Identification and characterization of a cDNA encoding a dolichyl pyrophosphate phosphatase located in the endoplasmic reticulum of mammalian cells.

The CWH8 gene in Saccharomyces cerevisiae has been shown recently (Fernandez, F., Rush, J. S., Toke, D. A., Han, G., Quinn, J. E., Carman, G. M., Choi, J.-Y., Voelker, D. R., Aebi, M., and Waechter, C. J. (2001) J. Biol. Chem. 276, 41455-41464) to encode a dolichyl pyrophosphate (Dol-P-P) phosphatase associated with crude microsomal fractions. Mutations in CWH8 result in the accumulation of Dol-P-P, deficiency in lipid intermediate synthesis, defective protein N-glycosylation, and a reduced growth rate. A cDNA (DOLPP1, GenBank accession number AB030189) from mouse brain encoding a homologue of the yeast CWH8 gene is now shown to complement the defects in growth and protein N-glycosylation, and to correct the accumulation of Dol-P-P in the cwh8Delta yeast mutant. Northern blot analyses demonstrate a wide distribution of the DOLPP1 mRNA in mouse tissues. Overexpression of Dolpp1p in yeast, COS, and Sf9 cells produces substantial increases in Dol-P-P phosphatase activity but not in dolichyl monophosphate or phosphatidic acid phosphatase activities in microsomal fractions. Subcellular fractionation and immunofluorescence studies localize the enzyme encoded by DOLPP1 to the endoplasmic reticulum of COS cells. The results of protease sensitivity studies with microsomal vesicles from the lpp1Delta/dpp1Delta yeast mutant expressing DOLPP1 are consistent with Dolpp1p having a luminally oriented active site. The sequence of the DOLPP1 cDNA predicts a polypeptide with 238 amino acids, and a new polypeptide corresponding to 27 kDa is observed when DOLPP1 is expressed in yeast, COS, and Sf9 cells. This study is the first identification and characterization of a cDNA clone encoding an essential component of a mammalian lipid pyrophosphate phosphatase that is highly specific for Dol-P-P. The specificity, subcellular location, and topological orientation of the active site described in the current study strongly support a role for Dolpp1p in the recycling of Dol-P-P discharged during protein N-glycosylation reactions on the luminal leaflet of the endoplasmic reticulum in mammalian cells.

The multisubunit complex, oligosaccharyltransferase (1), catalyzes the transfer of Glc 3 Man 9 GlcNAc 2 from dolichyl pyro-phosphate (Dol-P-P) 1 to appropriate asparagine residues during the co-translational N-glycosylation of nascent polypeptides in yeast and mammalian cells (2)(3)(4). During the primary N-glycosylation reaction, Dol-P-P is released on the luminal surface of the endoplasmic reticulum (ER). In order for Dol-P-P to be re-utilized as a glycosyl carrier lipid for additional rounds of lipid intermediate biosynthesis, it must be converted to dolichyl phosphate (Dol-P) and translocated to the cytoplasmic leaflet of the ER (4). Although it cannot yet be excluded that Dol-P-P, or perhaps Dol-P, diffuses transversely from the luminal leaflet to the cytoplasmic face by a protein-mediated mechanism, it is more likely that it is dephosphorylated on the luminal surface to form free dolichol that could more readily diffuse back to the cytoplasmic leaflet. Cytoplasmically oriented dolichol would then be re-phosphorylated by dolichol kinase (5,6). Recent studies (7,8) have shown that the CWH8 gene in Saccharomyces cerevisiae encodes a Dol-P-P phosphatase that actively converts Dol-P-P to Dol-P and is also capable of dephosphorylating Dol-P at a slower rate. Moreover, the yeast Dol-P-P phosphatase is recovered in crude microsomal fractions, but its subcellular location has not been established.
Although there have been numerous reports (4, 7-13) that crude microsomal fractions from a variety of mammalian cells contain enzymes that can hydrolyze exogenous Dol-P-P, the identity, specificity, number, and exact subcellular location of the enzymes catalyzing the hydrolysis of Dol-P-P have not been established. Recently, Inoue et al. (14) have described a mouse brain cDNA, referred to as gene 2-23, with a high degree of homology to yeast CWH8. The current study presents several lines of evidence that mouse 2-23 encodes a mammalian homologue of the CWH8 gene product from yeast. In this report, this gene is referred to as Dol-P-P phosphatase 1 (DOLPP1). The identification of the mammalian homologue as a Dol-P-P pyrophosphate phosphatase is based on its ability to complement defects in growth and protein N-glycosylation in the cwh8⌬ mutant (8,15) and to prevent the accumulation of Dol-P-P in the mutant cells. In addition, expression of DOLPP1 modified with either a carboxyl-terminal His 6 tag in the yeast mutants or an amino-terminal extension encoding the T7 bacteriophage epitope in Sf9 cells results in the appearance of a new polypeptide with an apparent molecular mass of 27 kDa and a substantial increase in the level of Dol-P-P phosphatase activity.
The mouse enzyme exhibits a marked preference for Dol-P-P over Dol-P and phosphatidic acid (PA), as reported for the yeast enzyme.
The results described in this paper provide solid evidence that DOLPP1 encodes the mammalian homologue of the Dol-P-P pyrophosphate phosphatase encoded by the CWH8 gene in S. cerevisiae, and represent the first experimental proof that this enzyme is located in the ER. The possible function of this novel lipid pyrophosphate phosphatase in the recycling of the glycosyl carrier lipid in the ER of brain and other mammalian cells is discussed.

EXPERIMENTAL PROCEDURES
Materials-Plasmid YEp352 containing the yeast gene CWH8 coding sequence and 240 bp of 5Ј-flanking sequence was a gift from Drs. Fabiana Fernandez and Markus Aebi (ETH, Zurich, Switzerland). The cDNA designated as m2-23 was a gift from Dr. S. Inoue (Tsukuba Research Institute, Banyu Pharmaceutical Co., Ltd. Ibaraki Prefecture, Japan). Yeast strains W303-1A, cwh8⌬, and lpp1⌬/dpp1⌬ were the generous gifts from Dr. George Carman, Department of Food Science, Rutgers University, New Brunswick, NJ. n-Octyl-␤-D-glucopyranoside was purchased from Calbiochem-Novabiochem. Triton X-100 was from Pierce. Trichloroacetonitrile, tetrabutylammonium hydroxide, and anhydrous acetonitrile were obtained from Aldrich. Nycodenz, dioleoylphosphatidic acid (lot number 129F8359), and diolein (D-8894) were obtained from Sigma. Dolichol (C95) was a generous gift from Dr. M. Mizuno, Kuraray Chemical Co. (Okayama, Japan). Carrier-free [ 32 P]phosphoric acid was purchased from American Radiolabeled Chemicals (St. Louis, MO). Econo-safe TM biodegradable scintillation counting mixture is a product of Research Products International (Mount Prospect, IL). Yeast nitrogen base, yeast extract, and BactoPeptone are products of BD Biosciences. Casamino acids is a product of Fisher. Antibodies were obtained from the following commercial sources and used according to instructions from the supplier: rabbit anti-calnexin polyclonal antibody and mouse anti-KDEL monoclonal antibody (StressGen Biotechnologies, Victoria, British Columbia, Canada); mouse anti-KDEL receptor monoclonal antibody, mouse anti-␤-COP monoclonal antibody, horseradish peroxidase-conjugated sheep antirabbit IgG, horseradish peroxidase-conjugated sheep anti-mouse IgG (Amersham Biosciences); and fluorescein isothiocyanate-conjugated goat anti-rabbit IgG and Texas Red-conjugated goat anti-mouse IgG (Molecular Probes, Eugene, OR). Anti-CPY antiserum was generously provided by Dr. Neta Dean, State University of New York, Stony Brook, NY. All other chemicals and reagents were purchased from standard commercial sources.
Affinity-purified Rabbit Anti-Dolpp1p Antibodies and Immunoblotting-Three New Zealand White rabbits were each immunized with 300 g of a synthetic peptide corresponding to amino acid residues 86 -105 of mouse Dolpp1p coupled to keyhole limpet hemocyanin as described previously (16). The antigen was injected intradermally in Freund's complete adjuvant, and rabbits were boosted three times at 3-week intervals with 300 g of peptide in Freund's incomplete adjuvant. An IgG fraction was prepared from preimmune and immune serum by specific binding to protein A-Sepharose CL-4B (Amersham Biosciences). IgG was affinity-purified by specific binding to a column consisting of the Dolpp1p peptide cross-linked to SulfoLink coupling gel (Pierce). For analysis of DolPP1 expression by immunoblotting, samples were analyzed by electrophoresis in 15% SDS-PAGE gels and immunoblotting essentially as described (17). Filters were blocked for 1 h with Sea Block (East Coast Biologics, North Berwick, ME) and washed in PBS-T (0.25% (v/v) Tween 20 in phosphate-buffered saline) followed by incubation for 2 h with rabbit anti-Dolpp1p polyclonal antibody (1 g/ml). The filters were washed with PBS-T and incubated for 45 min in a solution containing horseradish peroxidase-conjugated secondary antibody (sheep anti-rabbit IgG, Amersham Biosciences) diluted 1:2500 in PBS-T. The filters were washed and developed using ECL chemiluminescence reagents (Amersham Biosciences).
Recombinant DNA Methods-Plasmid and genomic DNA preparation, restriction enzyme digestion, and DNA ligations were performed by standard methods (19). Transformation of yeast (20) and Escherichia coli (19) were performed as described. The annealing temperature for the PCR was 55°C, and extension times were typically 2.0 min at 72°C. PCRs were routinely run for a total of 30 cycles. DNA sequencing reactions were performed by Davis Sequencing, LLC (Davis, CA). For blot hybridization of RNA, total RNA was isolated from mouse tissues as described (21). Blots were pre-hybridized for 30 min at 65°C in Rapid-Hyb buffer (Amersham Biosciences), followed by hybridization with a random hexamer-primed 32 P-labeled mouse cDNA fragment corresponding to the entire 717-bp coding region of DolPP1 for 3 h at 65°C. The blots were washed in 2ϫ SSC (1ϫ SSC ϭ 150 mM NaCl and 15 mM sodium citrate (pH 7)), 0.1% SDS at ambient temperature for 30 min, followed by stringent washes in 0.5ϫ SSC, 0.1% SDS at 65°C for 15 min and in 0.1 ϫ SSC, 0.1% SDS at 65°C for 15 min.
Yeast Culture-Yeast cultures were grown at 30°C in either 1% yeast extract, 2% BactoPeptone, and 2% dextrose or in yeast nitrogen base (6.7 g/liter), 50 mM sodium succinate (pH 5.0), 2% dextrose, amino acids (25 mg/liter), and purine and pyrimidine bases (25 mg/liter) as required to meet auxotrophic requirements for selective growth. Yeast transformants were screened on 2% agar plates containing 6.7 g/liter yeast nitrogen base, 0.5% casamino acids, 50 mM sodium succinate (pH 5.0), 2% dextrose, and 25 mg/liter adenine or leucine as required. The yeast strains used in this study and their respective genotypes are contained in Table I.
Expression of Mus musculus DOLPP1 and DOLPP1-His 6 in Yeast-A DNA fragment containing 240 bp upstream of the coding sequence of the yeast CWH8 gene was amplified by PCR using the appropriate oligonucleotide primers (forward primer, 5Ј-ACTGAATTCGGGTTTCT-GGGAAAAC-3Ј; reverse primer, 5Ј-CCCGGTACCGATATGATCCGAT-TCAAAACA-3Ј) and plasmid YEp352 containing the CWH8 gene (8,15) as template. The forward and reverse primers were designed to contain altered bases to insert flanking EcoRI and KpnI restriction sites to facilitate cloning (underlined above). The resulting PCR product was digested with EcoRI and KpnI and ligated into similarly prepared YEp352 (22) to generate a plasmid containing the putative yeast CWH8 promoter sequence, YEp352/yPro. A DNA fragment containing the entire murine 2-23 coding sequence flanked by KpnI and BamHI restriction sites (underlined in sequences below) was amplified by PCR using the appropriate oligonucleotide primers (forward, 5Ј-GGTCTCCGGGT-ACCATGGCAGCGGA-3Ј; reverse, 5Ј-ACGGGATCCTCACTGCAGTTT-TGTT-3Ј) and a cDNA containing the mouse 2-23 gene obtained from Tsukuba Research Institute, Banyu Pharmaceutical Co., Ltd. (Tsukuba, Ibaraki, Japan), as template. The PCR product was digested with KpnI/BamHI to yield a 0.74-kb DNA fragment. This fragment was then ligated into the KpnI/BamHI sites of plasmid YEp352/yPro to form the plasmid YEp352/yPro-DOLPP1. The correct DNA sequence of YEp352/yPro-DOLPP1 was confirmed by direct sequencing using the M13 primer sites in YEp352. This construct was then transformed into the pertinent yeast strains for the expression of the mouse DOLPP1 gene according to Schiestl and Gietz (20).
DOLPP1 carrying a carboxyl-terminal His 6 tag was prepared by PCR using appropriate oligonucleotide primers (forward, 5Ј-ACTGAAT-TCGGGTTTCTGGGAAAAC-3Ј; reverse, 5Ј-TCTAAGCTTTCAGTGGT-GGTGGTGGTGGTGCTGCAGTTTTGTT-3Ј) and pBS/yPro-DOLPP1 cDNA as template. The PCR product was digested with EcoRI/HindIII (underlined in sequences above) and ligated into similarly prepared YEp352. The construct was transformed into the pertinent yeast strains for the expression of DOLPP1-His 6 according to Schiestl and Gietz (20). Yeast microsomal fractions for the routine analysis of lipid phosphatase activities were prepared exactly as described previously (8). Assays of Dol-P-P, Dol-P, and PA phosphatase activities were performed as described previously (9,23). Measurement of Cellular Levels of Dol-P and Dol-P-P in Yeast Cells-Yeast cells (1 liter) were grown in YPD containing 2% dextrose to 1-4 A 600 and collected by centrifugation at 1,000 ϫ g, 20 min. The cells were resuspended in distilled water, sedimented again, resuspended in distilled water to 200 A 600 /ml, and incubated for 30 min at 30°C in 50 mM Tris-HCl (pH 7.4), 10 mM MgCl 2 containing 1 mg/ml lyticase (Sigma). The yeast spheroplasts were sedimented at 1,000 ϫ g for 20 min; the supernatant was decanted, and the pellet was mixed vigorously with 40 volumes of CHCl 3 /CH 3 OH (2:1, v/v). Dol-32 P and Dol-P-32 P (5,000 cpm of each) were added as recovery markers, and the insoluble material was sedimented in a clinical centrifuge. The organic layer was removed and reserved on ice. The insoluble pellet was sequentially extracted twice with 5 ml of CHCl 3 /CH 3 OH (2:1 v/v) and twice with 5 ml of CHCl 3 / CH 3 OH/H 2 O (10:10:3, v/v). The organic extracts were combined, taken to dryness by rotary evaporation under reduced pressure at 30°C, and then deacylated in 1 ml of toluene/CH 3 OH (1:1, v/v) containing 0.1 M KOH on ice for 30 min. Following deacylation, the reaction was neutralized with acetic acid, diluted with 5 ml of CHCl 3 /CH 3 OH (2:1 v/v), and partitioned with 1 ml of water. The aqueous layer was discarded, and the organic phase was subsequently partitioned twice with 1 ml of CHCl 3 /CH 3 OH/H 2 O (3:48:47, v/v). The organic phase was dried under a stream of N 2 gas and redissolved in 0.2 ml of CHCl 3 /CH 3 OH/H 2 O (10:10:3, v/v). Dol-P and Dol-P-P in the organic extracts were separated by ion exchange chromatography on a 20-ml column of DEAE-650M (Toyopearl, Supelco, Bellefonte, PA) equilibrated with CHCl 3 /CH 3 OH/ H 2 O (10:10:3, v/v) and eluted with a linear gradient of 0 to 0.2 M ammonium acetate (60 ml). The fractions containing Dol-P and Dol-P-P were pooled separately and supplemented with CHCl 3 and water to give a final composition of CHCl 3 /CH 3 OH/H 2 O (3:2:1, v/v). The phases were separated by a brief centrifugation, and the aqueous layer was discarded. The organic layer was then partitioned once with 1/5 volume of CHCl 3 /CH 3 OH/H 2 O (3:48:47, v/v), taken to dryness under a stream of N 2 gas, and redissolved in a small volume of CHCl 3 /CH 3 OH/H 2 O (10: 10:3, v/v). Aliquots representing equivalent portions of the initial extracts were spotted on 10 ϫ 20-cm Baker Si250 Silica Gel thin layer plates and developed in CHCl 3 /CH 3 OH/H 2 0/NH 4 OH (65:35:6:1,v/v). Following chromatography, the thin layer plates were allowed to air dry and were visualized by staining with anisaldehyde (24). The migration positions of radioactive standards were determined with a Bioscan System 200 Imaging Scanner (Bioscan Inc. Washington, D. C.).
Expression of T7-DOLPP1 in Insect Cells-A DNA fragment containing the coding sequence of the mouse 2-23 gene flanked by KpnI and HindIII and a 5Ј-extension encoding the T7 bacteriophage epitope was obtained by PCR (forward primer, 5Ј-GGGGTACCATGGCTAGCAT-GACTGGTGGACAGCAAATGGGTATGGCAGCGGACGGA-3Ј; reverse primer, 5Ј-TCTAAGCTTTCACTGCAGTTTTGTTCCA-3Ј) using the cDNA plasmid containing the mouse 2-23 coding sequence as a template. The PCR product was digested with KpnI/HindIII (underlined in sequences above), ligated into the same restriction enzyme sites of the pFASTBAC1 baculovirus vector, and expressed in insect Sf9 cells using the Bac-to-Bac Baculovirus Expression System (Invitrogen). Sf9 cells were routinely grown at 27°C in Sf900 II SFM (Invitrogen) containing penicillin (50 units/ml) and streptomycin (50 g/ml) (25). Cells were infected with a viral multiplicity of 5-10 and harvested after 48 h. To prepare a microsomal fraction, Sf9 cells were sedimented by centrifugation at 1,000 ϫ g for 10 min, resuspended in 20 ml of ice-cold 10 mM HEPES (pH 7.0) and 0.25 M sucrose per ml of packed cells, and lysed by sonication (four 15-s pulses) with a Kontes Micro Ultrasonic Cell Disruptor at 40% full power. Unbroken cells and cellular debris were removed by centrifugation at 1,000 ϫ g for 10 min at 4°C. Microsomes were then sedimented (100,000 ϫ g for 10 min) in a Beckman TL-100 tabletop ultracentrifuge and resuspended at a membrane protein concentration of ϳ10 mg/ml in 10 mM HEPES (pH 7.0), 0.25 M sucrose for further analysis.
Expression of pCHA7/DOLPP1 in COS Cells-To construct a Dolpp1p transient expression plasmid for mammalian COS cells, the GenBank TM expressed sequence tag division was searched for clones homologous to S. cerevisiae Cwh8p, yielding IMAGE clone 1974625, GenBank TM accession number AI876279. (This clone was obtained independently of gene 2-23 sequence described above.) Both strands of the cDNA were sequenced, which revealed that the clone was incomplete at the 5Ј end. Therefore, a 5Ј-RACE procedure was employed using 1 g of mouse total brain mRNA and a SMART RACE cDNA Amplification Kit (Clontech Laboratories, Inc., Palo Alto, CA). Briefly, Moloney murine leukemia virus-reverse transcriptase Superscript II (Invitrogen) was used to perform the reverse transcription according to the supplier's protocol. PCR was performed using the supplied universal primer mix, a gene-specific primer (5Ј-CTGGGAATGGCTGGAGGGCATCCCG-3Ј), and a nested primer (5Ј-CGCCCTGATTCAGTGCCAGTCCCCC-3Ј) to obtain the 5Ј terminus of mouse DOLPP1 cDNA. The PCR was performed using a PerkinElmer Life Sciences GeneAmp System 9600 using the Advantage 2 PCR kit (Clontech Laboratories, Inc., Palo Alto, CA) under conditions suggested by the supplier. The resulting fragment was subcloned into the plasmid pGEM-T-Easy (Promega, Madison, WI), sequenced, and shown to be complete at the 5Ј end. The entire open reading frame was then amplified from total mouse brain cDNA using primers corresponding to sequences within the 5Ј-RACE product and mouse expressed sequence tag IMAGE clone 1974625 (5Ј-ATG-GCAGCGGACGGACAGTGC-3Ј and 5Ј-TCACTGCAGTTTTGTTC-CAAG-3Ј). PCR conditions for the amplification consisted of the following: 95°C, 5 min, one cycle; 94°C, 30 s, 68bto 58°C ("touch down" protocol), 30 s, 72°C, 1 min for 10 cycles; 94°C, 30 s, 58°C, 30 s, 72°C, 1 min for 25 cycles; 72°C, 10 min. The resulting PCR product was subcloned into pCHA7 (26) using SalI and BamHI restriction sites at the 5Ј and 3Ј ends of the insert, respectively. The construct was verified by sequencing, and the open reading frame was found to be 100% identical to GenBank TM AB030189, which was deposited subsequently (14).
Simian COS-7L cells were maintained in Dulbecco's modified Eagle's medium (high glucose), supplemented with 10% (v/v) fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 0.25 g/ml amphotericin B. Cells were plated on Day 0 in 100-mm dishes and were transiently transfected on Day 1 using 6 g of DNA (pCHA7/DOLPP1) and 18 l of FuGENE 6 reagent (Roche Molecular Biochemicals) per plate. The cells were harvested 48 h post-transfection into ice-cold phosphate-buffered saline and pelleted at low speed. Cell pellets were resuspended in a homogenization buffer containing 50 mM Tris-HCl (pH 7.5), 0.25 mM sucrose, 5 mM EDTA, and 5 mM 2-mercaptoethanol and sonicated briefly. Nuclei and intact cells were removed by centrifugation at 500 ϫ g for 5 min at 4°C. The postnuclear supernatant was further centrifuged at 100,000 ϫ g for 1 h at 4°C in a Beckman Optima TLX ultracentrifuge. The resulting supernatant (S100) was reserved, and the membrane pellet (P100) was washed once and resuspended in homogenization buffer for immunoblotting or in 10 mM HEPES (pH 7.0), 0.25 M sucrose at a concentration of 10 mg of protein per ml for analysis of enzyme activity.
Subcellular Localization of Dolpp1p-COS-7L cells (six 100-mm dishes) were transfected using pCHA7/DOLPP1 as described above and washed twice in a buffer consisting of 10 mM triethanolamine, 10 mM acetic acid, 250 mM sucrose, 1 mM EDTA, and 1 mM dithiothreitol. Cells were harvested with a rubber policeman in 800 l of buffer containing 10 g/ml each of chymostatin, leupeptin, and pepstatin and homogenized by 15 passages through a 25-gauge needle on a 1-ml syringe. Nuclei and intact cells were removed by microcentrifugation at 1,200 ϫ g for 5 min at 4°C. The postnuclear supernatant was loaded on preformed Nycodenz gradients prepared exactly as described (27). Briefly, linear Nycodenz gradients were prepared for the Beckman SW 41 Ti rotor from initial discontinuous gradients (24,19,15, and 10% Nycodenz in 10 mM Tris-HCl (pH 7.4), 3 mM KCl, and 1 mM EDTA) that were allowed to diffuse in a horizontal position for 45 min at room temperature and then centrifuged for 4 h at 37,000 rpm in a Beckman L8-M ultracentrifuge to generate a nonlinear density gradient profile. The postnuclear supernatant was loaded on top of the gradients and centrifuged for 1.5 h at 37,000 rpm. Fifteen fractions were collected, and aliquots of each fraction were subjected to electrophoresis on SDS-PAGE gels. The distribution of Dolpp1p, calnexin (ER marker), KDEL receptor (Golgi marker), and ␤-COP (Golgi marker) in the gradients was determined by immunoblotting using ECL chemiluminescence reagents (Amersham Biosciences).
Immunofluorescence-COS-7L cells were grown on 12-mm diameter coverslips and transfected with pCHA7/DOLPP1 using FuGENE 6 as described above. At 48 h post-transfection, the cells were fixed for 15 min in 4% formaldehyde and permeabilized for 10 min with 0.05% Triton X-100 at room temperature. The cells were doubly stained using the rabbit anti-Dolpp1p antibody and either the anti-KDEL monoclonal antibody or anti-KDEL receptor monoclonal antibody. The secondary antibodies used were fluorescein isothiocyanate-conjugated goat antirabbit IgG and Texas Red-conjugated goat anti-mouse IgG. Immunofluorescence was visualized under a Zeiss Axiophot immunofluorescence microscope.
Preparation of Sealed Microsomal Vesicles and Estimation of Vesicle Integrity-Sealed microsomal vesicles were prepared from the pertinent yeast strains and evaluated for integrity as described previously (8,28).
General Analytical Methods-Protein was determined by the method of Rodriquez-Vico et al. (29) using a commercial protein assay reagent (BCA, Pierce) and bovine serum albumin as standard. Lipid-phosphorus analysis was according to Bartlett (30). Radioactivity was measured by liquid scintillation spectrometry in a Packard Tri-Carb TR-2100 Liquid Scintillation Analyzer (Packard Instrument Co., Meriden, CT) in the presence of Econo-Safe Biodegradable Counting Mixture (Research Products International, Mount Prospect, IL).

Mammalian Cells Express a Homologue of the Yeast CWH8
Gene-Recently, Inoue and co-workers (14) identified and sequenced 64 cDNA clones from a mouse brain cDNA library that suppress bacterial growth when expressed in E. coli. One of these clones (GenBank TM accession number AB030189) has a high degree of homology to the yeast CWH8 gene (GenBank TM accession number NP_011550), an ER Dol-P-P phosphatase with a luminally oriented active site (7,8). Fig. 1 compares the deduced amino acid sequences of yeast Cwh8p and the predicted amino acid sequences from the mouse brain cDNA (designated Dolpp1p). The sequences (Fig. 1, panel A) are 29.8% identical and 49.4% strongly similar. In addition, Dolpp1p contains the consensus lipid-phosphate phosphatase motif (Fig. 1,  panel B) defined by Stukey and Carman (31), suggesting that Dolpp1p is the mammalian homologue of yeast CWH8.
The DOLPP1 mRNA was found to be widely expressed in various mouse tissues as assessed by RNA blotting (Fig. 2). A single transcript migrating at 2.0 kb was observed in all tissues examined, with highest expression in brain, kidney, lung, and intestine and low but detectable levels in liver, spleen, and uterus. The broad tissue distribution is consistent with the ER Dol-P-P phosphatase having an essential function in the lipid intermediate pathway for protein N-glycosylation in all mammalian cells.
Overexpression of Dolpp1p Corrects Defects in Growth Rate and N-Glycosylation of CPY in the cwh8⌬ Mutant-The mutant yeast strain, cwh8⌬, grows at an abnormally slow rate and exhibits a defect in protein N-glycosylation (8, 15). To determine whether expression of mammalian DOLPP1 corrects the growth deficiency of the cwh8⌬ yeast strain, the DOLPP1 gene was subcloned into YEp352 under the control of the putative yeast CWH8 promoter and transformed into cwh8⌬. The effect on rate of growth in complete media was then assessed. As shown in Table II, the cell density of wild type yeast cultures typically doubles approximately every 2 h, whereas the doubling time (T 2ϫ ) for cwh8⌬ was ϳ4 h. Overexpression of DOLPP1 in cwh8⌬ reduces the doubling time to 3 h, similar to the effect of transformation with YEp352 containing the yeast CWH8 gene (T 2ϫ ϭ 2.5 h).
As reported previously (8,15), mutation of the CWH8 gene in yeast results in defects in protein N-glycosylation, as assessed by N-glycosylation of carboxypeptidase Y (CPY) (Fig. 3). Although SDS-PAGE of wild type yeast extracts revealed a single protein band corresponding to mature, fully N-glycosylated CPY (Fig. 3, lane 1), extracts from the cwh8⌬ mutant contain several bands corresponding to underglycosylated isoforms of CPY (Fig. 3, lane 2). (The migration positions of CPY lacking 1-4 oligosaccharide chains are indicated by the arrows.) Expression of YEp352 carrying either the DOLPP1 gene or the yeast CWH8 gene restored full glycosylation to CPY (Fig. 3,  lanes 3 and 4, respectively). The conclusion that the faster migrating bands represent underglycosylated isoforms of CPY was confirmed by the observation that digestion with endoglycosidase H converted all of the protein bands into the fastest migrating isoform (data not shown).
Overexpression of Either Dolpp1p or Cwh8p Reverses the Accumulation of Dol-P-P in the cwh8⌬ Mutant-Abnormally

TABLE II
Overexpression of either DOLPP1 or CWH8 partially restores the normal rate of growth in the cwh8⌬ yeast mutant Yeast strains were grown at 30°C in 1% yeast extract, 2% Bacto-Peptone, and 2% dextrose (YPD) to an A 600 of ϳ0.1 OD/ml. A 600 measurements were collected at hourly intervals. Doubling times (T 2ϫ ) were calculated using the formula: T 2ϫ ϭ (T ⅐ log2)/log(A T1 /A T0 ). Where A T0 is the A 600 of the culture at the start of the culture period, A T1 is the A 600 at some later time, and T is the total elapsed time. Total RNA (20 g) from various tissues was subjected to electrophoresis and transferred to a nylon membrane. Hybridization was performed using a 32 P-labeled DOLPP1 cDNA probe (upper panel) followed by washing as described under "Experimental Procedures." The filter was exposed to Amersham Biosciences Hyperfilm MP film with an intensifying screen at Ϫ85°C for 4 days. The integrity of RNA in each lane was demonstrated by stripping and reprobing the filter with a 32 P-labeled cDNA probe encoding mouse cyclophilin A (lower panel).
high levels of Dol-P-P accumulate in cwh8⌬ yeast cells, due to a lack of the Dol-P-P phosphatase activity responsible for the hydrolysis of the Dol-P-P released during the transfer of the precursor oligosaccharide from Glc 3 Man 9 GlcNAc 2 -P-P-Dol to nascent polypeptides in the lumen of the ER (8). To determine whether overexpression of Dolpp1p in cwh8⌬ will restore normal Dol-P-P levels, total lipid extracts were prepared from either wild type, cwh8⌬ plus YEp352, cwh8⌬ plus YEp352/ yPro-DOLPP1, or cwh8⌬ plus YEp352/CWH8 yeast cells and analyzed for Dol-P-P (Fig. 4, panel A, lanes 1-4). As reported earlier (8), the lipid extracts from cwh8⌬ cells (Fig. 4, panel A,  lane 2), contained a prominent anisaldehyde-positive compound with the chromatographic mobility of standard Dol 75 -P-P. In contrast to this result, virtually no Dol-P-P was detected in extracts from wild type cells (Fig. 4, panel A, lane 1) or from cwh8⌬ yeast cells expressing either the DOLPP1 gene (Fig. 4, panel A, lane 3) or the CWH8 gene (Fig. 4, panel A, lane  4). To confirm the identity of the anisaldehyde-positive compound as Dol-P-P, this lipid was quantitatively converted to a product chromatographically identical to Dol 75 -P by mild acid hydrolysis (2 h at 80°C in CHCl 3 , CH 3 OH, 2 N HCl, 10:10:3, v/v). It should be noted that the levels of Dol-P in the various yeast strains were not significantly different (Fig. 4, panel B,  lanes 1-4).
Increase in Dol-P-P Phosphatase Activity Upon Expression of Dolpp1p in lpp1⌬/dpp1⌬ Yeast or Sf9 Insect Cells-The results described above indicate that the DOLPP1 gene encodes a functional homologue of the yeast Dol-P-P phosphatase. To provide more direct biochemical evidence that Dolpp1p functions as a Dol-P-P phosphatase, microsomal fractions were prepared from the lpp1⌬/dpp1⌬ yeast mutant strain after transformation with YEp352/yPro-DOLPP1, and the rate of hydrolysis of Dol-P-P was assayed. The lpp1⌬/dpp1⌬ strain was chosen as the parental strain for this study in order to minimize the nonspecific phosphatase activities encoded by LPP1 and DPP1 (7,8,13). Microsomes from lpp1⌬/dpp1⌬/ DOLPP1 were found to enzymatically dephosphorylate Dol-P-P at a markedly higher rate than microsomes from the lpp1⌬/ dpp1⌬ (Fig. 5, panel A) parental strain. Preliminary studies determined that the Dol-P-P phosphatase activity encoded by DOLPP1 was active in the presence of n-octyl glucoside and slightly less sensitive to inactivation by Triton X-100 than Cwh8p. Divalent cations were not required for activity for the mammalian enzyme (data not shown).
To examine the specificity of Dolpp1p, yeast microsomes from various yeast strains were assayed for Dol-P-P, Dol-P, or PA phosphatase activity. As seen in Table III,  mixtures for the determination of Dol-P-P phosphatase contained yeast microsomes (10 g of protein) from either lpp1⌬/dpp1⌬/DOLPP1 (q) or lpp1⌬/dpp1⌬ (E) strain, 10 mM EDTA, 50 mM sodium citrate/sodium phosphate (pH 7.0), 0.6% OG, and 20 M [␤-32 P]Dol-P-P in a total volume of 0.02 ml. Following incubation at 30°C for the indicated periods, the amount of 32 P i released was determined as described previously (9,23). For the comparison of the different phospholipid substrates (panel B), reaction mixtures contained yeast microsomes from lpp1⌬/dpp1⌬/DOLPP1 (10 g of protein), 10 mM EDTA, 50 mM sodium citrate/sodium phosphate (pH 7.0), 0.6% OG, and the indicated concentration of either [␤-32 P]Dol-P-P (q), [ 32 P]Dol-P (E), or [ 32 P]PA (‚) in a total volume of 0.02 ml. After 2 min at 30°C, the amount of 32 P i released was determined as described previously (9,23) .   FIG. 3. Overexpression of DOLPP1 cDNA corrects hypoglycosylation of CPY in the yeast cwh8⌬ mutant. Total cell extracts from the various yeast strains were separated by SDS-PAGE and analyzed by Western blotting using ␣-CPY serum as described previously (56). strain results in a large increase of Dol-P-P phosphatase activity and a very small increase in Dol-P phosphatase activity. No increase in PA phosphatase was observed.
The substrate concentration curves depicted in Fig. 5 (panel  B) show that Dol-P-P (q) is rapidly dephosphorylated by Dolpp1p, whereas Dol-P (E) and PA (‚) are relatively poor substrates over the entire concentration range tested. This comparison indicates that Dolpp1p exhibits a relatively high affinity for Dol-P-P (apparent K m Ͻ10 M). When [␣,␤-32 P]Dol-P-P was used as substrate in Dol-P-P phosphatase assays, approximately equal quantities of 32 P i and [ 32 P]Dol-P were formed during the initial rate phase of the reaction, indicating that Dolpp1p is indeed a novel lipid pyrophosphate phosphatase exhibiting high specificity for Dol-P-P.
Expression of an amino-terminally tagged T7-DOLPP1 construct in the insect cell Sf9-baculovirus system yielded both T7-immunoreactive protein and increased enzymatic activity (Fig. 6, panel A, and Table IV). When a Dolpp1p construct containing a carboxyl-terminal histidine tag was expressed in the lpp1⌬/dpp1⌬ yeast strain, a protein of the expected M r (28 kDa) was observed (Fig. 6, panel B) with a corresponding increase in Dol-P-P phosphatase activity (data not shown). Similarly, these results indicate that the mammalian DOLPP1 cDNA reconstitutes Dol-P-P phosphatase activity not only in yeast but also in insect cells and that small extensions at the amino and carboxyl termini of the protein produced no gross effects on enzyme activity. Although stable detergent-solubilized preparations have been obtained, attempts to purify the Dolpp1p in an active form from baculovirus-infected insect cell membranes have been unsuccessful to date.
Expression and Subcellular Localization of Dolpp1p in COS Cells-Although mutations in the CWH8 gene affect the rate of lipid intermediate synthesis and protein N-glycosylation in the ER, the precise subcellular location of the yeast enzyme has not been firmly established. To explore the properties of the DOLPP1 expressed in mammalian cells, the full-length mouse brain DOLPP1 cDNA was inserted into the eukaryotic expression vector pCHA7 to yield the plasmid pCHA7/DOLPP1, and expression in COS cells was detected by enzyme assay and immunoblotting. When microsomes from transfected cells were assayed, a large (12-16-fold) increase in Dol-P-P phosphatase activity was seen compared with identical microsomal fractions from cells transfected with the empty vector alone (Table IV). It is noteworthy that no concomitant increase in Dol-P or PA dephosphorylation was observed, corroborating the specificity for Dol-P-P. Soluble (S100) and particulate (P100) fractions from DOLPP1-transfected COS cells were also analyzed by SDS-PAGE and immunoblotting using a rabbit polyclonal antibody raised against a Dolpp1p peptide (Fig. 7). This analysis revealed a single band in the P100 fraction (Fig. 7, lane 4) at an M r of 27 in accord with the molecular mass of Dolpp1p predicted from the DOLPP1 cDNA sequence (27.1 kDa). No bands were visible in the S100 fraction or the fractions from the vector-transfected cells (Fig. 7, lanes 1-3).
To determine further the subcellular location of Dolpp1p, transfected COS cells were fractionated on nonlinear Nycodenz gradients, using a method previously developed to provide effective separation of ER and Golgi complex proteins (27). As shown in Fig. 8, Ͼ95% of the immunoreactivity sedimented to the bottom of the gradient (upper panel), co-sedimenting with calnexin, a well established ER marker. A much smaller amount was found at the top of the gradient co-fractionating with the Golgi markers, KDEL receptor and ␤-COP (Fig. 8). The minor immunoreactivity in Golgi shown in Fig. 8 was not observed in every experiment and seemed to correlate with higher levels of Dolpp1p expression.
When COS cells overexpressing DOLPP1 were examined by immunofluorescence using a specific Dolpp1p antibody, a reticular and nuclear envelope pattern characteristic of ER localization was observed (Fig. 9). This pattern essentially completely overlapped with the ER marker (KDEL, upper panel) and showed no overlap with a Golgi marker (KDEL receptor, lower panel).
Topological Orientation of the Active Site of Dolpp1p-To determine whether the active site of Dolpp1p faces the lumen of the ER or the cytoplasm, protease-sensitivity studies were conducted on the Dol-P-P phosphatase activity in intact and un-  III Dol-P, Dol-P-P, and PA phosphatase activities assayed in microsomes from various yeast strains Phosphatase reaction mixtures contained microsomes (20 g of protein) from the indicated yeast strain, 10 mM EDTA, 0.6% OG, 25 mM citrate-sodium phosphate (pH 7), and 20 M [␤-32 P]Dol-P-P, [ 32 P]Dol-P, or [ 32 P]PA in a total of 0.02 ml. Following incubation for 2 min at 30°C, the rate of dephosphorylation of each substrate was assayed as described previously (9,23  IV Dol-P-P, Dol-P, and PA phosphatase activities assayed in microsomes from insect Sf9 and COS cell lines expressing Dolpp1p Phosphatase reaction mixtures contained microsomes from either Sf9 cells (20 g of protein) or COS cells (10 g of protein) following transformation with either empty vector or vector carrying DOLPP1, 10 mM EDTA, 0.6% OG, 25 mM citrate-sodium phosphate (pH 7), and 20 M either [␤-32 P]Dol-P-P, [ 32 P]Dol-P, or [ 32 P]PA in a total of 0.02 ml. Following incubation for 5 min at 30°C, the rates of dephosphorylation of each substrate were assayed as described previously (9,23). Two independent experiments using COS cells are shown.  (Table V) were treated with trypsin, the Dol-P-P phosphatase activity was extensively inactivated (74.8% of control). However, the phosphatase activity was found to be relatively resistant to proteolysis in intact ER vesicles (87.5% of control). Separate pilot experiments employing glucosidase I/II latency (28) established that the sealed vesicles used in this experiment were Ͼ98% intact throughout the time of incubation and that 0.2% Triton X-100 was sufficient to unseal Ͼ95% of the yeast ER vesicles. Thus, it appears that the mammalian Dol-P-P phosphatase has a trypsin-sensitive site, perhaps part of the reactive site of the enzyme, that is luminally oriented, as in the yeast enzyme (Fig. 10).

DISCUSSION
When Glc 3 Man 9 GlcNAc 2 is transferred from the glycosyl carrier lipid to appropriate asparagine residues in nascent proteins, Dol-P-P is formed on the luminal monolayer of the ER. This luminally oriented Dol-P-P must be converted to Dol-P and return to the cytoplasmic leaflet of the ER in order to participate in subsequent rounds of lipid intermediate biosynthesis (4). The CWH8 gene in S. cerevisiae has recently been shown to encode a Dol-P-P-specific pyrophosphate phosphatase that was proposed to function in a model for the recycling of the carrier lipid (7,8). This paper describes the identification and characterization of a mammalian homologue of the yeast CWH8 gene, DOLPP1, from a mouse brain cDNA library.
Although there have been numerous reports (7-10, 13, 32-41) describing Dol-P-P/Dol-P phosphatase activities in crude microsomal fractions from various mammalian tissues, the exact number, location, and specificity have not been rigorously defined. Previous studies have been complicated by an inability to study the enzymatic properties of these proteins independent of the activities of other contaminating phosphatases. Expression of Dolpp1p in the lpp1⌬/dpp1⌬ yeast strain has allowed us to characterize the brain Dol-P-P phosphatase in the absence of other endogenous lipid phosphatase activities hydrolyzing Dol-P-P and Dol-P. The comparisons reported in Tables III and IV show that expression of Dolpp1p in either yeast, insect Sf9 cells, or COS cells results in a substantial and highly specific increase in Dol-P-P phosphatase activity. The experiment described in Fig. 5 reveals that Dolpp1p exhibits a marked preference for Dol-P-P (K m Ͻ10 M), similar to Cwh8p (ϳ25 M) when expressed in the lpp1⌬/dpp1⌬ yeast double  A and D) and one of two markers, anti-KDEL antibody (ER marker, red, B and C) or anti-KDEL receptor (Golgi marker, red, E and F). Dolpp1p (green) shows a reticulated and nuclear envelope pattern typical of ER localization, an impression that is confirmed by the overlapped images (merge, C). Dolpp1p staining was distinct from Golgi staining (F). Weak nuclear labeling was evident in some preparations and was deemed nonspecific because it appeared in the presence of either anti-Dolpp1p or non-immune antibodies (not shown). Experiment shown is one of two yielding similar results.

TABLE V
Effect of trypsinization on Dol-P-P phosphatase activity in intact and disrupted microsomal vesicles from lpp1⌬/dpp1⌬/DOLPP1 yeast Sealed yeast microsomal vesicles (10 g of membrane protein) from lpp1⌬/dpp1⌬/DOLPP1 yeast were preincubated for 60 min at 21°C with or without 0.2% Triton X-100 in the presence or absence of trypsin (10 g/ml) in a reaction mixture containing 1 mM MgCl 2 , 10 mM sodium-HEPES (pH 7.4), 0.25 M sucrose. Proteolysis was stopped by the addition of 5 mM phenylmethylsulfonyl fluoride, and Dol-P-P phosphatase activity was determined as described under "Experimental Procedures." Microsomal vesicles were Ͼ98% intact as determined by glucosidase I/II latency as described previously (28 Recently, lipid phosphate phosphohydrolases (LPPs) have been an active subject of research due to their role in potential signaling pathways (31,42). The type 2 LPP isoforms share extensive sequence similarity, are Mg 2ϩ -independent, N-ethylmaleimide-insensitive, and show little substrate specificity hydrolyzing a variety of lipid-phosphate substrates, including phosphatidic acid, lysophosphatidic acid, ceramide 1-phosphate, sphingosine 1-phosphate, and diacylglycerol pyrophosphate (42). DOLPP1 shares some properties with mammalian LPPs but shows only modest sequence homology (ϳ10%) with this family of phosphohydrolases and differs markedly in its high degree of specificity for Dol-P-P.
DOLPP1 shares a high degree of sequence identity with CWH8 and is shown to be a functional homologue in the cwh8⌬ mutant yeast strain. This conclusion is based on the following observations that overexpression of mouse DOLPP1 in cwh8⌬ yeast cells: 1) complements the growth defect; 2) corrects the deficiency in protein N-glycosylation; and 3) reverses the accumulation of Dol-P-P resulting from defective Dol-P-P phosphatase activity. Moreover, as noted above, expression of the DOLPP1 gene in the lpp1⌬/dpp1⌬ yeast strain, Sf9 insect cells, or mammalian COS cells produces large increases in Dol-P-P phosphatase activity, indicating that Dolpp1p is an essential component of the brain enzyme. The presence of a consensus lipid phosphatase motif within the amino acid sequence and the observations that overexpression of DOLPP1 in yeast and Sf9 cells produces dramatic increases in the level of Dol-P-P phosphatase activity strongly suggest that Dolpp1p contains the catalytic site. Whether other components may be needed for full activity is a question that will only be answered upon successful solubilization, isolation, and reconstitution of the recombinant enzyme in active form.
Although the effects of mutations in the CWH8 gene on protein N-glycosylation, lipid intermediate synthesis, and the accumulation of Dol-P-P are consistent with the yeast enzyme being located in the ER, these observations are only indirect support for this conclusion. The immunofluorescence microscopy and subcellular fractionation studies described in this report represent the first direct evidence that Dolpp1p is located in the ER in COS cells. In addition, protease sensitivity studies indicate that the mammalian Dol-P-P phosphatase has a critical domain that is luminally oriented. The relationship of this enzyme to the Dol-P-P phosphatase activity previously reported (9,10) to be enriched in the Golgi compartment in brain is not clear. However, it seems likely that unrelated lipid-phosphate phosphatases with broad substrate specificity and overlapping subcellular distribution may have obscured the proper localization of this activity. In this regard, other researchers have reported that Dol-P-P phosphatase activity is found in the plasma membrane in rat liver (32) or throughout the plasma membrane, rough ER, smooth ER, Golgi, and lysosomal fractions in mouse L-1210 cells (33). A small and variable proportion (Ͻ5%) of Dolpp1p was seen in the Golgi upon overexpression in COS cells. This result raises the possibility that the enzyme might be partially localized in a Golgi compartment under some conditions. However, it is clear from these studies that the DOLPP1 gene product is predominantly localized in the ER, the site of lipid intermediate biosynthesis, and protein N-glycosylation.
The specificity, location in the ER, and luminally oriented active site of Dolpp1p meet all the criteria for a role in the recycling of Dol-P-P released during primary protein N-glycosylation reactions on the luminal surface. There is substantial evidence that cellular levels of Dol-P are a rate-controlling factor in the regulation of lipid intermediate biosynthesis (44 -50). The induction of cis-isoprenyltransferase activity correlates closely with the onset of developmental increases in Dol-P and Glc 3 Man 9 GlcNAc 2 -P-P-Dol biosynthesis (51,52), suggesting that de novo synthesis is an important source of Dol-P for lipid intermediate synthesis. However, considering the impressive effect (ϳ80% reduction in Glc 3 Man 9 GlcNAc 2 -P-P-Dol synthesis) produced by mutations in CWH8 (8,15), it seems likely that recycling of Dol-P-P plays an important role in maintaining Dol-P supplies for Glc 3 Man 9 GlcNAc 2 -P-P-Dol biosynthesis. An important future goal will be to determine whether the Dol-P formed via the recycling of Dol-P-P exists in a separate pool to initiate new rounds of lipid intermediate synthesis or intermixes with the pool acquired by the de novo pathway.
The hypothetical models for Dolpp1p and Cwh8p, based on the TMpred program (43), indicate that the proteins contain four transmembrane helices and are topologically identical (Ref. 8, Fig. 10). The transmembrane domains (TMD) and the large luminal domains between TMD 1 and 2 are highly conserved in the two proteins. The high degree of conservation in the TMDs suggests that these domains may be critical to the function of Dolpp1p phosphatase and could be involved in the interaction with the polyisoprenyl moiety of Dol-P-P. The two putative luminal domains in Dolpp1p and Cwh8p have a significantly higher degree of homology than the cytoplasmic domains and may form part of the reactive site. Studies are in progress to determine the precise domains forming the active site of the enzyme.
The identification of the DOLPP1 gene is an important development, with implications for the inherited metabolic diseases in humans that occur due to defects in the synthesis of the dolichol-linked oligosaccharide donor required for protein N-glycosylation. These disorders are collectively referred to as congenital disorders of glycosylation (CDG) (53)(54)(55). Considering the wide variety of enzymes that have already been described as causing CDG, and the severity of the protein Nglycosylation deficiency in the yeast CWH8 mutants, it seems inevitable that a human CDG related to human DOLPP1 deficiency will be discovered. The accumulation of Dol-P-P in the luminal leaflet resulting from defects in Dolpp1p could block protein N-glycosylation by end product inhibition and interfere with other ER functions by causing biophysical alterations due to anionic microheterogeneities. The description of this mam- FIG. 10. Predicted topological arrangement of mammalian Dolpp1p and yeast Cwh8p in the ER. The topology of potential transmembrane domains of Dolpp1p was predicted using the TMPred transmembrane helices prediction algorithm (43). Identical and highly conserved residues in Dolpp1p and Cwh8p are shaded black and gray, respectively. malian cDNA should facilitate the clinical diagnosis of the underlying metabolic disorder in patients of this type.