Mouse Transporter Protein, a Membrane Protein That Regulates Cellular Multidrug Resistance, Is Localized to Lysosomes¹

A novel protein with nucleoside transport activity has been identified recently by the ability of a partial cDNA isolated from mouse leukemia L1210 cells to complement a thymidine transport deficiency in the yeast Saccharomyces cerevisiae (1). The full-length cDNA (GenBank accession number U34259) encodes a small hydrophobic protein named MTP³ with 233 amino acids and four predicted transmembrane domains (1). MTP is highly conserved, with 97% identity at the amino acid level to a human protein of unknown function, HUMORF13 (GenBank accession number D14696), suggesting an important role in mammalian cells. Recombinant MTP localized to the plasma membrane and exhibited low levels of nucleoside transport activity when produced in truncated form but not when produced in full-length form in oocytes of Xenopus laevis (1). The functional results, together with the demonstration of MTP in intracellular membranes of liver subcellular fractions, led to the suggestion by Hogue et al. (1) that MTP plays a role in the transport of nucleosides and/or related molecules across intracellular membranes. MTP is structurally unrelated to the equilibrative nucleoside transporter and concentrative nucleoside transporter protein families, the molecular identities of which have been determined by isolation and functional characterization of cDNAs in several heterologous expression systems (reviewed in Refs. 2, 3, 4).

Recently, full-length MTP was produced in drug-sensitive strains of S. cerevisiae and found to mediate a multidrug resistance phenotype (5). Yeast cells with MTP exhibited a collateral: (a) increased resistance toward anthracyclines, carboxylic and neutral ionophores, dihydropyrines, and steroids; and (b) increased sensitivity toward hydrophobic cations (i.e., ethidium and tetraphenylphosphonium), 5-fluorouracil, 5-fluorouridine, and trifluoperazine. MTP was also shown to alter the subcellular distribution of steroids in yeast. These observations indicate that the multidrug resistance phenotype resulted from MTP-mediated alterations in subcellular distribution of drug in yeast and suggest that a similar role exists for MTP in mammalian cells (5).

Multidrug resistance can arise in mammalian cells by several different types of biochemical changes that alter sensitivity to cytotoxic drugs (6). One of the most well-studied and widely accepted mechanisms for the generation of multidrug resistance is the overproduction of the ATP-dependent drug pump, P-glycoprotein (7). P-glycoprotein works primarily by mediating the efflux of drugs from cells; however, many forms of multidrug resistance, including P-glycoprotein-dependent multidrug resistance, have been associated with the sequestration of drugs into subcellular compartments (6, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19). Interestingly, some studies have demonstrated that drug-resistant cell lines differ from parental cell lines by accumulating drugs in different subcellular compartments (9, 10, 11, 12, 13, 14, 15, 18). In some cases, lysosomes have been implicated in the subcellular compartmentalization of the drugs (9, 10, 11, 12, 13, 14); but in other cases, the vesicular compartment has not been defined clearly (15, 18). Compartmentalization of cationic amphiphilic drugs into acidic organelles was originally thought to be due only to the action of proton pumps (20), wherein the transmembrane proton gradient leads to the intraorganelle trapping of protonated drug. However, recent evidence for the presence of active drug transport processes in lysosomes (21) and the involvement of MTP in the subcellular redistribution of drugs in yeast (5) suggest that protein-mediated processes are responsible for the accumulation of certain drugs within acidic vesicles.

In light of the unusual functional characteristics of recombinant MTP in yeast, the present study was undertaken to define the intracellular location of MTP in mammalian cells. In our previous study (1), based on the observation that a MTP-like protein was discovered by immunoblot analysis in subcellular membrane fractions of rat liver enriched in Golgi membranes, we concluded that MTP normally resides in intracellular membranes. However, MTP is a low-abundance protein and could not be detected by either indirect immunofluorescence of cells or immunoblot analysis of solubilized membranes with MTP-specific antibodies in the cell line (mouse leukemia L1210) from which its cDNA was initially obtained. To circumvent this difficulty, we have transiently transfected baby hamster kidney cells (BHK21) with cDNA encoding an epitope-tagged version of MTP and then examined for localization by double-label indirect immunofluorescence using anti-tag antibodies and antibodies to various organellar proteins. In addition, we have examined subcellular fractions obtained from rat liver for the presence of native MTP by immunoblotting with anti-MTP antibodies. The highest levels of MTP were found in late endosomes and lysosomes, indicating that MTP is a resident protein of lysosomes, thereby substantiating the recent suggestion (5) that MTP plays a role in the subcellular distribution of drugs associated with multidrug resistance. Preliminary observations have been reported previously (22).

The molecular biology techniques used are described in Ausubel et al. (23). All DNA primers were synthesized with a Perkin-Elmer/Applied Biosystems Model 394 DNA synthesizer (Foster City, CA) using cyanoethyl chemistry. PCR products were sequenced using PRISM Dye Terminator chemistry with a Perkin-Elmer/Applied Biosystems Model 373A DNA sequencer (DNA Core Facility, Department of Biochemistry, University of Alberta).

pcDNA3/MTP3 was constructed by ligating the EcoRI-MTP3-NotI fragment from pcDMTP3 (1) into the pcDNA3 mammalian expression vector (Invitrogen, Carlsbad, CA).

The HA-tagged version of MTP (termed HA-MTP) was prepared using the PCR. An oligonucleotide primer was used to introduce a BamHI site (GGATCC), the nonapeptide (Y-P-Y-D-V-P-D-Y-A) HA epitope tag (24, 25), and a consensus ribosomal binding domain (GCCACCATG; Ref. 26) at the 5′ end of the MTP cDNA; this primer (5′-GGGATCCGCCACC-ATGTACCCATACGATGTTCCAGATTACGCTATGGTGTCCATGAGTTTCAAGCGG-3′; start site is underlined) was termed MTPHA1. A second oligonucleotide primer was used to introduce an EcoRI site (GAATTC) after the stop codon on the 3′ end of the MTP cDNA; this primer (5′-GCAGAGAATTCTCAGGCAGGCAGGTAAGGAGG-3′; start site is underlined) was termed MTP-P2. The HA-MTP construct was obtained by amplification from pcDMTP3 (1) using the oligonucleotides MTPHA1 and MTP-P2. The construct was digested with BamHI and EcoRI and subsequently ligated into pcDNA3 to produce pcDNA3/HA-MTP.

BHK21 cells were obtained from the American Tissue Culture Collection (Bethesda, MD) and were cultured in DMEM supplemented with 10% FCS. Cells were grown as adherent cultures at 37°C in 5% CO₂ and cultured for 20 passages. New cultures were reinitiated from Mycoplasma-free stock cultures stored in liquid nitrogen. All media and sera were obtained from Life Technologies (Burlington, Ontario, Canada). Cells were enumerated using an electronic particle counter (Coulter Electronics, Miami, FL).

BHK21 cells were plated at 2 × 10⁵ cells/well in six-well tissue culture dishes and grown for 18–24 h until ∼60–80% confluency was reached. The DNA used in the transfections was purified using Midi or Mini columns (Qiagen, Mississauga, Ontario, Canada) according to the manufacturer’s instructions. The cells were transfected with 1 μg of either pcDNA3/HA-MTP, pcDNA3/MTP3, or pcDNA3 and 5 μg of Lipofectamine (Life Technologies) according to the manufacturer’s instructions.

Cultures were grown on 12-mm coverslips (five coverslips/well) contained in 35-mm wells and transfected exactly as described above. At 24–48 h after transfection, cells were fixed and permeabilized with 100% methanol at −20°C for 6 min. The cells (on coverslips) were washed twice with PBS and incubated at room temperature for 1 h or 4°C overnight with 2% goat serum in PBS to block nonspecific binding of antibodies. The cells were then double stained with rat monoclonal anti-HA (from clone 3F10, 100 ng/ml; Roche Molecular Biochemicals, Laval, Quebec, Canada; Ref. 25) and rabbit polyclonal antibodies against either: (a) α-mannosidase II (anti-α-Man II, diluted 1:500; Ref. 27); (b) mannose-6-phosphate receptor (anti-M6PR, diluted 1:500; Ref. 28, 29); (c) lysosomal glycoprotein 110 (anti-lgp-110, diluted 1:100; Ref. 30); (d) rab5A (diluted 1:100; Santa Cruz Biotechnology, Santa Cruz, CA); or (e) a synthetic peptide (TFKRSRSDRFYSTRC) corresponding to residues 5–19 of MTP (anti-MTP, diluted 1:1000; Ref. 1). Cells were washed with 1 ml of PBS three times and incubated with either: (a) Cy5-conjugated donkey anti-rabbit IgG (diluted 1:250) and FITC-conjugated goat anti-rat IgG (diluted 1:250); (b) Cy5-conjugated donkey anti-rabbit IgG (diluted 1:250) and FITC-conjugated goat anti-mouse IgG (diluted 1:250); or (c) Cy5-conjugated goat-anti-mouse IgG (diluted 1:250) and FITC-conjugated goat anti-rat IgG (diluted 1:250). The antibody conjugates were from Jackson ImmunoResearch Laboratories (West Grove, PA). All antibodies were diluted in PBS, and incubations were for 30 min at room temperature. Coverslips were mounted onto slides in 1 mg/ml paraphenylenediamine/90% glycerol in PBS. Samples were examined with an Axioskop immunofluorescence microscope (Carl Zeiss, Toronto, Ontario, Canada) or an LSM-510 laser-scanning confocal microscope (Carl Zeiss). Images were acquired using the LSM-510 software and printed out on an NP-1600 sublimation printer (Codonics, Middleburg Heights, OH).

Golgi membranes were isolated according to the protocol of Morré and Morré (31) with minor modifications. Three freshly isolated rat livers were placed in ice-cold homogenization buffer [0.5 m sucrose, 50 mm Tris-HCl (pH 6.9), 5 mm β-mercaptoethanol, and 1% dextran (Sigma, 500,000 g/mol)], minced with scissors, and homogenized with a Polytron homogenizer (Brinkmann Instruments, Mississauga, Ontario, Canada). The homogenate was filtered through cheesecloth and centrifuged (5000 × g for 15 min at 4°C). Most of the supernatant was discarded, and the upper one-third of the pellet was resuspended in a small amount of supernatant. The resulting suspension was layered over a gradient solution (1.2 m sucrose in distilled water containing 3% dextran) and centrifuged (100,000 × g for 30 min at 4°C). The Golgi-enriched membranes were removed from the top of the 1.2 m sucrose gradient, collected by centrifugation (5000 × g for 15 min at 4°C), and resuspended in homogenization buffer.

Preparation of lysosomes was performed according to the instructions accompanying the density gradient media, Nycodenz (Life Technologies) and according to Graham et al. (32). Freshly isolated rat livers were placed in ice-cold buffer A [0.25 m sucrose, 5 mm Tris-HCl (pH 7.6), and 1 mm EDTA], cut into small pieces, and homogenized with a Polytron homogenizer. The homogenate was filtered through cheesecloth and centrifuged (1,000 × g for 10 min at 4°C), and the postnuclear supernatant was centrifuged (15,000 × g for 10 min at 4°C). The resulting 15,000 × g pellet was resuspended in 11 ml of buffer A, which was combined with an equal volume of 50% Nycodenz (w/v in buffer A). A discontinuous Nycodenz gradient was assembled (0.5 ml of 40% Nycodenz, 1.0 ml of 30% Nycodenz, 3.5 ml of 15,000 × g pellet in 25% Nycodenz, 2.0 ml of 23% Nycodenz, 2.0 ml of 20% Nycodenz, 2.0 ml of 15% Nycodenz, and 1.0 ml of 10% Nycodenz) and centrifuged (52,000 × g for 1.5 h at 4°C) in a swinging bucket rotor. Lysosome-enriched membranes were collected from the 15/20% Nycodenz interface.

Mitochondria were isolated from rat livers using the procedure of Rickwood et al. (33) and Fleischer et al. (34). Freshly isolated livers were placed in ice-cold homogenization buffer (0.3 m sucrose, 1 mm EGTA, 5 mm 3-(N-morpholino)-propane sulfonic acid, 5 mm potassium phosphate monobasic, 0.1% BSA (fatty acid free), and 1 mm phenylmethylsulfonyl fluoride, pH 7.4) and rinsed with homogenization buffer to remove excess blood. The remainder of the procedure was carried out on ice. The livers were minced and washed with homogenization buffer. Homogenization buffer (2 ml/g liver) was added, and the minced liver was homogenized for 1 min using a motorized Polytron homogenizer. The resulting homogenate was strained through cheesecloth and centrifuged (1,000 × g for 10 min at 4°C). The supernatant was decanted and then centrifuged (10,000 × g for 10 min at 4°C), and the supernatant from this centrifugation was discarded. Next, the crude mitochondrial pellet was separated from the pelleted RBCs and resuspended in homogenization buffer (35 ml). This centrifugation step was repeated until the RBC pellet was removed. The crude mitochondrial pellet was further purified using centrifugation (4,000 × g for 40 min at 4°C) through a continuous sucrose gradient (0.75–1.75 m sucrose in 1 mm EGTA, 5 mm 3-(N-morpholino)propanesulfonic acid, 5 mm potassium phosphate monobasic, and 0.1% BSA, pH 7.4). The resulting fractions were collected, resuspended in homogenization buffer, centrifuged (10,000 × g for 20 min at 4°C) and resuspended at concentrations of 10–20 mg/ml. The functionality of the mitochondria was determined by measurement of the respiratory control ratio using a Clarke oxygen electrode (33).

Golgi and lysosomal preparations were either used immediately or were frozen in liquid nitrogen and stored at −70°C. Mitochondria were frozen in liquid nitrogen (in homogenization buffer containing 10 mg/ml BSA) and stored at −70°C. The samples were tested for β-galactosidase (lysosome-specific enzyme) and succinate INT reductase (mitochondria-specific enzyme) as described in Graham et al. (32) and for galactosyl transferase (Golgi-specific enzyme) as described in Bergeron et al. (35). Protein concentrations were determined using the bicinchoninic acid assay (36).

Untransfected and transfected cells plated in 35-mm wells as described above were harvested by trypsinization and collected by centrifugation. Cells were lysed, and membranes were solubilized in 100 μl of 1× sample loading buffer [50 mm Tris-HCl (pH 6.8), 100 mm DTT, 2% SDS, 0.1% bromphenol blue, and 10% glycerol] as described in Maniatis et al. (37). The DNA in the samples was sheared using a 1-ml syringe and 22-gauge needle. Cell lysates were collected by centrifugation (15,800 × g for 3 min at 24°C). Rat liver samples prepared as described above were mixed with an equal volume of 2× sample loading buffer (37). All samples were heated at 65°C for 5 min and centrifuged (15,800 × g for 1 min at 24°C), and the resulting supernatants were subjected to electrophoresis on 12% SDS-polyacrylamide gels (38) along with prestained broad-range (M_r 6,000–175,000) protein markers (New England Biolabs, Mississauga, Ontario, Canada). Proteins were electroblotted onto polyvinylidene difluoride membranes (Immobilon-P; Millipore, Bedford, MA). Membranes were incubated overnight at 4°C, first with 5% skim milk powder in blocking buffer (Tris-buffered saline, 0.2% Tween 20) and then for 12–18 h with 5% skim milk powder in blocking buffer that also contained primary antibodies (either anti-MTP diluted 1:10,000 or anti-HA diluted 1:1,000). The membranes were then washed three times with blocking buffer and incubated for 2 h at room temperature with either goat anti-rabbit IgG or goat anti-rat IgG conjugated to horseradish peroxidase, diluted 1:10,000 (Jackson ImmunoResearch Laboratories). Membranes were washed with blocking buffer three times and visualized using the enhanced chemiluminescence (ECL) reagents (Amersham Pharmacia Biotech, Baie d’Urfé, Quebec, Canada) according to the manufacturer’s instructions. Membranes were stripped and reprobed according to the instructions for use with the ECL reagents.

Native MTP was not detected in mouse L1210 leukemia cells (the source of the cDNA library from which the MTP cDNA was obtained) or freshly isolated rat hepatocytes by either immunoblotting of membrane preparations or indirect immunofluorescence of cells (1). In the present work, immunocytochemical studies (data not shown) with antibodies against MTP also failed to detect native MTP in BHK21 and Chinese hamster ovary cells or in human cervical carcinoma HeLa cells, which were shown previously to contain MTP-like mRNA (1). Endogenous levels of MTP are below the level of detection by these methods. Thus, we asked whether recombinant MTP and an epitope-tagged version, HA-MTP, could be localized by transient transfection of BHK21 cells with plasmids that contained the cDNA encoding either full-length MTP (pcDNA3/MTP3) or the HA-tagged version of MTP (pcDNA3/HA-MTP). Solubilized membrane preparations of transfected cells were analyzed by immunoblotting with antibodies directed against a synthetic peptide derived from the MTP sequence (Fig. 1,A) or against the HA epitope (Fig. 1 B). The anti-MTP antibodies recognized a protein of M_r ∼20,000–22,000 in cells transfected with pcDNA3/MTP3 and a protein with a slightly higher apparent molecular weight in cells transfected with pcDNA3/HA-MTP. The anti-HA antibodies, which did not detect any proteins in cells transfected with pcDNA3/MTP3, recognized a single band of M_r ∼21,000–23,000 in cells transfected with pcDNA3/HA-MTP. The anti-MTP and anti-HA antibodies did not detect any proteins in preparations from cells transfected with pcDNA3 alone.

In the experiments of Fig. 2, double-label indirect immunofluorescence experiments using anti-MTP and anti-HA antibodies were carried out on cells transiently transfected as described above. These experiments demonstrated that both MTP and HA-MTP were localized to vesicular structures concentrated in the perinuclear region (Fig. 2, A, D, E, G, and H). As expected, HA-MTP-containing vesicles were detected by both anti-MTP and anti-HA antibodies (Fig. 2, A, D, and G). No such vesicular staining was observed in cells transfected with vector alone (Fig. 2, C, F, and I). These observations suggested that the HA epitope tag did not affect the targeting of MTP, because the vesicles in both HA-MTP- and MTP-producing cells were similar in both size and location. Recombinant HA-MTP and MTP were found only in vesicular structures and were not detected at the plasma membrane. To facilitate the use of rabbit antiorganelle antibodies for double-label indirect immunofluorescence, experiments were thereafter conducted with cells that produced the epitope-tagged recombinant protein, HA-MTP.

The size and location of the MTP- and HA-MTP-containing vesicular structures suggested localization to endosomes and/or lysosomes. In addition, treatment of MTP- and HA-MTP-transfected cells with chloroquine by procedures described elsewhere (29) resulted in the swelling of MTP- and HA-MTP-containing vesicles (data not shown), suggesting that these vesicles were acidic in nature and thus part of the endosomal/lysosomal pathway.

To verify that the MTP-containing vesicles were not part of other perinuclear membranes such as Golgi, cells transfected with either pcDNA3/HA-MTP or pcDNA3 were subjected to double labeling with anti-HA antibodies and antibodies raised against the Golgi membrane protein, α-ManII (27). The Golgi apparatus in cells transfected with either plasmid exhibited a normal appearance, being crescent-shaped and juxtanuclear (Fig. 3, C and D). The structures labeled with the anti-HA and the anti-α-ManII antibodies were clearly different, and consequently there was no colocalization between these antibodies (Fig. 3, A, C, D, E, and F). Thus, recombinant HA-MTP was not detected in the Golgi apparatus by this technique.

The experiments of Figs. 4 and 5 were undertaken to determine whether MTP could be detected in early or late endosomes, respectively. Antibodies against the early endosomal marker, rab5A (39), and the late endosomal marker, cation-independent M6PR (28, 29), were used with the anti-HA antibodies to examine, by double indirect immunofluorescence, cells that had been transfected with either pcDNA3/HA-MTP or pcDNA3. The rab5A protein is a member of the large GTP-binding family that is required for membrane fusion and regulates transport between the plasma membrane and early endosomes (39). The cation-independent M6PR mediates the transport of soluble lysosomal enzymes from the trans-Golgi network to lysosomes and is concentrated in late endosomes/prelysosomes (29). The production of HA-MTP did not modify the staining pattern of either anti-rab5A (Fig. 4, C and D) or anti-M6PR antibodies (Fig. 5, C and D). In cells transfected with pcDNA3/HA-MTP, the staining pattern of the anti-HA antibodies partially overlapped with that of the anti-M6PR antibodies (Fig. 5, A, C, and E) but not with that of the anti-rab5A antibodies (Fig. 4, A, C, and E). Thus, although recombinant HA-MTP could not be detected by immunofluorescence in early endosomes, it was present in late endosomes.

The experiments of Fig. 6, which were undertaken to determine whether MTP was also present in lysosomes, used anti-lgp-110, an antibody that recognizes the lysosomal membrane glycoprotein 110 (30). The anti-lgp110 antibodies recognized large vesicular structures in cells transfected with either plasmid (Fig. 6, C and D). In cells transfected with pcDNA3/HA-MTP, the anti-HA antibodies labeled many of the same large juxtanuclear vesicles as the anti-lgp-110 antibodies (Fig. 6, A, C, and E). The anti-HA antibodies did not recognize any structures in cells transfected with pcDNA3 (Fig. 6 B). These results demonstrated the presence of recombinant HA-MTP protein in lysosomes.

Six rat ESTs exhibiting high homology with MTP and its human equivalent were isolated in a recent study of the effects of nerve growth factor on rat cells (40). Two of these ESTs (105428 and 106134) correspond to the NH₂ terminus of MTP, the region against which anti-peptide antibodies were raised, and because they are almost identical to MTP, the anti-MTP polyclonal antibodies should recognize the rat equivalent of MTP. Thus, we examined subcellular fractions of rat liver prepared using established protocols (31, 32, 33, 34). The identities of the membrane fractions enriched for lysosomes, mitochondria, and Golgi were verified by measurement of β-galactosidase, succinate INT reductase, and galactosyl transferase activities, respectively (Table 1). The lysosome-, mitochondria- and Golgi-enriched membrane fractions each exhibited high activity for the particular marker enzyme known to be associated with that organelle. The presence of rat MTP in these fractions was determined by immunoblotting (Fig. 7). Cell lysates from BHK21 cells that were transiently transfected with either pcDNA3/MTP3 or pcDNA3 were included as positive and negative controls, respectively, for the detection of MTP. A band of M_r ∼20,000–22,000 was observed in the Golgi- and lysosome-enriched membrane fractions but not in the mitochondria-enriched membrane fraction. These results demonstrated the presence of the rat equivalent of MTP in the Golgi- and lysosome-enriched membranes.

The objective of this study was to identify the subcellular location of MTP, a highly conserved membrane protein identified by isolation of its cDNA from a mouse leukemia L1210/C2 cDNA library by functional complementation of a thymidine-transport defect in S. cerevisiae (1). The high degree of conservation between the human and mouse proteins implies an important function for this protein family. Interestingly, rat ESTs that exhibit high sequence identity to MTP were isolated in a study that indicated that the gene encoding the rat equivalent of MTP can be down-regulated by treatment of pheochromocytoma cells with nerve growth factor (40). MTP and HUMORF13 share features with several proteins with related sequences that were identified in the GenBank database using the BLAST algorithm (41). MTP and HUMORF13, like the human and murine “lysosome-associated protein transmembrane 5” proteins, hLAPTm5 and mLAPTm5 (GenBank accession numbers U51240 and U51239, respectively; Ref. 42), have COOH termini that contain the tyrosine-based motifs, YXXØ, where Y is tyrosine, X can be any amino acid, and Ø; is a bulky hydrophobic amino acid residue (Fig. 8). This tyrosine-based motif has been implicated as an internalization and a lysosomal targeting signal (43, 44, 45, 46, 47, 48). One of the tyrosine-based motifs in MTP (Fig. 8) is located nine amino acids downstream from the last putative transmembrane domain that is predicted to be facing the cytosol, an arrangement that resembles that of the tyrosine-based lysosomal targeting signal in Lamp1 (47). The importance of the tyrosine-based motifs in the targeting of MTP to lysosomes has not been established.

We have demonstrated that MTP resides in late endosomes and lysosomes by two independent approaches: localization of an epitope-tagged version of recombinant MTP in intracellular membranes of transiently transfected cultured cells by indirect immunofluoresence and identification of a native MTP-like molecule by immunoblotting of subcellular fractions from rat liver. In the immunofluoresence experiments, recombinant HA-MTP and MTP were both found in large juxtanuclear vesicular structures, indicating that the epitope-tagged protein was targeted to the same intracellular location as the untagged protein. HA-MTP was specifically shown to be present in late endosomes and lysosomes, but not in the Golgi apparatus nor in early endosomes. Because both recombinant HA-MTP and MTP were produced at high levels in BHK21 cells, transient levels of these proteins are expected in late endosomes en route to lysosomes. In the subcellular fractionation experiments, an MTP-like protein was found in the Golgi- and lysosome-enriched preparations from rat liver. In both the earlier study (1) and this study, it is likely that the rat equivalent of MTP was observed in Golgi-enriched fractions because such preparations are likely to contain late endosomes. Also, some lysosomal membrane proteins can be localized to late endosomes en route to their final destination (44). Based upon the observations reported in this study, we propose that MTP be redesignated mLAPTm4 and HUMORF13 be designated hLAPTm4. This change in nomenclature more accurately reflects characteristics of the protein and allows for simplified cross-species naming.

What is the physiological function of mLAPTm4 (i.e., MTP)? Lysosomes are responsible for turnover of intracellular macromolecules as well as extracellular material that enters cells through phagocytosis and endocytosis. Human lysosomes have several well-described transport systems that function to release breakdown products into the cytosol (49), at least one of which is specialized for transport of nucleosides (50). The lysosomal nucleoside transport process, which was studied in lysosomes isolated from human fibroblasts, has a preference for purine nucleosides (e.g., 2′-deoxyadenosine and inosine) over pyrimidine nucleosides (e.g., uridine), exhibits lower affinities for nucleosides than the plasma membrane nucleoside-transport processes, and is inhibited by the well-known transport inhibitors dipyridamole and nitrobenzylthioinosine. The evidence that mLAPTm4 has nucleoside transport activity is based on studies in heterologous expression systems where thymidine uptake was observed with the truncated proteins, MTP1 and MTPΔC (1).

There is no mLAPTm4 homologue in either the S. cerevisiae or the Caenorhabditis elegans genomic databases, although various ESTs with extraordinarily high homology have been identified in vertebrate genomes (e.g., mouse, rat, rabbit, and zebrafish). A recent study (5) discovered that production of recombinant full-length mLAPTm4 in yeast confers increased resistance or sensitivity to a wide variety of drugs. Drug-sensitive yeast producing recombinant mLAPTm4 were found to have increased resistance to daunorubicin, doxorubicin, erythromycin, progesterone, and rhodamine-123 and increased sensitivity to 5-fluorouracil, 5-fluorouridine, and trifluoperazine. Several studies in mammalian cells have demonstrated drug accumulation in lysosomes (8, 9, 10, 11, 12, 13, 14, 19, 20, 21, 51), probably attributable to a variety of mechanisms. The demonstration that mLAPTm4 is a resident protein of mammalian lysosomes, combined with its functional activity when produced in yeast, suggests that mLAPTm4 plays a role in sequestering structurally unrelated amphiphilic molecules in lysosomes. Drug compartmentalization and subsequent accumulation in lysosomes is an important determinant of the multidrug resistance phenotype (6, 8, 9, 10, 11, 12, 13, 14, 16); however, little is known about the resident transporter proteins of lysosomes that are responsible for this process. Identification of the physiological substrates of mLAPTm4 awaits development of a system in which the transport characteristics of the full-length, functional protein can be studied in isolation from other transporters, for example, by functional reconstitution of recombinant mLAPTm4 in proteoliposomes or by expression in heterologous systems.

We thank Drs. B. Reaves and P. Luzio for generous gifts of anti-lgp110 antibodies; Dr. M. Farquhar for gifts of anti-α-ManII and anti-M6PR antibodies; Drs. L. Pilarksi and R. Godbout for gifts of Cy5-conjugated secondary antibodies; Dr. S. Rice for the use of his immunofluorescence microscope; Dr. W. Colmers for supplying rat livers; Drs. A. Claude and P. Melançon for the galactosyl transferase assay protocol; Dr. X. J. Sun for assistance with confocal microscopy; D. Mowles for expert technical assistance; and Dr. I. Coe for helpful discussions.