JBC

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M409051200 on September 7, 2004

J. Biol. Chem., Vol. 279, Issue 49, 51482-51489, December 3, 2004
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
279/49/51482    most recent
M409051200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhang, M.
Right arrow Articles by Napoli, J. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhang, M.
Right arrow Articles by Napoli, J. L.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Elements in the N-terminal Signaling Sequence That Determine Cytosolic Topology of Short-chain Dehydrogenases/Reductases

STUDIES WITH RETINOL DEHYDROGENASE TYPE 1 AND CIS-RETINOL/ANDROGEN DEHYDROGENASE TYPE 1*

Min Zhang, Peirong Hu, and Joseph L. Napoli{ddagger}

From the Department of Nutritional Sciences and Toxicology, University of California, Berkeley, California 94720-3104

Received for publication, August 6, 2004 , and in revised form, September 1, 2004.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
High affinity, retinoid-specific binding proteins chaperone retinoids to manage their transport and metabolism. Proposing mechanisms of retinoid transfer between these binding proteins and membrane-associated retinoid-metabolizing enzymes requires insight into enzyme topology. We therefore determined the topology of mouse retinol dehydrogenase type 1 (Rdh1) and cis-retinoid androgen dehydrogenase type 1 (Crad1) in the endoplasmic reticulum of intact mammalian cells. The properties of Rdh1 were compared with a chimera with a luminal signaling sequence (11{beta}-hydroxysteroid dehydrogenase (11{beta}-HSD1)(1-41)/Rdh1(23-317); the green fluorescent protein (GFP) fusion proteins Rdh1(1-22)/GFP, Crad1(1-22)/GFP, and 11{beta}-HSD1(1-41)/GFP; and signaling sequence charge difference mutants using confocal immunofluorescence, antibody access, proteinase K sensitivity, and deglycosylation assays. An N-terminal signaling sequence of 22 residues, consisting of a hydrophobic helix ending in a net positive charge, anchors Rdh1 and Crad1 in the endoplasmic reticulum facing the cytoplasm. Mutating arginine to glutamine in the signaling sequence did not affect topology. Inserting one or two arginine residues near the N terminus of the signaling sequence caused 28-95% inversion from cytoplasmic to luminal, depending on the net positive charge remaining at the C terminus of the signaling sequence; e.g. the mutant L3R,L5R,R16Q,R19Q,R21Q faced the lumen. Experiments with N- and C-terminal epitope-tagged Rdh1 and molecular modeling indicated that a hydrophobic helix-turn-helix near the C terminus of Rdh1 (residues 289-311) projects into the cytoplasm. These data provide insight into the features necessary to orient type III (reverse signal-anchor) proteins and demonstrate that Rdh1, Crad1, and other short-chain dehydrogenases/reductases, which share similar N-terminal signaling sequences such as human Rdh5 and mouse Rdh4, orient with their catalytic domains facing the cytoplasm.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Retinol requires conversion into all-trans-retinoic acid to fulfill the vitamin A function of regulating gene expression that begins shortly after conception and continues throughout vertebrate life (1, 2). Retinoid-active SDRs1 recognize the major physiological form of retinol, retinol bound with CRBP, to generate retinal for irreversible conversion into all-trans-retinoic acid (3). The most active of the retinol dehydrogenases include mouse Rdh1 and its orthologs, rat Rodh1 and -2 and human RDH-E/RoDH4 (4-8). rRodh2 e.g. has lower apparent Km values with holo-CRBP than with unbound retinol and catalyzes retinal synthesis in the presence of excess apoCRBP. Dehydrogenation of retinol in the presence of excess CRBP with concentrations of both ligand and binding protein in the micromolar range indicates that the reaction proceeds through the SDR accessing CRBP-bound retinol because the CRBP-retinol complex has a Kd value ~0.1 nM, which would virtually eliminate unbound retinol. Close approach of the two has been confirmed by cross-linking of holo-CRBP with microsomal Rdh (9). Although Rdhs that recognize holo-CRBP also occur in cytoplasm, they are inhibited severely by a small excess of apoCRBP (10). Even uninhibited cytosolic enzymes make a quantitatively minor contribution to the total retinal-generating units. In contrast to SDR, the cytosolic alcohol dehydrogenase isozymes do not recognize CRBP-retinol in vitro (10, 11). Moreover alcohol dehydrogenase access to physiological amounts of unbound retinol in intact cells may be limited because retinol has very low solubility in aqueous media when not sequestered tightly with CRBP. Notably gene knock-out experiments did not provide evidence for an alcohol dehydrogenase contribution to retinol metabolism under normal circumstances, because Adh1-/-, Adh3-/-, Adh4-/-, and dual Adh1/Adh4-/- null mice showed no retinoic acid deficiency phenotype nor any disturbance of retinoid metabolism or compensatory response (12, 13).

The importance of high affinity, specific binding proteins chaperoning retinoids in vivo has been illustrated through gene ablation experiments (14). CRBP null mice e.g. seem morphologically normal but eliminate retinyl esters 6-fold faster than wild-type mice via an unknown mechanism (15). Mice null in the intestinal cellular retinol-binding protein type II die from vitamin A deficiency within 24 h after birth when delivered by dams fed a diet marginal in vitamin A (16). Mice null in the retinal pigment epithelium retinyl ester-binding protein RPE65 cannot produce 11-cis-retinol for production of rhodopsin, and mutations in RPE65 cause blindness (Leber's congenital amaurosis) (17, 18). Mutations in another retinal pigment epithelium-binding protein, cellular retinal-binding protein, cause night blindness and photoreceptor degeneration from inefficient metabolic processing of 11-cis-retinol (19). Mutations in hRdh5, which serves as one of the retinal pigment epithelium 11-cis-retinol dehydrogenases, are linked with the rare, autosomal recessive disease fundus albipunctatus, i.e. night blindness from delayed photopigment regeneration (20, 21). Serum retinol-binding protein null mice are unable to mobilize hepatic retinyl esters and suffer visual impairment (22). Proposing precise models of retinoid transfer between these retinoid-binding proteins and retinoid-metabolizing enzymes requires understanding the topology of the enzymes.

The retinoid-metabolizing, membrane-associated SDRs, including rRodh1, Rdh1, hRdh-E/Rodh4, and bovine 11-cis-retinol dehydrogenase/hRdh5/mRdh4, consist of N-terminal hydrophobic helices of ~18 residues flanked on the C terminus by net positive charges, midprotein hydrophobic helix-loop-helix transmembrane regions, and hydrophobic helices of ~23 residues close to their C termini. Several models have been proposed for their association with the ER. An early model suggested that all four of the hRoDH4 hydrophobic helices transected the ER membrane (7). A second model presented 11-cis-retinol dehydrogenase with its N- and C-terminal hydrophobic helices projecting through the ER membrane with the bulk of the protein facing the lumen (23). The SDR Crad1 and the 11-cis-retinol dehydrogenase orthologs Rdh4/5 also have been proposed as luminal facing (24). Another model presented rRodh1 as a type III (also known as reverse signal-anchor) protein with the N-terminal signaling sequence buried in the ER membrane but as a globular SDR projecting into the cytoplasm (25). In this model, the C-terminal helix associated with the surface of the ER. All these models cannot be accurate in view of the amino acid similarity among these closely related SDRs especially in their putative N-terminal signaling sequences.

The goals of this study were to determine whether the Rdh1 localizes exclusively to the ER membrane; the topology of Rdh1 by comparing its behavior in intact cells with luminal facing proteins for proteinase sensitivity, access to antibodies, and glycosylation; and the molecular features that establish ER retinoid dehydrogenase membrane expression locus and topology.


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
cDNA Constructs—Constructs were made by PCR (primers are detailed in the supplemental material) and were sequenced. pcDNA3/Rdh1 was used as the template for the Rdh1 mutants (4). Mutants and their primers (forward/reverse, respectively) included RD(1-18), F1/R1; RD(1-30), F2/R1; RD(289-317), F3/R2; and RD2, F1/R2. Point mutations were introduced into Rdh1 by overlapping PCR. Mutant glycosylation site at residue 213 was made using the primer pairs F3/R3 and F4/R1 to amplify the 5'- and 3'-sections of Rdh1. The gel-purified products were used as templates with primers F3/R1 to make the mutant V213N. Similarly mutant G251N was obtained with primer pairs F5/R4 and F3/R1 and was used to make the mutant L253T. Additional mutants and primer pairs included R19Q, F6/R5 and F3/R1; R21Q, F7/R6 and F3/R1; and R(19,21)2Q, F8/R7 and F3/R1. Mutant R(16,19,21)3Q was obtained with primer pairs F9/R8 and F3/R1 using the double mutant R(19,21)2Q as template. Mutants L(3,5)2R, L(3,5)2R/R(19,21)2Q, and L(3,5)2R/R(16,19,21)3Q were generated with primers F10/R1. Mutants with an arginine residue inserted after the second Rdh1 residue R3, R3/19Q, R3/21Q, R3/(19,21)2Q, and R3/(16,19,21)3Q, were made with the primer pair F11/R1.

To obtain the chimeric 11{beta}-HSD1(1-41)/Rdh1(23-317), a PCR product coding for the first 41 amino acid residues of human 11{beta}-HSD1 was amplified with primers F12/R9 and digested with EcoRI/StuI. The DNA fragment encoding amino acids 23-317 of Rdh1 was obtained with primers F13/R1 and digested with PmlI/XhoI. A sequence coding for the FLAG epitope (DYKDDDDK) was attached to the N or C termini of Rdh1 to generate FLAG-Rdh and Rdh-FLAG with primer pairs F14/R1 and F3/R10, respectively. Rdh1 mutants were subcloned into the EcoRI/XhoI site of pcDNA3 (Invitrogen).

GFP constructs were cloned into the EcoRI/NotI sites of pEGFP-N1 (BD Biosciences). To obtain Rdh(1-22)/GFP, a DNA fragment encoding the first 22 residues of Rdh1 was amplified with primers F3/R14 and digested with EcoRI/PvuII. A second DNA fragment encoding the coding region of GFP was amplified with primers F16/R11 and digested with PmlI/NotI. The two fragments were cloned. Rdh(1-18)/GFP was made with primers F3/R12, digested with EcoRI/DraI, and cloned with the second DNA fragment described above. Overlapping PCR was used to generate mutant GFP/Rdh(289-317) with primers F18/R13 and F19/R14 and cloned. Plasmid Crad1(1-22)/GFP was constructed with primers F3/R15 and F20/R11, and the chimeric DNA fragment was cloned.

Subcellular Fractionation and Western Blotting—COS cells (ATCC) were grown at 37 °C with 5% CO2 in Dulbecco's modified essential medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 µg/ml streptomycin. Cells on 100-mm plates were transfected with 8 µg of plasmid DNA using 40 µl of SuperFect (Qiagen). Twenty-four hours post-transfection, 5 x 107 cells were resuspended in 2 ml of ice-cold HB (150 µM MgCl2,10mM KCl, 10 mM Tris-HCl, pH 6.7) plus protease inhibitors and left on ice for 5 min. The cells were disrupted by nitrogen cavitation at 4 °C (80 p.s.i./30 min). The suspension was then homogenized by five gentle strokes of a Dounce grinder with a loose pestle. An additional 1/3 volume ice-cold HB with 1 M sucrose was added to the cell suspension and mixed well. The nuclei, cell debris, and unbroken cells were removed by centrifugation at 700 x g for 10 min. The postnuclear supernatant was centrifuged at 5000 x g for 10 min, and the pellet was resuspended with 5 ml of buffer B (250 mM sucrose, 1 mM EDTA, 10 mM Tris-HCl, pH 7.4) plus protease inhibitors. This homogenate was centrifuged at 5000 x g for 10 min to obtain a crude mitochondrial pellet. The crude mitochondria pellet was resuspended by five gentle strokes with a Dounce grinder, layered on a discontinuous sucrose gradient (30, 40, and 55%), and centrifuged at 150,000 x g with an SW41 rotor for 20 min at 4 °C. The mitochondria fraction was collected from the 40/55% interface, diluted with 5 volumes of buffer B, and centrifuged at 10,000 x g for 20 min. The resulting pellet was solubilized in buffer B. The microsomal fraction was obtained by centrifugation of the postmitochondrial supernatant at 100,000 x g for 1 h and was rehomogenized in buffer B. For Western blotting, 10 µg of protein were analyzed by 12.5% SDS-PAGE. Results were imaged with an ECL-Plus Western blotting detection system (Amersham Biosciences).

Deglycosylation—Deglycosylation was done with PNGase F (New England Biolabs). A mixture of denatured proteins (1x Glycoprotein Denature Buffer, 100 °C, 10 min) with 0.1 volume each of 10x G7 buffer and 10% Nonidet P-40 and 2 µl of PNGase F (500,000 units/ml) was incubated at 37 °C for 1 h. Mixtures were resolved by 12.5% SDS-PAGE and analyzed by Western blot.

Protease Protection—cDNA constructs were transfected into COS cells cultured in 6-well plates with SuperFect reagent (Qiagen). Twenty-four hours post-transfection, cells were washed three times in PBS and trypsinized with 300 µl of trypsin-EDTA (0.25%) for 10 min at 37 °C. Trypsinized cells were resuspended in PBS, washed twice with PBS, and permeabilized (5 min, 37 °C) by addition of 200 µl of SLO in PBS (1000 units/ml). Alternatively cells were treated with 200 µl of PBS or 200 µl of 0.5% Triton X-100 in PBS under the same conditions. Cells were then treated with proteinase K (50 µg/ml) at 37 °C for 15 min. Digestion was stopped by addition of 1 mM phenylmethylsulfonyl fluoride (Roche Applied Science) and 100 µl of 60 mM HEPES, pH 7.2, 1.5 M sucrose, 6 mM EDTA containing protease inhibitor mixture (BD Biosciences). Cells were lysed by sonication. Immunoblotting was done as described above using anti-Rdh1 or anti-GFP monoclonal antibody.

Antibodies and Fluorescence Markers—Subconfluent COS cells grown on coverslips were transfected with SuperFect reagent (Qiagen). Twenty-four hours post-transfection, cells were incubated in serum-free medium for 1 h. Cells were washed three times in SLO buffer (75 mM potassium acetate, 25 mM HEPES buffer, pH 7.2) and fixed with paraformaldehyde (4% in PBS) at room temperature for 15 min. After washing, cells were incubated for 15 min at room temperature with 500 units/ml activated SLO (Sigma). SLO was removed, and the cells were washed once and incubated in SLO buffer for 15 min at 37 °C to allow pore formation. Alternatively fixed cells were incubated in PBS or 0.2% Triton X-100 (containing 1% bovine serum albumin in PBS) for 15 min at room temperature. Some fixed cells were permeabilized with digitonin (5 µg/ml) for 15 min at 4 °C in 20 mM HEPES, pH 6.9, 0.3 M sucrose, 0.1 M KCl, 2.5 mM MgCl2, 1 mM EDTA.

Permeabilized cells were incubated at room temperature for 2 h with anti-Rdh1 (1:200, raised in rabbits against the peptide DRLSSNTKMIWDKASSEVK), anti-GFP (1:200, monoclonal antibody JL-8, BD Biosciences), or anti-calnexin (1:200, Sigma) followed by Alexa Fluor 488-conjugated goat anti-rabbit (5 µg/ml) or Alexa Fluor 594-conjugated goat anti-mouse immunoglobulins (5 µg/ml) for another 1 h. Fluorescein isothiocyanate-conjugated anti-FLAG M2 monoclonal antibody (10 µg/ml) was used for FLAG epitope detection (Sigma). Alexa Fluor 594-conjugated ConA or MitoTracker Red CMXRos (both from Molecular Probes) were used as ER or mitochondria markers, respectively. DAPI was used for nuclear counterstaining. Cells were mounted in ProLong Antifade reagents (Molecular Probes), viewed under a Zeiss 510 laser scanning confocal microscope, and processed with Zeiss Bitplane Imaris 3.3 software.

Molecular Modeling—The molecular model was generated with Swiss-Model Version 36.0002, based on three-dimensional structures of soluble SDR, using an optimal amino acid sequence alignment as described previously (26). The model relied on x-ray structures of human 17{beta}-hydroxysteroid dehydrogenase type I complexed with cofactor and/or substrate (39.8% identity; Protein Data Bank codes 1FDS [PDB] , 1A27 [PDB] , and 1EQU [PDB] ) (27-29), 20{beta}-hydroxysteroid dehydrogenase (51.5%; Protein Data Bank code 1HU4 [PDB] ) (30), and {beta}-ketoacyl-[acyl-carrier protein]-reductase (41% identity; Protein Data Bank codes 1I0E [PDB] and 1I0B [PDB] ) (31).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Only the ER Expresses Rdh1—The N terminus of Rdh1 has characteristics of a dual mitochondria/microsomal signaling sequence. The first 18 amino acid residues form a hydrophobic helix sufficiently long to span a lipid bilayer membrane and are flanked by the positively charged residues Arg19, Arg21, and Lys30 (Fig. 1). This resembles N-terminal signaling sequences in enzymes expressed dually by ER and mitochondria, such as CYP1A1, CYP2E1, and P450MT2 (32-34). We expressed Rdh1 in COS cells and determined its locus in subcellular fractions by immunoblotting and in intact cells by immunofluorescence. Immunoblotting (Fig. 1) revealed Rdh1 in the microsomal and mitochondria fractions, distributing with calnexin, an ER membrane protein (35). Cytochrome c, a mitochondria intermembrane space protein, appeared only in the mitochondria fraction. These results are consistent with reports of ER protein contamination of mitochondria even after their isolation by sucrose density gradient centrifugation (36, 37). The density of the Rdh1 signal relative to the calnexin signal remained the same in both the microsomal and mitochondrial fractions, indicating that only the ER expresses Rdh1. Immunofluorescence of intact cells confirmed these observations (Fig. 2, a and b). Rdh1 colocalized with ConA, an ER marker, but did not colocalize with MitoTracker.



View larger version (50K):
[in this window]
[in a new window]
  		FIG. 1.
Functions of Rdh1 hydrophobic regions in membrane targeting. Top, the N-terminal 30 amino acid residues of Rdh1 resemble dual mitochondria/ER membrane signaling sequences. Bottom left, mutants of the 317-amino acid residue SDR Rdh1 were created by deleting hydrophobic helices at the N-terminal (residues 1-18 or 1-30) and/or C terminus (residues 289-317). Bottom right, distributions of Rdh1 deletion mutants compared with the mitochondrial interspace protein cytochrome c and the ER membrane protein calnexin by Western blot in the 700 x g supernatants of COS cell homogenates (H), mitochondria fraction (Mt), microsomes (M), and cytosol (C).

 



View larger version (57K):
[in this window]
[in a new window]
  		FIG. 2.
Expression of Rdh1 and deletion mutants in intact cells. Rdh1 and deletion mutants were immunostained in fixed COS cells permeabilized with 0.2% Triton X-100. The left column shows immunostaining using anti-Rdh1 as the primary antibody and Alexa Fluor 488-conjugated goat anti-rabbit IgG as the secondary antibody (green). The middle column alternates showing Alexa Fluor 594-conjugated ConA (a, c, f, and h) or MitoTracker (Mito) Red CMXRos (b, d, e, g, and i) (both red) as microsomal and mitochondria markers, respectively, with the exception of RD(1-30), which shows only MitoTracker. The right column shows a merged image of the left and middle columns. Nuclei were stained with DAPI (blue): a and b, Rdh1; c and d, RD(1-18); e, RD(1-30); f and g, RD(289-317); h and i, RD2.

 
The N-terminal Signaling Sequence Targets Rdh1 to the ER—Rdh1 has hydrophobic helices at its N terminus (residues 1-18) and near its C terminus (residues 289-311). We made deletion mutants to determine whether one or both contribute to ER targeting (Fig. 1). Mutants RD(1-18) or RD(1-30), which lack the first 18 and 30 amino acid residues of Rdh1, respectively, did not localize to the ER. Surprisingly both distributed in the mitochondria fraction with cytochrome c and localized in mitochondria in intact cells (Figs. 1 and 2, c, d, and e). These data show that residues 1-18 and residues 1-30 are necessary for ER localization but do not demonstrate whether they are sufficient. The deletion mutant RD(289-317) produced immunoblot signals in both the microsomal and mitochondria fractions but predominantly in the mitochondria. Immunofluorescence confirmed this: only ~20% of transfected cells showed ER localization, whereas the rest exhibited the punctate signal of mitochondria expression (Figs. 1 and 2, f and g). Deletion of both the 18 N-terminal and the 28 C-terminal residues, mutant RD2, generated a protein that localized overwhelmingly in the mitochondria (Figs. 1 and 2, h and i). Thus, the N-terminal signaling sequence directs Rdh1 to the ER, the C-region hydrophobic helix may promote ER retention, and removing both does not generate a soluble protein.

To further define the function of the N- and C-terminal hydrophobic helices, we determined the loci of Rdh1/GFP fusion proteins by immunofluorescence. Native GFP had a diffuse pattern, distributing throughout the cell (not shown). Fusing the first 22 residues of Rdh1 to GFP (Rdh1(1-22)/GFP) produced a protein with an ER expression pattern, demonstrating that these residues are sufficient to cause ER targeting (Fig. 3). In contrast, the fusion protein Rdh1(1-18)/GFP, with only the first 18 N-terminal residues fused to GFP, showed a diffuse signal, indicating that the first 18 residues alone are not sufficient for ER targeting. The fusion protein of GFP and the last 29 Rdh1 amino acid residues, GFP/Rdh1(289-317), also showed a diffuse cytoplasmic and nuclear signal with an intense band in the nuclear membrane. These data show that the 22 N-terminal residues, but not the last 29 C-terminal residues, of Rdh1 are sufficient to direct and anchor the soluble protein GFP to the ER.



View larger version (76K):
[in this window]
[in a new window]
  		FIG. 3.
The N terminus of Rdh1 directs GFP to the microsomes. COS cells were transfected with Rdh1/GFP proteins as indicated on the left and permeabilized with 0.2% Triton X-100. GFP (green) and ConA (red) fluorescence was compared and merged. Nuclei were DAPI-stained (blue).

 
Rdh1 Faces the Cytoplasm—We assessed access of antibodies to Rdh1 expressed in COS cells treated with SLO, digitonin, or Triton X-100. SLO selectively permeabilizes the plasma membrane by binding to cholesterol and polymerizing to create pores ≥12 nm, permitting uptake of large molecules (38). Digitonin also creates pores in plasma membranes by interacting with cholesterol. To place the Rdh1 results into context with a luminal facing SDR, we made a chimera of the first 41 N-terminal residues of h11{beta}-HSD1 fused to the N terminus of GFP. These residues contain a signaling sequence, bounded near their N terminus by two lysine residues and bounded on their C terminus by two glutamate residues, that direct h11{beta}-HSD1 into the ER facing the lumen (39-41). PBS-treated COS cells transfected with either Rdh1 or h11{beta}-HSD1(1-41)/GFP exhibited faint, nonspecific signals when exposed to the appropriate antibodies (Fig. 4A, a and b). SLO- and digitonin-treated Rdh1-transfected cells showed strong fluorescent signals, indicating that permeabilization of the plasma membrane allowed access of anti-Rdh1 to Rdh1, i.e. the Rdh1 epitope faces the cytoplasm. In contrast, fluorescence of the luminal facing h11{beta}-HSD1(1-41)/GFP was observed only after Triton X-100 treatment, indicating that the antibody could not access the epitope until after solubilization of the ER membrane, i.e. the GFP epitope faced the lumen. Results similar to h11{beta}-HSD1(1-41)/GFP were obtained under the same conditions with antibodies against calnexin, a luminal facing ER protein (data not shown). We also contrasted the effects on GFP localization of the h11{beta}-HSD1 and Rdh1 signaling sequences. The chimera Rdh1(1-22)/GFP readily produced a GFP signal in transfected COS cells that was not altered by SLO or Triton X-100 treatment (Fig. 4A, c). SLO permeabilization of Rdh1(1-22)/GFP-transfected COS cells allowed access of the anti-GFP antibody to GFP as indicated by the red signal of the Alexa Fluor 594-conjugated goat anti-mouse IgG, and Triton X-100 treatment did not expand the red immunofluorescence pattern (Fig. 4A, d). A merge of the GFP and anti-GFP signals produced by Rdh1(1-22)/GFP indicates expression in the same cells (Fig. 4A, e). These results show a clear contrast between antibody access to the two GFP chimera, which differ only in the nature of their leader sequences.



View larger version (53K):
[in this window]
[in a new window]
  		FIG. 4.
Rdh1 anchors in the ER facing the cytoplasm. A, immunofluorescence in COS cells transfected with Rdh1 (a), 11{beta}-HSD1(1-41)/GFP (b), Rdh1(1-22)/GFP (c, d, and e). Cells were treated with PBS or permeabilized with SLO, digitonin (DIG), or Triton X-100 (TX). In a and b, the green and red signals, respectively, result from the fluors bound to the secondary antibodies used with anti-Rdh1 and anti-GFP. c shows GFP fluorescence (green). d shows the accessibility of anti-GFP antibody (red) to the GFP expressed in the chimera Rdh1(1-22)/GFP. e shows merged images of c and d. B, Western blots of Rdh1, 11{beta}-HSD1(1-41)/GFP, or Rdh1(1-22)/GFP obtained from COS cells after treatment in situ with proteinase K (PK). Cells were treated with PBS or were permeabilized with SLO or Triton X-100 (TX). GFP itself was run in the last gel. The arrowhead shows the GFP band. The arrow shows a nonspecific band.

 
To corroborate these data, we determined Rdh1 and GFP chimera sensitivity to proteinase K in intact cells. SLO permeabilization of COS cell plasma membranes allowed proteinase K access to Rdh1 but essentially did not allow access to h11{beta}-HSD1(1-41)/GFP (Fig. 4B). Complete digestion of h11{beta}-HSD1(1-41)/GFP occurred only after dissolving membranes with Triton X-100. SLO treatment of cells transfected with Rdh1(1-22)/GFP allowed cleavage by proteinase K of GFP from the Rdh1 signal sequence, indicating cytoplasm orientation. These data indicate that the N-terminal 22 residues are sufficient to target Rdh1 to the ER with cytoplasmic orientation. Notably we conducted proteinase K incubations at 37 °C. Many proteinase K digestions are done at 4 or 10 °C because of the efficiency of the protease. Some proteins, however, resist proteinase K, including GFP and various SDRs (25, 42, 43). The smaller bands in the h11{beta}-HSD1(1-41)/GFP and Rdh1(1-22)/GFP digestions show released but resistant GFP. Some h11{beta}-HSD1(1-41)/GFP degradation in SLO and PBS-treated cells suggests some membrane leaking.

The Locus of the C Terminus—The C-terminal region hydrophobic helix of Rdh1 might span the ER membrane with residues 306-317 entering the lumen or might not insert into the membrane with residues 306-317 exposed in the cytoplasm. To distinguish these possibilities, we performed immunofluorescence analysis with N- and C-terminal FLAG-tagged Rdh1 constructs expressed in COS cells (Fig. 5). Anti-FLAG antibody did not bind with the epitope in FLAG-Rdh unless the cells had been treated with Triton X-100, indicating that the N-terminal FLAG epitope was inaccessible to antibody, i.e. in the lumen. In contrast, anti-FLAG antibody recognized its epitope in Rdh-FLAG in SLO-permeabilized cells, indicating that the C-terminal FLAG epitope oriented toward the cytoplasm.



View larger version (70K):
[in this window]
[in a new window]
  		FIG. 5.
The Rdh1 C terminus faces the cytoplasm. COS cells were transfected with either N-terminal FLAG-tagged Rdh1 (FLAG-Rdh) or C-terminal FLAG-tagged Rdh1 (Rdh-FLAG). Cells were treated with PBS or were permeabilized with SLO or Triton X-100 (TX) and immunostained with fluorescein isothiocyanate-conjugated anti-FLAG antibody (green). Nuclei were stained with DAPI (blue).

 
Native Rdh1 Is Not Glycosylated but Undergoes Glycosylation If Forced to Face the ER Lumen—Because luminal proteins undergo glycosylation and cytoplasmic proteins do not, we used glycosylation as a measure of topology. We tested the sensitivity of Rdh1 to the glycosidase PNGase F because Rdh1 has two potential asparagine-containing glycosylation sites (71NKT and 161NVS). We found no electrophoretic migration differences between Rdh1 and Rdh1 treated with PNGase F, indicating that Rdh1 does not face the ER lumen (Fig. 6A, a). We also made two mutants of Rdh1, each with a single additional glycosylation site, V213N to generate 213NTS and L253T to generate 251NET, to determine whether sections of Rdh1 nearer the C terminus were exposed to the ER lumen. These mutants did not change migration upon PNGase F treatment (data not shown). Next we used the fusion protein h11{beta}-HSD1(1-41)/Rdh1(23-317) to determine whether an Rdh1 analog forced to face the ER lumen would undergo glycosylation. The chimera h11{beta}-HSD1(1-41)/Rdh1(23-317) mostly (~80%) faced the lumen as indicated by two additional and more slowly migrating electrophoretic bands from glycosylation of the two potential Rdh1 glycosylation sites and by PNGase F sensitivity of these bands (Fig. 6A, b). Thus, the h11{beta}-HSD1 signaling sequence reversed Rdh1 topology, and native Rdh1 undergoes glycosylation when facing the lumen.



View larger version (45K):
[in this window]
[in a new window]
  		FIG. 6.
Positive charges in the N terminus determine the topology of Rdh1. A, Western blots of Rdh1 (a), the chimera 11{beta}HSD1(1-41)/Rdh1 (23-317) (b), and mutant constructs (c-n) harvested from COS cells and treated with the glycosidase PNGase F. R3 denotes a mutant with an Arg residue inserted between residues Trp2 and Leu3. (See Fig. 1 for the N-terminal sequence of Rdh1.) The relative amounts of glycosylated forms were quantified by signal density scanning with an imaging densitometer (Bio-Rad). Data are the means ± S.D. of three transfections. B, COS cells expressing the mutant e (R(19,21)2Q) or k (R3/R(16,19,21)3Q) were treated with proteinase K (PK) with or without permeabilization of the plasma membrane with SLO, or total membrane digestion with Triton X-100 (TX). Immunoblotting of the 700 x g supernatant was done with anti-Rdh1. In both A and B, native (non-glycosylated) protein bands are indicated with arrowheads. Arrows show glycosylated bands.

 
Contributions of Charged Residues in the N-terminal Signaling Sequence—The "positive inside" rule states that a net positive charge at either end of a signaling sequence establishes topology by associating with the cytoplasm (44-46). A net positive charge at the N terminus of the signaling sequence will direct the bulk of the protein into the lumen; a net positive charge at the C terminus of the signaling sequence will direct the bulk of the protein into the cytoplasm. Therefore, we made a series of mutants that altered the number and positions of positive charges in the signaling sequence. ER localization of all mutants was confirmed by immunofluorescence with confocal microscopy (data not shown).

First we replaced the arginine residues in the signaling sequence of Rdh1 (Arg16, Arg19, and Arg21) with glutamine, a neutral, hydrophilic residue of similar size. We noted no glycosylation in the R -> Q mutants (Fig. 6A, c-f) except for the triple mutant R(16,19,21)3Q, which showed only slight glycosylation, ~5% of the total protein expressed. These results show that positive charges at the C-terminal end of a signaling sequence are not required to orient an ER membrane protein to the cytoplasm.

We next inserted an arginine residue near the N terminus of the Rdh1 signaling sequence between Trp2 and Leu3 (mutant R3) and in the R -> Q mutants (Fig. 6A, g-k). This single positive charge inverted Rdh1 to orientate luminally in varying degrees. The degree of inversion increased with decreases in the number of arginine residues at the end of the signaling sequence. These data indicate that positive charge(s) at the N-terminal end of an ER signaling sequence is required for luminal orientation and that positive charges at the C-terminal end of a signaling sequence can affect their efficiency in setting topology.

The next set of mutants replaced the two leucine residues near the N terminus of the signaling sequence, Leu3 and Leu5, with two arginine residues, mutant L(3,5)2R (Fig. 7A, l). Adding two positively charged residues at the N terminus of Rdh1 created a protein ~60% oriented to the ER lumen. Adding the two N-terminal positively charged residues while removing the positive charges at the C terminus of the signaling sequence created proteins that were substantially luminally oriented (Fig. 6A, m and n). These results confirm and extend the results with the single positive charges insertions described above.



View larger version (71K):
[in this window]
[in a new window]
  		FIG. 7.
The N-terminal hydrophobic domain anchors Crad1 in the ER facing the cytoplasm. A, the fusion protein Crad1(1-22)/GFP was expressed in COS cells and tested for localization in the ER with ConA. Nuclei were stained with DAPI (blue). B, the fusion protein was tested for sensitivity to proteinase K digestion (PK) in situ in the presence or absence of plasma membrane permeabilization with SLO or total cell membrane permeabilization with Triton X-100 (TX). The arrowhead shows GFP release from the fusion protein by protease digestion. The arrow shows a nonspecific signal.

 
A proteinase K access experiment reinforced these insights. SLO permeabilization allowed access of proteinase K to mutant R(19,21)2Q and to the non-glycosylated bands of mutant R3/R(16,19,21)3Q consistent with cytoplasmic orientation (Fig. 6B, mutants e and k, respectively). In contrast, the two glycosylated bands of R3/R(16,19,21)3Q were digested by proteinase K only after Triton X-100 treatment of the cells. Thus, these data show that the same SDR analog of Rdh1, i.e. R3/(16,19,21)3Q, under identical conditions (e.g. 37 °C) can tolerate exposure to proteinase K or succumb to proteolysis depending on luminal (glycosylated) or cytoplasmic (non-glycosylated) orientation, respectively.

Crad1 Faces the Cytoplasm—The first 28 residues of Crad1 differ from Rdh1 only in E20V. To determine whether this single residue would affect topology, we analyzed the chimera Crad1(1-22)/GFP. GFP fluorescence of Crad1(1-22)/GFP colocalized well with the ER marker just like Rdh1(1-22)/GFP (compare Figs. 7A and 3). SLO permeabilization rendered Crad1(1-22)/GFP susceptible to proteinase K digestion, identical to the outcome for Rdh1(1-22)/GFP (compare Figs. 7B and 4B). Cytoplasmic orientation was confirmed by deglycosylation analysis of native Crad1 from transfected COS cells: no N-linked glycosylation had occurred (data not shown). Thus, the E20V difference has no apparent impact on topology.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
This report has shown by four independent experimental approaches, immunofluorescence, antibody access, proteinase K sensitivity, and degree of glycosylation, applied to wild-type, chimera, and mutant proteins translated and processed within intact cells that Rdh1 anchors in the ER membrane facing the cytoplasm. Contrasting behavior contributed to this conclusion, i.e. between wild-type Rdh1 and a chimera consisting of an authentic luminal signaling sequence in place of the Rdh1 signaling sequence, viz. 11{beta}-HSD1(1-41)/Rdh1(23-317); between GFP fusion proteins with two different signaling sequences, Rdh1(1-22)/GFP and 11{beta}-HSD1(1-41)/GFP; and between Rdh1 and N-terminal residue mutants of Rdh1. The data establish that the first 22 residues of Rdh1 constitute a signaling sequence necessary and sufficient to anchor GFP in the ER membrane facing the cytoplasm. These 22 residues are essential for the ER localization and topology of Rdh1, but Rdh1 itself also requires C-terminal residues for optimal residence (stabilization?) in the ER membrane even though these C-terminal residues do not direct microsomal localization (Fig. 3, GFP/Rdh(289-317)).

X-ray crystallography has established that soluble SDRs occur as globular proteins (27-31). Alignment of Rdh1, 11{beta}-HSD1, and other ER-expressed SDRs (rRodh1 and Crad1) with several soluble SDRs (e.g. mouse lung carbonyl reductase) shows that the former have N-terminal extensions for membrane insertion and orientation beyond the core SDR structure (25, 26). The characteristic motifs of SDRs begin at Rdh1 residue 30, the start of the cofactor binding domain, and occupy the same relative positions in soluble and membrane-bound SDR. The catalytic residues, for example, SX11-12YX2SK, occupy the same relative positions. An N-terminal extension would not affect the three-dimensional core structure. A reasonable topology model for Rdh1, based on the data generated here and by analogy to the structures of soluble SDRs, shows the enzyme anchored in the ER through the N-terminal signaling sequence facing the cytoplasm (Fig. 8). This model allows for interaction between Rdh1 and the cytosolic CRBP and/or retinal dehydrogenases.



View larger version (55K):
[in this window]
[in a new window]
  		FIG. 8.
A model of Rdh1 and its relationship with the ER membrane. Rdh1 was modeled using Swiss-Model Version 36.0002 based on the three-dimensional structures of soluble SDRs (human 17{beta}-hydroxysteroid dehydrogenase type I, 20{beta}-hydroxysteroid dehydrogenase, and {beta}-ketoacyl-[acyl-carrier protein]-reductase) using an optimal amino acid sequence alignment as described previously (26). Blue, green, red, and orange denote the N-terminal, protein association, catalytic, and C-terminal hydrophobic residues, respectively. The right panel shows a right side, top view of the left panel. In the right panel perspective, the white arrow denotes the N-terminal helix, which projects into the back of the plane perpendicular to the white arrow.

 
The ER-associated retinoid/steroid-metabolizing SDRs (e.g. Rdh1, Crad1, and rRodh1) have C-terminal extensions of ~32 amino acid residues relative to soluble SDRs. These C-terminal extensions do not seem obligatory for membrane association because the ER-associated SDR h11{beta}-HSD1 has no C-terminal extension relative to soluble SDRs, and as mentioned above the C terminus of Rdh1 did not direct GFP to the ER. Although the N-terminal first 22 residues convert the soluble GFP into an ER-anchored protein and the C terminus does not, Rdh1 cannot be detected in the ER without its C terminus. The model shows that the C-terminal hydrophobic region projects from Rdh1 in a plane distinct from the projection of the N terminus (Fig. 8, orange residues). The shape of the C terminus hydrophobic region, which consists of a helix-turn-helix, also suggests that it would not transverse a membrane, and the accessibility of C-terminal FLAG-conjugated Rdh1 to anti-FLAG demonstrated projection of the C terminus into the cytoplasm. Manipulation of Rdh1 in the program RasMol revealed no perspective in which both the N terminus and C terminus would associate with the ER membrane unless the protein occurs in a membrane "pocket" that wraps around it. These data are consistent with the C terminus stabilizing Rdh1 once it associates with the ER membrane but not projecting into the membrane or being required for initial membrane association. Because SDRs occur as multimers, it is reasonable to suggest that the C terminus aids multimerization and thereby stability of Rdh1 (47). The right panel of the model shows the protein association helices of SDR (Fig. 8, green), which includes the three catalytic residues (red), and suggests how the C-terminal hydrophobic region of one Rdh1 molecule could interact with another SDR molecule.

The mutagenesis data made clear the need for positive charges near the N terminus of the leader sequence to establish luminal topology and the nonessential nature of positive charges at the C-terminal end of the leader sequence to establish cytoplasmic topology. Multiple positive charges in both loci, however, such as in mutants L(3,5)2R, R3/R19Q, and R3/R21Q, caused about equal distribution between luminal and cytoplasmic orientations, indicating that net positive charge at the C-terminal end of the signaling sequence, although not obligatory, can offset the influence of N-terminal positive charges. The precise location of the positive charges at the N-terminal end of the signaling sequence did not seem to matter for reversing orientation from cytoplasmic to luminal (compare the R3 series of mutants with the L(3,5)2R series), but two charged residues, rather than one, did make a small but reproducible difference in the total amount of Rdh1 oriented to the lumen (compare mutant R3/R(16,19,21)3Q with L(3,5)2R/R(16,19,21)3Q). These data are consistent with the positive inside rule in terms of luminal orientation (44-46) but also show that ER localization with cytoplasmic topology can occur in the absence of positive charge in the leader sequence. Apparently, for type III (reverse signal-anchor) proteins such as Rdh1, hydrophobicity of the single N-terminal transmembrane signaling sequence contributes more to cytoplasmic orientation than does net positive charges. This is consistent with studies of the ER epoxide hydrolase, which faces the cytoplasm and has no net charge at its N terminus and a net charge of -2 at the C-terminal end of its signaling sequence (48).

Our data provide additional support for a recently proposed model of protein insertion into the ER membrane (49). Hydrophobic signaling sequences at the N termini of ER proteins target them for membrane insertion during translation (50, 51). The SRP, a cytosolic RNA-protein complex (52), binds the signal sequence and arrests translation until the nascent polypeptide-ribosome complex has been translocated to the translocation channel (translocon), which includes the Sec61p complex and the translocating chain-associating membrane protein (53, 54). Upon binding to the SRP receptor, SRP releases the signal sequence to enter the membrane (55, 56). Besides membrane targeting, the signal sequence determines protein topology (57, 58). In the "head-on insertion" model, the signaling sequence inserts co-translationally with its N terminus projecting directly into the translocon, corresponding to the final topology of a reverse signal-anchor protein, i.e. Nexo/Ccyt (49). The signaling sequence inverts only if its charges are mismatched with positive charges in the translocon as in the case of positive charges at its N-terminal end. This reversal of the signaling sequence then pulls the elongating peptide chain through the translocon into the lumen where it undergoes glycosylation to effect retention.

The N-terminal signaling sequences of the subfamily of SDRs that metabolize retinoids have highly similar amino acid compositions, which include short signaling sequences flanked by residues important to cofactor binding. All these signal sequences have the same structure characterized by no net charge at the N terminus, a hydrophobic helix of ~18 residues, and a net positive charge at the C terminus. The data generated here indicate that none of these SDRs should face the lumen. Yet both hRdh5 and Crad1 have been reported to be luminally oriented (23, 24, 47, 59). These conclusions were based largely on partial resistance to and/or slow proteinase K digestion at 4 °C in the absence of Triton X-100 and the supposition that proteins cannot resist proteinase K digestion. Proteinase K susceptibility was not determined at higher temperatures or compared with luminal SDRs. Indeed Crad1 did show marked proteinase K-catalyzed degradation even at 4 °C after 45 min in the absence of Triton X-100. Our data indicate not only that Crad1 faces the cytoplasm but also that Crad1 and related retinoid-metabolizing SDRs do not have signal sequences consistent with the mechanism of establishing luminal topology.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grants DK36870 and DK047839. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

The on-line version of this article (available at http://www.jbc.org) contains the primers used for mutant constructs. Back

{ddagger} To whom correspondence should be addressed: Dept. of Nutritional Sciences and Toxicology, University of California, 119 Morgan Hall, MC#3104, Berkeley, CA 94720. Tel.: 510-642-0809; Fax: 510-642-0535; E-mail: jna{at}nature.berkeley.edu.

1 The abbreviations used are: SDR, short-chain dehydrogenase/reductase; 11{beta}-HSD1, 11{beta}-hydroxysteroid dehydrogenase, type 1; Crad1, mouse cis-retinol/androgen dehydrogenase type 1; CRBP, cellular retinol-binding protein type I; ER, endoplasmic reticulum; Rdh1, mouse retinol dehydrogenase type 1; SLO, streptolysin O; SRP, signal recognition particle; GFP, green fluorescent protein; PNGase F, peptide-N-glycosidase F; PBS, phosphate-buffered saline; ConA, concanavalin A; DAPI, 4,6-diamidino-2-phenylindole. Back



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Wolf, G. (1984) Physiol. Rev. 64, 873-938[Free Full Text]
  2. De Luca, L. M. (1993) Basic Life Sci. 61, 17-25[Medline] [Order article via Infotrieve]
  3. Napoli, J. L. (1996) FASEB J. 10, 993-1001[Abstract]
  4. Zhang, M., Chen, W., Smith, S. M., and Napoli, J. L. (2001) J. Biol. Chem. 276, 44083-44090[Abstract/Free Full Text]
  5. Chai, X., Boerman, M. H. E. M., Zhai, Y., and Napoli, J. L. (1995) J. Biol. Chem. 270, 3900-3904[Abstract/Free Full Text]
  6. Chai, X., Zhai, Y., Popescu, G., and Napoli, J. L. (1995) J. Biol. Chem. 270, 28408-28412[Abstract/Free Full Text]
  7. Gough, W. H., Van Ooteghem, S., Sint, T., and Kedishvili, N. Y. (1998) J. Biol. Chem. 272, 19778-19785
  8. Jurukovski, V., Markova, N. G., Karaman-Jurukovska, N., Randolph. R. K., Su, J., and Napoli, J. L., and Simon, M. (1999) Mol. Genet. Metab. 67, 62-73[CrossRef][Medline] [Order article via Infotrieve]
  9. Boerman, M. H. E. M., and Napoli, J. L. (1996) Biochemistry 34, 7027-7037
  10. Boerman, M. H. E. M., and Napoli, J. L. (1996) J. Biol. Chem. 271, 5610-5616[Abstract/Free Full Text]
  11. Kedishvili, N. Y., Gough, W. H., Davis, W. I., Parsons, S., Li, T. K., and Bosron, W. F. (1998) Biochem. Biophys. Res. Commun. 24, 191-196
  12. Molotkov, A., Deltour, L., Foglio, M. H., Cuenca, A. E., and Duester, G. (2002) J. Biol. Chem. 277, 13804-13811[Abstract/Free Full Text]
  13. Molotkov, A., Fan, X., Deltour, L., Foglio, M. H., Martras, S., Farres, J., Pares, X., and Duester, G. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 5337-5342[Abstract/Free Full Text]
  14. Noy, N. (2000) Biochem. J. 348, 481-495[CrossRef][Medline] [Order article via Infotrieve]
  15. Ghyselinck, N. B., Båvik, C., Sapin, V., Mark, M., Bonnier, D., Hindelang, C., Dierich, A., Nilsson, C. B., Håkansson, H., Sauvant, P., Azaïs-Braesco, V., Frasson, M., Picaud, S., and Chambon, P. (1999) EMBO J. 18, 4903-4914[CrossRef][Medline] [Order article via Infotrieve]
  16. E, X., Zhang, L., Lu, J., Tso, P., Blaner, W. S., Levin, M. S., and Li, E. (2002) J. Biol. Chem. 277, 36617-36623[Abstract/Free Full Text]
  17. Gollapalli, D. R., Maiti, P., and Rando, R. R. (2003) Biochemistry 42, 11824-11830[CrossRef][Medline] [Order article via Infotrieve]
  18. Mata, N. L., Moghrabi, W. N., Lee, J. S., Bui, T. V., Radu, R. A., Horwitz, J., and Travis, G. H. (2004) J. Biol. Chem. 279, 635-643[Abstract/Free Full Text]
  19. Saari, J. C., Nawrot, M., Kennedy, B. N., Garwin, G. G., Hurley, J. B., Huang, J., Possin, D. E., and Crabb, J. E. (2001) Neuron 29, 739-748[CrossRef][Medline] [Order article via Infotrieve]
  20. Yamamoto, H., Simon, A., Eriksson, U., Harris, E., Berson, E. L., and Dryja, T. P. (1999) Nat. Genet. 22, 188-191[CrossRef][Medline] [Order article via Infotrieve]
  21. Jang, G. F., Van Hooser, J. P., Kuksa, V., McBee, J. K., He, Y. G., Janssen, J. J., Driessen, C. A., and Palczewski, K. (2001) J. Biol. Chem. 276, 32456-32465[Abstract/Free Full Text]
  22. Quadro, L., Blaner, W. S., Salchow, D. J., Vogel, S., Piantedosi, R., Gouras, P., Freeman, S., Cosma, M. P., Colantuoni, V., and Gottesman, M. E. (1999) EMBO J. 18, 4633-4644[CrossRef][Medline] [Order article via Infotrieve]
  23. Simon, A., Romert, A., Gustafson, A. L., McCaffery, J. M., and Eriksson, U. (1999) J. Cell Sci. 112, 549-558[Abstract]
  24. Tryggvason, K., Romert, A., and Eriksson. U. (2001) J. Biol. Chem. 276, 19253-19258[Abstract/Free Full Text]
  25. Wang, J., Bongianni, J. K., and Napoli, J. L. (2001) Biochemistry 40, 12533-12540[CrossRef][Medline] [Order article via Infotrieve]
  26. Song, M.-S., Chen, W., Zhang, M., and Napoli, J. L. (2003) J. Biol. Chem. 278, 40079-40087[Abstract/Free Full Text]
  27. Ghosh, D., Pletnev, V. Z., Zhu, D. W., Wawrzak, Z., Duax, W. L., Pangborn, W., Labrie, F., and Lin, S. X. (1995) Structure 3, 503-513[Medline] [Order article via Infotrieve]
  28. Azzi, A., Rehse, P. H., Zhu, D.-W., Campbell, R. I., Labrie, F., and Lin, S.-X. (1996) Nat. Struct. Biol. 3, 665-668[CrossRef][Medline] [Order article via Infotrieve]
  29. Breton, R., Housset, D., Mazza, C., and Fontecilla-Camps, J. C. (1996) Structure (Lond.) 4, 905-915[Medline] [Order article via Infotrieve]
  30. Ghosh, D., Weeks, C. M., Grochulski, P., Duax, W. L., Erman, M., Rimsay, R. L., and Orr, J. C. (1996) Proc. Natl. Acad. Sci. U. S. A. 88, 10064-10068
  31. Fisher, M., Kroon, J. T., Martindale, W., Stuitje, A. R., Slabas, A. R., and Rafferty, J. B. (2000) Struct. Fold. Des. 15, 339-347
  32. Addya, S., Anandatheerthavarada, H. K., Biswas, G., Bhagwat, S. V., Mullick, J., and Avadhani, N. G. (1997) J. Cell Biol. 139, 589-599[Abstract/Free Full Text]
  33. Bhagwat, S. V., Biswas, G., Anandatheerthavarada, H. K., Addya, S., Pandak, W., and Avadhani, N. G. (1999) J. Biol. Chem. 274, 24014-24022[Abstract/Free Full Text]
  34. Robin, M. A., Anandatheerthavarada, H. K., Biswas, G., Sepuri, N. B. V., Gordon, D. M., Pain, D., and Avadhani, N. G. (2002) J. Biol. Chem. 277, 40583-40593[Abstract/Free Full Text]
  35. Bergeron, J. J., Brenner, M. B., Thomas, D. Y., and Williams, D. B. (1994) Trends Biochem. Sci. 19, 124-128[CrossRef][Medline] [Order article via Infotrieve]
  36. Kuwana, T., Mackey, M. R., Perkins, G., Ellisman, M. H., Lattérich, M., Schneider, R., Green, D. R., and Niemeyer, D. D. (2002) Cell 111, 331-342[CrossRef][Medline] [Order article via Infotrieve]
  37. Zong, W. X., Li, C., Hatzivassiliou, G., Lindsten, T., Yu, Q. C., Yuan, J., and Thompson, C. B. (2003) J. Cell Biol. 162, 59-69[Abstract/Free Full Text]
  38. Palmer, M., Vulicevic, I., Saweljew, P., Valeva, A., Kehoe, M., and Bhakdi, S. (1998) Biochemistry 37, 2378-2383[CrossRef][Medline] [Order article via Infotrieve]
  39. Ozols, J. (1995) J. Biol. Chem. 270, 2305-2312[Abstract/Free Full Text]
  40. Odermatt, A., Arnold, P., Stauffer, A., Frey, B. M., and Frey, F. J. (1999) J. Biol. Chem. 274, 28762-28770[Abstract/Free Full Text]
  41. Frick, C., Antanasov, A. G., Arnold, P., Ozols, J., and Odermatt, A. (2004) J. Biol. Chem. 279, 31131-31138[Abstract/Free Full Text]
  42. Bokman, S. H., and Ward, W. W. (1981) Biochem. Biophys. Res. Commun. 101, 1372-1380[CrossRef][Medline] [Order article via Infotrieve]
  43. Lapshina, E. A., Belyaeva, O. V., Chumakova, O. V., and Kedishvili, N. Y. (2003) Biochemistry 42, 776-784[CrossRef][Medline] [Order article via Infotrieve]
  44. von Heijne, G. (1986) EMBO J. 5, 3021-3027[Medline] [Order article via Infotrieve]
  45. Hartmann, E., Rapoport, T. A., and Lodish, H. F. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, 5786-5790[Abstract/Free Full Text]
  46. Gafvelin, G., Sakaguchi, M., Andersson, H., and von Heijne, G. (1997) J. Biol. Chem. 272, 6119-6127[Abstract/Free Full Text]
  47. Lidén, M., Tryggvason, K., and Eriksson, U. (2003) Mol. Asp. Med. 24, 403-409
  48. Eusebio, A., Friedberg, T., and Spiess, M. (1998) Exp. Cell Res. 241, 181-185[CrossRef][Medline] [Order article via Infotrieve]
  49. Goder, V., and Spiess, M. (2003) EMBO J. 22, 3645-3653[CrossRef][Medline] [Order article via Infotrieve]
  50. Blobel, G. (1980) Proc. Natl. Acad. Sci. U. S. A. 77, 1496-1500[Abstract/Free Full Text]
  51. Johnson, A. E., and van Waes, M. A. (1999) Annu. Rev. Cell Dev. Biol. 15, 799-842[CrossRef][Medline] [Order article via Infotrieve]
  52. Schatz, G., and Dobberstern, B. (1996) Science 271, 1519-1526[Abstract]
  53. Matlack, K. E. S., Mothes, W., and Rapoport, T. A. (1998) Cell 92, 381-390[CrossRef][Medline] [Order article via Infotrieve]
  54. Hegde, R. S., Rapoport, T. A., and Lingappa, V. R. (1998) Cell 92, 621-631[CrossRef][Medline] [Order article via Infotrieve]
  55. Keenan, R. J., Freymann, D. M., Stroud, R. M., and Walter, P. (2001) Annu. Rev. Biochem. 70, 755-775[CrossRef][Medline] [Order article via Infotrieve]
  56. Walter, P., and Johnson, A. E. (1994) Annu. Rev. Cell Biol. 10, 87-119[CrossRef][Medline] [Order article via Infotrieve]
  57. Wahlberg, J. M., and Spiess, M. (1997) J. Cell Biol. 137, 555-562[Abstract/Free Full Text]
  58. Rösch, K., Naeher, D., Laird, V., Goder, V., and Spiess, M. (2000) J. Biol. Chem. 275, 14916-14922[Abstract/Free Full Text]
  59. Romert, A., Tuvendal, P., Tryggvason, K., Dencker, L., and Eriksson, U. (2000) Exp. Cell Res. 256, 338-345[CrossRef][Medline] [Order article via Infotrieve]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit