Normal intestinal dietary fat and cholesterol absorption, intestinal apolipoprotein B (ApoB) mRNA levels, and ApoB-48 synthesis in a hypobetalipoproteinemic kindred without any ApoB truncation

Normal Intestinal Dietary Fat and Cholesterol Absorption, Intestinal Apolipoprotein B (ApoB) mRNA Levels, and ApoB-48 Synthesis in a Hypobetalipoproteinemic Kindred Without Any ApoB Truncation Judit I. Pulai, Maurizio Averna, Rai Ajit K. Srivastava, Mickey A. Latour, Ray E. Clouse, Richard E. Ostlund, and Gustav Schonfeld The purpose of this study was to characterize intestinal apolipoprotein B (apoB) metabolism in subjects with familial hypobetalipoproteinemia (FHBL), where segregation analysis supports linkage to the apoB gene but no apoB truncations are present. We investigated cholesterol and fat absorption, intestinal apoB mRNA synthesis and editing, as well as apoB-48 synthesis. Plasma triglycerides (TG) and retinyl palmitate in the chylomicron fractions were analyzed after 12 hours of fasting and then repeatedly for 14 hours after ingestion of a vitamin A-containing high-fat meal. Cholesterol absorption was assessed using a dual stable-isotope method. Mean peak times and concentrations and areas under the curve (AUCs) for fat absorption and mean percentages of cholesterol absorption were comparable in affected and nonaffected family members. Intestinal biopsies were extracted for total RNA and also incubated with 35S-methionine for measurements of apoB synthesis. Similar quantities of apoB mRNA were found to be expressed in the intestine in affected and control subjects by RNase protection assay. ApoB mRNA editing assay showed that the majority of apoB-100 mRNA was edited to the apoB-48 form to a similar extent in both groups. Virtually no apoB-100 protein was synthesized by the intestine in any subject, and apoB-48 protein synthesis was not significantly different in the affected individuals. These data are consistent with in vivo metabolism data that show normal production rates for liver-derived apoB-100 but increased apoB-100 fractional catabolic rates in affected members of this family. Thus, the molecular defect probably does not affect transcription, translation, or secretion of apoB-containing lipoproteins, but may instead affect their clearance.

Copyright © 1997 by W.B. Saunders Company HE PHENOTYPE OF low apolipoprotein B (apoB) and total and low-density lipoprotein (LDL) cholesterol may be operationally classified into three major lipoprotein disorders: (1) abetalipoproteinemia (ABL), which is due to deficiencies of microsomal triglyceride (TG) transfer protein, ~,2 resulting in defective lipoprotein assembly and low secretion rates of apoB-containing lipoproteins from both hepatocytes and enterocytes; (2) chylomicron retention disease (CRD), characterized by failure of chylomicron formation in enterocytes of unclear etiology3; and (3) familial hypobetalipoproteinemia (FHBL). The first two are inherited as autosomal recessive traits and manifest severe symptoms of fat malabsorption and its sequelae during infancy.4 FHBL is an autosomal codominant disorder that in turn may be subdivided into three genetic subtypes: kindreds in which the phenotype is due to ap0B truncationproducing mutations, kindreds in which the phenotype is genetically linked to the apoB locus but the specific molecular defect(s) is not known, and kindreds in which the phenotype is not linked to the apoB locus. 5-v FHBL homozygotes may have symptoms resembling ABL or CRD, but heterozygotes are usually asymptomatic. ApoB, a major protein of very-low-density lipoprotein (VLDL), LDL, and chy!omicrons , normally circulates in two forms in plasma. ApoB-100 is the full-length, 4,536-amino acid protein secreted from the liver on VLDL particles, and apoB-48, consisting of the N-terminal 2,153 amino acids of apoB-100, is secreted by enterocytes on chylomicrons. 8 ApoB-48 is the translation product of full-length apoB-100 mRNAs that are posttranscriptionally edited in enterocytes at nucleotide 6666 to form an in-frame stop codon. 9 Although several groups have reported on intestinal and hepatic apoB mRNA levels in ABL and CRD? °42 relatively little is known about the cellular mechanisms responsible for FHBL. We report on a kindred in which the FHBL phenotype is linked to the apoB gene locus but no apoB truncation has been detected at the protein level and no major genomic DNA

T

Metabolism, Vo146, No 9 (September), 1997:pp 1095-1100

rearrangements have been identified in the apoB gene. We examined intestinal cho!estcrol absorption and postprandial lipemia in three affected members of the D kindred, and compared the results with those of 16 normolipidemic "absorption controls." We also quantified apoB synthesis, apoB mRNA levels, and apoB mRNA editing in intestinal biopsy specimens in two affected and five normolipidemic "biopsy controls." SUBJECTS AND METHODS Subjects

The D kindred is a four-generation kindred that contains 49 individuals, 11 of whom are affected with FHBL as defined by apoB levels less than the fifth percentile for age, sex, and race 13 (range for affected, 17 to 37 mg/dL); their LDL cholesterol levels are 19 to 43 mg/dL. In the plasma of FHBL subjects, no truncated forms of apoB protein were detected by immunoblotting, and no major rearrangements of the apoB genomic DNA were visible on Southern blots (data not shown). Three affected and two normolipidemic family members and 14 healthy subjects volunteered for evaluations of fasting levels and postprandial lipid responses. Two affected family members and five other individuaIs underwent upper endoscopy and duodenal biopsies for evaluation of apoB mRNA levels and intestinal synthesis of apoB. None

From the Department of lnternal Medicine, Division of Atherosclerosis, Nutrition, and Lipid Research and Division of Gastroenterology, Washington University 4School of Medicine, St Louis, MO; and Institute di Medicina lnterna I, Universita ' di Palermo, Palermo, Italy. Submitted December 21, 1996; accepted January 30, 1997. Supported by National hzstitutes of Health Grant No. NIH HLRO142460, General Clinical Research Center Grant No. USPHS MO1RRO0036, and Diabetes Research and Training Center Grant No. NIH5P60DK20579. Address reprint requests to Gustav Schonfeld, MD, Washington University School of Medicine, Department of lnte/nal Medicine, 660 S Euclid Ave, Box 8046, St Louis, MO 63110. Copyright © 1997 by W.B. Saunders Company 0026-0495/97/4609-0022503.00/0

1095

1096

PULAI ET AL

of the subjects had endocrine, liver, or renal diseases, and none were taking medications. All subjects provided informed consent for a protocol approved by the Human Studies Committee of Washington University School of Medicine. Dietary habits of subjects participating in the fat tolerance tests were assessed by 3-day diaries, with the assistance of two experienced dieticians. Mean dietary intake was calculated by a computer program (Minnesota Nutrition Data System; University of Minnesota, Minneapolis, MN).

Intestinal ApoB-48 Synthesis and mRNA Levels Specimens were washed three times with DMEM, transferred to a 60 × 15-mm organ tissue culture dish (Becton Dickinson, Lincoln Park, NJ), and incubated with gentle agitation for 1 hour at 37°C in 300 ~aL freshly oxygenated DMEM containing 0.5 mCi [35S]methionine (SJ.204; Amersham, Arlington Heights, IL). At the end of the hour, the culture medium was removed and the specimens were washed three times with cold DMEM and homogenized with a Brinkmann Polytron apparatus (Westhury, NY) in 1 mL homogenization buffer. The homogenates were centrifuged at 40,000 rpm for 1 hour at 4°C in a Beckmann TL-100 ultracentrifuge (Beckmann Instruments, Palo Alto, CA). The resultant high-speed supernatant was removed, and apoB was immunoprecipitated with monospecific polyclonal antibody (R197-4). The immunoprecipitates were applied to 3% to 15% gradient sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) gels, and the gels were stained with Coomassie blue R250 treated with En3HANCE (NEN Research Products, Boston, MA) for 1 hour, followed by cold water precipitation for 30 minutes and 1% glycerol for 1 hour. The gels were dried and exposed to Kodak XAR-5 film (Eastman Kodak, Rochester, NY) at - 7 0 ° C for aut0radiography. The other five intestinal biopsy specimens (20 pg each) were homogenized in 1 mL RNAzol B Solution with a Brinkmann Polytron instrument, and total RNA was isolated. RNA samples were checked for integrity by electrophoresis. Fifteen micrograms of total RNA was dissolved in 25 gL sample buffer (50% formamide, 1 × MOPS buffer, and 6% formaldehyde). The mixture was incubated at 65°C for 10 minutes followed by incubation on ice for 5 minutes. Two microliters of RNA loading buffer was added to each RNA sample and e!ectrophoresed in 1.5% agarose gel that contained 6% formaldehyde. 28S and 18S bands were visualized by UV light. ApoB mRNA levels were determined by the RNase protection assay (Hyspeed RPA Kit; Ambion, Austin, TX). ApoB riboprobe was prepared on a linearized recombinant pGEM3Zf(+) plasmid containing 510 bp apoB cDNA (nucleotides 10201 to 10710) at PstI-EcoRI sites. Human [3-actin riboprobe was prepared using linearized recombinant plasmid obtained from Ambion. ApoB and [3-actin riboprobes were transcribed in vitro (Ambion MAXIscript Kit) according to the manufacturer's instructions using 32p-CTP (Amersham, Arlington Heights, IL). Hybridization was performed at 68°C for 10 minutes using 7.5 gg total RNA, followed by 30 minutes of ribonuclease A/T1 digestion at 37°C. The protected fragments were applied to a 6% polyacrylamide gel for 1 hour at 45 W. The

Vitamin A Fat-Tolerance and Cholesterol Absorption Tests After a 12-hour fast, subjects were given a fatty meal enriched with vitamin A 60,000 U/m 2 body surface area. After the meal, they ate no caloric foods for 14 hours but were allowed to drink water and/or nonalcoholic beverages. Blood samples were drawn before the meal and every hour until 6 hours and then every 2 hours until hour 14. Chylomicrons were separated by ultracentrifugation, and refinyl palmitare was assayed as previously described. 14 Results are presented as areas under the curve (AUCs) with baseline TG levels subtracted. 14 Cholesterol absorption was measured using stable nonradioactive tracers of cholesterol. 15 Briefly, [26,26,26,27,27,27-2H]cholesterol (30 mg) was administered orally and [23,24,25,26,27-13C]cholesterol (15 rag) was administered intravenously on day 0. On day 3, plasma samples were processed for chromatography-mass spectrometry by selected ion monitoring, and percent cholesterol absorption was calculated as the plasma ratio of 0ral to intravenous isotopic tracer.

Intestinal Biopsy Clinical indications for endoscopy for the biopsy controls were symptoms of acid reflux or histories compatible with peptic ulcer disease. None of the subjects had a history of malabsorption, inflammatory bowel disease, or bowel cancer. Endoscopies were performed in the outpatient videoendoscopy unit after overnight fasting. Endoscopic examinations of FHBL and control subjects were normal. Fifteen biopsy specimens were obtained from the duodenum by a gastroenterologist using standard biopsy forceps. Five fragments sent for routine anatomic diagnostic studies were normal in all these subjects. Five other pieces were immediately plunged into liquid nitrogen and approximately 30 minutes later into RNAzol B solution (Tel-Test, Friendswood, TX) for RNA extraction, and the last five pieces were collected in 1 mL methionine-free Dulbecco's modified Eagle's medium (DMEM) (D0422; Sigma Chemical, St Louis, MO) for apoB synthesis studies.

Table 1. Clinical Characteristics and Mean Lipid and Plasma Lipoprotein Levels in Subjects with FHBL and in Normal Controls Subject No.

Gender

Age

BMI

Cholesterol

TG

LDL

HDL

FHBL non-apoB truncation (D kindred) 111/10 111/12" 111/16 111/20"

F F F M

44 41 39 28

21.6 22.4 25.4 28.7

77 92 82 72

47 18 80 46

24 30 20 31

46 64 46 39

32 18 29 19

161 136 142 108

FHBL nonaffected (D kindred) II1~ 111/14

F F

50 40

28 23.7

183 151

89 121

108 81

57 52

91 69

158 200

Control(biopsy) 1" 2* 3* 4* 5*

F M M M M

53 71 65 52 69

21.7 27.7 32.7 20.9 28.0

210 153 215 93 157

79 121 177 141 50

135 101 146 46 100

50 28 34 19 47

101 95 126 70 75

131 98 123 69 121

NOTE. BMI is expressed in kg/m2; TG, cholesterol, LDL, HDL, apoAI, and apoB are expressed in rng/dL. *Underwent intestinal biopsy.

ApoB

ApoAI

1097

INTESTINAL apoB METABOLISM

Table 2. Mean Postprandial Lipemic Responses in FHBL Subjects and Normal Individuals Chylomicmns (retinyl palmitate) Group FHBL non-apoB truncation (D kindred) (n = 3) Control (absorption) FHBL nonaffected (D kindred) (n - 2) Normolipidemic controls (n = 14)* FHBL apoB truncation B > 48 (n = 8)* B < 48 (n - 6) SEM P

AUC (pg/dL • 14h)

Peak(mg/dL)

TGs

PeakTime(h)

AUC (mg/dL. 14h)

Peak(mg/dL)

Peak'Time(h)

10,357

1,933

8.3

1,419

221

7.2

9,233 9,185

1,617 1,328

6.2 7.2

609 1,536

175 308

6 6

13,137 12,447 998

1,568 1,591 126

8.6 7.6 0.31 .31

1,433 1,224 278

273 229 22.9 .58

5.5 5.6 0.2 .33

.57

.78

.15

NOTE. SEM is based on a pooled estimate of variance. *Previously published? 4

gel was dried and then exposed to x-ray film. AUCs were quantified using ImagePro Plus software (Jandel Scientific, San Rafael, CA).

ApoB mRNA Editing Assay The relative amounts of apoB-100 and apoB-48 mRNA were determined by primer extension analysisJ 6 Five micrograms of total RNA was reverse-transcribed with 50 U MuLV reverse transcriptase with primer PCR-B216 at 42°C for 30 minutes using a RNA-PCR kit (Perkin-Elmer, Norwalk, CT) following the manufacturer's instructions. After denaturation of the cDNA at 99°C for 5 minutes and then cooling at 4°C for 2 minutes, 2.5 U Taq DNA polymerase was added and 30 PCR cycles were performed at 92°C for 30 seconds and at 58°C

500 bpD,,400 bplm,-

for 1 minute in an automated temperature cycler (DNA engine; MJ Research, Watertown, MA). The amplified apoB cDNA product was purified with a Wizard PCR purification kit (Promega, Madison, WI). Two hundred nanograms of cDNA was mixed with 20 fmol ['yJ2P]ATP end-labeled synthetic 25mer (ATCATAATTATCTCTAATATACTGA), heated at 95°C for 5 minutes, and annealed at 70°C for 15 minutes in 250 mmol/L Tris hydrochlofide, 40 mmol/L MgC12, 150 mmol/L KC1, 5 mmol/L DTT (pH 8.5), and 500 #mol/L each of dATR dCTR dTTR and dGTR Then, 5 U AMV reverse transcriptase (Boehringer Mannheim, Indianapolis, IN) was added, and the mixture was incubated at 42°C for 1 hour. The samples were heated at 80°C for 3 minutes and electrophoresed on a 6% sequencing gel. Bands specifying apoB-100 and apoB-48

"~ ApoB

300 bpm,~

200 bp= ',~

"~ 13-actin

100 bp=,Fig 1. RNase protection assay fo r apoB and I~-actin in human intestinal biopsy samples. Lane 1, RNA size markers synthesized in vitro (Ambion); lanes 2 and 3, digested and undigested 13-actin riboprobes; lanes 4 and 5, digested and undigested apoB riboprobes; lane 6, unloaded; lanes 7 and 8, protected RNA fragments from FHBL affected members no. 111/20 and 111/12; lanes 9 to 13, protected RNA fragments from control individuals (no. 1 to 5); lane 14, protected fragments from hepatoma cell line. The protected apoB fragment is 510 bp and the protected 13-actin fragment 125 bp.

1098

PULAI ET AL

Table 3. Mean Intestinal ApoB mRNA Levels and Protein Synthesis in FHBL and Control Subjects Group Parameter FHBL non-apoB truncation (D kindred) (n = 2) Biopsy controls (n - 5) SEM P

ApoB mRNA

1,328 1,097 255

[3-Actin mRNA

3,857 2,400 460 .72

.16

ApoB/#-Actin

0.28 0.60 0.09 .13

ApoB Protein

421 447 143 .94

NOTE. Data represent the mean of arbitrary area units and area ratios of autoradiograms. SEM is based on a pooled estimate of variance.

were identified by autoradiography. The intensity of the bands was scanned by densitometry (ImagePro Plus; Jandel Scientific, San Rafael, CA). Synthetic oligonucleotide templates with edited (TAA) or unedited (CAA) sequence (35 mer) were used in the polymerase chain reaction (PCR) to produce standards for apoB-48 or apoB-100.

Statistical Analyses Parameters evaluated between affected subjects and those reported previously by this laboratory were analyzed using ANOVA. The Hartley F max test was used to determine differences in variation observed between data sets (ie, present experiment v those previously reported 14) for each variable. When significant differences were found, means were partitioned by Fisher's protected least-significant difference test. All data were analyzed using the SAS general linear models procedure. 17 Statements about significance are based on P less than .05 unless otherwise noted. RESULTS

Clinical characteristics of the FHBL subjects and the control groups are shown in Table 1. Dietary intakes were alike among the family members (data not shown). There were no significant

differences in body mass index (BMI) between affected and nonaffected members (24.5 v 26.1, P = .53), By definition, affected members had lower fasting total and LDL cholesterol and apoB levels than normolipidemic control (81 v 166, 26 v 102, and 25 v 90 mg/dL, respectively, P < .003). TG levels in affected FHBL versus control subjects were lower (48 v 111 mg/dL, P = .02), but HDL and apoAI levels were not different (49 v 41 mg/dL, P = .37 and 137 v 129 mg/dL, P = .73, respectively).

Impact of FHBL on CholesterolAbsorption and Postprandial Lipemia The cholesterol absorption test was performed on four affected family members, two nonaffected family members, and five healthy nonrelated individuals. Percent cholesterol absorption was comparable in the affected and control groups (47.5 _+ 6.3 v53.8 + 11.7). Postprandial lipemia was studied in three affected (no. III/!0, III/16, and III/20) and two nonaffected (no. III/14 and II/3) family members. In addition, the data were compared with previously published data on 14 FHBL heterozygotes harboring truncations ranging from apoB-31 to apoB-89 (six subjects with apoB < 48 and eight with apoB > 48) and 14 normolipidemic healthy individuals. 14 There were no significant differences between subjects with respect to mean AUCs, mean height of the peaks, or peak times for either chylomicron-retinyl palmitate or TG (Table 2). We evaluated whether the low-LDL phenotype was due to altered intestinal mRNA levels using the RNase protection assay. [3-Actin mRNA served as the internal control. The mean apoB/[3-actin ratio did not differ significantly between the two groups (Fig 1 and Table 3). Mean apoB mRNA editing activity oJ

<: < < < (O ~:

(Q_ -9o ¢ ~ -r-

cv -~'-

controls

ApoB-48

ApoB-100~

Probe Fig 2. ApoB mRNA editing in human intestinal biopsy samples. For each sample, 2 I~g total intestinal RNA was reverse-transcribed using a primer 3' to the edited site and amplified using oligonucleotides flanking the edited site. The amplified an d purified cDNA was used for primer extension analysis. TAA and CAA, the edited and unedited synthetic oligonucleotides, were used as controls. HepG2, the human hepatoma cell line, used as a control; 111/20and 111/12, FHBL affected individuals; controls, biopsy controls no. 2, 3, 4, and 5, respectively.

INTESTINAL apoB METABOLISM

1099

Table 4. Intestinal apoB mRNA Editing in FHBL and Control Subjects Group/ Parameter

ApoB-100

ApoB-48

Ape B-48/ B-100

FHBL non-apoB truncation (D kindred) (n - 2) Biopsy controls (n = 4) SEM P HepG2 cells

640 574 25.5 .75 1,712

1,580 2,257 274.1 .42 124

2.48 3.90 0.45 .18 0.07

NOTE. Data represent the mean of arbitrary area units and area ratios of autoradiograms. P values are for the affected v normal subject comparison. SEM is based on a pooled estimate of variance. HepG2 ceils were used as negative controls.

also was not altered (apoB-48/apoB-100 area ratio, 2.48 in FHBL subjects v 3.9 in controls). Liver (HepG2) RNA was used as a negative control, and approximately 7% of editing activity was detected (Fig 2 and Table 4). To evaluate apoB synthesis, equal amounts of total trichloroacetic acid precipitable proteins (--100,000 cpm) were immunoprecipitated from each sample with the anti-apoB polyclonal antibody and loaded onto 3% to 15% SDS-PAGE gel (Fig 3). Biopsy fragments produced almost exclusively apoB-48 (with < 1% apoB- 100), confirming that the intestinal apoB editing process was intact in these FHBL individuals. The 35S-methionine-labeled apoB-48 immunoprecipitated from high-speed supernatants appeared as double bands, similar to previous findings. !8,19 ApoB synthesis was similar in both normal and FHBL subjects (Table 3). Immunoprecipitation was repeated three times, and the results were well reproducible. DISCUSSION

To understand the mechanisms for low cholesterol and apoB levels in the affected members of the D kindred, we evaluated intestinal cholesterol absorption, chylomicron metabolism, intes-

final apoB mRNA levels, and apoB synthesis rates. The D kindred harbored no apoB truncation-producing mutation. Segregation analysis showed linkage to the apoB gene; however, no major genomic DNA rearrangements were found in either of two alleles by Southern blotting (Schonfeld G, unpublished observations, 1997). The normal development and health status of the affected subjects implied that no severe malabsorption was present. Nevertheless, we investigated the dietary fat and cholesterol absorption to access its possible contribution to the FHBL phenotype. Retinyl palmitate has been used as a marker for intestinally derived chylomicron metabolism.2°,2~ We observed comparable chylomicron retinyl palmitate and TG responses after the vitamin A oral fat loading test in controls and FHBL subjects. No differences were found in percent cholesterol absorption. These data suggest that the low-LDL phenotype was not due to altered absorption of intestinal fat and cholesterol, although individual differences were large and the number of subjects was small, making it difficult to detect subtle differences between groups. We further evaluated the intracellular mechanisms that might be responsible for FHBL in this family. At present, normal variations in apoB mRNA abundance in human intestine and liver are not well established. In ABL, both increased and reduced apoB mRNA levels have been reported in intestine, and increased apoB mRNA levels were reported in liver. 12,22 In homozygous FHBL, reduced apoB message levels were observed in the liver,23 but variations in intestinal apoB mRNA levels were not explored. Intestinal apoB mRNA levels have not been described in CRD or in heterozygous or non-apoBtruncation FHBL. Our non-apoB truncation-bearing putative heterozygotes exhibited intestinal apoB mRNA quantities that overlapped with normal levels. These data are compatible with the absence of gross rearrangements of genomic DNA where mRNA levels are expected to be very low to absent. 24 We

O

ApoB-48

Fig 3. ApoB synthesis in intestinal biopsy specimens. Intestinal biopsy samples were incubated with [aSS]methionine at 37°C for I hour. ApoB was immunoprecipitated from cell homogenates and applied to a 3% to 15% S!DS-PAGE gel in 2 affected family members (111/20 and III/ 12) and 5 control subjects. ApoB-48 was present as a doublet.

66 kD 46 kD

--II

[

O4

m

m

=

=

1

2

3

4

5

1100

PULAI ET AL

examined the efficiency of the apoB mRNA editing machinery 25,26 and documented similar editing rates in FHBL subjects and controls. Finally, we found normal rates of apoB-48 protein synthesis that were compatible with the normal apoB m R N A levels and apoB-48 editing activities. 27,2s FHBL is a heterogeneous disease at the molecular level; most of the well-characterized cases are due to genetic mutations leading to premature terminations of translation, in the D kindred, the molecular defect is unclear. Our results indicate that the FHBL phenotype in the D kindred is not due to promoter or enhancer defects in the apoB gene. Furthermore, apoB-!00 m R N A undergoes posttranscriptional modification, editing, polyadenylation, and splicing, and is capable of supporting normal synthetic rates of apoB-48 protein. We did not explore intracellular apoB degradation and secretion; however, our in vivo metabolic studies indicate that VLDL, IDL, and LDL apoB-100 production rates are comparable to controls and higher than truncation-producing heterozygotes. 29 The normal production rates imply that the apoB-100-producing machinery in the hepatocytes is normal and that no gross intracellular

packaging and secretion defects are present in the liver as they presumably appear to be in the intestinal cells. The size and density distributions and lipid composition of circulating apoB100-containing lipoproteins are also normal, but the fractional catabolic rates are about two to three times higher, 29 suggesting that enhanced clearance is the predominant mechanism for the low-LDL phenotype in this kindred. The remaining possibilities are that a structural variant of apoB is produced, resulting in high fractional catabolic rates, or that a set of particles with extremely rapid turnover are assembled that are not detected in our metabolic studies. Finally, the molecular defect responsible might be confined to the liver, which none of our studies have discovered to date.

ACKNOWLEDGMENT

The authors thank the members of the D kindred for their cooperation, and Diana Tessereau for assistance in obtaining consent for the intestinal biopsies and the blood samples. We are grateful to Mary Lou Rheinheimer for preparation of the manuscript.

REFERENCES

1. Sharp D, Blinderman L, Combs KA, et al: Cloning and gene defects in microsomal triglyceride transfer protein associated with abetalipoproteinaemia. Nature 365:65-69, 1993 2. Wetterau JR, Aggerbeck LE Bouma ME, et al: Absence of microsomal triglyceride transfer protein in individuals with abetalipoproteinemia. Science 258:999-1001, 1992 3. Bouma ME, Beucler I, Aggerbeck IR et al: Hypobetalipoproteinemia with accumulation of an apoprotein B-like protein in intestinal cells. J Clin Invest 78:398-410, 1986 4. Pessah M, Benlian P, Beucler I, et al: Anderson's disease-Genetic exclusion of the apolipoprotein-B gene in two families. J Clin Invest 87:367-370, 1991 5. Schonfeld G: The hypobetalipoproteinemias. Annu Rev Nutr 15:23-34, 1995 6. Linton ME Farese RV Jr, Young SG: Familial hypobetalipoproteinemia. J Lipid Res 34:521-541, 1993 7. Fazio S, Sidoli A, Vivenzio A, et al: A form of familial hypobetalipoproteinaemia is not due to a mutation in the apolipoprotein-B gene. J Intern Med 229:41-47, 1991 8. Young SG: Recent progress in understanding apolipoprotein-B. Circulation 82:1574-1594, 1990 9. Powell LM, Wallis SC, Pease RJ, et al: A novel form of tissue-specific RNA processing produces apolipoprotein B-48 in intestine. Cell 50:831-840, 1987 10. Levy E, Marcel yR, Deckelbaum J, et al: Intestinal apoB synthesis, lipids, and lipoproteins in chylomicron retention disease. J Lipid Res 28:1263-1274, 1987 11. Roy CC, Levy E, Green PHR, et al: Malabsorption, hypocholesterolemia, and fat-filled enterocytes with increased intestinal apoprotein B: Chylomicron retention disease. Gastroenterology 92:390-399, 1987 12. Glickman RM, Glickman JN, Magun A, et al: Apolipoprotein synthesis in normal and abetalipoproteinemic intestinal mucosa. Gastroenterology 101:749-755, 1991 13. Lipid Research Clinics: Population Studies Data Book: The Prevalence Study. US Department of Health and Human Services Publication No. (NIH) 80-1527, 1980 14. Averna M, Seip RL, Mankowitz K, et al: Postprandial lipemia in subjects with hypobetalipoproteinemia and a single intestinal allele for apoB-48. J Lipid Res 34:1957-1967, 1993 15. Bosner MS, Ostlund RE, Osofisan O, et al: Assessment of percent cholesterol absorption in humans with stable isotopes. J Lipid Res 34:1047-1053, 1993

16. Driscoll DM, Wynne JK, Wallis SC, et al: An in vitro system for the editing of apolipoprotein B mRNA. Cell 58:519-525, 1989 17. SAS Institute: SAS/STAT User's Guide: Statistics Version 6.10. Cary, NC, SAS Institute, 1995 18. Krul ES, Tang JJ, Kettler TS, eta!: Lengths of truncated forms of apolipoprotein-B (ApoB) determine their intestinal production. Biochem Biophys Res Commun 189:1069-1076, 1992 19. sasak VW, Buller HA, Reinhold R: Biosynthetic precursor (214 kDa) of apolipoprotein B-48 is not secreted by Caco-2 cells and normal human intestine. Biochem J 264:365-370, 1989 20. Cohu JS, Johnson EJ, Millar JS, et al: Contribution of ApoB-48 and ApoB-100 triglyceride-rich lipoproteins (TRL) to postprandial increases in the plasma concentration of TRL triglycerides and retinyl esters. J Lipi d Res 34:2033-2040, 1993 21. Cohn JS, McNamara JR, Krasinski SD, et at: Role of triglyceride rich lipoproteins from the liver and intestine in the etiology of postprandial peaks in plasma triglyceride concentration. Metabolism 38:484-490, 1989 22. Lackner KJ, Monge JC, Gregg RE, et al: Analysis of the apolipoprotein B gene and messenger ribonucleic acid in abetalipoproteinemia. J Clin Invest 78:1707-1712, 1986 23. Ross RS, Gregg RE, Law SW, et al: Homozygous hypobetalipoproteinemia: A disease distinct from abetalipoproteinemia at the molecular level. J Clin Invest 8:590-595, 1987 24. Cooper DN: Human gene mutations affecting RNA processing and translation. Ann Med 25:11-17, 1993 25. Patel S, Pessah M, Beucler I, et al: Chylomicron retention disease: Exclusion of apolipoprotein B gene defects and detection of mRNA editing in an affected family. Atherosclerosis 108:201-207, 1994 26. Lopezmiranda J, Kam N, Osada J, et al: Effect of fat feeding on human intestinal apolipoproteiu B mRNA levels and editing. Biochim Biophys Acta 1214:143-147, 1994 27. Davidson NO, Magun AM, Brasitus TA, et al: Intestinal apolipoprotein A-1 and B-48 metabolism: Effects of sustained alterations in dietary triglyceride and mucosal cholesterol flux. J Lipid Res 28:388402, 1987 28. Black DD, Hay RV, Rohwernutter PL, et al: Intestinal and hepatic apolipoprotein-B gene expression in abetalipoproteinemia. Gastroenterology 101:520-522, 1991 29. Latour MA, Patterson BW, Pulal J, et al: The metabolism of apolipoprotein B-100 in a hypobetalipoproteinemic kindred without a truncated form of apoB. J Lipid Res 38:592-599, 1997