Demonstration of Biphasic Effects of Docosahexaenoic Acid on Apolipoprotein B Secretion in HepG2 Cells
Abstract Oleic acid (OA) stimulates apolipoprotein B (apoB) secretion from HepG2 cells by protecting the nascent protein from rapid intracellular degradation. In contrast, the n-3 fatty acids, docosahexaenoic acid (DHA) and eicosapentaenoic acid, have been shown to reduce apoB secretion by increasing its intracellular degradation in rat hepatocytes. We attempted to determine if OA and DHA have these opposite effects at the same point in the secretory pathway for apoB or if they act at different points in HepG2 cells. Unexpectedly, we found that when DHA (0.2 mmol/L) was incubated with HepG2 cells for 2 hours, it stimulated both triglyceride (TG) synthesis and apoB secretion significantly (the “stimulatory effect”). The stimulatory effect of DHA on apoB secretion was associated with decreased intracellular degradation of newly synthesized apoB. These acute effects of DHA on TG synthesis and apoB secretion paralleled those previously demonstrated with OA. After DHA was removed from the medium, however, both TG synthesis and apoB secretion rapidly decreased to a level that was significantly less than the control level (the “inhibitory effect”). At the same time, intracellular apoB degradation was significantly increased, and this degradation was efficiently prevented by proteasome inhibitors. Removal of DHA from the incubation resulted in inhibition of the incorporation of endogenous fatty acids into TG. In contrast, removal of OA from the media was not associated with any such inhibitory effect. The inhibitory effect of DHA on basal apoB secretion persisted at least 8 hours. These studies suggest that incubation of HepG2 cells with DHA has biphasic effects on TG synthesis and apoB secretion: an initial stimulation of TG synthesis is followed by inhibition of TG synthesis and increased apoB degradation. Although the stimulatory effect of DHA is apparent during short incubations of HepG2 cells, both effects would be expected to occur during long incubations, since fatty acid uptake by cells is rapid and efficient. Thus, long incubations of HepG2 cells with DHA could result in overall reduced apoB secretion compared with cells incubated in bovine serum albumin. If these findings are extrapolated to the in vivo situation, they can explain the ability of dietary n-3 fatty acids, which would be delivered to the liver intermittently, to reduce very low density lipoprotein secretion.
- Received April 7, 1997.
- Accepted June 24, 1997.
N-3 fatty acids, EPA and DHA, decrease plasma TG levels in humans,1,2 and previous studies in isolated rat hepatocytes suggested that compared with OA, n-3 fatty acids are poor substrates for TG synthesis.3,4 This hypothesis was supported by evidence suggesting that n-3 fatty acids inhibited the activities of two enzymes in the TG synthesis pathway: diacylglycerol acyl transferase and phosphatidate phosphohydrolase.5,6 However, the results of other studies raised doubts about the proposed mechanism of action of n-3 fatty acids. Thus, works by Lang and Davis7 and Wang et al8 demonstrated that the n-3 fatty acids, DHA and EPA, were at least as potent as OA in their ability to acutely stimulate TG synthesis in rat hepatocytes. On the other hand, those studies and others indicated that these n-3 fatty acids decrease the secretion of apoB-containing lipoproteins from rat hepatocytes,7,8 rat liver,9 rat hepatoma cells,10 and HepG2 cells.11 Lang and Davis7 proposed that n-3 fatty acids, by unknown mechanisms, impaired the assembly and/or secretion of apoB-containing, TG-rich VLDL from rat hepatocytes, despite the acute stimulation of TG synthesis. These findings were confirmed by Wang et al8 who showed that n-3 fatty acids stimulated apoB degradation in rat hepatocytes. These latter studies7–11 support the findings that diets rich in fish oil suppress VLDL production in human.12
Our previous studies in HepG2 cells13,14 led us to hypothesize that newly synthesized TG, derived from exogenous OA or human plasma VLDL, is the critical core lipid that determines the secretion of apoB-containing lipoproteins from this cell line. This raised the following question: if n-3 fatty acids acutely stimulate TG synthesis,7,8 why do they not increase apoB-lipoprotein assembly and secretion? These studies were conducted in an attempt to answer this question.
[4,5-3H]leucine (135 Ci/mmol, catalog no. TRK.683), [2-3H]glycerol (1.0 Ci/mmol, catalog no.TRA.118), and [2-14C]acetic acid, sodium salt (57 mCi/mmol, catalog no. CFA.14) were purchased from Amersham Corp. Monospecific anti-human apoB antiserum was raised in a rabbit. Protein A-Sepharose CL-4B was obtained from Pharmacia LKB Biotechnology, Inc. Minimum essential medium , nonessential amino acids, sodium pyruvate, and penicillin/streptomycin were from Life Technologies. Fetal bovine serum was from Integen. Leucine-free medium was generated from a minimum essential selection kit (Life Technologies, catalog no. 300-9050AV). Leupeptin and pepstatin A were from Peninsula Laboratories, Inc. BSA, DHA (free acid), OA (sodium salt), and acetyl-leucyl-leucyl-norleucinal were from Sigma Chemical Co. Lactacystin was provided by ProScript, Inc. Triacsin D was from Fujisawa. All other chemicals were of the highest purity available.
Growth of Cells
HepG2 cells, obtained from American Type Culture Collection, were grown in 35-mm dishes in minimal essential medium containing 0.1 mmol/L nonessential amino acid, 1 mmol/L sodium pyruvate, penicillin (100 U/ml), streptomycin (100 μg/ml), and 10% fetal bovine serum. When cells were 90% confluent, the medium was changed to serum-free minimal essential experimental medium as described below.
Preparation of DHA Sodium Salt, DHA-BSA Complex, and OA-BSA Complex
Preparations were made according to the method of Miller et al15 with minor modifications. Briefly, 25 mg of DHA free acid was mixed with 0.6 mL of ethanol and 0.025 mL of 5 mol/L NaOH. The mixture was dried under gas nitrogen. After drying, it was dissolved in 3 mL of 150 mmol/L NaCl and incubated in a water bath at 60°C for 5 minutes. Four milliliters of 24% BSA was introduced into this solution and stirred at 4°C for 15 hours. The final concentration of DHA was 10.87 mmol/L complexed with 13.7% BSA. Before experiments, DHA was diluted into the experimental medium, and BSA was adjusted so that the ratio of DHA to BSA was 2:1. OA-BSA complexes were prepared according to Dixon et al.16
Determination of Secretion and Intracellular Degradation of ApoB
To determine apoB secretion, HepG2 cells were labeled with [3H]leucine in the presence of BSA alone, DHA (0.2 mmol/L) complexed with BSA, or OA (0.2 mmol/L) complexed with BSA for 2 hours. The labeling medium was collected, and cells were chased in serum-free medium for 40 minutes. The chase medium was then collected. In some experiments, all the cells were then labeled a second time with [3H]leucine in the presence of BSA alone (no fatty acids) for 2 hours and the labeling medium was collected. ApoB secretion into the medium was determined by immunoprecipitation.
Pulse-chase experiments were carried out to determine apoB degradation. HepG2 cells were preincubated with BSA alone, BSA/DHA, or BSA/OA for 1 hour, pulse-labeled with [3H]leucine for 10 minutes, and chased for various periods of time (10, 20, and 60 minutes). ApoB that had been secreted into the medium and apoB remaining in the cells were determined by immunoprecipitation.
Determination of Lipid Synthesis
To determine TG synthesis, HepG2 cells were incubated with [3H]glycerol for 2 hours. To determine synthesis of endogenous fatty acids, HepG2 cells were incubated with [14C]acetate for 2 hours. The medium was then removed, and the cells were extracted with hexane/isopropanol (3:2) for lipids. Thin-layer chromatography was used to separate lipid species. Lipid spots were counted in a scintillation counter. Total synthesis of endogenous fatty acids was defined as the sum of 14C incorporated into labeled TG, phosphatidylcholine, phosphatidylethanolamine, and free fatty acids.
Immunoprecipitation of apoB was carried out exactly according to the method of Dixon et al.16 Medium or cell lysate samples were mixed with NET buffer (150 mmol/L NaCl, 5 mmol/L EDTA, 50 mmol/L Tris-HCl, pH 7.4, 0.5% Triton X-100, and 0.1% sodium dodecyl sulfate) and an excess amount of anti-apoB antiserum, and the mixture was incubated at 4°C on a shaker for 10 hours. Protein A-Sepharose CL-4B was added to the mixture, the incubation was continued for another 3 hours, and the beads were washed extensively with NET buffer. ApoB was extracted from the protein A pellet with sample buffer by boiling for 4 minutes. An aliquot of sample was run on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The gel was treated with autofluor (National Diagnostics) and, after drying, was exposed to a film at −80°C.
Protein concentration was determined by the bicinchoninic acid method (Pierce Chemical Co). Total protein synthesis and secretion were determined with the combination of trichloroacetic acid precipitation and scintillation counting. The significance of difference between two groups was determined by Student’s t test. P<.05 was considered to be significant.
DHA Has Biphasic Effects on Secretion of ApoB From HepG2 Cells
When HepG2 cells were incubated for 2 hours with [3H]leucine in the presence of BSA (0.75%) alone, BSA complexed with DHA (0.2 mmol/L), or BSA complexed with OA (0.2 mmol/L), both DHA and OA stimulated apoB secretion 3-fold (Fig 1⇓, Label). However, when Hep G2 cells that had been treated with either BSA, OA, or DHA were chased in the serum-free medium with BSA for 40 minutes (fatty acids were absent in the chase medium), apoB secretion rates were strikingly different in the plates previously treated with fatty acids (Fig 1⇓, Chase). Cells that had been treated with OA during the initial 2 hours (−OA) continued to secrete more labeled apoB compared with control cells. In contrast, cells that had received DHA during the initial 2 hours (−DHA) secreted significantly less labeled apoB compared with control cells. Thus, when DHA was present in the incubation medium, it stimulated apoB secretion; when DHA was withdrawn from the incubation medium, its stimulatory effect on apoB secretion was rapidly converted to a inhibitory effect. Therefore, unlike OA, the effect of DHA on apoB secretion was absolutely dependent on its presence in the media.
Higher concentrations of DHA (0.4 or 0.8 mmol/L) or a higher DHA-BSA molar ratio (4:1) generated similar results (unpublished data). Additionally, in our system, 0.4 and 0.8 mmol/L DHA were found to be toxic to the cells as judged by total protein synthesis. Therefore, 0.2 mmol/L DHA and a 2:1 DHA:BSA molar ratio were used in most of the experiments presented in this study.
Stimulatory Effect of DHA on ApoB Secretion Inhibited by Triacsin D, an Inhibitor of TG Synthesis
The stimulatory effect of OA on apoB secretion is associated with the ability of OA to stimulate TG synthesis.13,14 To test whether DHA stimulates apoB secretion by the same mechanism, the following experiment was carried out. HepG2 cells were incubated for 2 hours with BSA alone, BSA plus DHA (0.1 and 0.2 mmol/L), or BSA plus OA (0.1 or 0.2 mmol/L) in the presence of [3H]leucine (100 μCi/ml) and in absence or presence of Triacsin D, an inhibitor of TG synthesis (12.5 μmol/l). Triacsin D inhibits fatty acyl CoA synthase, The medium was collected to determine apoB secretion. As shown in Fig 2⇓, DHA, similar to OA, significantly stimulated apoB secretion at either dose; Triacsin D significantly blocked the stimulatory effects of both DHA and OA on apoB secretion.
Inhibitory Effect of DHA on ApoB Secretion Is Persistent
In this experiment, we determined the duration of the inhibitory effect of DHA on apoB secretion. Hep G2 cells were labeled with [3H]leucine in the presence of BSA alone, BSA plus OA, or BSA plus DHA for 2 hours, the media were removed for determination of apoB secretion (Fig 3⇓, Label 1), and all cells were incubated for additional 8 hours in serum-free medium containing only BSA. After this, cells were labeled with [3H]leucine for 2 hours in the absence of fatty acids, and apoB secretion was determined (Fig 3⇓, Label 2). Both OA and DHA stimulated apoB secretion 4-fold during the first labeling period. During the second labeling period, cells that received OA during the original 2-hour labeling period (−OA) secreted amounts of labeled apoB that were identical to the levels of apoB secreted by cells that had originally received BSA; in contrast, cells that had received DHA during the initial 2-hour incubation period (−DHA) secreted significantly less apoB after an additional 8-hour incubation in the absence of any fatty acid compared with the cells that had initially received BSA. Thus, the inhibitory effect of DHA on apoB secretion that was observed immediately after removal of the fatty acid from the medium (Fig 1⇑) persisted for at least 8 hours; OA pretreatment was not associated with any inhibition of apoB secretion, although its stimulatory effects on apoB secretion disappeared by 40 minutes of chase without OA).14,16
Overall Secretion of ApoB From Cells Transiently Exposed to DHA Less Than From Control Cells
Since the inhibitory effect of DHA on apoB persisted for several hours, we hypothesized that the overall secretion of apoB from cells that had transiently received DHA would be lower than that from control cells. The following experiment was carried out to test this hypothesis.
HepG2 cells were labeled for 1 hour with [3H]leucine in the presence of BSA, and the labeling medium was collected and changed to fresh BSA-containing labeling medium every 1 hour for an additional 3 hours. A second set of cells were labeled with [3H]leucine in the presence of BSA/DHA for 1 hour. The medium was collected, and the DHA-treated cells were divided into two groups. The first group was labeled again in the presence of BSA/DHA and changed to fresh BSA/DHA-containing labeling medium every 1 hour for an additional 3 hours; the second group was labeled in the presence of BSA and changed to fresh BSA-containing labeling medium every 1 hour for an additional 3 hours. In this experiment, therefore, apoB secretion was determined hourly for 4 hours. Fig 4A⇓ shows the result. Under control conditions (BSA treatment), the rate of apoB secretion reached steady state during the 2nd hour; with DHA treatment and when DHA was continuously present, the rate of apoB secretion was accelerated and was at steady-state by 1 hour. However, if DHA was present only during the 1st hour but was not present (DHA/−DHA) during the following hours, the rate of apoB secretion was higher than that in control BSA-only cells during the 1st hour but lower than the control levels afterward. More importantly, total apoB secretion rates during the 4-hour incubation were significantly different between the groups (Fig 4B⇓) When DHA was persistently present, apoB secretion was significantly higher than the control level; however, if DHA was withdrawn after the initial 1-hour incubation, total apoB secretion over 4 hours was significantly lower than the control level.
Inhibitory Effect of DHA on ApoB Secretion Converted to Stimulatory Effect When DHA Added Back to Incubation Medium
When unperturbed HepG2 cells were initially exposed to DHA, apoB secretion was stimulated. This raised the following question: Would HepG2 cells still respond to the addition of DHA during the inhibitory phase induced by a prior exposure to DHA? The following experiment was carried out to answer this question. HepG2 cells were labeled with [3H]leucine in the presence of BSA alone or DHA complexed to BSA for 2 hours. and the medium was collected for determination of apoB secretion. All cells were then incubated with BSA alone for an additional 2 hours. After this, cells were labeled with [3H]leucine for 2 hours. DHA was added back to half of the cells that had received DHA during the initial labeling period. Medium was collected at the end of the second labeling period to determine apoB secretion. The results are depicted in Fig 5⇓. As reported in earlier experiments, DHA stimulated apoB secretion significantly during the initial 2-hour labeling period (Label 1); during the second 2-hour labeling period; however, apoB secretion in the presence of only BSA was significantly lower in the cells first treated with DHA (−DHA) compared with cells that received only BSA during the initial labeling period (Label 2). On the other hand, when DHA was present in the medium during the second labeling (Label 2; DHA), cells that had been initially incubated with DHA were stimulated to secrete apoB once again. Thus, exposure to new DHA could overcome the inhibitory effects of a prior DHA exposure.
Effect of DHA on TG Synthesis Is Biphasic
The inhibitory effects on apoB secretion seen after removal of DHA from the media suggested parallel effects on TG synthesis. Therefore, HepG2 cells were incubated with BSA, BSA/OA, or BSA/DHA for 2 hours in the presence of [3H]glycerol. A second group of HepG2 cells were preincubated with either BSA, BSA/OA, or BSA/DHA (but no labeled glycerol) for 2 hours, and then incubated with BSA-only media containing [3H]glycerol for an additional 2 hours. OA and DHA stimulated TG synthesis equally when present in the initial labeling medium (Fig 6A⇓). After OA was removed from the medium (−OA), TG synthesis in cells previously treated with OA was equal to that in cells incubated in BSA alone throughout (Fig 6B⇓). In contrast, cells that had been initially incubated in DHA (−DHA) synthesized significantly less TG during the second 2-hour period. These results indicate that the biphasic effects of DHA on apoB secretion are paralleled by biphasic effects on TG synthesis. Furthermore, this experiment confirms the very strong association between TG synthesis and apoB secretion in HepG2 cells.
Effect of DHA on Endogenous Synthesis of Fatty Acids and TG
The inhibitory effect associated with removal of DHA on TG synthesis shown in Fig 6⇑ raised two possibilities: (1) removal of DHA is associated with inhibition of the synthesis of endogenous fatty acids and (2)removal of DHA is associated with inhibition of the incorporation of endogenous fatty acids into TG. To examine these possibilities, the following experiment was carried out. HepG2 cells were labeled with [14C]acetate for 2 hours in the presence of BSA alone, BSA/OA, or BSA/DHA. A second group of cells were first preincubated with BSA alone, BSA/OA, or BSA/DHA for 2 hours (without [14C]acetate) and then incubated with BSA alone for 40 minutes. The cells were labeled with [14C]acetate for 2 hours in medium containing only BSA. After each labeling, lipids were extracted and separated with thin-layer chromatography. The results (Fig 7A⇓) indicate that when present in the medium, both DHA (65±11×104 versus 93±14×104 cpm/mg of protein, DHA versus BSA) and OA (32±6×104 versus 93±14×104 cpm/mg of protein, OA versus BSA) significantly decreased endogenous fatty acid synthesis compared with BSA, confirming the result of Gibbons et al.17 More importantly, after removal of DHA from the incubation media (Fig 7B⇓; −DHA), cells preincubated with DHA synthesized similar amounts of endogenous fatty acids compared with cells that had always received BSA alone. This result suggests strongly that inhibition of the synthesis of endogenous fatty acids is unlikely to be the reason for the decreased TG synthesis observed after removal of DHA from the incubation medium.
Despite the fact that DHA did not affect the total amount of endogenously synthesized fatty acid, the synthesis of TG from endogenous fatty acids might be impaired by the removal of DHA from the incubation medium. Fig 7C⇑ and 7D⇑ shows that in BSA-treated cells, 32±2.1% of endogenously synthesized fatty acids was incorporated into TG, when DHA was present in the incubation medium, 46±5.2% of endogenously synthesized fatty acids was incorporated into TG, and after DHA was removed from the incubation medium, only 18±2.0% of endogenously synthesized fatty acids was incorporated into TG. On the other hand, similar proportions of endogenously synthesized fatty acids were esterified into TG whether OA was present in (30±3.9%) or absent from (33±3.4%) the incubation medium. These results indicate that incorporation of fatty acids into TG was specifically inhibited after the removal of DHA from the incubation medium.
DHA Inhibition of ApoB Secretion Associated With Increased Intracellular Degradation
ApoB secretion is regulated mainly at a post-translational level: the rate of apoB secretion is determined mostly by a regulatable intracellular degradation process in a pre-Golgi compartment.18–24 We therefore carried out a series of experiments to determine if removal of DHA increased apoB degradation and, if so, where the site(s) of that degradation would be.
HepG2 cells were preincubated with BSA or BSA/DHA for 1 hour, pulse-labeled with [3H]leucine for 10 minutes, and chased up to 60 minutes. In the cells preincubated with DHA, the fatty acid was also present in the labeling medium but was either absent (−DHA) or present (DHA) in the chase medium. In cells treated with only BSA, newly synthesized apoB was rapidly degraded; thus, after chase for 60 minutes, only 21.0±1.9% (mean±SEM, n=4) of initial apoB was detected in the cells (Fig 8A⇓; BSA). When DHA was present in the pulse and chase medium, it protected intracellular apoB from the rapid degradation; 35.0±2.5% (mean±SEM, n=4) of initially labeled apoB was detected in the cells after a 60-minute chase (Fig 8A⇓; DHA). When DHA was absent only from the chase medium, more apoB degradation was observed compared with cells treated with only BSA; thus 13.0±1.1% (mean±SEM, n=4) of initial apoB was detected in the cells at the end of the chase (Fig 8A⇓; −DHA). These results suggest that the biphasic effect of DHA on apoB secretion resulted from its effect on intracellular degradation of apoB rather than on a redistribution of apoB between cells and medium.
Although the endoplasmic reticulum and the cytoplasm have been identified as sites for degradation of newly synthesized apoB in HepG2 cells,18–24 apoB degradation in rat hepatocytes, which is stimulated by n-3 fatty acid,8 appears to occur in a post-endoplasmic reticulum compartment.25 Therefore, we sought to examine the possibility that DHA-induced apoB degradation could occur in the Golgi or lysosomal compartments in HepG2 cells. In the first experiment, HepG2 cells were labeled with [3H]leucine for 1 hour in the presence of monensin, which inhibits the transport from the Golgi to the plasma membrane, plus one of three agents (BSA, BSA/DHA, or BSA/OA), and chased in the presence of monensin plus the same agents for 1 hour to allow newly synthesized apoB to reach the Golgi compartment. After this, cells that received BSA were incubated with BSA-containing medium for an additional 1 hour. Cells that received fatty acids (either OA or DHA) were divided into two groups: the first group was incubated in the appropriate BSA/fatty acid-containing medium (DHA and OA) for another 1 hour, and the second group was changed to BSA-only medium (−DHA and −OA) for 1 hour. Monensin was present during all incubations, and no detectable secretion occurred (data not shown). If DHA induced apoB degradation in the Golgi compartment, we would detect less apoB when DHA was removed from the incubation medium during the last hour. The result shown in Fig 8B⇑, however, suggested that once apoB entered the Golgi compartment, it was stable whether or not DHA was present in the incubation medium.
Next we examined the possibility that DHA-induced apoB degradation occurs in the lysosomal compartment. HepG2 cells were labeled with [3H]leucine for 2 hours in the presence of BSA or BSA/DHA. BSA treated cells were chased in serum-free medium for 1 hour; DHA-treated cells were divided into three groups: the first group was chased in DHA-containing medium; the second group was chased in serum-free medium; and the third group was chased in serum-free medium to which 20 mmol/L NH4Cl was added. ApoB secreted into the medium and remaining in the cells was determined. Fig 8C⇑ (top) shows that DHA stimulated apoB secretion during the labeling period as shown in previous experiments. During the chase (Fig 8C⇑, middle), DHA continued to stimulate apoB secretion; when DHA was removed from the chase medium (−DHA), apoB secretion was less than in control cells. The addition of NH4Cl to the chase medium from which DHA was removed (−DHA/+NH4Cl) did not increase apoB secretion. A similar pattern of results was observed for intracellular apoB (Fig 8C⇑, bottom). These results suggest that DHA-induced apoB degradation was not associated with the lysosomal compartment.
Recently, the cytosolic ubiquitin-proteasome pathway has been reported to be responsible for apoB degradation in HepG2 cells.18,23 Since the results described above indicated that the increased degradation of apoB observed after removal of DHA was not associated with either the Golgi or lysosomal compartments, we carried out experiments to determine if this increased apoB degradation was associated with the proteasome pathway.
HepG2 cells were preincubated with BSA plus lactacystin or DHA plus lactacystin for 1 hour, pulse-labeled for 10 minutes, and chased for up to 60 minutes. Lactacystin was present in both the labeling medium and the chase medium. DHA was present in the labeling medium but was either present (DHA) or absent (−DHA) in the chase medium. The result is shown in Fig 8D⇑. Lactacystin efficiently protected apoB from degradation in all the cells: 68.0±3.6%, 73.0±6.2%, and 59.0±6.3% (mean±SEM, n=4) of initial apoB was detected in the cells treated with lactacystin/BSA, lactacystin/DHA, and lactacystin/−DHA, respectively. These protective effects are significant in comparison with that observed in the absent of lactacystin (shown in Fig 8A⇑) and suggest that apoB degradation, including that induced by the removal of DHA from the chase medium, was associated, at least in part, with the proteasome pathway.
Dietary intake of n-3 fatty acids has long been known to reduce plasma TG levels.1,2 Many studies3–12 have tried to elucidate the underlying mechanisms; in general, investigators have found that n-3 fatty acids have the ability to decrease the secretion of apoB-containing lipoproteins from hepatocytes. Those studies seemed to be consistent with demonstrated inhibitory effects of n-3 fatty acids on TG synthesis3–6,9 and provided a logical explanation for the beneficial effect of n-3 fatty acids in vivo. The present studies, however, suggest that the effects of the n-3 fatty acid DHA on the secretion of apoB-containing lipoproteins may be more complex than appreciated.
We demonstrated that when present in the medium, DHA stimulated both TG synthesis and apoB secretion from HepG2 cells. The discrepancy between our present results and those from some earlier studies5,6,9 cannot be explained fully at the present time. In those studies, DHA was found to inhibit phosphatidate phosphohydrolase and diacylglycerol acyl transferase, thereby inhibiting TG synthesis. Several other studies have, however, shown that n-3 fatty acids are at least as potent as OA in stimulating TG synthesis.7,8,26,27 On the other hand, despite the stimulation of TG synthesis by n-3 fatty acids seen in those more recent studies, decreased apoB secretion was observed in two of the three studies in which it was measured.7,8 The presence of DHA in the media and the concomitant ongoing synthesis of TG may be the keys to the difference between our results and those of Lang and Davis7 and Wang et al.8 The uptake of fatty acids by hepatocytes has been reported to be very rapid: the vast majority of fatty acids in the medium is taken up by cells within the first few hours of incubation.28 Investigators7,8 who did not observe a stimulatory effect of DHA on apoB secretion (despite stimulation of TG synthesis) used study designs that included long periods of incubation during which it was likely that no significant quantity of DHA remained in the media for much of those incubations. For example, in the studies by Davis et al7 rat hepatocytes were incubated with a single dose of DHA for 18 hours, and the effect of DHA on apoB secretion was observed over this entire period of time. The apoB secretion observed in that study, therefore, represented the overall effect of DHA: secretion may have been increased early but decreased (compared with control cells) for the majority of the incubation time, as we showed in Figs 1⇑ and 3⇑. We used short incubation protocols (1 or 2 hours) in most experiments so that media DHA concentrations were maintained at high levels for a significant fraction of the incubations. Indeed, an acute stimulatory effect of EPA, another n-3 fatty acid, on apoB secretion from HepG2 cells was demonstrated by Kurokawa et al26 using similar incubation protocols. In another study, Wong et al11 demonstrated that EPA and DHA decreased apoB secretion even during short-term incubation (3 hours) from HepG2 cells. However, in those studies OA failed to stimulate apoB secretion. We also cannot exclude the possibility that the differences between the present investigations and those earlier studies reflect cell-specific differences between HepG2 cells and rat hepatocytes. Indeed, it has been reported that the large cytoplasmic TG pool, which is very inactive in HepG2 cells,17,29 turns over rapidly and contributes greatly to VLDL-TG synthesis in rat hepatocytes.28,30,31
A novel finding of the present study was the biphasic effect of DHA on the secretion of apoB-lipoproteins from HepG2 cells. We demonstrated that the acute, stimulatory effect of DHA on apoB secretion from HepG2 cells is rapidly converted to an inhibitory effect after removal of the n-3 fatty acid from the medium. This late inhibitory effect was not observed with OA. The inhibitory effect of DHA on apoB secretion persisted for at least 8 hours. Furthermore, inhibition of apoB secretion could be rapidly converted to stimulation with the addition of fresh, DHA-containing medium. In preliminary experiments, we found that EPA had biphasic effects on apoB secretion from HepG2 cells (data not shown).
There are at least three potential explanations for the biphasic effect of DHA on apoB secretion; all of them derive from our observation that DHA has a biphasic effect on TG synthesis as well. First, DHA may inhibit the synthesis of endogenous fatty acids. Inhibition of the synthesis of endogenous fatty acids by OA has been reported in a recent study.17 Our results (Fig 7A⇑ and 7B⇑) suggest, however, that this is unlikely as the basis for DHA-induced apoB degradation. Second, when present in the medium (and being actively taken up by the cells), DHA might induce enzymes in the β oxidation pathway. The basal level of β oxidation in HepG2 cells is reportedly very low,17 and stimulation by n-3 fatty acids might drive a greater portion of endogenous fatty acids toward β oxidation and away from TG synthesis after DHA was removed from the medium. This could reduce basal substrate availability for TG synthesis. Our results suggest that this is also unlikely since the total amount of newly synthesized fatty acids that accumulated in the cells after DHA was removed from the incubation was not different from control cells. Third, DHA may have dual effect on enzymes in the TG synthesis pathway, such as diacylglycerol acyl transferase and phosphatidate phosphohydrolase. Being a substrate, DHA could activate these enzymes as it was being actively taken up into the cells. On the other hand, DHA might directly inhibit these enzymes,5,6 with the inhibitory effect becoming dominant when DHA’s role as a substrate diminished after it was removed from the medium. Our results provide some indirect evidence in support of this possibility. Thus, less fatty acid was incorporated into TG when DHA was removed from the medium in cells previously incubated with that n-3 fatty acid (Fig 7C⇑ and 7D⇑).
ApoB degradation in rat hepatocytes was reported to occur in the Golgi compartment25 and is activated by DHA and EPA.8 Our results suggested that the DHA-induced apoB degradation in HepG2 cells does not occur in the Golgi compartment. In addition, lysosomes are also unlikely as the site for the DHA-activated degradation of apoB. Rather, our results suggest that proteasomes are important for DHA-induced apoB degradation in HepG2 cells. This is consistent with the fact that these studies also demonstrated that DHA-derived TG, similar to OA-derived TG, plays a critical role in the regulation of apoB secretion from HepG2 cells. The effect of DHA on TG synthesis closely paralleled its effect on apoB secretion. When HepG2 cells were coincubated with DHA and Triacsin D, a potent inhibitor of fatty acyl-CoA synthase, DHA failed to stimulate apoB secretion. These results, along with our prior findings,13,14 support our hypothesis that increased TG availability, by increasing the interaction between microsomal transfer protein and newly synthesized apoB,32 targets apoB for secretion and prevents its degradation by the ubiquitin-proteasome pathway.
In summary, the present studies demonstrate that the effect of DHA on apoB secretion from HepG2 cells is biphasic: DHA stimulates apoB secretion when present continuously in the medium; this effect is similar to that observed with OA. Withdrawal of DHA from the medium is associated rapidly with inhibition of apoB secretion; this effect is not observed with OA. The overall effect of DHA on apoB secretion may depend on the relative periods of exposure of cells to media with DHA and media from which DHA has been withdrawn (or depleted). The chronic effect of dietary intake of n-3 fatty acids seen in vivo in humans and other species probably reflects n-3 fatty acid-induced inhibition of hepatic TG synthesis during postabsorptive periods, particularly the long overnight period, when endogenous fatty acids are the key substrate. Our results raise the possibility that dietary supplementation with n-3 fatty acids might be most effective if provided once daily, rather than in divided doses.
Selected Abbreviations and Acronyms
|BSA||=||bovine serum albumin|
This work was supported by grants HL36000, HL21007, and HL55638 from the National Institutes of Health.
Harris WS. Fish oils and plasma lipid and lipoprotein metabolism in humans: a critical review. J Lipid Res. 1989;30:785–807.
Rustan AC, Nossen JO, Christiansen EN, Drevon CA. Eicosapentaenoic acid reduces hepatic synthesis and secretion of triacylglycerol by decreasing the activity of acyl CoA:1,2 diacyl glycerol acyltransferase. J Lipid Res. 1988;29:1417–1426.
Lang CA, Davis RA. Fish oil fatty acids impair VLDL assembly and/or secretion by cultured rat hepatocytes. J Lipid Res. 1990;31:2079–2086.
Wang HX, Chen X, Fisher EA. N-3 fatty acids stimulate intracellular degradation of apoprotein B in rat hepatocytes. J Clin Invest. 1993;91:1380–1389.
Fox JC, Hay RV. Inhibition of triglyceride secretion by N-3 polyunsatured fatty acid in cultured rat hepatoma cells. Arteriosclerosis. 1987;7:495a. Abstract.
Wong SH, Fisher EA, Marsh JB. Effects of eicosapentaenoic and docosahexaenoic acids on apoprotein B mRNA and secretion of very low density lipoprotein in HepG2 cells. Arteriosclerosis. 1989;9:836–841.
Nestel PJ, Connor WE, Reardon MF, Connor S, Wong S, Boston R. Suppression by diets rich in fish oil of very low density lipoprotein production in man. J Clin Invest. 1984;74:82–89.
Wu X, Sakata N, Lui E, Ginsberg, HN. Evidence for a lack of regulation of the assembly and secretion of apolipoprotein B-containing lipoprotein from HepG2 cells by cholesteryl ester. J Biol Chem. 1994;269:12375–12382.
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.
Dixon J, Furukawa S, Ginsberg HN. Oleate stimulate secretion of apoB-containing lipoproteins from HepG2 cells by inhibiting early intracellular degradation of apoB. J Biol Chem. 1991;266:5080–5086.
Gibbons GF, Khurana R, Odwell A, Seelaender MCL. Lipid imbalance in HepG2 cells: active synthesis and impaired mobilization. J Lipid Res. 1994;35:1801–1808.
Furukawa S, Sakata N, Ginsberg HN, Dixon JL. Studies of the sites of intracellular degradation of apolipoprotein B in HepG2 cells. J Biol Chem. 1992;267:22630–22638.
Sato R, Imanaka T, Takatsuki A, Takano T. Degradation of newly synthesized apoB100 in a pre-Golgi compartment. J Biol Chem. 1990;265:11880–11884.
Adeli K. Regulated degradation of apolipoprotein B in semipermeable HepG2 cells. J Biol Chem. 1994;269:9166–9175.
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.
Fisher EA, Zhou M, Mitchell DM, Wu X, Omura S, Wang H, Goldberg AL, Ginsberg HN. The degradation of apolipoprotein B100 is mediated by a ubiquitin-proteasome pathway and involves heat shock protein 70. J Biol Chem. 1997;272:20427–20434.
Wu X, Sakata N, Lele KM, Zhou M, Ginsberg HN. A two-site model for apoB degradation in HepG2 cells. J Biol Chem. 1997;272:11575–11580.
Wang CN, Hobman TC, Brindley DN. Degradation of apolipoprotein B in cultured rat hepatocytes occurs in a post-endoplasmic reticulum compartment. J Biol Chem. 1995;270:24924–24931.
Homan R, Grossman JE, Pownall HJ. Differential effects of eicosapentaenoic acid and oleic acid on lipid synthesis and secretion by HepG2 cells. J Lipid Res. 1991;32:231–241.
Wu X, Shang A, Jiang H, Ginsberg HN. Low rates of apoB secretion from HepG2 cells result from reduced delivery of newly synthesized triglyceride to a “secretion-coupled” pool. J Lipid Res. 1996;37:1198–1206.
Chao FF, Stiers DL, Ontko JA. Hepatocellular triglyceride synthesis and transfer to lipid droplets and nascent very low density lipoproteins. J Lipid Res. 1986;27:1174–1181.
Yang LY, Kuksis A, Myher JJ, Steiner G. Origin of triglyceride moiety of plasma very low density lipoproteins in the rat: structural studies. J Lipid Res. 1995;36:125–136.
Wu X, Zhou M, Huang LS, Wetterau J, Ginsberg HN. Demonstration of a physical interaction between microsomal triglyceride transfer protein and apolipoprotein B during the assembly of apoB-containing lipoproteins. J Biol Chem. 1996;271:10277–10281.