Level of Apolipoprotein B mRNA Has an Important Effect on the Synthesis and Secretion of Apolipoprotein B–Containing Lipoproteins
Studies on Transfected Hepatoma Cell Lines Expressing Recombinant Human Apolipoprotein B
Abstract The effect of apoB mRNA level on hepatic apoB production has not been studied extensively, primarily because the steady state level of apoB mRNA cannot be altered on a short-term basis. We studied the effect of vastly different apoB mRNA levels on the synthesis and secretion of apoB-containing lipoproteins using rat hepatoma (McA-RH7777) cell lines transfected with cDNA constructs encoding human apoB53 (the amino-terminal 53% of the protein; hapoB53) or apoB100 (hapoB100). Among the three hapoB53-transfected cell lines, the relative steady state levels of the hapoB53 mRNA were 10:2.5:<0.1. Correspondingly, the relative concentration of the intracellular hapoB53 protein was 8:3:1 and of the medium hapoB53 (accumulated over a period of 18 hours) was 12:4:1, which positively correlates with the hapoB53 mRNA levels. The expression level of hapoB53 did not affect the buoyant density of lipoproteins containing hapoB53 (d=1.06 to 1.21 g/mL) or endogenous rat apoB100 (d<1.06 g/mL). When cell lines containing high or intermediate hapoB53 mRNA levels were compared, there was an eightfold increase in the synthesis and a twofold increase in the secretion efficiency of hapoB53. Analysis of the synthesis and secretion of lipids revealed that in cells producing high levels of hapoB53, triglyceride synthesis (twofold) and secretion (twofold to threefold) were also increased. Furthermore, with the three hapoB100-transfected cells we also observed an increase in apoB100 synthesis (threefold), apoB100 secretion efficiency (twofold), triglyceride synthesis (fourfold to fivefold), and triglyceride secretion (fourfold to fivefold) in the cells expressing high levels of hapoB100. In all the cell lines examined, secretion efficiency of endogenous rat apoA-I was not affected by transfection. Together these data suggest that secretion of apoB-containing triglyceride-rich lipoproteins can be influenced by the level of apoB mRNA or the rate of apoB translation.
- Received November 30, 1994.
- Accepted July 21, 1995.
The recognition of the atherogenic nature of plasma apoB such as VLDL and LDL, has stimulated investigation into mechanisms that regulate the biosynthesis of apoB.1 In humans, apoB is primarily synthesized in the liver and intestine, and its primary structure has been completely characterized at the levels of both gene and protein.2 The level of apoB gene expression in the liver seems to be tightly controlled; the steady state levels of apoB mRNA remain relatively constant under conditions in which the level of apoB protein production changes significantly.3 4 5 6 7 Sequence elements or regions that enhance or reduce the apoB gene expression have been mapped,8 9 10 but the mechanisms whereby a relatively constant apoB mRNA level is maintained in the liver remain unknown.
Posttranslational regulation of apoB secretion has been demonstrated to be an important mechanism that governs the amount of apoB secreted.1 Many research groups working with different systems have reported that the amount of apoB synthesized in the liver cells is in excess of the amount secreted.11 Under conditions that are unfavorable to apoB secretion, a great proportion of the newly synthesized apoB proteins is degraded intracellularly.12 Although the mechanism that regulates apoB intracellular degradation is poorly understood, it has been shown that the secretion efficiency (the amount secreted versus the amount synthesized) of apoB is augmented when substrates for biosynthesis of lipids, including triglycerides,13 14 cholesteryl esters,15 16 and phospholipids,17 are abundant. It is generally believed that the lipid availability plays a major role in the posttranslational regulation of apoB-containing lipoprotein secretion.
Secretion of apoB-containing lipoproteins may not be solely regulated by the availability of lipid. Accumulating evidence suggests that the steady state level of hepatic apoB mRNA may change, although the magnitude of change is small (less than twofold). Increases in apoB mRNA levels were observed with HepG2 cells on the addition of 25-hydroxycholesterol,15 low concentrations of amino acids,18 or VLDL.19 In all cases, secretion of apoB100 by the cells was increased, suggesting a positive correlation between the levels of apoB mRNA and apoB secretion. Altered apoB mRNA concentrations were also observed during in vivo studies with animals and humans. Elevated apoB mRNA levels were found in the liver of diabetic rats given insulin for 7 days.20 Elevated levels of the apoB mRNA and intracellular apoB protein were also found in liver biopsies from patients with abetalipoproteinemia,21 a disease in which the assembly and secretion of apoB-containing lipoproteins is defective.22 On the other hand, reduced apoB mRNA levels were observed in liver biopsies from some patients homozygous for familial hypobetalipoproteinemia,23 an autosomal codominant disorder associated with abnormally low plasma apoB concentrations.24 Recently, phenotypes similar to hypobetalipoproteinemia including reduced apoB mRNA levels were observed in transgenic mice homozygous for a truncated mouse apoB protein.25 It remains to be determined whether changes in the levels of apoB mRNA have significant effects on the production of apoB-containing lipoproteins.
In principle there are at least two approaches that can be taken to achieve significant alteration of the apoB mRNA levels. One approach is to disrupt one allele of the apoB gene to reduce the level of the gene product. Targeted apoB gene disruption has been achieved in HepG2 cells26 and shown to reduce the level of apoB transcript, but the effect of the disruption on lipoprotein production has not been pursued. Alternatively, transgenic mice or cell lines that express different steady state levels of apoB mRNA can be generated. Transgenic mice expressing different levels of human apoB100 (hapoB100) have been created, and a positive correlation between the transgene copy number and plasma hapoB100 concentration has been suggested.27 28 It remains to be determined whether plasma levels of the recombinant hapoB found in the transgenic mice are attributable to transgene expression levels or owing to the failure to clear the hapoB-containing lipoproteins, or both. Using rat hepatoma McA-RH7777 cells (American Type Culture Collection) stably transfected with human apoB48 (hapoB48), we have recently demonstrated that the synthesis and secretion of endogenous rat apoB100 (rapoB100) are inhibited by hapoB48 overexpression.29
In this study, we used McA-RH7777 cells that express different steady state levels of two hapoB proteins, namely hapoB53 (the amino-terminal 53% of apoB100) and hapoB100, to determine the effects of apoB mRNA level on hepatic lipogenesis and production of both hapoB- and rapoB-containing lipoproteins. Our results suggest that the level of hapoB mRNA or the rate of hapoB translation has a profound effect on the level of hapoB-containing lipoprotein secretion.
Cell culture medium, serum, and G418 (the neomycin analog Geneticin) were obtained from Life Technologies. Reagents for polyacrylamide gel electrophoresis and nitrocellulose membranes for blotting were obtained from Bio-Rad. Paragon agarose gels were obtained from Beckman, Ltd. Monoclonal antibody 1D1 was a gift from R.W. Milne and Y.L. Marcel (University of Ottawa Heart Institute). Polyclonal antibodies specific for rapoB, apoA-I, and apoE were gifts from R. Davis (San Diego State University) and T. Ohnishi (University of Alberta). Peroxidase-conjugated anti-mouse or anti-rabbit immunoglobulin G antibodies and the enhanced chemiluminescent (ECL) western blotting detection system were obtained from Amersham International. Sheep anti-hapoB antiserum was obtained from Boehringer Mannheim. Protein A–Sepharose CL–4B beads were obtained from Pharmacia-LKB Biotechnology, Inc. [32P]dCTP, [35S]methionine (Tran35S-label, 1100 Ci/mmol), and [3H]glycerol were purchased from ICN Biomedicals Inc or Amersham.
Cell Culture and Transfection
McA-RH7777 was cultured in DMEM containing 10% fetal bovine serum and 10% horse serum in 100-mm dishes (Falcon). The cells were transfected with the expression plasmid pB53L-L30 or pB100L-L31 and pSV2neo to generate stable transformants as described previously. In these apoB constructs, codon CAA (Gln) at position 2153 of the hapoB cDNA was mutagenized into CTA (Leu) (pB53L-L is shown in Fig 1A⇓). We demonstrated previously that in McA-RH7777 cells transfected with pB53L-L30 but not with pB100L-L31 this mutation substantially reduced formation of hapoB48. Production of hapoB53 or hapoB100 by individual clones was determined by Western blot analysis (see below), and stable cell lines that showed a positive reaction were maintained in the medium containing 200 μg/mL G418. To determine the effect of transfection on the growth rate of McA-RH7777 cells, we plated the transfected cells at 3×105 per dish (60-mm) and then monitored cell protein concentrations at 24, 48, and 72 hours after plating by the method of Lowry et al.32 Because the plating density (number of cells per dish) of McA-RH7777 cells has been shown to have a profound effect on lipoprotein synthesis and secretion,29 33 we performed all experiments at similar plating densities with the stably transfected cell lines.
Preparation of RNA and RNase Protection Assay
Total RNA was prepared from the hapoB53-expressing cell lines by the method of Chirgwin et al.34 Fifteen micrograms of total RNA was hybridized to a 32P-labeled riboprobe encompassing sequences from –85 to +121 of the hapoB gene in a buffer containing 80% formamide, 40 mmol/L PIPES, pH 6.4, 0.4 mol/L NaCl, and 1 mmol/L EDTA at 55°C for 16 hours. The hybrids were then digested with ribonuclease T1 at a final concentration of 4 μg/mL for 1 hour at 30°C. The reaction was terminated by addition of SDS and proteinase K and incubated at 37°C for 15 minutes. After extraction with phenol/chloroform the RNA was recovered by ethanol precipitation with the use of 10 μg of yeast tRNA as a carrier, dissolved in formamide loading buffer, and fractionated on an 8% polyacrylamide sequencing gel. The relative intensities of the bands were determined densitometrically with the use of a CAMAG chromatography scanner, and the area under each peak was integrated using a CAMAG SP4290 integrator.
Immunoblot Analysis of Apolipoproteins
Transfected cells were plated in 60-mm Primaria dishes (Becton Dickinson & Co) and grown to subconfluency (60% to 70%, 0.6 to 0.7 mg of cell protein). The cells were incubated with 3 mL of serum-free DMEM for 16 hours, and the conditioned medium was collected and supplemented with EDTA (0.5 mmol/L) and phenylmethylsulfonyl fluoride (0.015%). Lipoproteins in the medium were absorbed onto fumed silica (Cab-O-Sil) as previously described,17 and apolipoproteins were eluted into 200 μL sample buffer (6 mol/L urea, 2% SDS). Samples derived from equal amounts of cell proteins were resolved by electrophoresis on a 5% polyacrylamide gel containing 0.1% SDS (SDS-PAGE) and transferred onto nitrocellulose membranes for immunoblotting as described previously.31 The relative intensities of the bands were measured using a CAMAG scanner as described above.
Ultracentrifugation of Lipoproteins
Cells (100-mm dishes, 60% to70% confluent) were incubated with 8 mL of serum-free DMEM for 16 hours. The conditioned media from two dishes were combined, and the lipoproteins were fractionated by ultracentrifugation in a salt gradient as described previously.35 In some experiments, separation of lipoproteins containing rapoB100 or hapoB53 was accomplished by a single-spin ultracentrifugation at d=1.05 g/mL for 24 hours at 50 000 rpm, 10°C (Ti 70.1 rotor). Two fractions (d<1.05 and d>1.05 g/mL) were collected, and lipoproteins were concentrated with Cab-O-Sil for electrophoresis and immunoblotting as described above.
Agarose Gel Electrophoresis
Conditioned serum-free media were concentrated tenfold with the use of Centricon-10 concentrators (Amicon Inc). Aliquots of the concentrated samples were fractionated by electrophoresis on agarose gels (Paragon, Beckman, Ltd), and lipoproteins were transferred onto nitrocellulose membranes and blotted for hapoB, rat apoE, or apoA-I, as previously described.36 Human LDL and rat HDLs prepared by sequential flotation were used as β- and α-migrating lipoprotein standards, respectively.
Metabolic Labeling of Lipids and Lipid Analysis
For metabolic labeling experiments, the cells (60-mm) were washed twice with serum-free DMEM and incubated with [3H]glycerol (10 μCi/mL) in the same medium for up to 3 hours. At the indicated times, cells and media were collected and lipids were extracted with chloroform/methanol (1:1, by volume). The extracted lipids were dried under nitrogen gas and subjected to TLC on 20×20-cm silica plates as described previously.17 Triglyceride and phosphatidylcholine bands were identified by comparison with lipid standards and scraped from the plates; radioactivity associated with the lipid samples was quantified by scintillation counting.
For lipid mass measurement, cells (100-mm) were cultured in serum-free DMEM for 16 to 18 hours. Lipids were extracted from the conditioned media combined from two dishes and subjected to triglyceride or total cholesterol analyses using enzymatic reagents (Wako Pure Chemical Industries, Ltd) according to the manufacturer’s instructions. For semiquantitative analysis of medium lipids, the conditioned media were fractionated into d<1.05 and d>1.05 g/mL fractions, and lipids extracted from the two fractions were resolved by HPTLC as described previously.17
Metabolic Labeling of Apolipoproteins
For measuring the synthetic rate of apoB, cells (60-mm) were washed twice with methionine-deficient DMEM and then incubated with 1 mL of medium supplemented with 200 μCi [35S]methionine for up to 30 minutes. The labeled apoB proteins within the cells were recovered at indicated times during the labeling period by immunoprecipitation as previously described.35 Pulse-chase experiments for apoB and apoA-I were performed according to McLeod et al.31 In experiments in which hapoB53-transfected cells were used the length of pulse was 30 minutes, and the chase medium contained 2 mmol/L methionine.
Reuptake of secreted apolipoproteins was tested by the following experiment. Cells (100-mm) were labeled with [35S]methionine (1.1 mCi/5 mL methionine-deficient medium) for 2 hours and then incubated with DMEM containing 2 mmol/L methionine for 2 hours (chase). The chase medium was added to fresh cells and incubated for up to 4 hours. Cells and medium were collected at the indicated time points, apoB and apoA-I were immunoprecipitated, and the radioactivity associated with the apolipoproteins was quantified.
Characterization of Human ApoB53-Expressing Cell Lines
From 18 G418-resistant McA-RH7777 cell clones transfected with the pB53L-L plasmid,30 we detected 4 clones that contained hapoB53 within the cell and secreted the protein into the medium, namely B53-1, B53-10, B53-14 (Fig 1B⇑), and B53-4 (not shown). Densitometric estimation of the amount of hapoB53 visualized on the immunoblot showed that the relative intracellular levels of hapoB53 among the four cell lines were 8:3:1 for B53-1 (and B53-4):B53-14:B53-10, whereas the relative amounts of hapoB53 accumulated in the medium during an 18-hour period were 12:4:1 for B53-1 (and B53-4):B53-14:B53-10. Since clones B53-4 and B53-1 produced similar levels of the hapoB53, we chose the latter to represent the high hapoB53-expressing cell line for the subsequent studies. Transfection had no adverse effect on cell growth as monitored by measurement of the cell protein mass over a period of 3 days; the doubling time for cell protein mass was approximately 24 hours for all the transfected cell lines and the nontransfected McA-RH7777 cells.
The differences in cell and medium hapoB53 levels among the transfected cell lines are most likely attributable to different copy numbers of the transgene. Analysis of the steady state levels of the hapoB53 mRNA by RNase protection assay (Fig 1C⇑) revealed that the mRNA level was directly correlated with the amount of hapoB53 protein synthesized and secreted. Estimation of the RNase protected bands (Fig 1C⇑) by densitometry showed that the relative hapoB53 mRNA levels were 10:2.5:<0.1 for B53-1 (and B53-4):B53-14:B53-10. The effect of overexpression of hapoB53 on the production of rapoB100 and rapoB48 was determined by immunoblotting with the use of an antibody specific to rapoB (Fig 1D⇑). When cell lines that represent high (B53-1), intermediate (B53-14), and low (B53-10) expression of hapoB53 were compared, we found that neither the steady state concentrations of intracellular rapoB100 (Fig 1D⇑, left) nor the amount of secreted rapoB100 (Fig 1D⇑, right) was affected by the level of transgene expression. The steady state levels of rapoB100 secretion in transfected cells were decreased by ≈50% compared with the nontransfected control cells. The reduction in rapoB100 secretion was independent of transgene expression levels, the relative amount was 0.53:0.49:0.52:1 for B53-1:B53-14:B53-10:McA-RH7777 as determined by densitometry of immunoblots (not shown). Since there was very little rapoB48 detected in McA-RH7777 cells (Fig 1D⇑), the effect of overexpression of hapoB53 on rapoB48 production could not be accurately assessed.
We examined the buoyant density of the hapoB53-containing lipoproteins secreted by the hapoB53-expressing cells. Previously we have shown that the buoyant density of apoB-containing lipoproteins is inversely related to the length of apoB polypeptides.30 31 Fig 2A⇓ demonstrates that the majority of the hapoB53-containing lipoproteins secreted by the cells have densities resembling HDL (d=1.06 to 1.21 g/mL), and their density distribution is independent of hapoB53 expression level. This result confirms that the buoyant density of apoB-containing lipoproteins is determined by apoB length. Notably, although the construct was designed to eliminate the formation of hapoB48, a protein approximately the same size as apoB48 (≈220 kD) was observed (see Fig 2A⇓, B53-10 fractions). This apoB48-like protein has been observed in previous studies,30 31 and its production is probably attributable to premature polyadenylation of the hapoB mRNA in these cells.37 38 We then compared the density distribution of rapoB100 in the media. Fig 2B⇓ demonstrates that rapoB100 secreted by the transfected cells is primarily associated with LDL/VLDL particles (d<1.06 g/mL), and its density distribution is comparable to that in nontransfected cells (Fig 2B⇓, bottom). Flotation of the medium lipoproteins (Fig 2C⇓, 2D⇓, and 2E⇓) confirmed that rapoB100 was confined to the d<1.05 g/mL fraction, whereas hapoB53 and endogenous apoE and apoA-I were associated with d>1.05 g/mL fractions. The absence of an effect on the buoyant density of rapoB100 together with the lack of an inhibitory effect on accumulation of endogenous apoE or apoA-I in the medium (Fig 2D⇓ and 2E⇓) suggest that expression of hapoB53 exerted little adverse effect on the metabolism of host apolipoproteins. In addition, analysis of total medium proteins resolved by SDS-PAGE and stained with Coomassie blue did not reveal significant changes in overall protein secretion from the transfected cells (data not shown). Because buoyant density of lipoprotein particles reflects their lipid content, and because the high hapoB53-expressing cells (B53-1) secrete more hapoB53, they may also secrete more lipids.
Lipid Analysis with Human ApoB53-Expressing Cells
We quantified the intracellular content of three major lipids (triglyceride, total cholesterol, and phospholipids) and the content of these lipids in the media by enzymatic measurement (see the Table⇓) and also analyzed the overall composition of intracellular and medium lipids by HPTLC (Fig 3A⇓). Data summarized in the Table⇓ show that among the three hapoB53-expressing cell lines that were cultured at ≈3 mg cell protein/100-mm dish in serum-free media, there were no significant differences in the intracellular pool size of total cholesterol or phospholipids. However, the intracellular triglyceride concentration in B53-1 (22.7 μg/1 mg cell protein) was 16% lower than in B53-14 (27.0 μg/1 mg cell protein) and 25% lower than in B53-10 (30.5 μg/1 mg cell protein) (see the Table⇓). The lower triglyceride level and the similar cholesterol or phospholipid levels in B53-1 were also observed when these cells were cultured in media containing 20% serum (data not shown). These experiments excluded the possible effects of culture conditions (eg, serum-free medium) on the observed lipid contents. The intracellular lipid composition determined by the enzymatic measurements was confirmed by the semiquantitative HPTLC analysis (Fig 3A⇓); lanes 9 through 11 of the chromatogram show that triglyceride concentrations were decreased in B53-1 compared with B53-14 and B53-10.
Secretion of triglycerides into the media by clone B53-1 was 2.7-fold and 2.2-fold higher, respectively, than by clones B53-10 and B53-14 (see the Table⇑). Similarly, secretion of phospholipids by B53-1 was 2.4-fold higher than by B53-10 but was not different from that by B53-14 (see the Table⇑). The low level of medium lipids secreted by the transfected cells precluded an accurate estimation by enzymatic assays. Therefore, medium lipids were analyzed semiquantitatively by HPTLC (Fig 3A⇑, lanes 3 through 8). Because lipoproteins containing hapoB53 or rapoB100 could be separated by density ultracentrifugation (Fig 2C⇑ and 2D⇑), we fractionated the medium into d<1.05 and d>1.05 g/mL fractions and extracted lipids from each fraction. The majority of the secreted phosphatidylcholine, cholesterol, and triglycerides was associated with d>1.05 g/mL fractions (Fig 3A⇑, lanes 4, 6, and 8). As was the case for mass measurement, visual inspection of the HPTLC plate suggested that secretion of the three major lipids by clone B53-1 was higher than by clones B53-14 and B53-10 (Fig 3A⇑, lanes 4, 6, and 8). There was no difference in secretion of lipids associated with the d<1.05 g/mL fractions (Fig 3A⇑, lanes 3, 5, and 7); the decreased concentration of cholesterol in the d<1.05 g/mL fraction in cell line B53-1 (Fig 3A⇑, lane 3) was not reproducibly observed.
Using agarose gel electrophoresis and immunoblot analysis of the whole medium lipoproteins, we observed that hapoB53, like hapoB100, was associated primarily with β-migrating lipoproteins (Fig 3B⇑, left). The hapoB53-associated lipoproteins contained virtually no rat apoA-I or apoE, which was associated primarily with α-migrating lipoproteins (Fig 3B⇑, middle and right). Thus, the increased lipid level that was associated with the d>1.05 g/mL fractions is most likely attributable to the increased number of hapoB53-containing lipoproteins secreted by the high expressor cells.
Lipid synthesis and secretion were examined using the high (B53-1) and low (B53-10) expressors by [3H]glycerol incorporation experiments (Fig 4⇓). During a 3-hour period, triglyceride and phosphatidylcholine were the species that exhibited the highest [3H]glycerol incorporation among the cell and medium lipids. The rate of [3H]glycerol incorporation into cell triglyceride during the first 2 hours was approximately twofold greater in clone B53-1 than in clone B53-10, whereas the rate of [3H]glycerol incorporation into phosphatidylcholine between the two cell lines was similar during the first 2-hour labeling period (Fig 4⇓, left). Likewise, the rate of [3H]glycerol incorporation into secreted triglyceride in clone B53-1 was threefold (at 2 hours) and 4.5-fold (at 3 hours) higher than in clone B53-10, whereas [3H]glycerol incorporation into secreted phosphatidylcholine was increased by 2.5-fold during the 2-hour incubation period (followed by a turnover of 3H-labeled phosphatidylcholine) (Fig 4⇓, right). Taken together, the results of these metabolic labeling experiments are generally consistent with the mass measurement of the medium lipid levels.
Kinetic Studies of ApoB with Human ApoB53-Expressing Cells
We first determined the rate of apoB synthesis by [35S]methionine labeling experiments and found a linear incorporation of [35S]methionine into hapoB53 and rapoB100 during the first 30 minutes (data not shown). Thus, we used the radioactivity associated with apoB at the end of a 30-minute labeling period as a measure of apoB synthesis. When the rate of apoB synthesis between clones B53-1 and B53-14 (Fig 5⇓, top, time zero left and center) were compared, we found that the rate of hapoB53 synthesis in B53-1 was eightfold higher than in B53-14 (84 300 versus 10 700 cpm/30 minutes per milligram of cell protein). This increased hapoB53 synthesis may account for the 2.7-fold increase in cell hapoB53 mass and threefold increase in accumulation of hapoB53 in the medium (Fig 1B⇑). The rate of rapoB100 synthesis was similar between clones B53-1 and B53-14 (31 360 versus 29 980 cpm/30 minutes per milligram of cell protein); these results are in agreement with the unchanged steady state levels of intracellular or medium rapoB100 (Fig 1D⇑). Kinetics of apoB synthesis was also conducted with clone B53-10 (low expressor) and nontransfected McA-RH7777 cells in a separate experiment. The incorporation of [35S]methionine into hapoB53 in clone B53-10, however, was too low to permit accurate measurements. Nevertheless, incorporation of [35S]methionine into rapoB100 during a 2-hour period was identical in the low expressor and nontransfected control cells (32 900 and 33 800 cpm/dish per 2-hour period, respectively).
Fig 5⇑ summarizes the data obtained from pulse-chase experiments of hapoB53 and rapoB100 with clones B53-1 and B53-14. Accumulation of 35S-labeled hapoB53 in the medium at the end of a 4-hour chase was 24.6-fold higher in B53-1 than in B53-14 (46 900 versus 1900 cpm/1 mg cell protein), whereas the increase in 35S-labeled rapoB100 accumulation in B53-1 was less than twofold compared with B53-14 (4084 versus 2155 cpm/1 mg cell protein). The pulse-chase data (Fig 5⇑, top, left and center) indicate that the secretion efficiency (the percentage of maximum labeled apolipoproteins that was secreted) of hapoB53 in clone B53-1 (38%) was increased by greater than twofold compared with clone B53-14 (17%). (The values at the maximum labeling and 4-hour points were used for calculating hapoB53 secretion efficiency.) Like synthesis, the secretion of rapoB100 (Fig 5⇑, top, right) in clones B53-1 and B53-14 was also similar (secretion efficiency of 13% and 7%, respectively). In a separate experiment, low secretion efficiency of rapoB100 was also observed in the low expressor (B53-10) and in nontransfected cells (17% and 24%, respectively; data not shown). Such low secretion efficiency of rapoB100 for McA-RH7777 cells cultured under serum-free conditions was not unexpected; values ranging from 20% to ≤10% have been reported.31 35 39 Degradation of newly synthesized apoB was observed in all the cells examined (data from clones B53-1 and B53-14 are shown in Fig 5⇑, top; data from clone B53-10 and McA-RH7777 cells are not shown). At the end of a 4-hour chase, 71% and 64% of total (cells+medium) 35S-labeled hapoB53 and 37% and 20% of total 35S-labeled rapoB100, respectively, were recovered in clones B53-1 and B53-14. Finally, the secretion efficiency of endogenous apoA-I was not affected by the level of hapoB53 expression. Approximately 35% of total labeled apoA-I was secreted into the media at the end of a 2-hour chase by the transfected cells (B53-1 and B53-14) or nontransfected McA-RH7777 cells (Fig 6⇓).
To exclude the possibility of cellular reuptake of apoB-containing lipoproteins during the 4-hour chase period, we performed a control experiment in which the conditioned media containing the labeled apolipoproteins were added to fresh cells. Data obtained from this experiment (not shown) indicated that within the 4-hour incubation time frame, reuptake of labeled apoB or apoA-I by the cells was negligible.
Lipid and ApoB Metabolism in Human ApoB100-Transfected Cells
We extended our studies using McA-RH7777 cells stably transfected with hapoB100 to further investigate the effects of apoB overexpression on the synthesis and secretion of hepatic apoB and lipid. Three stable transformants were obtained (Fig 7⇓); one (B100-18) contained a high intracellular steady state level and secreted high quantities of hapoB100 into the media, whereas the others (B100-24 and B100-36) expressed very low levels of the protein (Fig 7A⇓). Densitometric estimation showed that the relative amount of hapoB100 accumulated in the media during an 18-hour period was 30:5:1 for B100-18:B100-24:B100-36. Significant quantities of the hapoB48-like protein were detected in the cells and media of all three transfected cell lines. The intracellular steady state level of rapoB100 and its secretion were not affected by the transfection (Fig 7B⇓) neither was the density distribution of hapoB100 (Fig 7C⇓) altered. Steady state levels of rapoB secreted by the transfected cells were comparable to nontransfected cells, achieving a ratio of 1.1:1.1:1.3:1 for B100-18:B100-24:B100-36:McA-RH7777 as determined by densitometry (data not shown).
When the high (B100-18) and low (B100-24) expressor cell lines were compared there was a 3.5-fold increase in the synthesis of apoB100 and apoB48 as determined by [35S]methionine incorporation experiments (Fig 8⇓, time zero as seen on left). It is relevant to mention that the antibody used in these experiments reacts with both hapoB and rapoB, thus the immunoprecipitated 35S-labeled apoB100 contained rapoB100 and hapoB100 (≈550 kD), whereas 35S-labeled apoB48 contained rapoB48 (trace amount) and hapoB48-like proteins (≈220 kD). Secretion of apoB100 and apoB48 was increased in the high expressor cell line. The maximum radioactivity associated with medium apoB100 during the chase period was 5.7-fold higher in clone B100-18 than in clone B100-24 (15 400 versus 2700 cpm/1 mg cell protein), whereas radioactivity associated with medium apoB48 was 4.1-fold higher in B100-18 than in B100-24 (21 200 versus 5100 cpm/1 mg cell protein). The secretion efficiency of apoB100 was increased by twofold in the high expressor compared with the low expressor, ≈40% in clone B100-18 and ≈20% in clone B100-24. Pulse-chase analysis demonstrated that the secretion efficiency of endogenous apoA-I between clone B100-18 and B100-24 was identical (data not shown).
Finally, synthesis and secretion of lipids between clones B100-18 and B100-24 were determined by [3H]glycerol-labeling experiments. Data summarized in Fig 9⇓ demonstrate that synthesis and secretion of radiolabeled triglycerides in clone B100-18 were increased by four- to fivefold compared with B100-24 (Fig 9⇓, top). Incorporation of [3H]glycerol into phosphatidylcholine, however, was similar in the two cell lines (Fig 9⇓, bottom).
Taken together, data from the hapoB53- and hapoB100-expressing cell lines all indicate that the level of apoB expression directly correlates with the synthesis and secretion of apoB-containing lipoproteins, and increased apoB expression may also have a stimulatory effect on hepatic triglyceride synthesis.
Despite a number of studies3 4 5 6 7 that have demonstrated that the steady state level of apoB mRNA remains relatively unaltered under conditions in which the level of apoB100 secretion changes dramatically, accumulating evidence suggests that the steady state level of apoB mRNA can be changed and that the altered apoB mRNA levels sometimes positively correlate with the level of apoB secretion.15 18 19 In the present study, we used an in vitro system to assess the interrelation between the secretion of apoB-containing lipoproteins and the level of apoB gene expression. The transfected cells examined exhibited vastly different synthetic rates of hapoB53 or hapoB100 proteins. The data clearly indicate that in the transfected cells, the amount of hapoB secreted and hapoB secretion efficiency (the proportion of total apoB that is secreted) parallel the level of hapoB synthesis; if more hapoB is produced, then more hapoB-containing lipoproteins are secreted. This observation was somewhat unexpected, since we originally thought that in liver cells the availability of lipids would limit the amount of apoB secreted, with the remainder diverted to degradation.1 Under conditions in which the hepatic lipid supply was limiting, the output of hapoB-containing lipoproteins was expected to be restricted even if more hapoB molecules were synthesized in the transfected cell lines. However, the current results suggest that the availability of lipids may not be the sole limiting factor that governs the efficiency of apoB secretion. Rather, our data suggest that the production of apoB-containing triglyceride-rich lipoproteins can be stimulated by the increased level of apoB translation.
Unlike other apolipoproteins (eg, apoE and apoA-I) that can be secreted from the cells without association with significant amount of lipids, apoB requires proper assembly with lipids into lipoproteins for secretion. With the use of hepatoma cell lines expressing various truncated hapoB forms, several research groups31 35 39 40 41 have provided evidence for a direct correlation between the apoB length and the amount of lipids assembled. Among all the hapoB forms examined (including hapoB53) that are long enough to assemble a lipid core, a significant proportion of the nascent apoB is subjected to degradation. Supplementation of the culture medium with precursors (particularly oleate) for lipid synthesis could stimulate apoB secretion. However, no manipulation has been shown to preclude degradation and permit secretion of all of the nascent apoB. It is conceivable that the inability of hepatoma cell lines to fully secrete apoB results from an inadequate lipid supply. Thus, we predicted that restriction of lipid availability in the transfected cells would result in a constant output of rapoB. However, the current studies demonstrate that under defined metabolic conditions (ie, serum-free medium), the level of secretion of hapoB is to a large extent in proportion to the level of hapoB synthesis. Because the buoyant densities of the secreted lipoproteins containing hapoB53 or hapoB100 were not altered (Figs 2 & 7), we could exclude the possibility that the high expressors reduced the amount of lipids per lipoprotein particle to accommodate the increased number of apoB molecules. Likewise, since the secretion of rapoB100 (Figs 1D⇑ and 7B⇑) was unchanged relative to expression levels of the human proteins, we ruled out the possibility that the increased secretion of hapoB-containing lipoproteins is achieved at the expense of rapoB100 secretion (discussed in more detail below). Instead, results of the [3H]glycerol incorporation experiments suggest that triglyceride synthesis in cells overexpressing hapoB53 (Figs 3A⇑ and 4⇑ and the Table⇑) or hapoB100 (Fig 9⇑) is increased. An increase in triglyceride synthesis, however, was not matched by a corresponding increase in triglyceride levels in cells expressing high levels of hapoB53, and in fact there was a 25% reduction in triglyceride level in these cells relative to the low hapoB53 expressor (see the Table⇑). This reduction may be in part attributable to mobilization of triglyceride stores for lipoprotein assembly and secretion when hapoB53 is overexpressed. However, since the rate of [3H]glycerol uptake by the cells and the pool size of intermediates such as glycerol-3-phosphate were not determined, we cannot conclude with certainty that hapoB overexpression results in upregulated triglyceride synthesis. The interrelation between hepatic apoB synthesis and lipogenesis has not been noted previously and merits further study.
The positive correlation between apoB mRNA levels and apoB synthesis suggests that there might be a cause and effect relation. For instance, the steady state levels of the hapoB53 mRNA, cell hapoB53 protein, and medium hapoB53 protein in the high expressor cells (B53-1) were fourfold, 2.7-fold, and threefold higher, respectively, than in the intermediate expressor cells (B53-14) (Fig 1⇑), which was accompanied by an eightfold increase in hapoB53 synthesis (Fig 5⇑). Because synthesis of apoB occurs on the ribosome bound to the rough endoplasmic reticulum, the increased apoB mRNA concentrations may stimulate apoB synthesis by increasing the number of apoB mRNA–ribosome complexes bound to the translation and translocation channels on the endoplasmic reticulum. As recently suggested by Chen et al,42 who used a “square lattice model,” the rate of protein synthesis involving cotranslational translocation across the endoplasmic reticulum membrane (such as apoB) is determined by the concentration of mRNA. It is noteworthy, however, that the greatly increased apoB synthesis (eightfold for hapoB53 and threefold for apoB100) in the transfected cells did not result in more than a twofold increase in apoB secretion efficiency (Figs 5⇑ and 8⇑), indicating that a significant proportion of the newly synthesized apoB is still degraded posttranslationally.
Overexpression of hapoB48 in McA-RH7777 cells has been shown to decrease the synthesis and secretion of endogenous rapoB100.29 In the present study, a reduction in steady state levels of secreted rapoB100 was observed in hapoB53-expressing cells. The reason for this reduction is not entirely clear. Expression of hapoB53, unlike hapoB48,29 did not seem to influence synthesis of rapoB100, neither was the reduction in rapoB100 secretion dependent on the levels of expression of hapoB53. The effect could not be attributable to an impairment in general metabolism because the steady state level of total protein (including apoA-I and apoE) was not affected. Neither could the reduction be explained by a decrease in overall protein secretion because the secretion efficiency of rat apoA-I observed in the hapoB53-expressing cells was comparable with that of nontransfected cells. The present data suggest, however, that expression of hapoB53 results in decreased secretion efficiency of rapoB100 (24% in McA-RH7777 and 7% to 17% in the transfected cells). In comparison with hapoB48- and hapoB53-transfected cells, such a reduction in steady state levels of rapoB100 secretion did not occur in hapoB100-transfected cells. The lack of an effect of apoB expression on endogenous rapoB100 secretion was also observed with cells expressing other large hapoB forms such as hapoB72 and hapoB64 (S. Selby and Z. Yao, unpublished observation, 1995). It is unlikely that the differential effect of varying hapoB expression on endogenous rapoB100 synthesis is due to the level of hapoB expression in the cells, because the decreased rapoB100 secretion was observed in clone B53-10, which produced very low levels of hapoB53 (Fig 1B⇑), but was not observed in clone B100-18, which produced very high levels of hapoB100 (Fig 7A⇑). An alternative explanation is that the inhibitory effect on rapoB100 secretion is a function of the apoB length, much like that seen with the buoyant density31 and secretion kinetics1 of the truncated hapoB-containing lipoproteins. The effect of truncated hapoB expression on full-length rapoB100 synthesis in McA-RH7777 cells needs to be further determined.
Although our data from the in vitro transfection experiments suggest that vastly different apoB mRNA levels may have a profound effect on apoB synthesis and secretion, our model does not represent the in vivo situation. At the present time, under no metabolic conditions have the hepatic apoB mRNA levels been shown to vary by >50%. However, our in vitro data may provide explanations for the in vivo observations from transgenic mice that harbor different copy numbers of the apoB transgene. In transgenic mice that express hapoB100 specifically in the liver, there is a positive correlation between the transgene copy number and plasma LDL apoB100 concentration.27 28 In these animals, overexpression of hapoB100 does not result in a decrease in mouse apoB in the plasma.27 In addition to possibilities such as decreased catabolism or uptake of apoB, the increased plasma levels of the recombinant hapoB100 in the transgenic mice may be attributable to an increased hepatic secretion. The present data obtained from in vitro transfection studies are compatible with the transgenic mice studies. Furthermore, our data strongly suggest that an increased apoB gene expression can indeed result in an enhanced apoB secretion and that the level of apoB secretion may not be passively controlled by lipid availability.
Selected Abbreviations and Acronyms
|cpm||=||counts per minute|
|DMEM||=||Dulbecco’s modified Eagle’s medium|
|hapoB||=||human apolipoprotein B|
|HPTLC||=||high-performance thin-layer chromatography|
|rapoB||=||rat apolipoprotein B|
|SDS-PAGE||=||SDS–polyacrylamide gel electrophoresis|
This work was supported by the Medical Research Council of Canada (grant MT-11559). Dr Yao is a Research Scholar of the Heart and Stroke Foundation of Canada. We wish to thank A. Brooks for performing the RNase protection assay, Y. Zhao for expert technical assistance, R.S. McLeod and H.J. Verkade for support during the early stage of the project, and J.E. Vance for a critical reading of the manuscript. We also thank T. Ohnishi, S. Yokoyama, R. Milne, Y. Marcel, and R. Davis for antibodies.
Chan L. Apolipoprotein B, the major protein component of triglyceride-rich and low density lipoproteins. J Biol Chem.. 1992;267:25621-25624.
Pullinger CR, North JD, Teng BB, Rifici VA, Rifici VA, de Brito AER, Scott J. The apolipoprotein B gene is constitutively expressed in HepG2 cells: regulation of secretion by oleic acid, albumin, and insulin, and measurement of the mRNA half-life. J Lipid Res.. 1989;30:1065-1077.
Dashti N, Williams DL, Alaupovic P. Effects of oleate and insulin on the production rates and cellular mRNA concentrations of apolipoproteins in HepG2 cells. J Lipid Res.. 1989;30:1365-1372.
Sorci-Thomas M, Hendricks CL, Kearns MW. HepG2 cell LDL receptor activity and the accumulation of apolipoprotein B and E in response to docosahexanoic acid and cholesterol. J Lipid Res.. 1992;33:1147-1156.
Kushwaha RS, McMahan A, Mott GE, Carey KD, Reardon CA, Getz GS, McGill HC. Influence of dietary lipids on hepatic mRNA levels of proteins regulating plasma lipoproteins in baboons with high and low levels of large high density lipoproteins. J Lipid Res.. 1991;32:1929-1940.
Brooks AR, Nagy BP, Taylor S, Simonet WS, Taylor JM, Levy-Wilson B. Sequences containing the second-intron enhancer are essential for transcription of the human apolipoprotein B gene in the livers of transgenic mice. Mol Cell Biol.. 1994;14:2243-2256.
Paulweber B, Sandhofer F, Levy-Wilson B. The mechanism by which the human apolipoprotein B gene reducer operates involves blocking of transcriptional activation by hepatocyte nuclear factor 3. J Biol Chem.. 1993;266:21956-21961.
Kardassis D, Zannis VI, Cladaras C. Organization of the regulatory elements and nuclear activities participating in the transcription of the human apolipoprotein B gene. J Biol Chem.. 1992;267:7956.
Dixon J, Ginsberg HN. Regulation of hepatic secretion of apolipoprotein B-containing lipoproteins: information obtained from cultured liver cells. J Lipid Res.. 1993;34:167-179.
Borén J, Rustaeus S, Wettesten M, Andersson M, Wiklund A, Olofsson S-O. Influence of triacylglycerol biosynthesis rate on the assembly of apoB-100–containing lipoproteins in Hep G2 cells. Arterioscler Thromb.. 1993;13:1743-1754.
Craig WY, Nutik R, Cooper AD. Regulation of apoprotein synthesis and secretion in the human hepatoma HepG2: the effect of exogenous lipoprotein. J Biol Chem.. 1988;263:13800-13890.
Dashti N. The effect of low density lipoproteins, cholesterol, and 25-hydroxycholesterol on apolipoprotein B gene expression in HepG2 cells. J Biol Chem.. 1992;267:7160-7169.
Cianflone KM, Yasruel Z, Rodriguez MA, Vas D, Sniderman AD. Regulation of apoB secretion from HepG2 cells: evidence for a critical role for cholesterol ester synthesis in the response to a fatty acid challenge. J Lipid Res.. 1990;31:2045-2055.
Yao Z, Vance DE. The active synthesis of phosphatidylcholine is required for very low density lipoprotein secretion from rat hepatocytes. J Biol Chem.. 1988;263:2998-3004.
Zhang Z, Sniderman AD, Kalant D, Vu H, Monge JC, Tao Y, Cianflone K. The role of amino acids in apoB100 synthesis and catabolism in human HepG2 cells. J Biol Chem.. 1993;268:26920-26926.
Wu X, Sakata N, Dixon J, Ginsberg HN. Exogenous VLDL stimulates apolipoprotein B secretion from HepG2 cells by both pre- and post- translational mechanisms. J Lipid Res.. 1994;35:1200-1210.
Sparks JD, Zolfaghari R, Sparks CE, Smith HC, Fisher EA. Impaired hepatic apolipoprotein B and E translation in streptozotocin diabetic rats. J Clin Invest.. 1992;89:1418-1430.
Lackner KJ, Monge JC, Gregg RE, Hoeg JM, Triche TJ, Law SW, Brewer HB Jr. Analysis of the apolipoprotein B gene and messenger ribonucleic acid in abetalipoproteinemia. J Clin Invest.. 1986;78:1707-1712.
Wetterau JR, Aggerbeck LP, Bouma M-E, Eisenberg C, Munck A, Hermier M, Schmitz J, Gay G, Rader DJ, Gregg RE. Absence of microsomal triglyceride transfer protein in individuals with abetalipoproteinemia. Science.. 1992;258:999-1001.
Ross RS, Gregg RE, Law SW, Monge JC, Grant SM, Higuchi K, Triche TJ, Jefferson J, Brewer HB Jr. Homozygous hypobetalipoproteinemia: a disease distinct from abetalipoproteinemia at the molecular level. J Clin Invest.. 1988;81:590-595.
Homanics GE, Smith TJ, Zhang SH, Lee D, Young SG, Maeda N. Targeted modification of the apolipoprotein B gene results in hypobetalipoproteinemia and developmental abnormalities in mice. Proc Natl Acad Sci U S A.. 1993;90:2389-2393.
Farese RV, Flynn LM, Young SG. Modification of the apolipoprotein B gene in HepG2 cells by gene targeting. J Clin Invest.. 1992;90:256-261.
Linton MF, Farese RV, Chiesa G, Grass DS, Chin P, Hammer RE, Hobbs HH, Young SG. Transgenic mice expressing high plasma concentrations of human apolipoprotein B100 and lipoprotein(a). J Clin Invest.. 1993;92:3029-3037.
Callow MJ, Stoltzfus LJ, Lawn RM, Rubin EM. Expression of human apolipoprotein B and assembly of lipoprotein(a) in transgenic mice. Proc Natl Acad Sci U S A.. 1994;91:2130-2134.
Hussain MM, Zhao Y, Kancha RK, Blackhart BD, Yao Z. Characterization of recombinant human apoB48-containing lipoproteins in rat hepatoma McA-RH7777 cells transfected with apoB48 cDNA. Arterioscler Thromb Vasc Biol.. 1995;15:485-494.
Yao Z, Blackhart BD, Johnson DF, Taylor SM, Haubold KW, McCarthy BJ. Elimination of apolipoprotein B48 formation in rat hepatoma cell lines transfected with mutant human apolipoprotein B cDNA constructs. J Biol Chem.. 1992;267:9858-9867.
McLeod RS, Zhao Y, Selby S, Westerlund J, Yao Z. Carboxyl terminal truncation impairs lipid recruitment by apolipoprotein B100 but does not affect secretion of the truncated apolipoprotein B-containing lipoproteins. J Biol Chem.. 1994;269:2852-2862.
Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem.. 1951;193:265-275.
Yao Z, Blackhart BD, Linton MF, Taylor SM, Young SG, McCarthy BJ. Expression of carboxyl-terminally truncated forms of human apolipoprotein B in rat hepatoma cells. J Biol Chem.. 1991;266:3300-3308.
Yao Z, Lauer SJ, Sanan DA, Fazio S. Apo-AIV is secreted on discrete HDL particles by the rat hepatoma cell line McA-RH7777 transfected with apoA-IV cDNA. Arterioscler Thromb.. 1993;13:1476-1486.
Johnson DF, Innerarity TL. Human apolipoprotein B cryptic polyadenylation sites are activated in McA-RH7777 cells. Circulation. 1993;88(suppl I):I-132. Abstract.
Heinemann T, Metzger S, Fisher EA, Breslow JL, Huang L-S. Alternative polyadenylation of apolipoprotein B RNA is a major cause of B-48 protein formation in rat hepatoma cell lines transfected with human apoB-100 minigenes. J. Lipid Res.. 1994;35:2200-2211.
White A, Graham DL, LeGros J, Pease RJ, Scott J. Oleate-mediated stimulation of apolipoprotein B secretion from rat hepatoma cells: a function of the ability of apolipoprotein B to direct lipoprotein assembly and escape presecretory degradation. J Biol Chem.. 1992;267:15657-15664.
Spring DJ, Chen-Liu LW, Chatterton JE, Elovson J, Schumaker VN. Lipoprotein assembly: apolipoprotein B size determines lipoprotein core circumference. J Biol Chem. 1992;267: 14839-14845.
Chen L, Zhang J, Chan L. Apolipoprotein B synthesis: a square lattice model. J Lipid Res.. 1994;35:84-92.