(Received for publication, February 24, 1997, and in revised form, May 27, 1997)
From the Mammalian Cell and Molecular Biology Laboratory, Department of Biology and Molecular Biology Institute, San Diego State University, San Diego, California 92182-0057
Stable plasmid-driven expression of the
liver-specific gene product cholesterol 7-hydroxylase
(7
-hydroxylase) was used to alter the cellular content of
transcriptionally active sterol response element binding protein 1 (SREBP1). As a result of stable expression of 7
-hydroxylase,
individual single cell clones expressed varying amounts of mature
SREBP1 protein. These single cell clones provided an opportunity to
identify SREBP1-regulated genes that may influence the assembly and
secretion of apoB-containing lipoproteins. Our results show that in
McArdle rat hepatoma cells, which normally do not express
7
-hydroxylase, plasmid-driven expression of 7
-hydroxylase results
in the following: 1) a linear relationship between (i) the cellular
content of mature SREBP1 and 7
-hydroxylase protein, (ii) the
relative expression of 7
-hydroxylase mRNA and the mRNA's encoding the enzymes regulating fatty acid, i.e. acetyl-CoA
carboxylase and sterol synthesis, i.e. HMG-CoA reductase,
(iii) the relative expression of 7
-hydroxylase mRNA and
microsomal triglyceride transfer protein mRNA, a gene product that
is essential for the assembly and secretion of apoB-containing
lipoproteins; 2) increased synthesis of all lipoprotein lipids
(cholesterol, cholesterol esters, triglycerides, and phospholipids);
and 3) increased secretion of apoB100 without any change in apoB
mRNA. Cells expressing 7
-hydroxylase contained significantly
less cholesterol (both free and esterified). The increased cellular
content of mature SREBP1 and increased secretion of apoB100 were
concomitantly reversed by 25-hydroxycholesterol, suggesting that the
content of mature SREBP1, known to be decreased by
25-hydroxycholesterol, mediates the changes in the lipoprotein assembly
and secretion pathway that are caused by 7
-hydroxylase. These data
suggest that several steps in the assembly and secretion of
apoB-containing lipoproteins by McArdle hepatoma cells may be
coordinately linked through the cellular content of mature SREBP1.
Apolipoprotein B100 (apoB)1 is an unusually large (>500 kDa) amphipathic protein responsible for the assembly and secretion of plasma lipoproteins by the liver and intestine (reviewed in Refs. 1-3). Its concentration in plasma, as a component of LDL, is a major determinant of susceptibility to the development of atherosclerotic cardiovascular disease (4, 5). Hepatic derived apoB100-containing lipoproteins are the precursors of plasma LDL (6). Hepatic assembly and secretion of apoB-containing lipoproteins require an orchestration of many seemingly independent processes as follows: 1) the production of component lipids (cholesterol, cholesterol esters, triglycerides, and phospholipids); 2) the synthesis of apoB, a uniquely large polypeptide containing multiple amphipathic structural domains that irreversibly associate with phospholipids (7); 3) translocation across the endoplasmic reticulum that requires an intralumenal protein complex consisting of MTP and PDI (8); and 4) the assembly of VLDL particle within the endoplasmic reticulum (1, 2, 9).
Previous studies showed a coordinate induction of the synthesis of all
VLDL lipids and the assembly and secretion of apoB-containing lipoproteins in response to changes in metabolic state (e.g.
during dietary carbohydrate overload (10) and fetal development (11)). In contrast, fasting caused a coordinate repression of all of these
processes (12, 13). These findings led us to propose that in a
coordinate manner, metabolic signals control multiple steps of the VLDL
assembly/secretion pathway. Using a somatic cell genetic approach, we
examined this possibility. Previous experiments showed that expression
of the liver-specific gene 7-hydroxylase (EC 1.14.13.17) in
non-hepatic Chinese hamster ovary cells increased the transcription of
the LDL receptor gene (14), a process now known to be regulated by the
SREBP family of transcription factors (15, 16). Sterol-responsive
transcriptional regulation is invoked by a proteolytic cleavage of the
initial translational SREBP gene product that resides in the
endoplasmic reticulum membrane as a trans-membrane loop (17, 18). When the sterol content of cells is insufficient to meet metabolic demands,
proteolysis of the N-terminal domain releases the DNA binding domain of
SREBP(s) (mature forms), allowing them to enter the nucleus and
activate transcription of genes having cognitive promoter elements.
Expression of 7
-hydroxylase in cells that do not normally express it
increases the cellular content of mature SREBP1 (see "Results").
Moreover, since McArdle rat hepatoma cells express all of the genes
required to assemble and secrete apoB-containing lipoproteins (19-21),
but lack 7
-hydroxylase (see "Results"), by stably expressing
7
-hydroxylase via plasmid transfection in McArdle hepatoma cells, we
were able to increase the cellular content of mature SREBP1 to
different extents in single cell clones. These cells provide a unique
opportunity to examine how SREBP1 may affect each of the steps of the
VLDL assembly/secretion pathway. The results indicate that SREBP1
coordinately regulates many of the individual lipogenic and protein
processing steps.
All chemicals used for biochemical techniques were purchased
from Sigma, VWR, Fisher, or Boehringer Mannheim. Cell culture medium
was obtained from Life Technologies, Inc. and fetal bovine serum from
Gemni. Restriction enzymes were purchased from New England Biolabs or
Promega. The cDNA probes were obtained from the following: hamster
MTP was a generous gift from Drs. Richard Gregg, John Wetterau, David
Gordon, and their colleagues at Bristol-Myers Squibb (22); hamster
HMG-CoA reductase was obtained from ATCC; rat 7-hydroxylase was a
generous gift from Dr. David Russell (23); and acetyl-CoA carboxylase
was a generous gift from Dr. Tim Osborne (24). A mouse hybridoma cell
line producing monoclonal antibody against human SREBP-1 (a generous
gift from Dr. Tim Osborne) was propagated in mice, and the resulting
IgG from the ascites fluid was obtained (25). A rabbit polyclonal
antibody produced from a synthetic SREBP2 peptide was a generous gift
from Dr. Michael Briggs.
McArdle RH-7777 hepatoma
cells, a gift from Dr. Tom Innerarity, were cultured in Dulbecco's
modified Eagle's medium (DMEM) containing 25 mM glucose,
15% heat-inactivated fetal bovine serum, and antibiotics (penicillin,
100 units/ml; streptomycin, 100 units/ml; and fungizone, 500 µg/ml).
The cells were grown at 37 °C in an atmosphere of air with 5%
CO2. A plasmid was prepared from pcDNA3 (Invitrogen) in
which the coding region of rat 7-hydroxylase was placed in the
EcoRI-digested linker region. This plasmid contains a
cytomegalovirus promoter and encodes neomycin resistance. Stable expression of the plasmid in McArdle RH-7777 cells was achieved by
calcium phosphate precipitation and selection for neomycin resistance
(G418, 400 µg/ml) (26). Single cell clones of neomycin-resistant cells were obtained.
Poly+A RNA was isolated from cells using a modification of the guanidinium isothiocyanate method, as described (27). Two to 5 µg of the resulting mRNA was separated by 0.8% agarose gel electrophoresis, transferred to a nylon membrane, and probed with nick-translated 32P-cDNA probes prepared from gel-purified inserts.
SREBP Analysis of Nuclei and Membrane FractionsCells were harvested on ice in cold phosphate-buffered saline containing a protease inhibitor mixture (1 mM phenylmethylsulfonyl fluoride, 100 µg/ml aprotinin, and 50 µg/ml leupeptin) using a rubber policeman. Nuclei and membrane fractions were obtained using the method described by Wang et al. (17). Cells were centrifuged at 1000 rpm for 10 min, and the pellets were resuspended in 10 volumes of cell homogenization buffer (10 mM HEPES-KOH at pH 7.6, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM dithiothreitol, 1 mM EDTA, and proteinase inhibitors). The cells were disrupted by passage through a 22-gauge needle (15 times) and then centrifuged at 1000 rpm for 10 min. The crude nuclear pellet was extracted with an equal volume of nuclear extraction buffer (20 mM HEPES-KOH at pH 7.6, 25% glycerol, 0.5 M NaCl, 1.5 mM MgCl2, 1 mM EDTA, and proteinase inhibitors) and centrifuged at 12,000 rpm for 30 min. The supernatant was used as nuclear extract for immunoblotting analysis. The microsomal membrane fraction was obtained by further centrifugation of the supernatant obtained from the nuclear pellet by ultracentrufugation at 45,000 rpm for 2 h using a TLA45 rotor (Beckman).
Western Blot AnalysisWestern blotting was performed as described (28). Following SDS-PAGE (1-15% gradient), the gels were electroblotted onto nitrocellulose membranes. The nonspecific binding sites of the membranes were blocked using 10% defatted dried milk, followed by addition of the appropriate primary antibody. The relative amount of primary antibody bound to the proteins on the nitrocellulose membrane was detected with the species-specific horseradish peroxidase-conjugated IgG. Blots were developed using the enhanced chemiluminescence detection kit (Amersham Corp.).
[35S]Methionine Labeling and ImmunoprecipitationWild-type McArdle RH-7777 and SLW-1 cells were grown to 85% confluency on 60-mm plates, after which the cells were cultured in serum-free DMEM for 24 h. On the 2nd day, the cells were labeled for 4 h with [35S]methionine (100 µCi/ml) containing serum-free DMEM in the presence of either 1 mM oleate/BSA or BSA (0.8%) alone. After labeling, cells and media were harvested as described (28), and cells were washed once with phosphate-buffered saline and solubilized in 1 ml of TETN buffer (25 mM Tris at pH 7.5, 5 mM EDTA, 250 mM NaCl, 1% Triton X-100, and proteinase inhibitors). Proteins were immunoprecipitated using specific antibodies, as described (28).
Analysis of Lipid BiosynthesisCells were incubated in serum-containing medium with [1-14C]acetate (5 µCi, specific activity 47 mCi/mol) for 2 h, extracted with chloroform/methanol (2:1, v/v), separated on silica gel TLC plates, and the radioactivity in phospholipids, free cholesterol, triglycerides, and cholesterol esters was quantitated (29). The cellular content of free and esterified cholesterol was determined in cells cultured to ~85% confluency by gas liquid chromatography, as described (14).
Statistics AnalysisResults are given as means ± S.D. Linear regression analysis was determined by Sigma Plot computer program, and significance of correlation constants was determined by Student's t test. Values of p < 0.05 were considered to be significant. The number of data points obtained from individual single cell clones used for linear correlations differed from experiment to experiment due to loss of a particular single cell clone. All data points were used and none were selectively deleted.
McArdle cells
were transfected with a plasmid that expresses neomycin resistance and
a cytomegalovirus promoter-driven expression of the coding region of
rat 7-hydroxylase. Following antibiotic selection, single cell
clones were picked and grown. One single cell clone of McArdle cells
stably expressing 7
-hydroxylase (SLW-1 cells) was used for detailed
metabolic studies, whereas single cell clones were used to establish
metabolic relationships (see below). Single cell clones of transfected
cells expressed 7
-hydroxylase mRNA and protein of the size
expected from the plasmid used (Fig. 1,
A and B). While wild-type McArdle rat hepatoma
cells express all of the gene products required for VLDL assembly and
secretion (19-21), they show no detectable expression of
7
-hydroxylase mRNA or protein (Fig. 1, A and
B).
Cellular Content of Mature SREBP1 Varies Linearly with 7
Single cell clones of McArdle cells
transfected with pcDNA-7 were subjected to Western blot analysis
using repetitive immunodetection with antibodies to SREBP1,
7
-hydroxylase, and albumin. Since the amount of albumin in either
the cell or medium was unchanged (see below), albumin served as an
internal standard for recovery and blotting. The cellular content of
mature SREBP1 was increased, whereas the content of the precursor form
was decreased in SLW-1 cells stably expressing 7
-hydroxylase (Fig.
2A). Incubating SLW-1 cells
and wild-type McArdle cells with 25-hydroxycholesterol, an established
inhibitor of SREBP1 proteolytic processing (17), decreased the content
of mature SREBP1 by about 90%, whereas the precursor form was slightly
increased (Fig. 2B). Moreover, analysis of all of the single
cell clones showed that the cellular content of mature SREBP1 varied as
a linear function of 7
-hydroxylase protein (Fig. 2C).
Additional studies show that the cellular content of SREBP2 was not
significantly affected by the expression of 7
-hydroxylase (data not
shown).
Secretion of ApoB100 Varies as a Linear Function of 7
The secretion of apoB100, apoB48,
and albumin by McArdle and SLW-1 cells was compared by Western blotting
of culture medium (Fig. 3A).
The blots were scanned by densitometry, and the amount of apoB100
secreted by SLW-1 cells was increased by 2-fold, compared with
wild-type McArdle cells (p < 0.05). There were no
significant differences in the secretion of either apoB48 or albumin by
the two different cell types. Since the expression of apoB mRNA was unaffected (Fig. 3B), the increased secretion of apoB100 was
the result of a post-transcriptional event. Analysis of the single cell
clones showed that the secretion of apoB100 varied as a linear function
of 7-hydroxylase protein (Fig. 3C).
Expression of MTP and 7
The content of MTP mRNA was significantly increased
in SLW-1 cells expressing 7-hydroxylase as compared with wild-type
McArdle cells. Using three individual plates of cells from each group, there was a 2.2-fold increase in the content of MTP mRNA relative to
-actin in SLW-1 cells compared with wild-type McArdle cells (Fig.
4A, p < 0.05). Moreover, the relative abundance of MTP mRNA varied in
proportion to the relative abundance of 7
-hydroxylase mRNA (Fig.
4B). These data indicate that expression of 7
-hydroxylase in McArdle cells induces the expression of MTP mRNA.
Expression of 7
Compared with wild-type cells, in
SLW-1 cells the synthesis of all VLDL lipids (cholesterol, cholesterol
esters, triglycerides, and phospholipids) was significantly increased
as determined by the incorporation of [14C]acetate (Fig.
5A). Since the cellular
content of mature SREBP is sensitive to the sterol content of cells
(15-18), we determined the cellular content of free and esterified
cholesterol in McArdle and SLW-1 cells (Fig. 5B). In SLW-1
cells, expressing 7-hydroxylase, the cellular content of both free
and esterified cholesterol was 80 and 50% of the levels in McArdle
cells (Fig. 5B). This experiment was repeated two times and
similar results were obtained.
Expression of HMG-CoA Reductase and Acetyl-CoA Carboxylase Varies as a Linear Function of 7
Recent reports
indicate that several genes, whose products control both sterol and
fatty acid synthesis, are targets for SREBP-induced transcription (15,
24, 30-33). The expression of the mRNA's encoding HMG-CoA
reductase, the rate-limiting enzyme in cholesterol biosynthesis, and
acetyl-CoA carboxylase, the rate-limiting enzyme in fatty acid
synthesis, was increased in SLW-1 cells compared with wild-type McArdle
cells (Fig. 6A). In contrast,
the relative abundance of SREBP1 mRNA was similar in both cell
types. Moreover, the expression of HMG-CoA reductase (Fig.
6B) and acetyl-CoA carboxylase mRNA (Fig. 6C)
varied as a linear relationship with the expression of 7-hydroxylase
mRNA. These data suggest that the expression of genes responsible
for the biosynthesis of cholesterol and the fatty acid components in
VLDL is coordinately influenced by the expression of 7
-hydroxylase
in McArdle cells.
Expression of 7
Oleic acid stimulates the
synthesis of glycerolipids and increases the secretion of apoB100 when
added to the culture medium of McArdle hepatoma cells (2). To examine
if 7-hydroxylase-mediated induction of glycerolipid synthesis (Fig.
5) "saturates" the lipid requirement for maximal secretion of
apoB100, we examined the effect of oleic acid on the cellular content
and secretion of apoB100 in wild-type and SLW-1 cells (Fig.
7). In McArdle and SLW-1 cells, oleic
acid increased the cellular content of apoB100 by 2.5-fold
(p < 0.01) and 1.5-fold (p < 0.05),
respectively (Fig. 7A). Adding oleic acid to the serum-free
BSA-containing medium increased the secretion of apoB100 by McArdle
cells (Fig. 7B). The oleic acid stimulation of apoB100
secretion by McArdle cells was specific since it did not significantly
affect the secretion of apoB48 or total protein (data not shown). These
data confirm those of others showing that in McArdle hepatoma cells,
oleic acid stimulates apoB100 secretion but not apoB48 secretion (34). Moreover, although oleic acid increased the secretion of apoB100 by
wild-type McArdle cells, it had no affect on the secretion of apoB100
by SLW-1 cells expressing 7
-hydroxylase (Fig. 7B). Since
oleic acid did significantly increase the cellular content of apoB100
in SLW-1 cells, albeit less than the increase observed in wild-type
McArdle cells (Fig. 7A), the unaltered apoB100 secretion in
SLW-1 cells cannot be due to impaired apoB100 synthesis. It appears
that SLW-1 cells respond similarly to oleic acid as do primary cultured
hepatocytes: increased content of cellular apoB but no change in
secretion (35, 36).
Increased Secretion of ApoB100 in SLW-1 Cells Is Blocked by 25-Hydroxycholesterol
To examine the possibility that
7-hydroxylase increases apoB100 secretion through a SREBP1-mediated
process, McArdle and SLW-1 cells were cultured with and without
25-hydroxycholesterol (2 µg/ml) plus cholesterol (10 µg/ml) for
24 h. The proteolytic processing of transmembrane SREBP precursors
to the mature transcriptionally active forms is inhibited by
25-hydroxycholesterol (17). Similar to its ability to decrease the
cellular content of mature SREBP1 in SLW-1 cells (Fig. 2B),
25-hydroxycholesterol also decreased the secretion of apoB100 to levels
that were similar to those of wild-type McArdle cells (Fig.
8). In the absence of
25-hydroxycholesterol, SLW-1 cells secreted 3.1-fold more apoB100
(p < 0.05) than McArdle cells. In the presence of
25-hydroxycholesterol, both cell types secreted similar amounts of
apoB100. Although 25-hydroxycholesterol did not significantly affect
the secretion of apoB100 in wild-type McArdle cells, it decreased the
secretion of apoB100 in SLW-1 cells by 71% (p < 0.02). The secretion of apoB48 and albumin by both groups of cells was
similar and unaffected by 25-hydroxycholesterol. This experiment was
repeated twice and similar results were obtained. The combined data
suggest that the augmented secretion of apoB100 displayed by McArdle
cells expressing 7
-hydroxylase can be reversed to levels detected in
wild-type McArdle cells by 25-hydroxycholesterol. The additional
finding that 25-hydroxycholesterol also restores the level of mature
SREBP1 in SLW-1 cells to those of McArdle cells suggests that SREBP1
mediates the augmented secretion of apoB100 caused by expression of
7
-hydroxylase.
SREBP was first identified as the sterol-responsive transcription factor that mediates the regulation of human LDL receptor gene (15). Its rat homologue ADD1 was identified as a transcription factor involved in adipocyte differentiation (37), suggesting that it may play a diverse role in regulating the expression of genes involved in the biosynthesis and metabolism of several classes of lipids. Subsequent studies show that many of the genes encoding isoprenoid/cholesterol biosynthetic enzymes are SREBP-responsive: HMG-CoA synthase (17), HMG-CoA reductase (31), isopentyfarnasyldiphosphate synthase (30), and squalene synthase (32). More recently, it has been reported that the genes for fatty acid synthase (33) and acetyl-CoA carboxylase (24) are both SREBP targets. The data derived from our studies now extend the metabolic regulatory importance of SREBP(s) to having a significant influence on controlling several steps of the VLDL assembly/secretion pathway.
Through the expression of 7-hydroxylase in McArdle hepatoma cells,
we were able to obtain single cell clones having a fairly large
variation in the cellular content of mature SREBP1 that varied as a
linear function of 7
-hydroxylase protein content (Fig.
2C). It is interesting to note that in vivo the
cellular content of mature SREBP2 is increased in the livers of
hamsters treated with bile acid sequestrants, which induce the
expression of 7
-hydroxylase (38). The mechanism through which
7
-hydroxylase expression increases the cellular content of mature
SREBP1 (this study) or SREBP2 in hamsters is not firmly established but
may involve altering the signaling of SREBP cleavage by regulatory sterols (39).
The following data and interpretations support the conclusion that
SREBP1 coordinately regulates the assembly and secretion of
apoB-containing lipoproteins. Since 7-hydroxylase expression via
plasmid transfection is the manipulation we employed to obtain SLW-1
cells, changes in lipoprotein assembly and secretion are the resulting
consequence of this liver-specific gene (23). These consequences
include linear relationships between 7
-hydroxylase mRNA and the
expression of mRNAs encoding the enzymes regulating fatty acid and
sterol synthesis (acetyl-CoA carboxylase and HMG-CoA reductase) and
MTP. As a result of the increased expression of these mRNAs and
perhaps others (not studied), SLW-1 cells display increased synthesis
of all lipoprotein lipids (cholesterol, cholesterol esters,
triglycerides, and phospholipids). These changes in lipid biosynthesis
and MTP expression may be responsible for the increased secretion of
apoB100. Moreover, as shown by the direct linear relationship between
the cellular content of 7
-hydroxylase and mature SREBP1 (Fig.
3C), it is likely that the changes in the expression of
genes regulating sterol and fatty acid biosynthesis and the assembly of
apoB-containing lipoproteins (i.e. MTP) caused by
7
-hydroxylase expression are mediated through the variation of the
cellular content of mature SREBP1. The additional findings that 1)
expression of 7
-hydroxylase (SLW-1 cells) decreased the cellular
content of both free and esterified cholesterol (Fig. 5B),
and 2) 25-hydroxycholesterol concomitantly decreases the cellular
content of mature SREBP1 in SLW-1 cells and decreases the secretion of
apoB100 to the same levels exhibited by wild-type McArdle cells provide
compelling evidence that the cellular content of mature SREBP1 is
responsible for these changes. Based on this rationale, we propose that
SREBP1 is a common regulatory mediator responsible for coordinate
regulation of the lipoprotein assembly and secretion pathway in McArdle
hepatoma cells. It is important to point out that in other cell types
(e.g. liver in vivo (38)), SREBP2 may mediate
changes in lipoprotein assembly and secretion. Recent studies in our
lab show that in a differentiated hepatoma cell line found to express
liver-specific genes not expressed by McArdle cells, SREBP2 induces the
expression of the MTP gene.2
Thus, depending upon the cell type and physiologic conditions, SREBP1
or SREBP2 may act to mediate changes in lipoprotein assembly and
secretion. This coordinate regulation provides a means for the
efficient utilization of both lipids and apoB100 for transport from the
liver to peripheral tissues.
It is now established that under most conditions, hepatic secretion of
apoB is regulated post-transcriptionally by a process that determines
the portion of de novo synthesized apoB that either enters
the lipoprotein assembly/secretion pathway or is degraded in the
endoplasmic reticulum (1, 2, 9). Studies using cultured hepatocytes and
perfused livers of rats showed that the translocation of apoB across
the endoplasmic reticulum is inefficient (40). One pool consisting of
incompletely translocated apoB appeared to be diverted into the
degradation pathway, whereas the pool of apoB that is completely
translocated into the lumen can ultimately be secreted as a lipoprotein
particle (40). Additional studies using human hepatoma HepG2 cells (41)
and McArdle rat hepatoma cells (42) showed that the availability of
glycerolipids (triglycerides and phospholipids) affects the efficiency
of apoB translocation across the endoplasmic reticulum. It appears that when the availability of lipids is insufficient, translocation of
de novo synthesized apoB is disrupted causing it to be
degraded. The majority of the degradation of translocation-arrested
apoB occurs co-translationally in HepG2 cells (43). While the
availability of lipid is required for efficient translocation and
utilization of apoB (9), the presence and activity of MTP is essential. In Chinese hamster ovary cells, which do not normally express MTP,
essentially all of the apoB synthesized from a stably expressed plasmid
is translocation arrested and rapidly degraded by a ALLN-inhibitable process (26, 28). In addition, blocking the functional activity of MTP
also blocks the secretion of apoB by hepatoma cells, which do express
MTP (44). Our results showing that the increased secretion of apoB100
correlated with MTP and mature SREBP1, but was not associated with any
detectable change in apoB mRNA (Fig. 3B), suggest that a
post-transcriptional mechanism is responsible. The additional finding
that oleic acid increases the secretion of apoB100 by wild-type McArdle
hepatoma cells, but has no effect in McArdle cells expressing
7-hydroxylase (Fig. 7), supports the proposal that SREBP1 induction
of genes regulating lipogenesis saturates the requirement for lipids in
the lipoprotein assembly pathway of SLW-1 cells. Our findings that in
SLW-1 cells oleic acid increased the cellular content of apoB100, but
not its secretion, suggest that the intracellular degradation was
blocked. The inability of oleic acid to increase the secretion of
apoB100 in SLW-1 cells suggests that one or more of the processes
required to assemble and secrete apoB100 lipoproteins other than lipid
availability is limiting. Based on these data it is tempting to
speculate that in McArdle cells, when the availability of lipids is in
excess of what is required for lipoprotein assembly, MTP becomes
rate-limiting for apoB100 translocation and the subsequent assembly and
secretion of apoB100-containing lipoproteins.
Our results may explain why oleic acid stimulation of lipogenesis is not associated with increased apoB secretion in primary rat hepatocytes (35, 36), whereas oleic acid increases the secretion of apoB by hepatoma cells (HepG2 (45-47) and McArdle (34, 48, 49)). In hepatoma cells, glycerolipid biosynthesis may be limiting for apoB secretion, whereas in primary hepatocytes other processes may be limiting.
There are other examples in which limitations in the availability of lipids in addition to triglycerides restrict the ability of the liver cell to assemble and secrete apoB-containing lipoproteins. Decreased production of cholesterol and cholesterol esters caused by drugs that competitively inhibit HMG-CoA reductase and/or ACAT also decrease the assembly and secretion of apoB-containing lipoproteins in some (50, 51) but not all (52) studies. Decreased production of phosphatidylcholine (via choline deficiency (53) and altered head-group precursors (54)) also blocks the secretion of apoB-containing lipoproteins. It seems reasonable to propose that a deficiency in the availability of one or more of the lipid components required for the intracellular processing of apoB and assembly of the lipoprotein particle can impair the overall pathway. Optimal utilization of lipids and apoB for lipoprotein assembly requires that each essential protein and lipid component be in sufficient supply relative to the others to assemble the lipoprotein particle.
Coordinate changes of the individual processes required for VLDL assembly and secretion are invoked in response to physiologic state (reviewed in Ref. 55). For example, carbohydrate "overload" leads to a coordinate increase in the synthesis of all lipoprotein lipids and an amplification of all of the processes required to increase the secretion of apoB-containing lipoproteins (10). Conversely, fasting leads to the opposite coordinated responses: decreased apoB translocation, increased apoB degradation, decreased lipid biosynthesis, and decreased lipoprotein secretion (12, 13). Fetal nutritional development also shows a coordinated change in apoB processing, lipogenesis, and lipoprotein secretion (11). These findings have led us to propose that one or more metabolic signals is responsible for the coordinate regulation of lipid biosynthesis and the intracellular processing of apoB.
Our present data show that expression of 7-hydroxylase in McArdle
cells alters the cellular content of mature SREBP1 in a manner that
coordinately changes the availability of lipids and the processes
required to efficiently assemble apoB-containing lipoproteins. We also
observed that in McArdle cells expressing 7
-hydroxylase,
25-hydroxycholesterol decreases both the cellular content of mature
SREBP1 and the secretion of apoB100. Our findings may explain the
well-established observation that in patients treated with bile acid
sequestrants, known to increase the expression of 7
-hydroxylase,
there is a concomitant increase in the secretion of VLDL
triglycerides (56, 57). Furthermore, in some patients, there
appears to be a direct correlation between the rate of bile acid
synthesis, a parameter linked to 7
-hydroxylase activity, and
VLDL-triglyceride secretion (56, 57). The combined data suggest that
changes in cholesterol metabolism may alter VLDL assembly and secretion
by 1) affecting the amount of cholesterol and cholesterol esters that
are available for VLDL assembly and 2) by altering the cellular content
of mature SREBP. It is through the second parameter (cellular content
of mature SREBP) that coordinate response of the gene products
controlling the availability of glycerol lipids (i.e.
triglycerides and phospholipids) and MTP is linked to the first
parameter (cholesterol metabolism). Optimal production of VLDL in
mammals may require coordinate induction of all of the gene products
necessary for the synthesis of individual lipids and for their assembly
with apoB (e.g. MTP). This may be comparable to other
species (e.g. birds) in which VLDL secretion by the liver is
optimized by a coordinate regulation of gene expression by estrogen
(58-61). However, it is important to emphasize that mammalian VLDL
production may not be optimal under all conditions. Variation in
physiologic conditions may determine which of the many processes are
rate-limiting for VLDL assembly and/or secretion.
We gratefully acknowledge the following
people who provided reagents for these studies: Drs. Timothy Osborne
(antibody to SREBP1 and cDNA for acetyl-CoA carboxylase), Mike
Briggs (antibody to SREBP2), David Russell (original cDNA for rat
7-hydroxylase used for newly constructed expression plasmid), Aldons
Lusis for the cDNA for rat apoB, and David Gordon, Richard
Gregg, John Wetterau, and their colleagues at Bristol-Myers Squibb for
the MTP cDNA. We thank Drs. Huda Shubeita and John Trawick for help
in preparing the plasmid expressing 7
-hydroxylase. In addition, we
thank Christian A. Drevon for his reading of the manuscript and helpful
suggestions.