In polarized cells, the delivery of numerous membrane proteins from the trans-Golgi network to the basolateral surface depends on specific sequences located in their cytoplasmic domain. We have previously shown that the insulin-like growth factor-II/mannose 6-phosphate receptor (IGF-II/MPR) exhibits a polarized cell surface distribution in the human colon adenocarcinoma (Caco-2) cell line in which there is a threefold enrichment on the basolateral surface. To investigate the role of residues in the cytoplasmic region of the receptor that facilitates its entry into the basolateral sorting pathway, we generated stably transfected Caco-2 cell lines expressing various mutant bovine IGF-II/MPRs. The steady-state surface distribution of mutant receptors was analyzed by subjecting filter-grown cell monolayers to incubation with iodinated IGF-II/MPR-specific antibody or to indirect immunofluorescence and visualization by confocal microscopy. Together, these results demonstrate that the sorting of the IGF-II/MPR to the basolateral cell surface depends on recognition of sequences located in its cytoplasmic region that are distinct from the Tyr-based internalization and dileucine-dependent endosomal trafficking motifs.
- intracellular trafficking
- insulin-like growth factor receptor
- polarized cells
epithelial cells carry out a variety of vectorial transport and secretory processes that depend on the polarized distribution of proteins and lipids on their cell surface. The plasma membrane of these cells is divided into two morphologically, functionally, and biochemically distinct cell surface domains: 1) an apical domain that faces the exterior of the organism and 2) a basolateral domain that faces the internal environment. Epithelial cells are able to selectively direct newly synthesized membrane or secretory proteins to either of these domains, and polarity is maintained by the continuous sorting of apical and basolateral components in the secretory and endocytic pathways (48). Sorting signals that specify basolateral surface expression have been localized to the cytoplasmic domain of numerous membrane proteins (22, 29, 44). A number of these basolateral sorting signals are colinear with the signals for coated pit localization and belong to two subgroups: 1) an essential tyrosine residue in the context of NPXY or YXXØ (where Ø is a bulky hydrophobic residue) (28, 41, 42) and2) a dileucine motif (13, 18, 41, 43). In addition to directing basolateral expression, signals from both classes have been shown, in some cases, to also function in endocytosis from the plasma membrane and in mediating lysosomal sorting from the trans-Golgi network (TGN) to endosomes and lysosomes (18, 29,43).
The 300-kDa insulin-like growth factor-II/mannose 6-phosphate receptor (IGF-II/MPR) is a multifunctional protein that delivers newly synthesized lysosomal enzymes to the lysosome and regulates the circulating levels of IGF-II by mediating its uptake and lysosomal degradation. In higher eukaryotic cells, newly synthesized soluble acid hydrolases acquire mannose 6-phosphate (Man-6-P) residues on their N-linked oligosaccharides. In the Golgi, phosphomannosyl residues serve as high-affinity ligands for binding to the IGF-II/MPR. The removal of acid hydrolases from the secretory pathway occurs when the receptor-lysosomal enzyme complex enters into clathrin-coated pits and vesicles for delivery from the TGN to an acidified late endosomal compartment. The acidic pH of this compartment induces the complex to dissociate. Released lysosomal enzymes are then delivered to lysosomes, whereas the receptors either return to the Golgi to repeat the process or move to the plasma membrane where the IGF-II/MPR functions to internalize extracellular ligands via a recapture pathway (7, 23,33, 45). Numerous studies have been performed to identify those signals in the cytoplasmic region of the IGF-II/MPR that mediate its intracellular trafficking in nonpolarized cells. A conserved casein kinase II site followed by a dileucine motif (DDSDEDLL) at the COOH terminus of the IGF-II/MPR is important for sorting lysosomal enzymes to the lysosome (5, 6, 21, 27), whereas an aromatic-based sequence (YKYSKV) located 24–29 residues from the transmembrane region is essential for rapid internalization of the receptor (4,20, 27).
Our recent studies have demonstrated that in the human intestinal epithelial cell line Caco-2, the steady-state polarized secretion of lysosomal enzymes into the apical medium is facilitated by the IGF-II/MPR selectively endocytosing lysosomal enzymes from the basolateral surface (47). We have previously demonstrated that the IGF-II/MPR is enriched threefold on the basolateral surface relative to its expression on the apical cell surface (9). This polarized surface expression of the IGF-II/MPR to the basolateral surface is not limited to Caco-2 cells, because studies by Prydz et al. (40) report that the IGF-II/MPR could be detected on the basolateral, but not the apical, surface of Madin-Darby canine kidney (MDCK) cells. However, nothing is known about the structural determinants of the receptor that mediate its polarized distribution in epithelial cells. To determine whether sequences in the cytoplasmic region of the IGF-II/MPR serve as a targeting signal to the basolateral cell surface, mutant forms of the bovine IGF-II/MPR containing truncations in its COOH-terminal cytoplasmic region were expressed in polarized human Caco-2 cells and the surface distribution of the bovine MPRs was determined. Our results demonstrate that residues essential for entry of the IGF-II/MPR into the basolateral sorting pathway are located within residues 36–50 of the cytoplasmic domain and are distinct from both the aromatic-based motif required for rapid internalization and from the dileucine motif required for endosomal sorting.
MATERIALS AND METHODS
The following reagents were obtained commercially as indicated: EXPRE35S35S 35S-protein labeling mix (1,200 Ci/mmol, NEN Life Science); fetal bovine serum (FBS; HyClone Laboratories); DMEM and trypsin-EDTA (GIBCO-BRL Life Technologies); and protein A-Sepharose, glucose 6-phosphate (Glc-6-P), and Man-6-P (Sigma). Caco-2 cells were kindly provided by Dr. Ward Olsen of Veterans Administration Hospital (Madison, WI).
Caco-2 cells (passages 76–96) were grown in DMEM (25 mM glucose) supplemented with 20% heat-inactivated FBS, 4 mM glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin in a humidified atmosphere containing 5% CO2. To determine the cell surface distribution of the IGF-II/MPR, cells were grown as epithelial layers by high density seeding (3.4 × 105cells/cm2) onto Transwell polycarbonate membrane filter inserts (Costar). The formation and integrity of monolayers were assessed by the development of a significant transepithelial electrical resistance of 250–300 ohms/cm2 over the resistance of the filter alone. Resistance readings were measured with a Millicell-ERS Voltohmeter (Millipore). All polarity studies were performed >10 days after plating.
Transfection of Caco-2 cells.
Generation of the bovine wild-type and mutant (YAYA, Val8, Asn21, Leu75, and Arg124) IGF-II/MPR constructs was described previously (27). The construct encoding the Asp36 mutant IGF-II/MPR was described previously (4). A mutant IGF-II/MPR lacking the COOH-terminal 113 residues was generated by substituting Ala51 of the cytoplasmic region with a stop codon using a single-step polymerase chain reaction-based method [forward primer, 5′-CGTCCATCACGGGCTCCAGCA3′; reverse mutagenic primer, 5′-CGACGCGTTCATCACGGCGGCTGGATCTCCTCCATC3′ (stop codon underlined)]. The stop codon at Ala51 was followed by a second stop codon and an Mlu I site. The resulting PCR product was digested with BsrG I and Mlu I and cloned into the corresponding sites of the IGF-II/MPR. The region of the construct synthesized by PCR was confirmed by DNA sequencing. These constructs were placed in the SFFV-neo vector (14) that uses the Friend spleen focus-forming virus 5′-long terminal repeat to promote transcription of the cDNA and also contains the neomycin resistance gene that confers resistance to the antibiotic G418. Caco-2 cells were seeded in 100-mm dishes and transfected 48 h after plating with 35 μg of Xba I-linearized plasmid DNA using a modification of the calcium phosphate method. Briefly, precipitates were formed by adding equal volumes of DNA in 250 mM CaCl2and 2× HeBS buffer [1× HeBS consists of (in mM) 140 NaCl, 5 KCl, 0.75 Na2HPO4, 6 dextrose, and 25 HEPES, pH 7.05] and the mixture was incubated at room temperature for 30 min. The precipitated DNA was added to Caco-2 cells in DMEM, and cells were incubated for 6 h at 37°C. Cells were then treated with 15% DMSO in HeBS for 1 min, rinsed, and then placed in DMEM containing 20% FBS. After 72 h, cells were passaged and seeded to 100-mm dishes using serial dilutions. Selection was started 48 h later using 500 μg/ml of G418 sulfate (GIBCO-BRL Life Technologies) and continued for 10–14 days. Clones were isolated either by using cloning rings or by serial dilution in 96-well plates, and cells were grown in the continuous presence of 350 μg/ml G418. The expression level of the recombinant bovine IGF-II/MPR in each of the clones was determined by quantitative Western blot analysis (see below).
Caco-2 cells were starved for 15 min in DMEM lacking methionine and cysteine (GIBCO-BRL Life Technologies) containing 10% heat-inactivated FBS (DMEM-FBS). Cells were then incubated in DMEM-FBS containing EXPRE35S35S 35S-protein labeling mix (50 μCi/ml) for 19 h. Cells were solubilized for 1 h on ice in buffer containing 0.1 M Tris, pH 8.0, 0.1 M NaCl, 10 mM EDTA, Triton X-100 (1% vol/vol), sodium deoxycholate (0.1% wt/vol), aprotinin (1% vol/vol), antipain (4 μg/ml), benzamidine (20 μg/ml), and 2 μg/ml each of leupeptin, chymostatin, and pepstatin. The IGF-II/MPR was purified from the resulting cell lysates by pentamannosyl phosphate-agarose affinity chromatography (8,17). Protein-encoding domains 1–13 of the extracytoplasmic region of the bovine IGF-II/MPR was purified from transiently transfected COS-1 cells metabolically labeled with EXPRE35S35S 35S-protein labeling mix as described previously (10).
Purified IGF-II/MPRs were incubated at 4°C for 16–24 h with protein A-Sepharose plus anti-IGF-II/MPR polyclonal antibody (B2.5) or monoclonal antibodies (32g, 56f, 70h, and 86f7) plus rabbit anti-mouse IgG. IGF-II/MPRs were purified from [35S]methionine-labeled cells by pentamannosyl phosphate-agarose affinity chromatography (8, 17). After recovery by centrifugation, the protein A-Sepharose beads were washed four times with buffer containing 0.1 M Tris (pH 8.0), 0.1 M NaCl, 10 mM EDTA, and 1% Triton X-100 and washed once in buffer containing 20 mM Tris (pH 8.0) and 20 mM NaCl. Bound proteins were eluted by the addition of Laemmli sample buffer and analyzed on 7.5 or 9% SDS polyacrylamide gels under reducing conditions. The radiolabeled bands were analyzed using a PhosphorImager (Storm 860; Molecular Dynamics) with ImageQuant (version 4.1) software.
Western blot analysis.
Caco-2 cells were solubilized for 1 h on ice in buffer containing 0.1 M Tris, pH 8.0, 0.1 M NaCl, 10 mM EDTA, Triton X-100 (1% vol/vol), sodium deoxycholate (0.1% wt/vol), aprotinin (1% vol/vol), antipain (4 μg/ml), benzamidine (20 μg/ml), and 2 μg/ml each of leupeptin, chymostatin, and pepstatin. The total amount of protein in the resulting cell lysate was determined using the Bradford protein assay as recommended by the manufacturer (Bio-Rad). The resulting cell lysates were subjected to quantitative Western blot analysis as described previously (51), except that, after incubation with the 86f7 monoclonal antibody, the membranes were probed with horseradish peroxidase-conjugated goat anti-mouse IgG (Pierce) and the proteins were visualized using enhanced chemiluminescence (ECL) as recommended by the manufacturer (Pierce). Bands were quantified using an Ambis radioanalytical imaging system.
Surface binding and internalization of the monoclonal antibody 86f7.
86f7 hybridoma media was adjusted to 50% ammonium sulfate, and after centrifugation, the concentrated antibody was dialyzed against PBS. 86f7 antibody was iodinated using IODO-GEN (Pierce) to a specific activity of 5–13 μCi/μg. Transfected Caco-2 cells were grown on 12- or 24-mm Costar Transwell filters for 10–14 days. Cells were rinsed with binding buffer (PBS containing 0.1 mM CaCl2, 1.0 mM MgCl2, and 0.5% BSA), and iodinated 86f7 was then added to the apical or basolateral surface at a concentration of 0.5–1.0 μg/ml in serum-free DMEM containing 1% BSA and incubated for 2 h at 37°C. After cells were chilled to 4°C in an ice-water bath, media was removed from both surfaces and counted to confirm the integrity of the monolayer. Media was then combined and an aliquot was precipitated with 10% trichloroacetic acid and compared with a control for nonprecipitable counts. Cells were then washed three times each with ice-cold binding buffer and wash buffer (binding buffer without BSA). Surface-bound 86f7 was removed by washing in ice-cold low pH buffer (50 mM glycine, pH 2.8, 150 mM NaCl), rinsed with wash buffer, and cells were removed from the filters by treatment with 0.1 N NaOH. Total protein was determined from the cell lysate by the Lowry method (Bio-Rad DC protein assay). Radioactivity in the trichloroacetic acid-soluble media, low pH wash, and cell lysates were counted in a γ-counter and normalized to total protein to give degraded, surface, and internalized values, respectively. Nontransfected Caco-2 cells were used as a control for nonspecific binding. An internalization index was determined by the sum of internalized and degraded values divided by the surface-bound values and serves to normalize for variations in the level of receptor expression among cell lines.
Surface binding and internalization of β-glucuronidase.
To determine surface ratios and extent of internalization of the endogenous human IGF-II/MPR, similar experiments as described above were performed on nontransfected Caco-2 cells except that the iodinated lysosomal enzyme β-glucuronidase (2.0 nM) was used in place of the monoclonal antibody 86f7. β-glucuronidase was purified and iodinated as described previously (9). Nonspecific binding was determined by incubation in the presence of 10 mM Man-6-P. To determine the amount of surface-bound ligand, cells were washed with 10 mM Glc-6-P (nonspecific ligand) followed by 10 mM Man-6-P (specific ligand).
Steady-state surface distribution of the IGF-II/MPR.
Transfected Caco-2 cells were grown in 12- or 24-well tissue culture plates to postconfluence. On days 7 and 8 of growth, cells were washed and chilled to 4°C with binding buffer and incubated with 0.5–0.65 μg 125I-86f7 in binding buffer for 3 h at 4°C to measure the amount of receptor on the cell surface. To determine total IGF-II/MPR levels, parallel wells were incubated with 125I-86f7 in the presence of 0.1% saponin. Cells were washed three times each with ice-cold binding buffer and wash buffer and then harvested in 0.1N NaOH. Radioactivity was measured in a γ-counter and normalized to total protein. The amount of cell surface receptors (measured in the absence of saponin) was calculated as a percentage of the total receptors (measured in the presence of saponin). Nontransfected cells were used as a control for nonspecific binding.
Transfected Caco-2 cells were grown on 12-mm Costar Transwell polycarbonate filters (pore size, 0.4 mm) for 10 days before immunostaining. Cells were fixed as described previously (13) in a 1:9 methanol/acetone solution followed by rehydration with PBS. A 1:200 dilution of the monoclonal antibody 86f7 (specific for the bovine IGF-II/MPR) was added to both the apical and basolateral surfaces for 1 h at room temperature. After washing with PBS, both surfaces were incubated with 1:100 dilution of a fluorescein isothiocyanate-conjugated goat anti-mouse antibody (Pierce) for 1 h at room temperature. Both antibodies were diluted in PBS containing 3% goat serum. Nontransfected cells were incubated with the polyclonal antibody B2.5 (raised against bovine liver IGF-II/MPR) followed by fluorescein isothiocyanate-conjugated goat anti-rabbit antibody (Pierce). After washing, the filters were removed and mounted in VECTASHIELD (Vector Laboratories). Images were obtained with a Bio-Rad MRC 600 laser scanning confocal imaging system mounted on a Nikon Optiphot microscope using a 100× objective. Data analysis was accomplished using MetaMorph software.
Expression of the bovine IGF-II/MPR in human Caco-2 cells.
To distinguish recombinant bovine IGF-II/MPRs expressed in Caco-2 cells from the endogenous human receptor, species-specific antibodies were identified. We obtained four monoclonal antibodies generated against the bovine IGF-II/MPR. To determine whether these antibodies recognize the human IGF-II/MPR produced by Caco-2 cells, Caco-2 cells were metabolically labeled with [35S]methionine and the IGF-II/MPR was purified by pentamannosyl phosphate-agarose affinity chromatography. [35S]methionine-labeled protein encompassing domains 1–13 of the extracytoplasmic region of the bovine IGF-II/MPR was used as a positive control (10). Equal aliquots of either the purified Caco-2 IGF-II/MPR or the purified bovine IGF-II/MPR (domains 1–13) were incubated with various antibodies plus protein A-Sepharose. Figure 1 A shows that the four monoclonal antibodies (32g, 56f, 70h, and 86f7) efficiently recognized the purified bovine IGF-II/MPR encompassingdomains 1–13 but were unable to immunoprecipitate the purified human Caco-2 IGF-II/MPR. In contrast, the polyclonal antibody B2.5 efficiently recognized both the human Caco-2 and bovine IGF-II/MPRs. As a control, the supernatants from samples in Fig. 1 A lanes 1, 3–7, 9, and 15 were reimmunoprecipitated with the polyclonal antibody B2.5 to demonstrate that intact Caco-2 or bovine IGF-II/MPR were, in fact, present in these samples (Fig. 1 B). These results demonstrated that the monoclonal antibodies 32g, 56f, 70h, and 86f7 exhibit species-specificity in that they recognize the bovine, but not the human, IGF-II/MPR. In addition, the observation that these monoclonal antibodies precipitate a construct that lacks the cytoplasmic region of the IGF-II/MPR (construct encodes domains 1–13 of the extracytoplasmic region, see Fig. 1) demonstrates that mutations placed in the cytoplasmic region of the receptor will not inhibit antibody recognition. We have previously mapped the epitope of the 32g, 56f, and 86f7 monoclonal antibodies to domain 5of the extracytoplasmic region of the IGF-II/MPR (3). Thus the use of these monoclonal antibodies allowed for further specific analyses of recombinant bovine IGF-II/MPRs expressed in Caco-2 cells.
Studies from a number of laboratories have demonstrated that the cytoplasmic domain of various proteins contains determinants required for their polarized distribution in epithelial cells (12, 13, 16,41, 42). To identify the residues essential for the polarized expression of the IGF-II/MPR, Caco-2 cells were transfected with the bovine IGF-II/MPR cDNA containing various mutations in the cytoplasmic domain. Four mutations (Val8, Asn21, Lev75, Arg124) involved the introduction of a termination codon at various positions along the cytoplasmic domain, which resulted in the generation of constructs containing truncations of the cytoplasmic domain, and one mutation (YAYA) consisted of amino acid substitutions in which tyrosine residues at positions 24 and 26 were replaced with alanine (Fig.2). These aromatic residues were previously shown to be important for the internalization of the receptor via clathrin-coated pits (4, 20, 27). Stably transfected Caco-2 clonal cell lines were generated for each of the various constructs shown in Fig. 2, and the level of receptor expression was determined by quantitative Western blotting (10) for each cell line. Figure3 shows a representative Western blot of the wild-type and mutant bovine IGF-II/MPRs expressed in stably transfected Caco-2 cells. Differences in mobility of the various truncated receptors were readily discernable, with decreasing apparent molecular weight correlating with the increasing number of residues missing from the cytoplasmic domain. The absence of a band in the “nontransfected” lane confirms the specificity of the monoclonal antibody 86f7 for the bovine, but not the endogenous human, IGF-II/MPR (see Fig. 1).
Cell surface expression of the IGF-II/MPRs.
The steady-state cell surface distribution of the recombinant IGF-II/MPRs was measured. The percentage of each construct on the cell surface (representing apical plus basolateral surfaces) was determined by incubating the stably transfected Caco-2 cells with iodinated 86f7 monoclonal antibody in the absence or presence of saponin. In the absence of saponin, only the receptors on the cell surface will be recognized in the intact cells, whereas in the presence of saponin, which partially permeabilizes membranes, the total receptor population is available for antibody recognition. Table1 shows that the percentage of the Leu75, Arg124, and YAYA receptors on the cell surface is similar to that of the wild-type receptor. In contrast, Val8 and Asn21 constructs express a significantly higher percentage (62 and 47%, respectively) of receptor on the cell surface at steady state than does the wild-type receptor, which is consistent with the loss of the endocytosis motif (Y24KYSKV29) in these constructs (see Fig. 2).
Polarized surface distribution and internalization.
To determine whether the mutant receptors maintained their polarized distribution, the iodinated 86f7 monoclonal antibody was added to either the apical or basolateral medium of Caco-2 cells grown on filter inserts, and incubations were carried out at 37°C to measure the ability of the receptor to undergo endocytosis from the cell surface followed by a 4°C incubation to measure surface expression of the receptor. The results demonstrate that the recombinant wild-type bovine receptor exhibits a similar polarized surface distribution and internalization to the endogenous IGF-II/MPR (Table 1). Leu75, Arg124, and YAYA constructs are similar to the wild-type receptor in their basolateral enrichment on cell surface. In addition, Leu75 and Arg124constructs internalize antibody from the basolateral surface to a similar extent as the wild-type receptor. In contrast, the YAYA construct exhibits a decreased level of internalization from the basolateral surface, consistent with alteration of the endocytosis signal in this mutant (see Fig. 2). Additional truncations of the cytoplasmic domain (Val8 and Asn21) result in a dramatic change in the distribution of the receptor on the cell surface compared with that of the wild-type receptor: both Val8 and Asn21 constructs are expressed predominantly at the apical cell surface, and limited internalization is observed (Table 1), which is again consistent with the loss of the endocytosis signal in these mutants (Fig. 2). These results were confirmed using the Man-6-P-containing ligand, β-glucuronidase: Val8 and Asn21 constructs differed dramatically from the wild-type receptor in that little [125I]β-glucuronidase was internalized from the basolateral surface (data not shown).
Biochemical studies were confirmed using confocal microscopy (Figs.4 and 5). Results show that the recombinant wild-type bovine receptor exhibits a similar surface distribution the endogenous human IGF-II/MPR (Fig.4 b, e, and h compared witha, d, and g) and show that the substitution of Tyr24 and Tyr26 with alanine residues (YAYA construct, Fig. 4 c, f, andi) does not significantly alter the steady-state surface distribution of the receptor. Similarly, the removal of up to 89 residues of the COOH terminus of the receptor does not significantly alter the steady-state surface distribution of the receptor (Fig. 5). In contrast, Val8 and Asn21 constructs are detected only on the apical surface (Fig. 5). These results demonstrate that residues 21–74 of the cytoplasmic region of the IGF-II/MPR contain the basolateral targeting sequence.
Although analysis of the YAYA construct revealed that efficient basolateral targeting of the IGF-II/MPR does not depend on the Tyr residues at positions 24 and 26 contained within the motif (Y24KYSKV29) known to be important for rapid internalization, it is possible that the basolateral targeting sequence partially overlaps with the internalization motif observed for other proteins (28, 39, 41). To test this hypothesis, two additional truncation mutants, Asp36 and Ala51, were generated (Fig. 2), and stably transfected Caco-2 clonal cell lines expressing these constructs were isolated (Fig. 3). Analysis of the Caco-2 cells grown on filter inserts using the iodinated 86f7 monoclonal antibody revealed that the Ala51 construct was similar to that of the wild-type receptor in the percent found on the cell surface, its enrichment on the basolateral surface, and in its ability to internalize the antibody from the basolateral surface (Table 1). In contrast, the Asp36 construct was enriched on the apical surface to a similar extent as the Val8 and Asn21 constructs (Table 1). However, unlike the Val8 and Asn21constructs, the total amount of the Asp36 construct expressed on the cell surface at steady state was only slightly elevated compared with that of the wild-type receptor. In addition, the Asp36 construct found at the basolateral surface was able to internalize the iodinated antibody (Table 1). These results are consistent with the Asp36 construct being able to undergo endocytosis from the basolateral cell surface due to the presence of the internalization motif located at positions 24–29. Confocal microscopy, which showed the predominantly apical expression of the Asp36 construct in contrast to the basolateral enrichment of the Ala51 construct, confirmed the biochemical studies (Fig. 6). Together, these results demonstrate that the sequence D36ENETEWLMEEIQPP50 located within the cytoplasmic region of the IGF-II/MPR contains residue(s) critical for the basolateral surface expression of the receptor.
IGF-II/MPR mediates the lysosomal targeting of Man-6-P-containing soluble acid hydrolases and IGF-II. Although much information is available concerning the sorting pathways traversed by the IGF-II/MPR in the targeting of lysosomal enzymes to the lysosome in nonpolarized cells (34) and in the uptake of IGF-II, which has been shown to function as an autocrine growth factor in intestinal epithelial cells (49, 50, 52), very little is known about the trafficking of the receptor and its ligands in polarized epithelial cells. Our recent studies have shown that the IGF-II/MPR exhibits a polarized plasma membrane distribution with a threefold enrichment on the basolateral surface of Caco-2 cells (9). In addition, we have found that the IGF-II/MPRs expressed on the two cell surfaces are functionally distinct: unlike the receptor on basolateral membranes, the IGF-II/MPR on the apical surface cannot endocytose lysosomal enzymes (9). Furthermore, we have shown that the secretion-recapture pathway in which the IGF-II/MPR internalizes secreted lysosomal enzymes from the basolateral surface plays a critical role in establishing the steady-state polarized distribution of phosphorylated lysosomal enzymes (enrichment of secretion into the apical medium) in intestinal epithelial cells (47). To begin to understand the mechanism underlying this functional difference, we have investigated the sorting of the IGF-II/MPR in polarized Caco-2 cells by expressing mutant forms of the receptor in stably transfected cells and monitoring their steady-state surface distribution and cell surface internalization. Our observation that the bovine IGF-II/MPR when expressed in human Caco-2 cells exhibits a similar trafficking pattern to the endogenous receptor, coupled with our ability to distinguish the recombinant MPRs from the endogenous receptor, has made this approach a viable system to evaluate the effects that a specific region of the IGF-II/MPR has on its polarized distribution in intestinal epithelial cells.
Numerous studies have illustrated that sorting of membrane proteins to the basolateral plasma membrane is determined by the presence of specific residues in the cytoplasmic domain that serve as sorting signals (12, 13, 16, 41, 42). These cytoplasmic basolateral targeting signals are typically found to be dominant over apical sorting determinants. Evidence for this conclusion comes from the observation that mutant forms of basolateral membrane proteins lacking their cytosolic signals are sorted predominantly to the apical surface (19, 32, 39). We have observed a similar finding with the IGF-II/MPR: deletion of all but 7 (Val8construct), 20 (Asn21 construct), or 35 (Asp36construct) residues of the receptor's 163 residue cytosolic region resulted in a redistribution from the basolateral surface to nearly exclusive expression on the apical surface (Table 1 and Figs. 5 and 6). A comparison of the Asp36, Ala51, Leu75, and Arg124 constructs revealed that the removal of 128, but not 113, 89, or 40, amino acids from the COOH terminus of the receptor resulted in the loss of basolateral expression and the predominant polarized distribution of the receptor at the apical cell surface (Table 1 and Figs. 5 and 6). These results demonstrate that amino acids contained within the region of the cytoplasmic domain encompassing residues 36–50 are essential for entry of the IGF-II/MPR into the basolateral sorting pathway. The cytosolic tail of the IGF-II/MPR has been shown to be phosphorylated at Ser85 and Ser156 (30) and contains a conserved casein kinase II site followed by a dileucine motif (D154DSDEDLLHV163) at the COOH terminus as well as an aromatic-based sequence (Y24KYSKV29) located 24–29 residues from the transmembrane region that has been shown to be important for rapid internalization of the receptor (27) (see Fig. 2). Analysis of a mutant in which the tyrosine residues at positions 24 and 26 were replaced with alanine residues (YAYA construct) revealed no significant alteration in the steady-state polarized surface distribution of the receptor (Table 1 and Fig. 4). These results demonstrate that the IGF-II/MPR basolateral sorting signal differs from the class of known basolateral sorting signals that are tyrosine based, because Tyr24 and Tyr26 are the sole tyrosine residues in the cytosolic region of the bovine IGF-II/MPR (25). In addition, our preliminary data (D. A. Wick and N. M. Dahms, unpublished data) show that Val29, but not Asn30, is critical for internalization of the receptor in Caco-2 cells, which is in agreement with previous studies (4). Analysis of the Asp36 mutant revealed that Val29 and N30 are not involved in the basolateral targeting signal (Table 1 and Fig. 6). Together, IGF-II/MPR contains a basolateral sorting signal that is not colinear with its phosphorylation sites, Tyr-based internalization signal, or dileucine motif known to be required for lysosomal enzyme targeting via endosomal compartments (5, 6, 21, 27), and thus IGF-II/MPR is a member of a growing number of proteins (2, 16, 24, 35, 37) that contain basolateral signals that differ from Tyr- and Leu-based internalization motifs.
The sequence D36ENETEWLMEEIQPP50 located within the cytoplasmic domain of the IGF-II/MPR, which this study has identified as being critical for the basolateral surface expression of the receptor, is highly conserved; a comparison of the bovine, human, rat, mouse, and chicken sequences reveals changes only atpositions 36 (Asn in chicken), 48 (Ala in chicken), and 49 (not conserved). On analysis of basolateral targeting sequences found in other proteins, this 15-residue region of the IGF-II/MPR bears a striking similarity to the basolateral targeting motif recently identified for the transmembrane growth factor, stem cell factor (SCF) (46). Within the 36-residue cytoplasmic domain of SCF, a single leucine residue is required for basolateral targeting of SCF, and the presence of an acidic cluster NH2terminus to the leucine residue enhances the efficiency of basolateral sorting [K1KKQSSLTRAVENIQINEEDNEISMLQQKEREFQEV36(critical residues underlined)]. Unlike the invariant chain that uses Met-Leu as a dihydrophobic basolateral sorting signal (36), substitution of the methionine residue adjacent to the critical leucine residue did not affect basolateral targeting of SCF (46). Thus the SCF basolateral targeting signal represents a novel motif in that a monomeric, rather than a dimeric, leucine residue mediates basolateral sorting. Additional studies will be required to determine whether the IGF-II/MPR utilizes a basolateral targeting motif similar to that of SCF or the invariant chain.
A recent study by Distel et al. (12) has shown that the cation-dependent MPR (CD-MPR) contains a basolateral sorting determinant in its cytoplasmic region that differs from the signals known to be involved in its internalization from the cell surface and in its sorting of lysosomal enzymes to the lysosome. The authors identified the 12 amino acids (QRLVVGAKGMEQ) juxtaposed to the transmembrane region as being essential for basolateral targeting of the CD-MPR in MDCK cells. The authors hypothesized that the RXXV sequence within this region of the CD-MPR is of particular importance, because a similar motif has been shown to be important for basolateral delivery of a truncated version of the poly-Ig receptor (1). In addition, the authors state that a similar motif is present near the transmembrane region of the IGF-II/MPR (REMV located 5–8 residues from the transmembrane region). Our results clearly demonstrate that the basolateral sorting signal of the IGF-II/MPR in Caco-2 cells does not reside in the 35 residues adjacent to the transmembrane region of the receptor and thus the IGF-II/MPR utilizes a different basolateral sorting signal than the CD-MPR. However, it will be necessary to determine whether the CD-MPR utilizes the same sorting signal in Caco-2 cells as in MDCK cells, because it has been reported that sorting signals can be interpreted differently by different polarized cells (42).
In summary, we have shown that the IGF-II/MPR utilizes a structural determinant in its cytoplasmic region for basolateral delivery that differs from that used for rapid internalization from the cell surface and for sorting of lysosomal enzymes in endosomal compartments. We have shown that this basolateral sorting signal is located 36–50 residues from the transmembrane region and contains a dihydrophobic sequence (Leu-Met) within a cluster of acidic residues. However, additional studies will be required to identify the specific residue(s) in this region that are essential for entry of the IGF-II/MPR into the basolateral sorting pathway. In nonpolarized cells, it has been demonstrated that the cytoplasmic region of the IGF-II/MPR is recognized by the clathrin adaptor molecules AP-1 and AP-2 (15, 31) as well as by the newly identified protein TIP47 (11, 38) that function to facilitate the intracellular trafficking of the receptor. Future studies will be directed toward identifying the components of the cellular machinery of polarized epithelial cells that mediate the basolateral sorting of the IGF-II/MPR.
We thank Dr. Stuart Kornfeld and Dr. Peter Lobel for their generous gift of mutant IGF-II/MPR plasmids and Dr. Donald Messner for providing the bovine IGF-II/MPR-specific monoclonal antibodies.
First published October 3, 2001;10.1152/ajpgi.00028.2001
This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases DK-44200. This work was done during the tenure of an Established Investigatorship from the American Heart Association (to N. M. Dahms).
Address for reprint requests and other correspondence: N. M. Dahms, Medical College of Wisconsin, Dept. of Biochemistry, 8701 Watertown Plank Rd., Milwaukee, WI 53226 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2002 the American Physiological Society