Hepatitis C Virus Nonstructural Proteins Inhibit Apolipoprotein B100 Secretion*

Host genes involved in lipid metabolism are differentially regulated during the early stages of hepatitis C virus (HCV) infection. The majority of lipids synthesized in the liver are exported to other tissues in the form of lipoproteins. The formation of these lipoproteins is dependent upon the association of triglycerides with apolipoprotein B100. Using the HCV subgenomic replicon expression system, we show that secretion of apoB100 is significantly reduced. Inhibition of apoB100 degradation by ALLN did not improve secretion. Triglyceride levels as well as microsomal triglyceride transfer protein mRNA and activity levels were reduced in replicon-expressing cells, indicating potential reasons for the observed decrease. Further evidence is presented for the interaction between the HCV nonstructural protein 5A and apoB100. These results provide further insight into the alteration of lipid metabolism by HCV.

Host genes involved in lipid metabolism are differentially regulated during the early stages of hepatitis C virus (HCV) infection. The majority of lipids synthesized in the liver are exported to other tissues in the form of lipoproteins. The formation of these lipoproteins is dependent upon the association of triglycerides with apolipoprotein B100. Using the HCV subgenomic replicon expression system, we show that secretion of apoB100 is significantly reduced. Inhibition of apoB100 degradation by ALLN did not improve secretion. Triglyceride levels as well as microsomal triglyceride transfer protein mRNA and activity levels were reduced in replicon-expressing cells, indicating potential reasons for the observed decrease. Further evidence is presented for the interaction between the HCV nonstructural protein 5A and apoB100. These results provide further insight into the alteration of lipid metabolism by HCV. 3 infection is a major health problem worldwide. HCV infection causes chronic hepatitis in up to 60 -80% of infected adults (1) and is associated with steatosis, cirrhosis, and hepatocellular carcinoma (2). The HCV viral genome consists of a positive sense single-stranded RNA 9.6 kb in length and encodes a polyprotein precursor of ϳ3000 amino acids (3). The resulting polyprotein is cleaved into at least three structural proteins (the core protein and the envelope proteins E1 and E2) and a variety of nonstructural proteins (p7 through NS5A/B) (3,4). Many of the nonstructural proteins are essential for productive viral replication (5).

Hepatitis C virus (HCV)
The development of selectable subgenomic HCV RNA replicons has led to important advances in the field of HCV. Recently, several groups have reported the full-length infectious tissue culture system (6 -8), which is likely to further our understanding of the infectious process. The subgenomic replicons are bicistronic constructs composed of the HCV internal ribosomal entry site (IRES), the neomycin phosphotransferase (neo) gene, and the encephalomyocarditis IRES, which mediates the translation of the HCV nonstructural proteins NS3 through NS5, followed by the 3Ј-noncoding region (9).
Chronic HCV is characterized by several histological features of the liver such as bile duct damage, lymphoid follicles, low serum cholesterol levels, and, in ϳ50 -60% of the cases, steatosis (10,11). Recently, genomic analysis conducted on the livers of HCV-infected chimpanzees revealed that the transcript levels of genes involved in lipid metabolism and homeostasis are altered (12,13).
The liver is a major site of lipid synthesis. Many of the synthesized lipids are exported from the liver in the form of apolipoprotein-containing particles known as very low density lipoproteins (VLDLs), precursors of low-density lipoproteins. The VLDLs are composed of a nonpolar core of triglycerides and cholesteryl esters surrounded by a monolayer amphipathic coating of protein, phospholipids, and cholesterol (14,15). VLDLs are formed in the luminal space within the endoplasmic reticulum (ER), whereas the lipids that are required for their formation are synthesized on the cytosolic side of the ER membrane (16). The ER is also the center of HCV translation and replication activities, resulting in the ER stress response and the induction of oxidative stress (17).
The microsomal triglyceride transfer protein (MTP) and apolipoprotein B100 (apoB) are major regulators of VLDL assembly in the liver and have been shown to physically interact during VLDL assembly (18 -22). ApoB is the major structural protein associated with the VLDLs and low density lipoproteins (23). In the presence of sufficient lipids, mainly triglycerides, apoB is efficiently assembled into secretion-competent VLDL particles (23). If, however, apoB is poorly lipidated, the majority of newly synthesized apoB is degraded by the proteasome in the cytosol (24) or by ER-resident proteases (25,26). MTP stabilizes apoB by lipidation, and the subsequently lipidated apoB fuses with triglyceride-rich particles, leading to the formation of VLDLs (27,28). MTP is present in the ER lumen as a heterodimeric protein consisting of the active 97-kDa monomer linked with a protein disulfide isomerase (29). It transfers lipids from various donor lipid sites to acceptor sites and plays a fundamental role in the cotranslational lipidation of apoB as it enters the ER lumen, thus preventing degradation (28). Although MTP is capable of transferring all classes of lipids found in apoB lipoproteins (triglyceride, cholesteryl esters, free cholesterol, and phospholipids), in vitro analyses show that it strongly prefers triglycerides and cholesteryl esters (30).
During the early stages of HCV infection, host genes involved in lipid metabolism are differentially regulated (12). HCV infection causes the formation of hepatocellular lipid droplets onto which HCV structural and nonstructural proteins have been shown to colocalize (12). In addition, the HCV core protein induces steatosis by itself and colocalizes with the cytoplasmic globular lipid structures (31). The HCV core protein also reduces the activity of MTP (31). Even though core can associate with several cellular proteins, direct association of the HCV core protein with apoB or MTP has not been demonstrated (32). In addition, the HCV nonstructural protein NS5A colocalizes with the core protein on lipid droplets and interacts with apolipoprotein AI and AII (protein components of high-density lipoproteins) (33). In chronic HCV infection, hypobetalipoproteinemia is prevalent, especially in patients infected with genotype 3 (34).
In this study, we examine the role of the HCV nonstructural proteins on lipid export via apoB using the HCV subgenomic replicon system. Secretion of apoB via lipoproteins was drastically reduced in the cells containing the HCV subgenomic replicon, whereas in cell lines cured of the replicon, apoB secretion was restored. A potential reason for the decrease in apoB secretion could be due to the lower triglyceride levels that we detected. Our studies further revealed that MTP activity was decreased in the subgenomic replicon-containing cells. In addition, MTP transcript levels and promoter activity were reduced. These results provide evidence that HCV subgenomic replicons inhibit lipoprotein secretion by interfering with the formation of secretioncompetent apoB lipoproteins via inhibition of MTP. Moreover, apoB is shown here to interact with the HCV nonstructural protein NS5A, suggesting that apoB may be a target for HCV in the context of the HCV subgenomic replicon.

MATERIALS AND METHODS
Cell Culture, Immunoprecipitations, and Reporter Assays-Huh-7, FCA4 (gift of C. Seeger, Institute for Cancer Research, Fox Chase Center, PA), and K2040 (gift of Michael Gale, University of Texas Southwestern Medical Center) human hepatoma cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. Huh-7.5 cured cells were a gift from C. Rice at Rockefeller University. The K2040 cured cells were cured as described by Tardif et al. (35). For metabolic labeling, Huh-7, FCA4, and K2040 cells (and the cured cells) were labeled for the indicated times with 100 Ci of trans 35 S label (specific activity 1,175 Ci/mmol; MP Biomedicals, Inc.) per milliliter of the above medium. In some experiments 40 g/ml N-acetylleucyl-leucyl-norleucinal (ALLN) were added 4 h prior to the labeling. Cellular lysates and media were immunoprecipitated with goat antiapolipoprotein B polyclonal antibody (Chemicon International) or rabbit anti-human albumin antibody (ICN Biomedicals, Inc.) at 4°C for 16 h. The immune complexes were analyzed by electrophoresis using a 4 -15% SDS-polyacrylamide gradient gel (Bio-Rad). For pulse-chase experiments, indicated cells were labeled with trans 35 S label and chased over various time points in complete Dulbecco's modified Eagle's medium. Cells and chase medium were collected and immunoprecipitated as stated above. ALLN (40 g/ml) was added to the medium in some cases prior to the pulse as well as during the chase periods. Oleic acid conjugated to bovine serum albumin (30 l/ml medium) (Sigma) was added to the medium 24 h prior to the pulse in some cases. For transfections, cells at ϳ70% confluency were transfected using Lipofectamine 2000 reagent (Invitrogen). For protein-protein interaction studies, indicated cellular lysates were prepared as described previously (36). Lysates were immunoprecipitated with anti-apoB antibody followed by immunoblotting with anti NS5a (gift of C. Cameron, Pennsylvania State University) and anti-NS4a (Virogen). The HCV NS5A coding sequence and the NS5A mutants were generated as described by Gong et al. (37). For MTP promoter-luciferase reporter (gift of R. A. Davis) (38) assays, lysates were assayed as described previously (36). A ␤-galactosidase expression plasmid was used as an internal control for transfection.
Overall Triglyceride Levels-Triglycerides were quantitated using the L-type TG H kit (Wako Chemicals USA, Inc.) according to the manufacturer's protocol. Briefly, 3.68 ϫ 10 6 cells per reaction (Huh-7, FCA4 and K2040 cells) were collected. Lipids were extracted as described previously (39) and then resuspended in 200 l of 2% Triton X-100 in phosphate-buffered saline. 40 l of lipids were added to 533.3 l of enzyme color A and incubated at 37°C for 5 min. 266.6 l of color enzyme B were then added, and the samples were incubated for 5 min at 37°C. The absorbance was then read at 595 nm. The reactions were standardized using lipid calibrator (Wako Chemicals USA, Inc.). Triplicate reactions were carried out with each cell line.
Measurement of MTP Activity-Huh-7, FCA4, and K2040 cellular lysates were divided into 250-l aliquots and sonicated on ice. Proteins were quantitated by Bradford assay. MTP activity was determined using the MTP activity kit (Roar Biomedical, New York, NY) according to the manufacturer's protocol. Briefly, 150 g of protein were combined with 6 l each of suspensions of donor and acceptor particles. The change of fluorescent intensity was monitored (excitation at 465 nm and emission at 535 nm). The reaction was allowed to proceed over a 720-min time course.
Quantitative Real Time PCR-Total RNA was extracted using RNA STAT-60 reagent (Tel-Test, Inc.) according to manufacturer's standard procedures. 1 g of total RNA was preincubated with oligo(dT) primer (Invitrogen) at 70°C for 5 min, put on ice for 5 min, and then used to generate a cDNA library using ImProm-II reverse transcriptase (Fisher). 1.5 l of cDNA library were added to each 25-l PCR reaction using QuantiTect SYBR Green PCR reagent (Qiagen). Quantifications were done in triplicate and repeated three separate times. Amplification reactions were performed using primers for either MTP (forward 5Ј-GGAGCTTCCCCAAGAAATGAAT-3Ј and reverse 5Ј-GAACTC-GACGGACAATTTTGCT-3Ј) or hypoxanthine phosphoribosyltransferase (forward 5Ј-TTGGTGGAGATGATCTCTCAAC-3Ј and reverse 5Ј-AGCTTGCGACCTTGACCAT-3Ј). Relative transcript levels were calculated using the ⌬⌬Ct method as specified by the manufacturer.

HCV Subgenomic Replicon Decreases Intact Synthesis and Secretion of apoB-
To study the effect of the HCV nonstructural proteins on apoB synthesis and secretion, we first set out to determine whether we could detect apoB synthesis and secretion in the HCV subgenomic repliconcontaining cells. Huh-7 cells have been shown to secret lipoprotein particles that contain apoB100 (16). ApoB synthesis and secretion were determined by metabolically labeling Huh-7 cells (control) and FCA4 cells (HCV replicon-containing cells) with [ 35 S]methionine followed by immunoprecipitation with an apolipoprotein B antibody. The immunoprecipitates were then subjected to SDS-PAGE. The results show intact apoB (fully lipidated, 550 kDa) in both the lysates (Fig. 1A, lanes 1 and 2).
To detect apoB secretion out of the cells via lipoproteins, [ 35 S]methi- onine-labeled Huh-7 and FCA4 culture media were immunoprecipitated with the anti-apoB antibody. Intact apoB was detected in the Huh-7 medium, whereas there was no evidence for apoB secretion into the FCA4 culture medium (Fig. 1A, lanes 3 and 4). These results indicate that apoB synthesis occurs in Huh-7 as well as the FCA4 cells. However, apoB secretion was significantly decreased in HCV-expressing FCA4 cells.
To determine whether our observations are due to clonal differences between the parental Huh-7 cells and the Huh-7 cells harboring the HCV replicon, cured replicon cells (cured by prolonged treatment with interferon ␣) were also examined by the same method described above (Fig. 1B). Two separate cured cell lines (Huh-7.5 and K2040 cured) were both able to restore apoB secretion into the media (Fig. 1B, lanes 3-6). Consequently, the decrease in apoB secretion can be attributed to the presence of the HCV replicon.
To further analyze apoB synthesis, we pulse-labeled Huh-7, FCA4, and K2040 cells with [ 35 S]methionine from 15 to 60 min, followed by an immunoprecipitation with antibodies against either apoB or albumin. K2040 is another HCV subgenomic replicon-containing cell line that was used to rule out possible clonal effects of FCA4. At each time point there were lower levels of intact apoB in the FCA4 and K2040 cells compared with the levels in the Huh-7 cells (Fig. 1C, top row). In addition, albumin protein synthesis remained constant at the various time points among all three cell lines, indicating a specific inhibitory effect on apoB synthesis in the presence of the HCV subgenomic replicon (Fig.  1C, bottom row).
Secretion Level of apoB Was Decreased in the Presence of the HCV Nonstructural Proteins-Lipoprotein secretion is reduced when synthesis of apoB is inhibited or when degradation of apoB is elevated because of a lack of lipidation (24,41,42). To demonstrate the secretion rate of apoB100 in the presence of the HCV subgenomic replicon, Huh-7, FCA4 and K2040 cells were pulse-labeled with [ 35 S]methionine and then chased for up to 2 h. The cell lysates as well as the culture media were monitored for intact apoB as well as albumin by immunoprecipitation using anti-apoB or anti-albumin antibodies. We detected slightly lower levels of apoB in the FCA4 and K2040 cellular lysates as compared with the Huh-7 cellular lysates (Fig. 2, portion marked cellular). However, the secretion rate of apoB was dramatically reduced in the FCA4 and K2040 cells as compared with the secretion rate in the Huh-7 cells (Fig. 2, portion marked secreted). We were able to detect a small amount of apoB secretion in the FCA4 and K2040 cells at later chase times, indicating that apoB secretion can occur in these cells but at considerably lower levels (Fig. 2). Also, general protein secretion was not affected in the replicon-containing FCA4 and K2040 cells as determined by the secretion of albumin into the culture media (Fig. 2, portion marked  secreted, row marked apoB).
The degradation of apoB due to poor lipidation occurs via the proteasome in the cytosol and by proteases in the endoplasmic reticulum (16). ALLN, a cysteinyl protease inhibitor, has been shown to help block both of these activities (43). We tested whether apoB secretion would increase in the replicon-containing cells if apoB degradation were inhibited. Huh-7, FCA4, and K2040 cells were pulse-labeled with [ 35 S]methionine in the presence of ALLN, chased for up to 2 h, and then immunoprecipitated with anti-apoB or anti-albumin antibodies. We were able to detect synthesized intact apoB in all the cellular lysates in the presence of ALLN, indicating that ALLN does not inhibit apoB synthesis ( Fig. 3 portion marked cellular). ApoB secretion by FCA4 and K2040 cells was not increased when degradation of apoB was inhibited by ALLN (Fig. 3, portion marked secreted, row marked apoB). These cells were, however, able to secrete albumin at levels similar to those of Huh-7 cells (Fig. 3, portion marked secreted, row marked albumin). These results suggest that degradation of apoB is not the major reason for the inability of HCV replicon-containing cells to secrete apoB into the culture medium.
Triglyceride Levels Decreased in the Presence of the HCV Subgenomic Replicon-ApoB synthesis and ultimate secretion as part of a lipoprotein complex depends upon proper lipidation of apoB, in particular, lipidation by triglycerides. Using a colorimetric assay, we determined the amount of triglycerides in Huh-7, FCA4, and K2040. We detected an ϳ30% reduction of triglycerides in the replicon-containing cells compared with Huh-7 cells (Fig. 4). The replicon-containing cells therefore have less triglycerides available to lipidate apoB for secretion.
Oleic Acid Treatment Does Not Restore apoB Secretion-We examined the effect of oleic acid treatment of the Huh-7 cells, the repliconcontaining cells, and the cured cells on apoB secretion. Following treatment of the cells with oleic acid for 24 h, the cells were pulse-labeled with [ 35 S]methionine followed by immunoprecipitation with either an anti-apoB antibody or an anti-albumin antibody. As we have shown previously, without oleic acid treatment (Fig. 1) there were slightly lower levels of synthesized apoB in the replicon-containing cellular lysates compared with the Huh-7 and cured cellular lysates (Fig. 5). More importantly however, there was no evidence of restored apoB secretion in the replicon-containing cells, although albumin secretion was similar in all of the cell lines (Fig. 5). These results indicate that even in the presence of increased levels of free fatty acids, which should increase the amount of triglycerides available for lipoprotein synthesis and secretion, apoB secretion is not restored in the HCV subgenomic replicon-containing cells.
HCV Subgenomic Replicon Reduces MTP Activity, Transcript Levels, and Promoter Activity-MTP transports the lipids from the ER lumen to the growing apoB polypeptide chain in the lumen of the ER, allowing FIGURE 2. ApoB secretion is lower in HCV subgenomic replicon cells. Effect of HCV on apoB secretion rates. Huh-7, FCA4, and K2040 cells were pulse-labeled with 35 S for 20 min and chased for the indicated times. The cells were lysed, and the lysates (cellular) and the media (secreted) were immunoprecipitated as stated in the Fig. 1 legend.   FIGURE 3. ApoB secretion is not increased when apoB degradation is inhibited in HCV subgenomic replicon cells. Huh-7, FCA4, and K2040 cells were treated with ALLN for 4 h and then pulse-labeled for 20 min and chased for the indicated times. The cells (cellular) and media (secreted) were then treated and immunoprecipitated as described for Fig. 1. proper translocation and folding to occur (44). In addition, MTP is also necessary for the formation of protein-free triglyceride droplets in the ER lumen that fuse with the apoB-containing particles prior to secretion as lipoproteins (45). Using the subgenomic replicon system, we determined the effect of HCV nonstructural proteins on MTP in the absence of the core protein. An MTP activity assay was performed on Huh-7, FCA4, and K2040 cellular lysates. There was an observable reduction in MTP activity in the FCA4 and K2040 cellular lysates (Fig. 6A). In addition, MTP transcript levels determined by quantitative real time PCR were ϳ80 -96% lower in the K2040 and FCA4 cells, respectively (Fig.  6B). A luciferase assay using an MTP-luciferase promoter construct revealed that, in the presence of the HCV nonstructural proteins, MTP promoter activity was reduced 74% in the K2040 cells and 94% in the FCA4 cells compared with the Huh-7 cells (Fig. 6C). In a transient transfection assay, Huh-7 cells were transfected with the HCV subgenomic replicon RNA (BM4 -5) or with a control RNA. The MTP promoter activity levels were reduced 40% in the BM4 -5 transfected cells compared with those in the control cells (Fig. 6D). Interestingly, in the cured cell line Huh-7.5 the MTP promoter activity was found to be slightly greater than that in Huh-7 cells expressing wild-type MTP promoter activity levels (data not shown). These results suggest that MTP activity, transcript level, and promoter activity are reduced in the presence of the HCV nonstructural proteins. This reduction will severely impact the lipidation of apoB by triglycerides in the cells and therefore cause a decrease in apoB secretion.
HCV Nonstructural Protein, NS5A, Interacts with apoB-The HCV nonstructural protein NS5A has been shown to be capable of interacting with various components of cellular lipid metabolic pathways (33). NS5A not only localizes to the ER and Golgi apparatus but also colocalizes with the HCV core protein on cytoplasmic lipid droplets (33). To determine whether NS5A interacts with apoB, Huh-7 cells were transiently transfected with an NS5A expression vector. The cellular lysates were then immunoprecipitated with an anti-apoB antibody and subsequently immunoblotted with anti-NS5A serum. The results show the interaction of NS5A with apoB (Fig. 7A, lanes 1 and 2). To confirm this interaction in the context of the HCV subgenomic replicon, cellular lysates from FCA4 and K2040 cells were immunoprecipitated with an anti-apoB antibody followed by immunoblotting using an anti NS5A serum (Fig. 7A, lanes 3 and 4). In addition, cellular lysates from FCA4 and K2040 cells were treated in the same way as described above, except that immunoblotting was done using an anti-NS4 antibody. No interaction was detected (data not shown). These results confirm the interaction between NSA5A and apoB in the context of the subgenomic replicon.
We further analyzed this interaction by mapping the apoB binding region of NS5A. We transiently transfected Huh-7 cells with NS5A deletion mutants (Fig. 7B, bottom portion). We only detected an interaction with mutant 1, which contains the C-terminal region of NS5A (Fig. 7B). This indicates that the C-terminal portion of NS5A may be important for binding; however, it may require a certain conformation for binding because there was no binding with mutant 2, which contains a smaller portion of the NS5A C-terminal region.

DISCUSSION
Alteration of lipid metabolism has been shown to occur in the presence of HCV gene expression. Multiple cellular genes involved in lipid metabolism are differentially regulated in HCV acutely infected chimpanzees (12,13). In animal models, the HCV core protein is steatogenic when expressed in the liver of mice (46). In this study, by using the HCV subgenomic replicon system we have determined that the HCV nonstructural proteins NS3 through NS5B are able, in the absence of core protein, to alter lipid metabolism and, in particular, are able to reduce the activity and the transcript and protein levels of MTP, a major regulator of the assembly and secretion of triglyceride-rich lipoprotein particles. In addition, we have determined that one of the HCV nonstructural proteins, NS5A, interacts with apoB.
Interestingly, steatosis occurs when there is an accumulation of triglycerides. We, however, observed a modest decrease in triglyceride synthesis in replicon-expressing cells. Steatosis linked to HCV has been demonstrated to be much more prevalent in genotype 3 infections as compared with genotype 1 infections in European countries (34,(47)(48)(49)(50). However, in Japan hepatic steatosis is common in patients with HCV genotype 1b infection (10). It has been suggested that there may be some differences in the amino acid sequences in the genotype 1b HCV clones between Japan and European countries (11). The HCV subgenomic replicons that were used in this study represent genotype 1b described by Guo et al. (54), which represents the European lb genotype. In addition, in vitro and in vivo studies have shown that the HCV core protein can induce steatosis in transfected cells (55) and transgenic mice (46). The core protein is absent in the HCV subgenomic replicon-containing cells. This study demonstrates that HCV subgenomic replicon expression is capable of altering lipid metabolism in cultured cells.
We have developed a system to monitor apoB synthesis and secretion in the presence of the HCV subgenomic replicon. Using this system, we detected slightly reduced levels of intact apoB synthesis in the repliconcontaining cells. We also determined that in these same cells the secre-tion of apoB was very low. This effect is specific to the presence of the active HCV subgenomic replicon, because apoB secretion does occur efficiently in cured cell lines. Because apoB degradation occurs when apoB is not fully lipidated by triglycerides, we specifically monitored apoB secretion in the presence of ALLN, a cysteinyl protease inhibitor that has been determined previously to help block apoB degradation. Even when apoB degradation was reduced in the replicon cells, secretion of apoB did not occur. Some evidence suggests that apoB is produced in excess of the needs for VLDL secretion and that at least 70% of synthesized apoB undergoes intracellular proteolytic degradation (57). In our experiments we always detected lower amounts of secreted apoB relative to the synthesized levels of apoB. The addition of ALLN alone did not change the overall synthesis and secretion of apoB in the replicon-containing cells, indicating that degradation of apoB is not the major cause for the lack of apoB secretion in these cells. Our results also correlate well with a recent study conducted on the serum lipids of  . NS5A interacts with apoB. A, Huh-7 cells were transiently transfected with and without NS5A expression plasmid. Cellular lysates from these cells and FCA4 and K2040 cells were then immunoprecipitated (IP) with an apoB antibody followed by a Western blot (WB) assay using an anti-NS5A antibody as described under "Materials and Methods." B, Huh-7 cells were transiently transfected with the NS5A expression plasmid or with one of the NS5A deletion mutant (Mut) expression plasmids (bottom) and then immunoprecipitated (IP) and Western blotted (WB) as described above. Wt, wild-type. patients with genotype 1b HCV infection in Japan in which serum apoB levels were significantly reduced compared with levels in patients infected with the hepatitis B virus or an HCV genotype 2a (11).
Lipidation of apoB by triglycerides is regulated at least at three independent steps, namely the synthesis of triglycerides, the transfer of triglycerides across the ER membrane, and the delivery of triglycerides from the ER membrane to apoB by MTP (16). The process and regulation of the transfer of triglycerides across the ER membrane is not well understood (16). In the presence of the HCV subgenomic replicon we have detected a decrease in triglyceride levels. Lower triglyceride levels do help to explain the reduction detected in apoB synthesis and secretion. However, why we are seeing this decrease in triglycerides is difficult to determine. The decrease in triglyceride levels is small and, therefore, the impact on overall apoB secretion is also small and is not believed to be the major cause for the lower levels of apoB secreted. In addition, when oleic acid was added, apoB secretion could not be restored in the replicon-expressing cells, suggesting that the decrease in apoB secretion is likely due to low MTP activity rather than a lack of available triglycerides. Further investigation is currently underway to better understand the impact free fatty acids have on apoB secretion in the context of the HCV subgenomic replicon.
The production by the liver of apoB-containing lipoproteins is also regulated by the relative amount of de novo synthesized apoB that is translocated into the ER and assembled into a secretion-competent lipoprotein (14,44,58,59). The rate of VLDL secretion by the liver is closely linked to the expression and activity of MTP (51)(52)(53)56). The HCV core protein inhibits the MTP-mediated transfer of triglycerides to the growing nascent apoB via an unknown mechanism (40). Our results provide evidence that the expression of the HCV nonstructural proteins from the HCV subgenomic replicon reduces MTP activity, providing evidence for the reduction in apoB secretion in these cells. Because apoB is secreted as a component of triglyceride-rich lipoprotein particles, inhibition of MTP would therefore decrease the addition of triglycerides to apoB-containing particles, thus reducing secretion. In addition, unlike the core protein, which does not appear to impact MTP expression levels and has no direct association with MTP or apoB, we have demonstrated that the expression of the subgenomic replicon leads to dramatic reduction of the transcript levels as well as the promoter activity of MTP. It is possible that the repression of the MTP promoter is not due to direct association with an HCV nonstructural protein but involves the interaction of a nonstructural protein(s) with other factors involved in MTP transcription. This hypothesis is supported by the observations that in the cured HCV cell lines the MTP promoter activity is restored, demonstrating the regulatory role of HCV nonstructural proteins in the observed phenomenon.
The HCV nonstructural protein NS5A is a multifunctional phosphoprotein that is believed to have various functions in HCV replication, interferon resistance, and HCV pathogenesis (4,33). It has recently been determined that NS5A associates with apolipoproteins AI and AII (33,55). These apolipoproteins regulate hepatic lipase-mediated hydrolysis of triglycerides and phospholipids in high density lipoprotein particles. In addition, previous studies by others have suggested an association of NS5A with the ER and the Golgi apparatus (33). NS5A co-localizes with the HCV core protein on cytoplasmic lipid droplets (33). We have now further implicated NS5A in contributing to altered lipid metabolism by establishing an interaction between NS5A and apoB at the C-terminal region of NS5A. At this time it is not known if this interaction is direct or indirect via other proteins or lipids; however, we were able to detect this interaction in the absence of other HCV proteins and were not able to detect any interaction between apoB and NS4 in FCA4 and K2040 cells. The interaction of apoB and NS5A does establish a connection between lipid secretion and HCV nonstructural proteins in the context of the subgenomic replicon.
In summary, this study demonstrates the ability of HCV NS proteins to alter the course of apoB synthesis, assembly, and secretion pathways in Huh-7 cells. At the cellular level there was a significant and notable reduction of apoB secretion, whereas at the molecular level MTP mRNA and MTP activity were greatly diminished. The results shed light on the potential role of HCV gene expression in modifying lipid export and contribute to the spectrum of liver disease pathogenesis associated with chronic hepatitis C.