© 1999 American Heart Association, Inc.
Atherosclerosis and Lipoproteins |
Anderson's Disease
Exclusion of Apolipoprotein and Intracellular Lipid Transport Genes
Correspondence to Dr Lawrence P. Aggerbeck, Centre de Génétique Moléculaire, Centre National de la Recherche Scientifique, Avenue de la Terrasse, Bât. 26, 91198 Gif-sur-Yvette, France. E-mail aggerbeck{at}cgm.cnrs-gif.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Abstract—Anderson's disease is a rare, hereditary hypocholesterolemic syndrome characterized by chronic diarrhea, steatorrhea, and failure to thrive associated with the absence of apo B48–containing lipoproteins. To further define the molecular basis of the disease, we studied 8 affected subjects in 7 unrelated families of North African origin after treatment with a low-fat diet. Lipid loading of intestinal biopsies persisted, but the pattern and extent of loading was variable among the patients. Electron microscopy showed lipoprotein-like particles in membrane-bound compartments, the densities (0.65 to 7.5 particles/µ2) and the mean diameters (169 to 580 nm) of which were, in general, significantly larger than in a normal fed subject (0.66 particles/µ2, 209 nm mean diameter). There were also large lipid particles having diameters up to 7043 nm (average diameters from 368 to 2127 nm) that were not surrounded by a membrane. Rarely, lipoprotein-like particles 50 to 150 nm in diameter were observed in the intercellular spaces. Intestinal organ culture showed that apo B and apo AIV were synthesized with apparently normal molecular weights and that small amounts were secreted in lipid-bound forms (density <1.006 g/mL). Normal microsomal triglyceride transfer protein (MTP) and activity were also detected in intestinal biopsies. Segregation analyses of 4 families excluded, as a cause of the disease, significant regions of the genome surrounding the genes for apo AI, AIV, B, CI, CII, CIII, and E, as were the genes encoding 3 proteins involved in intracellular lipid transport, MTP, and fatty acid binding proteins 1 and 2. The results suggest that a factor other than apoproteins and MTP are important for human intestinal chylomicron assembly and secretion.
Key Words: Anderson's disease • chylomicron retention disease • hypocholesterolemia • apolipoprotein B • malabsorption
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
There are 3 inherited lipoprotein deficiency states, characterized by hypocholesterolemia, which affect apo B–containing lipoproteins: abetalipoproteinemia, familial hypobetalipoproteinemia, and Anderson's disease or chylomicron retention disease.1
Abetalipoproteinemia is a rare disease. This autosomal recessive disorder manifests itself in infancy, and although the clinical expression is variable, it is characterized by profound hypocholesterolemia, hypotriglyceridemia, lipid malabsorption, diarrhea, retinitis pigmentosa, acanthocytosis, spinocerebellar degeneration, and the complete absence of apo B–containing lipoproteins. A defect in lipoprotein assembly is due to mutations in the gene encoding the large subunit of the microsomal triglyceride transfer protein (MTP) (see reference 2 for a review).
Familial hypobetalipoproteinemia has an autosomal codominant form of transmission. Subjects who are homozygous for "null-alleles," in which plasma apo B is absent, are phenotypically similar to subjects having abetalipoproteinemia (see reference 3 for a review). Subjects who are homozygous but in which truncated forms of apo B are found in the plasma are, in general, asymptomatic clinically, as are subjects who are heterozygous for the disease. However, in both cases there are decreased plasma and LDL cholesterol levels. The disease arises from mutations in the apo B gene. Based on the levels of plasma apo B, the frequency of the heterozygous form of familial hypobetalipoproteinemia has been estimated to be 1/500 to 1/1000 in Western populations.3
Anderson's disease and chylomicron retention disease are the terms used to describe subjects having hypobetalipoproteinemia with selective absence of apo B48.1 Thirty-five cases4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 have been reported, including the following 2 series: one from France consisting of 10 cases12 under the name of Anderson's disease and one from Canada consisting of 8 cases14 15 under the name of chylomicron retention disease. Subjects with these disorders exhibit the clinical manifestations initially described by Anderson and colleagues4 that consist of a malabsorption syndrome with steatorrhea and growth retardation. Endoscopy shows a white stippling-like hoar frosting covering the mucosal surface of the small intestine. Neurological signs, although variable, consist most frequently of a loss of reflexes. Biochemically, there is an absence of apo B48–containing lipoproteins and no postprandial chylomicrons are detected, whereas lipoproteins containing apo B100 are present, although in decreased amounts. There are low plasma levels of HDL as well as low levels of total lipids, cholesterol, phospholipids, carotenoids, and lipid soluble vitamins. There are, however, normal fasting triglyceride levels. There is little acanthocytosis and no retinitis pigmentosa. Both absorption of luminal fatty acids and their consecutive esterification by the epithelial cells appear normal. On institution of a low-fat diet supplemented with lipid soluble vitamins (A and E) and essential fatty acids, normal growth resumes with abatement of the gastrointestinal symptomatology. Departure from the low-fat diet results in rapid relapse and recurrence of symptoms.
Immunoenzymatic12 14 and organ culture studies15 have shown the presence of apo B48 in enterocytes, which suggests that mutations in the apo B gene are not the cause of the disease. Further, genetic linkage studies have excluded the apo B gene in 8 cases of Anderson's disease in 3 families17 18 and in 2 cases (1 family) of chylomicron retention disease.19 The apo B messenger RNA appeared to be correctly edited in 2 cases of chylomicron retention disease in 1 family,19 suggesting that a defect in intestinal apo B editing is not involved. Intestinal ultrastructural studies have shown that villi are present in normal number and length but that the enterocytes are overloaded with fat droplets. In contrast to abetalipoproteinemia, the presence of chylomicron and VLDL-sized particles has been noted in membrane-bound compartments in the enterocytes of some patients,7 8 14 18 20 21 suggesting that lipoprotein assembly may take place. Roy and colleagues14 proposed the name chylomicron retention disease for the disorder in their patients. The molecular basis of Anderson's disease (chylomicron retention disease) has not been established.
The purpose of this study was to investigate the roles of the major apolipoproteins and 3 intracellular lipid transport proteins in the pathogenesis of Anderson's disease. We performed an ultrastructural study of the enterocytes to assess the nature of the lipid accumulations in the cells of our patients as compared with those in other patients with Anderson's disease and chylomicron retention disease. We further used intestinal organ culture to evaluate the biosynthesis and secretion of major apolipoproteins and MTP. Finally, we looked for evidence of genetic linkage between apolipoprotein and intracellular lipid transport genes and Anderson's disease using highly polymorphic genetic markers.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Pedigrees
The protocols used in these studies were approved by the Institut National de la Santé et de la Recherche Médicale as part of a Biomedical Project (RBM 93018) and by a bioethics committee, Comité Consultatif de Protection des Personnes dans la Recherche Biomédicale de Paris-Hôpital Bichat-C. Bernard, registration No. 94002. Informed, written consent was obtained from all participants or from their legal guardians. The diagnosis of Anderson's disease was made on the basis of the clinical findings (diarrhea, steatorrhea, and malabsorption), endoscopy (presence of a "white hoary" layer or "gelée blanche" on the small intestine mucosa), the intestinal biopsy (vacuolated enterocytes that stain positively with oil red O), the levels of plasma cholesterol (hypocholesterolemia), the presence of apo B100 and the absence of apo B48 in the plasma, and the absence of chylomicrons and apo B48 in the plasma after a fat feeding. The parents of the patients were asymptomatic. The patients with pedigrees 100 (Sa.B. and So.B.), 200 (M.K.), 400 (A.B.), 500 (A.C.), and 600 (Y.Z.); Figure 1
; and
M.B. have been described in the
literature.5 7 9 12 16 17 The pertinent clinical
and biochemical features of the patients have been summarized in Table 1. The causes of the deaths of the
3 siblings of M.K. are unknown. Previously unpublished plasma lipid and
apolipoprotein values for the father and mother, respectively, of M.K.
are cholesterol, 5.77 and 4.42 mmol/l (normal=4 to 7);
triglyceride, 3.43 and 1.26 mmol/l (normal=0.5 to 2);
apo B, 151 and 94 mg/dL (normal=60 to 100); and apo AI, 135 and 131
mg/dL (normal=120 to 160). Because no details concerning the siblings
of M.B. are available, other than the fact that an older brother and 2
older sisters may be affected, the pedigree of this patient is not
shown.
|
|
The propositus of pedigree 300 (O.A.) was born January 29, 1985, the third of 3 children from Algerian parents. The major symptom was intermittent diarrhea that began at 4 months of age. At 2 years of age, a diagnosis of celiac disease was made and a gluten free diet was begun. She presented to the hospital at age 6 years with growth retardation (-4 SD in weight and -0.5 SD in height), steatorrhea, diarrhea, and abdominal swelling. The deep tendon reflexes were normal. Osteoporosis was present and the bone age was estimated, radiographically, to be 3 years. Hematological analysis showed an anemia without the presence of acanthocytosis. Plasma lipid values (mmol/l) were triglycerides 0.65; cholesterol, 2.25; and phospholipid, 1.25 (N=1.8 to 3.2). Plasma apoprotein values (mg/dL) were apo AI, 50; apo AII, 55 (normal=40 to 60); and apo B, 62. The vitamin A level was 1.43 µmol/l (N=1.1 to 2.9) and the vitamin E level was 4 µmol/l (N=14 to 32). Endoscopy showed a "white hoary frosting" (gelée blanche) and the intestinal biopsy showed fat loaded enterocytes in an otherwise normal mucosa. Plasma cholesterol levels of the father and the mother were 4.1 and 4.4 mmol/l, respectively. The diagnosis of Anderson's disease was made on the criteria outlined above after elimination of other possible causes of malabsorption. After installation of a low-fat diet and vitamin A and E supplementation, digestive symptoms abated.
Intestinal Biopsy Specimens
Intestinal biopsy specimens (~5 to 10 mg each) were obtained
between 8 and 10 AM during upper gastrointestinal endoscopy
for follow-up from patients with Anderson's disease who had fasted for
12 to 15 hours. Biopsies were obtained with a pediatric endoscopic
capsule in the second or third part of the duodenum. The patients had
been treated with a low-fat diet supplemented with vitamins A and E for
at least 6 months leading up to the endoscopic procedure. Intestinal
biopsy specimens were obtained from 11 fasted normal subjects who had
been following their normal diet and who were undergoing
endoscopy for unrelated causes. An intestinal biopsy was also obtained
from one normal subject 1.5 hours after 50 mL olive oil was instilled
into the stomach. The preparation of ultrathin sections of biopsies for
electron microscopy and western blots for MTP protein and the assay for
MTP activity were performed as previously described (references 22 and
23, respectively). The 99% confidence intervals for normal MTP
activity were calculated using the Sigma-Stat statistical
analysis software (Jandel Scientific) from the measurements of
MTP activity in 5 normal subjects. The particle densities and mean
particle diameters in biopsies were measured on photographic
enlargements (6000 to 96 000) of electron micrographs and statistical
analysis was performed with the SigmaStat software. Particle
density measurements were made in several different regions of the
biopsy and included several hundred particles. Between 115 and 561
particles were measured for the calculation of the particle diameters
for each patient. Organ culture of intestinal fragments and
metabolic labeling protocols were begun <30 minutes after
biopsy.22 24 25 The incorporation of labeled methionine
into the cellular proteins was evaluated by precipitating an aliquot
with trichloroacetic acid (TCA) followed by scintillation counting.
Immunoprecipitation of biopsy homogenates or culture media
was performed with polyclonal antibodies directed against the protein
of interest. Rabbit polyclonal antibodies to apo B and apo AIV were the
gifts of Dr A. Mazur of Institut National de la Recherche
Agronomique, Champanelle, France and Dr. L. Lagrost, Institut
National de la Santé et de la Recherche Médicale, Dijon,
France. Antibodies to MTP were as previously described.23
Treatment of the immunoprecipitates with endo-ß-N-acetyl
glucosaminidase H and endoglycosidase
F/peptide-N4-(acetyl-ß-glucosaminyl)-asparagine
amidase (PNGase F) to assess the extent of glycosylation of apo AIV was
performed according to the manufacturer's protocol (Boehringer
Mannheim). After immunoprecipitation, samples were analyzed
using sodium dodecyl sulfate (SDS) polyacrylamide gels
as previously described.22 24 25 26 For the quantitative
analysis of apo AIV and apo B, all of the immunoprecipitated
material was loaded onto the SDS gel. After
autoradiography, densitometry was performed and the
values were normalized with respect to the amount of TCA precipitable
incorporated material in the biopsies. Chylomicrons in the biopsy or
secreted into the biopsy culture medium were isolated by isopycnic
centrifugation at density 1.006 g/mL at 40 000 rpm in
a 50Ti rotor for 1 hour at 4°C, as previously
described.27
DNA Isolation and Genotyping
Genomic DNA was extracted from peripheral blood
leukocytes from patients and from their families (pedigrees 100, 200,
300, and 400) using standard techniques.28
Oligonucleotides for genotyping MTP, apo B, apo CII,
and apo CIII were as previously described.29 30 31 32 For these
genes and for the other genes (apo AI, apo AIV, apo E, and FABP 1 and
2), highly polymorphic microsatellite-type markers spanning the
regions where these genes are located33 34 were also used
(see Tables 3 to 6).
|
|
|
|
DNA amplification was performed by polymerase chain reaction (PCR) in a Hybaid thermocycler using a standard program of denaturation at 94°C for 10 minutes followed by 30 cycles of amplification consisting of denaturation at 94°C for 30 seconds, annealing at 55°C for 30 seconds, and extension at 72°C for 30 seconds and terminated by 5 minutes at 72°C. PCR products were separated on 6% denaturing polyacrylamide gels28 31 and then transferred to Hybond N+ nylon membranes. The membrane was probed with a digoxigenin-labeled (CA)12 probe. Chemiluminescent detection with Lumigen-PPD,CSPD was performed according to the manufacturer (Boehringer Mannheim) using an antidigoxigenin antibody labeled with alkaline phosphatase. The allele sizes were scored relative to each other.
Linkage Analyses
Linkage analyses between the microsatellite
polymorphic markers and Anderson's disease were performed by
computer using the Linkage program package version
5.2.35 36 37 An autosomal recessive mode of transmission was
assumed (see Discussion). Significant evidence of linkage to the
disease requires a LOD score of 3 or more (1000 to 1 odds). Exclusion
of a locus is generally accepted for negative LOD scores of 2 or
more.36 37 Further, the negative test is considered
significant for values of the recombination fraction, 
, for which
the LOD score has negative values of 2 or more, and the disease gene is
excluded from this region of the genome.38 39
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Intestinal Ultrastructure
Intestinal biopsy sections from fasted patients with Anderson's disease are heterogeneous in appearance. First, for a given patient, villi in different biopsy sections are affected to a variable extent, although all sections contain at least 1 abnormal villus (abnormal meaning containing heavily lipid laden enterocytes). Second, the percent of lipid-laden villi varies among the patients. Thus, although 100% of the villi are lipid-laden in the case of Y.Z., only 50% are lipid-laden in the cases of M.K. and O.A., and the other patients have values between these 2 extremes (Table 2
). When <100% of the villi are
lipid-laden, the remaining villi appear completely normal. Third, the
region of the villus that is lipid-laden also varies among the
patients. They may be affected characteristically on both sides (Sa.B.,
So.B., M.K., A.B., A.C., and Y.Z.) or only on 1 side (M.B.) or on 1
side in some sections and on both sides in others (O.A.) (Table 2). In all the biopsy sections from all the patients, the cells
in approximately the inferior one-third of the villus show
no accumulation of lipids. Finally, even in the regions of the
villi that contain lipid-laden enterocytes, there are always a few (a
minority) morphologically normal appearing cells.
|
Biopsy sections from both normal fasted and fed subjects and
Anderson's disease subjects were examined extensively by electron
microscopy. Most enterocytes from normal fasted subjects have no
intracellular and intercellular lipoprotein particles and exhibit a
flat, nondistended Golgi apparatus (Figure 2A). In contrast, enterocytes from a
normal fed subject show intracellular (white arrows in Figure 2B) and intercellular VLDL and chylomicron-sized particles
(black arrows in Figure 2B). The intracellular lipoproteins are
situated in membrane-bound vesicles and are seen budding from or
located close to the Golgi apparatus. The lipoproteins are
very heterogeneous in size, with diameters up to
approximately 600 nm and a mean diameter of 209 nm (Table 2
).
The extracellular lipoprotein particles are also very
heterogeneous with a size distribution (mean diameter of
226 nm) roughly similar to that of the intracellular particles (Table 2 and Figure 3).
|
|
The heterogeneity in the biopsy sections of Anderson's
disease subjects visible on light microscopic examination is also
apparent when the biopsies are examined by electron microscopy.
Extensive observation in the region of the villus tips of the biopsies
(sections from different tissue blocks) of all the patients show that,
in some regions (as represented by the biopsy from Sa.B.,
Figure 2C), the enterocytes have an intracellular architecture
similar to that found in normal fasted subjects. Intracellular and
intercellular lipoprotein-like particles are not readily apparent, and
the Golgi apparatus is flat and nondistended. In contrast,
in other regions of the same biopsy (and this is the case for
all the patients), the enterocytes contain large amounts of lipid
particles that are scattered throughout the cytoplasm and alter the
normal architecture of the cells. This lipid-laden morphology, which is
typical of all the patients studied, is demonstrated by the biopsy from
M.K. (Figure 2D). Many of these particles are clearly situated
in membrane-bound compartments (white arrows, Figure 2D). Other
larger particles appear to be lipid droplets located in the cytoplasm
(labeled L, Figure 2D).
Examination of these regions at higher magnification more clearly
shows the heterogeneous nature of the lipid particles, as
demonstrated by the biopsy from So.B. (Figure 2E). There are
large lipid droplets, up to 7043 nm in diameter that appear to be
located within the cytoplasm (labeled L, Figure 2E). These
droplets are not in a membrane-bound compartment; there is neither a
translucent space nor a cell membrane surrounding the droplets. The
density of these particles (expressed as the number of particles per
square micron) is variable among the patients (from 0.09 to 0.71
particles/µ2) and is significantly larger than
the very limited number of intracytoplasmic lipid particles apparent in
a normal fed subject (0.03 particles/µ2) (Table 2). The mean diameters of the lipid particles vary greatly from
368 to 2127 nm in the Anderson's disease subjects and are smaller
(except for Sa.B.) than the mean diameter of the intracytoplasmic lipid
particles in normal fed subjects (2271 nm) (Table 2). No
lipid-like particles were observed in the case of O.A. In addition to
these large lipid droplets, there are smaller chylomicron and
VLDL-sized particles. These particles are surrounded by a translucent
space within a membrane-bound compartment (white arrows, Figure 2E). Frequently, there are 1 or a few particles (<ten) in these
compartments, which are reminiscent of dilated, vesiculated channels of
the smooth endoplasmic reticulum but could also be transverse sections
of Golgi apparatus. In addition, there are large
membrane-bound compartments that contain many (50 or more) particles
that are very heterogeneous in size. When clearly
identifiable, the Golgi apparatus (stars in Figure 2E) is empty, although frequently distended, with membrane-bound
compartment containing particles in close juxtaposition. These
membrane-bound lipid particles resemble lipoprotein particles seen in a
normal fed subject; (Figure 2B) however, their size distribution
is frequently skewed toward larger sizes (Figure 3). The mean
diameter for all the Anderson's disease subjects is 305 nm,
significantly larger as compared with the mean diameter of 209 nm in a
normal fed subject (Table 2). Considered individually, the mean
particle diameters of all but 2 of the Anderson's disease subjects are
significantly larger than those of the normal fed subject (Table 2). The mean particle diameter of A.C. was not
significantly different from that of the normal fed subject, whereas
that of A.B. was significantly smaller (Table 2). Even among the
Anderson's disease subjects, the mean lipoprotein-like particle size
varies significantly. Further, in Anderson's disease, the average
density of the membrane-bound lipoprotein-like particles, 2.34
particles/µ2, greatly exceeds that found in the
enterocytes of a normal subject, 0.66
particles/µ2, after a meal (Table 2).
Also, when considered individually, the density of membrane-bound
lipoprotein-like particles in Anderson's disease subjects (except for
Sa.B. and O.A.) is significantly larger than that of the normal fed
subject (Table 2).
The extracellular spaces in the biopsies of the patients are
generally devoid of lipoprotein particles. Sometimes, however, small
amounts of intercellular lipoprotein-like particles are observed (as
shown for A.C., black arrows, Figure 2F), which have a range of
diameters from 50 to 150 nm with a mean diameter of 63±19 nm, not
significantly different from normal fasted subjects in which the mean
diameter of intercellular lipoprotein particles is 83±45 nm. Thus the
ultrastructural results show that the enterocytes of Anderson's
disease patients have large amounts of lipoprotein-like particles that
are larger than lipoprotein-like particles in normal fed subjects and
that they secrete very limited amounts of particles, which have
diameters resembling small chylomicrons or VLDL.
Biosynthesis of Apolipoproteins and MTP in Intestinal Organ
Culture
Immunoprecipitation with polyclonal antibodies to apo B of the
total homogenate of the organ culture of intestinal
biopsies from Anderson's disease patients (Sa.B., So.B., M.K., O.A.,
A.C.) shows the synthesis of an apo B48 identical in size to that of
normal control subjects. The results for one of the patients (A.C.) are
shown in Figure 4A. Densitometric
scanning reveals that there is 3- to 5-fold more apo B48 present in
the biopsies of the patients as compared with those of normal subjects
when corrected for the amount of TCA precipitable incorporated
radioactivity. Some intracellular degradation of the apo B48 may occur
as evidenced by the lower molecular weight bands in the patient's
biopsy as compared with the normal biopsy. Analysis of the
culture media shows that all of the patients secrete a small amount apo
B48 having an apparent molecular weight identical to that of the apo
B48 secreted by normal subjects (as illustrated by the results with
A.C. in Figure 4A). Secretion of apo B48 from the biopsies of
both affected and normal subjects began with periods of pulse or chase
longer than 30 minutes. Preparative ultracentrifugation
reveals that the apo B48 secreted (after a 2-hour pulse followed by a
1-hour chase) by patients Sa.B., So.B., O.A., M.K., and normal subjects
floats like chylomicrons at a density of <1.006 g/mL (as shown for
So.B. in Figure 5A). No apo B100
was detected in the biopsy or culture medium.
|
|
As shown by the results obtained from A.C. (Figure 4B),
immunoprecipitation of biopsy homogenates from 5 patients
(M.K., So.B, Sa.B., O.A., and A.C. [4 families]) with a polyclonal
antibody to apo AIV showed that the patients synthesized and secreted
apo AIV having a molecular weight identical to that found in normal
subjects. Two proteins were found in the biopsies of both patients and
normal subjects. The larger of the 2 proteins corresponded to the 1
secreted into the medium and had an electrophoretic mobility identical
to pure plasma apo AIV. As quantified by immunoprecipitation and
densitometric scanning, there were increased amounts of the proteins in
the biopsies of Sa.B. (3.9-fold), So.B. (8.8-fold), and A.C.
(1.7-fold), as compared with the biopsy of the normal subject when
corrected for the amounts of trichloroacetic acid precipitable protein
incorporated radioactivity. Although there appears to be an increased
amount of the lower molecular weight protein relative to higher
molecular protein in the biopsy of A.C. compared with the normal biopsy
(Figure 4B), examination of several normal biopsies showed that
the lower molecular weight protein may predominate in normal persons as
well.
To determine whether the 2 intracellular protein forms were glycosylated differently, immunoprecipitated apo AIV was treated with endo-ß-N-acetyl glucosaminidase H or endoglycosidase F/PNGase F to remove high mannose and complex carbohydrates, respectively. No effect was observed for either endoglycosidase for biopsies from Sa.B. and So.B., suggesting that differences in N-glycosylation were not the origin for the 2 bands (results not shown). The origin of the lower molecular weight protein is unknown at present (proteolysis, alternative splicing).
Analysis of the culture medium by
ultracentrifugation and gel electrophoresis was
performed for Sa.B., So.B., and M.K. and showed that apo AIV was
present with apo B48 in a chylomicron-like fraction that floated at
a density <1.006 g/mL, as shown for M.K. in Figure 5B. Further,
when this chylomicron-like fraction was immunoprecipitated with
antibodies to apo B, both apo B48 and apo AIV could be detected as
shown in the case of M.K. in Figure 5C. The protein immediately
below apo AIV corresponds to that observed in biopsies (Figure 4B) and sometimes may appear in the medium. The other
proteins are unidentified at present. These results indicate, in
agreement with the ultrastructural results, that a small amount of
chylomicron-like particles can be secreted by the enterocytes of
Anderson's disease patients.
Because MTP has been shown to play an integral role in lipoprotein
assembly, we examined biopsy homogenates for MTP protein on
western blots and for MTP activity. The heavy (97 kDa) subunit was
readily detected in the 3 patients, Sa.B., So.B., and O.A., from
pedigrees 100 and 300 (results not shown). The in vitro assay results
revealed that the MTP activities in the biopsy homogenates
from these same patients were within the 99% confidence intervals
established by the in vitro assays of 5 normal subjects (Figure 6). These data indicate that a defect in
MTP was highly unlikely as the molecular basis of the disease.
|
indicates patient
O. A.; •, Sa. B.; and 
, So. B. The shaded area
represents the 99% confidence interval around the mean of the
MTP activities of 5 normal subjects.
Exclusion of the Apolipoprotein and Lipid Transfer Protein
Genes
We next used highly polymorphic microsatellite markers
to perform linkage studies between several candidate genes and
Anderson's disease in 5 cases (So.B., Sa.B., M.K., O.A., and A.B.) in
4 families, pedigrees 100, 200, 300, and 400 (Figure 1).
Because there are essentially no apo B48–containing lipoproteins, the
levels of plasma LDLs are reduced in the plasma of patients with
Anderson's disease, and given the small number of patients with the
disease, it is important to evaluate linkage to the apo B gene in all
cases to assess the role of this protein and to determine possible
heterogeneity in the cases. The apo B gene was studied
using a tetranucleotide repeat located within the gene and
with 2 highly polymorphic markers, D2S305 and D2S220, immediately
surrounding the location of the gene on chromosome 2 (Table 3). No evidence for linkage was found
with the tetranucleotide repeat and the negative LOD scores
of the 2 other markers clearly excluded any association for a distance
of at least 5 cM on either side of the reported location of the
gene. These results, combined with the intestinal organ culture
results, clearly rule out the apo B gene as the origin of the disease
in these families.
Apo AIV and apo AI are major intestinal apoproteins and are present
in chylomicrons. There are increased amounts of these proteins in the
enterocytes of patients with Anderson's disease12 and
chylomicron retention disease,14 and the plasma levels of
these apolipoproteins as well as those of HDL are reduced. We therefore
evaluated the relationship of these major apolipoprotein genes to the
disease. No highly polymorphic markers have been reported in the
apo AIV or apo AI genes on chromosome 11. However, a polymorphic
marker, apo CIII-IVS3, does exist within the apo CIII gene. We
therefore investigated this marker as well as a series of
polymorphic markers surrounding the apo AIV, apo AI, and apo CIII
gene cluster (Table 4). The apo CIII gene
is clearly excluded. A region of approximately 15 cM in length around
the location of the apo AI-CIII-AIV gene cluster is also excluded as
being linked to the disease on the basis of the LOD scores of the
various recombination fractions (Table 4). These results, along
with the organ culture results, effectively exclude mutations in the
apo AIV, apo AI, and apo CIII gene cluster as being the cause of the
disease.
Marked immunoenzymatic labeling of the C apoproteins in enterocytes and
decreased plasma levels of these proteins have been noted in
Anderson's disease.12 These proteins are also chylomicron
constituents; therefore, we evaluated the relationship to the disease
of the other C apoproteins along with apo E. Within the apo CII gene on
chromosome 19 there is a polymorphic dinucleotide
repeat. We studied this marker as well as 3 other polymorphic
markers spanning a region of 18 cM containing the apo CI, apo CII, and
apo E genes (Table 5). Linkage
analysis shows no evidence of segregation between Anderson's
disease and this region of chromosome 19. The apo CII gene is clearly
excluded by the negative LOD scores found for the marker within the
gene. A region of 24 cM around the apo CII gene, which includes the apo
CI and apo E genes, is also excluded.
Because a mutation in one of the major apolipoprotein genes did not appear to be the cause of Anderson's disease, we also performed linkage studies on 3 intracellular lipid transport proteins that are potential candidates for the origin of the disease. MTP plays a major role in triglyceride-rich lipoprotein assembly, and mutations in the heavy subunit are known to cause abetalipoproteinemia.2 Although we have found apparently normal MTP activity in 3 patients, a genetic test would provide definitive evidence for the lack of linkage to the disease.
The patterns of inheritance in pedigrees 100, 200, and 300 for
the polymorphic marker MTPIVS10 located within the MTP gene are
shown in Figure 7. In each pedigree, the
presence of different genotypes in affected subjects (pedigree
100) or of an identical genotype in affected and unaffected
subjects (pedigrees 200 and 300) effectively rules out the MTP gene
locus as the site for the mutation causing the disease. We have found
that the MTP gene is closely linked to the marker D4S406 (Dannoura et
al, unpublished results, 1998). Analysis of the LOD
scores as a function of the recombination fractions in the 4 families
between the disease and the locus D4S406 (Table 6) also excludes this locus in agreement
with the above results. These results, combined with the results
obtained with the other highly polymorphic markers on chromosome 4,
exclude an extended area of approximately 30 cM surrounding the MTP
gene on chromosome 4 (Table 6) as a cause of the disease.
|
Fatty acid binding proteins perform important intracellular lipid
transport functions that may be essential for
triglyceride-rich lipoprotein production.
Chromosome 4 also contains the gene for intestinal fatty acid binding
protein, FABP2. Highly polymorphic markers spanning a region of
more than 15 cM on either side of the reported location for the FABP2
gene were analyzed. The LOD scores and recombination fractions
(Table 6) effectively exclude the FABP2 gene as well as any
other genes located in an area of 30 cM surrounding the gene.
No highly polymorphic markers have been described in the FABP1
gene, which is located on chromosome 2. We therefore examined several
markers spanning a region of 12 cM around the reported location of the
gene. On the basis of the negative LOD scores and the recombination
fractions for all the polymorphic markers studied (Table 3),
a region of 20 cM surrounding the location of the FABP1 gene is
excluded as being linked to the disease.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The study of inherited disorders of apo B–containing lipoproteins have provided important insights into the roles of apolipoprotein (apo B in the case of hypobetalipoproteinemia) and non-apolipoprotein (MTP in the case of abetalipoproteinemia) factors in normal human chylomicron and VLDL assembly and secretion. In the present report, we have performed biochemical, cell biological, and ultrastructural analyses to assess the role of apolipoprotein and non-apolipoprotein factors in the pathogenesis of Anderson's disease. We have, in addition, specifically tested the hypothesis that 1 of the major apolipoprotein genes or that 1 of 3 important intracellular lipid transport genes is linked to Anderson's disease.
We first investigated the intestinal biopsies of 8 affected subjects from 7 families. Our patients had been maintained on a low-fat, lipid-soluble vitamin-supplemented diet for at least 6 months before their biopsy, which was performed after 12 to 15 hours fasting. Even after this treatment and abatement of the gastrointestinal symptoms, biopsies remained lipid-laden with lipoprotein- and lipid-like particle densities exceeding those found in the normal fed subject. Three major conclusions derive from the ultrastructural analysis. The first is that there are chylomicron- and VLDL-sized particles (approximately 300 nm in diameter, range 169 to 580 nm) in membrane-bound compartments as has been described for several cases of chylomicron retention disease.14 20 Second, although rare, smaller particles (63 nm mean diameter) can be found in the intercellular spaces of affected subjects suggesting that secretion may occur. Third, there are other, larger intracellular lipid particles (from 368 to 2127 nm mean diameter) that do not appear to be located in a membrane-bound compartment. All the patients have these 2 types of particles.
The identity of the membrane-bound compartment that contains the
lipoprotein-sized particles is not entirely clear, nor is the
composition of the particles. Because the cells are so heavily
lipid-laden, the location of the vesicles is not useful for the
identification of the membrane-bound compartment. There is a remarkable
resemblance between some of the particle-bearing membrane bound
compartments observed in the present study and the images of
putative second-step triglyceride-rich particles found in
the endoplasmic reticulum of transgenic apo B deficient mice (see
Figures 9 to 11 in reference 40) in which assembly does not occur due
to the absence of apo B. The particles not situated in membrane
vesicles, also described in other cases of Anderson's disease and
chylomicron retention disease,14 16 18 are also
reminiscent of the lipid droplets found in the cytoplasm in transgenic
mice deficient in apo B synthesis (see Figure 10 in reference 40 and
Figure 4 in reference 41) or observed in cases of
abetalipoproteinemia,14 22 where assembly does not occur
due to the absence of MTP. Based on the ultrastructural observations in
transgenic mice deficient in apo B synthesis,40 41 the
lipid droplets in the cytoplasm of our patients could derive from the
breakdown of membrane-bound compartments that contain unassembled lipid
particles (putative second step triglyceride-rich
particles).
However, both the transgenic apo B deficient mouse model and patients with abetalipoproteinemia differ from patients with Anderson's disease. In Anderson's disease, as well as chylomicron retention disease, immunochemical analysis of enterocytes has shown the presence of increased amounts of apo B,12 14 and we have shown that MTP is present and active in Anderson's disease (reference 23 and this work). Further, the ultrastructural aspects of enterocytes in cases of abetalipoproteinemia, in which apo B is present intracellularly but MTP is inactive, and Anderson's disease are very different. Large nonmembrane-bound lipid droplets predominate in abetalipoproteinemia, whereas smaller membrane-bound particles predominate in Anderson's disease and chylomicron retention disease. The presence of apo B48 (this study and reference 15) and apo AIV (this study) in chylomicron fractions isolated from biopsies of patients with Anderson's disease or chylomicron retention disease suggest that assembly of some kind of lipid-rich lipoprotein particle can take place. Further, as opposed to 3 cases of abetalipoproteinemia with established mutations in MTP that we have studied (unpublished results), biopsies from patients with Anderson's disease secrete some apo B48 along with apo AIV into the culture medium in a chylomicron-density range. If transport to the Golgi apparatus is defective, assembled or partially assembled particles, as well as putative second step triglyceride-rich lipid particles, could accumulate in an endoplasmic reticulum compartment. Alternatively, transport of post-Golgi secretory vesicles could be defective.
Some studies of biopsies obtained from fasted patients have suggested that the lipoprotein-sized particles are present in both the endoplasmic reticulum and the Golgi apparatus.7 8 14 20 Other studies, however, suggest that the Golgi is normal but empty16 18 and that the particles are located in the endoplasmic reticulum only.18 After feeding, increased amounts of particles have been reported to be in the smooth endoplasmic reticulum,18 20 whereas the Golgi became narrow and empty.20 Strich et al18 have suggested that the particles in Anderson's disease are, in fact, in the endoplasmic reticulum and that they do not reach the Golgi, thus explaining the decreased glycosylation (decreased incorporation of mannose) that has been observed in chylomicron retention disease.15 Decreased incorporation of mannose has also been reported in vitamin A deficiency.42 Anderson's and chylomicron retention disease patients are frequently deficient in vitamin A; however, the role of vitamin A and glycosylation in Anderson's disease has not been studied.
In our cases, the Golgi apparatus is often distended, but it is generally empty. However, particles in membrane-bound compartments are in close proximity and sometimes appear to be part of the Golgi stack or to be budding off the lateral aspect of the Golgi. It is noteworthy that, in our patients, the particles in the membrane-bound compartments are larger and more numerous than the membrane-bound particles in normal fed subjects. Double immunogold labeling of apolipoproteins and cell compartment markers may allow the identification of the cell compartment (reticulum, Golgi, post-Golgi) to which the particle laden membrane-bound vesicles belong.43 A biochemical approach to determine whether the lipoprotein-like particles do indeed reach the Golgi apparatus in Anderson's disease patients would be to determine whether the apo B48 contains complex carbohydrates, the acquisition of which occurs in the Golgi apparatus. In preliminary results, we have found that treatment with endoglycosidase H and PNGase F of metabolically labeled apo B48 from Anderson's disease patients does indeed show that the protein has acquired complex carbohydrates. This acquisition of complex carbohydrates does not occur in brefeldin-treated biopsies nor in abetalipoproteinemic patients in which transport of apo B–containing lipoproteins to the Golgi is blocked or does not occur (unpublished results). Quantitation of the extent of complex glycosylation could help determine what proportion of apo B48 does indeed attain the Golgi apparatus.
For the linkage analysis, we were able to study 4 families. Highly polymorphic microsatellite markers, most frequently the (CA)n type, were used to establish the haplotypes in all the family members. These tandem repeats of a simple short nucleotidic sequence offer several advantages. They are abundant and uniformly distributed along the human genome.31 33 34 They exhibit, in general, a greater degree of polymorphism compared with the other genetic markers such as restriction fragment length polymorphisms. Finally, a large number of these markers have been situated with respect to one another and with respect to many genes with considerable accuracy within each chromosome.33 34
The choice of the genetic model and an appreciation of the genetic homogeneity of the cases are important considerations. The diagnostic criteria for Anderson's disease or chylomicron retention disease are quite specific and encompass biochemical, endoscopical, histological, ultrastructural, and clinical aspects as described in the Methods section. All the patients reported in the literature and all the patients studied in this report exhibited these characteristics at the time of diagnosis. The parents of all the patients studied here were asymptomatic.
Anderson's disease and chylomicron retention disease are clearly distinguishable from abetalipoproteinemia and homozygous hypobetalipoproteinemia (null alleles), particularly by the presence of apo B100–containing lipoproteins but also by the absence of acanthocytosis, retinitis pigmentosa, and severe neurological symptoms. Anderson's disease is also clearly distinguishable from heterozygous familial hypobetalipoproteinemia and homozygous familial hypobetalipoproteinemia (with truncated apo B) by the presence of diarrhea, malabsorption, and steatorrhea. Finally, the mode of inheritance (recessive versus dominant for familial hypobetalipoproteinemia) and the clear involvement of the apo B gene in familial hypobetalipoproteinemia further differentiate the disorders.
Thirty-five cases of Anderson's disease or chylomicron retention disease in 26 families have been reported in the literature.4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 Of these cases, 16, from ten different families,5 7 9 12 13 16 17 18 are of North African origin (Algeria, Morocco, Tunisia). Ten cases are from 9 Canadian families.6 11 14 15 The 9 remaining cases are from 6 different countries.4 8 10 19 20 21 Of the 26 families, consanguinity is found in 9, no consanguinity is present in 12, and information regarding consanguinity is not available in 5.
The autosomal recessive model used here for the genetic analysis of Anderson's (chylomicron retention) disease is based, therefore, on the following: there are 7 families with 2 or more cases, the parents are phenotypically normal (clinically asymptomatic with normal lipid values), 14 girls and 21 boys are affected. Of the 4 families used for the linkage analysis in this study, 3 have consanguinity. All 7 families that we studied here are of North African origin.
In addition to the absence of apo B48–containing lipoproteins, major biochemical hallmarks of the patients described with Anderson's disease or chylomicron retention disease include the low levels of plasma LDL and HDLs and decreased plasma levels of apo AI. The role of the various apolipoproteins in chylomicron assembly and secretion is still poorly understood in spite of considerable progress. Other than their participation as structural elements, apolipoproteins have not been attributed other functions important for chylomicron assembly and secretion. However, given the exceptional diversity in the physiological functions of some apolipoproteins, it would not be surprising that they play roles, either directly or indirectly, in the assembly and secretion processes.
Given the complete absence of apo B48–containing lipoproteins in the disease, the apo B gene has long been considered a candidate. Further, given the rarity of the disease and possible heterogeneity, it is important to evaluate all cases as to the involvement of the apo B gene. No evidence of linkage was found between the disease and the microsatellite repeat located in the apo B gene nor in the region of chromosome 2 surrounding the gene. Further, the intestinal organ culture shows the synthesis and secretion of a normal sized apo B48. These results effectively rule out the apo B gene locus as the origin of the disease in these 4 families. Pessah et al17 in 5 cases in 2 families, Strich et al18 in 3 cases in a single family, and Patel et al19 in 2 cases in a single family used variable number tandem repeat probes in the apo B gene to rule out its association with Anderson's disease in their cases. These results combined with the present results now exclude the apo B gene in 13 cases from 7 families with Anderson's disease.
Apo B48 is not, however, the only apolipoprotein expressed by the intestine. Analysis of mRNA synthesis in the human intestine has shown the presence of mRNA for apolipoproteins AI, AIV, E, CII, and CIII.44 45 46 47 In addition to apo B48, apo AI, apo AIV, apo CII, and apo CIII have been detected in human intestine by immunochemical techniques.12 14 43 48 49 50 51 Apo AI and apo AIV are major intestinal apoproteins and have been detected in increased amounts in enterocytes of patients with Anderson's disease12 or chylomicron retention disease,14 as have the C apoproteins.12 These proteins have been reported to be components of human chylomicrons.52 Further, decreased plasma levels of these proteins have been reported in patients.12 14 It has been suggested that apo E has a function in the hepatic VLDL assembly and secretion cascade.53 The role, if any, of apoproteins other than apo B, in human intestinal chylomicron and VLDL assembly is unknown. In fact, the majority of current knowledge concerning lipoprotein assembly has been derived from transformed hepatic cell lines. Intestinal chylomicron assembly and secretion may well differ, in certain respects, from hepatic triglyceride-rich lipoprotein assembly and secretion. The existence of Anderson's or chylomicron retention disease is, perhaps, one of the best indications of this difference. Thus, to clearly establish the relationship of these apoproteins to Anderson's disease, we evaluated their genes as well as large regions of the chromosomes surrounding the genes for linkage to the disease.
The results reported here exclude the genes of all of these proteins (apo AI, apo AIV, apo B, apo E, apo CI, apo CII, and apo CIII) as being linked to Anderson's disease. In addition, the LOD scores as a function of the recombination fractions effectively exclude any genes located in defined regions around these genes: approximately 18 cM around the apo B gene on chromosome 2; approximately 15 cM around the apo AI, apo CIII, apo AIV gene cluster on chromosome 11; and approximately 24 cM around the apo CII, apo E, and apo CI genes on chromosome 19.
Given that the major apolipoproteins produced in the intestine are effectively excluded as causes for Anderson's disease, we investigated intracellular lipid transport proteins that could play indispensable roles in lipoprotein assembly and secretion. MTP protein has recently been shown to be implicated in lipoprotein assembly, and mutations in the large subunit of this complex have been found to be the molecular basis for several cases of abetalipoproteinemia.2 In previous work, we showed that MTP protein and activity were present in 1 case (M.K.) of Anderson's disease23 from pedigree 200. Given the rareness of the disease and the important role of MTP in lipoprotein assembly, it is important to clearly establish the lack of association of the disease with the MTP complex in more than 1 subject. In this study we have tested 3 additional cases from 2 different families, pedigrees 100 and 300. MTP large subunit proteins were clearly present in the intestinal biopsies of all 3 subjects as well as MTP activity. It was necessary to demonstrate that the MTP protein was active because we have detected the presence of a normal-sized inactive MTP protein with a false sense mutation in a case of abetalipoproteinemia.54 The linkage analysis definitively ruled out the MTP gene as well as any other genes located in a 30 cM region around the gene.
Interestingly, the gene for intestinal fatty acid binding protein is located in this region. This protein, which is important for facilitating intracellular uptake and trafficking of fatty acids to the site of triacylglycerol biosynthesis, is also a potential candidate for the disease. Although 3 studies6 8 15 have suggested that intracellular triglycerides are formed and constitute the majority of the intracellular lipid, it is clear that alterations in the levels of fatty acid binding protein might affect lipoprotein biosynthesis.55 However, we find no evidence of linkage between the FABP2 gene and Anderson's disease. Another fatty acid binding protein that is also expressed in the intestine is hepatic fatty acid binding protein, FABP1. Use of polymorphic markers spanning the region where the gene is located revealed no evidence of linkage and the LOD scores effectively ruled out a 20-cM region of the chromosome surrounding the location of the gene. Taken together, these results suggest that fatty acid transport mediated by fatty acid binding proteins 1 and 2 should be normal in subjects with Anderson's disease.
In conclusion, the results of the present study clearly show that chylomicron-like particles can be observed intracellularly in membrane-bound vesicles in enterocytes in Anderson's disease as has been reported in chylomicron retention disease.14 We have further quantitated the variability in the ultrastructural aspects and defined the characteristics of the lipid and lipoprotein-like particles in our patients as well as showing that lipoprotein-like particles are also present, although rarely, in the intercellular spaces in biopsies. Small amounts of normal-sized apoproteins B and AIV are secreted into the medium in complexes having densities similar to chylomicrons suggesting that lipoprotein assembly may occur and that a small amount of secretion may take place in agreement with the ultrastructural observations and with some reports of other patients with Anderson's disease.4 18 However, increased amounts of normal sized apo B48 and apo AIV are present intracellularly. Finally, defects in the genes encoding the apolipoproteins expressed in the intestine (AI, AIV, B, CI, CII, CIII, and E) are excluded as causes of Anderson's disease as are mutations in the genes encoding 3 intracellular lipid transport proteins (MTP, FABP1, and FABP2). Exclusion of these genes clearly suggest the importance of another factor for human intestinal chylomicron secretion. Future studies would profit from an assessment of the basolateral secretion of proteins other than apolipoproteins to determine whether the defect in this disease specifically affects lipoproteins or affects basolateral enterocyte secretion in general. Finally, a genome-wide genetic linkage analysis of a larger group of families with affected subjects is justifiable given current technology.
|
Acknowledgments |
|---|
This study was supported financially by the Center National de la Recherche Scientifique (CNRS), by the Institut National de la Santé et de la Recherche Médicale (INSERM), by the Caisse Nationale de l'Assurance Maladie des Travailleurs Salariés (CNAMTS) grant number 4AIC01, and by the Societé Nationale Française de Gastroentérologie. We also thank M. Gegauff, Laboratoire d'Anatomie Pathologique, Hôpital Necker-Enfants Malades, Paris, for assistance.
|
Footnotes |
|---|
Centre de Génétique Moléculaire, Centre National de la Recherche Scientifique, Associé à l'Université Pierre et Marie Curie (A.H.D., L.P.A.), U327 Institut National de la Santé et de la Recherche Médicale, Faculté de Médecine Xavier Bichat, (A.H.D., N.B.-V., N.V., M.-E.S.-B.), and Departement de Pédiatrie (V.A., J.S.) and U393 Institut National de la Santé et de la Recherche Médicale (P.A.), Hôpital Necker-Enfants Malades, Paris, France; and Department of Metabolic Diseases, Bristol-Myers Squibb, Princeton, NJ (J.R.W.).
Received October 21, 1998; accepted February 6, 1999.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Kane JP, Havel RJ. Disorders of the biogenesis and secretion of lipoproteins containing the B apolipoproteins. In: Scriver CR, Beaudet AR, Sly WS, Valle D, Stanbury JB, Wyngaarden JB, Frederickson DS, eds. The Metabolic Basis of Inherited Disease. New York: McGraw-Hill;1989:1139–1164.
- Gordon DA. Recent advances in elucidating the role of the microsomal triglyceride transfer protein in apolipoprotein B lipoprotein assembly. Curr Opin Lipid. 1997;8:131–137.[Medline] [Order article via Infotrieve]
- Linton MF, Farese RV, Young SG. Familial hypobetalipoproteinemia. J Lipid Res. 1993;34:521–541.[Medline] [Order article via Infotrieve]
- Anderson C, Townley RRW, Freeman M, Johansen P. Unusual causes of steatorrhoea in infancy and childhood. Med J Aust. 1961;11:617–622.
- Lamy M, Frezal J, Polonovski J, Rey J, Jos J, Nézelof C, Herrault A, Cohen-Solal J. Diarrhée chronique par trouble du transport intra-cellulaire des lipides. Arch Fr Pédiatr. 1967;24:1079. Abstract.
- Silverberg M, Kessler J, Neumann PZ, Wiglesworth FW. An intestinal lipid transport defect. A possible variant of hypo-beta-lipoproteinemia. Gastroenterology. 1968;54:1271. Abstract.
- Polonovski C, Navarro J, Fontaine JL, de Gouyon F, Saudubray JM, Cathelineau L. Maladie d'Anderson. Ann Pediatr (Paris). 1970;17:342–354.[Medline] [Order article via Infotrieve]
- Partin JC, Schubert WK. Jejunal mucosal biopsy studies in two new types of hypo-beta lipoproteinemia: light, electron microscopical, and biochemical analysis. Gastroenterology. 1970;58:1022. Abstract.
- Costil J. Maladie d'Anderson. In: Journées Parisiennes de Pediatrie. Paris: Flammarion; 1976:229–239.
-
Scott BB, Miller JP, Losowsky MS.
Hypobetalipoproteinemia: a variant of
the Bassen-Kornzweig syndrome. Gut. 1979;20:163–168.
[Abstract/Free Full Text] - Gauthier S, Sniderman A. Action tremor as a manifestation of chylomicron retention disease. Ann Neurol. 1983;14:591.
- Bouma ME, Beucler I, Aggerbeck LP, Infante R, Schmitz J. Hypobetalipoproteinemia with accumulation of an apoprotein B-like protein in intestinal cells. Immunoenzymatic and biochemical characterization of seven cases of Anderson's disease. J Clin Invest. 1986;78:398–410.
- Polanco I, Mellado MJ, Lama R, Larr