Function and properties of chimeric MPR 46-MPR 300 mannose 6-phosphate receptors.

The two known mannose 6-phosphate receptors (MPR 46 and MPR 300) mediate the transport of mannose 6-phosphate-containing lysosomal proteins to lysosomes. Endocytosis of extracellular mannose 6-phosphate ligands can only be mediated by MPR 300. Neither type of MPR appears to be sufficient for targetting the full complement of lysosomal enzymes to lysosomes. The complements of lysosomal enzymes transported by either of the two receptors are distinct but largely overlapping. Chimeric receptors were constructed in which the transmembrane and cytoplasmic domains of the two receptors were systematically exchanged. After expression of the chimeric receptors in cells lacking endogenous MPRs the binding of ligands, the subcellular distribution and the sorting efficiency for lysosomal enzymes were analyzed. All chimeras were functional, and their subcellular distribution was similar to that of wild type MPRs. The ability to endocytose lysosomal enzymes was restricted to receptors with the lumenal domain of MPR 300. The efficiency to sort lysosomal enzymes correlated with the lumenal and cytoplasmic domains of MPR 300. In contrast to the wild type receptors, a significant fraction of most of the chimeric receptors was misrouted to lysosomes, indicating that the signals determining the routing of MPRs have been fitted for the parent receptor polypeptides.

distribution of the two MPRs between TGN and endosomes seems to be identical but dependent on the cell type (4 -8).
MPR 300 can bind mannose 6-phosphate (Man6P)-containing ligands at the cell surface and endocytose extracellular Man6P-containing ligands. In addition, MPR 300 binds and internalizes the insulin-like growth factor II, which does not contain Man6P residues. This property appears to be important for controlling the serum level of insulin-like growth factor II (12). The insulin-like growth factor II binding region has been localized to the N-terminal portion of the repeating unit 11 (13,14). MPR 46 does not bind Man6P ligands at the cell surface and does not internalize extracellular Man6P-containing ligands (15)(16)(17). Ligand binding to MPR 46 is improved by the presence of divalent cations (2). Different pH optima for ligand binding have been reported: pH 6.0 -7.4 for MPR 300 (18 -21) and pH 6.0 -6.3 for MPR 46 (21,22). In other studies similar pH optima from pH 5.8 to 7 have been observed for both receptors (23). In both MPRs the ligand binding domain resides in the lumenal, extracellular portion of the receptor (24,25). In MPR 300 ligands bind to the repeating units 3 and 9 (26). Arginine residues appear to be involved in the ligand binding (27,28).
The signals in the MPRs needed for endocytosis from the cell surface and efficient incorporation into clathrin-coated pits in the TGN reside in the cytoplasmic tails of the receptors. The signal for the rapid internalization of MPR 300 at the plasma membrane has been localized to the pentapeptide KYSKV (residues 24 -29 in its 163-residue cytoplasmic tail) (29 -31). MPR 46 appears to contain in its cytoplasmic tail of 67 residues several distinct signals for rapid internalization (32,33) and binding to clathrin adaptors (34). Both receptors have a leucine-based sorting signal close to their C terminus, which is critical for their ability to sort lysosomal enzymes (32, 35 -37).
The question of why two MPRs are present in mammalian cells remains unanswered. Analysis of cell lines deficient in MPR were obtained from mice with a targeted disruption of the respective MPR gene (23, 38 -40) or from tumors (41). Analysis of these cell lines has shown that each of the two MPRs contributes to the targeting of newly synthesized lysosomal enzymes and that the complement of lysosomal enzymes that are transported by either receptor are distinct but largely overlapping. However, neither type of MPR appears to be sufficient for targeting of lysosomal enzymes to lysosomes. This has been demonstrated by expression of either MPR 46 or MPR 300 in cells that lack endogenous MPRs (42). In the receptor-deficient cells the bulk of lysosomal enzymes (Ͼ85%) are missorted into the secretions. Overexpression of MPR 46 to levels up to five times higher than the wild type level corrects missorting of only two-thirds of the newly synthesized lysosomal proteins. A 2-fold overexpression of MPR 300 is sufficient to fully correct the missorting of lysosomal enzymes. However, a substantial fraction of the lysosomal enzymes are targetted to lysosomes along a secretion recapture pathway, which depends on MPR 300-mediated endocytosis and is of minor importance in cells that express both types of MPR.
To analyze whether the transmembrane and/or cytoplasmic domain of one MPR can substitute for the respective domains in the other MPR, chimeric receptors were constructed in which the transmembrane and cytoplasmic domains of the two receptors were exchanged systematically. After expression in cells that are deficient in both MPRs, the ability of chimeric receptors to bind ligands, their subcellular distribution, and their sorting efficiency for lysosomal ligands was compared with that of wild type MPRs. All chimeras were functional and displayed a subcellular distribution pattern similar to the wild type MPRs. The ability to endocytose ligands depended on the presence of the lumenal domain of MPR 300. Furthermore, the sorting efficiency of receptors carrying the lumenal domain and/or the cytoplasmic domain of MPR 300 was markedly better than that of other receptors.

Constructs-
The chimeras were named corresponding to the origin of the lumenal, transmembrane, and cytoplasmic domain of either MPR 300 (L) or MPR 46 (S) (see Fig. 1). Chimeras were constructed from the human cDNAs by PCR using the overlap extension method (43,44). Fragments of the genes that are to be recombined are generated in separate PCR reactions (primers 1 and 2 and primers 3 and 4). The primers are designed so that the ends of the products contain complementary sequences. When the PCR products are mixed, denatured, and reannealed, the strands having the matching sequences at their 3Ј ends overlap and act as primers for each other. Extension of this overlap by DNA polymerase produces a molecule in which the original sequences are fused together. The recombinant product is PCR-amplified in the presence of primers for its 5Ј and 3Ј ends (primers 1 and 4). For expression all cDNAs were subcloned to the pMPSVHE or pMPSVEH vector (45). For subcloning into MPR 300, an MluI restriction site was introduced behind the stop codon at position 7626. Chimeras containing the lumenal MPR 300 portion were subcloned into a pMPSV-derived vector containing MPR 300 cDNA by an internal NdeI site at position 6724 and the recombinant MluI site. Chimeras containing the lumenal MPR 46 portion were subcloned into a pMPSV-derived vector containing MPR 46 cDNA using an internal BglII site at position 545 and a multicloning EcoRI site at the 3Ј end. All constructs were sequenced in the regions encompassing mutations using the ABI 373A system. The sequences of the primers used in this study are available on request.
Cells and Cell Culture-Mouse embryonic fibroblasts (MEF) were grown in Dulbecco's minimal essential medium and 10% fetal calf serum. Baby hamster kidney cells (BHK21) were grown in the same medium supplemented with 5% fetal calf serum. Transfection of BHK21 was performed as reported (24), and transfection of MEF deficient in MPR 46 and MPR 300 (mpr Ϫ MEF) was as described (42).
Expression Levels-For determination of relative amounts of human MPR 46 or MPR 300 wild type and chimeras, the iodinated monoclonal antibodies 21D3 and 2C2 recognizing the lumenal epitopes, MPR 46 or MPR 300, respectively, were used in the presence of saponin (46). For the determination of surface-associated MPRs, the incubation with the monoclonal antibodies was performed in the absence of saponin. Expression levels of MPR chimeras were determined as described (42).
The amount of MPR 46 in control MEF and mpr Ϫ /MPR 46 clone II and IV was determined by Western blot analysis using a tail-specific antiserum against MPR 46. Correlation of the binding of the tailspecific antiserum with that of [ 125 I]21D3 showed that binding of 13400 cpm[ 125 I]21D3/mg of cell protein is equivalent to the endogenous level of MPR 46 in control MEF. For clones II-IV a high correlation between the binding of the polyclonal antiserum and the monoclonal 21D3 antibody was observed (r ϭ 0.99).
The amount of MPR 300 was calculated from the radioactivity incorporated into MPR 300 during a 16-h metabolic labeling in the presence of [ 35 S]methionine referred to that incorporated into MPR 300 by control MEF. r ϭ 0.98 for the binding of monoclonal 2C2 antibody and the incorporation of [ 35 S]methionine for clones A-C.
Lysosomal Enzyme Assays-Lysosomal acid phosphatase was measured as described using p-nitrophenylphosphate as substrate (47). Other lysosomal enzymes were detected using fluorometric assays as described (48).
Metabolic Labeling and Immunoprecipitation-Confluent dishes (35 mm) of MEF or BHK cells were incubated in methionine-free medium for 1 h and then labeled for the indicated times with 1.85 MBq [ 35 S]methionine (Amersham Pharmacia Biotech) in minimum essential medium containing 4% dialyzed fetal calf serum. For the chase the media were supplemented with 0.25 mg/ml methionine. Immunoprecipitation was performed from cells and media as described (49). ␤-Glucuronidase was immunoprecipitated with antiserum specific for mouse ␤-glucuronidase that was kindly provided by R. Swank (Roswell Cancer Institute, Buffalo, NY). The human MPR 300 was immunoprecipitated with antiserum specific for the human MPR 300 (50). For MPR 46 immunoprecipitation antiserum specific for the human MPR 46 (51) was used. All immunoprecipitates were subjected to SDS-PAGE and fluorography. Radioactivity incorporated into the polypeptides was quantified by densitometry using Hewlett-Packard Scann Jet 4c/T (Palo Alto, CA) and the program WinCam 2.2 or the phosphoimager (FuJix Bas 1000) and programs macMAS and image reader.
Stability of MPR and MPR Chimeras-For the analysis of the stability of the receptor polypeptides, cells expressing the respective MPRs were metabolically labeled with [ 35 S]methionine. For wt MPR 300 and chimeras a pulse of 1 or 2 h was followed by a chase of 8 h, and for MPR 46 and chimeras a pulse of 6 h was followed by a chase of 16 h. The chase was in the presence or absence of the protease inhibitors leupeptin and pepstatin, 0.1 mM each. Immunoprecipitation was done as described above.
Western Blot Analysis-Samples were prepared and subjected to Western blot analysis as described (51). For cathepsin D 50 g of total cellular protein were subjected to Western blot analysis, and the blot was probed with antiserum specific for cathepsin D. The blots were probed using an ECL light-based immunodetection system (Amersham Pharmacia Biotech).

Endocytosis of [ 35 S]Arylsulfatase A-Fibroblasts
(mpr Ϫ MEF) that were transfected with human ASA (52) were grown to confluency on 10-cm dishes and labeled overnight with 5.5 MBq [ 35 S]methionine. The medium was precipitated with 50% (w/v) ammoniumsulfate, dissolved in Tris-buffered saline and dialyzed overnight against Tris-buffered saline. The [ 35 S]ASA was purified using a column to which an ASAspecific antibody (20B1) had been coupled (53). An aliqout of each fraction was counted for radioactivity, and the ASA activity was measured (54). Fractions containing active ASA were pooled. [ 35 S]Arylsulfatase A was used as a tracer of MPR-dependent endocytosis as described (55). Confluent dishes (35 mm) were washed twice with phosphatebuffered saline and then incubated for 4 h at 37°C with 1 ml of Dulbecco's minimal essential medium supplemented with 20000 cpm of [ 35 S]arylsulfatase A plus 40 milliunits of unlabeled ASA, 10% fetal calf serum in the presence and absence of 5 mM mannose 6-phosphate. The cells were washed five times with ice-cold Hanks' buffer, trypsinized, and centrifuged. The pellet was dissolved in Tris-buffered saline/0.05% Triton X-100 and sonicated. Media and cells were counted for radioactivity, and the [ 35 S]arylsulfatase A was immunoprecipitated (56) from cells for control.

Construction and Expression of MPR 46/MPR 300
Chimeras-MPR 300 and MPR 46 are type I transmembrane proteins. For the construction of the receptor chimeras the lumenal, transmembrane, and cytoplasmic domains of the two receptors were systematically exchanged yielding a total of six chimeras (Fig. 1). The domains were designated with L (large) or S (small) according to their origin from MPR 300 or MPR 46, respectively. In the three-letter code for the chimeras the first letter designates the lumenal, the second the transmembrane, and the third the cytoplasmic domain.
The chimeric receptors were expressed in embryonic mouse fibroblasts that lack endogenous MPR 46 and MPR 300 (mpr Ϫ MEF) (23). Stably expressing clones were selected, and the level of the chimeras was quantified by measuring the binding of 125 I-labeled antibodies directed against the lumenal domain of MPR 46 (monoclonal antibody 21D3) or of MPR 300 (monoclonal antibody 2C2) to permeabilized cells. The level of the chimeric receptors containing the lumenal domain of MPR 46 (MPR 46 chimeras SSL, SLL, and SLS) was 0.8 -2.2-fold that of endogenous MPR 46 in mouse fibroblasts (Table I). For chimeras containing the lumenal domain of MPR 300 (MPR 300 chimeras LLS, LSS, and LSL) the expression level was 1-4-fold that of endogenous MPR 300 (Table II). Clones expressing similar levels of wt MPR 46 (SSS) or wt MPR 300 (LLL) were used for comparison (Tables I and II).
Localization of the Receptor Chimeras-The fraction of receptors accessible to the antibodies at the cell surface varied between 8 and 19% for wt MPR 46 and MPR 46 chimeras (Table  I) and 10 and 22% for MPR 300 and MPR 300 chimeras (Table  II). Thus the fraction of receptor present at the cell surface is not or only little affected by the domain composition of the chimeric receptors.
Based on indirect immunofluorescence the intracellular distribution of MPR 46, MPR 300, and receptor chimeras was comparable. The bulk of receptors was concentrated in structures surrounding the nucleus. Vesicular structures containing receptors were scattered throughout the cytoplasm with a decreasing frequency toward the periphery of the cells. Morphological signs of an accumulation within the endoplasmic reticulum or within lysosomal structures (as defined by the lysosomal membranes glycoprotein LAMP 1) were not noted in any of the receptor chimeras.
Molecular Properties of the Receptor Chimeras-After metabolic labeling of cells with [ 35 S]methionine, a membrane fraction was prepared and solubilized with Triton X-100. The detergent extract was passed over a mannose 6-phosphatesubstituted affinity matrix. The receptor polypeptides were immunoprecipitated from the fractions of unbound material and material that was eluted from the affinity matrix with 5 mM mannose 6-phosphate. The apparent size of the receptor chimeras differed from that of the wild type receptors according to different sizes of the cytoplasmic domains of MPR 300 (163 amino acids) and MPR 46 (67 amino acids). Exchange of the transmembrane domains did not affect the apparent size.
The fraction of bound receptor was similar for wild type MPRs (98%) and chimeric receptors (95-100%), indicating that the folding of the lumenal domain that mediates the binding of mannose 6-phosphate was not affected by the type of the transmembrane and/or the cytoplasmic domain (Fig. 2).
For some of the chimeras carrying the lumenal domain of MPR 300, notable differences in the expression level as determined by binding of [ 125 I]2C2 and the amount of [ 35 S]methionine incorporated into the receptor polypeptides during a short term metabolic labeling were noted (not shown). This pointed to differences in the stability of the chimeric and wild type receptors.
For the determination of the stability, wild type and chimeric receptors were expressed in BHK cells. This was necessary, because of the unequal ability of the wild type and chimeric receptors to mediate targetting of lysosomal enzymes to lysosomes (see below). As a consequence the steady state levels of lysosomal enzymes including that of lysosomal proteinases depends on the receptor construct expressed in the mpr Ϫ MEF. Hence the stability of the receptors in mpr Ϫ MEF is affected by the transfer of the receptors to lysosomes and by the sorting efficiency of the receptors for lysosomal proteinases. In BHK cells, which sort lysosomal proteinases efficiently because of their endogenous MPR 46 and MPR 300, the stability of the transfected receptors depends less if at all on the sorting function of the transfected receptor.
The different cell lines were pulse-labeled for 1 or 6 h with   Quantitation of the labeled receptor polypeptides in these cells (Fig. 3) revealed that the stability of MPR 46 chimeras was greatly reduced (estimated half-life, 9 -14 h). The stability of MPR 300 chimeras depended on the type of the cytoplasmic domain. The stability of the two MPR 300 chimeras carrying the cytoplasmic tail of the small receptor was greatly reduced with apparent half-lives of 2.8 h (LLS) and 2.4 h (LSS), whereas the stability of the LSL chimera was comparable with that of wild type MPR 300. The presence of the proteinase inhibitors during the chase had a stabilizing effect on MPR 46 and MPR 300 chimeras with shortened half-lives (compare lanes 2 and 3 in Fig. 3), indicating that they are delivered to lysosomes, where they are degraded. The apparent half-lives of MPR 46 chimeras increased to 16 -29 h and that of MPR 300 chimeras increased to 3.6 -3.8 h.
Endocytic Activity of Receptor Chimeras-A major functional difference between the two MPRs is the ability of cell surface associated MPR 300 to mediate internalization of mannose 6-phosphate containing polypeptides. In spite of its expression at the cell surface and rapid cycling between internal membranes and the cell surface, MPR 46 is incapable of mediating internalization of ligands.
To follow the internalization of mannose 6-phosphate-containing polypeptides, the various cell lines were incubated for 4 h in the presence of [ 35 S]arylsulfatase A, a lysosomal enzyme known to contain mannose 6-phosphate residues. The internalized radioactivity was quantified and characterized by SDS-PAGE and fluorography. Cells expressing wt MPR 300 or MPR 300 chimeras internalized 8 -20% of the offered [ 35 S]arylsulfatase A (Table III). The cell associated radioactivity comigrated with an [ 35 S]arylsulfatase A standard. The internalization depended on the expression of MPR 300 constructs and was sensitive to inhibition by 5 mM mannose 6-phosphate. An endocytotic index was calculated by referring the amount of internalized arylsulfatase A to the amount of receptor polypeptides at the cell surface. Compared with the endocytic index of recombinant wt MPR 300, that of the receptor chimeras was 42% (LLS), 54% (LSS), and 127% (LSL) for the different MPR 300 chimeras (Table III) (24). The MPRs were immunoisolated from the flow through (1) and wash fractions (2)(3)(4)(5) and the Man6P eluate (6 -8). Shown are the fluorograms of the immunoisolated receptors and the percentage of receptor polypeptide recovered in the Man6P eluate.

Sorting of Newly Synthesized Lysosomal Enzymes by the
Receptor Chimeras-As a measure for the sorting function of the receptor chimeras, the accumulation of lysosomal enzymes in the medium was followed. When mpr Ϫ MEF are incubated for 24 h in the medium, about 90% of the total ␤-hexosaminidase activity are recovered in the medium, and only 10% are recovered in the cells. Expression of wt MPR 300 increases the fraction of ␤-hexosaminidase recovered in the cells up to 80% of total, depending on the level of MPR 300 expression (Fig. 4,  bottom panel). The ability of MPR 300 chimeras LLS, LSL, and LSS to reduce the fraction of secreted and to increase the fraction of intracellularly retained ␤-hexosaminidase was comparable with that of wt MPR 300.
The results were different for MPR 46 chimeras. Although the steady state concentration of wild type MPR 46 can reach up to the 5-fold level of that of endogenous MPR 46, the fraction of secreted ␤-hexosaminidase decreases from 90% only to about 40%. The sorting efficiency of the SLS chimera was comparable with that of wt MPR 46, whereas that of the SSL and SLL chimeras was clearly better than that of wt MPR 46 (Fig. 4, top  panel).
The sorting of a newly synthesized lysosomal enzyme was directly followed for [ 35 S] ␤-glucuronidase. The latter was immunoprecipitated from cells and medium after incubating the cells for 6 h in the presence of [ 35 S]methionine and subsequently for 16 h in chase medium. The fraction of secreted [ 35 S]␤-glucuronidase was reduced from 93% in nontransfected cells to 3-34% in cells expressing MPR 300 or MPR 300 chimeras (Table IV). In the cells expressing MPR 46 or MPR 46 chimeras the fraction of secreted [ 35 S] ␤-glucuronidase decreased only to 46 -74% (Table IV). Again the sorting efficiency of the SSL and SLL chimeras was better than that of wild type MPR 46, in spite of their lower expression level.
Analysis of cathepsin D by Western blot revealed that expression of MPR 300 or MPR 300 chimeras induced normal intracellular steady state concentrations of cathepsin D. Moreover, the relative distribution of the precursor, single chain intermediate and double chain mature form was as in control MEF (Fig. 5, bottom panel). Because proteolytic processing to the intermediate and mature forms depends on the transport of the cathepsin D precursor to endosomes/lysosomes, these data further indicate that cathepsin D is correctly targeted in cells expressing MPR 300 chimeras, and endosomes/lysosomes in these cells contain the proteinases required for the processing.
Expression of MPR 46 and MPR 46 chimeras was less efficient in correcting processing of cathepsin D (Fig. 5, top panel). We ascribe this to the insufficient proteolytic capacity (which itself depends on the MPR mediated targetting of lysosomal proteinases) rather than to a failure to deliver the intracellularly retained cathepsin D to lysosomes. This was supported by indirect immunofluorescence, which showed in these cells a colocalization of cathepsin D with the lysosomal marker LAMP 1 as in controls (not shown). Traces of mature cathepsin D were only detectable in cells expressing MPR 46 chimeras SSL and SLL (Fig. 5, top panel), pointing to their higher sorting efficiency. DISCUSSION The structural and functional homology of the two MPRs made it likely that their domains are exchangeable and that chimeric receptors would target newly synthesized lysosomal enzyme to lysosomes. The data of the present study demonstrate that indeed the chimeric receptors representing the various possible combinations between the lumenal, transmem- a Confluent dishes (35 mm) were incubated for 4 h at 37°C with 1 ml of medium containing 40 milliunits of ASA plus 6 milliunits of [ 35 S]ASA (40,000 cpm/ml) in the presence and absence of 5 mM Man6P. The medium was taken off, noninternalized surface-bound ASA was dissociated by trypsin treatment, and the amount of internalized radioactivity in the cells was determined. In the presence of 5 mM Man6P the internalization of [ 35 S]ASA was inhibited to Ͻ0.5% of offered radioactivity.
b The ratio between the internalized [ 35 S]ASA and the amount of receptor expressed at the cell surface (see Table II) was calculated to obtain a measure for the endocytotic efficiency of the surface receptors. The value obtained for wt MPR 300 (LLL) was set as 1, and the values for the MPR 300 chimeras were expressed as percentage of the former. brane, and cytoplasmic domains are functional.
A striking difference between MPR 46 and MPR 300 is the inability of the former to mediate internalization of Man6Pcontaining ligands. Only under specific conditions such as extracellular pH of 6.5, high ligand concentration, and strong overexpression of the receptor, was a low endocytic activity of MPR 46 detectable (16). The analysis of the chimeric receptors demonstrates that the ability to endocytose Man6P-containing ligands strictly depends on the presence of the lumenal domain of MPR 300. Replacing the transmembrane or cytoplasmic domain of MPR 300 by the respective domain of MPR 46 had no or only a moderate effect on the endocytic activity. Thus the transmembrane and cytoplasmic domains of the two receptors are both functional in endocytosis in spite of their poor structural homology.
The sorting efficiency of MPR 300 chimeras was comparable with that of wt MPR 300. Depending on the expression level, 80% or more of the newly synthesized lysosomal enzymes were retained in the cells. It is known from earlier studies that in cells expressing only MPR 300 part of newly synthesized lysosomal enzymes are not directly transferred from the secretory route to endosomes/lysosomes but are first secreted and than recaptured by MPR 300 mediated endocytosis (42). Because of the endocytic activity of MPR 300 chimeras, it is likely that their high sorting efficiency results in part from their ability to recapture secreted lysosomal enzymes.
The sorting efficiency of MPR 46 for lysosomal enzymes is generally lower than that of MPR 300. Even at the highest levels of expression that was achieved (about 5-fold of that of endogenous MPR 46) about 50% or more of newly synthesized ␤-hexosaminidase and ␤-glucuronidase accumulates in the secretions (42). This can be attributed in part to the inability of MPR 46 to recapture secreted lysosomal enzymes. It came therefore as a surprise that the sorting efficiency of MPR 46 chimeras carrying the cytoplasmic domain of MPR 300 (SSL and SLL) was much higher that that of wt MPR 46, in spite of the inability of these chimeras to mediate receptor-mediated endocytosis of lysosomal enyzmes.
A higher sorting efficiency can result either from an accelerated recycling between the sites, where ligands are bound (e.g. in the TGN) and where they are released (e.g. in endosomes) or from an altered intracellular routing, which reduces the delivery of ligands at a site (e.g. early endosomes), from where the ligands can escape into the extracellular space. It remains to be determined by which mechanism the cytoplasmic domain of MPR 300 improves the sorting efficiency of the respective MPR 46 chimeras.
It was not possible to resolve differences in the intracellular distribution of wt and chimeric receptors by indirect immunofluorescence. The bulk of the receptors was concentrated in the perinuclear area in vesicular, tubular, and cisternal structures. None of the receptors accumulated in the endoplasmic reticulum, which is in agreement with the failure to detect abnormally glycosylated receptors. However, we also failed to see an accumulation of receptors in lysosome-like structure, although several of the chimeras have shorter half-lives and apparently are subject to lysosomal proteolysis.
The stability of MPR 300 is much lower than that of MPR 46. Presence of the cytoplasmic and/or the transmembrane domain of MPR 46 in MPR 300 chimeras did not increase the stability, indicating that the lumenal domain of MPR 300 is a dominant determinant for the stability. This is in agreement with the earlier observation that loss of MPR 300 occurs mainly through nonlysosomal proteolysis of its lumenal domain. The latter is released into the extracellular fluid, indicating that proteolysis occurs at the cell surface or within the secretory route (57). Presence of the cytoplasmic domain of MPR 46 in MPR 300 chimeras even decreased significantly the half-time of the chimeras, and this was at least in part due to misrouting of the chimeras to lysosomes. This observation was unexpected because all sequences that control the escape of MPR 46 from the lysosomal route and thereby prevent lysosomal degradation have been localized to the cytoplasmic tail of MPR 46 (58,59). One possible explanation for the failure of MPR 46 cytoplasmic tail to stabilize the LLS and LSS chimeras could be that the quaternary structure of the chimeras is different from the quarternary structure of MPR 46 and that this impairs the efficiency of MPR 46 cytoplasmic tail to mediate escape from the route to lysosomes. In fact, MPR 46 exists as dimer and tetramer (60 -62), whereas MPR 300 exists as monomer, which dimerizes upon binding of multivalent ligands (63). Thus, MPR 300 is expected to return from endosomes to the Golgi as a monomer. The quaternary structure of MPR 300 chimeras and the efficiency of MPR 46 cytoplasmic tail to avoid misrouting to lysosomes of monomeric and oligomeric reporter molecules need to be determined.
The high stability of MPR 46 was significantly reduced by the presence of the transmembrane and/or the cytoplasmic domain of MPR 300. MPR 46 chimeras were at least partly  Tables I and II) were metabolically labeled for 6 h and then chased for 16 h. Mouse ␤-glucuronidase was immunoprecipitated from the cells and the medium and analyzed by SDS-PAGE and fluorography. ␤ Glucuronidase was recovered in the cells as a 72-kDa mature form and in the secretions as a 75-kDa precursor form. The percentage of newly synthesized ␤-glucuronidase recovered in the secretions was calculated after densitometric analysis. Control  Homogenates (50 g of protein) from cultured mouse embryonic fibroblasts were analyzed by Western blotting with a cathepsin D-specific polyclonal antiserum. Control MEF and mpr Ϫ MEF were included for control. For MPR 46 wt clone IV and chimeric MPR 46 clones II and for MPR 300 wt clone C and chimeric MPR 300 clones B were analyzed. P, I, and M indicate the positions of the precursor, immediate, and mature forms of cathepsin D, respectively. misrouted to lysosomes. As discussed above for MPR 300 chimeras, this may be related to differences in the oligomeric state of MPR 46 and MPR 300 and their chimeras. Preliminary experiments with the two MPR 46 chimeras carrying the transmembrane domain of MPR 300 (SLS and SLL) indicate that they form higher oligomers than the wild type MPR 46. This again makes it necessary to analyze the sorting efficiency of the cytoplasmic tails of MPR 46 and MPR 300 as a function of the oligomeric state of the transmembrane protein to which they are fused.
In summary, the analysis of the receptor chimeras has shown that all are functional in terms of lysosomal enzyme sorting. The cytoplasmic domain of MPR 46 was less efficient in mediating transfer of lysosomal enzymes to lysosomes than that of MPR 300, whereas the lumenal domain of MPR 300 was necessary and sufficient to mediate endocytosis of lysosomal enzymes. The inability of most of the receptor chimeras to avoid efficiently misrouting to lysosomes indicates that the efficiency of the trafficking signals residing in the cytoplasmic domain and possibly other domains has been fitted for their parent receptor polypeptide.