The Assembly of Hepatic Very Low Density Lipoproteins: Evidence of a Role for the Golgi Apparatus Klara Valyi-Nagy, Carla Harris, and Larry L. Swift* Department of Pathology, Vanderbilt University School of Medicine, Nashville, Tennessee 37232-2561

ABSTRACT: Previous studies from our laboratory have suggested that the assembly of lipoproteins by the liver is not completed in the rough endoplasmic reticulum but continues while the particles are en route to or within the Golgi apparatus. To investigate further the role of the Golgi apparatus in lipoprotein assembly, mice were injected with [3H]glycerol and killed 7.5 to 45 min after injection. Microsomes and Golgi apparatus-rich fractions were isolated from the livers and separated into membrane and content fractions. TG within microsomal and Golgi membranes were labeled, rapidly reaching peak specific activity within 7.5 min of isotope injection. The specific activity of TG in microsomal membranes decreased to approximately 40% of peak values by 45 min, whereas the specific activity of TG in the Golgi membranes decreased to approximately 30% of peak values by 45 min. To determine whether the turnover of the Golgi membrane TG pool was dependent on microsomal TG transfer protein (MTP), mice were gavaged with an MTP inhibitor, and the labeling experiments were repeated. Inhibition of MTP attenuated the turnover of newly synthesized Golgi membrane TG by approximately 50% and the turnover of microsomal membrane TG by approximately 40%. Based on the rapid turnover of the Golgi membrane TG pool and the attenuation of the turnover of this pool by MTP inhibitor, we propose that lipid is transferred from the Golgi membrane to luminal lipoproteins in an MTP-dependent manner. The results support our hypothesis that assembly of VLDL continues within the Golgi apparatus. Paper no. L8950 in Lipids 37, 879–884 (September 2002).

A number of studies have suggested that the assembly of VLDL by the liver occurs via a two-step process (1–3). In this model, lipoprotein assembly is initiated in the rough endoplasmic reticulum (ER) with the synthesis of apolipoprotein B (apoB). The first step involves the association of apoB with phospholipids and core material, primarily cholesterol ester, to form a small HDL-like particle. In the second step additional core material, primarily TG, is added to the HDL-like particle to form a VLDL. The two-step model implies that assembly is completed within the ER of the cell. However, considerable evidence

*To whom correspondence should be addressed at Department of Pathology, C 3321 Medical Center North, Vanderbilt University School of Medicine, Nashville, TN 37232-2561. E-mail: [email protected] Abbreviations: apoB, apolipoprotein B; DGAT, DAG acyltransferase; ER, endoplasmic reticulum; MTP, microsomal TG transfer protein. Copyright © 2002 by AOCS Press

suggests that the Golgi apparatus plays more than a passive role in the assembly process (4–7). For example, Bamberger and Lane (4), in studies with chick hepatocytes, concluded that assembly of TG with apoproteins occurs in the Golgi. Likewise Higgins’ studies in rat liver suggested that the transGolgi regions appeared to be the major intracellular sites of assembly of apoB with TG (5). More recent studies by Stillemark et al. (8) suggest that assembly of apoB-48-containing VLDL is not completed in the rough ER. Pulse-chase studies combined with subcellular fractionation showed that apoB48 VLDL did not accumulate in the rough ER, indicating either that completed apoB-48 VLDL was rapidly transferred out of the ER or that assembly steps occurred post-ER. Previous studies in our laboratory suggest a critical role for the Golgi apparatus in the assembly process. By comparing the composition of nascent VLDL (d < 1.006 g/mL) recovered from the rough ER with VLDL recovered from the Golgi apparatus-rich fractions of rat livers, we concluded that an additional assembly step occurs while the particles are en route to the Golgi apparatus or within the Golgi apparatus (9). We estimated that up to 50% of the total VLDL TG and 30–40% of the phospholipid are added after the particle leaves the rough ER. The source of the lipid and the mechanism by which lipid is added to the particles is unknown. More recent studies in our laboratory (10) have documented size heterogeneity of nascent VLDL within the Golgi apparatus of mouse liver. We reported that approximately 50% of the apoB within the lumen of mouse liver Golgi apparatus was associated with particles recovered in the d 1.006–1.210 g/mL fraction. Furthermore, using radioisotope tracer studies we found that more than 80% of the newly synthesized apoB found in the serum was recovered with the VLDL fraction. This suggested that lipid is added to the nascent particles within the Golgi apparatus before secretion. This study was undertaken to investigate further the role of the Golgi apparatus in the assembly of VLDL by the liver. In vivo lipid labeling experiments with [3H]glycerol have led to the identification of a pool of newly synthesized TG within hepatic Golgi membranes. This pool, distinct from the pools in the rough and smooth ER, has a rapid turnover rate. Furthermore, the turnover of this pool is attenuated in the presence of a microsomal TG transfer protein (MTP) inhibitor. We hypothesize that this pool of TG is derived from the ER membranes and is used in the lipidation of VLDL within the Golgi apparatus in an MTP-dependent process.

879

Lipids, Vol. 37, no. 9 (2002)

880

K. VALYI-NAGY ET AL.

EXPERIMENTAL PROCEDURES Animals. Male ICR mice (approximately 30 g) were purchased from Harlan Industries (Indianapolis, IN). The animals were maintained in the Vanderbilt University Animal Care facility on food (Wayne Lab Blox; Allied Mills, Inc., Chicago, IL) and water ad libitum for at least 5 but no longer than 10 d prior to the experiments. All procedures using mice were approved by the Animal Care and Use Committee at Vanderbilt University. Isolation of subcellular fractions. Golgi apparatus-rich fractions were isolated from the mouse livers by the method of Swift et al. (10). Livers (usually three, with a total mass of 5–6 g) were minced finely with scalpels, placed in homogenizing buffer (1.5 mL/g) consisting of 0.1 M phosphate buffer, pH 7.3; 0.25 M sucrose; 1% dextran; and 0.01 M MgCl2 and homogenized at a setting of 45 for 20 s using a Virtishear homogenizer. The homogenate from three livers was placed in one tube and centrifuged in an SW40 rotor at 900 × g for 10 min, at which point the speed was increased to 6,000 × g and the samples centrifuged an additional 30 min. The supernatant was discarded except when used for the isolation of microsomes (see below). Homogenizing buffer (1.7 mL) was then added to the pellet. The top one-third of the pellet was gently dislodged from the remainder of the pellet, resuspended, and layered on 8 mL of 1.2 M sucrose. The samples were centrifuged in the SW40 rotor at 3,300 × g for 10 min, after which the speed was increased to 13,000 × g for 10 min and finally increased to 83,000 × g for 45 min. The material banding on the 1.2 M sucrose pad was carefully removed, diluted with ice-cold distilled water, and pelleted using the SW40 rotor (6,000 × g for 30 min). The pellet was resuspended in 15 mM Tris and 0.154 M NaCl, pH 7.4, and the content and membrane fractions were recovered by sodium carbonate treatment followed by centrifugation as described below. Microsomes were isolated by centrifugation of the first supernatant at 114,000 × g for 60 min using an SW41 rotor. The pellet was washed by resuspending it in PBS; it was then pelleted under the same conditions. The final microsomal pellet was resuspended in PBS, quick-frozen, and stored at −80°C. Cytosolic fat was defined as the fat that floated during the initial isolation of microsomes. It was recovered from the top of the tube after tube slicing and washed in PBS under the same conditions as used for isolation. Separation of membrane and luminal content fractions. Golgi and microsomal fractions were incubated on ice with an equal volume of 0.2 M sodium carbonate, pH 11.0, for 60 min (11). The final concentration of protein in the sample was approximately 0.5–1.0 mg/mL. The membranes were pelleted using an Optima TLX tabletop ultracentrifuge with a TLA 120.2 rotor (511,000 × g, 30 min). The supernatant was dialyzed against 0.154 M NaCl and 0.01% EDTA, pH 7.4, and was used to isolate luminal lipoproteins. Isolation of nascent subcellular and serum lipoprotein fractions. The d < 1.006 g/mL fraction was isolated using the TLA 120.2 rotor. The samples were centrifuged for 3 h at 511,000

Lipids, Vol. 37, no. 9 (2002)

× g, and the top 250–300 µL was removed after tube slicing. The density of the infranatant was raised to 1.210 g/mL using solid KBr, and the d 1.006–1.210 g/mL fraction was floated by centrifuging for 4.5 h at 511,000 × g in the TLA 120.2 rotor. The top 250–300 µL was removed after tube slicing, and both the d < 1.006 and the d 1.006–1.210 g/mL fractions were dialyzed against 0.01% EDTA, pH 7.4, with a final change against distilled water prior to lyophilization. Serum lipoproteins were isolated at the same densities using the same conditions with the final fractions being dialyzed and lyophilized. Radioisotope incorporation studies. Mice were anesthetized with ketamine/xylazine and injected via the retroorbital plexus with 25–30 µCi [2-3H]glycerol (16.5 Ci/mmol; NEN Life Science Products, Inc., Boston, MA). At 7.5, 15, 30, and 45 min after injection, the animals were killed by exsanguination from the inferior vena cava. The livers were removed, trimmed of excess fat and connective tissue, and rinsed. Subcellular fractions and nascent lipoproteins were isolated as described above. Aliquots of serum (100 µL) were taken for isolation of serum lipoprotein fractions as described above. MTP inhibitor studies. MTP inhibitor (BMS 197636, kindly provided by R. Gregg, Bristol Myers Squibb, Princeton, NJ) was dissolved in 10% M-Pyrol (N-methyl-2-pyrrolidone) 80% water, 5% Cremophor EL, and 5% ethanol at a concentration of 6.25 mg/mL. The inhibitor was administered to mice by gavage (20 µg/g) 60 min prior to injection of [3H]glycerol. The remainder of the experiment was as described above. To establish the effectiveness of the MTP inhibitor, hepatic VLDL TG production rates were assessed using Triton WR1339 (12). Isolation of lipids. Membrane and cytosolic fat fractions were delipidated using the method of Folch et al. (13). Lipoprotein fractions were delipidated using ethanol/ether (14). Neutral and polar lipids were separated using small silicic acid columns (15). Neutral lipids were separated into individual lipid classes by TLC on silica gel 60A thin-layer plates (Whatman K-6; Fisher Scientific, Atlanta, GA) with petroleum ether/ethyl ether/acetic acid (80:20:1, by vol) as the developing solvent. TG were scraped from the plate into scintillation vials. Scintillation fluid (Bio-Safe II; Research Products International Corp., Mount Prospect, IL) was added, and radioactivity was determined using a Beckman LS6800 scintillation counter. Analytical and enzymatic assays. Protein was determined by the bicinchoninic acid method (Pierce, Rockford, IL), modified to eliminate interference by lipid (16) and using BSA as standard. TG were quantified using an enzymatic assay (Raichem, San Diego, CA) adapted to microtiter plates. Aliquots of the lipid fractions were dissolved in isopropanol, and 20 µL was added to each well. Water (80 µL) was added followed by 100 µL of the color reagent. Plates were incubated at 37°C for 10 min and read at 520 nm using a BioRad Model 550 microplate reader. The assay was linear from 0.5 to 10.0 µg TG. Glycerol was used as a standard. Statistics. Data are presented as the mean ± SEM and were analyzed by an unpaired Student’s t-test using GraphPad InStat (GraphPad Solftware, San Diego, CA).

ADDITION OF TG TO NASCENT VLDL IN LIVER

881

RESULTS [3H]Glycerol incorporation studies. [3H]Glycerol was incorporated rapidly into Golgi and microsomal membrane TG, reaching peak specific activity in both membranes within 7.5 min of isotope injection (Fig. 1A). The specific activity of TG in the microsomal membranes decreased to approximately 40% of peak value by 45 min, while the specific activity of TG in Golgi membrane decreased to about 30% of peak value within the same time. The appearance of labeled TG in Golgi VLDL followed a pattern similar to that of the TG within the Golgi membrane; however, the specific activity of Golgi VLDL TG was less than 50% of that in the Golgi membrane at each time point (Fig. 1B). The specific activity of TG in the d < 1.006 g/mL lipoproteins recovered from the microsomal fraction remained relatively low with a peak near 15 to 30 min. The specific activity of the luminal TG was lower than in the membrane, similar to that observed in the Golgi. Labeled TG did not appear in the serum VLDL fraction until the 15-min time point, and increased to the 30-min time point, consistent with peak intracellular specific activity and its trafficking out of the cell (Fig. 2A). Incorporation of [3H]glycerol in MTP-inhibited animals. The MTP inhibitor BMS 197636 was administered to mice FIG. 2. Effects of the microsomal TG transfer protein (MTP) inhibitor BMS 197636 on hepatic TG production in mice. BMS 197636 was administered to mice by gavage (20 µg/g), and 60 min later [2-3H]glycerol was injected. The appearance of labeled TG in serum VLDL was monitored (A), and hepatic TG production was determined using the Triton method (B). Data represent the mean ± SEM. (A) Control, n = 3; MTP inhibitor (MTPi), n = 3. (B) Control, n = 5; MTPi, n = 3.

FIG. 1. Incorporation of [3H]glycerol into TG of mouse hepatic fractions. Mice were injected with 25–30 µCi [2-3H]glycerol, and hepatic fractions were isolated at different times after injection. Lipids were extracted, and TG mass and radioactivity were determined as described in the Experimental Procedures section. (A) Golgi and microsomal membranes; (B) Golgi and microsomal VLDL. Data represent the mean ± SEM. Golgi, n = 5; microsomes, n = 4.

by gavage, and 60 min later hepatic TG production rates were determined. The MTP inhibitor decreased hepatic TG production rates by greater than 85% [127.4 ± 4.7 vs. 17.1 ± 5.9 (mean ± SEM) µmol/kg/h, n = 3/group] (Fig. 2B). To verify the block in hepatic TG production, the inhibitor was given to mice and 60 min later [3H]glycerol was injected. Less than 5% of the radioactivity found in the serum VLDL of control animals was recovered in the MTP-inhibited mice, confirming the block of MTP activity and secretion of VLDL TG (Fig. 2A). Inhibition of MTP did not affect the incorporation of newly synthesized TG into microsomal membranes, as the specific activity of the TG in these membranes at 7.5 min was identical to that found in the control (noninhibited) animals. However, inhibition of MTP did attenuate the turnover of newly synthesized TG in both the microsomal and Golgi membranes (Fig. 3). In microsomal membranes the turnover was reduced by approximately 40%, whereas in the Golgi membranes turnover was reduced nearly 50%. The inhibitor dramatically altered the incorporation of newly synthesized TG into the cytosolic pool. In control animals the cytosolic TG pool reached peak specific activity approximately 30 min after [3H]glycerol injection (Fig. 4). In contrast, in animals given the inhibitor the specific activity of Lipids, Vol. 37, no. 9 (2002)

882

K. VALYI-NAGY ET AL.

FIG. 3. Effects of the MTP inhibitor BMS 197636 on turnover of newly synthesized TG in microsomal and Golgi membranes. Control mice and mice given the BMS 197636 (20 µg/g) were injected with [2-3H]glycerol, and hepatic Golgi and microsomal membranes were isolated at 7.5 and 45 min postinjection. TG mass and radioactivity were determined as described in the Experimental Procedures section, and specific activity was calculated. The specific activity of TG at 45 min is expressed as percentage of specific activity at the 7.5-min time point. Data represent the mean ± SEM. Control, n = 4; MTPi n = 3. *,**Significantly different from control at P = .007 and .003, respectively. For abbreviations see Figure 2.

the cytosolic TG reached peak value within 7.5 min, and the specific activity at this time point was identical to the specific activity at 30 min in the control animals (Fig. 4). In addition, the specific activity of the cytosolic TG in the presence of the inhibitor remained constant over the experimental period, suggesting little movement out of this pool. Hepatic membrane and cytosol TG content. Golgi membranes recovered after sodium carbonate treatment and release of luminal lipoproteins contained over twice as much TG as membranes recovered from microsomal fractions (P = 0.0004) (Table 1). Inhibition of MTP led to a significant decrease in the amount of TG in the Golgi membrane and a slight, but not significant, increase in the amount of TG in the microsomal membrane. The cytosolic TG pool increased more than 130% when MTP was inhibited. It is important to note that in these studies MTP was inhibited from 67.5 to 105 min. DISCUSSION A growing body of evidence suggests that the assembly of VLDL is not completed within the ER of the cell but continues en route to or within the Golgi apparatus (8,10,17). Previ-

FIG. 4. Incorporation of [3H]glycerol into TG in the cytosolic pool. Control mice and mice given the MTP inhibitor (BMS 197636, 20 µg/g) were injected with [2-3H]glycerol, and hepatic cytosolic fat was isolated at various times postinjection. TG mass and radioactivity were determined, and specific activity was calculated. Data are expressed as mean ± SEM. Control, n = 4; MTPi, n = 4. *Significantly different from control at P < .005. For abbreviations see Figure 2.

ous studies from our laboratory have indicated that as much as 50% of the TG and 30% of the phospholipid may be added in the Golgi apparatus of the cell (9). More recent studies from our laboratory have identified a population of apoB-containing HDL within the Golgi apparatus of mouse liver that are apparently secreted as VLDL, suggesting that neutral lipid is added within the Golgi apparatus (10). The results presented here provide additional evidence for a direct role of the Golgi apparatus in the assembly process. Our studies demonstrate the presence of a pool of newly synthesized TG within the membranes of the Golgi apparatus. Based on membrane protein, the level of TG within this membrane is greater than that in the membrane of the ER (Table 1). The turnover of TG in the Golgi membrane pool is rapid and is similar to that found with the TG in the membrane of the microsomal fraction. Furthermore, the turnover of this Golgi membrane pool is attenuated by the MTP inhibitor BMS 197636. We suggest that this pool of TG is used in the lipidation of lipoproteins within the Golgi apparatus in an MTP-dependent process. The presence of TG within Golgi membranes is not a new finding. Hebbachi and Gibbons (18) incubated rat hepatocytes for 4 h in the presence of labeled oleate and glycerol and reported that the membranes of both cis- and trans-Golgi were more heavily labeled with TG than those of the rough or smooth ER. The trans-Golgi membrane had nearly twice as

TABLE 1 TG Content of Hepatic Fractionsa Group Control MTP inhibited a

Golgi membranes (mg/mg protein)

Microsomal membranes (mg/mg protein)

Cytosol (µg/g liver)

0.062 ± 0.007a,b 0.031 ± 0.004b

0.027 ± 0.003a 0.032 ± 0.004

198.0 ± 13.0c 461.2 ± 42.9c

Data represent the mean ± SEM from eight preparations. Values with the same roman superscript are significantly different at the following P values: a0.0004; b0.0007; c<0.0001.

Lipids, Vol. 37, no. 9 (2002)

ADDITION OF TG TO NASCENT VLDL IN LIVER

much labeled TG per mg subcellular vesicle protein as the smooth ER membrane and almost 10 times as much as the rough ER membrane. The cis-Golgi membrane contained about the same amount of labeled TG as the smooth ER membrane but eight times as much as the rough ER. Our studies demonstrate that the mass of TG in the Golgi membrane per mg protein is two to three times that found in the membranes of the rough and smooth ER (Table 1). Since TG is not a structural component of membranes and since the Golgi membrane is not recognized as a major site of TG synthesis, the presence of this amount of TG within the Golgi membrane led to questions about its origins and disposition. To address these questions, kinetic studies with radioisotopic tracers were performed. The labeling kinetics of the Golgi membrane TG were similar to those found with microsomal membranes (Fig. 1). A number of studies have shown that TG is synthesized in the ER and trafficks primarily to the cytosol, although some may enter the VLDL assembly pathway (19,20). Our results suggest that newly synthesized TG also trafficks to the Golgi membrane, yet the pathway by which this occurs is unknown. It is unlikely that the Golgi membrane TG derives from the cytosolic pool since the specific activity of the Golgi pool is higher than that of the cytosol at the early time points. Furthermore, we have been unable to document significant level of DAG acyltransferase (DGAT) activity in Golgi membranes. The presence of DGAT in the Golgi membrane might be expected if TG were transported from the cytosolic pool via a process of hydrolysis and subsequent re-esterification analogous to that occurring with the ER (21). It seems most likely that the transfer of membrane TG from the ER to the Golgi occurs via vesicular transport (22,23). Moreau et al. (24), using an in vitro transfer assay, reported that little labeled TG was transferred to the Golgi fractions (24). However, their results were collected using an in vitro system in which lipid transfer was less than 1% of the total lipids. Transfer in vivo may be quite different. Previous studies from our laboratory (9) led us to conclude that as much 50% of the VLDL TG may be added to nascent lipoproteins in the Golgi apparatus. The finding of newly synthesized TG within the Golgi membrane led us to hypothesize that this TG serves as a pool for addition to nascent lipoproteins. The mechanism by which this TG is added to nascent Golgi lipoproteins is unknown, but there is evidence that MTP may be present in cis- and trans-Golgi fractions of liver (18,25,26). In addition, a recent study demonstrates the presence of MTP in the Golgi apparatus of rat enterocytes (27). We then studied the effects of MTP inhibition on the synthesis of TG and distribution of newly synthesized TG within the liver. Prior to initiating these studies, we demonstrated that rates of hepatic TG output, as measured using the Triton method, were reduced by more than 85% when mice were given the inhibitor at a dose of 20 µg/g body weight (Fig. 2B). In addition, our radioisotope incorporation studies were consistent with the Triton studies, demonstrating a marked reduction in the amount of newly synthesized TG appearing in the serum pool over the 45-min time period (Fig. 2A).

883

The inactivation of MTP led to a doubling of the cytosolic TG pool and a slight, albeit statistically insignificant, increase in the microsomal membrane TG pool (Table 1). In contrast, the pool of TG in the Golgi membrane was decreased by nearly 40% when MTP was inactivated. As might be predicted, inhibition of MTP blocked the entry of newly synthesized TG into the lumen of the ER (data not shown). Surprisingly, in the presence of the inhibitor the specific activity of the cytosolic TG reached its peak more rapidly than in control samples, and the specific activity of this pool remained constant over the 45-min period (Fig. 4). These data suggest that inhibition of TG transfer to the lumen of the ER leads to more rapid transfer of newly synthesized TG into the cytosolic pool. Furthermore, the fact that the specific activity of this pool remained constant over the 45-min time period suggests that the inactivation of MTP blocks the transfer of TG from the cytosol to the ER membrane. Hebbachi and Gibbons (18) reported a similar result in studies with rat hepatocytes. They suggested that MTP inhibition blocks the transfer of cytosolic TG to the ER membrane by a route that involves the luminal leaflet of the ER membrane. As noted, this would cause an accumulation of TG not only in cytosol but also in the ER membrane. It is also possible that the inhibition of MTP affects the hydrolysis of TG in the cytosolic pool or even its reesterification within the ER membrane. The decrease in TG in the Golgi membrane in relation to the increased TG in the ER membrane is an important finding. It suggests that if Golgi membrane TG does derive from ER membranes, this process is not a simple blebbing of the ER but must involve some sorting, as suggested by Moreau et al. (24), and the MTP alters or inhibits the sorting or transport in some manner. Inhibiting MTP not only decreased the TG in the Golgi membrane but also attenuated the turnover of newly synthesized Golgi membrane TG (Fig. 3), which suggests that MTP may play a role in the transfer of TG from the Golgi membrane to its contents. Hebbachi et al. (25) reported that inhibiting MTP with BMS 200150 resulted in a delayed removal of apoB-100 and apoB-48 from the Golgi membranes, and Hebbachi and Gibbons (18) showed a delayed removal of newly synthesized TG from the membranes of cis- and trans-Golgi. Alternatively, inhibiting MTP could block the assembly of lipoprotein in the ER, leading to a decrease in lipoproteins in the Golgi apparatus. The transfer of TG from the Golgi membrane to nascent lipoproteins within the Golgi complex would then also be blocked because of a lack of acceptor lipoproteins within the lumen. In this case, the direct involvement of MTP within the Golgi apparatus would not be required. Regardless of the mechanism, the results support a role for the Golgi apparatus in adding lipid (TG) to nascent lipoproteins. Although our studies do not permit us to differentiate between occurrences in the cis- or trans-Golgi, it is possible that the lipidation occurs early within the Golgi apparatus as suggested by Cartwright and Higgins (7). Indeed, if MTP is located within the Golgi complex by way of escape from the ER, we might expect it to be found within the cis Golgi elements. Lipids, Vol. 37, no. 9 (2002)

884

K. VALYI-NAGY ET AL.

These studies provide additional evidence that the Golgi apparatus participates in the assembly of nascent lipoproteins by the liver. The finding of a pool of newly synthesized TG within the Golgi membrane fraction provides a plausible mechanism by which lipid is added to nascent lipoproteins. Our finding that the inhibition of MTP attenuates the turnover of the Golgi membrane TG pool further supports the hypothesis that this pool is involved in lipoprotein assembly within the Golgi. The pathway by which TG is delivered to the Golgi membranes and the precise mechanisms by which lipid is added to the particles being formed are under study. ACKNOWLEDGMENTS This work was supported by National Institutes of Health grants, HL57984 and DK26657. We gratefully acknowledge Richard Gregg, Bristol Myers-Squibb, for the generous gift of MTP inhibitor and John Wetterau, Bristol Myers-Squibb, for helpful comments in the preparation of the manuscript.

REFERENCES 1. Alexander, C.A., Hamilton, R.L., and Havel, R.J. (1976) Subcellular Localization of B Apoprotein of Plasma Lipoproteins in Rat Liver, J. Cell Biol. 69, 241–263. 2. Boren, J.S., Rustaeus, S., and Olofsson, S.-O. (1994) Studies on the Assembly of Apolipoprotein B-100 and B-48-Containing Very Low Density Lipoproteins in McA-RH7777 Cells, J. Biol. Chem. 269, 25879–25888. 3. Olofsson, S.-O., Asp, L., and Borén, J. (1999) The Assembly and Secretion of Apolipoprotein B-Containing Lipoproteins, Curr. Opin. Lipidol. 10, 341–346. 4. Bamberger, M., and Lane, M. (1990) Possible Role of the Golgi Apparatus in the Assembly of Very Low Density Lipoprotein, Proc. Natl. Acad. Sci. USA 87, 2390–2394. 5. Higgins, J. (1988) Evidence That During Very Low Density Lipoprotein Assembly in Rat Hepatocytes Most of the Triacylglycerol and Phospholipid Are Packaged with Apolipoprotein B in the Golgi Complex, FEBS Lett. 232, 405–408. 6. Boström, K., Borén, J., Wettesten, M., Sjöberg, A., Bondjers, G., Wiklund, O., Carlsson, P., and Olofsson, S.-O. (1988) Studies on the Assembly of ApoB-100-Containing Lipoproteins in HepG2 Cells, J. Biol. Chem. 263, 4434–4442. 7. Cartwright, I., and Higgins, J. (1995) Intracellular Events in the Assembly of Very Low Density Lipoprotein Lipids with Apolipoprotein B in Isolated Rabbit Hepatocytes, Biochem. J. 310, 897–907. 8. Stillemark, P., Borén, J., Andersson, M., Larsson, T., Rustaeus, S., Karlsson, K.-A., and Olofsson, S.-O. (2000) The Assembly and Secretion of Apolipoprotein B-48-Containing Very Low Density Lipoproteins in McA-RH7777 Cells, J. Biol. Chem. 275, 10506–10513. 9. Swift, L.L. (1995) The Assembly of Very Low Density Lipoprotein in Rat Liver: A Study of Nascent Particles Recovered from the Rough Endoplasmic Reticulum, J. Lipid Res. 36, 395–406. 10. Swift, L.L., Valyi-Nagy, K., Rowland, C., and Harris, C. (2001) Assembly of Very Low Density Lipoproteins in Mouse Liver: Evidence for Heterogeneity of Particle Density in the Golgi Apparatus, J. Lipid Res. 42, 218–224. 11. Fujiki, Y., Hubbard, A.L., Fowler, S., and Lazarow, P.B. (1982) Isolation of Intracellular Membranes by Means of Sodium Carbonate Treatment: Application to Endoplasmic Reticulum, J. Cell Biol. 93, 97–102.

Lipids, Vol. 37, no. 9 (2002)

12. Swift, L.L., Farkas, M.H., Major, A.S., Valyi-Nagy, K., Linton, M.F., and Fazio, S. (2001) A Recycling Pathway for Resecretion of Internalized Apolipoprotein E in Liver Cells, J. Biol. Chem. 276, 22965–22970. 13. Folch, J., Lees, M., and Sloane Stanley, G.H. (1957) A Simple Method for the Isolation and Purification of Total Lipides from Animal Tissues, J. Biol. Chem. 226, 497–509. 14. Bersot, T., Brown, W., Levy, R., Windmueller, H., Fredrickson, D., and LeQuire, V. (1970) Further Characterization of the Apolipoproteins of Rat Plasma Lipoproteins, Biochemistry 9, 3427–3433. 15. Dittmer, J., and Wells, M. (1969) Quantitative and Qualitative Analysis of Lipids and Lipid Components, Methods Enzymol. 14, 482–530. 16. Morton, R., and Evans, T. (1992) Modification of the Bicinchoninic Acid Protein Assay to Eliminate Lipid Interference in Determining Lipoprotein Protein Content, Anal. Biochem. 204, 332–334. 17. Asp, L., Claesson, C., Borén, J., and Olofsson, S.-O. (2000) ADP-Ribosylation Factor 1 and Its Activation of Phospholipase D Are Important for the Assembly of Very Low Density Lipoproteins, J. Biol. Chem. 275, 26285–26292. 18. Hebbachi, A.-M., and Gibbons, G. (1999) Inactivation of Microsomal Triglyceride Transfer Protein Impairs the Normal Redistribution but Not the Turnover of Newly Synthesized Glycerolipid in the Cytosol, Endoplasmic Reticulum and Golgi of Primary Rat Hepatocytes, Biochim. Biophys. Acta 1441, 36–50. 19. Gibbons, G.F., and Wiggins, D. (1995) Intracellular Triacylglycerol Lipase: Its Role in the Assembly of Hepatic Very-LowDensity Lipoprotein (VLDL), Adv. Enzyme Regul. 35, 179–198. 20. Salter, A.M., Wiggins, D., Sessions, V.A., and Gibbons, G.F. (1998) The Intracellular Triacylglycerol/Fatty Acid Cycle: A Comparison of Its Activity in Hepatocytes Which Secrete Exclusively Apolipoprotein (apo) B100 Very-Low-Density Lipoprotein (VLDL) and in Those Which Secrete Predominantly apoB48 VLDL, Biochem. J. 332, 667–672. 21. Gibbons, G.F., Islam, K., and Pease, R.J. (2000) Mobilisation of Triacylglycerol Stores, Biochim. Biophys. Acta 1483, 37–57. 22. Bannykh, S., and Balch, W. (1998) Selective Transport of Cargo Between the Endoplasmic Reticulum and Golgi Compartments, Histochem. Cell Biol. 109, 463–475. 23. Palade, G.E. (1975) Intracellular Aspects of the Process of Protein Transport, Science 189, 347–354. 24. Moreau, P., Rodriguez, M., Cassagne, C., Morré, D.M., and Morré, D.J. (1991) Trafficking of Lipids from the Endoplasmic Reticulum to the Golgi Apparatus in a Cell-Free System from Rat Liver, J. Biol. Chem. 266, 4322–4328. 25. Hebbachi, A.-M., Brown, A.-M., and Gibbons, G. (1999) Suppression of Cytosolic Triacylglycerol Recruitment for Very Low Density Lipoprotein Assembly by Inactivation of Microsomal Triglyceride Transfer Protein Results in a Delayed Removal of apoB-48 from Microsomal and Golgi Membranes of Primary Rat Hepatocytes, J. Lipid Res. 40, 1758–1768. 26. White, D., Bennett, A., Billett, M., and Salter, A. (1998) The Assembly of Triacylglycerol-Rich Lipoproteins: An Essential Role for the Microsomal Triacylglycerol Transfer Protein, Br. J. Nutr. 80, 219–229. 27. Levy, E., Stan, S., Delvin, E., Menard, D., Shoulders, C., Garofalo, C., Slight, I., Seidman, E., Mayer, G., and Bendayan, M. (2002) Localization of Microsomal Triglyceride Transfer Protein in the Golgi, J. Biol. Chem. 277, 16470–16477.

[Received October 13, 2001, and in revised form June 13, 2002; revision accepted September 18, 2002]