Biochimica et Biophysica Acta 1768 (2007) 1325–1341www.elsevier.com/locate/bbamem

Review

Transport and transporters in the endoplasmic reticulum

Miklós Csala a, Paola Marcolongo b, Beáta Lizák a, Silvia Senesi b, Éva Margittai a, Rosella Fulceri b,Judit É. Magyar a, Angelo Benedetti b, Gábor Bánhegyi a,b,⁎

a Department of Medical Chemistry, Molecular Biology and Pathobiochemistry, Semmelweis University,Pathobiochemistry Research Group of The Hungarian Academy of Sciences and Semmelweis University, Budapest, Hungary

b Department of Physiopathology, Experimental Medicine and Public Health, University of Siena, Siena, Italy

Received 28 December 2006; received in revised form 8 March 2007; accepted 15 March 2007Available online 24 March 2007

Abstract

Enzyme activities localized in the luminal compartment of the endoplasmic reticulum are integrated into the cellular metabolism bytransmembrane fluxes of their substrates, products and/or cofactors. Most compounds involved are bulky, polar or even charged; hence, theycannot be expected to diffuse through lipid bilayers. Accordingly, transport processes investigated so far have been found protein-mediated. Theselective and often rate-limiting transport processes greatly influence the activity, kinetic features and substrate specificity of the correspondingluminal enzymes. Therefore, the phenomenological characterization of endoplasmic reticulum transport contributes largely to the understanding ofthe metabolic functions of this organelle. Attempts to identify the transporter proteins have only been successful in a few cases, but recentdevelopment in molecular biology promises a better progress in this field.© 2007 Elsevier B.V. All rights reserved.

Keywords: Nucleotide sugar transporters; Solute carrier family 35; Glucose 6-phosphate translocase; TAP; Translocon; Permeability

Contents

1. Transported molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13261.1. Sugars and derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1326

1.1.1. Glucose. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13261.1.2. Ascorbate/dehydroascorbate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13271.1.3. Glucose 6-phosphate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1327

1.2. Phosphate and sulfate ions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13291.3. Nucleotide sugars and derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1329

Abbreviations: ER, endoplasmic reticulum; SER, smooth ER; RER, rough ER; EST, expressed sequence tags; DIDS, 4,4′-diisothiocyanostilbene-2,2′-disulfonicacid; NEM, N-ethylmaleimide; G6Pase, glucose 6-phosphatase; PDI, protein disulfide isomerase; G6P, glucose 6-phosphate; H6PDH, hexose 6-phosphatedehydrogenase; 11β-HSD1, 11β-hydroxysteroid dehydrogenase type 1; GSD1, glycogen storage disease type 1; G6PT, G6P translocase; vG6PT, variant G6Ptranslocase; SR, sarcoplasmic reticulum; NaPi or NPT, sodium/phosphate transporter; UDP, uridine diphosphate; UDP-Glc, UDP-glucose; UDP-GlcNAc, UDP-N-acetylglucosamine; UDP-Gal, UDP-galactose; UDP-GalNAc, UDP-N-acetylgalactosamine; UDP-Xyl, UDP-xylose; NSTs, nucleotide sugar transporters; SLC35,solute carrier family 35; Glc, glucose; Man, mannose; GlcNAc, N-acetylglucosamine; UGGT, UDP-Glc glycoprotein glucosyltransferase; UGTrel1 (SLC35B1), UDP-Gal transporter related protein 1; AtUTr1, Arabidopsis thaliana UDP-Gal/UDP-Glc transporter; UDP-GlcA, UDP-glucuronic acid; SITS, 4-acetamido-4′-isothiocyanostilbene-2,2′-disulfonic acid; DEPC, diethyl pyrocarbonate; UGTrel7 (SLC35D1), UDP-Gal transporter related protein 7; UGT, UDP-glucuronosyl-transferase; Ero1p, endoplasmic reticulum oxidoreductin 1 protein; Erv1p, essential for respiration and viability 1 protein; Fmo1p, flavin-containing monooxygenase;Flc, flavin carrier; CoA, coenzyme-A; Ac-CoA, acetyl-CoA; AT-1, Ac-CoA transporter; STS, steroid sulfatase; ARSC, arylsulfatase C; ES, estrone sulfatase; GSH,glutathione; GSSG, glutathione disulfide; RyR1, ryanodine receptor type 1; MHC, major histocompatibility complex; TAP, transporter associated with antigenprocessing; HEPES, 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid⁎ Corresponding author. Tel.: +36 1 2662615; fax: +36 1 2662615.E-mail address: banhegyi@puskin.sote.hu (G. Bánhegyi).

0005-2736/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.bbamem.2007.03.009

1326 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

1.3.1. UDP-glucose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13301.3.2. UDP-galactose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13301.3.3. UDP-xylose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13311.3.4. UDP-glucuronic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13311.3.5. UDP-N-acetylglucosamine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13321.3.6. UDP-N-acetylgalactosamine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1332

1.4. Adenine nucleotides and dinucleotides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13321.4.1. ATP/ADP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13321.4.2. NAD+/NADP+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13331.4.3. FAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1333

1.5. Acetyl-CoA and carnitine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13331.5.1. Acetyl-CoA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13331.5.2. Carnitine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1333

1.6. Conjugates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13341.7. Oligopeptides. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1334

1.7.1. Glutathione . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13341.7.2. Antigenic oligopeptides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1335

2. Non-specific transport in the ER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13353. Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1336

The endoplasmic reticulum (ER) is a membrane network [1]found in every nucleated cell. The morphologically distinctsmooth and rough ER (SER and RER) are formed by the samecontinuous membrane as the nuclear envelope. Its internalcompartment, the ER lumen, is completely separated from thecytosol. However, the luminal enzyme activities related tocarbohydrate metabolism, biotransformation, steroid metabo-lism and protein processing [2] are integrated in the cellularmetabolism, and strongly connected to the cytosolic processes.This compartmentation often narrows the specificity of luminalenzymes because several potential substrates cannot pass thebarrier. The transport of selected substrates across the ERmembrane is an additional point where the enzyme activity canbe potentially regulated. It is, therefore, doubtless that the ERfunctions cannot be properly revealed without understanding therelated transport processes, which, in turn, requires theidentification of the participating membrane proteins. Certainproteins involved in transport processes across ER membranehave been molecularly identified. Nonetheless, for the majorityof the molecules that admittedly cross the ER membrane theinvolved transport protein(s) are still molecularly undefined,which is probably due to technical difficulties. The classicstrategy based on separation, purification and fractionalreconstitution of solubilized ER membrane proteins did notlead to the expected success, with the exception of the transloconpore components. Screening cDNA libraries or EST databasesfor homologues of previously cloned transporters seems to be amore fruitful approach. Although some transport activities haveonly been characterized functionally, our knowledge about thetrans-membrane traffic in the ER is growing gradually. Thisreview focuses on the ER transport activities, which connectluminal and extraluminal metabolic processes by facilitatingsubstrate and product fluxes. Our aim was to provide a summaryof the available information in the field.

1. Transported molecules

1.1. Sugars and derivatives

1.1.1. GlucoseThe assumption that glucose crosses the microsomal

membrane by simple diffusion [3] proved false and it hasbecome clear that glucose is unable to cross cellularmembranes. There are two possible routes for glucose generatedin the ER lumen to be exported to the blood. It has not beenclarified whether glucose leaves the cell by vesicular transportor it is secreted through two consecutive transport steps throughthe ER and plasma membranes.

Glucose transporters known at molecular level are integralproteins of the plasma membrane. Facilitated transport ofglucose across the plasma membrane is catalyzed by a family ofproteins referred to as GLUTs. At least 12 GLUT isoforms areknown [4], having different kinetic properties, specificity andtissue distribution.

Much less is known about glucose transport across the ERmembrane. The existence of a microsomal facilitative transportsystem – different from GLUTs – was confirmed in severalways. Meissner and Allen [5] studied glucose transport in ratliver microsomes using rapid filtration technique. Pre-loadedmicrosomes released 70% of glucose within 20 s. Nevertheless,about 30% of microsomes showed a much lower glucosepermeation rate (t1/2=3 min). It was concluded that theexistence of microsomal vesicles with different permeabilityproperties could be due to the restricted number of glucosechannels in the ER.

Two components of the heterogeneous glucose transportwere characterized in rat liver microsomes [6]. The dominantrapid phase of glucose transport had a t1/2 of a few seconds andit was inhibited by pentamidine, cytochalasin B and 4,4′-

1327M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

diisothiocyanostilbene-2,2′-disulfonic acid (DIDS) [7]. Theslow component of glucose traffic [6] reached the steady-statelevel in 10 min and became saturable at around 100 mM glucoseconcentration. It was inhibited by the thiol alkylating agent N-ethylmaleimide (NEM), but not by the inhibitors of the rapidtransport (pentamidine or cytochalasin B).

The fluctuations of glucose concentration in the cytosol andin the ER lumen of HepG2 cells were detected in situ, by usingfluorescence resonance energy transfer-based nanosensors [8].The steady-state glucose levels or the kinetics of glucose uptakein these two compartments were similar, indicating the presenceof a high-capacity, bidirectional glucose transporter in the ERmembrane. However, the different sensitivity to cytochalasin Band the different relative kinetics for galactose uptake andrelease suggests that distinct proteins facilitate glucose traffic inthe plasma membrane and ER membrane.

Skepticism about the existence of the ER glucose transport issupplied by the fact that genetic deficiency of the ER glucosetransporter has not been unequivocally demonstrated. Analternative model for the glucose export from the ER wassuggested on the basis of findings in GLUT2 homozygousknockout mice [9]. These mice showed normal hepatic glucoseproduction indistinguishable from the glucose efflux observedin the wild-type. The proposed glucose export pathway wastemperature-sensitive, not inhibitable by cytochalasin B or byintracellular traffic inhibitors, such as brefeldin A andmonensin, but sensitive to progesterone, which is known toslow down exocytosis [10]. This study was broadened withfurther experiments on GLUT2-null hepatocytes by pulse-labeling technique [11]. In control and GLUT2-deficienthepatocytes a constant fast release of glucose was observedwith similar rate, even though the GLUT2-null cells accumu-lated glucose in their cytosol. This rapid glucose release couldbe impaired by the addition of nocodazole, which causesmicrotubule disruption, progesterone or low temperature.Glucose export from the cytosol proceeds at a slower rate,and was inhibited by phloretin and low temperature. These dataindicate the coexistence of two glucose-export pathways: themajor pathway that releases glucose through a vesiculartransport, while a separated minor pathway that relies on afacilitated diffusion process. However, the existence of thevesicular transport mechanism is still controversial [12]. It wasshown that nocodazole inhibits glucose transport directly, via amicrotubule independent pathway in adipocytes.

Surprisingly, it was also reported that glucose is unable tocross the microsomal membrane [13]. This striking interpreta-tion was presumably due to the use of mannitol as a postulatedextravesicular space marker, however it was evidenced thatmannitol penetrates the ER membrane [7].

Although a great amount of evidence support the existenceand functional role of an ER glucose transporter, it still remainsto be identified. Waddell et al. [14] described a 52-kDa proteinin rat liver microsomes, which was identified as a glucosetransporter protein, because it cross-reacted with an antibodyraised against the human erythrocyte glucose transporter and itweakly bound cytochalasin B. This microsomal protein wasthen purified and antiserum was raised against it. The antiserum

inhibited glucose 6-phosphatase (G6Pase) activity and glucoserelease from microsomes; hence this 52 kDa protein wasthought to be T3, the glucose transporter of the G6Pase system.Rat liver cDNA library was also screened using this antiserum[15]. The isolated clone showed a sequence similarity with themembers of GLUT family. The 52-kDa protein was, therefore,termed GLUT7. Its comparison with GLUT2 revealed threeregions where the amino acid sequences of the two proteinswere identical and six extra amino acids in GLUT7 containingan ER retention signal, which strongly suggested that theprotein resided in the ER membrane. However, this work wasretracted later, and the clone termed GLUT7 was judged as acloning artifact [16].

1.1.2. Ascorbate/dehydroascorbateAscorbate-dependent enzymatic reactions are present in the

lumen of the ER (e.g. prolyl hydroxylases). Since ascorbate –even in species able to synthesize this molecule – is producedonly in the liver or kidney, transport system(s) are needed tosatisfy the cofactor demand of these enzymes [17]. Althoughthe ER ascorbate/dehydroascorbate transporter is unknown atmolecular level, functional studies in rat liver microsomesrevealed the existence of a preferential protein-facilitateddiffusion of dehydroascorbate. Dehydroascorbate uptake iscis-inhibited and trans-stimulated by glucose and affected byglucose transport inhibitors, which indicates the possibleinvolvement of a yet unidentified hexose transporter. Thepresence of intravesicular reducing compounds increases,while extravesicular reducing environment decreases dehy-droascorbate influx [18]. The uptake of ascorbate showssimilar time course to its oxidation in rat liver microsomalvesicles: a rapid burst phase followed by a slower process,which suggests that it enters the vesicles after being oxidizedto dehydroascorbate. This assumption is further supported bythe findings that ascorbate influx is effectively hindered bycertain inhibitors of ascorbate oxidation (proadifen, econazoleor quercetin) [19].

Dehydroascorbate taken up by the ER can be reduced back toascorbate at the expense of glutathione [20] or protein thiols[21,22]. Both reactions can be catalyzed by the luminal proteindisulfide isomerase (PDI). Ascorbate entrapped in the lumencan leave the cell through the secretory pathway, and contributeto the maintenance of vitamin C level in the blood [23].

1.1.3. Glucose 6-phosphateGlucose 6-phosphate (G6P) is the substrate for at least two

enzymes in the ER lumen: G6Pase [for review, see 24] andhexose 6-phosphate dehydrogenase (H6PDH) [24–26]. G6Paseis present in cells capable of glucose production. It plays acrucial role in blood glucose homeostasis because it generatesglucose by hydrolyzing G6P derived either from glycogenolysisor gluconeogenesis. The enzyme is a membrane-bound proteinin the ER [27], and its catalytic site is localized in the lumen ofthis organelle [28–32]. H6PDH is a ubiquitous microsomalenzyme and it was shown to regulate the redox state of pyridinenucleotides within the ER [33]. By this means, it suppliesNADPH to 11β-hydroxysteroid dehydrogenase type 1 (11β-

1328 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

HSD1) [33–35], a luminal ER enzyme that plays a key role inpre-receptorial activation of glucocorticoids [36].

The consensus that G6P transport across the ERmembrane isrequired for both G6Pase and H6PDH activities was a result oflong debate. According to the “conformational hypothesis” ofG6Pase [37,38], transport of G6P is not required to explain thelatency and specificity of the enzyme activity in nativemicrosomes. In contrast, the “substrate transport hypothesis”,forwarded three decades ago, postulated the existence of ahighly specific G6P translocase, termed T1, in the ER/microsomal membrane [39]. The “substrate transport hypoth-esis” has been supported by a variety of experimental [3,32,40–42] and pathological [43–45] evidence. However, the finaldecisive evidence was provided when a transmembrane ERprotein that allows the entry of G6P in the luminal space wasmolecularly identified [46]. In addition, it is presently wellestablished that ER G6P transport is also a prerequisite forH6PDH activity [33,34,42], and indirectly for the cortisonereductase activity of 11β-HSD1 [33–35]. Accordingly, trans-port of G6P could be demonstrated in ER-derived vesicles(microsomes) from practically all cell types investigated so far.

Phenomenological characterization of G6P transport in livermicrosomes revealed a bidirectional facilitated diffusion withhigh capacity and low affinity [40,42]. The observed highcapacity reflects a relatively high number of transporters permembrane surface unit in the liver ER. These features of G6Ptransport are in accordance with the similar kinetic parametersof G6Pase, which make the system able to produce glucose at asufficient rate to maintain blood glucose level. In contrast,H6PDH has a relatively high affinity (Km=5 μM) to G6P[33,34,42]. The balance between passive G6P uptake andluminal hydrolysis determines the size of a metabolically activemicrosomal G6P pool [41]. Microsomal G6P transport isinhibited by a variety of compounds, including the aniontransport inhibitor DIDS [40,47,48], thiol reagents [49],mercaptopicolinic acid [50], certain protease inhibitors (i.e.tosyl-lysyl chloromethane and tosylphenylalanylchloro-methane) [51], fatty acids [44], acyl-CoAs [52,53], hydroxyni-trobenzaldehyde [54], chlorogenic acid and its derivates [54–57]. Chlorogenic acid derivates, such as S3483, are particularlypowerful (Ki =1 μM) and highly selective, thus, became usefulexperimental tools in G6P-transport-related studies [e.g. 32–34,42,58].

Genetic deficiency of G6Pase activity was already describedin the 1950s [59], and termed type 1 glycogen storage disease(GSD1). It was discovered later that liver microsomes preparedfrom certain GSD1 patients contained functional G6Paseenzyme. In these cases, G6Pase activity was found to bereduced only when it was measured in native microsomes, whileit was normal after membrane disruption [43,44]. It wasconcluded that genetic deficiency of a putative G6P transportercan also result in GSD1 phenotype [43,44]. Hence, twosubtypes of GSD1 were distinguished: GSD1a caused byG6Pase enzyme defect and GSD1b caused by G6P transporterdefect. Accordingly, little or no G6P transport activity wasmeasured in liver microsomes obtained from GSD1b patients[45]. Defective G6P transport leads to insufficient substrate

supply to G6Pase in gluconeogenic tissues; therefore, it mimicsthe true enzyme deficiency. Moreover, metabolic derangementis accompanied by neutropenia and functional defects ofpolymorphonuclear leukocytes and monocytes in GSD1bpatients [24]. In fact, polymorphonuclear leukocytes fromGSD1b patients exhibit impaired chemotaxis as well asdiminished respiratory burst and phagocytotic activities.Chemical inhibition of G6P transport with S3483 mimickedsome leukocyte defects of GSD1b patients and resulted inapoptosis of human neutrophils [58]. It was hypothesized thatG6P transport may have a role in the antioxidant defense ofneutrophils [58].

The cDNA sequence coding the human ER G6P translocase(G6PT) was discovered on the basis of homology with abacterial hexose-phosphate transporter [46]. Human G6PT is asingle copy gene consisting of 9 exons spanning approximately5.3 kb DNA [60–62], at chromosome 11q23 [60,63]. Homo-zygous mutant G6PT has been found in more than 95% of theinvestigated GSD1b patients and at least 80 separate G6PTmutations have been described in 117 GSD 1 patients [see 64for a recent review]. Murine and rat cDNA sequences codingG6PT share 93–95% homology with the human sequence [65],and the pathology of GSD1b has been essentially reproduced inG6PT−/− mice [66,67].

G6P uptake was investigated in microsomes prepared fromCOS-1 cells over expressing human G6PTs. Microsomes wereincubated with [14C]G6P and luminal accumulation of radio-activity was detected. Isotope accumulation was only evidentwhen the cells were co-transfected for the G6Pase enzyme [61],thus, it was concluded – as previously suggested [65,68] – thatG6Pase protein was somehow required for the G6PT activity.Nevertheless, this indirect measurement of G6P uptake does notprove the hypothesized regulatory relationship between G6PTand G6Pase. It is now clear that in the presence of G6Pase, themajority of luminal radioactivity is [14C]glucose retained evenby liver microsomal vesicles [41]. An evident microsomaltransport of G6P was subsequently observed in COS-7 cellsexpressing human G6PT, but not the G6Pase enzyme [69].

Northern blot analysis revealed at least two G6PT mRNAs.The “liver”mRNA contains eight out of nine exons (without exon7) and “brain” mRNA contains all the nine exons [70]. The latterone was also found in heart and skeletal muscle [62,71] andnamed “variant” vG6PT [71]. ThemRNAencodingG6PT ismostabundant in liver and kidney, but is also present at lower levels invarious other tissues including, large intestine, small intestine,stomach, testis and neutrophils [46,62,70–73]. As mentionedabove, vG6PT is expressed in brain, heart, and skeletal muscle.

Predicted Mw of human G6PT protein is approximately46 kDa, while human vG6PT protein should contain 22additional amino acids coded by exon 7. Western blot analysisusing an antibody against the N-terminal 17 amino acids ofG6PT revealed a liver microsomal protein, termed P46, but itsapparent Mw was not reported [74]. A FLAG-tagged proteinoverexpressed in COS-1 cells, transfected with the human livercDNA coding G6PT, appeared to have an Mw of 37 kDa [75].Subsequently, antibodies to hydrophilic peptide sequences ofhuman G6PT protein have been produced [69]. Using this

1329M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

antibody, human (and rat) liver G6PT protein was shown tomigrate in SDS PAGE with an apparent Mw of 33 kDa [69].Several lines of evidence, however, indicate that its real Mw is46 kDa, and the observed unorthodoxmigration of G6PT proteinin SDS-PAGE is admittedly due to its high hydrophobicity [69].G6PT was also immunodetected in kidney microsomes, but notin microsomes derived from human fibrocytes, rat spleen andlung and a variety of cell lines [69]. An immunoreactive proteinwith a slightly higher Mw, likely vG6PT, was detected in brainand skeletal muscle microsomes [69].

G6PT is a rather hydrophobic protein embedded in the ERmembrane, its transmembrane topology is, however, stillcontroversial. Two models have been proposed, one with 10[75] and the other with 12 [46,76] transmembrane alpha-helices.Bothmodels agreewith the findings that the two extremities of theprotein are on the cytosolic side of the ER, and that the protein isnot glycosylated. Despite mammalian G6PTs are phylogeneti-cally related to phosphoester/Pi antiporters or to G6P receptors inbacteria/plant, they appear to operate as facilitative transporters.G6P transport in liver microsomes is bidirectional [40,42], andappears to be independent of the [Pi] gradient between themicrosomal lumen and the incubation medium (Fulceri andBenedetti, unpublished data). Moreover, entrapping Pi within themicrosomal lumen (i.e. with Pb2+ [32] or Ca2+ [77] ions) does notreduce the rate of G6P hydrolysis/transport.

The role of G6PT (or vG6PT) in G6P transport across the ERmembrane has been proven in several cell types, nevertheless,in some cases the protein(s) facilitating G6P permeation remainto be identified. A G6PT-independent G6P transport activitywas found in fibroblasts. It could also be detected inmicrosomes of skin fibroblasts from GSD1b patients havinghomozygous mutations of the G6PT gene and it was shown tobe insensitive to G6PT inhibitor chlorogenic acid [78]. Up-regulation of the yet-unidentified transporter(s) might accountfor the improvement of metabolic functions or lack ofneutrophil dysfunctions [64] observed in some GDS1b patients.

1.2. Phosphate and sulfate ions

Bidirectional transport of inorganic phosphate [Pi] across theendo/sarcoplasmic reticulum has been repeatedly observed anddescribed. In the cytosol, Pi is mainly generated as the productof ATP hydrolysis; while in the lumen of the ER, it is produceddominantly from G6P or UDP.

The concentration of phosphate ions (i.e. HPO42− and

H2PO4−) greatly influences the size of the rapidly releasable

intracellular calcium pool in a wide variety of cell types. Lowconcentration of Pi, in the range reported to occur in the cytosol,favors Ca2+ uptake and storage, via Ca2+ and Pi co-accumula-tion, in ER vesicles derived from liver [79] and several non-hepatic tissues [80,81].

The presence of a Pi transporter in the sarcoplasmicreticulum (SR) membrane of mammalian skeletal muscle fibershas been reported [82]. The Pi transporter operates by handlingCa2+ uptake into the SR, where it has been suggested that Ca2+

and Pi co-precipitate within the lumen. Moreover, the Pitransporter is active at physiological concentrations of myo-

plasmic Pi and Ca2+. It could be partially inhibited byphenylphosphonic acid at concentrations that have only asmall effect on the SR Ca2+ pump, indicating that the Pitransporter and the SR Ca2+ pump are separate proteins [83,84].

The existence of a Pi translocase in the ER membrane wasproposed in 1975 by Arion [39]. According to his model,termed “substrate-transport model”, G6Pase enzymatic systemconsists of the G6Pase enzyme and three transporters: T1, T2and T3 mediating the traffic of G6P, Pi and glucose,respectively. Accordingly, GSD1 was divided into fourtheoretical subtypes, corresponding to defects in the G6Pasecatalytic unit (la), the G6P transporter (lb), a putative phosphatetransporter (1c), and a putative glucose transporter (1d) [85].

Pi produced from G6P could be entrapped in the lumen ofintact microsomes by forming an insoluble complex with Pb2+

ions. This provided evidence for the luminal localization ofG6Pase activity and for the substrate transport model [32]. Theobservations suggested a very short (near 1 s) half-life of Piefflux from microsomal vesicles.

Kinetic studies performed on liver microsomes indicated theexistence of a common transporter for Pi, PPi, and carbamoylphosphate, which is separate from the G6P translocase [3] [3].Nevertheless, evidence was also provided for a microsomal Pitransport protein that does not transport PPi [86]. Permeabilityof the microsomal membrane to Pi is not altered in GSD1b [45],nor is it affected by G6P [87], supporting the independence of Pitransporter from G6P transporter.

The molecular identity of the protein(s) involved in Pitransport across the ER membrane is still undefined. Mamma-lian cells contain several Pi transporters. The one present in theinner mitochondrial membrane is a 34-kDa protein thatmediates the exchange of monobasic phosphate for OH−[88].Pi transporters of the plasma membrane carry out Na+-dependent transport, and belong to at least three differentfamilies: NaPi-I (also called NPT1), NaPi-II and NaPi-III [89].

It was reported that antibodies directed against themitochondrial Pi transporter cross-react with a microsomalprotein of matching size in human liver [90]. But these resultscould be explained by contamination of the microsomal fractionby mitochondrial particles.

A novel candidate for ER Pi transporter has been namedNPT4 [91]. NPT4 gene codes for a Na+/Pi co-transporter, whichbelongs to the anion-cation symporter family. NPT4 isexpressed in liver and kidney and the protein is localized inthe ER membrane. Mutation in one allele of NPT4 was detectedin one case of GSD1c, and it was suggested that the diminishedG6Pase activity was due to the reduced efficiency Pi transportacross the ER membrane.

Possibility that Pi and sulfate transport could be mediated bya common transporter has been excluded as neither Pi nor avariety of phosphate transport inhibitors affected sulfatetransport in liver microsomes [92].

1.3. Nucleotide sugars and derivatives

Nascent polypeptides are translocated into the lumen of theER, where many of them undergo co- or posttranslational

1330 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

glycosylation. The initial oligosaccharide moiety to betransferred to N-glycosylation sites of the native proteins ispartially synthesized on the cytosolic side of the ER membraneas a dolichol derivative. Yet UDP-glucose (UDP-Glc) and UDP-N-acetylglucosamine (UDP-GlcNAc) and UDP-galactose(UDP-Gal) are substrates of luminal glycosyltransferases ofthe ER. Since nucleotide sugars are synthesized primarily in thecytosol these charged solutes must enter the ER lumen so theycan participate in protein and lipid glycosylation.

Several nucleotide sugars, such as UDP-Glc, UDP-GlcNAc,UDP-Gal, UDP-N-acetylgalactosamine (UDP-GalNAc), UDP-xylose (UDP-Xyl) were shown to be transported into the lumenof microsomal vesicles prepared from rat liver and yeast.Nevertheless, in some instances the biochemical function of theobserved transport activities in the ER is not clear. It may reflectincomplete sorting of the corresponding Golgi proteins duringmembrane biogenesis. Alternatively, it might be explained bythe overlapping ligand specificity of ER nucleotide sugartransporters, e.g. a UDP-Glc transporter may also facilitateUDP-Xyl and UDP-GalNAc fluxes.

Nucleotide sugar transporters (NSTs) belong to the solutecarrier family 35 (SLC35) [93]. They are hydrophobic proteinsof 320–400 amino acid residues forming ten transmembranehelices with C- and N-terminal regions exposed to the cytosol.NSTs are antiporters, which mediate the uptake of nucleotidesugars from the cytosol into the lumen of the Golgi and/or theER in exchange for the corresponding nucleoside monopho-sphates. According to the proposed mechanism, nucleosidediphosphates generated by most nucleotide-sugar-dependentglycosyltransferases are converted subsequently by nucleosidediphosphatase activities to nucleoside monophosphates, whichserve as counter-substrates for the nucleotide sugar antiport.However, recent studies showed that nucleotide sugar trans-porters allow the entry of nucleotide sugars into the luminalcompartments in nucleoside-diphosphatase-deficient Sacchar-omyces cerevisiae. Nucleoside diphosphates could exit to thecytosol by an unknown mechanism in these mutants [94].

1.3.1. UDP-glucoseGlc3Man9GlcNAc2 structure is co-translationally transferred

to most newly synthesized polypeptides in the ER. Thisoligosaccharide moiety is immediately trimmed to a “mono-glucosylated” (GlcMan9GlcNAc2) form, which directs thenascent protein to calnexin and calreticulin. In addition tofacilitating the folding process, these chaperones retain theimmature protein in the ER lumen until it achieves its nativeconformation. The protein is detached from the chaperoneswhen its peripheral glucose residue is removed by glucosidaseII. If the protein has been correctly folded, it can leave the ER byvesicular transport else it is reglucosylated by UDP-Glcglycoprotein glucosyltransferase (UGGT), which leads to itsreassociation with calnexin and calreticulin. This mechanismreferred to as quality control prevents the release of misfoldedproteins from the ER. UGGT plays an essential role in theprocess by allowing the glycoprotein to re-enter the foldingcycle. The luminal UDP-Glc consumption of UGGT [95]requires the entry of UDP-Glc from the cytosol, which is

mediated presumably by an NST protein. Although severalNSTs have been described in eukaryotes, most of them arerelated to protein glycosylation and polysaccharide biosynthesisin the Golgi apparatus, and the NST involved in supplying thesubstrate for UGGT has not yet been identified.

Inward UDP-Glc transport was only functionally character-ized in RER vesicles derived from rat liver. It was found to bearthe features of a carrier-mediated transport, such as saturation,specificity, temperature-dependence and countertransport. Inhi-bition and trans-stimulation studies indicated that the postulatedmicrosomal UDP-Glc carrier system transports other uridinenucleotide sugars and 5′-UMP too, while cytosine or guanosinenucleotides and non-5′-uridine monophosphates are poorsubstrates [96,97]. The uptake of UDP-Glc into the lumen ofthe endomembrane system was also proven by the observationthat UDP-Glc is constitutively released through the secretorypathway from physiologically relevant tissues [98]. Tempera-ture-dependent, substrate specific and saturable UDP-Glctransport was also detected in ER-containing microsomesfrom S. cerevisiae [99].

UDP-Gal transporter related protein 1 (UGTrel1 orSLC35B1) [100] is a promising candidate UDP-Glc transporterin the ER. Although the substrate specificity and subcellularlocalization of human, murine and rat UGTrel1 proteins havenot been reported yet, the S. cerevisiae homologue, Hut1,showed weak, yet significant, transport of UDP-Gal [101].Further studies revealed that Hut1 is localized in the ER, and thehomologues from Schizosaccharomyces pombe and Arabidop-sis thaliana transport UDP-Glc as well as UDP-Gal [102,103].The phenotype caused by disrupted hut1 gene in fission yeastresembles the phenotype observed in mutant cells deficient inER quality control proteins, which strongly suggests the role ofHut1 in the protein processing in the ER [102].

AtUTr1, the UDP-Gal/UDP-Glc transporter from A. thali-ana, was studied after heterologous expression in yeast. It wasshown to transport UDP-Glc 200 times faster than UDP-Gal.The C terminus of the protein contains an ER retention signalfor membrane proteins (KKXX) [103]. The localization of theprotein was analyzed by a chimera, fusing the green fluorescentprotein to the C terminus of AtUTr1. The reticulated distributionof AtUTr1-GFP indicated its localization in the ER rather thanin the Golgi apparatus. It should be noticed that the induction ofAtUTr1 was observed in the unfolded protein response [104].

1.3.2. UDP-galactoseThe UDP-Gal supply for the galactosyltransferases partici-

pating in galactosylation of glycosphingolipids and proteins inthe Golgi lumen is ensured by an antiporter allowing the entryof UDP-Gal and the exit of UMP [105,106]. It was reported thatUDP-Gal transport activity is exclusively localized to Golgifractions and absent from the ER in rat liver [107]. However,this is contradicted by the presence of ceramide galactosyl-transferase in the ER that requires UDP-Gal transport from thecytosol into the lumen for the galactosylation of ceramides anddiglycerides. Ectopically expressed human UDP-Gal transpor-ter co-operated functionally in the ER with ceramide galacto-syltransferase enzyme either endogenous or co-expressed (in

1331M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

Caco-2 or CHOlec8 cells, respectively). It was demonstratedusing subcellular fractionation and double label immunofluor-escence microscopy that ceramide galactosyltransferase retainsUDP-Gal transporter in the ER to ensure the substrate supply[108].

In addition – as it was discussed in the previous section –AtUTr1 from A. thaliana was also shown to transport UDP-Galin the ER [104].

1.3.3. UDP-xyloseXylosylation is involved in proteoglycan biosynthesis. UDP-

Xyl can be generated from UDP-glucuronic acid (UDP-GlcA)within the lumen of the ER by the concerted action of UDP-GlcA transporter and the intraluminally oriented UDP-GlcAcarboxy-lyase [109]. Nevertheless, it was suggested that UDP-Xyl transport is not restricted to the Golgi but is also present inthe ER [106].

Investigations on permeabilized chondrocytes providedevidence for UDP-Xyl transport into the ER. These cells cansynthesize chondroitin sulfate by utilizing UDP-Xyl eitherexogenous or generated from UDP-GlcA in situ. It wasconfirmed using electron microscopic autoradiography thatthe incorporation of labeled xylose from exogenous UDP-[14C]Xyl began in the ER and continued in the Golgi apparatus [109].

Comparison of RER- and Golgi-derived rat liver subcellularfractions revealed that UDP-Xyl transport was 2–5 times fasterin the latter. Accordingly, the specific activity of xylosyltrans-ferase was much higher in Golgi vesicles and endogenousacceptors for xylosylation were only present in this fraction[110].

The role of UDP-Xyl transport in the hepatic ER might berelated to glucuronidation. Addition of UDP-Xyl was shown tostimulate UDP-glucuronosyltransferase (UGT) activity in ratliver microsomes and in isolated permeabilized hepatocytes.The enhanced UGT activity is due to stimulation of UDP-GlcAuptake across the ER membrane and to elimination of thereaction product (UDP and/or UMP) from the ER lumen. UDP-Xyl is a remarkable trans-stimulator and a weak cis-inhibitor ofmicrosomal UDP-GlcA uptake. On the other hand, cytosolicUDP-GlcA strongly trans-stimulates the efflux of UDP-Xyl,whereas cytosolic UDP-Xyl does not trans-stimulate the effluxof UDP-GlcA. Microsomal UDP-Xyl influx was markedlystimulated by luminal UMP and UDP [111]. UDP-GlcA uptakeis greatly enhanced by luminal UDP-GlcNAc and UDP-Xyl inthe ER while hardly affected in Golgi vesicles [112]. The resultssuggest that an antiport mechanism is involved in the UDP-Xyltraffic across the ER membrane.

1.3.4. UDP-glucuronic acidUDP-GlcA is synthesized in the cytosol and utilized within

the ER lumen by UGT isoenzymes in a glucuronyl grouptransfer reaction called glucuronidation. UDP-GlcA is arelatively large polar molecule with double negative charge(phosphoryl and glucuronyl groups); hence its passive diffusionacross lipid bi-layers can be excluded. The assumptions thatglucuronidation does not require UDP-GlcA transport into theER lumen [110] or that this nucleotide-sugar is transported into

the ER membrane rather than into the luminal space [113] wereconfuted. It was proven that UDP-GlcA traffic across the ERmembrane is a bidirectional carrier-mediated translocation[114,115]. The carrier-mediated UDP-GlcA transport acrossthe ER membrane was also shown to be required and rate-limiting for glucuronidation in intact microsomal vesicles aswell as in the intact ER of permeabilized hepatocytes [116].

The different properties of UDP-GlcA transport in ER andGolgi derived vesicles led to the conclusion that separate carrierproteins should exist in the two organelles [112,117]. It was alsoshown that UDP-GlcA and UDP-Glc are transported via twodistinct mechanisms across the ER membrane [115]. The yet-unidentified carrier protein can be inhibited by general aniontransport inhibitors, 4-acetamido-4′-isothiocyanostilbene-2,2′-disulfonic acid (SITS), DIDS, probenecid [113] and by thethiol-alkylating agent NEM administered at the cytosolic side ofthe ER membrane [114,118]. Some histidyl-specific irreversibleinhibitors, especially the hydrophobic ones, such as diethylpyrocarbonate (DEPC) or p-bromophenacyl bromide alsoinactivate the UDP-GlcA uptake [119]. Kinetic analysis ofUGT activity in intact and permeabilized microsomes indicatedthat UDP-GlcNAc facilitates the transport of UDP-GlcA fromthe cytosol to the ER lumen [120]. The transport was alsoshown to be activated by Mg2+ and GTP (but not ATP) [117].The kinetic properties of UDP-GlcA transport indicated thepresence of a two component UDP-GlcA transporting system inthe ER. The high affinity (Km=1.6 μM) transporter is moresensitive to inhibition by SITS and furosemide as well as totrans-activation by UDP-GlcNAc than the low affinity(Km=38 μM) transporter [121].

Microsomal UDP-GlcA uptake detected by rapid filtrationtechnique shows a characteristic time-course. It starts with anovershoot phase, during which luminal UDP-GlcA leveltransiently exceeds the amount of UDP-GlcA present in thevesicles at equilibrium. This phenomenon cannot be observed athigher UDP-GlcA concentrations or after long (more than 2 h)preincubation of the microsomes, but it can be reconstituted inpreincubated microsomal vesicles by preloading with a highconcentration (5 mM) of UDP-Glc [114]. It was concluded thatUDP-GlcA transport is trans-stimulated by a compound orcompounds contained by fresh microsomal vesicles. Accord-ingly, UDP-GlcA release was effectively trans-stimulated by 5′-uridine nucleotides (UDP-GlcA, UDP-Glc, UDP-GlcNAc,UDP-Xyl, UMP, UTP) [114]. It was also demonstrated thatthe transport of UDP-GlcA and some uridine nucleotides istrans-stimulated by certain phenol glucuronides (p-nitrophenylglucuronide and phenolphthalein glucuronide) [122]. Theproposed antiport mechanism could co-operate with the UGTsby exchanging cytosolic UDP-GlcA for the glucuronidesformed in the ER lumen. On the basis of mutual transporttrans-activations between various uridine nucleotides, modelsystems were postulated in the ER, where UDP-GlcNAc orUDP-Xyl functions as a shuttle compound in helping UDP-GlcA to enter and UDP and UMP to leave the ER lumen[111,123].

It was suggested that the oligomer complex formation ofUGT enzyme proteins in the ER membrane affects the

1332 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

UDPGlcNAc-dependent stimulation of glucuronidation byfacilitating UDP-GlcA uptake [124]. The UGT oligomer itselfcould function as a UDP-GlcA carrier or channel [118].However, this hypothesis is contradicted by some observations.Firstly, the membrane-impermeant protein-modifying agentdiazobenzenesulphonate abolished the normal stimulation ofglucuronidation by UDP-GlcNAc in intact microsomes while itonly inhibited the basal rate of glucuronidation when themembrane was permeabilized [125]. Secondly, binding studiesindicated that UGTs are not accessible to NEM or UDP-GlcNAc, when they are in their membrane environment.Thirdly, microsomes from UGT1-deficient Gunn rat livershowed normal UDPGA uptake indicating that UGTs of the1A family do not function as UDPGA carriers [126].Collectively, a separate NEM- and diazobenzenesulphonate-sensitive UDP-GlcA transporter might mediate the stimulationof UGTs by UDP-GlcNAc in intact microsomes [127].

Recent findings indicate the permeability of translocon poresabundant in the RER membrane to organic anions includingUDP-GlcA, thus it might be considered as a non-specific low-activity UDP-GlcA permease [128].

Search of the EST database for genes related to the humanUGTrel1 revealed a novel human transporter (UGTrel7 orSLC35D1). It was heterologously expressed in yeast cells and itwas shown to facilitate the transport of both UDP-GlcA andUDP-GalNAc in yeast microsomes. SLC35D1 was alsotransiently expressed in CHO-K1 cells, and immunostainingindicated its co-localization with an ER marker protein. It wastherefore concluded that the novel transporter might participatein glucuronidation by mediating the UDP-GlcA supply of UGTs[93,129].

1.3.5. UDP-N-acetylglucosamineUDP-GlcNAc is required for the synthesis of O- and N-

glycoproteins in the ER lumen [130]. UDP-GlcNAc transloca-tion was observed across the membrane of vesicles derived fromrat liver RER and Golgi apparatus. The translocation istemperature-dependent, saturable, selective, capable of trans-stimulation, operational against a concentration gradient, andprotein-dependent (i.e. inhibited by treatment of vesicles withproteases). The amount of vesicle-associated UDP-GlcNac atequilibrium is directly proportional to the volume of thevesicular space, and permeabilized microsomes are unable toretain UDP-GlcNAc. Consequently, the microsomal uptake ofUDP-GlcNAc meets the criteria for bidirectional, carrier-mediated translocation [107,131].

The transport of UDP-GlcNAc in the ER – similarly to thatof UDP-Xyl – may be connected to glucuronidation. It wassuggested that UDP-GlcA influx coupled to UDP-GlcNAcefflux and UDP-GlcNAc influx coupled to UMP efflux,combined with intravesicular consumption of UDP-GlcA,form a mechanism that leads to the observed stimulation ofglucuronidation by UDP-GlcNAc [123].

Though UDP-GlcNAc transporter-deficient mammaliancells have not been isolated, an aerobic milk yeast mutant(Kluyveromyces lactis mnn2-2) was identified with thisdeficiency [132], and complementation of the genetic defect

led to the identification of the transporter gene [133]. Due to alimited degree of similarity, this gene and the subsequentlyidentified canine and human UDP-GlcNAc transporter genes[134,135] were assigned to different subfamilies of SLC genefamily: 35A and 35B, respectively [93]. The human UDP-GlcNAc transporter was shown to be localized in the Golgimembrane when expressed in CHO cells [135].

In conclusion, functional evidence supports the existence ofan ER UDP-GlcNAc transporter, while transporters with thisactivity were identified at molecular level only in the Golgi.

1.3.6. UDP-N-acetylgalactosamineThe gene of a novel NST related to the human UDP-Gal

transporter-related isozyme 1 was identified by using ESTdatabase search and named hUGTrel7. The protein washeterologously expressed in S. cerevisiae and its transporterfunction was characterized. The uptake of UDP-GlcA andUDP-GalNAc was enhanced in the microsomes prepared fromthe hUGTrel7-expressing yeast compared with the controlmicrosomes. It was, therefore, concluded that hUGTrel7transports both UDP-GlcA and UDP-GalNAc [129]. Theresults of double-staining immunofluorescence indicate thathUGTrel7 is localized in the ER in mammalian cells [129], so itmay participate in glucuronidation and/or chondroitin sulfatebiosynthesis in this organelle.

1.4. Adenine nucleotides and dinucleotides

1.4.1. ATP/ADPChaperone-mediated protein folding in the lumen of the ER

requires ATP as an energy source. Phosphorylation of someluminal proteins was also observed. ATP must first betranslocated across the ER membrane before it can serve assubstrate in these luminal reactions.

The transport of ATP was demonstrated in RER-derivedvesicles prepared from rat liver and canine pancreas. Themicrosomes showed a saturable translocation of ATP, whichwas protein mediated, since it was inhibited by treatment ofintact vesicles with pronase, NEM or DIDS. HPLC analysis andthe use of a nonhydrolyzable analog proved that the entire ATPmolecule was translocated [136]. The transport of ATP and itsinhibition by DIDS was also demonstrated in yeast ER [137].

The ATP transporters of the yeast and of rat liver ER werereconstituted into proteoliposomes [138,139]. A 68-kDa proteinfrom the ER of S. cerevisiae (Sac1p) was supposed to have anintimate role in the transport of ATP into the ER lumen [139].However, it turned out that Sac1p is not an ATP transporteritself but an important regulator of the transport process,probably by controlling ER phosphoinositides. Purified Sac1preconstituted in proteoliposomes did not catalyze ATP uptake[140].

In rat liver, a 56-kDa protein was identified as an ATPtransporter by photoaffinity labeling and partial purification[141]. Transport activity of the protein could be demonstrated inreconstitution experiments. In a more recent study ATPtransporter from rat liver RER was solubilized and reconstitutedinto phosphatidylcholine liposomes. The RER proteoliposomes

1333M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

and intact RER vesicles had an apparent Km value in the lowmicromolar range. ATP transport was time- and temperature-dependent, inhibited by DIDS, but was not affected byatractyloside, a specific inhibitor of mitochondrial ADP/ATPcarrier. ATP transport was cis-inhibited strongly by ADP andweakly by AMP. ADP-preloaded RER proteoliposomes showeda specific increase of ATP transport activity while it was notobserved in AMP-preloaded ones. These results stronglysuggest that ATP and ADP are translocated by an antiportmechanism across rat liver RER membrane and the transporteris apparently different from the corresponding mitochondrialprotein [142].

1.4.2. NAD+/NADP+

Pyridine-nucleotide-dependent oxidoreductases are knownto be present in the ER lumen [25,143–146]. Accordingly, aseparate pyridine nucleotide pool was detected in ER-derivedmicrosomal vesicles [33,147]. Since the enzymes participatingin pyridine nucleotide biosynthesis are present in the cytosoland mitochondria, a pyridine nucleotide transport mechanismshould be supposed. However, microsomal vesicles maintaintheir luminal pyridine nucleotides for long time and the luminalNAD(P)(H)-dependent oxidoreductase activities are latent dueto the limited access of added pyridine nucleotides to the activesites [34,148,149]. These circumstances rule out the possibilityof a uniport-type facilitated diffusion. Further studies areneeded to clarify the transport mechanism and to identify thetransporter.

1.4.3. FADProtein disulfide formation is a crucial step in the folding

process of proteins that traverse the secretory pathway. Aprotein relay delivers electrons from cargo proteins to oxygen(or alternative electron acceptors) in the ER. Two majorcomponents of the system are ER oxidoreductin 1 protein(Ero1p) [150,151], and PDI [152].

Ero1p is an FAD-binding protein, and its activity is highlydependent also on free FAD levels in the luminal compartmentof the ER [153]. FAD is also required for the activity of twoother proteins presumably involved in oxidative protein folding,Erv1p and Fmo1p [154,155].

The presence of a FAD transport system in the ER membraneis supported by ample biochemical evidence. The rapid entry ofFAD and the consequent oxidation of luminal proteins weredemonstrated in yeast microsomes [156]. Biochemical char-acterization of FAD transport in rat liver microsomes revealed asimilar mechanism. The observed bidirectional FAD transportdid not require exogenous energy sources and was inhibited byanion transport inhibitors and atractyloside in either direction[157]. The FAD transport activities in yeast and mammalianmicrosomes showed nearly identical kinetic characteristics. Thereported features of FAD transport suggest that the freecytosolic FAD is allowed to equilibrate with the ER luminalFAD pool by a facilitated diffusion.

This assumption was further supported by the identificationof flavin carrier (Flc) proteins required for FAD transport intothe ER in Candida albicans. Though Flc family exhibit

significant sequence conservation, they do not exhibit homol-ogy to other carrier protein families. Each Flc protein has 9–10transmembrane domains, suggesting that these integral mem-brane proteins can function directly as FAD transporters [158].

Despite the similar FAD transport process in yeast andmammals and the conserved FAD-dependent oxidative proteinfolding machinery, the FLC gene family is fungal-specific, i.e. itsmembers were detected in every fungal genome examined to datebut their homologues have not been found in other eukaryoticgenomes. Therefore, the identification of FAD transporters inhigher eukaryotic species requires further investigation.

1.5. Acetyl-CoA and carnitine

1.5.1. Acetyl-CoASynthesis of sialic acid or sialyl residues of gangliosides and

glycoproteins utilizes acetyl-CoA (Ac-CoA) as the donor ofacetyl group in the transfer catalyzed by acetyltransferases inthe lumen of the Golgi apparatus [159]. Therefore, Ac-CoAsynthesized in the cytosol must reach this compartment.

An Ac-CoA transporter (AT-1) was cloned in 1997. Theprotein contains multiple transmembrane sequences and aleucine zipper motif. It consists of 549 amino acids and has amolecular mass of 60.9 kDa. Immunohistochemistry suggestedthat AT-1 is expressed in the ER membrane [160]. Rat andmouse homologues were also cloned [161,162] and thegenomic structure and promoter regions of the mouse Ac-CoA transporter gene (slc33a1) were determined [163].Although the tightly membrane bound and unstable transporterprotein was never purified, transmembrane hidden Markovmodel (TMHMM) analysis predicts that it contains approxi-mately 11 membrane-spanning domains [164].

Despite the highly tissue-specific distribution of acetylatedgangliosides, AT-1 is expressed ubiquitously. Two majortranscripts (3.3 kb and 4.3 kb) are present in all tissues. Asingle 3.0-kb transcript is ubiquitously expressed in adulttissues of mice, including brain [162]. The expression of AT-1 isdevelopmentally regulated; increased expression occurs duringearly embryonic stages. To date, no human diseases involvingAT-1 have been described [165]. The AT-1 gene is highlyconserved phylogenetically, suggesting that it is particularlysignificant. The precise physiological roles of this transporterprotein, however, remain to be elucidated.

1.5.2. CarnitineAcyl-CoAs generated in the cytosol are substrates for several

metabolic pathways in mitochondria, peroxisomes and ER.Acyl-CoA thioesters are utilized in the ER lumen by – amongothers – a diacylglycerol acyltransferase that is involved intriacylglycerol synthesis and an acyl-CoA:cholesterol acyl-transferase involved in cholesterol metabolism. Since longchain acyl-CoA cannot be transported directly, its acyl group istranslocated as acyl-carnitine into the above mentionedorganelles. These shuttle systems include transferase enzymesfor the transfer of acyl groups from CoA to carnitine and viceversa, and a transporter for the translocation of acyl-carnitineand carnitine [166]. Accordingly, carnitine acetyltransferase and

1334 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

carnitine palmitoyltransferase activities were also detected inthe ER lumen [166].

Existence of transport systems for carnitine and acyl-carnitine in the ER membrane was proposed on the basis ofmicrosomal experiments. The membrane of intact livermicrosomes was impermeable to palmitoyl-CoA, while bothpalmitoyl-carnitine and free carnitine entered the vesiclesefficiently. The transport process was different from themitochondrial inner membrane carnitine/acylcarnitine exchangecarrier, since it was inhibited by mersalyl but not by NEM orsulfobetaine [167].

The carnitine transporter was solubilized from rat livermicrosomes and reconstituted into liposomes. The reconstitutedcarnitine transporter catalyzed a first-order uniport reactioninhibited by HgCl2 and DIDS. The reconstituted transporteralso catalyzed carnitine efflux from the proteoliposomes; theefflux was stimulated by externally added long-chain acyl-carnitines. The most probable interpretation of this result is thatthe same transporter that catalyses the uniport of carnitine maybehave as an antiporter for carnitine/long chain acyl-carnitine.Ornithine, arginine, glutamine and lysine were also taken up bythe reconstituted liposomes, though with lower efficiencycompared to carnitine. Thus, the ER carnitine transporter mayalso function as an amino acid transporter [168].

The presence of an acyl-carnitine shuttle mechanism clearlysuggests the need for an ER luminal pool of CoA. However,questions regarding the source and maintenance of this pool invivo remain to be answered.

1.6. Conjugates

Some conjugated products of biotransformation need to betransported across the ER membrane because they are producedor cleaved by luminal enzymes. Glucuronides and sulfoconju-gates, which are the most relevant from this aspect, are chargedat physiological pH. The existence of conjugate transporters inthe ER, therefore, can be postulated. Glucuronides generated byUGTs leave the lumen and appear in the cytosol for beingpumped into the bile or into the blood by ABC transporters ofthe plasma membrane. On the other hand, the entry ofglucuronides in the ER might also have biochemical impor-tance, as they can be hydrolyzed by the luminal β-glucur-onidase [169]. Although no specific disease is known to beassociated with the defective transport of glucuronides acrossthe ER membrane, such abnormality was detected in ajaundiced patient [170]. Accumulation of newly synthesizedbilirubin glucuronide was found in the lumen of livermicrosomes prepared from the patient. Moreover, the sameliver microsomes released 1-naphthol glucuronide at normalrate. It was concluded that the ER export might be regulated byspecific transporters and play a role in the sorting ofglucuronides, destined for export through bile canalicular orbasolateral plasma membranes [170]. The presence of a protein-mediated, facilitated glucuronide traffic was further supportedby the observation that bilirubin diglucuronide is able to crossnative microsomal membranes in contrast to model ERphospholipid membranes [171].

It was found in rat liver microsomes that the membranetraffic of certain phenol-glucuronides is mutually trans-stimulated by UDP-GlcA. It was, therefore, suggested that anantiport mechanism can assist the UGTs by allowing the entryof a substrate (UDP-GlcA) in exchange for the product(glucuronide) of the enzyme [122]. A protein mediated transportof estradiol glucuronide was demonstrated in rat liver micro-somes. The transport was shown to be bi-directional, time- andtemperature-dependent and it could be inhibited by DIDS. Thelack of ATP- or glutathione-dependence indicated that thismechanism is distinct from the glucuronide transporters of theplasma membrane. Cis-inhibition of the unknown ER transpor-ter suggested that its substrate specificity was directed towardselected sulfo- and glucuronoconjugates [172]. Competitiveinhibition between various glucuronides for microsomal trans-port was further investigated with direct transport measure-ments. The collected data suggest the existence of multipleglucuronide transporters of overlapping specificities. The maindeterminant of the glucuronide binding to the transporter seemsto be the size of the aglycone [173].

Similarly to glucuronides, sulfoconjugates, especially thoseof steroid hormones must be translocated from the cytosol in theER lumen to have access to an enzyme’s active site. Steroidsulfatase (STS) – also called arylsulfatase C (ARSC) or estronesulfatase (ES) – has a mushroom-like tertiary structure with atransmembrane stem and a luminal cap that contains the catalyticsite near the internal surface of the ER membrane [174].

It is possible that the hydrophobic deconjugated products ofSTS (e.g. steroid hormones) and even the sulfated substrates cancross the membrane through a hydrophobic tunnel formed bythe two transmembrane helices of the enzyme. Nevertheless, thesulfoconjugate transporter of the ER membrane has not beenidentified.

1.7. Oligopeptides

1.7.1. GlutathioneRedox environment of the ER lumen is of great importance

for the oxidative folding of secretory proteins. Glutathione(GSH) and glutathione disulfide (GSSG) constitute the mostimportant redox buffer of animal cells both in the cytosol and inorganelles. The ratio of (GSH)/(GSSG) in the cytosol is 30–100:1 resulting in a redox potential of about −230 mV. Thelumen of the ER is more oxidized (−180 mV) with a 1–3:1 ratioof (GSH)/(GSSG) [175]. The high proportion of the totalpopulation of glutathione was found to be in mixed disulfideswith proteins, which may play role as a GSH reserve and as acomponent of a redox buffering system [176]. Two possiblemechanisms were presumed to maintain the oxidative environ-ment of the ER lumen. The first is the preferential uptake ofGSSG, or the efflux (or exocytosis) of the reduced form.Alternatively, ER-resident enzymes could produce oxidizingcompounds in the lumen.

Investigation of the microsomal transport of the redox couplesupported the second possibility. A bi-directional, saturableGSH transport was found in rat liver microsomal vesicles, whichcould be inhibited by the anion transport blockers flufenamic

1335M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

acid and DIDS [177]. A part of the GSH taken up by the vesicleswas metabolized to GSSG in the lumen. Microsomal membranewas virtually impermeable to GSSG. This result was contrary toprevious findings indicating the preferential transport of GSSG[175], which can be explained by the preparation and storage ofmicrosomes in a reducing buffer.

Rapid permeation of GSH and GSSG across SR membranewas demonstrated in skeletal muscle. It was also proved that thelocal redox potential regulate ryanodine receptor type 1 (RyR1)due to the alterations in the oxidation state of its hypersensitivethiol groups. Thiol oxidation by reactive oxygen species,GSSG, and other thiol reagents activate, while reducing agents,such as GSH, dithiothreitol and mercaptoethanol inhibit thechannel. Accordingly, RyR1 from skeletal muscle can functionas a transmembrane redox sensor and the transport of GSH/GSSG is involved in regulating of this channel [178]. It wasfound that GSH transport across ER/SR membrane correlatedwith the abundance of RyR1. The transport was fastest inmuscle terminal cisternae, fast in muscle microsomes and slowin liver, heart and brain microsomes. GSH influx could beinhibited by RyR1 blockers, and the inhibitory effect wascounteracted by RyR1 agonists [179]. These results suggest thatRyR1 may act as glutathione transporter on its own, or mayinteract with a putative GSH/GSSG transporter. The uptake ofradiolabeled GSH into SR vesicles was demonstrated, using animproved rapid-filtration assay where intravesicular glutathionewas trapped by the efflux blocker lanthanum chloride. Theuptake could be blocked by RyR inhibitor ruthenium red andryanodine. An SR-like GSH transport appeared in microsomesobtained from a HEK-293 cell line transfected with RyR1 [180].These observations strongly suggest that RyR1 itself mediatesGSH transport through the SR membranes of skeletal muscle.

1.7.2. Antigenic oligopeptidesThe peptides needed for antigen presentation are generated

by the proteasome complex in the cytosol. They need to betransported into the ER lumen where they are further processedand loaded onto major histocompatibility complex (MHC) classI molecules before they are finally displayed on the cell surface.The macromolecular peptide-loading complex [181] containsthe transporter associated with antigen processing (TAP) thatmediates the entry of antigenic peptides into the lumen of theER [182]. TAP is a heterodimer of TAP1 and TAP2 and belongsto the ATP-binding casette superfamily [183]. Each subunitcontains a transmembrane domain that binds to oligopeptidesand forms the translocation pore, and a cytosolic nucleotide-binding domain which provides energy for translocations byATP hydrolysis [182]. Expression of TAP – as a key factor ofantigen presentation – is upregulated during infection. How-ever, tumors and viruses may escape immune surveillance bysuppressing TAP function [184].

2. Non-specific transport in the ER

It has been repeatedly observed that microsomal vesiclesderived from the ER exhibit a basal permeability to variouscompounds – including xenogenous molecules – which

presumably do not have strictly specific transporters. A varietyof structurally unrelated compounds is (or should be) able tocross the ER membrane, but only a few transporters have beenreported so far [185]. This non-specific permeability is usuallyattributed to the damage or to the improper orientation of themembrane. However, recent observations indicate the role oftranslocon protein channel in the permeation of small moleculesacross the ER membrane.

Co-translational protein translocation and integration into themembrane of the ER occur at sites termed translocon (for areview, see [186]). This pore is formed by the heterotrimericprotein complex Sec61αβγ that forms an aqueous pore. This isthe largest pore in the ER membrane, with an estimateddiameter of 40–60 Å in the ribosome-bound state and a smallerdiameter of 9–15 Å in the ribosome-free state [187].

The tight ribosome-channel junction on the cytosolic sidecan prevent the passage of small molecules. In addition, thepermeability of the pore is regulated by the luminal chaperoneBiP. The structure of the pore was elucidated by X-raycrystallography, which revealed a permeability barrier insidethe channel. The channel has an hourglass shape, and a small á-helical segment seems to act as a plug, which is removed (i.e.the pore is opened) by the binding of ribosomes [188].

A recent report demonstrated the persistent binding ofribosomes after termination of translocation. Thus, the ERmembrane might contain a large number of ribosome-boundtranslocons with their pores unoccupied by polypeptides [189].In this empty state, the complex seems to allow the passage ofsmall molecules.

Electrophysiological studies, using pancreatic rough micro-somes, demonstrated that polarized molecules like glutamateand 4-(2-hydroxyethyl)piperazine-1-ethanesulfonic acid(HEPES) can permeate the translocon channel. A large increasein membrane conductance was detected after addition ofpuromycin, an antibiotic that releases nascent polypeptidechains from the ribosomes [190].

It was proved with fluorescent probes that secretory proteinsare in an aqueous milieu when undergoing translocation acrossthe translocon channels [191]. The pore size was determined byapplying fluorescent quenchers of different sizes, and it wasshown that I− and NAD+ can cross the channel [192].

The permeation of a small neutral molecule, 4-methyl-umbelliferyl-α-D-glucopyranoside was tested by measuring therate of its hydrolysis in the ER. The reaction is catalyzed by aluminal α-glucosidase, hence, it is dependent on the substrate'sentry into the ER. The basal entry of the dye into the ER ofChinese hamster ovary-S cells was increased by the addition ofpuromycin and pactamycin [193,194]. This effect was inhibitedby cycloheximide, an elongation and release inhibitor antibiotic[194]. It was concluded that translocon channels are permeableto 4-methyl-umbelliferyl-α-D-glucopyranoside as long as emptyribosomes remain bound to them.

The role of translocon pores in the basal Ca2+ leak from theER was studied in pancreatic acinar cells, and a puromycin-induced Ca2+ efflux was found [195]. It was also demonstratedthat puromycin depletes the ER calcium store and this effectcould be prevented by anisomycin, an inhibitor of peptidyl

1336 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

transferase [196]. Calcium leak through the translocon channelswas further evidenced, and it was demonstrated that this leakagecan activate the store-operated Ca2+ channels, which suggeststhe possibility of a translocon-related calcium signaling [197].

Contribution of translocon peptide channel to the permeationof low molecular mass anions, such as UDP-GlcA was alsodemonstrated in rat liver microsomes. Puromycin increased theactivity of the luminal UGTs in intact rough, but not in smoothmicrosomes. This effect could be inhibited by pretreatement withanisomycin or an antibody against Sec61 translocon component.It was concluded that translocons may take part in the substratesupply of luminal enzymes, in the transport of variousxenobiotics, FAD and exogenous biotin derivatives [128].

3. Conclusion

Main metabolic functions of the ER, such as glucoseproduction, pre-receptorial steroid reactivation (including corti-sone reduction and steroid sulfate hydrolysis), glucuronidationand deglucuronidation, protein synthesis, oxidative folding andglycosylation, antigen processing and presentation, are allassociated to transmembrane traffic ofmetabolites. It is, therefore,doubtless that characterization of ER transport is indispensable forthe understanding of ER biochemistry. All transport activitiesstudied so far have been found to be protein-mediated. Moreover,competition-patterns and comparison of inhibitor-profiles oftenindicate that the studiedmolecules are not transported by the sameprotein.Nevertheless, it cannot be excluded that ER transporters–similarly to several ER enzymes – are specific to a group ofmolecules sharing common structural features rather than to aparticular compound. In other terms, the large number ofobserved transport activities might be attributed to a few proteinswith wide and perhaps overlapping substrate ranges. Identifica-tion of these ER transporters is a major goal of on-goinginvestigation. Hopefully, the progress in this field will beaccelerated by the recent development in molecular biologytechniques and by the growing protein and nucleic acid databases.Cloning and sequencing of G6PT and TAP subunits were majorachievements that not only revealed the pathomechanism ofGSD1b and certain immunologic diseases but also inspiredfurther research of the compartmentalized metabolism of the ER.

Acknowledgments

Special thanks to Peter Mayinger for his helpful advice. Thiswork was supported by the Hungarian Ministry of Welfare (ETT182/2006 and 183/2006), by the National Office of Researchand Technology (NKFP 1A/056/2004), by the NationalScientific Research Fund (OTKA TS049851 and T048939),by Szentágothai János Knowledge Center, by the ItalianMinistry of University and Research (FIRB, RBAU014PJA),and by the University of Siena (PAR-progetti 2005).

References

[1] G.E. Palade, The endoplasmic reticulum, J. Biophys. Biochem. Cytol. 2(1956) 85–98.

[2] A. Benedetti, G. Banhegyi, A. Burchell, Endoplasmic Reticulum: AMetabolic Compartment, IOS Press, Amsterdam, 2005.

[3] W.J. Arion, A.J. Lange, H.E. Walls, L.M. Ballas, Evidence for theparticipation of independent translocation for phosphate and glucose 6-phosphate in the microsomal glucose-6-phosphatase system. Interactionsof the system with orthophosphate, inorganic pyrophosphate, andcarbamyl phosphate, J. Biol. Chem. 255 (1980) 10396–10406.

[4] M. Uldry, B. Thorens, The SLC2 family of facilitated hexose and polyoltransporters, Pflugers Arch. 447 (2004) 480–489.

[5] G. Meissner, R. Allen, Evidence for two types of rat liver microsomeswith differing permeability to glucose and other small molecules, J. Biol.Chem. 256 (1981) 6413–6422.

[6] G. Banhegyi, P. Marcolongo, A. Burchell, A. Benedetti, Heterogeneity ofglucose transport in rat liver microsomal vesicles, Arch. Biochem.Biophys. 359 (1998) 133–138.

[7] P. Marcolongo, R. Fulceri, R. Giunti, A. Burchell, A. Benedetti,Permeability of liver microsomal membranes to glucose, Biochem.Biophys. Res. Commun. 219 (1996) 916–922.

[8] M. Fehr, H. Takanaga, D.W. Ehrhardt, W.B. Frommer, Evidence for high-capacity bidirectional glucose transport across the endoplasmic reticulummembrane by genetically encoded fluorescence resonance energy transfernanosensors, Mol. Cell. Biol. 25 (2005) 11102–11112.

[9] M.T. Guillam, R. Burcelin, B. Thorens, Normal hepatic glucoseproduction in the absence of GLUT2 reveals an alternative pathway forglucose release from hepatocytes, Proc. Natl. Acad. Sci. U. S. A. 95(1998) 12317–12321.

[10] E.J. Smart, Y. Ying, W.C. Donzell, R.G. Anderson, A role for caveolin intransport of cholesterol from endoplasmic reticulum to plasma mem-brane, J. Biol. Chem. 271 (1996) 29427–29435.

[11] M. Hosokawa, B. Thorens, Glucose release from GLUT2-null hepato-cytes: characterization of a major and a minor pathway, Am. J. Physiol.:Endocrinol. Metab. 282 (2002) E794–E801.

[12] J.C. Molero, J.P. Whitehead, T. Meerloo, D.E. James, Nocodazole inhibitsinsulin-stimulated glucose transport in 3T3-L1 adipocytes via a micro-tubule-independent mechanism, J. Biol. Chem. 276 (2001) 43829–43835.

[13] A. Romanelli, J.F. St-Denis, H. Vidal, S. Tchu, G. van deWerve, Absenceof glucose uptake by liver microsomes: an explanation for the completelatency of glucose dehydrogenase, Biochem. Biophys. Res. Commun.200 (1994) 1491–1497.

[14] I.D. Waddell, H. Scott, A. Grant, A. Burchell, Identification andcharacterization of a hepatic microsomal glucose transport protein, T3of the glucose-6-phosphatase system? Biochem. J. 275 (1991) 363–367.

[15] I.D. Waddell, A.G. Zomerschoe, M.W. Voice, A. Burchell, Cloning andexpression of a hepatic microsomal glucose transport protein. Compar-ison with liver plasma-membrane glucose-transport protein GLUT 2,Biochem. J. 286 (1992) 173–177.

[16] A. Burchell, A re-evaluation of GLUT 7, Biochem. J. 331 (1998) 973.[17] G. Banhegyi, L. Braun, M. Csala, F. Puskas, J. Mandl, Ascorbate

metabolism and its regulation in animals, Free Radical Biol. Med. 23(1997) 793–803.

[18] G. Banhegyi, P. Marcolongo, F. Puskas, R. Fulceri, J. Mandl, A.Benedetti, Dehydroascorbate and ascorbate transport in rat livermicrosomal vesicles, J. Biol. Chem. 273 (1998) 2758–2762.

[19] M. Csala, V. Mile, A. Benedetti, J. Mandl, G. Banhegyi, Ascorbateoxidation is a prerequisite for its transport into rat liver microsomalvesicles, Biochem. J. 349 (2000) 413–415.

[20] W. Wells, D. Xu, Y. Yang, P. Rocque, Mammalian thioltransferase(glutaredoxin) and protein disulfide isomerase have dehydroascorbatereductase activity, J. Biol. Chem. 265 (1990) 15361–15364.

[21] M. Csala, L. Braun, V. Mile, T. Kardon, A. Szarka, P. Kupcsulik, J.Mandl, G. Banhegyi, Ascorbate-mediated electron transfer in proteinthiol oxidation in the endoplasmic reticulum, FEBS Lett. 460 (1999)539–543.

[22] G. Nardai, L. Braun, M. Csala, V. Mile, P. Csermely, A. Benedetti, J.Mandl, G. Banhegyi, Protein-disulfide isomerase- and protein thiol-dependent dehydroascorbate reduction and ascorbate accumulation in thelumen of the endoplasmic reticulum, J. Biol. Chem. 276 (2001)8825–8828.

1337M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

[23] J.M. Upston, A. Karjalainen, F.L. Bygrave, R. Stocker, Efflux of hepaticascorbate: a potential contributor to the maintenance of plasma vitamin C,Biochem. J. 342 (1999) 49–56.

[24] E. van Schaftingen, I. Gerin, The glucose-6-phosphatase system,Biochem. J. 362 (2002) 513–532.

[25] T. Takahashi, S.H. Hori, Intramembraneous localization of rat livermicrosomal hexose-6-phosphate dehydrogenase and membrane per-meability to its substrates, Biochim. Biophys. Acta 524 (1978)262–276.

[26] J. Ozols, Isolation and the complete amino acid sequence of lumenalendoplasmic reticulum glucose-6-phosphate dehydrogenase, Proc. Natl.Acad. Sci. U. S. A. 90 (1993) 5302–5306.

[27] H.G. Hers, J. Berthet, L. Berthet, C. De Duve, The hexose-phosphatasesystem. III. Intracellular localization of enzymes by fractional centrifuga-tion, Bull. Soc. Chim. Biol. (Paris) 33 (1951) 21–41.

[28] A. Leskes, P. Siekevitz, G.E. Palade, Differentiation of endoplasmicreticulum in hepatocytes: I. Glucose-6-phosphatase distribution in situ, J.Cell Biol. 49 (1971) 264–287.

[29] A. Leskes, P. Siekevitz, G.E. Palade, Differentiation of endoplasmicreticulum in hepatocytes: II. Glucose-6-phosphatase in rough micro-somes, J. Cell Biol. 49 (1971) 288–302.

[30] A. Benedetti, R. Fulceri, M. Comporti, Calcium sequestration activity inrat liver microsomes. Evidence for a cooperation of calcium transportwith glucose-6-phosphatase, Biochim. Biophys. Acta 816 (1985)267–277.

[31] A. Benedetti, R. Fulceri, M. Ferro, M. Comporti, On a possible role forglucose-6-phosphatase in the regulation of liver cell cytosolic calciumconcentration, Trends Biochem. Sci. 11 (1986) 284–285.

[32] I. Gerin, G. Noel, E. Van Schaftingen, Novel arguments in favor of thesubstrate-transport model of glucose-6-phosphatase, Diabetes 50 (2001)1531–1538.

[33] S. Piccirella, I. Czegle, B. Lizak, E. Margittai, S. Senesi, E. Papp, M.Csala, R. Fulceri, P. Csermely, J. Mandl, A. Benedetti, G. Banhegyi,Uncoupled redox systems in the lumen of the endoplasmic reticulum.Pyridine nucleotides stay reduced in an oxidative environment, J. Biol.Chem. 281 (2006) 4671–4677.

[34] G. Banhegyi, A. Benedetti, R. Fulceri, S. Senesi, Cooperativity between11beta-hydroxysteroid dehydrogenase type 1 and hexose-6-phosphatedehydrogenase in the lumen of the endoplasmic reticulum, J. Biol. Chem.279 (2004) 27017–27021.

[35] A.G. Atanasov, L.G. Nashev, R.A. Schweizer, C. Frick, A. Odermatt,Hexose-6-phosphate dehydrogenase determines the reaction direction of11beta-hydroxysteroid dehydrogenase type 1 as an oxoreductase, FEBSLett. 571 (2004) 129–133.

[36] A. Odermatt, A.G. Atanasov, Z. Balazs, R.A.S. Schweizer, L.G. Nashev,D. Schuster, T. Langer, Why is 11[beta]-hydroxysteroid dehydrogenasetype 1 facing the endoplasmic reticulum lumen?: Physiological relevanceof the membrane topology of 11[beta]-HSD1, Mol. Cell. Endocrinol. 248(2006) 15–23.

[37] J.F. St-Denis, B. Comte, D.K. Nguyen, E. Seidman, K. Paradis, E. Levy,G. van de Werve, A conformational model for the human livermicrosomal glucose-6-phosphatase system: evidence from rapid kineticsand defects in glycogen storage disease type 1, J. Clin. Endocrinol.Metab. 79 (1994) 955–959.

[38] G. van de Werve, A. Lange, C. Newgard, M.C. Mechin, Y. Li, A.Berteloot, New lessons in the regulation of glucose metabolism taught bythe glucose 6-phosphatase system, Eur. J. Biochem. 267 (2000)1533–1549.

[39] W.J. Arion, B.K. Wallin, A.J. Lange, L.M. Ballas, On the involvement ofa glucose 6-phosphate transport system in the function of microsomalglucose 6-phosphatase, Mol. Cell. Biochem. 6 (1975) 75–83.

[40] R. Fulceri, G. Bellomo, A. Gamberucci, H.M. Scott, A. Burchell, A.Benedetti, Permeability of rat liver microsomal membrane to glucose 6-phosphate, Biochem. J. 286 (1992) 813–817.

[41] G. Banhegyi, P. Marcolongo, R. Fulceri, C. Hinds, A. Burchell, A.Benedetti, Demonstration of a metabolically active glucose-6-phosphatepool in the lumen of liver microsomal vesicles, J. Biol. Chem. 272 (1997)13584–13590.

[42] I. Gerin, E. Van Schaftingen, Evidence for glucose-6-phosphate transportin rat liver microsomes, FEBS Lett. 517 (2002) 257–260.

[43] Y. Igarashi, S. Kato, K. Narisawa, K. Tada, Y. Amano, T. Mori, S.Takeuchi, A direct evidence for defect in glucose-6-phosphate transportsystem in hepatic microsomal membrane of glycogen storage disease typeIB, Biochem. Biophys. Res. Commun. 119 (1984) 593–597.

[44] K. Tada, K. Narisawa, Y. Igarashi, S. Kato, Glycogen storage disease typeIB: a new model of genetic disorders involving the transport system ofintracellular membrane, Biochem. Med. 33 (1985) 215–222.

[45] P. Marcolongo, G. Banhegyi, A. Benedetti, C.J. Hinds, A. Burchell, Livermicrosomal transport of glucose-6-phosphate, glucose, and phosphate intype 1 glycogen storage disease, J. Clin. Endocrinol. Metab. 83 (1998)224–229.

[46] I. Gerin, M. Veiga-da-Cunha, Y. Achouri, J.F. Collet, E. Van Schaftingen,Sequence of a putative glucose 6-phosphate translocase, mutated inglycogen storage disease type Ib, FEBS Lett. 419 (1997) 235–238.

[47] M.A. Zoccoli, R.R. Hoopes, M.L. Karnovsky, Identification of a rat livermicrosomal polypeptide involved in the transport of glucose 6-phosphate.Labeling with 4,4′-diisothiocyano-1,2-diphenyl[3H]ethane-2,2′-disulfo-nic acid, J. Biol. Chem. 257 (1982) 3919–3924.

[48] M.A. Zoccoli, M.L. Karnovsky, Effect of two inhibitors of aniontransport on the hydrolysis of glucose 6-phosphate by rat livermicrosomes. Covalent modification of the glucose 6-P transportcomponent, J. Biol. Chem. 255 (1980) 1113–1119.

[49] E. Clottes, A. Burchell, Three thiol groups are important for the activity ofthe liver microsomal glucose-6-phosphatase system. Unusual behavior ofone thiol located in the glucose-6-phosphate translocase, J. Biol. Chem.273 (1998) 19391–19397.

[50] J.D. Foster, A.M. Bode, R.C. Nordlie, Time-dependent inhibition ofglucose 6-phosphatase by 3-mercaptopicolinic acid, Biochim. Biophys.Acta 1208 (1994) 222–228.

[51] M. Speth, H.U. Schulze, Protease inhibitors but not proteases inhibit theglucose-6-phosphatase of native rat liver microsomes, Biochem.Biophys. Res. Commun. 183 (1992) 590–597.

[52] R. Fulceri, A. Gamberucci, H.M. Scott, R. Giunti, A. Burchell, A.Benedetti, Fatty acyl-CoA esters inhibit glucose-6-phosphatase in ratliver microsomes, Biochem. J. 307 (1995) 391–397.

[53] G. Mithieux, C. Zitoun, Mechanisms by which fatty-acyl-CoA estersinhibit or activate glucose-6-phosphatase in intact and detergent-treatedrat liver microsomes, Eur. J. Biochem. 235 (1996) 799–803.

[54] W.J. Arion, W.K. Canfield, F.C. Ramos, P.W. Schindler, H.J. Burger, H.Hemmerle, G. Schubert, P. Below, A.W. Herling, Chlorogenic acid andhydroxynitrobenzaldehyde: new inhibitors of hepatic glucose 6-phos-phatase, Arch. Biochem. Biophys. 339 (1997) 315–322.

[55] H. Hemmerle, H.J. Burger, P. Below, G. Schubert, R. Rippel, P.W.Schindler, E. Paulus, A.W. Herling, Chlorogenic acid and syntheticchlorogenic acid derivatives: novel inhibitors of hepatic glucose-6-phosphate translocase, J. Med. Chem. 40 (1997) 137–145.

[56] W.J. Arion, W.K. Canfield, F.C. Ramos, M.L. Su, H.J. Burger, H.Hemmerle, G. Schubert, P. Below, A.W. Herling, Chlorogenic acidanalogue S 3483: a potent competitive inhibitor of the hepatic and renalglucose-6-phosphatase systems, Arch. Biochem. Biophys. 351 (1998)279–285.

[57] A.W. Herling, H.J. Burger, D. Schwab, H. Hemmerle, P. Below, G.Schubert, Pharmacodynamic profile of a novel inhibitor of the hepaticglucose-6-phosphatase system, Am. J. Physiol. 274 (1998) G1087–G1093.

[58] R. Leuzzi, G. Banhegyi, T. Kardon, P. Marcolongo, P.L. Capecchi, H.J.Burger, A. Benedetti, R. Fulceri, Inhibition of microsomal glucose-6-phosphate transport in human neutrophils results in apoptosis: a potentialexplanation for neutrophil dysfunction in glycogen storage disease type1b, Blood 101 (2003) 2381–2387.

[59] G.T. Cori, C.F. Cori, Glucose-6-phosphatase of the liver in glycogenstorage disease, J. Biol. Chem. 199 (1952) 661–667.

[60] P. Marcolongo, V. Barone, G. Priori, B. Pirola, S. Giglio, G. Biasucci, E.Zammarchi, G. Parenti, A. Burchell, A. Benedetti, V. Sorrentino,Structure and mutation analysis of the glycogen storage disease type 1bgene, FEBS Lett. 436 (1998) 247–250.

[61] H. Hiraiwa, C.J. Pan, B. Lin, S.W. Moses, J.Y. Chou, Inactivation of the

1338 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

glucose 6-phosphate transporter causes glycogen storage disease type 1b,J. Biol. Chem. 274 (1999) 5532–5536.

[62] I. Gerin, M. Veiga-da-Cunha, G. Noel, E. Van Schaftingen, Structure ofthe gene mutated in glycogen storage disease type Ib, Gene 227 (1999)189–195.

[63] B. Annabi, H. Hiraiwa, B.C. Mansfield, K.J. Lei, T. Ubagai, M.H.Polymeropoulos, S.W. Moses, R. Parvari, E. Hershkovitz, H. Mandel, M.Fryman, J.Y. Chou, The gene for glycogen-storage disease type 1b mapsto chromosome 11q23, Am. J. Hum. Genet. 62 (1998) 400–405.

[64] D. Melis, R. Fulceri, G. Parenti, P. Marcolongo, R. Gatti, R. Parini, E.Riva, R. Della Casa, E. Zammarchi, G. Andria, A. Benedetti, Genotype/phenotype correlation in glycogen storage disease type 1b: a multicentrestudy and review of the literature, Eur. J. Pediatr. 164 (2005) 501–508.

[65] B. Lin, B. Annabi, H. Hiraiwa, C.J. Pan, J.Y. Chou, Cloning andcharacterization of cDNAs encoding a candidate glycogen storage diseasetype 1b protein in rodents, J. Biol. Chem. 273 (1998) 31656–31660.

[66] L.Y. Chen, J.J. Shieh, B. Lin, C.J. Pan, J.L. Gao, P.M. Murphy, T.F. Roe,S. Moses, J.M. Ward, E.J. Lee, H. Westphal, B.C. Mansfield, J.Y. Chou,Impaired glucose homeostasis, neutrophil trafficking and function in micelacking the glucose-6-phosphate transporter, Hum. Mol. Genet. 12 (2003)2547–2558.

[67] S.Y. Kim, A.D. Nguyen, J.L. Gao, P.M. Murphy, B.C. Mansfield, J.Y.Chou, Bone marrow-derived cells require a functional glucose 6-phosphate transporter for normal myeloid functions, J. Biol. Chem. 281(2006) 28794–28801.

[68] K.J. Lei, L.L. Shelly, C.J. Pan, J.B. Sidbury, J.Y. Chou, Mutations in theglucose-6-phosphatase gene that cause glycogen storage disease type 1a,Science 262 (1993) 580–583.

[69] S. Senesi, P. Marcolongo, T. Kardon, G. Bucci, A. Sukhodub, A.Burchell, A. Benedetti, R. Fulceri, Immunodetection of the expression ofmicrosomal proteins encoded by the glucose 6-phosphate transportergene, Biochem. J. 389 (2005) 57–62.

[70] C. Middleditch, E. Clottes, A. Burchell, A different isoform of thetransport protein mutated in the glycogen storage disease 1b is expressedin brain, FEBS Lett. 433 (1998) 33–36.

[71] B. Lin, C.J. Pan, J.Y. Chou, Human variant glucose-6-phosphatetransporter is active in microsomal transport, Hum. Genet. 107 (2000)526–529.

[72] J.J. Shieh, C.J. Pan, B.C. Mansfield, J.Y. Chou, A potential new role formuscle in blood glucose homeostasis, J. Biol. Chem. 279 (2004)26215–26219.

[73] R. Fulceri, T. Kardon, G. Banhegyi, W.F. Pralong, A. Gamberucci, P.Marcolongo, A. Benedetti, Glucose-6-phosphatase in the insulinsecreting cell line INS-1, Biochem. Biophys. Res. Commun. 275(2000) 103–107.

[74] Y. Li, M.C. Mechin, G. van de Werve, Diabetes affects similarly thecatalytic subunit and putative glucose-6-phosphate translocase ofglucose-6-phosphatase, J. Biol. Chem. 274 (1999) 33866–33868.

[75] C.J. Pan, B. Lin, J.Y. Chou, Transmembrane topology of human glucose6-phosphate transporter, J. Biol. Chem. 274 (1999) 13865–13869.

[76] J. Almqvist, Y. Huang, S. Hovmoller, D.N. Wang, Homology modelingof the human microsomal glucose 6-phosphate transporter explains themutations that cause the glycogen storage disease type Ib, Biochemistry43 (2004) 9289–9297.

[77] R. Fulceri, G. Bellomo, A. Gamberucci, A. Benedetti, Liver glucose-6-phosphatase activity is not modulated by physiological intracellular Ca2+concentrations, Biochem. J. 275 (1991) 805–807.

[78] R. Leuzzi, R. Fulceri, P. Marcolongo, G. Banhegyi, E. Zammarchi, K.Stafford, A. Burchell, A. Benedetti, Glucose 6-phosphate transport infibroblast microsomes from glycogen storage disease type 1b patients:evidence for multiple glucose 6-phosphate transport systems, Biochem. J.357 (2001) 557–562.

[79] R. Fulceri, G. Bellomo, A. Gamberucci, A. Benedetti, MgATP-dependentaccumulation of calcium ions and inorganic phosphate in a liver reticularpool, Biochem. J. 272 (1990) 549–552.

[80] R. Fulceri, G. Bellomo, A. Gamberucci, A. Romani, A. Benedetti,Physiological concentrations of inorganic phosphate affect MgATP-dependent Ca2+ storage and inositol trisphosphate-induced Ca2+ efflux

in microsomal vesicles from non-hepatic cells, Biochem. J. 289 (1993)299–306.

[81] X.J. Meng, R.T. Timmer, R.B. Gunn, R.F. Abercrombie, Separate entrypathways for phosphate and oxalate in rat brain microsomes, Am. J.Physiol.: Cell Physiol. 278 (2000) C1183–C1190.

[82] M.W. Fryer, J.M. West, D.G. Stephenson, Phosphate transport into thesarcoplasmic reticulum of skinned fibres from rat skeletal muscle, J.Muscle Res. Cell Motil. 18 (1997) 161–167.

[83] H.I. Stefanova, S.D. Jane, J.M. East, A.G. Lee, Effects of Mg2+ and ATPon the phosphate transporter of sarcoplasmic reticulum, Biochim.Biophys. Acta 1064 (1991) 329–334.

[84] D.R. Laver, G.K.E. Lenz, A.F. Dulhunty, Phosphate ion channels insarcoplasmic reticulum of rabbit skeletal muscle, J. Physiol. (London)535 (2001) 715–728.

[85] Y.T. Chen, A. Burchell, Glycogen storage diseases, in: C.R. Scriver, A.L.Beaudet, W.S. Sly, D. Valle (Eds.), The metabolic and molecular bases ofinherited disease, McGraw-Hill Inc, New York, 1995, pp. 935–965.

[86] R.C. Nordlie, H.M. Scott, I.D. Waddell, R. Hume, A. Burchell, Analysisof human hepatic microsomal glucose-6-phosphatase in clinical condi-tions where the T2 pyrophosphate/phosphate transport protein is absent,Biochem. J. 281 (1992) 859–863.

[87] W. Xie, G. van de Werve, A. Berteloot, Probing into the function of thegene product responsible for glycogen storage disease type Ib, FEBS Lett.504 (2001) 23–26.

[88] G.M. Gibb, G.P. Reid, J.G. Lindsay, Purification and characterization ofthe phosphate/hydroxyl ion antiport protein from rat liver mitochondria,Biochem. J. 238 (1986) 543–551.

[89] A. Werner, L. Dehmelt, P. Nalbant, Na+-dependent phosphate cotran-sporters: the NaPi protein families, J. Exp. Biol. 201 (1998) 3135–3142.

[90] I.D. Waddell, J.G. Lindsay, A. Burchell, The identification of T2; thephosphate/pyrophosphate transport protein of the hepatic microsomalglucose-6-phosphatase system, FEBS Lett. 229 (1988) 179–182.

[91] D. Melis, A.C. Havelaar, E. Verbeek, G.P. Smit, A. Benedetti, G.M.Mancini, F. Verheijen, NPT4, a new microsomal phosphate transporter:mutation analysis in glycogen storage disease type Ic, J. Inherit. Metab.Dis. 27 (2004) 725–733.

[92] M. Csala, S. Senesi, G. Banhegyi, J. Mandl, A. Benedetti, Characteriza-tion of sulfate transport in the hepatic endoplasmic reticulum, Arch.Biochem. Biophys. 440 (2005) 173–180.

[93] N. Ishida, M. Kawakita, Molecular physiology and pathology of thenucleotide sugar transporter family (SLC35), Pflugers Arch. 447 (2004)768–775.

[94] C. D'Alessio, J.J. Caramelo, A.J. Parodi, Absence of nucleosidediphosphatase activities in the yeast secretory pathway does not abolishnucleotide sugar-dependent protein glycosylation, J. Biol. Chem. 280(2005) 40417–40427.

[95] Y. Ito, S. Hagihara, I. Matsuo, K. Totani, Structural approaches to thestudy of oligosaccharides in glycoprotein quality control, Curr. Opin.Struct. Biol. 15 (2005) 481–489.

[96] M. Perez, C.B. Hirschberg, Topography of glycosylation reactions in therough endoplasmic reticulum membrane, J. Biol. Chem. 261 (1986)6822–6830.

[97] F. Vanstapel, N. Blanckaert, Carrier-mediated translocation of uridinediphosphate glucose into the lumen of endoplasmic reticulum-derivedvesicles from rat liver, J. Clin. Invest. 82 (1988) 1113–1122.

[98] E.R. Lazarowski, D.A. Shea, R.C. Boucher, T.K. Harden, Release ofcellular UDP-glucose as a potential extracellular signaling molecule,Mol. Pharmacol. 63 (2003) 1190–1197.

[99] O. Castro, L.Y. Chen, A.J. Parodi, C. Abeijon, Uridine diphosphate-glucose transport into the endoplasmic reticulum of Saccharomycescerevisiae: in vivo and in vitro evidence, Mol. Biol. Cell 10 (1999)1019–1030.

[100] N. Ishida, N. Miura, S. Yoshioka, M. Kawakita, Molecular cloning andcharacterization of a novel isoform of the human UDP-galactosetransporter, and of related complementary DNAs belonging to thenucleotide-sugar transporter gene family, J. Biochem. (Tokyo) 120 (1996)1074–1078.

[101] M. Kainuma, Y. Chiba, M. Takeuchi, Y. Jigami, Overexpression of HUT1

1339M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

gene stimulates in vivo galactosylation by enhancing UDP-galactosetransport activity in Saccharomyces cerevisiae, Yeast 18 (2001)533–541.

[102] H. Nakanishi, K. Nakayama, A. Yokota, H. Tachikawa, N. Takahashi, Y.Jigami, Hut1 proteins identified in Saccharomyces cerevisiae and Schi-zosaccharomyces pombe are functional homologues involved in theprotein-folding process at the endoplasmic reticulum, Yeast 18 (2001)543–554.

[103] L. Norambuena, L. Marchant, P. Berninsone, C.B. Hirschberg, H. Silva, A.Orellana, Transport ofUDP-galactose in plants. Identification and functionalcharacterization of AtUTr1, an Arabidopsis thaliana UDP-galactos/UDP-glucose transporter, J. Biol. Chem. 277 (2002) 32923–32929.

[104] F. Reyes, L. Marchant, L. Norambuena, R. Nilo, H. Silva, A.Orellana, AtUTr1, a UDP-glucose/UDP-galactose transporter fromArabidopsis thaliana, is located in the endoplasmic reticulum and up-regulated by the unfolded protein response, J. Biol. Chem. 281 (2006)9145–9151.

[105] N.J. Kuhn, A. White, Evidence for specific transport of uridinediphosphate galactose across the Golgi membrane of rat mammarygland, Biochem. J. 154 (1976) 243–244.

[106] C.B. Hirschberg, P.W. Robbins, C. Abeijon, Transporters of nucleotidesugars, ATP, and nucleotide sulfate in the endoplasmic reticulum andGolgi apparatus, Annu. Rev. Biochem. 67 (1998) 49–69.

[107] M. Perez, C.B. Hirschberg, Translocation of UDP-N-acetylglucosamineinto vesicles derived from rat liver rough endoplasmic reticulum andGolgi apparatus, J. Biol. Chem. 260 (1985) 4671–4678.

[108] H. Sprong, S. Degroote, T. Nilsson, M. Kawakita, N. Ishida, P. van derSluijs, G. van Meer, Association of the Golgi UDP-galactose transporterwith UDP-galactose:ceramide galactosyltransferase allows UDP-galac-tose import in the endoplasmic reticulum, Mol. Biol. Cell 14 (2003)3482–3493.

[109] A.E. Kearns, B.M. Vertel, N.B. Schwartz, Topography of glycosylationand UDP-xylose production, J. Biol. Chem. 268 (1993) 11097–11104.

[110] N. Nuwayhid, J.H. Glaser, J.C. Johnson, H.E. Conrad, S.C. Hauser, C.B.Hirschberg, Xylosylation and glucuronosylation reactions in rat liverGolgi apparatus and endoplasmic reticulum, J. Biol. Chem. 261 (1986)12936–12941.

[111] X. Bossuyt, N. Blanckaert, Uridine diphosphoxylose enhances hepaticmicrosomal UDP-glucuronosyltransferase activity by stimulating trans-port of UDP-glucuronic acid across the endoplasmic reticulummembrane, Biochem. J. 315 (1996) 189–193.

[112] X. Bossuyt, N. Blanckaert, Differential regulation of UDP-GlcUAtransport in endoplasmic reticulum and in Golgi membranes, J. Hepatol.34 (2001) 210–214.

[113] S.C. Hauser, J.C. Ziurys, J.L. Gollan, A membrane transporter mediatesaccess of uridine 5′-diphosphoglucuronic acid from the cytosol into theendoplasmic reticulum of rat hepatocytes: implications for glucuronida-tion reactions, Biochim. Biophys. Acta 967 (1988) 149–157.

[114] X. Bossuyt, N. Blanckaert, Carrier-mediated transport of intact UDP-glucuronic acid into the lumen of endoplasmic-reticulum-derived vesiclesfrom rat liver, Biochem. J. 302 (1994) 261–269.

[115] A. Radominska, C. Berg, S. Treat, J.M. Little, R. Lester, J.L. Gollan, R.R.Drake, Characterization of UDP-glucuronic acid transport in rat livermicrosomal vesicles with photoaffinity analogs, Biochim. Biophys. Acta1195 (1994) 63–70.

[116] X. Bossuyt, N. Blanckaert, Carrier-mediated transport of uridinediphosphoglucuronic acid across the endoplasmic reticulum membraneis a prerequisite for UDP-glucuronosyltransferase activity in rat liver,Biochem. J. 323 (1997) 645–648.

[117] C.L. Berg, A. Radominska, R. Lester, J.L. Gollan, Membrane transloca-tion and regulation of uridine diphosphate-glucuronic acid uptake in ratliver microsomal vesicles, Gastroenterology 108 (1995) 183–192.

[118] S. Ikushiro, Y. Emi, S. Kimura, T. Iyanagi, Chemical modification of rathepatic microsomes with N-ethylmaleimide results in inactivation of bothUDP-N-acetylglucosamine-dependent stimulation of glucuronidation andUDP-glucuronic acid uptake, Biochim. Biophys. Acta 1428 (1999)388–396.

[119] E. Battaglia, A. Radominska-Pandya, A functional role for histidyl

residues of the UDP-glucuronic acid carrier in rat liver endoplasmicreticulum membranes, Biochemistry 37 (1998) 258–263.

[120] S.C. Hauser, B.J. Ransil, J.C. Ziurys, J.L. Gollan, Interaction of uridine5′-diphosphoglucuronic acid with microsomal UDP-glucuronosyltrans-ferase in primate liver: the facilitating role of uridine 5′-diphospho-N-acetylglucosamine, Biochim. Biophys. Acta 967 (1988) 141–148.

[121] E. Battaglia, S. Nowell, R.R. Drake, M. Mizeracka, C.L. Berg, J.Magdalou, S. Fournel-Gigleux, J.L. Gollan, R. Lester, A. Radominska,Two kinetically-distinct components of UDP-glucuronic acid transport inrat liver endoplasmic reticulum, Biochim. Biophys. Acta 1283 (1996)223–231.

[122] G. Banhegyi, L. Braun, P. Marcolongo, M. Csala, R. Fulceri, J. Mandl, A.Benedetti, Evidence for an UDP-glucuronic acid/phenol glucuronideantiport in rat liver microsomal vesicles, Biochem. J. 315 (1996)171–176.

[123] X. Bossuyt, N. Blanckaert, Mechanism of stimulation of microsomalUDP-glucuronosyltransferase by UDP-N-acetylglucosamine, Biochem.J. 305 (1995) 321–328.

[124] S. Ikushiro, Y. Emi, T. Iyanagi, Protein–protein interactions betweenUDP-glucuronosyltransferase isozymes in rat hepatic microsomes,Biochemistry 36 (1997) 7154–7161.

[125] B. Haeger, R. de Brito, T. Hallinan, Diazobenzenesulphonate selectivelyabolishes stimulation of glucuronidation by UDP-N-acetylglucosamine,Biochem. J. 192 (1980) 971–974.

[126] T. Kobayashi, J.E. Sleeman, M.W. Coughtrie, B. Burchell, Molecular andfunctional characterization of microsomal UDP-glucuronic acid uptakeby members of the nucleotide sugar transporter (NST) family, Biochem.J. 400 (2006) 281–289.

[127] B. Burchell, P.J. Weatherill, C. Berry, Evidence indicating that UDP-N-acetylglucosamine does not appear to stimulate hepatic microsomal UDP-glucuronosyltransferase by interaction with the catalytic unit of theenzyme, Biochim. Biophys. Acta 735 (1983) 309–313.

[128] B. Lizak, I. Czegle, M. Csala, A. Benedetti, J. Mandl, G. Banhegyi,Translocon pores in the endoplasmic reticulum are permeable to smallanions, Am. J. Physiol.: Cell Physiol. 291 (2006) C511–C517.

[129] M. Muraoka, M. Kawakita, N. Ishida, Molecular characterization ofhuman UDP-glucuronic acid/UDP-N-acetylgalactosamine transporter, anovel nucleotide sugar transporter with dual substrate specificity, FEBSLett. 495 (2001) 87–93.

[130] C. Abeijon, C.B. Hirschberg, Intrinsic membrane glycoproteins withcytosol-oriented sugars in the endoplasmic reticulum, Proc. Natl. Acad.Sci. U. S. A. 85 (1988) 1010–1014.

[131] X. Bossuyt, N. Blanckaert, Functional characterization of carrier-mediated transport of uridine diphosphate N-acetylglucosamine acrossthe endoplasmic reticulum membrane, Eur. J. Biochem. 223 (1994)981–988.

[132] C. Abeijon, E.C. Mandon, P.W. Robbins, C.B. Hirschberg, A mutantyeast deficient in Golgi transport of uridine diphosphate N-acetylgluco-samine, J. Biol. Chem. 271 (1996) 8851–8854.

[133] C. Abeijon, P.W. Robbins, C.B. Hirschberg, Molecular cloning of theGolgi apparatus uridine diphosphate-N-acetylglucosamine transporterfrom Kluyveromyces lactis, Proc. Natl. Acad. Sci. U. S. A. 93 (1996)5963–5968.

[134] E. Guillen, C. Abeijon, C.B. Hirschberg, Mammalian Golgi apparatusUDP-N-acetylglucosamine transporter: molecular cloning by phenotypiccorrection of a yeast mutant, Proc. Natl. Acad. Sci. U. S. A. 95 (1998)7888–7892.

[135] N. Ishida, S. Yoshioka, Y. Chiba, M. Takeuchi, M. Kawakita, Molecularcloning and functional expression of the human Golgi UDP-N-acetylglucosamine transporter, J. Biochem. (Tokyo) 126 (1999) 68–77.

[136] C.A. Clairmont, A. De Maio, C.B. Hirschberg, Translocation of ATP intothe lumen of rough endoplasmic reticulum-derived vesicles and itsbinding to luminal proteins including BiP (GRP 78) and GRP 94, J. Biol.Chem. 267 (1992) 3983–3990.

[137] P. Mayinger, D.I. Meyer, An ATP transporter is required for proteintranslocation into the yeast endoplasmic reticulum, EMBO J. 12 (1993)659–666.

[138] E. Guillen, C.B. Hirschberg, Transport of adenosine triphosphate into

1340 M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

endoplasmic reticulum proteoliposomes, Biochemistry 34 (1995)5472–5476.

[139] P. Mayinger, V.A. Bankaitis, D.I. Meyer, Sac1p mediates theadenosine triphosphate transport into yeast endoplasmic reticulumthat is required for protein translocation, J. Cell Biol. 131 (1995)1377–1386.

[140] K.U. Kochendorfer, A.R. Then, B.G. Kearns, V.A. Bankaitis, P.Mayinger, Sac1p plays a crucial role in microsomal ATP transport,which is distinct from its function in Golgi phospholipid metabolism,EMBO J. 18 (1999) 1506–1515.

[141] S. Kim, S. Shin, J. Park, Identification of the ATP transporter of rat liverrough endoplasmic reticulum via photoaffinity labeling and partialpurification, Biochemistry 35 (1996) 5418–5425.

[142] S.J. Shin, W.K. Lee, H.W. Lim, J. Park, Characterization of the ATPtransporter in the reconstituted rough endoplasmic reticulum proteolipo-somes, Biochim. Biophys. Acta 1468 (2000) 55–62.

[143] C. Bublitz, S. Steavenson, The pentose phosphate pathway in theendoplasmic reticulum, J. Biol. Chem. 263 (1988) 12849–12853.

[144] C. Bublitz, C.A. Lawler, S. Steavenson, The topology of phosphoglu-conate dehydrogenases in rat liver microsomes, Arch. Biochem. Biophys.259 (1987) 22–28.

[145] Y. Hino, S. Minakami, Hexose-6-phosphate and 6-phosphogluconatedehydrogenases of rat liver microsomes. Involvement in NADPH andcarbon dioxide generation in the luminal space of microsomal vesicles, J.Biochem. (Tokyo) 92 (1982) 547–557.

[146] J. Ozols, Lumenal orientation and post-translational modifications of theliver microsomal 11 beta-hydroxysteroid dehydrogenase, J. Biol. Chem.270 (1995) 2305–2312.

[147] C. Bublitz, C.A. Lawler, The levels of nicotinamide nucleotides in livermicrosomes and their possible significance to the function of hexosephosphate dehydrogenase, Biochem. J. 245 (1987) 263–267.

[148] S.H. Hori, T. Takahashi, Latency of microsomal hexose-6-phosphatedehydrogenase activity, Biochim. Biophys. Acta 496 (1977) 1–11.

[149] T. Takahashi, S.H. Hori, Alterations in the latency of hepatic microsomalhexose-6-phosphate dehydrogenase under various in vivo and in vitroconditions, Biochim. Biophys. Acta 672 (1981) 165–175.

[150] A.R. Frand, C.A. Kaiser, The ERO1 gene of yeast is required foroxidation of protein dithiols in the endoplasmic reticulum, Mol. Cell 1(1998) 161–170.

[151] M.G. Pollard, K.J. Travers, J.S. Weissman, Ero1p: a novel and ubiquitousprotein with an essential role in oxidative protein folding in theendoplasmic reticulum, Mol. Cell 1 (1998) 171–182.

[152] B. Wilkinson, H.F. Gilbert, Protein disulfide isomerase, Biochim.Biophys. Acta 1699 (2004) 35–44.

[153] B.P. Tu, J.S. Weissman, Oxidative protein folding in eukaryotes:mechanisms and consequences, J. Cell Biol. 164 (2004) 341–346.

[154] C.S. Sevier, J.W. Cuozzo, A. Vala, F. Aslund, C.A. Kaiser, A flavoproteinoxidase defines a new endoplasmic reticulum pathway for biosyntheticdisulphide bond formation, Nat. Cell Biol. 3 (2001) 874–882.

[155] J.K. Suh, L.L. Poulsen,D.M. Ziegler, J.D. Robertus,Yeast flavin-containingmonooxygenase generates oxidizing equivalents that control protein foldingin the endoplasmic reticulum, Proc. Natl. Acad. Sci. U. S. A. 96 (1999)2687–2691.

[156] B.P. Tu, J.S. Weissman, The FAD- and O(2)-dependent reaction cycle ofEro1-mediated oxidative protein folding in the endoplasmic reticulum,Mol. Cell 10 (2002) 983–994.

[157] M. Varsanyi, A. Szarka, E. Papp, D. Makai, G. Nardai, R. Fulceri, P.Csermely, J. Mandl, A. Benedetti, G. Banhegyi, FAD transport and FAD-dependent protein thiol oxidation in rat liver microsomes, J. Biol. Chem.279 (2004) 3370–3374.

[158] O. Protchenko, R. Rodriguez-Suarez, R. Androphy, H. Bussey, C.C.Philpott, A screen for genes of heme uptake identifies the FLC familyrequired for import of FAD into the endoplasmic reticulum, J. Biol.Chem. 281 (2006) 21445–21457.

[159] A. Varki, S. Diaz, The transport and utilization of acetyl coenzyme A byrat liver Golgi vesicles. O-acetylated sialic acids are a major product, J.Biol. Chem. 260 (1985) 6600–6608.

[160] A. Kanamori, J. Nakayama, M.N. Fukuda, W.B. Stallcup, K. Sasaki, M.

Fukuda, Y. Hirabayashi, Expression cloning and characterization of acDNA encoding a novel membrane protein required for the formation ofO-acetylated ganglioside: a putative acetyl-CoA transporter, Proc. Natl.Acad. Sci. U. S. A. 94 (1997) 2897–2902.

[161] R.S. Bora, A. Kanamori, Y. Hirabayashi, Cloning and characterization ofa putative mouse acetyl-CoA transporter cDNA, Gene 238 (1999)455–462.

[162] R.S. Bora, S. Ichikawa, A. Kanamori, Y. Hirabayashi, cDNA cloning ofputative rat acetyl-CoA transporter and its expression pattern in brain,Cytogenet. Cell Genet. 89 (2000) 204–208.

[163] R.S. Bora, S. Ichikawa, A. Kanamori, Y. Hirabayashi, Genomic structureand promoter analysis of putative mouse acetyl-CoA transporter gene,FEBS Lett. 473 (2000) 169–172.

[164] A. Krogh, B. Larsson, G. von Heijne, E.L. Sonnhammer, Predictingtransmembrane protein topology with a hidden Markov model: applica-tion to complete genomes, J. Mol. Biol. 305 (2001) 567–580.

[165] Y. Hirabayashi, A. Kanamori, K.H. Nomura, K. Nomura, The acetyl-CoAtransporter family SLC33, Pflugers Arch. 447 (2004) 760–762.

[166] R.R. Ramsay, R.D. Gandour, F.R. van der Leij, Molecular enzymology ofcarnitine transfer and transport, Biochim. Biophys. Acta 1546 (2001)21–43.

[167] J.M. Gooding, M. Shayeghi, E.D. Saggerson, Membrane transport offatty acylcarnitine and free L-carnitine by rat liver microsomes, Eur. J.Biochem. 271 (2004) 954–961.

[168] A. Tonazzi, M. Galluccio, F. Oppedisano, C. Indiveri, Functionalreconstitution into liposomes and characterization of the carnitinetransporter from rat liver microsomes, Biochim. Biophys. Acta 1758(2006) 124–131.

[169] M. Csala, G. Banhegyi, L. Braun, R. Szirmai, A. Burchell, B. Burchell,A. Benedetti, J. Mandl, Beta-glucuronidase latency in isolated murinehepatocytes, Biochem. Pharmacol. 59 (2000) 801–805.

[170] I.D. Waddell, K. Robertson, A. Burchell, R. Hume, B. Burchell, Evidencefor glucuronide (small molecule) sorting by human hepatic endoplasmicreticulum, Mol. Membr. Biol. 12 (1995) 283–288.

[171] S.D. Zucker, W. Goessling, A.G. Hoppin, Unconjugated bilirubinexhibits spontaneous diffusion through model lipid bilayers and nativehepatocyte membranes, J. Biol. Chem. 274 (1999) 10852–10862.

[172] E. Battaglia, J. Gollan, A unique multifunctional transporter translocatesestradiol-17beta-glucuronide in rat liver microsomal vesicles, J. Biol.Chem. 276 (2001) 23492–23498.

[173] M. Csala, A.G. Staines, G. Banhegyi, J. Mandl, M.W. Coughtrie, B.Burchell, Evidence for multiple glucuronide transporters in rat livermicrosomes, Biochem. Pharmacol. 68 (2004) 1353–1362.

[174] F.G. Hernandez-Guzman, T. Higashiyama, W. Pangborn, Y. Osawa, D.Ghosh, Structure of human estrone sulfatase suggests functional roles ofmembrane association, J. Biol. Chem. 278 (2003) 22989–22997.

[175] C. Hwang, A.J. Sinskey, H.F. Lodish, Oxidized redox state of glutathionein the endoplasmic reticulum, Science 257 (1992) 1496–1502.

[176] R. Bass, L.W. Ruddock, P. Klappa, R.B. Freedman, A major fraction ofendoplasmic reticulum-located glutathione is present as mixed disulfideswith protein, J. Biol. Chem. 279 (2004) 5257–5262.

[177] G. Banhegyi, L. Lusini, F. Puskas, R. Rossi, R. Fulceri, L. Braun, V. Mile,P. di Simplicio, J. Mandl, A. Benedetti, Preferential transport ofglutathione versus glutathione disulfide in rat liver microsomal vesicles,J. Biol. Chem. 274 (1999) 12213–12216.

[178] W. Feng, G. Liu, P.D. Allen, I.N. Pessah, Transmembrane redox sensor ofryanodine receptor complex, J. Biol. Chem. 275 (2000) 35902–35907.

[179] M. Csala, R. Fulceri, J. Mandl, A. Benedetti, G. Banhegyi, Ryanodinereceptor channel-dependent glutathione transport in the sarcoplasmicreticulum of skeletal muscle, Biochem. Biophys. Res. Commun. 287(2001) 696–700.

[180] G. Banhegyi, M. Csala, G. Nagy, V. Sorrentino, R. Fulceri, A. Benedetti,Evidence for the transport of glutathione through ryanodine receptorchannel type 1, Biochem. J. 376 (2003) 807–812.

[181] J. Koch, R. Tampe, The macromolecular peptide-loading complex inMHC class I-dependent antigen presentation, Cell. Mol. Life Sci. 63(2006) 653–662.

[182] B. Lankat-Buttgereit, R. Tampe, The transporter associated with antigen

1341M. Csala et al. / Biochimica et Biophysica Acta 1768 (2007) 1325–1341

processing: function and implications in human diseases, Physiol. Rev. 82(2002) 187–204.

[183] M. Herget, R. Tampe, Intracellular peptide transporters in human-compartmentalization of the “peptidome”, Pflugers Arch. 453 (2007)591–600.

[184] R. Abele, R. Tampe, Modulation of the antigen transport machinery TAPby friends and enemies, FEBS Lett. 580 (2006) 1156–1163.

[185] S. Le Gall, A. Neuhof, T. Rapoport, The endoplasmic reticulummembrane is permeable to small molecules, Mol. Biol. Cell 15 (2004)447–455.

[186] A.R. Osborne, T.A. Rapoport, B. van den Berg, Protein translocation bythe Sec61/SecY channel, Annu. Rev. Cell Dev. Biol. 21 (2005) 529–550.

[187] A.E. Johnson, M.A. van Waes, The translocon: a dynamic gateway at theER membrane, Annu. Rev. Cell Dev. Biol. 15 (1999) 799–842.

[188] T.A. Rapoport, V. Goder, S.U. Heinrich, K.E. Matlack, Membrane–protein integration and the role of the translocation channel, Trends CellBiol. 14 (2004) 568–575.

[189] M.D. Potter, C.V. Nicchitta, Endoplasmic reticulum-bound ribosomesreside in stable association with the translocon following termination ofprotein synthesis, J. Biol. Chem. 277 (2002) 23314–23320.

[190] S.M. Simon, G. Blobel, A protein-conducting channel in the endoplasmicreticulum, Cell 65 (1991) 371–380.

[191] K.S. Crowley, S. Liao, V.E. Worrell, G.D. Reinhart, A.E. Johnson,

Secretory proteins move through the endoplasmic reticulum membranevia an aqueous, gated pore, Cell 78 (1994) 461–471.

[192] B.D. Hamman, J.C. Chen, E.E. Johnson, A.E. Johnson, The aqueous porethrough the translocon has a diameter of 40–60 A during cotranslationalprotein translocation at the ER membrane, Cell 89 (1997) 535–544.

[193] D. Heritage, W.F. Wonderlin, Translocon pores in the endoplasmicreticulum are permeable to a neutral, polar molecule, J. Biol. Chem. 276(2001) 22655–22662.

[194] A. Roy, W.F. Wonderlin, The permeability of the endoplasmic reticulumis dynamically coupled to protein synthesis, J. Biol. Chem. 278 (2003)4397–4403.

[195] R.B. Lomax, C. Camello, F. Van Coppenolle, O.H. Petersen, A.V.Tepikin, Basal and physiological Ca(2+) leak from the endoplasmicreticulum of pancreatic acinar cells. Second messenger-activatedchannels and translocons, J. Biol. Chem. 277 (2002) 26479–26485.

[196] F. Van Coppenolle, F. Vanden Abeele, C. Slomianny, M. Flourakis, J.Hesketh, E. Dewailly, N. Prevarskaya, Ribosome–translocon complexmediates calcium leakage from endoplasmic reticulum stores, J. Cell Sci.117 (2004) 4135–4142.

[197] M. Flourakis, F. Van Coppenolle, V. Lehen'kyi, B. Beck, R. Skryma, N.Prevarskaya, Passive calcium leak via translocon is a first step for iPLA2-pathway regulated store operated channels activation, FASEB J. 20(2006) 1215–1217.