Apolipoprotein B100 quality control and the regulation of hepatic very low density lipoprotein secretion

Eric Fisher,Elizabeth Lake,Roger S McLeod

Biochemistry&Molecular Biology,Faculty of Medicine,Dalhousie University,Halifax,NS B3H 4R2,Canada.

Apolipoprotein B100 quality control and the regulation of hepatic very low density lipoprotein secretion

Eric Fisher,Elizabeth Lake,Roger S McLeod✉

Biochemistry&Molecular Biology,Faculty of Medicine,Dalhousie University,Halifax,NS B3H 4R2,Canada.

Apolipoprotein B(apoB)is the main protein component of very low density lipoprotein(VLDL)and is necessary for the assembly and secretion of these triglyceride(TG)-rich particles.Following release from the liver,VLDL is converted to low density lipoprotein(LDL)in the plasma and increased production of VLDL can therefore play a detrimental role in cardiovascular disease.Increasing evidence has helped to establish VLDL assembly as a target for the treatment of dyslipidemias.Multiple factors are involved in the folding of the apoB protein and the formation of a secretion-competent VLDL particle.Failed VLDL assembly can initiate quality control mechanisms in the hepatocyte that target apoB for degradation.ApoB is a substrate for endoplasmic reticulum associated degradation (ERAD)by the ubiquitin proteasome system and for autophagy.Efficient targeting and disposal of apoB is a regulated process that modulates VLDL secretion and partitioning of TG.Emerging evidence suggests that significant overlap exists between these degradative pathways.For example,the insulin-mediated targeting of apoB to autophagy and postprandial activation of the unfolded protein response(UPR)may employ the same cellular machinery and regulatory cues.Changes in the quality control mechanisms for apoB impact hepatic physiology and pathology states,including insulin resistance and fatty liver.Insulin signaling,lipid metabolism and the hepatic UPR may impact VLDL production,particularly during the postprandial state.In this review we summarize our current understanding of VLDL assembly,apoB degradation,quality control mechanisms and the role of these processes in liver physiology and in pathologic states.

apolipoprotein B,very low density lipoprotein,proteasome,ubiquitin,autophagy,steatosis

APOLIPOPROTEIN B AND THE ASSEMBLY OF VERY LOW DENSITY LIPOPROTEINS

Apolipoprotein(apo)B provides the structural framework for the assembly of triglyceride(TG)-rich lipoproteins in the liver and the intestine(reviewed in[1-8]).Each apoB-containing lipoprotein particle contains a single molecule of apoB,either apoB100,or apoB48(the N-terminal 48%of apoB1002).Increased hepatic secretion of apoB or decreased removal from the plasma can increase the plasma concentration of apoB[9],leading to an increased risk for developing atherosclerosis[10-13].Therefore,interventions aimed at reduction of particle number may be even more important than reduction of the lipid contents(including cholesterol)of the particles[14,15].

Although the details of assembly and secretion of apoB-containing lipoproteins are not completelydefined,three components are recognized as necessary for efficient assembly:apoB,an adequate supply of lipids and the microsomal TG transfer protein (MTP)[3,16,17].ApoB is unique in its ability to associate with neutral lipids from cellular stores,and high levels of fatty acid substrates for TG biosynthesis can increase hepatic very low density lipoprotein(VLDL) assembly[18-23].Mutations in the amino terminus of apoB that prevent the initiation of assembly also reduce hepatic VLDL secretion[24,25].Recent studies showed that reduction of apoB mRNA by antisense oligonucleotide(ASO)techniques reduced VLDL production in HepG2 cells[26],in mice[27,28],in cynomolgus monkeys[29]and in humans[30-32].Thus,gene silencing approaches to treat dyslipidemias at the level of hepatic VLDL assembly have now reached clinical application[33].

Assembly of hepatic VLDL has been the subject of intense research for more than 25 years and experiments in hepatoma cells in culture have added greatly to our knowledge of liver biology.When exogenous fatty acid(usually oleate)[34-37]is provided in the culture medium,rat hepatoma cells assemble and secrete authentic VLDL.This model has suggested that VLDL assembly occurs in two steps,one co-translational(the first step)and another post-translational (the second step)[36,38].The larger amount of TG is added during the second step,perhaps via fusion of a primordial apoB lipoprotein particle with an apoB-free TG droplet in the secretory pathway[39,40].Although this is an attractive mechanism,it is also recognized that assembly in primary hepatocytes[41,42]may proceed without two discrete steps.

The microsomal TG transfer protein,MTP,facilitates apoB secretion[43-46]in part by addition of lipid to the polypeptide as apoB translocates into the lumen of the ER(Fig.1A).The initial association of apoB with phospholipids may occur independently of MTP as MTP is not required for the secretion of the N-terminal apoB18[47].MTP is particularly important for the translocationoffull-lengthapoB100,but maynot berequired for translocation of the shorter apoB48 that is also synthesized by rodent hepatocytes[48-50].Nevertheless, this does not exclude a role for MTP in the lipidation of apoB48.MTP appears to be required for VLDL maturation[35,51,52]and can be found in the Golgi apparatus in addition to the ER[53],consistent with a post-ER second step of lipidation[54].MTP is subject to diurnal regulation[55]via the genes Clock and Shp.Mutation of these genes causes hypertriglyceridemia in mice,since Shp normally suppresses the MTP promoter and this is released when Clock is absent.Shp levels change in a diurnal manner and disruption of this cycle can cause increased VLDL output by the liver.Perhaps this is due to a mismatch in the normal feeding-fasting cycle with MTP levels.In apoB100-only ob/ob mice,adenovirusmediated overexpression of MTP stimulated an increase in VLDL-TG secretion without increasing apoB100 secretion[56].This suggests that MTP allows greater incorporation of TG into existing apoB-containinglipoproteins,thus yielding larger particles,rather than increasing the number of secreted particles.MTP can also transfer cholesteryl ester(CE)from its site of synthesis in the ER(by ACAT1,ACAT2)onto nascent VLDL and thereby relieve product inhibition to enhance cholesterol esterification[57].

MTP became an obvious pharmacologic target in attempts to reduce hepatic lipoprotein production. However,one of the concerns associated with pharmacological inhibition or knockout of MTP has been that this can cause hepatic lipid accumulation.Liver-specific knockout of MTP caused hepatic steatosis but did not affect insulin sensitivity or glucose tolerance[58]. Selective inhibition of the TG transfer activity of MTP has been suggested as an alternative approach to this type of therapy[59].These authors restored MTP in knockout mice using adenoviruses with either drosophila MTP,which transfers only phospholipids, or with human MTP,which transfers both TG and PL.Both of the constructs restored some of the plasma lipoproteins and both attenuated liver TG accumulation by stimulating β-oxidation without enhancing lipogenesis.This experiment suggests that PL transfer alone may be sufficient to restore a low level of lipoprotein assembly.Thus,selective reduction of TG transfer may be an alternative approach to reducing VLDL output.MTP inhibition leading to TG accumulation in liver may not always occur with knockdown of the MTP mRNA.Combined therapy with MTP and DGAT2 siRNA were shown to reduce both plasma cholesterol and hepatic TG levels in mice[60].

VLDL assembly is regulated by apolipoproteins other than apoB[61],by lipid availability[62]and by signal transduction mechanisms reporting on organismal nutrition[8,63,64].This review will focus on the elements of apoB regulation that affect protein folding,and the consequences of failed assembly.In our view,this process is in itself regulatory,and only becomes obvious when the fidelity of substrate production is compromised.The overwhelming evidence supports a surveillance process that repeatedly samples the assembly environment and only triggers apoB intracellular destruction after repeated attempts at proper assembly have failed.Indeed,we believe that the studies of apoB degradation have revealed a process at the interface between physiological regulation of afunctional lipid transport entity and pathological destruction of a potentially cytotoxic macromolecular aggregate.Hepatic lipotoxicity is relevant to fatty liver disease,hepatic insulin resistance,inflammation and impaired ER function.Over the course of this review we will examine the consequences of apoB production in various models and speculate on the role of apoB in these assorted pathologies.

Fig.1ER-associated degradation of ApoB.A:Cotranslational lipid recruitment.Following engagement between the ribosome and translocation channel,apoB crosses the ER bilayer and enters the ER lumen.There,the growing polypeptide begins to acquire lipid ligands,forming a nascent lipoprotein.Lipidation of apoB is an MTP-dependent process.MTP is present in the lumen as a heterodimer with PDI.The ER chaperone BiP is associated with apoB regardless of apoB lipidation status.Efficient translocation allows the growing apoB peptide to begin forming a secretioncompetent lipoprotein.Additional protein folding machinery associates with nascent apoB,as discussed in the text.B:Translocation arrest and substrate selection.If the supply of lipid ligands is inadequate to sustain the assembly of this apoB-containing lipoprotein,translocation of apoB into the ER lumen becomes delayed,or arrested.When inefficient lipidation occurs,hydrophobic regions of lumenal apoB may be exposed to the aqueous environment in the lumen.This leads to further BiP binding(and other luminal chaperones),prolonged exposure of apoB to the inside of the translocation channel,and sudden exposure of apoB motifs to the cytosol(‘‘looping out’’).Cytosolically exposed apoB associates with cytosolic chaperones such as Hsp70 and Hsp90.With regards to substrate selection for ERAD,one,two or all three of these events may contribute to triggering downstream events in ERAD.C:Ubiquitination and retrotranslocation of apoB.p97 has been shown to facilitate the proteasomal degradation of apoB. Gp78(and possibly other ubiquitin ligases,including Hrd1)mediates the ubiquitination of apoB.The presence of a‘‘threshold’’number of ubiquitin moieties on apoB likely provokes the p97 complex to act on the target substrate.There may exist‘‘sub-ERAD’’levels of apoB ubiquitination,either in the form of short ubiquitin chains,or low number of chains per polypeptide.p97-dependent retrotranslocation of ERAD substrates appears to require temporary deubiquitination of the target peptide.The substrate then becomes ubiquitinated once more in order to stimulate further proteasomal substrate trafficking and processing events.It is unknown through which channel/mechanism apoB is removed from the ER,nor is it known which(if any) cytosolic ubiquitin‘‘shuttling’’proteins are involved in delivering apoB to the proteasome.D:Retrotranslocation via ER-associated lipid droplet.A secondary ERAD pathway has been for shown for apoB.It depends on MTP activity and the formation of a primordial apoB-containing lipoprotein.‘‘ApoB crescents’’have been described as an intersection of proteasomal and autophagic degradation of apoB.Movement of apoB from the ER membrane to the ER-associated lipid droplets requires Derlin-1.The membrane-associated hairpin protein Ubxd8 was found to mediate the extraction of apoB from these lipid droplet-associated crescents by p97.It is a remarkable discovery that lipidated apoB,found in an abnormal,cytosolic lipid dropletassociated state,is extracted in linear form and can be processed by the proteasome.

APOLIPOPROTEIN B PRODUCTION AND DEGRADATION

Factors promoting lipid availability during VLDL assembly

The pathways of neutral lipid mobilization for assembly into VLDL are still poorly understood. Lehner and colleagues have recently reviewed luminal lipid metabolism and lipoprotein assembly[62]and Sturley and Hussain have reviewed the formation of lipid droplets on either the cytoplasmic or luminal side oftheERmembrane[65].Thisisaveryactiveareaofcurrent investigation,which will only be summarized here.

Several laboratories have explored the roles of TG hydrolases in mobilization of TG for VLDL assembly. Caviglia[66]showed that the TG hydrolase ABHD5/ CGI-58 is involved in the mobilization of stored TG for VLDL.Knockdown of the enzyme in McARH7777 cells using RNAi reduced apoB,TG and CE secretion.The secreted TG were enriched in saturated fatty acids at the expense of oleic acid,suggesting that this hydrolase may be involved in selective fatty acid mobilization.Ces-1/Es-x is another member of the TG hydrolase family that appears to be involved in the partitioning of regulatory fatty acids[67].Knockout of this gene in mice increased hepatic lipogenesis and increased secretion of VLDL.The knockout animals developed obesity,fatty liver,and insulin resistance,even on a chow diet.Furthermore,release of polyunsaturated fatty acids(PUFAs)from intracellular TG stores was decreased and sterol regulatory elementbinding protein(SREBP)1c-mediated lipogenesis was increased.Provision of dietary n-3 fatty acids was able to modulate the increased lipogenesis in this model.Ko et al[68]showed in McA-RH7777 cells that Ces1/Es-X prevents TG accumulation by enhancing β-oxidation, without having an effect on secretion of apoB and TG.Ectopic expression of hepatic lipase in the ER of McA-RH7777 cells decreased the incorporation of fatty acids into cellular TG and CE and decreased the secretion of apoB-containing lipoproteins[69],although the physiologic significance of this observation is not yet clear.The modified hepatocytes displayed an increase in PPARα expression and enhanced β-oxidation,suggesting that fatty acids generated by hydrolysis of TG in the ER may upregulate cellular utilization of energy.Thus,studies in these model systems indicate that intracellular lipolysis of TG can modulate lipid deposition and apoB secretion from hepatocytes.

Adipose TG lipase(ATGL)generates many of the FFAs found in the liver and may contribute to the characteristics of the ER stress response[70].In this study,ATGL knockout or wild-type mice were treated with tunicamycin to induce ER stress.ATGL knockout animals accumulated more TG than wild-type animals even though lipogenesis,β-oxidation and VLDL secretion were decreased by tunicamycin in both groups.The knockout of ATGL prevented the increase in several ER stress and inflammatory markers,while fatty acid analysis indicated that absence of ATGL prevented the increase in the PA:OA ratio that was found in the wild-type mice.Thus,it appears that ER stress-associated changes in the hepatic ER fatty acid profile may contribute to inflammation.

In rats,overexpression of PGC-1α increased cellular mitochondrial content and function[71].This was also observed in isolated hepatocytes,where increased fatty acid oxidation was detected.Consistent with a decrease in TG storage and secretion,apoB and MTP expression were also reduced.Thus,PGC-1α appears to regulate fatty acid availability for VLDL secretion through its effects on mitochondrial β-oxidation.

CideB is an ER and cytosolic lipid droplet(CLD)-associated protein that is involved in TG mobilization for VLDL assembly.When CideB was depleted in mice,hepatic TG was increased and VLDL secretion was reduced.In addition,the secreted VLDL particles were TG-poor[72].Restoration of VLDL output was achieved by expression of CideB containing both ER and LD-associating domains,but not when either domain was absent.CideB also interacts directly with apoB,suggesting an integral role for CideB in providing effective lipid availability for assembly[72]. Recently,it was reported that CideB-null mice also displayed marked upregulation of hepatic perilipin 2/ ADRP[73],which promotes storage,rather than mobilization,of TG in cytosolic lipid droplets.Knockdown of perilipin 2 in this model enhanced VLDL production and reduced TG accumulation,suggesting opposing roles for CideB and perilipin 2 in regulating hepatic lipid availability for VLDL assembly[73].Gain and loss of function analysis has also established a role for the phosphatidic acid phosphohydrolase lipin-1 in hepatic TG metabolism[74].Loss of lipin-1 function did not impair TG synthesis,nor did overexpression enhance TG synthesis.However,increasing lipin-1 decreased TG secretion and knockout increased TG secretion, suggesting that this enzyme plays a role in VLDL assembly,perhaps via modulation of TG availability.

Several of the soluble apolipoproteins have been implicated in the regulation of VLDL assembly and secretion,although the mechanisms of these effects are not yet clear.Studies in humans have indicated thatapoCIII plays a central role in the dysregulation that leads to hypertriglyceridemia and small dense LDL[75].Particles secreted from the liver are enriched in apoCIII and deficient in apoE in hypertriglyceridemic subjects.These particles are also cleared more slowly from the plasma and the LDL particles that result from their catabolism are more dense,and by definition more atherogenic.Overexpression of apoCIII in transgenic micealsoincreasedhepaticTGlevels andinducedinsulin resistance[76].Sundaram and colleagues showed that apoCIII has an intracellular role in VLDL assembly[77]. In McA-RH7777 cells they showed that overexpression of apoCIII increased apoB and VLDL secretion,but only in culture conditions where TG levels in the cells were also elevated.In this model system the enhancement was accompanied by an increase in the expression of MTP,and suggested that apoCIII was involved in the second step of lipidation.

ApoAIV has also been implicated in the assembly of VLDL[78].In a model system where the apoAIV was trapped in the ER by appending to it the KDEL retention signal,secretion of apoB and TG were decreased. Interestingly,apoAIV increased the amount of secreted TG per particle without affecting MTP expression, suggesting that apoAIV may modulate apoB trafficking,allowing the particles to acquire additional TG. Plasma apoAII may also affect VLDL production rate[79].Genetic polymorphisms in the apoA5 promoter that affect TG metabolism have been associated with coronary artery disease[80].Thus,several of the soluble apolipoproteins may have an influence on VLDL production,but exactly how is not yet clear.

It has been known for some time that there is an inverse relationship between plasma HDL-C levels and TG levels in humans,and this is particularly apparent in the ABCA1-deficient state.The effect of HDL metabolism on hepatic VLDL production has been explored recently.Chung et al.[81]used a gene silencing technique to reduce ABCA1 levels in McARH7777 cells,which caused the accumulation of large HDL in the medium.This also caused an increase in the secretion of TG-rich VLDL1,as a consequence of a decrease in the activation of PI3 kinase.These changes were corrected by the addition of normal HDL to the cell culture,suggesting that it is the characteristics of the HDL that affects the response of the hepatocyte.Wiersma showed that HDL can be taken up by hepatocytes in culture and components of HDL can be secreted in VLDL[82].By overexpression or knockout of SRBI,VLDL-apoB and VLDL-TG secretion were increased or decreased,respectively. This was also found in vivo,where changes in MTP, SREBP-2 and liver cholesterol content were consistent with the changes in VLDL production.Thus,there is accumulating evidence of an effect of HDL metabolism on liver production of atherogenic lipoproteins.

Cellular trafficking events play an integral role during the assembly and secretion(or degradation)of nascent VLDL.Willnow has reviewed the role of sortilins in lipoprotein metabolism[83].Sortilin(Sort1)is located on chromosome 1p13.3 and SNPs in this locus have been associated with hypercholesterolemia and myocardial infarction.Sort1 is an intracellular sorting receptor for apoB in the Golgi apparatus and knockout of Sort1 reduces the secretion of apoB-containing lipoproteins from the liver.This resulted in decreased plasma cholesterol levels and reduced atherosclerotic lesions in LDLr-null mice[84].Conversely,overexpression of Sort1 increased plasma cholesterol and apoB-lipoprotein secretion.

The role of Sort1 and ER stress has been further explored by Ai and colleagues[85].In contrast to the previous work these authors found that Sort1 expression was decreased in obese mice and that restoration of Sort1 expression decreased hepatic TG and apoB secretion.In addition,inhibition of mTOR reduced the level of ER stress and increased Sort1 expression, suggesting that induction of ER stress through the mTOR signaling pathway may contribute to dyslipidemia in obese states.Strong and colleagues have shown that sortilin reduces apoB secretion by binding to the secretory protein in the Golgi apparatus and delivering apoB to the lysosome for degradation[86].

Insulin stimulates an interaction between Sort1 and apoB via PI3 kinase in McA-RH7777 cells[87],suggesting that Sort1 may be involved in the normal insulin response in the hepatocyte.A direct interaction between Sort1 and apoB was detected by inhibition of lysosomal degradation,which also increased apoB secretion.This study suggests that the Sort1-apoB interaction and lysosomal degradation play roles in the insulin-dependent regulation of apoB secretion.

Trafficking of apoB inthesecretorypathwaymay also be impacted by PCSK9[88],as overexpression of PCSK9 in multiple mouse models increased plasma apoB,cholesterolandTGduetoincreasesinbothLDLandVLDL. The authors suggested that interaction of PCSK9 with apoB in the hepatocyte may inhibit autophagy.In the secretory pathway PCSK9 may shuttle between the LDL receptor and apoB,thereby coordinately regulating lipoprotein import and export.

In intestinal Caco-2 cells,overexpression of Sar1b increased chylomicron production and enhanced the esterification of fatty acids into all lipids[89].ApoB48 synthesis was increased as were the activities of MGAT,DGAT and MTP.Sec23 interactions withSar1b were demonstrated by immunoprecipitation. Since these proteins in turn interact with SCAP and SREBP1c,these complexes may be implicated in the enhancement of lipid biosynthesis.Although these observations may not be directly relevant to hepatic apoB metabolism,complementary pathways may be present in the hepatocyte and the enterocyte.

Mechanism(s)of apoB degradation

Although changes in the levels of apoB transcript[90]and translation of the mRNA[91-93]can play a role in apoB production,secretion of apoB is primarily regulated at the level of protein degradation[94].Degradation of apoB can occur either co-translationally or after the full-length protein has entered the secretory pathway. Whether apoB is assembled into a secretory lipoprotein particle or is degraded before reaching the extracellular space depends on the interaction between apoB and components of the assembly and degradation systems. Degradation may be initiated when part of the apoB protein is recognized as malfolded by the quality control machinery of the ER due to incomplete lipidation, initiating delivery of the protein to the cytosolic proteasome[95].The proteasome is often involved in the removal of abnormally folded cellular proteins in the process termed ER-associated degradation(ERAD)[96]. Although it may seem wasteful that normal,functional proteins such as apoB should be degraded before performing their primary function,perhaps as much as 25% of newly synthesized polypeptides may be removed by this surveillance system[97,98].Surveillance can remove nascent proteins without initiating ER stress and the unfolded protein response(UPR)[99],which can compromise cell function and jeopardize cell survival[100-102]. ApoB can be degraded by autophagy when the nascent lipoprotein fails late-stage quality control surveillance and in response to hepatic insulin signaling[103].Hepatic autophagy is highly integrated with lipid homeostasis[104]. In addition,selective autophagy removes and degrades ERAD-related protein complexes from the ER,by newly emerged mechanisms[105,106].

Since the first demonstration of apoB degradation in 1987[107],an enormous amount of effort has been expended in the study of this process.However,it was not until nearly 10 years later that the involvement of the cytosolic proteasome was demonstrated[108]. Much of the evidence for the role of the proteasome in apoB degradation has been derived from studies of the HepG2 cell because degradation of apoB by the ubiquitin-proteasome pathway is particularly active in this cell line due to inefficient VLDL assembly.Although this cell model does have shortcomings,other human liver cell lines do not appear to have significant advantages over HepG2[109,110].ERAD of apoB is not unique to the HepG2 cell line and can also be demonstrated in the rat hepatoma McA-RH7777[111].In primary hepatocytes, however,the role of the proteasome is less clear because the proteasome inhibitors(MG132,lactacystin and ALLN)used to demonstrate the involvement of the proteasome can be toxic to these cells[42].

Although apoB secretion is modulated by intracellular degradation,this may not always be equivalent to regulation.It is important to determine whether quality control surveillance machinery in the ER targets and destroys apoB in a reactive manner,in response to‘‘malfolding’’events during VLDL assembly,or whether apoB is removed by degradation via a proactive,and therefore potentially regulated process. Since multiple coordinated events are required for apoB to achieve its‘‘native’’,secretion-competent conformation(including sustained translocation for the duration of apoB translation,disulfide bond formation and rearrangement,glycosylation,lipid ligand acquisition and intracellular trafficking)it is clear that the QC machinery will be reactive to the fidelity of these biosynthetic events.However,the modulation of apoB production in response to whole body metabolism in changing nutritional states suggests that there are proactive surveillance mechanisms that must remove nascent apoB when curtailing VLDL output is warranted.The remainder of this section will review apoB quality control and degradation mechanisms.

Luminal and cytosolic triggers for proteasomal degradation of apoB

Conditions that limit the assembly of VLDL such as reduced availability of lipid ligands[112]or MTP inhibition[113]can trigger the proteasomal degradation of apoB.The translocation efficiency of apoB is tied to its necessity for co-translational acquisition of lipid ligands.When lipidation is inadequate,apoB can undergo‘‘translocation arrest.’’Considerable evidence supports the hypothesis that proteasomal degradation occurs as a consequence of apoB translocation arrest, causing regions of the polypeptide to become exposed to the cytosol[114-118].Admittedly,this may reflect the relatively low efficiency of assembly in hepatoma cell systems.ApoB may exist transiently in a bitopic topology(exposed to both the ER lumen and the cytosol), simultaneously accessible to both assembly and degradation pathways[117].Jiang and colleagues[119]have shown that the presence of MTP during apoB translation prolongs the window for lipid recruitment in vitro.Three‘‘lesions’’likely occur simultaneously,or in extremelyclose succession:(1)exposure of poorly lipidated apoB regions in the ER lumen,(2)prolonged residence in the translocon,and(3)exposure of apoB motifs to the cytosol(depicted inFig.1B).Given its bitopic topology,translocation arrested apoB may even behave as an ERAD substrate that is membrane-integrated,despite the final product of VLDL being a luminal,secreted lipoprotein[120].These characteristics of apoB are important upon consideration that mechanisms of ERAD substraterecognitionandprocessingareoftendeterminedby lesion location;specifically whetheran unfoldedor malfolded segment of the polypeptide is in the ER lumen (ERAD-L),withinthe bilayer(ERAD-M)orinthecytoplasm(ERAD-C)[121,122].

The ER chaperones BiP/Grp78,calreticulin,PDI and Grp94 have been found crosslinked to apoB in HepG2 cells,regardless of apoB lipidation status[123].However, the ER-resident chaperone BiP was reported to increase binding to apoB when the interaction between nascent apoB and MTP was disrupted[124].BiP is considered a principal indicator of ER stress,as its levels increase incrementally via transcriptional upregulation.Mild to moderate levels of fatty acid-induced ER stress increase both BiP levels and VLDL assembly[125].Therefore,it seems that BiP levels may not regulate apoB ERAD but more likely contribute to stabilizing poorly lipidated apoB motifs prior to their removal from the ER.Low, medium and high levels of BiP do not correlate linearly, with respect to substrate fate.However,it appears that prolonged apoB-BiP binding may facilitate the ERAD of apoB[124].

There are ten different PDI species in mammals that are critical to the differences in ERAD substrate recognition[126].This study in yeast,which have 5 PDIs,showed that PDI-1 interacted with exogenously expressed apoB proteins and its chaperone activity facilitated apoB degradation.In contrast,other ERAD substrates were not dependent on PDI-1 but were dependent on the redox activity in the ER lumen.Thus,chaperone interactions withintheERcandeterminesubstratefates.Itiscurrently unclear whether apoB is a client of the N-glycan sugar timing system that governs glycoprotein quality control[102].WhileapoBdoesassociatewithcalnexinandcalreticulin under‘‘normal’’conditions,the difference between secretion and degradation of apoB is primarily determined by lipid availability.This does not preclude, however,the possibility that lipid-dependent folding events of the apoB protein might be connected to a glucose-trimming mechanism of quality control.ApoB also associates withthe ER luminal protease ER-60 in HepG2 cells[127].Subsequent studies of ER-60 overexpression and proteolytically inactive ER-60 suggested that ER-60 may cleave apoB translocation-arrested apoB in the ER lumen[128,129].It is possible that luminal cleavage events are coordinated with retrotranslocation of poorly lipidated apoB polypeptides into the cytosol.Studies in HepG2 cells have revealed that the N-terminal region of apoB may not be removed from the ER into the cytosol[130],but may instead be cleaved and then either degraded or secreted[131].

Inefficiently translocated apoB is associated with the cytosolic chaperones Hsp70[118]and Hsp90[132].It appears that cytosolic heat shock proteins are necessary to support degradation.Since increased Hsp70 expression enhances apoB ubiquitination when the lipid supply is unaltered,apoB may be stabilized by Hsp70 in the cytosol during stages where it is subject to quality control mechanisms.There are likely other ubiquitin proteasome system proteins in the cytosol that bind to apoB as well,beyond those of the heat shock family.

In some experimental systems,apoB chains that were degraded by the proteasome had already translocated into the ER lumen and interacted with calnexin[133],implying that they had attained full-length and acquired significant glycosylation prior to degradation.Nevertheless,if translocation arrest cannot be demonstrated in vivo,apoB can still be degraded from within the secretory pathway.Inhibition of the MEKERK pathway[134]increases the efficiency of VLDL assembly in HepG2 cells and decreases apoB ubiquitination[135].In our experiments,cellular apoB is largely exposed to the cytosol[114,136]while smaller recombinant apoB proteins(such as apoB29)are not.Fusion proteins containing apoB29 and the apoB segment from B37-42 were partially arrested[114].The β1 domain (defined as between the C-termini of apoB22 and apoB42)has been implicated in translocation arrest, susceptibility to proteasomal degradation and LpB assembly[116,137].Our studies suggest that not all β1 regions increase the susceptibility of apoB to degradation,but that specific regions of the β1 domain (including B37-42)may mediate proteasomal degradation and core lipid recruitment[114].These unique sequences in apoB may mediate proteasomal degradation by increasing exposure to the cytosol or by slowing translocation sufficiently to initiate the ERAD. Intriguingly,intra-membrane cleavage of ERAD substrates has emerged as a novel proteolytic mechanism, and this process may yet have relevance to apoB[138].

Ubiquitination of apoB

ApoB could be targeted to ERAD by multiple ubiquitin ligases and with more than one means of escape from the ER.Overexpression of the ubiquitin ligase gp78 in HepG2 cells caused increased ubiquitinationand degradation of apoB combined with a requisite decrease in secretion[139].Conversely,we observed apoB secretion to increase upon modest,siRNA-mediated depletion of gp78 in HepG2 cells[135]. Furthermore,our knockdown of gp78 paradoxically shifted the density of secreted apoB-containing lipoproteins from HepG2 cells into the VLDL range,away from typically observed majority of LDL.Thus,in this model,ubiquitination is not only responsive to lipid recruitment during assembly but influences the ability of nascent apoB as it progresses through the secretory pathway.ApoB has also been found to associate with the ER-resident E3 ligase Hrd1,as well as the purported retrotranslocation mediator Derlin-1[124].Full length apoB is degraded by the proteasome,as are co-translationally truncated apoB polypeptides of many different sizes.E3 ligase overlap and/or cooperativity is likely,and not without precedent.The cell may employ different machinery to detect and target these distinct apoB substrates to ERAD.The specifics of apoB ubiquitination are poorly understood and warrant further investigation.

Retrotranslocation of apolipoprotein B

As the field of ERAD and substrate retrotranslocation advanced,so has our understanding of apoB as an ERAD substrate.For some time there has been an active(and technically challenging)search to identify a‘‘retrotranslocon’’[140].What has resulted from this search is an increased understanding of ER function, and an appreciation of the diverse nature of ERAD substrates and the means by which they are processed. Some substrates can be threaded through retrotranslocons as linear polypeptides,while others appear to move through to the cytosol while still folded or in multimeric complexes[141].Presented below are data pertaining to apoB retrotranslocation and other relevant recent findings that have enhanced our understanding of the ER.

Partially translated apoB associates with both the Sec61 translocon and the proteasome[142],suggesting that retrotranslocation of arrested apoB could occur through the same pore that mediates translocation.Ubiquitinated apoB can found in the Sec61 complex in HepG2 cells[143]. It is tempting to speculate that the‘‘hydrophobicity’’threshold of the hydrophobic core in the Sec61 channel may play a role in apoB quality control[144].This attribute of the Sec61 core was described in the context of nascent transmembrane domain integration into the bilayer.As apoB undergoes translocation arrest,its densely hydrophobicepitopescouldtriggerasimilarprocessthatmoves apoB laterally into the ER bilayer,perhaps into an ERAD-dedicated multimeric complex.

The ER-membrane spanning proteins Derlin-1 and the ubiquitin ligase Hrd1 facilitate movement of ERAD substrates to the cytosol,yet it has been unclear whether theseproteins form a putative retrotranslocation channel themselves,or are required in another manner. Studies in mammalian cells and C.Elegans suggested that Derlin-1 proteins served as components in a retrotranslocation channel[145,146].More substrates for Derlin-1-dependent retrotranslocation have since emerged, lending support to this route out of the ER[147-149].A unique photocrosslinking technique was used to show that amino acids of the Hrd1 protein within the transmembrane domain come into direct contact with ERAD substrates[150,151].The protein Usa1(known as HERP in mammals)supports the interaction of multiple Hrd1 proteins,which may in turn facilitate the formation of relatively large,dynamic retrotranslocation channels capable of moving partially folded protein cargo through to the cytosol[152].Recently,photocrosslinking and mutagenesis of yeast Der1 revealed that‘‘recognized’’substrates are transferred from luminal receptors Hrd3 and Yos9 and moved through the ER membrane in a manner that requires transmembrane domains of Der1.Further, Der1TMdomainsfacilitatetheubiquitinationofthesubstrate(CPY*)in the cytosol by the Hrd1 ubiquitin ligase[153].ApoB has been shown to associate with mammalian Derlin-1 and Hrd1 at the ER in HepG2 cells,yet their requirement for nascent apoB ERAD is unclear[124].

The homo-hexamer AAA-ATPase p97 has been shown to facilitate removal of ubiquitinated apoB from the ER for proteasomal degradation(seeFig. 1C)[124,154].Reduction of p97 by siRNA in HepG2 cells delayed apoB turnover and accumulated apoB at the ER,while apoB secretion efficiency was unaltered. The effect of the p97 knockdown on apoB could be abolished by either knocking down the ubiquitin ligase gp78 or enhancing VLDL assembly pharmacologically[135].While these studies revealed a cotranslational role for p97 in apoB metabolism,Suzuki and colleagues have found an alternative,lipid droplet-dependent pathway for retrotranslocation of lipidated‘‘apoB-crescents’’in Huh7 hepatoma cells[155].Reduction of Derlin-1 caused apoB to accumulate only in the ER lumen,while depletion of tail-anchored protein Ubxd8 caused apoB to accumulate on ER-associated lipid droplets and the ER.This apoB accumulation was suppressed by inhibition of MTP activity,suggesting that these apoB ERAD substrates require at least partial lipidation to enter this particular pathway. Remarkably,Ubxd8 recruits p97 for the extraction of apoB from lipid droplets for proteasomal degradation (Fig.1D)[155].Membrane-embedded domains of Derlin-1 associate directly with ERAD substrates andthe ubiquitin ligase Hrd1,as shown by site-specific photocrosslinking[153].Thus,it is possible that Hrd1 participates in processing apoB substrates as part of this ERAD protein complex.The role of p97 in apoB ERAD now appears to be twofold:retrotranslocation of nascent apoB from the ER and extraction of partially-lipidated apoB from lipid droplets.These are two distinct subclasses of apoB ERAD substrates that share some overlapping machinery.

Ubxd8 has been shown to regulate lipid droplet homeostasis in yeast[156].Unsaturated fatty acids cause Ubxd8 to dissociate from ubiquitinated Insig-1 and oligomerize,which in turn increases Insig-1 levels at the ER and prevents the activation of SREBP.However, saturated fats do not provoke this response,resulting in inadequate regulation of lipids.The inability of Ubxd8 to‘‘sense’’saturated fats has been proposed as a means for hepatic metabolism dysregulation in the face of a Western-style diet[157].Given that the Scap/SREBP proteins are essential for the progression of diabetic fatty liver[158]it would be fascinating to know whether Ubxd8 modulatesneutrallipidavailability forapoB production in a pathological setting,in addition to its involvement in trafficking apoB through the cytoplasmic lipid droplet(CLD)compartment during ERAD.

In addition to serving as storage organelles for excess fatty acids and cholesterol,which are stored as TG and cholesterol esters,respectively[159],CLDs are also protein storage depots[160].Lipid droplets provide an intersection of lipid metabolism and protein quality control.Ancient ubiquitous protein 1(AUP1) may facilitate the delivery of ERAD substrates to lipid droplets en route to the proteasome[161],while Ubxd8-depleted Huh7 cells accumulate many different ubiquitinated proteins in addition to apoB[155].HMG CoA reductase is delivered to lipid droplets after it is dislocated from the ER membrane and before it is degraded by the proteasome[162].

How the biogenesis of lipid droplets is organized to support both assembly and degradation of apoB is unclear.At some point between substrate recognition and degradation,non-exchangeable apoB polypeptides must be stripped of their lipid ligands,which,in vitro, bind irreversibly to sequences in the β1 domain[163]. The answer to this question may emerge from probing VLDL assembly and lipid droplet dynamics in multiple hepatocyte models.

Role of deubiquitinating enzymes in modulating ERAD substrate fate

Although it was previously presumed that ubiquitin was only used to target non-functional proteins for disposal,ubiquitin-protein conjugates are now recognized to be dynamic structures.Deubiquitinating enzymes (DUBs)are responsible for the removal of ubiquitin from substrate proteins.The human genome encodes approximately 100 DUBs in 5 families[164].Proteomic analysis of DUBs and their associated proteins by Sowa and colleagues revealed six DUBs likely to interact with the AAA-ATPase p97[165].The role of DUBs in p97-dependent ERAD has recently been reviewed[166].Dislocation of ERAD substrates requires p97 and a dimeric complex of Npl4 and Ufd1,which binds to ubiquitin in addition to p97[167].Ubiquitin chain trimming on ERAD substrates may change the affinity for ubiquitin-binding effectors.This has been suggested to contribute to the transfer of the substrate from p97 to effectors such as the Bag6 holdase complex[168]to possibly facilitate transfer from the ER to the proteasome.

Ataxin-3 is a DUB associated with the neurodegenerative disorder spinocerebellar ataxia type 3[169].Work by Zhong and Pittman in HEK293T cells suggested that ataxin-3 binds to p97 and reduces its interaction with Ufd1,therefore,causing accumulation of ERAD substrates in the ER[170].In contrast,evidence has also shown that ataxin-3 may promote p97-associated deubiquitination to facilitate the transfer of ERAD substrates to the proteasome[171].Expression of an inactive ataxin-3 mutant caused accumulation of ubiquitinated proteins associated with p97,suggesting that substrate transfer from p97 to the proteasome was blocked.

The mammalian DUB YOD1 is a component of the p97 complex by way of its UBX domain[172]. Expressionof catalyticallyinactive YOD1 causedaccumulation of a model ERAD substrate and of polyubiquitinated proteins.Interaction between YOD1 and other components of ERAD such as Derlin-1 and Ubxd8 were also reported.Recently,it was demonstrated that YOD1 also regulates the degradation of a nonubiquitinated ERAD substrate,cholera toxin A1 (CTA1)[173].Using siRNA knockdown of YOD1 and a retrotranslocation assay,they concluded that YOD1, but not other DUBs(ataxin-3 and USP14),blocks CTA1 retrotranslocation.Since CTA1 is a nonubiquitinated substrate,they suggested that YOD1 inhibits retrotranslocation by deubiquitination of components of the ERAD machinery[173].Notably,gp78-dependent ubiquitination of the ERAD-associated chaperone Bag6 cofactor Ubl4 caused inactivation of the Bag6 chaperone,while conversely the DUB Usp13 removed ubiquitin conjugates from Ub14 and preserved Bag6 function[174].Liu et al.also demonstrated that Usp13 acts on ERAD substrates downstream of p97 since deglycosylation of TCR was not affected by Usp13 knockdown.These data illustrate the emerging complexity and dynamics of ubiquitination in ERAD.

The DUB enzyme USP25 has been recently implicated in ERAD.Blount and colleagues showed that exogeneous USP25 interacts with the E3 ligase Hrd1 and with p97[175].The catalytic activity of USP25 rescued an ERAD substrate(CD3)from proteasomal degradation,while knockdown of USP25 resulted in a decrease in levels of an ERAD substrate.Deubiquitination of ER membrane proteins has been suggested to‘‘sharpen’’substrate discrimination by modulating the affinity of E3 ligases for their substrates′ubiquitin chains[176]. Taken together,ubiquitin chain remodeling appears to be prominent during ERAD substrate fate determination.

DUB activity may play a significant role in apoB quality control processes.Ubiquitination of apoB appears to be responsive to apoB folding and topology, which is governed by lipid recruitment.Removal of ubiquitin conjugates from apoB may rescue partiallyassembled lipoproteins from degradation and allow apoB toreachsecretioncompetence.Also,DUBactivity is likely required to facilitate apoB retrotranslocation as well as efficient entry into the proteasome core.E3 ligases have an emerging role in regulating autophagy[177].Indeed,broad spectrum DUB inhibition with PR619 has been reported to activate autophagy in oligodendroglial cells[178].Further down the VLDL assembly line,DUB activity could participate in selecting and processing apoB for degradation by autophagy.As the ubiquitin landscape is revealed,it will be fascinating to observe the full extent to which deubiquitination is involved in VLDL assembly and protein quality control in general.

Clearance of nascent apoB-containing lipoproteins by autophagy

Autophagy is a rapidly growing area of cell biology. Dozens of proteins are actively involved in autophagy, across many cell types and tissues.The role of autophagy in hepatic apoB control quality control was recently reviewed[179],as has the broader role of autophagy in regulating lipid homeostasis in physiological and pathological settings[104].Defective hepatic autophagy has been reported to associate with chronic ER stress and insulin resistance in diet-induced and genetic models of obesity in the mouse[180].The relationship between hepatic autophagy,insulin signalling,lipid metabolism and apoB quality control is complex and incompletely understood.Autophagy is a recently appreciated mechanism for apoB disposal;however, it is unclear whether ERAD and autophagy are distinct processes,with different cues and physiological roles, or whether the two processes represent parts of a continuum of cellular degradation pathways.Summarized below are the observed mechanisms of apoB autophagy.

Omega-3 fatty acids were shown to induce hepatic autophagy of apoB and reduced VLDL secretion,in a process termed pre-secretory,post-ER proteolysis or PERPP[103].Blocking lipid peroxidation in rat hepatoma and primary rodent hepatocytes restored VLDL secretion.It was later reported that PUFA-derived reactive oxygen species(ROS)were responsible for apoB aggregation and degradation by autophagy[181]. In McA-RH7777 cells,maturation of pre-VLDL to fully lipidated VLDL in the Golgi is impeded by providing the omega-3 fatty acid DHA,resulting in autophagy of apoB[182].

In primary mouse hepatocytes,the insulin-dependent degradation of apoB by autophagy was shown to require signaling through class II PI3-kinase[183]. However,in McA-RH7777 cells,class I and III PI3-kinases were apparently involved,with overexpression of the negative insulin regulator phosphatase and tensin homolog(PTEN)and mutant Vps34(required for autophagy)showing that insulin-dependent apoB degradation involves a form of autophagy[184]. Interestingly,the flavonoid naringenin caused rapid, proteasome-independent apoB degradation and reduced apoB secretion from HepG2 cells via PI3K and MEK/ERK signalling,independent of the insulin receptor[185].Although it remains unconfirmed,autophagy is the likely degradation mechanism elicited by naringenin.ApoB overproduction was prevented by naringenin treatment of LDL receptor knock-out mice with diet-induced insulin resistance[186].These observations demonstrate that apoB autophagy is a complex,regulated and dynamic process.Fig.2displays apoB as it moves from the ER to the Golgi and is either subject to degradation by autophagy or completes the second bulk lipidation event and is secreted from the cell.Also shown inFig.2,in broad strokes, are the movement of lipids among the ER,Golgi,lipid droplets and autophagosomes that are involved in apoB metabolism.

Ohsaki and colleagues have described apoB‘‘crescents’’[187,188],which increase when autophagy or the proteasome are inhibited and decrease when MTP is reduced.The authors suggested that these structures are lipid droplets arrested in the ER membrane by binding to apoB and propose that apoB processing and lipid droplet formation are closely linked processes.Further,it has been proposed that these cytosolic lipid droplets are a site of convergence for proteasomal and autophagic apoB degradation.Thus, ERAD and autophagy may degrade apoB starting attemporally and spatially distinct exit points from the secretory pathway,yet both require the lipid droplet to process doomed apoB and its lipid cargo.

Fig.2Schematic representation of VLDL production,apoB autophagy and lipid cargo movement.Following synthesis,the apoB protein remains associated with the ER bilayer(left-hand side).This pre-VLDL particle is trafficked to the Golgi,where apoB then receives lipid ligands in bulk to form a full-sized VLDL particle(bottom,right).ApoB topology shifts and the VLDL particle is released from membrane association,followed by rapid secretion from the hepatocyte(top right).MTP is necessary for the formation of apoB-free lipid droplets(left,bottom),but MTP enzymatic activity is not required to facilitate the posttranslational bulk transfer of lipids onto apoB in the Golgi.ApoB is subject to late-stage quality control at the Golgi,where apoB can be removed from the secretory pathway by autophagy machinery.Movement of lipid metabolites is possible from cytosolic lipid droplets(CLD)to and from the ER bilayer,or to autophagosomes.Metabolic regulatory cues will determine the relative rate at which lipid metabolites are mobilized to the ER for export as VLDL,stored in the cytosol,oxidized in the mitochondria(not shown)or catabolized by autophagy.When apoB is destroyed via autophagy,its lipid cargo may become recycled and enter lipid pools that are oxidized,stored or re-mobilized to the ER.

Glucosamine induces a PERK-dependent decrease in translation and an increase in autophagic and proteasomal degradation to clear apoB[189-191].In McARH7777,apoB colocalized with the autophagy marker LC3 under stress conditions(tunicamycin or glucosamine)[191].Signaling through the PERK arm of the UPR was required to remove apoB by ER stress-induced autophagy.Notably,this stress could be relieved with the small molecule chaperone 4-phenylbutyric acid (PBA).This pathway was present in primary rat and hamster hepatocytes at basal conditions,but was absent from HepG2 cells unless the proteasome was also inhibited and extracellular fatty acid was present.It is notable that different cell types dispose of apoB in response to a single stressor,glucosamine,by different mechanisms.

As with ERAD,the mechanism of substrate selection is unclear in the apoB autophagy(PERPP)pathway. LC3,a critical protein component of autophagosome formation,bindsp62thatcontainsanubiquitinassociated (UBA)domain and is involved in ubiquitin-dependent substrate selection for autophagy[192].Binding of p62 to apoB-polyubiquitin chains at the Golgi membrane could conceivably allow the regulated removal of apoB-containing lipoproteins from the secretory pathway. Intriguingly,p62 activates Nrf2[193],an‘‘antioxidant response’’transcription factor that can be activated by the UPR(and is proposed to be the‘‘fourth arm’’of the UPR)[194].PerhapstheinsulinandROS-inducedclearance of apoB are regulated by a p62-Nrf2 axis that connects oxidative stress,the UPR and VLDL assembly.

This concludes the overview of cellular degradative mechanisms that target nascent apoB.These protein quality control processes exhibit crosstalk between one another,as well as sharing regulatory cues withother metabolic processes that impact apoB production.From an apoB-centric point of view,Section 3 will discuss the coordinated regulation of VLDL production by insulin signalling,lipid metabolism and the hepatic UPR in physiological and pathological settings.

APOLIPOPROTEIN B DEGRADATION -PHYSIOLOGY AND PATHOLOGY

The unfolded protein response(UPR)and the hepatic ER

The unfolded protein response(UPR)is a multipronged signalling response which was initially characterized as the ability of the ER to sense and adapt to the stress of an increased protein folding burden[195].Activation of the UPR is observed in disease states affecting many cell types and tissues. However,it is not always clear whether the ER stress and resulting cellular response are the cause,or an effect,of the disease.Complicating matters is the emerging distinction between physiological and pathological ER stress[196,197].The ER is a dynamic and finely-tuned organelle that serves as the major site of lipid and protein biosynthesis in the cell[198]and the UPR pathways are highly integrated into energy metabolism[197].While the role of ER stress in apoB production in vivo is not fully known,recent studies have revealed that several components of the UPR impact VLDL output from the liver[199].

Activation of the UPR,generally speaking,reduces new protein synthesis while expanding the protein folding and degradative capacity of the ER.The IRE1α arm of the UPR stimulates lipid biosynthetic pathways to expand the ER membrane to provide space for the elevated protein folding/degradative requirements.IRE1α also has a constitutive role in regulating hepatic lipid homeostasis,particularly through modulating postprandial,insulin-mediated de novo lipogenesis[200].Indeed,XBP1,directly downstream of IRE1α, has been described as a bona fide transcription factor in hepatic lipogenesis[201].Thus,activation of IRE1α may provide additional lipids to support VLDL assembly. Enhanced degradation of ERAD substrates by the ubiquitin proteasome system and autophagy could degrade apoB and decrease the amount of apoB molecules reaching secretion competence.Thus,during ER stress,there is simultaneous activation of processes that either support,or limit,VLDL assembly.Consideration of the characteristics of each individual ER stressor being deployed experimentally(pharmacologic,dietinduced,or genetic)is essential to prediction of the net effect on ER homeostasis.

Free fatty acids(FFA)are the most physiologically relevant hepatic ER stressors.A widely cited example of reduced VLDL output is during FFA-induced hepatic ER stress[125].While induction of ER stress decreased VLDL production in animal and cell models,the authors emphasized that apoB demonstrated a dose-and timedependent parabolic response to FFA.At mild to moderate levels of FFA-induced ER stress,apoB secretion was increased,demonstrating a window of ER stress activation wherein lipid availability is able to outstrip the concomitant decline in protein folding capacity in the ER.

Recent studies have found connections between the hepatic IRE1α arm of the UPR and VLDL output. Increased IRE1α-XBP1s activation induced PDI expression,which in turn increased MTP activity and apoB secretion,while liver-specific deletion of IRE1α impaired VLDL assembly in mice[202].The catalytic activity of PDI was dispensable for this particular function in lipoprotein assembly,suggesting that PDI increased VLDL assembly by modulating MTP activity rather than through its enzymatic function.

There is mounting evidence to suggest that the hepatic UPR is involved in maintaining normal glucose and lipid metabolism.An insulin-mediated,nutrientsensing signalling node,has been described that involves mammalian Target Of Rapamycin Complex 1(mTORC1)[203].This analysis revealed a metabolic transcriptional program with a broad regulatory connections between insulin and the UPR.mTORC1 is activated in the fed state[204].Meanwhile,mTORC1 inhibition by rapamycin blocked postprandial activation of hepatic IRE1α,and prevented XBP1-dependent lipogenesis in rats[205].XBP1 is required for insulin-stimulated activities of SREBP1[200].Nuclear levels of the lipogenic regulator SREBP1c were also decreased by the mTOR inhibition by rapamycin[158].Proteolytic processing and nuclear translocation of the SREBP proteins is needed for development of diabetic fatty liver and hypertriglyceridemia in mice[158].Mouse models of obesity showed that elevated hepatic ER stress,SREBP activation,increased lipid synthesis and overproduction of VLDL were mTORC1-dependent[85].Moreover,hepatic mTORC1 is elevated in ob/ob mice[206].Amazingly,a mouse model of liverspecific,inducible XBP1 revealed that XBP1 expression alone is sufficient to trigger a switch into the postprandial phenotype,even in the absence of caloric intake[207].

As mentioned in section 2a,repression of sortilin-1 by ER stress induction(via the UPR transcription factor named ATF3)caused increased VLDL output from the liver,providing an example of increased apoBsecretion during ER stress in vivo[85].Further,the suppression of sortilin-1 in ob/ob mice was found to be induced via increased hepatic mTORC1 activity. Taken together,these data suggest that increased mTORC1 activity and hepatic ER stress can provide an environment suitable for elevated VLDL production,via decreased apoB degradation and elevated insulin-mediated synthesis of lipid ligands for apoB. The next section delves into the unfortunate bifurcation of hepatic insulin signalling and the resulting pathological impact on VLDL production during insulin resistance.

Reciprocal effects of insulin resistance and fatty liver on apoB metabolism

Haas et al.have reviewed the regulation of apoB metabolism by insulin[208].Abnormalities of both intestinal and liver lipoprotein production underlie the complex hyperlipidemia that occurs in insulin resistance[64].Verges reviewed the mechanisms behind VLDL overproduction during insulin resistance[209]and Sparks et al.have reviewed the role of selective hepatic insulin resistance[63].Here,we shall examine how apoB production is affected by insulin resistance,accumulation of fatty liver and activation of the hepatic UPR.

A unique half insulin-sensitive/half insulin-resistant response arises in the livers of diabetic rats[210].Hepatic glucose production normally ceases after the feedinginduced increase in circulating insulin.Liver insulin resistance results in sustained glucose production[211]. On the other hand,the lipogenic pathways in the liver remain insulin-sensitive[210],promoting lipid biosynthesis.De novo lipogenesis(DNL)increases production of fatty acids in response to insulin.When hepatic glucose production continues in parallel with insulin-sensitive DNL,TG accumulation increases production of VLDL[212].Non-alcoholic fatty liver disease(NAFLD) and increased levels of VLDL-derived atherogenic lipoproteins result from these metabolic abnormalities in insulin-resistant individuals[213].Nutrient excess contributes to the chronic over-production of liver lipids by frequently flooding the organ with dietary substrates, thus creating a self-sustaining,pathological pattern of gene regulation.

Pour and Adeli[214]explored how insulin decreases the apoB mRNA translation when insulin is a global inducer of translation.ApoB mRNA accumulates in cytoplasmic complexes called P bodies,which contain silenced mRNAs.ApoB mRNA within P bodies could be observed 4-16 h after insulin treatment and their presence correlated with the decrease in apoB translation.Insulin resistance may prevent the movement of apoB mRNA into this inactive pool and perhaps allow for the‘‘over-translation’’of apoB as a prelude to the overproduction of VLDL.Sparks and colleagues[215]have extensively studied hepatic insulin resistance and the production of VLDL.Recently,they showed that although both MTP expression and apoB secretion are decreased by insulin under normal conditions, insulin was still able to suppress apoB secretion when MTP was overexpressed.This suggests that the decrease in apoB secretion is,in part,independent of the insulin effect on MTP.

Tsai et al.[216]found that hepatic inflammation underlies the increased VLDL production during insulin resistance in the diet-induced,insulin-resistant hamster.Fructose feeding of hamsters for as little as 4 days caused a decrease in Inhibitor of NF-kappaB (IkB)levels in the liver,while activation of IkB kinase (IkK)increased apoB synthesis and decreased degradation.IkK lies upstream of NF-kB,and has been shown to inhibit insulin signaling.The postprandial state is associated with increased inflammation and oxidative stress markers in circulation.Human studies have shown specific dietary fats can affect the nature of this response[217,218].Chronic inflammation and ER stress appear to be linked mechanistically in the liver during obesity and type 2 diabetes[219,220],with crosstalk amongst the UPR and inflammatory networks being relevant to several human diseases[221].

Basciano and colleagues[222]studied LXRα activation in hamsters with the agonist T0901317.After 7 days the agonist increased TG,apoB and VLDL-TG secretion via an increase in lipogenesis,promoted by stimulation of SREBP-1c,fatty acid synthase(FAS)and stearoyl-CoA desaturase(SCD).The agonist stabilized apoB for secretion without an effect on MTP.They suggested that dysregulation of hepatic insulin signaling can be a major mechanism for enhanced VLDL assembly and secretion with this drug.Interestingly, in a transgenic mice study,the LXR-SREBP1-PLTP regulatory axis was found to govern VLDL size by increasing TG synthesis and phospholipid availability for VLDL assembly[223].The ratio of hepatic phosphatidyl choline(PC)to phosphatidyl ethanolamine(PE) has also been shown to modulate membrane integrity and steatosis[224].LXR supports maintainence of ER membrane integrity by driving the expression of Lpcat3[225],whose enzymatic activity contributes unsaturated fatty acyl chains to the phospholipids comprising the ER bilayer.Moreover,restoring the imbalanced PC/PE ratio in ER stressed obese rodent livers improved calcium homeostasis,ameliorated ER stress and improved glucose homeostasis[226].These data providean intriguing link between hepatic ER stress,lipid synthesis and apoB production.

Irrespective of the extent of obesity,the metabolic dysfunction characterized by insulin resistance and VLDL overproduction is reflective of hepatic TG content[227].High levels of insulin,in the absence of insulin resistance,do not increase VLDL production in humans[228].PTEN is a negative regulator of PI3 kinase signaling in the insulin pathway.Qiu and colleagues[229]showed that liver-specific knockout of PTEN in mice decreased hepatic apoB and MTP mass and increased liver TG levels,concomitant with a decrease in apoB secretion.These changes were mediated by an increase in the degradation of the nascent apoB protein.They suggested that inappropriate decreases in hepatic PTEN expression may contribute to hepatic steatosis. Liver specific knockout of PTEN leads to enhanced hepatic insulin signaling with increased secretion of TG and apoB[230].If liver TG stores are normalized by inhibiting DGAT1,DGAT2 or SREBP1c,insulin signaling is retained but TG and apoB secretion are reduced.Acute reduction of PTEN does not reduce liver TG stores and also does not reduce apoB secretion despite enhanced insulin signaling.These observations suggest that hepatic TG is the dominant regulator of apoB secretion.

One report elegantly demonstrated the difference between excess de novo lipogenesis(high fructose feeding)and exogenous lipid oversupply(high fat feeding)in rats[231].Both high fructose and high fat diets caused steatosis and blunted insulin sensitivity within days.Interestingly,ER stress was detected in the fructose-fed animals but not in the high fat-fed animals,suggesting that the stress was associated with the lipogenesis,and was not an immediate component of the early fatty liver disease or insulin resistance induced by the high-fat diet[231].Increased mRNA expression of both MTP and apoB was observed in NAFLD subjects[232].However,in this study the most severe insulin resistance was associated with reduced MTP,suggesting that as the liver disease worsens,this compensatory mechanism is lost.This data implies that VLDL secretion persists during pathological hepatic ER stress and NAFLD,perhaps until progressive liver damage impacts the capacity of the secretory pathway to produce VLDL.

While in vivo evidence does not support the existence an ER stress-dependent decrease in VLDL output as a causative mechanism of fatty liver,perturbed apoB protein quality control during the progression of insulin resistance and NAFLD may still have unique,deleterious consequences.Augmenting apoB production is sufficient to either worsen or improve ER homeostasis[233].Su et al.proposed that increased intracellular apoB is a link between ER stress and hepatic insulin resistance.Dietary up-regulation of apoB,or its direct overexpression,caused ER stress and decreased insulin signaling in the liver,suggesting that apoB dysregulation in itself is sufficient to perturb ER homoeostasis. Furthermore,reduction of apoB using siRNA prevented ER stress and hepatic insulin resistance. Defective hepatic autophagy has been observed to promote ER stress and insulin resistance during obesity[180].It is conceivable that inefficient turnover of apoB selected for autophagy contributes to the disruption of hepatic metabolism,perhaps by creating a backlog of lipid cargo.

Continuous apoB production inflicts a constant, unique metabolic burden on the hepatocyte and the loss of insulin-dependent suppression of hepatic apoB levels may confer an additional burden,on top of that already posed by chronic over-nutrition.Successful moderation of apoB production is dependent on ER homeostasis,while conversely,hepatic ER homeostasis is dependent,in part,on the moderate and unimpeded export of lipidated apoB.Abrogation of VLDL output will cause neutral lipids to accumulate[234]. Combined with β-oxidation,VLDL output serves as a means to consume and partition energy resources within hepatocytes such that lipotoxic intermediary metabolites do not accumulate(specifically diacylglycerol(DAG)and ceramide).It has been suggested that proper liver homeostasis via continuous export of TG(and apoB)is maintained at the expense of an atherogenic plasma lipid profile[235].In this respect,apoB is perhaps a molecular‘‘steatosis suppressor,’’to borrow semantics from cancer researchers.

The physiological UPR and metabolic homeostasis

The UPR transcription factors that maintain ER homeostasis and help elicit the hepatic feeding response also have a significant impact on VLDL output and the metabolic status of the liver.Postprandial,insulinmediated suppression of VLDL output coincides with physiological activation of the UPR.However,hepatic insulin resistance is also associated with chronic UPR activation and VLDL overproduction during the postprandial period.These observations underscore the important differences between a physiological and pathological UPR.While the chronic,steady-state hepatic ER stress associated with metabolic diseases is well-documented,the short term effects of a single large meal on hepatic ER stress remain largely unexplored.When the‘‘postprandial UPR’’is superimposed onto an already-stressed ER,the resulting UPR may represent a pathological insult,whereby increased inflammation and excessive lipogenesis ensue.The mammalian postprandial UPR is an intriguing concept requiring experimental characterization under physiological and pathological conditions.

The chaperone component of ER homeostasis is precisely regulated;likened to a‘‘double-edged sword,’’where too much or too little can be dangerous[235].The enhancement of chaperone function is an intriguing potential therapeutic target for many human diseases,especially given the well-documented agerelated decline in chaperone regulation[236,237].The chemical chaperone 4-phenyl butyric acid(PBA)was able to normalize ER homeostasis and VLDL secretion in mice subjected to chronic oleate exposure[125]. Interestingly,PBA may improve insulin signalling and β-cell function in humans with lipid-induced insulin resistance[238].Treatment of humans with the chemical chaperone tauroursodeoxycholic acid(TUDCA) improved insulin sensitivity in the liver and skeletal muscle[239].Restoration of effective protein quality control may represent a point at which persistent cycles of pathological ER stress,inflammation and insulinresistance might be interrupted[240].Whether modulation of the UPR and chaperoning capacity in general will provide clinical benefits for metabolic diseases remains to be seen.

Concluding remarks

ApoB is a fascinating and complex substrate for cellular quality control.Since apoB production is continuous and regulated by many direct and indirect metabolic factors,the liver must be able to both maintain sufficient apoB to dispense with excess cellular TG when necessary,and to prevent the accumulation of apoB when hepatocellular nutrient levels are low. However,over time and in the presence of chronic levels of excess dietary carbohydrate and fat,the fidelity of insulin signalling can diminish,as does appropriate metabolite partitioning and protein quality control(Fig.3).Overflow levels of cytosolic lipid stores in the liver as well as chronic inflammation in the organ can create an environment in which VLDL production is continuous,rather than episodic and can contribute to extrahepatic pathologies such as atherosclerosis.Further work on the hepatic quality control mechanisms will be required to delineate causality and to increase the potential for new and improved therapeutic interventions.

Fig.3Factors affecting the net output of VLDL.This diagram illustrates the distinct,yet interconnected,processes that combine to determine the net rate of VLDL output.The figure serves as a qualitative schematic to envision factors that influence the regulatory and quality control decisions during VLDL assembly.Each of these factors have been shown to impact the fate of nascent apoB independently,however under physiological conditions the status of insulin signalling,lipid homeostasis,protein folding capacity and degradative capacity act in concert.In pathological states,these four processes can become discordant such that hepatic VLDL output is inappropriate.

[1] Shelness GS,Ledford AS.Evolution and mechanism of apolipoprotein B-containing lipoprotein assembly.Curr Opin Lipidol 2005;16:325-32.

[2] Avramoglu RK,Adeli K.Hepatic regulation of apolipoprotein B.Rev.Endocr.Metab Disord 2004;5:293-301.

[3] Hussain MM,Shi J,Dreizen P.Microsomal triglyceride transfer protein and its role in apoB-lipoprotein assembly. J Lipid Res 2003;44:22-32.

[4] Davidson NO,Shelness GS.Apolipoprotein B:mRNA editing,lipoprotein assembly,and presecretory degradation.Annu Rev Nutr 2000;20:169-93.

[5] Shelness GS,Ingram MF,Huang XF,DeLozier JA. Apolipoprotein B in the rough endoplasmic reticulum: Translation,translocation and the initiation of lipoprotein assembly.J Nutr 1999;129:456S-62S.

[6] Davis RA.Cell and molecular biology of the assembly and secretion of apolipoprotein B-containing lipoproteins by the liver.Biochim Biophys Acta 1999;1440: 1-31.

[7] Chan L.Apolipoprotein B,the major protein component of triglyceride-rich and low density lipoproteins.J Biol Chem 1992;267:25621-4.

[8] Rutledge AC,Su Q,Adeli K.Apolipoprotein B100 biogenesis:A complex array of intracellular mechanisms regulating folding,stability,and lipoprotein assembly. Biochem Cell Biol 2010;88:251-67.

[9] Havel RJ,Kane JP.Disorders of the biogenesis and secretion of lipoproteins containing the B apolipoproteins. In:Scriver CR,Beaudet AL,Sly WA,Valle D(Eds)The Metabolic and Molecular Bases of Inherited Disease Volume II.McGraw-Hill,Inc;1995:1853-1885.

[10]Sniderman AD,Furberg CD,Keech A,Lennep JERv, Frohlich J,Jungner I,et al.Apolipoproteins versus lipids as indices of coronary risk and as targets for statin treatment.Lancet 2003;361:777-80.

[11]Sniderman AD,Pierre ACS,Cantin B,Dagenais GR, Despres JP,Lamarche B.Concordance/discordance between plasma apolipoprotein B levels and the cholesterol indexes of atherosclerotic risk.Am J Cardiol 2003; 91:1173-7.

[12]Pierre ACS,Cantin B,Dagenais GR,Despres JP, Lamarche B.Apolipoprotein-B,low-density lipoprotein cholesterol,and the long-term risk of coronary heart disease in men.Am J Cardiol 2006;97:997-1001.

[13]Benn M,Nordestgaard BG,Jensen GB,Tybjaerg-Hansen A.Improving prediction of ischemic cardiovascular disease in the general population using apolipoprotein B: The copenhagen city heart study.Arterioscler Thromb Vasc Biol 2007;27:661-70.

[14]Sniderman A,Couture P,Graaf JD.Diagnosis and treatment of apolipoprotein B dyslipoproteinemias. Nat.Rev.Endocrinol.2010;6:335-46.

[15]Sniderman AD,Graaf JD,Couture P,Williams K,Kiss RS,Watts GF.Regulation of plasma LDL:The apoB paradigm.Clinical Science 2009;118:333-9.

[16]Gordon DA,Jamil H.Progress towards understanding the role of microsomal triglyceride transfer protein in apolipoprotein-B lipoprotein assembly.Biochim Biophys Acta 2000;1486:72-83.

[17]Wetterau JR,Lin MC,Jamil H.Microsomal triglyceride transfer protein.Biochim Biophys Acta 1997;1345:136-50.

[18]Zhang YL,Hernandez-Ono A,Ko C,Yasunaga K,Huang LS,Ginsberg HN.Regulation of hepatic apolipoprotein B-lipoprotein assembly and secretion by the availability of fatty acids.I.differential response to the delivery of fatty acids via albumin or remnant-like emulsion particles. J Biol Chem 2004;279:19362-74.

[19]Sniderman AD,Cianflone K.Substrate delivery as a determinant of hepatic apoB secretion.Arterioscler Thromb 1993;13:629-36.

[20]Taghibiglou C,Carpentier A,Iderstine SCV,Chen B, Rudy D,Aiton A,et al.Mechanisms of hepatic very low density lipoprotein overproduction in insulin resistance.evidence for enhanced lipoprotein assembly, reduced intracellular apoB degradation,and increased microsomal triglyceride transfer protein in a fructose-fed hamster model.J Biol Chem 2000;275:8416-25.

[21]Kummrow E,Hussain MM,Pan M,Marsh JB,Fisher EA. Myristic acid increases dense lipoprotein secretion by inhibiting apoB degradation and triglyceride recruitment. J Lipid Res 2002;43:2155-63.

[22]Zammit VA,Lankester DL.Oleate acutely stimulates the secretion of triacylglycerol by cultured rat hepatocytes by accelerating the emptying of the secretory compartment. Lipids 2001;36:607-12.

[23]Heath RB,Karpe F,Milne RW,Burdge GC,Wootton SA, Frayn KN.Dietary fatty acids make a rapid and substantial contribution to VLDL-triacylglycerol in the fed state. Amer J Physiol Endo Metab 2007;292:E732-9.

[24]Chen Z,Fitzgerald RL,Schonfeld G.Hypobetalipoproteinemic mice with a targeted apolipoprotein(apo) B-27.6-specifying mutation:In vivo evidence for an important role of amino acids 1254-1744 of ApoB in lipid transport and metabolism of the apoB-containing lipoprotein.J Biol Chem 2002;277:14135-45.

[25]Burnett JR,Zhong S,Jiang ZG,Hooper AJ,Fisher EA, McLeod RS,et al.Missense mutations in APOB within the betaalpha1 domain of human APOB-100 result in impaired secretion of ApoB and ApoB-containing lipoproteins in familial hypobetalipoproteinemia.J Biol Chem 2007;282:24270-83.

[26]Liao W,Ning G.Knockdown of apolipoprotein B,an atherogenic apolipoprotein,in HepG2 cells by lentivirus-mediated siRNA.Biochem Biophys Res Commun 2006;344:478-83.

[27]CrookeRM,GrahamMJ,LemonidisKM,WhippleCP,Koo S,Perera RJ.An apolipoprotein B antisense oligonucleotide lowers LDL cholesterol in hyperlipidemic mice without causing hepatic steatosis.J Lipid Res 2005;46:872-84.

[28]Soutschek J,Akinc A,Bramlage B,Charisse K,Constien R,Donoghue M,et al.Therapeutic silencing of an endogenous gene by systemic administration of modified siRNAs.Nature 2004;432:173-8.

[29]Zimmermann TS,Lee AC,Akinc A,Bramlage B, Bumcrot D,Fedoruk MN,et al.RNAi-mediated gene silencing in non-human primates.Nature 2006;441: 111-4.

[30]Akdim F,Visser ME,Tribble DL,Baker BF,Stroes ES, Yu R,et al.Effect of mipomersen,an apolipoprotein B synthesis inhibitor,on low-density lipoprotein cholesterol in patients with familial hypercholesterolemia.Am J Cardiol 2010;105:1413-9.

[31]Visser ME,Akdim F,Tribble DL,Nederveen AJ,Kwoh TJ,Kastelein JJ,et al.Effect of apolipoprotein-B synthesis inhibition on liver triglyceride content in patients with familial hypercholesterolemia.J Lipid Res 2010;51: 1057-62.

[32]Raal FJ,Santos RD,Blom DJ,Marais AD,Charng MJ, Cromwell WC,et al.Mipomersen,an apolipoprotein B synthesis inhibitor,for lowering of LDL cholesterol concentrations in patients with homozygous familial hypercholesterolaemia:A randomised,double-blind,placebo-controlled trial.Lancet 2010;375:998-1006.

[33]Norata GD,Tibolla G,Catapano AL.Gene silencing approaches for the management of dyslipidaemia. Trends Pharmacol Sci 2013;34:198-205.

[34]McLeod RS,Wang Y,Wang S,Rusi+a-Eol A,Links P, Yao Z.Apolipoprotein B sequence requirements for hepatic very low density lipoprotein assembly.J Biol Chem 1996;271:18445-55.

[35]Wang Y,McLeod RS,Yao Z.Normal activity of microsomal triglyceride transfer protein is required for the oleate-induced secretion of very low density lipoproteins containing apolipoprotein B from McA-RH7777 cells.J Biol Chem 1997;272:12272-8.

References of 36-240 are available online as a supplementary file.

Received 15 January 2014,Accepted 15 February 2014,Epub 28 March 2014

✉Corresponding author:Prof.Roger McLeod,Department of Biochemistry&Molecular Biology,Faculty of Medicine,Dalhousie University,PO Box 15000,Rm 9C,Tupper Medical Building,5850College Street,Halifax,NS B3H 4R2,Canada.Tel/Fax:+902-4947013/ +902-4941355,E-mail:Roger.McLeod@dal.ca.

The authors reported no conflict of interests.

ⓒ2014 by the Journal of Biomedical Research.All rights reserved.

10.7555/JBR.28.20140019