenon is probably due to a different ability of the apoB mutants to form stable lipoprotein particles and deliver lipids and fat-soluble vitamins from the intestine and the liver to the peripheral tissues. However, the metabolic alterations associated with these apoB mutations are only partially understood.
In this report we describe the clinical and genetic characterization of a kindred with a mutation of the apoB gene causing a truncated apoB form designated apoB-87Padova. In addition, we evaluated the in vivo LDL metabolism and the in vitro ability of LDL containing apoB-87Padova to interact with the LDL receptor. The results indicate that apoB-87–containing LDL is readily recognized by the LDL receptor and in vivo is removed from circulation at a faster rate than normal LDL. We propose that the mutation causing apoB-87 does not substantially alter the delivery of lipids to peripheral tissues despite the low circulating levels of apoB and cholesterol and may in fact be protective against atherosclerosis.
Human Subjects and Analysis of Lipoprotein and Apolipoprotein
The proband (Fig 1⇓ and Table 1⇓: subject II.3), a 42-year-old white female, was referred to our lipid clinic for persistent hypocholesterolemia. A complete clinical evaluation was performed in the proband, the homozygous sister (II.2), and a heterozygous individual (I.5) while they were inpatients on the metabolic ward of the Institute of Internal Medicine of the University of Padua. A general biochemical (SMAC-22, Technicon) and hematological profile was performed in each family member, in addition to lipid, lipoprotein, and apolipoprotein characterization. All subjects gave informed consent for the investigation, which was approved by the Clinical Review Committee of the University of Padua.
For lipoprotein and apolipoprotein analysis, blood was collected into sterile tubes containing Na2EDTA (3 mmol/L final concentration), and plasma was separated after low-speed centrifugation. Aprotinin (1000 kallikrein inhibitory units per milliliter) and sodium azide (7 mmol/L) were added to the plasma samples to inhibit protease activity. Plasma lipoproteins were isolated by preparative ultracentrifugation,5 using an L5-65 ultracentrifuge (Beckman Instruments Inc) and a Ti 50.3 rotor at densities of 1.006, 1.019, 1.063, and 1.210 g/mL, to obtain VLDL, IDL, LDL, and HDL, respectively. Triglyceride6 and cholesterol7 determinations on plasma and lipoprotein fractions were performed by manual enzymatic methods, using buffers and enzymes obtained from Biochemia (Boehringer AG). Protein concentrations were quantified according to the method of Lowry et al.8 ApoB levels were assessed by an ELISA.9 ApoE phenotyping was performed by isoelectric focusing of delipidated VLDL.10 Apolipoproteins were separated from each lipoprotein fraction by electrophoresis in 3% to 15% or 3% to 10% SDS-PAGE gradient slab gels with a Bio-Rad Protean I apparatus.11 Before electrophoresis, lipoproteins were dialyzed and delipidated as described.12 Gels were stained with 0.1% Coomassie brilliant blue R-250 or with silver (Bio-Rad Silver Stain Kit). In some cases, the apolipoproteins separated by SDS-PAGE were electrophoretically transferred onto nitrocellulose membranes for Western blotting.13 A sheep anti-human apoB polyclonal antibody raised in our laboratory was used to discriminate apoB bands; further apoB characterization was obtained using a variety of anti-apoB Mabs previously described14 and kindly provided by Dr Yves Marcel (Ottawa, Canada). The Mabs and the epitope recognized on apoB were 1D1 (aa 401 to 582), 2D8 (aa 1297 to 1482), 4G3 (aa 2980 to 3080), 5E11 (aa 3441 to 3568), BSol 20 (aa 3926 to 4005), BSol 16 (aa 4154 to 4189), BSol 2 (aa 4293 to 4355), and BSol 7 (aa 4525 to 4536).
ApoB-87 LDL particle size was evaluated after separation by nondenaturing gradient PAGE, as described.15 16 Densitometric evaluation of gel bands was obtained after scanning Coomassie-stained gels on an LKB Ultrascan XL laser densitometer with the GSXL software program (Pharmacia LKB Instruments Inc).
Genomic DNA Amplification and Sequence Analysis
Genomic DNA was prepared from leukocytes as previously reported.17 The PCR18 was used to amplify a region of the apoB gene encompassing exons 27 and 28. Oligonucleotide primer sequences used for this amplification were as follows: primer A, 5′CGTCCTACTGTTAgGAtcCTAATAAAATAC 3′ and primer B, 5′ CTCGCTCTTGGaaGCtTGTCACTCATTAGG 3′, which correspond to the apoB genomic sequences beginning 48 nucleotides upstream of exon 27 and 40 nucleotides downstream from exon 28, respectively. Lower-case letters denote primer mismatches with the native apoB sequence included to allow for directional cloning. The PCR products of the predicted size were isolated from 2% NuSieve/1% agarose gels (FMC) and either cloned into the BamHI-HindIII sites of pGem3Z (Promega) and sequenced19 or sequenced directly.20
LDL Turnover Study
Two normal adult volunteers and one apoB-87–homozygous subject (the proband's sister, S.B., II.2) were studied while inpatients on the metabolic ward of the Institute of Internal Medicine of the University of Padua. All subjects were free of renal, hepatic, hematologic, and thyroid abnormalities. No medication was allowed starting 4 weeks before the turnover study. Potassium iodide (200 mg TID orally) was begun the day before the injection and continued throughout the study to prevent thyroid uptake of radioactive iodine. A specific written informed consent was obtained from all the subjects before the study.
For the in vivo kinetic study, fasting plasma was obtained by plasmapheresis from a normal control subject (300 mL) and from the homozygous patient II.2 (1200 mL), both carrying an apoE 3/3 phenotype. LDL (d=1.019 to 1.063 g/mL) from each of the two subjects was isolated by sequential preparative ultracentrifugation at 4°C using a 60 Ti rotor (Beckman).5 The LDL was dialyzed against 0.15 mol/L NaCl-0.3 mmol/L Na2EDTA-10 mmol/L Tris, pH 7.0, and concentrated. One milliliter of each isolated LDL fraction containing 4.5 to 5.0 mg of total protein was labeled separately with 125I (apoB-87 LDL) and 131I (normal LDL) using the iodine monochloride method, as previously described.21 22 Unbound iodine was removed by gel filtration in an 8×300-mm column of G-10 Sephadex (Pharmacia) using 0.15 mol/L NaCl-0.3 mmol/L Na2EDTA, pH 7.0, in the elution process, followed by dialysis against the same buffer. Lipid labeling, as determined by chloroform-methanol (2:1, vol/vol) extraction, was less than 4% of the lipoprotein-bound radioactivity. The radioactive iodine distribution in the labeled LDL was analyzed by 3% to 10% gradient SDS-PAGE: radioactivity associated with apoB in normal LDL and apoB-87 LDL was 84% and 82% of the total, respectively. Less than 4% of the total radioactivity was recovered from the portion of the gel corresponding to the normal migration of apoE and albumin in both preparations. Human serum albumin (1% final concentration) was added to the radiolabeled lipoproteins, and the samples were sterilized by membrane filtration (Millex-GV 0.22 μm, Millipore).
At 8 am, 28 μCi of normal 131I-labeled LDL and 40 μCi of apoB-87 125I-labeled LDL were injected in rapid sequence into the antecubital vein of each subject. Blood samples were collected at 10 minutes, 30 minutes, and 1, 2, 3, 6, 9, and 12 hours, and then daily (fasting) for 1 week. The blood was collected in EDTA (3 mmol/L), cooled in crushed ice, and plasma was separated by low-speed centrifugation. LDL (d=1.019 to 1.063 g/mL) was obtained by preparative ultracentrifugation using a Beckman 40.3 rotor. Cholesterol, triglyceride, and apoB concentrations were measured in each plasma and LDL sample. Plasma and LDL radioactivity was quantitated in a Packard 5230 Autogamma spectrometer. The FCR of LDL was calculated by multicompartmental analysis of the LDL radioactivity over time curve using a computer curve-fitting program (SAAM 27)23 on a VAX-11/780 computer system (Digital Equipment Corp). The LDL-apoB pool size was calculated by multiplying the apoB concentration in LDL by the plasma volume. ApoB concentration was the mean of multiple determinations of fasting samples obtained during the study; plasma volume was assumed to be 4.5% of body weight. Production rate (PR) was calculated using the formula .
HSFs were grown from explants of skin biopsies obtained from normolipidemic healthy individuals. Cells were grown in monolayers and maintained in 75-cm2 plastic flasks at 37°C in a humidified atmosphere of 95% air, 5% CO2 in modified Eagle's medium supplemented with 10% fetal calf serum, nonessential amino acid solution (1%, vol/vol), penicillin (100 U/mL), streptomycin (100 μg/mL), tricine buffer (20 mmol/L, pH 7.4), and NaHCO3 (24 mmol/L). For all experiments, cells from the stock flasks were dissociated with 0.05% trypsin-0.02% EDTA at confluence (5 to 15 passages), seeded in 35-mm plastic Petri dishes (1.5×105 cells), and used just before reaching confluence, usually 6 days after plating. The medium was changed every 2 to 3 days. Cell viability, assessed by Trypan blue exclusion, was always >95%. For lipoprotein competition experiments, confluent cells were preincubated 24 hours at 37°C in medium containing 10% human lipoprotein-deficient serum to induce LDL receptors.24 The preincubation period was followed by the addition of fresh medium containing normal human 125I-LDL protein (7.5 μg/mL) along with increasing concentrations of unlabeled control or B-87 LDL. The cells were incubated for a further 5 hours at 37°C. For determining total uptake (binding+internalization) of LDL, cell monolayers were directly digested in 0.1N NaOH after standard washing procedures; one aliquot was counted for the cell-associated radioactivity and another aliquot was used for cell protein.8 LDL degradation was measured as the accumulation of noniodide trichloroacetic acid–soluble 125I in the incubation medium in excess of that occurring in the absence of cells.24 25 Nonspecific uptake and degradation were determined by adding a 100-fold excess of unlabeled normal LDL. Each experimental point represents the average value of triplicate incubations. The amount of unlabeled LDL required to displace 50% of the 125I-labeled ligand was calculated by linear regression analysis of the logarithm of concentration (micrograms of protein per milliliter) versus probits and read from a probit transformation table.26
Pedigree Analysis and Clinical Findings of the Proband and Sister S.B.
Hypocholesterolemia was first reported in the proband during a clinical evaluation for hypertension in 1981. When the lipoprotein evaluation was extended to the available relatives (Table 1⇑), a sister (S.B.) was found to have LDL cholesterol in the same range as the proband. Multiple determinations of LDL cholesterol and plasma apoB never exceeded 10 mg/dL (0.25 mmol/L) and 13 mg/dL, respectively, in both patients. Both parents originated from a small village in the northeastern Italian Alps, but no direct evidence of consanguinity was found in the kindred. All first-degree relatives of these two patients had hypobetalipoproteinemia (LDL cholesterol levels below the 5th percentile of the same age class). Siblings of the mother of the proband, as well as of the father (who died at the age of 63 years of cancer), showed decreased LDL cholesterol and apoB levels.
On physical examination, both the proband and sister S.B. were found to have only a moderate liver enlargement; liver ultrasonography indicated also the presence of liver steatosis in both individuals. Two liver needle biopsies, performed in the proband in 1982 and 1991, showed the presence of fat droplets in ≈25% of hepatocytes but no evidence of progression to cirrhosis over time. Fat malabsorption signs (steatorrhea, diarrhea) were absent in both individuals. Histological examination of intestinal cells obtained by endoscopic jejunal biopsy of the proband (fasting samples) did not show the presence of lipid-particle accumulation in the epithelial cells. To test the ability to acquire lipids from intestine, a fat load was administered to the proband, with a standardized fatty meal (25% cream, 10 g cocoa, 1 egg yolk, 50 mL skimmed milk, 25 g d-xylose; total fat amount, 1 g/kg body weight). After 4 hours, plasma triglyceride levels increased from 30 to 196 mg/dL, with a parallel elevation of VLDL triglycerides from 3 to 112 mg/dL (0.03 to 1.26 mmol/L). Plasma vitamin E concentration in the proband patient was normal (34.7 μg/mL).
No classic neurological manifestation was found (ie, reduction of deep tendon reflexes, presence of Romberg's sign, ataxic gait); however, both sisters had a slower nerve-conduction velocity when evaluated by electromyography. Visual acuity and night vision were not compromised, nor did ocular fundus examination report any sign of pigmentary retinal degeneration. Red blood cell count and hemoglobin concentration were in the normal range in both individuals, but acanthocytosis was present in 10% to 15% of red cells after analysis by scanning electron microscopy.
Identification of ApoB-87
The apolipoprotein content of lipoprotein fractions was evaluated by analytical gradient SDS-PAGE and Western blot analysis in both the proband and the sister S.B. (II.2). An abnormal band with a molecular weight of ≈470 kD was found in VLDL, IDL, and LDL; it reacted with polyclonal anti-apoB antibody. Following the centile system used for apoB species classification, the new apoB was designated apoB-87Padova. No apoB-87Padova protein was detected in HDL and in the >1.21-g/mL-density fraction. In subject S.B., apoB-87 was virtually the sole apoB species observed in LDL (Fig 2⇓). A similar pattern was present in the proband (S.N., II.3): apoB-87 was the main apoB form in VLDL, IDL, and LDL. Trace amounts of apoB-48 were found in VLDL and IDL of both sisters during fasting.
All other subjects in the family identified as FHBL carriers showed small amounts of apoB-87 in plasma and VLDL (Table⇑s 1 and 2). Therefore, the defect causing apoB-87Padova was apparently present in both parents' families.
Epitope mapping of apoB-87 was carried out by the Western blotting technique, using the described panel of eight Mabs. Mabs recognizing the portion of apoB homologous to apoB-48 (1D1 and 2D8) and the putative LDL receptor binding region of apoB-100 (4G3 and 5E11) readily reacted with apoB-87. The three Mabs that recognize the terminal portion of normal apoB-100 (BSol 16, BSol 2, and BSol 7) did not react against apoB-87. The carboxyl-terminal epitope recognized in apoB-87 protein placed the C-terminus of this molecule between amino acids 3926 and 4005 (Mab BSol 20), showing that the molecule is considerably shorter than the full-length 4536 amino acid mature apoB-100. Therefore, in agreement with the size indicated by SDS-PAGE, the pattern of Mab reactivity predicted a protein size between 85.6% and 88.3% of normal apoB-100.
The genetic mutation in apoB-87 was determined by sequence analysis of amplified genomic DNA. Oligonucleotide primers flanking the 27th and 28th exons of the apoB gene were used to amplify genomic DNA from the proband. The amplified 500-bp DNA was subcloned and sequenced in both directions. The nucleotide sequences of normal apoB and the proband at the region of the intron-exon junction of exon 28 are illustrated in Fig 3⇓. The normal apoB genomic DNA sequence in this region, which contains three consecutive Gs, is replaced by a sequence with a single G deletion in the proband. This deletion occurs immediately after the consensus AG dinucleotide at the 3′ end of intron 27 and effectively deletes the first base of exon 28. This G deletion at the 5′ end of exon 28, representing nucleotide 12 032 of the 14.1-kb apoB mRNA, shifts the translational reading frame, with the subsequent addition of 37 unique amino acids followed by termination at amino acid 3978 (Fig 4⇓). The predicted mutant apoB isoprotein is 87.7% of the full-length apoB-100. The sequence of the additional 37 amino acids present in apoB-87 has not been previously reported. It is of interest that five of the last eight amino acids added to apoB are prolines, which may be important in the conformation and clearance of the mutant apoB protein.
LDL Analysis and In Vivo LDL Turnover
Anthropometric features, lipoprotein values, and apolipoprotein values of the study subjects are given in Table 3⇓. The analysis of lipid-protein composition of the two LDL preparations used for the labeling is shown in Table 4⇓. Compared with normal LDL, apoB-87 LDL from S.B. was moderately protein poor and triglyceride enriched. The average diameter of LDL from patient S.B., as evaluated by nondenaturing polyacrylamide gradient gel chromatography, was 25.0 nm compared with 27.1±0.4 nm in 20 normal healthy individuals studied in our laboratory.16
The catabolic curves of normal and apoB-87 LDL in the three subjects of the study are illustrated in Fig 5⇓. The symbols represent the observed values of radioactivity remaining in the LDL compartment, while the lines show the best fit of the multicompartmental model. After simultaneous injection of the two LDL preparations, a similar metabolic behavior was observed in all three study subjects. In the two normal individuals (Fig 5B and 5C⇓⇓), 125I-LDL containing apoB-87 was catabolized at a faster rate than normal 131I-LDL. The half-life of B-87 particles was about one third of normal LDL (9 and 15 hours in normal subjects 1 and 2, respectively, versus 33 hours in both individuals). Catabolism of normal 131I-LDL in the homozygous patient (Fig 5A⇓) was only slightly faster than that observed in the control subjects and, again, apoB-87–containing LDL was removed at a faster rate, with a decay curve similar to that observed in the two normal volunteers.
Determination of the FCR gave additional information regarding the kinetic behavior of the LDL in the subjects under investigation (Table 5⇓). Normal LDL in normal subjects had FCRs of 0.47 and 0.35 d−1, which are in the range of the results of previous metabolic studies in normal human subjects.22 A comparable, even if slightly higher, FCR value was obtained for normal LDL in the hypobeta patient.
FCRs of apoB-87 LDL in normal individuals were more than twice those of normal LDL: values in control subject 1 were 1.10 versus 0.47 d−1 and in subject 2 were 0.82 versus 0.35 d−1. In patient S.B., the FCR of apoB-87 LDL (0.99 d−1) paralleled the values obtained in the normal subjects and was approximately double the FCR of normal LDL.
These results indicate that LDL containing apoB-87 is rapidly catabolized both in normolipidemic individuals and the FHBL subject. In addition, normal LDL is cleared by the hypobeta patient at a normal rate and therefore, despite the low LDL cholesterol, this patient does not show a particular upregulation of the LDL removal pathway.
Table 6⇓ shows the kinetic parameters of normal LDL in normal individuals and apoB-87 LDL in the hypobeta patient (S.B.). LDL apoB concentration in S.B. was ≈2.5% (1/40) of the values in normal control subjects, and the apoB-87 production rate was substantially decreased, being ≈7% of the normal LDL apoB production rates obtained in control individuals. Therefore, the reduced LDL concentration in the apoB-87 hypobeta patient is due to both decreased LDL production and increased apoB-87 LDL catabolism.
In Vitro Competition Assay
To determine whether the indications obtained from the in vivo turnover could be reproduced in vitro on cells expressing the LDL receptor, we studied the ability of LDL from the proband to interact with HSFs. The results are shown in Fig 6⇓. LDL from the proband was much more effective (2.5- to 3-fold) in competing with control 125I-LDL for uptake and degradation by cultured fibroblasts than was LDL from two normolipidemic subjects. The concentration of LDL required for 50% reduction of LDL uptake and degradation of control 125I-LDL by HSFs (IC50) for the proband's LDL versus normolipidemic subjects' LDL was 3.9 versus 9.7 and 2.1 versus 7.0 for uptake and degradation, respectively.
Over the last 6 years, more than 25 unique truncations of apoB were described in association with FHBL.3 These studies pointed out that unlike other mutations on the apoB gene causing hypercholesterolemia (familial defective apoB Arg3500→Gln), FHBL is quite heterogeneous at the molecular level. Most of the patients so far described are heterozygotes or compound heterozygotes, which makes it difficult to assess the impact of each mutation on the phenotypic expression of the disease. In addition, the metabolic alterations underlying the clinical picture are only partially understood. Patients homozygous for the same defect have only been described in three other kindreds. Huang et al27 reported a subject homozygous for a large 694-bp deletion of the apoB gene, which results in a complete absence of apoB in plasma and a severe clinical picture with fat malabsorption and neurological disturbances. Hardman et al28 and Malloy et al29 evaluated a homozygous patient with a truncated apoB species (apoB-50), which is rapidly removed from VLDL with virtually no conversion to LDL. More recently, Young et al30 reported an asymptomatic patient homozygous for a nonsense mutation resulting in a formation of apoB-45.2. This individual does not present vitamin E deficiency and neurological abnormalities. In contrast, the apoB-50 homozygote had virtually undetectable levels of vitamin E and severe neurological symptoms.
The present study reports the first FHBL homozygous subjects carrying a nearly full-length truncated apoB protein. The marked reductions of LDL cholesterol and plasma apoB, together with mild acanthocytosis and liver steatosis, are the only signs that characterize apoB-87Padova homozygotes on the clinical ground. Heterozygous individuals are asymptomatic, with LDL cholesterol concentrations corresponding to ≈50% of normal values. Comparing, therefore, this kindred with the previously described homozygous individuals,4 it is possible to note a progressive decrease of the clinical manifestations of the disease as the length of the abnormal apoB increases. This finding is consistent with the concept that only very short apoB species are unable to deliver efficiently lipids and fat-soluble vitamins to the peripheral tissues. By contrast, when the mutant protein approaches the size of apoB-48, as in the patient carrying apoB-50,28 clinical consequences are much milder.29 This idea of a correlation between apoB length and ability to form functionally efficient particles is strengthened by the recent report describing a linear relation between rate of secretion of total apoB and size of the molecule.31 In apoB-87 homozygotes, the apoB-48–containing lipoprotein pathway appears substantially preserved, plasma triglycerides increase sixfold after a fat load, and no signs of lipid accumulation are found in the cells of the jejunal mucosa.
The homozygous defect causing FHBL in this kindred provides a unique opportunity to investigate the metabolic consequences of the deletion of the terminal portion of the apoB molecule. The in vivo study was performed on patient S.B., in whom apoB-87 was the only apoB detectable in LDL. The data obtained in this kinetic study have two major implications. First, the observation that normal apoB-100 LDL is normally removed by the FHBL patient indicates that despite low plasma apoB and cholesterol levels, the LDL receptor pathway is not upregulated. Thus, the intracellular cholesterol pool should be normal in these patients, probably because of an efficient delivery of exogenous cholesterol via the apoB-48–containing particles, which are rapidly taken up via apoE. These results are consistent with previous in vivo studies, which pointed out the key role of an effective apoE function in the regulation of the LDL receptor activity.22
The second observation is that apoB-87 LDL has a catabolism twice as fast as normal LDL in the FHBL patient as well as in the normal control subjects. This implies that the low levels of LDL in this kindred are due at least in part to rapid removal from circulation. The calculated LDL apoB production is also markedly decreased in the apoB-87–homozygous patient. Reduced entry of apoB in the LDL compartment can be due to either decreased total production of apoB or increased removal of apoB-87 particles throughout the VLDL-IDL-LDL cascade, or both. The design of the present kinetic study cannot answer this question; however, the decreased apoB-100/apoB-87 ratio in LDL of heterozygotes compared with more buoyant fractions (Table 2⇑) is consistent with the idea of a preferential removal of apoB-87 during the delipidation process. Parhofer et al32 investigated the apoB metabolism in apoB-89/apoB-100 heterozygotes after endogenous labeling using 13C-leucine. In both subjects, apoB-89 and apoB-100 were produced at similar rates. ApoB-89 catabolism was faster (1.4 to 3 times) than apoB-100 in all the apoB-containing fractions (VLDL, IDL, and LDL) of the same individual. Compared with apoB-100, the conversion rate of apoB-89 from VLDL to LDL was reduced (84.9% versus 22.8%, respectively). The entry of B-89 in the LDL compartment was therefore greatly decreased (0.43 to 3.24 mg/kg−1·d−1) compared with B-100, while B-89 LDL FCRs were 1.5 to 2.3 times higher than B-100 LDL FCRs. The data obtained from the analysis of the two different mutations are therefore consistent with the concept that the truncation of the terminal part of apoB-100 may affect not only LDL catabolism but the entire metabolic cascade of VLDL-IDL-LDL.
The mutation in apoB-87Padova occurs downstream of the binding recognition site of apoB for the LDL receptor. The in vitro competition assays performed on HSFs show that apoB-87 LDL is more actively taken up and degraded than apoB-100 LDL; the 2.5-fold difference is similar to that observed in the in vivo turnover. Thus, fast catabolism of this truncated B species is mediated by the LDL receptor pathway.
The reason for these changes is not yet completely understood. The simplest explanation would be an increased apoE/apoB molar ratio in LDL, enhancing the affinity of the LDL containing the truncated variants for the apoB,E receptor. In our homozygous patients, although the apoE concentration was not increased in LDL due to the low concentration of apoB, the apoE/apoB ratio in this fraction showed a sixfold increase. Why the apoB-87 particle can allocate more apoE and whether apoE has an important role in the increased LDL binding remain to be clarified. A more intriguing hypothesis invokes a direct role of the last portion of apoB-100 in a modulation of LDL receptor binding. In the case of apoB-89, Krul et al33 reported that specific monoclonal antibodies, which inhibit the apoB binding in vitro, prevent the B-89 LDL uptake by fibroblast receptors, whereas an anti-apoE antibody does not. In addition, Fantappiè et al34 observed that the interaction of different murine monoclonal antibodies with human LDL can change the binding to the LDL receptor in opposite directions. Using an Mab specific for the carboxyl-terminal region of apoB-100 (aa 4082 to 4306), they found a threefold increase in LDL binding to fibroblast receptors. These data favor the idea that truncation itself is responsible for the increased clearance.
Even though this is a very stimulating hypothesis, it must be considered with caution. Multiple mechanisms can be involved at the same time in the interaction between receptors and lipoprotein particles. LDL has been shown to interact with the receptors on coated pits, depending on size. Chappell et al35 found a threefold to fourfold increased maximum binding when small LDL (19.7 nm) was compared with large LDL (27 nm), according to a lattice model for LDL binding. ApoB-87 LDL may therefore also have a better interaction with the receptor because of its smaller diameter.
In conclusion, the combined information on clinical consequences, molecular defect, and metabolic characteristics of this mutant apoB provides new insights on the structure and physiological function of this unique apolipoprotein and stimulates future studies on the role of the terminal portion of apoB-100 in LDL metabolism.
Selected Abbreviations and Acronyms
|FCR||=||fractional catabolic rate|
|HSF||=||human skin fibroblast|
|PAGE||=||polyacrylamide gel electrophoresis|
|PCR||=||polymerase chain reaction|
This work was supported in part by the National Research Council of Italy (CNR Progetto Finalizzato Invecchiamento, Sottoprogetto 3 (96.3677); Progetto Bilaterale Italia-USA 92.00387.ct04). We thank all the members of the apoB-87 kindred for their interest, help, and cooperation. We remain indebted to Dr Y. Marcel and R. Milne for providing us the anti-apoB monoclonal antibodies, Dr S. Zambon for performing the LDL size determination, and Dr F. Visioli for the vitamin E evaluation. We also thank S. Pigozzo and R. Marin for their excellent technical assistance.
Kane JP, Havel RJ. Disorder of biogenesis and secretion of lipoproteins containing the B apolipoprotein. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic Basis of Inherited Disease. New York, NY: McGraw-Hill Publishing Co; 1989:1139-1164.
Havel RJ, Eder HA, Bragdon JH. The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest. 1955;34:1345-1353.
Wahlefeld AW. Triglycerides determination after enzymatic hydrolysis. In: Bergmeyer HU, ed. Methoden der Enzymatischen Analyse. 3rd ed. Weinheim, Germany: Verlag Chemie; 1974:1878-1982.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem. 1951;193:265-275.
Warnick GR, Mayfield C, Albers JJ, Hazzard WR. Gel isoelectric focusing method for specific diagnosis of familial hyperlipoproteinemia type 3. Clin Chem. 1979;25:279-284.
Gabelli C, Stark DG, Gregg RE, Brewer HB Jr. Separation of apolipoprotein B species by agarose-acrylamide gel electrophoresis. J Lipid Res. 1986;27:457-460.
Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci U S A. 1979;76:4350-4354.
Marcel YM, Innerarity TL, Spilman C, Mahley RW, Protter AA, Milne RW. Mapping of human apolipoprotein B antigenic determinants. Arteriosclerosis. 1987;7:166-175.
Krauss RM, Burke DJ. Identification of multiple subclasses of plasma low density lipoproteins in normal humans. J Lipid Res. 1982;23:97-104.
Manzato E, Zambon S, Zambon A, Cortella A, Sartore G, Crepaldi G. Levels and physicochemical properties of lipoprotein subclasses in moderate hypertriglyceridemia. Clin Chim Acta. 1993;218:57-65.
Sambrook H, Fritsch EF, Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Press; 1989.
Saiki RK, Gelfand DH, Stoffel S, Sharf SJ, Higuchi R, Horn GT, Mullis KB, Erlich HA. Primer-directed enzymatic amplification of DNA with thermostable DNA polymerase. Science. 1988;239:487-491.
Sanger F, Nicklen S, Coulsen AR. DNA sequencing with chain terminating inhibitors. Proc Natl Acad Sci U S A. 1977;74:5463-5467.
Gabelli C, Gregg RE, Zech LA, Manzato E, Brewer HB Jr. Abnormal low density lipoprotein metabolism in apolipoprotein E deficiency. J Lipid Res. 1986;27:326-333.
Berman M, Weiss M. SAAM Manual. Bethesda, MD: US Dept of Health, Education, and Welfare publication NIH 78-180; 1977:1-198.
Goldstein JL, Basu SK, Brown MS. Receptor-mediated endocytosis of low density lipoproteins in cultured cells. Methods Enzymol. 1982;98:241-260.
Corsini A, Roma P, Sommariva D, Fumagalli R, Catapano AL. Autoantibodies to the low density lipoprotein receptor in a subject affected by severe hypercholesterolemia. J Clin Invest. 1986;78:940-946.
Fisher RA, Yates F. Statistical Tables for Agricultural and Medical Research. 4th ed. Edinburgh, Scotland: Oliver & Boyd; 1953:1-60.
Huang LS, Ripps ME, Korman SH, Deckelbaum RJ, Breslow JL. Hypobetalipoproteinemia due to an apolipoprotein B gene exon 21 deletion derived by Alu-Alu recombination. J Biol Chem. 1989;264:11394-11400.
Hardman DA, Pullinger CR, Hamilton RL, Kane JP, Malloy MJ. Molecular and metabolic basis for the metabolic disorder normotriglyceridemic abetalipoproteinemia. J Clin Invest. 1991;88:1722-1729.
Malloy MJ, Kane JP, Hardman DA, Hamilton RL, Dalal KB. Normotriglyceridemic abetalipoproteinemia: absence of the B-100 apolipoprotein. J Clin Invest. 1981;67:1441-1450.
Young SG, Bihain B, Flynn LM, Sanan DA, Ayrault-Jarrier M, Jacotot B. Asymptomatic homozygous hypobetalipoproteinemia associated with apolipoprotein B45.2. Hum Mol Genet. 1994;3:741-744.
Parhofer KG, Barret PHR, Aguilar-Salinas CA, Richter WO, Schonfeld G. The length of truncated apolipoprotein B correlates linearly with its in vivo secretion rate. Circulation. 1994;90(suppl I):I-185. Abstract.
Parhofer KG, Barret PHR, Bier DM, Schonfeld G. Lipoproteins containing the truncated apolipoprotein, apo B-89, are cleared from human plasma more rapidly than apo B-100–containing lipoprotein in vivo. J Clin Invest. 1992;89:1931-1937.
Krul ES, Kinoshita M, Talmud P, Humphries SE, Turner S, Goldberg AC, Cook K, Boerwinkle E, Schonfeld G. Two distinct truncated apolipoprotein B species in a kindred with hypobetalipoproteinemia. Arteriosclerosis. 1989;9:856-868.
Fantappiè S, Corsini A, Sidoli A, Uboldi P, Granata A, Zanelli T, Rossi P, Marcovina S, Fumagalli R, Catapano AL. Monoclonal antibodies to human low density lipoprotein identify distinct areas on apolipoprotein B-100 relevant to low density lipoprotein-receptor interaction. J Lipid Res. 1992;33:1111-1121.
Chappell DA, Fry GL, Waknitz MA, Berns JJ. Ligand size as a determinant for catabolism by the low density lipoprotein (LDL) receptor pathway: a lattice model for LDL binding. J Biol Chem. 1991;266:19296-19302.