|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Lipid Research, Vol. 45, 155-163, January 2004
Copyright © 2004 by American Society for Biochemistry and Molecular Biology

* Divisions of Atherosclerosis, Nutrition, and Lipid Research, Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110
Gastroenterology, Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110
Published, JLR Papers in Press, September 16, 2003. DOI 10.1194/jlr.M300275-JLR200
2 Permanent address of G. Li: Department of Cardiology, Second Affiliated Hospital, Chongqing University of Medical Sciences, Chongqing, China. ![]()
1 To whom correspondence should be addressed. e-mail: zchen{at}im.wustl.edu
| ABSTRACT |
|---|
|
|
|---|
The greater-than-expected decreases in apoB-100 production rates of FHBL heterozygous humans appear to be attributable to a defect in secretion rather than in the synthesis of apoB-100 from the unaffected apoB allele.
Supplementary key words VLDL secretion gene targeting animal model
| INTRODUCTION |
|---|
|
|
|---|
In recent years, we (13, 14) and other investigators (1517) have developed several lines of Apob-modified mice to model the FHBL condition. These mice provided new insights into the FHBL syndrome and the structure-function relationship of apoB-100 (1316). However, apoB-48, instead of apoB-100, is the major protein product of the normal mouse apoB allele as a result of the presence of apoB mRNA-editing machinery in the mouse liver (18, 19). Thus, the currently available FHBL mouse models are not suitable for studying the effect of apoB-truncation mutations on the synthesis and secretion of apoB-100 from the normal apoB allele in the liver.
In this study, we bred our previously generated apoB-38.9-only mice (Apob38.9/38.9) (13) with apoB mRNA-editing catalytic enzyme-1 (Apobec-1) knockout mice (Apobec-1-/-) (20) and with apoB-100-only mice (Apob100/100) (21) to generate two different lines of apoB-38.9 heterozygous mice in which only apoB-38.9 and apoB-100 are produced by the liver. These mice provided us an opportunity to determine directly the effects of the apoB-38.9 mutation on the synthesis and secretion of apoB-100 and to explore the underlying mechanisms at the cellular level.
| METHODS |
|---|
|
|
|---|
Mice
The generation of the apoB-38.9-producing mouse has been described (13). This mouse carries a single nucleotide deletion in its Apob, leading to the formation of a premature stop codon at residue 1,767. The apoB-100-only mouse was purchased from Jackson Laboratories (Bar Harbor, ME). This mouse was generated by Farese et al. (21). It carries a targeted missense mutation at codon 2,153 that prevents the formation of apoB-48 resulting from apoB mRNA editing. Both lines of mice have a mixed genetic background, with 50% C57BL/6 and 50% 129/SvJ (13, 21). The mice doubly heterozygous for the apoB-38.9- and apoB-100-specifying alleles (Apob100/38.9) were generated by crossbreeding of Apob38.9/38.9 and Apob100/100 mice. The Apob100/38.9 and Apob 100/100 mice obtained from offspring of the intercross of Apob100/38.9 mice were used in this study. All of these offspring had a mixed genetic background, with 50% C57BL/6 and 50% 129/SvJ. The Apob100/38.9 mice were also crossed with LDL-receptor knockout (Ldlr-/-) and apoE knockout (Apoe-/-) mice (Jackson Laboratories) to produce Apob100/100/Ldlr-/-, Apob100/38.9/Ldlr-/-, Apob100/100/Apoe-/-, and Apob100/38.9/Apoe-/- mice. The Ldlr-/- and Apoe-/- mice had a C57BL/6 genetic background; thus, the resulting Apob100/100/Ldlr-/-, Apob100/38.9/Ldlr-/-, Apob100/100/Apoe-/-, and Apob100/38.9/Apoe-/- mice were predicted to have a mixed genetic background, with 75% C57BL/6 and 25% 129/SvJ.
The Apobec-1-/- mouse was generated by Hirano et al. (20). It has a C57BL/6 genetic background. Thus, offspring (Apobec-1-/+/Apob38.9/+) produced from crossbreeding of Apob38.9/38.9 with Apobec-1-/- mice had a mixed genetic background, with 75% C57BL/6 and 25% 129/SvJ. Apobec-1-/-/Apob+/+ and Apobec-1-/-/Apob38.9/+ mice generated from intercrossing of Apobec-1-/+/Apob38.9/+ mice were used for this study.
Only offspring from intercross breeding were used in this study, and in most cases, littermates were used to minimize variation in the strain percentage.
All mice were weaned at 3 weeks of age, housed in a specific-pathogen-free barrier facility with a 12 h light/dark cycle, and fed a regular mouse chow diet (Ralston Purina, St. Louis, MO).
Fast-protein liquid chromatography fractionation and Western blot analysis of plasma apoB
For Western blot analysis, mouse plasma was subjected to electrophoresis on 312% gradient SDS-PAGE gels under reducing conditions and electrotransferred onto Immobilon-P (Millipore Corp., Bedford, MA). Western blot analyses were carried out using rabbit antisera raised against a GST fusion protein containing amino acids 26 to 289 of mouse apoB (13, 14) and an enhanced chemiluminescence (ECL) Western blot detection kit (Amersham Pharmacia Biotech, Inc.). The ECL signals were quantified by analyzing the density of the protein bands on Kodak X-Omat film using SigmaGel computer software (SPPS Science Corp., Chicago, IL).
Fast-protein liquid chromatography (FPLC) fractionation of plasma lipoproteins was carried out as described (13, 14) using 200 µl of plasma obtained from Apob100/100/Ldlr-/- or Apob100/38.9/Ldlr-/- mice. Cholesterol contents in each fraction were determined enzymatically, and apoB contents were determined by Western blot analysis.
In vivo study
To determine the in vivo secretion rates of apoB and triglycerides, mice (12 weeks old) were fasted for 4 h and administered Triton WR-1339 (Sigma) (500 mg/kg body weight) and [35S]Promix (300 µCi/g body weight) via intravenous injection. Blood samples (
80 µl each) were taken using heparinized capillary tubes at 0 h and at 0.5, 1, 2, and 3 h thereafter. VLDL was isolated from 20 µl of plasma in 200 µl of EDTA-saline-KBr (d = 1.009) using 250 µl Polyallomer Centrifuge tubes (Beckman) and a 42.2 Ti rotor (Beckman). Aliquots of VLDL and infranatant were electrophoresed on SDS-PAGE gels. 35S activity in apoB-100 of VLDL samples and albumin of infranatant were quantified on a GS-525 PhosphorImager system using a low ß-screen (Bio-Rad). The relative amount of [35S]apoB-100 in each plasma sample was expressed as the ratio of [35S]apoB-100 to [35S]albumin.
Triglyceride concentrations in the plasma samples were also determined. The values obtained at the 0 time point were subtracted from the values obtained at later times.
Hepatocyte culture study
Primary hepatocytes were isolated from mouse liver by perfusion of a collagenase solution via the portal vein (13, 14). Viability of the cell was
80% as determined by Trypan exclusion. Cells were plated onto six-well plates (0.6 x 106 cells/well) coated with poly-D-lysine (Sigma) and incubated at 37°C under 5% CO2 in 10% FBS/DMEM. After 1 h of attachment, cell monolayers were washed twice and incubated in 10% FBS/DMEM until use. All experiments involving cultured hepatocytes were commenced at 78 h after the cells were cultured. After this initial culture period, cells were washed three times with PBS and incubated in Met- and Cys-free DMEM for 30 min to deplete the cellular pool of Met and Cys. Thereafter, the medium was replaced with 1 ml of Met- and Cys-free DMEM containing 200 µCi of [35S]Promix with or without oleic acid (OA) (0.5 mM; to give anOA/BSA ratio of 3.6) (14), and cells were labeled for the specified time period to determine the rates of apoB secretion.
For pulse-chase experiments, the cells were pulsed with [35S]Promix for 45 min as described above. Thereafter, they were washed twice with PBS and incubated in 1 ml of DMEM containing 10 mM Met and 3 mM Cys for the specified time period with or without the specified additives.
Immunoprecipitation and quantification of 35S-labeled apoB
Immunoprecipitation of the labeled apoB in cell lysate and cultured media was carried out using rabbit anti-mouse apoB antisera as described (13, 14). Rabbit anti-human albumin antisera (Sigma) were also included in the immunoprecipitation mixture to immunoprecipitate mouse albumin simultaneously. The immunoprecipitated proteins (apoB and albumin) were resolved on a 312% gradient gel under reducing conditions. The gels were dried, and the radioactivity associated with apoB and albumin was quantified on a GS-525 PhosphorImager system using a low ß-screen (Bio-Rad). Our initial studies found no difference in the rates of synthesis and secretion of albumin between the Apobec-1-/-/Apob+/+ and Apobec-1-/-/Apob+/38.9 and between the Apob100/100 and Apob100/38.9 hepatocytes. Thus, the 35S-labeled albumin signal was used to correct the [35S]apoB-100 signal in each lane.
Miscellaneous procedures
Liver triglyceride contents were determined as described (13, 14). Cellular protein contents were determined using a modified Lowry method (22). Plasma triglyceride concentrations were determined using an enzymatic kit (WAKO Chemicals USA, Inc., Richmond, VA). Student's t-test and ANOVA analysis were performed to determine the significance of the differences.
| RESULTS |
|---|
|
|
|---|
|
To determine whether the low levels of plasma apoB-100 in Apobec-1-/-/Apob+/38.9 and Apob100/38.9 mice were caused by enhanced clearance or reduced secretion rates of apoB-100-containing lipoproteins, we crossed the Apob100/38.9 mice with Ldlr-/- and Apoe-/- mice, thereby blocking apoE-mediated plasma VLDL remnant and LDL receptor-mediated LDL catabolism, respectively. As shown in Fig. 2
, apoB-100 levels in Apob100/38.9/Apoe-/- mice amounted to only 1520% of those of Apob100/100/Apoe-/- mice (Fig. 2A), similar to the results comparing Apob100/38.9/Apoe+/+ to Apob100/100/Apoe+/+ mice. This suggests that apoE in the apoB-38.9 mice did not affect either the secretion of VLDL from the liver or its catabolism. Disruption of LDL receptor function yielded similar results (Fig. 2B). FPLC analysis showed that plasma LDL cholesterol concentrations in Apob100/38.9/Ldlr-/- mice were also
25% of those the Apob100/100/Ldlr-/- mice (Fig. 3)
. This suggested that LDL receptor-mediated uptake and catabolism of VLDL did not account for the apparent low rates of VLDL secretion in the apoB-38.9 mice. Together, these results provided physiological evidence indicating that low rates of hepatic apoB-100 VLDL secretion may be the major mechanism for the lower than expected levels of plasma apoB-100 in the Apobec-1-/-/Apob+/38.9 and Apob100/38.9 mice.
|
|
|
50% compared with those of Apobec-1-/-/Apob+/+ and Apob100/100 mice, respectively (Fig. 4B), which was similar to the level in humans (12). To assess the potential difference in the characteristics of apoB-100 VLDL particles secreted by Apob100/100 and Apob100/38.9 mice, VLDLs were isolated from the plasma of Apob100/100 and Apob100/38.9 mice at 2 h after injection with Triton WR-1339. Western blot analysis confirmed that VLDL from the Apob100/38.9 mice contained almost no apoB-38.9 (data not shown). Chemical analyses revealed that the VLDL produced by the Apob100/38.9 mice had a significantly higher triglyceride content than did the VLDL of Apob100/100 mice (20.1 ± 1.46 vs. 14.9 ± 1.27 mg/mg VLDL protein; P < 0.001, n = 3 mice). There was no significant difference in cholesterol contents in VLDL particles between the Apob100/38.9 and Apob100/100 mice (4.65 ± 0.42 vs. 4.10 ± 0.41 mg/mg VLDL protein; P > 0.05, n = 3 mice).
Synthesis and secretion of apoB-100 by cultured hepatocytes
To determine the mechanism responsible for the dramatic reductions in secretion rates of apoB-100 by livers of Apobec-1-/-/Apob+/38.9 and Apob100/38.9 mice, we determined the rates of intracellular accumulation and secretion of newly synthesized apoB-100 by cultured hepatocytes. On continuous metabolic labeling in the absence of OA, the rates of accumulation of [35S]apoB-100 in the cell and in the medium of Apob100/38.9 hepatocytes were approximately the expected 50% of those of the Apob100/100 hepatocytes (Fig. 5)
. However, supplementation of the labeling media with 0.5 mM OA, although significantly stimulating the secretion of apoB-100 by Apob100/100 hepatocytes, did not affect apoB-100 secretion by the Apob100/38.9 hepatocytes (Fig. 5). Thus, in the presence of OA, the rate of secretion of apoB-100 by Apob100/38.9 hepatocytes was only 25% of that of Apob100/100 hepatocytes (Fig. 5). The total amounts of [35S]apoB-100 accumulating in the cells and media at each time point were not affected by OA. Together, these data suggest that the secretion, but not the synthesis, of apoB-100 from the apoB-100-specifying allele may be impaired in the Apob100/38.9 hepatocytes.
|
Reduced secretion efficiency of apoB-100 in Apob100/38.9 hepatocytes
Pulse-chase experiments were carried out to examine the kinetics and efficiency of secretion of the newly synthesized apoB-100 in the Apob100/38.9 hepatocytes. In the absence of OA, there were no differences between Apob100/38.9 and Apob100/100 hepatocytes during the chase period in rates of secretion, intracellular decay, and total recovery of [35S]apoB-100 during the pulse period (Fig. 6)
. However, in the presence of OA, a significantly greater percentage of initially 35S-labeled apoB-100 synthesized during the pulse period was secreted into the media after a 2 h chase by Apob100/100 hepatocytes than by Apob100/38.9 hepatocytes (62 vs. 41%; P < 0.02, n = 3). Rates of intracellular decay of apoB-100 were similar in hepatocytes from the two genotypes (Fig. 6). Consequently, the total recovery of [35S]apoB-100 was significantly lower in the Apob100/38.9 than in the Apob100/100 hepatocytes (Fig. 6), suggesting greater proportional degradation of apoB-100 in the Apob100/38.9 hepatocytes.
|
|
|
| DISCUSSION |
|---|
|
|
|---|
Two lines of apoB-38 heterozygous mice that produced only apoB-38.9 and apoB-100 were created in this study. The first line is deficient in Apobec-1 and thus incapable of editing apoB RNA; the second line is doubly heterozygous for an engineered apoB-100-only allele and the apoB-38.9-specifying allele. The apoB-38.9 mutation caused a 70% reduction in plasma apoB-100 in these mice. Plasma apoB-100 levels in the Apob100/38.9/ApoE-null and Apob100/38.9/LDL-receptor-null mice were 8085% lower than those of their respective controls, i.e., Apob100/100/ApoE-null and Apob100/100/LDL-receptor-null mice, indicating that the low plasma levels of apoB-100 in the Apob100/38.9 mice result mainly from a lower than expected rate of apoB-100 secretion. Indeed, by using [35S]Met/Cys to label newly synthesized proteins and Triton WR-1339 to block the lipolysis/uptake of apoB-100 VLDL (2326), we demonstrated that apoB-100 secretion rates in both of the apoB-38.9 heterozygotes were reduced to 20% of those of their respective Apobec-1-/-/Apob+/+ or Apob100/100 littermates. To the best of our knowledge, the present study is the first that has used appropriate FHBL mouse models to confirm the previously reported lower than expected production rates of apoB-100 observed in FHBL heterozygous humans based on metabolic kinetic modeling (912).
Like apoB-100 secretion, hepatic triglyceride secretion rates were also reduced in these apoB-38.9 heterozygotes, but only to 50% of control levels. This may be attributable to the fact that the apoB-38.9 protein retains some capacity for transporting triglycerides, albeit in reduced amounts (13, 14). In addition, our data on the lipid composition of newly secreted apoB-100 VLDL particles in the Apob100/38.9 mice strongly suggested that increased amounts of triglycerides were packed into the apoB-100 VLDL particles. Similar to these mice, VLDL triglyceride secretion rates in the apoB-2 human heterozygotes were 40% of those of control subjects, albeit there was an 80% reduction in apoB-100 production rates (12). Thus, the apoB-100 produced by the unaffected apoB allele appears to recruit more triglycerides to compensate for the overall decreased triglyceride-exporting capacity of the liver. Nevertheless, this compensatory mechanism is apparently insufficient, because fatty livers developed in heterozygous FHBL humans (2730) as well as in our Apobec-1-/-/Apob+/38.9 and Apob100/38.9 mice.
A greater than expected decrease in the synthesis of apoB-100 from the apoB-100 allele could have contributed to the >50% reduction in apoB-100 secretion rates from our mice and FHBL heterozygous humans. However, we (13, 14) and others (15, 16) have previously shown that the presence of a truncated apoB allele in the mouse liver does not affect the steady-state levels of mRNA transcripts of the unaffected apoB allele. Moreover, in the present study, our continuous labeling experiments demonstrated that the rates of accumulation of 35S-labeled apoB-100 synthesized by the Apob100/38.9 or Apobec-1-/-/Apob+/38.9 hepatocytes were approximately the expected 50% compared with those in the hepatocytes derived from Apob100/100 or Apobec-1-/-/Apob+/+ mice, respectively. Similarly, in our pulse-chase experiments, the amounts of 35S-labeled apoB-100 synthesized by the Apob100/38.9 hepatocytes during the pulse period were
50% of the amounts synthesized by the Apob100/100 hepatocytes (data not shown). Together, these results suggest that the reduced secretion rates of apoB-100 from the unaffected apoB allele of FHBL heterozygous humans and mice probably did not result from a larger than expected decrease in apoB-100 synthetic rates. Rather, they were more likely attributable to impaired secretion of apoB-100 by the hepatocytes. This conclusion was supported by the results of our continuous labeling and pulse-chase experiments. In the presence of OA in the culture media, which is more likely to reflect the situation in vivo, apoB-100 secretion rates in Apob100/38.9 hepatocytes amounted to only 30% of those of the controls. Furthermore, the efficiency of secretion of newly synthesized apoB-100 was significantly lower in Apob100/38.9 hepatocytes than in Apob100/100 hepatocytes. In the absence of OA, the rates of secretion of apoB-100 in Apobec-1-/-/Apob+/38.9 or Apob100/38.9 hepatocytes were
50% of those of the hepatocytes from Apobec-1-/-/Apob+/+ or Apob100/100 mice. The addition of 0.5 mM OA in the culture media enhanced apoB-100 secretion in Apobec-1-/-/Apob+/+ or Apob100/100 hepatocytes but not in the Apobec-1-/-/Apob+/38.9 or Apob100/38.9 hepatocytes. It is not clear why the larger than expected differences in apoB-100 secretion between Apobec-1-/-/Apob+/38.9 or Apob100/38.9 hepatocytes and their corresponding controls did not emerge in the absence of exogenous OA. However, it is important to note that because of the presence of fatty liver in the apoB-38.9 mice, cellular triglyceride levels of the cultured Apobec-1-/-/Apob+/38.9 or Apob100/38.9 hepatocytes were 1.5- to 2.0-fold higher than those of the respective controls. Apparently, the status of lipid availability for VLDL formation may be very different between these cells. It is possible that even in the absence of OA in the culture media, the availability of apoB-100 was a limiting factor for VLDL production in the Apobec-1-/-/Apob+/38.9 and Apob100/38.9 hepatocytes; thus, VLDL secretion in these hepatocytes was no longer responsive to another increase in cellular triglyceride availability resulting from OA supplementation to the culture media. It is unlikely that the apparently reduced secretion of apoB-100 by the Apob100/38.9 or Apobec-1-/-/Apob+/38.9 hepatocytes was attributable to enhanced reuptake of the newly secreted apoB-100, as heparin and RAP did not diminish the difference in total recovery of newly synthesized apoB-100 between Apob100/38.9 and Apob100/100 hepatocytes.
The assembly and secretion of apoB-100 VLDL is a complex process during which a significant amount of apoB-100 is subjected to presecretory degradation (31, 32). The proteasomal pathway plays an important role in the intracellular degradation of apoB-100 in hepatoma cells (31, 32). In primary hepatocytes, nonproteasome-mediated degradation pathways, such as the post-endoplasmic reticulum protein degradation pathway (3234), may play a more important role in the presecretory degradation of apoB-100. A larger fraction of the apoB-100 was subjected to intracellular degradation in the Apob100/38.9 hepatocytes than in the Apob100/100 hepatocytes. It is not known whether this is the cause or the consequence of the reduced efficiency of apoB-100 secretion in these cells. The LDL receptor has been implicated in mediating the presecretory degradation of apoB (35, 36). However, our results from the Apob100/38.9/LDL-receptor-null mice indicated that even in the absence of the LDL receptor, apoB-100 secretion in the Apob100/38.9 mice was still much lower than 50% of that in the Apob100/100 mice, suggesting that the LDL receptor pathway is not responsible for the enhanced presecretory degradation of apoB-100 in our apoB-38.9 mice.
Another potential explanation for the reduced secretion efficiency of apoB-100 in apoB-38.9 heterozygous hepatocytes is that apoB-38.9 may directly interfere with the assembly and/or secretion of apoB-100 VLDL. However, numerous in vitro studies have shown that overexpression of apoB-100 or of truncated apoB variants in rat (3739) or human (40, 41) hepatoma cells does not affect the secretion rates of the endogenous apoB-100. Nor does overexpression of apoB-17 in liver using recombinant adenoviruses affect the levels of apoB-48 and apoB-100 in mouse plasma (41). Nevertheless, more in vivo studies are needed on the effects of additional apoB truncation-specifying mutations on the secretion of apoB-100.
| ACKNOWLEDGMENTS |
|---|
Manuscript received June 20, 2003 and in revised form August 28, 2003.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
Z. Chen, E. P. Newberry, J. Y. Norris, Y. Xie, J. Luo, S. M. Kennedy, and N. O. Davidson ApoB100 is required for increased VLDL-triglyceride secretion by microsomal triglyceride transfer protein in ob/ob mice J. Lipid Res., September 1, 2008; 49(9): 2013 - 2022. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Degrace, L. Demizieux, Z.-y. Du, J. Gresti, L. Caverot, L. Djaouti, T. Jourdan, B. Moindrot, J.-C. Guilland, J.-F. Hocquette, et al. Regulation of Lipid Flux between Liver and Adipose Tissue during Transient Hepatic Steatosis in Carnitine-depleted Rats J. Biol. Chem., July 20, 2007; 282(29): 20816 - 20826. [Abstract] [Full Text] [PDF] |
||||
![]() |
R. R. Sankatsing, S. W. Fouchier, S. de Haan, B. A. Hutten, E. de Groot, J. J.P. Kastelein, and E. S.G. Stroes Hepatic and Cardiovascular Consequences of Familial Hypobetalipoproteinemia Arterioscler. Thromb. Vasc. Biol., September 1, 2005; 25(9): 1979 - 1984. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. J. Whitfield, P. H. R. Barrett, F. M. van Bockxmeer, and J. R. Burnett Lipid Disorders and Mutations in the APOB Gene Clin. Chem., October 1, 2004; 50(10): 1725 - 1732. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. R. Garbow, X. Lin, N. Sakata, Z. Chen, D. Koh, and G. Schonfeld In vivo MRS measurement of liver lipid levels in mice J. Lipid Res., July 1, 2004; 45(7): 1364 - 1371. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||