HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 33, 25188-25193, August 18, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/33/25188    most recent M001138200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Krise, J. P. Articles by Pfeffer, S. R. Search for Related Content PubMed PubMed Citation Articles by Krise, J. P. Articles by Pfeffer, S. R. Social Bookmarking                      What's this?

Quantitative Analysis of TIP47-Receptor Cytoplasmic Domain Interactions

IMPLICATIONS FOR ENDOSOME-TO-TRANS GOLGI NETWORK TRAFFICKING^*

Jeffrey P. Krise, Paul M. Sincock^§, Joke G. Orsel^¶, and Suzanne R. Pfeffer

From the Department of Biochemistry, Stanford University School of Medicine, Stanford, California 94305-5307

Received for publication, February 10, 2000, and in revised form, May 26, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

TIP47 (tail-interacting protein of 47 kDa) binds to the cytoplasmic domains of the cation-independent and cation-dependent mannose 6-phosphate receptors and is required for their transport from late endosomes to the trans Golgi network in vitro and in vivo. We report here a quantitative analysis of the interaction of recombinant TIP47 with mannose 6-phosphate receptor cytoplasmic domains. Recombinant TIP47 binds more tightly to the cation-independent mannose 6-phosphate receptor (K_D = 1 µM) than to the cation-dependent mannose 6-phosphate receptor (K_D = 3 µM). In addition, TIP47 fails to interact with the cytoplasmic domains of the hormone-processing enzymes, furin, phosphorylated furin, and metallocarboxypeptidase D, as well as the cytoplasmic domain of TGN38, proteins that are also transported from endosomes to the trans Golgi network. Although these proteins failed to bind TIP47, furin and TGN38 were readily recognized by the clathrin adaptor, AP-2. These data suggest that TIP47 recognizes a very select set of cargo molecules. Moreover, our data suggest unexpectedly that furin, TGN38, and carboxypeptidase D may use a distinct vesicular carrier and perhaps a distinct route for transport between endosomes and the trans Golgi network.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Mannose 6-phosphate receptors (MPRs)¹ bind to newly synthesized, soluble lysosomal enzymes and carry them from the Golgi complex to prelysosomes (1). The low pH within this compartment leads to the release of the receptor-bound hydrolases; MPRs can then return to the Golgi complex to initiate another round of lysosomal enzyme delivery. Two MPRs have been described to date (2). One is a 300-kDa transmembrane glycoprotein that also contains a high affinity binding site for insulin-like growth factor II. This receptor is known as the cation-independent MPR (CI-MPR) because it does not require divalent cations for ligand binding in vitro. A second, smaller, 46-kDa MPR requires divalent cations for ligand binding and is referred to as the cation-dependent MPR (CD-MPR). Both receptors share common pathways of intracellular transport and bind to an overlapping set of lysosomal enzymes (2).

We study the recycling of MPRs from endosomes back to the Golgi complex. We have established a cell-free system that reconstitutes this process in vitro (3, 4) and have identified a number of proteins that are required for MPR recycling. The Rab9 GTPase (5), its effector p40 (6), the NSF ATPase, its adaptor -SNAP (7), microtubules, cytoplasmic dynein, and a protein named mapmodulin (7) are all required for the delivery of MPRs from endosomes to the Golgi complex.

We recently discovered a protein named TIP47 (tail-interacting protein of 47 kDa) that binds to the cytoplasmic domains of the CI- and CD-MPRs and is required for their recycling to the Golgi complex both in vitro and in vivo (8). TIP47 recognizes a Phe-Trp motif in the CD-MPR (8) that is required for retrieval of this receptor from endosomes and diversion from lysosomes (9). It is not yet known what TIP47 recognizes in the CI-MPR, which does not contain a related, Phe-Trp motif. Nevertheless, TIP47 does not recognize tyrosine or dileucine motifs in either the CD- or CI-MPRs (8), a feature that distinguishes it from clathrin AP-1 and AP-2 adaptor complexes (10).

In addition to MPRs, other proteins also cycle between endosomes and the Golgi complex. For example, TGN38 is a ~90-kDa transmembrane glycoprotein of unknown function that is localized to the trans Golgi network (11). Its 33-amino acid cytoplasmic domain contains the sequence, YQRL that is essential for TGN localization (12-15). Although TGN38 is found predominantly in the TGN at steady state, the protein cycles continuously between the cell surface, endosomes, and the TGN (12, 16). Similarly, furin (17) and metallocarboxypeptidase D (CPD) (18) are TGN-localized, membrane-bound enzymes that function in the post-translational processing of secreted proteins. Like TGN38, furin and CPD also appear to cycle with some frequency between the cell surface, endosomes, and the TGN (17-21).

Since TIP47 recognizes MPRs and is required for their retrieval from endosomes, we think of TIP47 as a cargo selection device for endosome-to-Golgi transport. It was therefore of interest to determine if TIP47 interacts with other proteins that traverse similar intracellular transport pathways. We report here that TIP47 fails to interact with the cytoplasmic domains of TGN38, furin, phosphorylated furin, or CPD. These observations suggest that these proteins may utilize a distinct type of transport vesicle for their transport from endosomes to the Golgi complex.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

His-tagged TIP47 was purified after expression in Eschericha coli as described (8). Recombinant cytoplasmic domains (GST-CI-MPR, GST-CI-MPR Y24A/Y26A/V29A, GST-CD-MPR, GST-LDL receptor, GST-furin (gift of Gary Thomas, Vollum Institute, Portland, OR), GST-carboxypeptidase (gift of Oleg Varlamov, Albert Einstein University, New York), and GST-HIV-Nef (gift of Didier Trono, University of Geneva)) were prepared and purified as described (8) followed by S100 gel filtration chromatography. GST-TGN-38 (gift of Sharon Milgram, University of North Carolina) was purified as follows. Bacteria were lysed by incubation with lysozyme and 0.1% Triton X-100 (in HKE buffer (50 mM Hepes, pH 7.0, 100 mM KCl, 2 mM EDTA)) (instead of a French press), and the cleared lysate (100 ml) was loaded onto a 5-ml, SP-Sepharose cation exchange resin. The resin was washed with 50 ml of HKE buffer without Triton X-100 and eluted with 15 ml of 50 mM Hepes, pH 8.0, 1 M KCl, 2 mM EDTA, and 0.01% Triton X-100. The eluate was applied onto 500 µl of glutathione-Sepharose with recycling, and the column was washed with 50 ml of HKE, pH 8.0. GST-TGN38 was then eluted by resuspension in 2 × 1.25 ml of elution buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, and 10 mM reduced glutathione) plus 0.01% Triton X-100. Glutathione was removed by passage over a P6DG column (Bio-Rad) equilibrated in 25 mM Hepes, pH 7.5, 50 mM KCl, 1 mM MgCl₂, 0.01% Triton X-100. All protein preparations contained a small amount of free GST; GST background was minimal and subtracted.

In Vitro Phosphorylation of GST-Furin-- Phosphorylation of GST-furin was achieved by incubating GST-furin (20 µM), ATP (200 µM), and casein kinase II (250 units/ml, New England Biolabs, Beverly, MA) in 20 mM Tris-HCl, 50 mM KCl, and 10 mM MgCl₂. The reaction was incubated at 30 °C for 1 h, after which glutathione-Sepharose was added to capture the GST-furin. With small scale reactions, quantitation of [-³²P]ATP incorporation suggested incorporation of 2 mol of phosphate/mol of cytoplasmic domain; under identical conditions, GST was not a substrate for the kinase.

Texas Red Fluorescent Labeling of TIP47-- Texas Red^® sulfonyl chloride (Molecular Probes, Inc., Eugene, OR; 60 µl at 10 mg/ml in anhydrous dimethyl formamide), was combined with 2 ml of ice-cold TIP47 (3 mg/ml in 25 mM Hepes, pH 7.5). The reaction was mixed end-over-end for 45 min at 4 °C. Labeled TIP47 was subsequently purified on a Sepharose S-100 size exclusion column to remove unincorporated dye. Fractions containing labeled TIP47 were pooled and analyzed for protein content; Texas Red^® content was determined by absorbance at 596 nm. The molar ratio of Texas Red^® to TIP47 was calculated to be 1.9.

Determination of Active Molecules and Calculation Assumptions-- All of our calculations assume that His-tagged TIP47 has a monomeric molecular mass of 52 kDa. However, recombinant TIP47 chromatographs as a somewhat heterodisperse oligomer (8); thus, all of our K_D determinations may be underestimates of the actual binding affinity. TIP47 protein binding capacity is highly dependent upon expression and storage conditions. Most active preparations were obtained from cells induced to express the protein at low cell densities (A₆₀₀ = 0.1) and contained glycerol during the purification and for storage. We estimate that TIP47 preparations contain 10% active molecules as determined by the fraction that were capable of binding to saturating amounts of the CI-MPR cytoplasmic domain. It is possible that the tendency of recombinant TIP47 to oligomerize is responsible for the relatively low recovery of active molecules. Nevertheless, we observe the same selectivity for MPR wild type and mutant forms when partially purified, native TIP47 is compared with the recombinant protein studied here. Finally, Texas Red labeling does not influence the activity of our preparation; control immunoblot experiments using unlabeled recombinant TIP47 and assuming 10% active molecules yielded identical K_D values.

Fluorescent TIP47 Binding-- Binding was determined in 25 mM Hepes, pH 7.5, 150 mM KCl, 1 mM MgCl₂, and 1 mg/ml bovine serum albumin in a 500-µl total volume. Receptor cytoplasmic domain GST fusion proteins (0-15 µM) were mixed with TIP47 (50 nM active protein) and incubated for 1 h at room temperature followed by the addition of 50 µl of glutathione-Sepharose (1:1 slurry) for an additional 45 min. The slurry was subsequently passed through an empty 1-ml column fitted with a porous support and washed rapidly with 1 ml of binding buffer to remove unbound TIP47. GST fusion proteins were eluted from glutathione-Sepharose by a 15-min incubation in 500 µl of elution buffer. The eluate was analyzed for fluorescent labeled TIP47 using _ex = 596 nm and _em = 611 nm with an AMINCO Bowman series 2 Luminescence spectrometer (Urbana, IL).

Cytosolic TIP47 Binding-- Receptor cytoplasmic domain-GST fusions (3.6 nmol) were mixed end-over-end with 950 µl of K562 cytosol (8.4 mg/ml) in 0.01% Triton X-100 plus 1 mM phenylmethylsulfonyl fluoride, 10 µg/ml leupeptin, 40 µg/ml aprotinin, 1 µM pepstatin A for 90 min. Subsequently, 100 µl of glutathione-Sepharose beads (1:1 slurry) were added for another 60 min. Beads were pelleted at low speed and washed five times with 1 ml of HKE buffer (pH 8.0). Bound proteins were eluted by a 15-min incubation of the beads in 100 µl of elution buffer containing 50 mM reduced glutathione. The elution was repeated, eluates were combined, and 40 µl was analyzed by immunoblotting for TIP47.

AP-1 and AP-2 Binding-- Clathrin adaptors AP-1 and AP-2 were purified from bovine brain clathrin-coated vesicles as described (22) and purified to homogeneity using hydroxyapatite (23). AP-1 and AP-2 concentrations were determined as follows. Known amounts of a marker protein were compared by SDS-polyacrylamide gel electrophoresis with various amounts of AP-1 and AP-2; the amount of the AP-1 and AP-2 µ-chains were quantified by densitometric scanning relative to the standard. Receptor cytoplasmic domain-GST fusion constructs (2.0 µM) were incubated with AP-2 (50 nM) in 25 mM HEPES, pH 7.5, 50 mM KCl, 1 mM MgCl₂, 1 mg/ml bovine serum albumin, and 0.01% Triton X-100. After incubation for 1 h at room temperature, glutathione-Sepharose beads (30 µl of a 50% slurry) were added for 1 h. Beads were then pelleted at low speed and washed three times in HKE buffer, pH 8.0. Bound proteins were eluted by incubation with elution buffer for 15 min and separated by 12% SDS-polyacrylamide gel electrophoresis. After transfer to nitrocellulose, AP-1 and AP-2 were detected by Western blotting analysis using the anti--adaptin antibody (100:3) or anti--adaptin monoclonal antibody (100:2) (Sigma).

Other Methods-- Protein was determined using Bio-Rad reagent and bovine serum albumin as a standard.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We established a fluorescence-based assay to examine the interaction of recombinant TIP47 with receptor cytoplasmic domains. For this purpose, we attached Texas Red to free amino groups present on bacterially expressed, N-terminally His-tagged TIP47 (TxRed-TIP47). As shown in Fig. 1, as many as 18 mol of Texas Red could be incorporated into TIP47. However, at high levels of dye substitution, a significant amount of fluorescence quenching was observed, possibly due to protein aggregation. We therefore utilized a molar ratio of 2:1 (Texas Red:TIP47) for the experiments described here.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Incorporation of Texas Red dye onto TIP47. Labeling was carried out as described under "Experimental Procedures" using increasing concentrations of Texas Red.

The binding assay was carried out as follows. TxRed-TIP47 was incubated with the cytoplasmic domains of mannose 6-phosphate receptors, in the form of GST fusion proteins. Receptor-bound TIP47 was then collected on glutathione-Sepharose and quantified using a fluorimeter after elution of the resin with excess glutathione.

Fig. 2 shows that the specificity of TIP47 binding determined previously using an immunoblot assay (8) can be recapitulated using the fluorescence-based assay. TIP47 bound to the CD-MPR cytoplasmic domain in strong preference to the LDL receptor cytoplasmic domain. Moreover, TIP47 failed to interact with a CD-MPR in which the FW motif was replaced by two alanine residues. Thus, the fluorescence assay was well suited to investigate the affinity of the binding interaction and the selectivity of TIP47 for a variety of potential transmembrane cargo proteins.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Texas Red-labeled TIP47 shows the expected specificity for binding CD-MPRs. TIP47 binding to GST-cytoplasmic domain constructs representing wild type CD-MPR (filled circles), CD-MPR F18A/W19A (FWAA) (open diamonds), CD-MPR Y45A/V48A (YV AA) (open circles), or LDL receptor (open triangles) is shown.

Equilibrium binding constants for TIP47-receptor cytoplasmic domain interactions were then determined using the fluorescence assay (Fig. 3). These experiments revealed that TIP47 interacts more strongly with the CI-MPR cytoplasmic domain, with an apparent K_D of 1 µM. A weaker interaction with the CD-MPR cytoplasmic domain (K_D = 3 µM) may be due to the possible oligomerization of this cytoplasmic domain construct (24).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   TIP47 binds the CI-MPR (filled circles) with slightly higher affinity than the CD-MPR (shaded circles) and fails to bind the LDL receptor (filled triangles). The solid lines represent theoretical curve fits to experimental data according to the equation describing bimolecular reversible binding (41). Estimates for KD were generated using a curve fitting algorithm (SigmaPlot, SPSS Inc., Chicago, IL). CI-MPR and LDL-R KD values have been confirmed independently in at least 10 experiments; S.D. was 4-10% for individual determinations.

We also performed experiments to determine the rates of association and dissociation of TIP47 with both CI-MPR and CD-MPR cytoplasmic domains. TIP47 association was measured by adding the protein to receptor cytoplasmic domains prebound to glutathione-Sepharose. Rapid mixing and separation was achieved by carrying out the incubations in a syringe that could be cleared rapidly by air pressure. These experiments confirmed that binding was complete in 15 min (data not shown).

TIP47 dissociation was measured by forming TIP47-receptor cytoplasmic domain complexes, collecting them on glutathione-Sepharose, and then washing the samples under continuous flow and monitoring the amount of TIP47 remaining on the beads as a function of time. From these experiments, we determined the TIP47 dissociation rate to be 0.9-1.0 × 10² min¹. The relatively slow dissociation rate ensured that the brief wash step for the removal of unbound TIP47 would result in a minimal loss of MPR-bound TIP47.

TIP47 Binding Specificity-- The specificity of TIP47 binding was examined by carrying out binding reactions using other GST-cytoplasmic domain constructs. In this study, we included HIV-Nef, a 206-residue, myristoylated cytoplasmic protein that enhances the endocytosis of the CD4 glycoprotein by linking CD4 to the clathrin AP-2 adaptor complex at the plasma membrane (25-27). Nef can also link CD4 to -COP (28). HIV-1-Nef recruits clathrin adaptors via a dileucine based signal near its C terminus and -COP via a diacidic sequence, also near its C terminus (26, 28). Since Nef binds AP-2 and -COP, it was worth checking if it also bound TIP47.

As shown in Fig. 4, fluorescent TIP47 failed to bind the cytoplasmic domains of furin and carboxypeptidase D, under conditions in which significant binding to the CI-MPR was detected. Compared with the CI-MPR (K_D = 0.8 µM), furin and CPD displayed apparent K_D values of 50 and 70 µM, respectively. It is hard to imagine that these proteins are ever found at this high a level in the membranes of mammalian cells. Similarly, HIV-Nef bound TIP47 with an apparent K_D of 12.2 µM (not shown). Altogether, the lack of interaction of furin and carboxypeptidase D with TIP47 was unexpected, because these proteins cycle between endosomes and the trans Golgi network.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   TIP47 interacts exclusively with MPRs. TIP47's ability to bind GST fusion constructs containing the cytoplasmic domains of CI-MPR (closed circles) or CPD (closed triangles) (upper panel) or furin (open circles) or phosphorylated furin (closed circles) (lower panel) is shown. KD estimates for the LDLR (Fig. 3), furin, and CPD are associated with large errors due to weakness of interaction under the experimental conditions employed. For these proteins, the reported KD most likely overestimates the true affinity. Reactions were carried out three times; representative curves are shown.

Furin sorting has been shown to be dependent upon the specific phosphorylation of its cytoplasmic domain (29). Moreover, interaction of furin with a key sorting protein named PACS-1 is dependent upon furin phosphorylation (30). Thus, we tested whether phosphorylated furin cytoplasmic domain showed enhanced interaction with TIP47. Fig. 4 shows that the phosphorylated furin cytoplasmic domain did not show any difference in its ability to interact with TIP47 as nonphosphorylated furin cytoplasmic domain.

Another important substrate to test was TGN38. Although the function of this protein is not known, it is a steady-state resident of the trans Golgi network and continually cycles between the cell surface and the TGN (31). Initial experiments suggested strong binding of recombinant GST-TGN38 cytoplasmic domain to Texas Red-labeled TIP47. However, experiments using other fluorescently labeled control proteins strongly suggested that the observed binding was entirely nonspecific, in contrast to the other proteins tested (data not shown). It is important to note that the TGN38 cytoplasmic domain is more difficult to purify than all of the other constructs and can only be eluted from glutathione-Sepharose in the presence of detergent. Also, the TGN38 cytoplasmic domain-GST fusion protein has a pI of 8.9 compared with all of the other constructs, which have predicted pI values of 5.5-6. Thus, we sought an alternative binding assay strategy that would decrease potential nonspecific binding.

We reasoned that unlabeled cytosolic TIP47, present within a mixture of cytosolic proteins, would provide a physiologically relevant test of TIP47-TGN38 cytoplasmic domain interaction. Thus, GST-TGN38 cytoplasmic domain was incubated with cytosolic proteins, and the bound material was tested for the presence of TIP47 by immunoblotting. As shown in Fig. 5, only background levels of binding were observed for cytosolic TIP47 interaction with LDL receptor and TGN38 cytoplasmic domains. In contrast, significant TIP47 bound to the CI-MPR cytoplasmic domain under identical conditions. These experiments show that TIP47 does not interact with the cytoplasmic domain of TGN38 under relatively physiological conditions.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   TIP47 fails to bind TGN38. GST-TGN38, GST-CI-MPR, or GST-LDL receptor cytoplasmic domains were incubated with K562 cytosol for 90 min at room temperature. Bound proteins were collected using glutathione-Sepharose and analyzed by immunoblot using rabbit anti-TIP47 antibody and chemiluminescent detection. This experiment was confirmed in three experiments; a representative gel and quantitation are shown.

Receptor Cytoplasmic Domains Are Functional-- All of the experiments described thus far rely on the proper folding and functionality of the recombinant, GST fusion proteins. To verify that the proteins used in this study were active, we tested their capacities to interact with the clathrin adaptor, AP-2. AP-2 binds to tyrosine-based signals in receptor cytoplasmic domains and mediates their endocytosis via clathrin-coated vesicles (10).

To monitor potential AP-2 association, receptor cytoplasmic domain-GST fusion constructs were incubated with purified, bovine AP-2; bound AP-2 was then quantified after collection of receptor-AP2 complexes using glutathione-Sepharose beads followed by immunoblot analysis. Fig. 6A shows an experiment carried out to validate the assay procedure. As expected, AP-2 failed to bind GST alone and bound efficiently to the CI-MPR cytoplasmic domain. Binding was entirely dependent on the presence of a tyrosine-based motif, because when a CI-MPR construct was employed in which Tyr²⁴, Tyr²⁶, and Val²⁹ were converted to alanine, no binding was observed. Lack of binding to this construct was not due to its misfolding, since it was fully capable of interacting strongly with TIP47 (Fig. 6B) and AP-1 (data not shown).

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Receptor cytoplasmic domains are functional in that they bind AP-2. A, immunoblot analysis of AP-2 binding shows the expected requirement for a tyrosine-based signal. Binding was determined using either GST, GST-CI-MPR cytoplasmic domain, or GST-CI MPR cytoplasmic domain in which Tyr24, Tyr26, and Val29 are converted to alanines. AP-2 was monitored using anti--adaptin antibodies. B, comparison of TIP47 binding (dark bars) and AP-2 binding (lighter bars) to the indicated constructs. AP-2 binding was determined by scanning the immunoblot shown in A. C, AP-2, but not TIP47, binds LDL receptor, TGN38, and furin. Symbols are as in B. AP-2 binding was determined as in A and B. The relative amounts of TIP47 binding shown in A and B were taken from the experiments presented in Figs. 3-5. This experiment was carried out twice with identical results.

We then used this assay to compare the interaction of our various cytoplasmic domain preparations with either TIP47 or AP-2. As shown in Fig. 6C, while the CD-MPR bound significantly more TIP47 than LDL receptor, TGN38, or furin cytoplasmic domains (dark bars), LDL receptor, TGN38, and furin all bound more strongly to AP-2 than the CD-MPR (lighter bars). The cytoplasmic domain of CPD did not bind AP-2 in these experiments (not shown); thus, we cannot rule out the possibility that this construct is misfolded. Nevertheless, our data demonstrate that the cytoplasmic domains of LDL receptor, TGN38, and furin are fully functional in terms of their capacities to interact with cytoplasmic domain-interacting proteins. Moreover, they confirm the high level of selectivity of TIP47 for mannose 6-phosphate receptors.

In these experiments, CI-MPR bound more AP-2 than CD-MPR. It is important to note that GST fusion proteins will dimerize via GST. Moreover, it was possible that in the context of a GST fusion, the smaller, 67-residue CD-MPR cytoplasmic domain is less accessible than the larger, 163-residue, CI-MPR cytoplasmic domain. If true, this would suggest that our estimate for TIP47-CD-MPR interaction might be an underestimate of the true affinity.

To explore this possibility in greater detail, we compared the binding of AP-1 and AP-2 to both MPR cytoplasmic domains. As shown in Fig. 7, AP-1 actually bound preferentially to the CD-MPR cytoplasmic domain, whereas AP-2 showed preference for the CI-MPR cytoplasmic domain. AP-1 also bound to furin but not LDL receptor (data not shown), as would be expected given previous studies on the sorting of these two proteins (cf. Ref. 29). These data add confidence to our conclusions that the cytoplasmic domains are fully folded and have distinct affinities for interactions with TIP47.

View larger version (37K): [in this window] [in a new window]   Fig. 7.   Comparison of adaptor binding to CI- and CD-MPR cytoplasmic domains. Upper panels, immunoblot analysis of AP-1 (left) or AP-2 (right) binding to GST-CI-MPR and GST-CD-MPR cytoplasmic domains. Adaptor binding was detected by anti--adaptin (for AP-1) or anti--adaptin (for AP-2) antibodies and quantitated by densitometry (lower panel). The amount of binding to GST was subtracted to obtain the relative binding shown in the lower panel. This experiment was carried out in duplicate (AP-1) or triplicate (AP-2); a representative example is shown.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

We have shown here that TIP47 binds highly selectively to MPR cytoplasmic domains. Recombinant TIP47 binds to the CI-MPR and CD-MPR with apparent K_D values of 1 and 3 µM, respectively. Our initial estimates for the affinity of recombinant TIP47 for MPRs are in the same range as that observed for other cargo adapters (cf. Ref. 23) and would be consistent with the intracellular concentrations of these proteins. The K_D values are probably more accurate than those we reported previously (8) because of the greater sensitivity and reliability of the assay employed.

Surprising were our findings that TIP47 does not interact with TGN38, furin, or CPD. Like MPRs, these three proteins cycle continuously between the plasma membrane, endosomes, and the Golgi complex. Since only a small percentage of total cell surface proteins are transported back to the Golgi complex (32), it might have been expected that this subset of proteins would share a common route of intracellular transport.

Clues to the existence of multiple pathways for endosome-to-TGN transport are already numerous and suggest new levels of complexity in the endosomal system. Shiga toxin (33) and TGN38 (31), for example, appear to return to the Golgi complex by a process that bypasses late endosomes entirely and occurs via a compartment referred to as the "recycling endosome" (Fig. 8). Mallet and Maxfield (34) showed further that TGN38 and furin are differentially sorted in endosomes and use distinct routes for their return to the Golgi. Moreover, it appeared from this work that both furin and TGN38 were transported from compartments that were distinct from those containing MPRs (34).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Speculative model for endosome-to-TGN transport. Mallet and Maxfield (34) have shown that TGN38 and furin appear to recycle via compartments distinct from MPRs. If MPRs are segregated into a patch of membrane by interaction with TIP47, simple fission of a late endosome will yield two distinct late endosome types. Rab9 would then facilitate MPR transport to the TGN, Rab11 would drive TGN38 recycling, and other proteins would facilitate furin recycling to the trans Golgi network. For details, see "Discussion."

The data of Mallet and Maxfield (34) fit well with our findings. If TGN38, furin and MPRs are present in different compartments, they must in some way, be differently sorted. When thinking about endosomal sorting, it is important to keep in mind the dynamics of this membrane-bound compartment. It is now fairly well established that endosomes undergo homotypic fusion and also membrane fission; for example, hyperactivity of the Rab5 GTPase can alter the balance between these events for early endocytic compartments (35). Sorting into different endosome classes could be accomplished by protein association with different adapter complexes. As shown in Fig. 8, binding of MPR cytoplasmic domains by TIP47 could lead to their segregation from other cargoes. Subsequent organelle fission could then segregate the two cargo classes into distinct membrane-bound vesicles. Failure to bind TIP47 and/or binding instead to other proteins may very well be responsible for the sorting of furin and TGN38 into different compartments.

Other proteins that are likely to be important for furin sorting include APB-280 (36) and PACS-1 (30), which bind furin in the endocytic pathway. Moreover, as stated earlier, furin phosphorylation is clearly important for its intracellular transport and for its binding to PACS-1 (29, 37, 38). Our observation that furin phosphorylation did not influence TIP47 interaction supports a model in which PACS-1 and TIP47 play distinct roles in the cell. TGN38 is likely to utilize Rab11 for transport from recycling endosomes to the TGN (39), unlike MPR transport, which is dependent upon Rab9 (5, 40). More work will be needed to elucidate the pathway-specific molecules driving these probably distinct intracellular transport pathways.

Taken together, all of these findings suggest that transport from the endocytic pathway to the biosynthetic pathway is a carefully regulated process that uses an unexpectedly diverse number of proteins and possibly also a larger than expected number of transport events. Much more work will be needed to define the molecular requirements for each of these pathways and the physiological advantages of this complexity. A challenge for the future will be to identify any coat proteins with which TIP47 may interact in its role as a cargo selection device for MPR trafficking from endosomes to the Golgi complex.

	ACKNOWLEDGEMENTS

We are grateful to Drs. Gary Thomas, Oleg Varlamov, Lloyd Fricker, Vincent Piguet, Didier Trono, and Sharon Milgram for generous gifts of GST-fusion constructs and Elva Diaz for helping to initiate this project.

	FOOTNOTES

^* This research was supported by a research grant from the National Institutes of Health (DK37336).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.

Supported by a postdoctoral fellowship from the Pharmaceutical Research and Manufacturers of America Foundation.

^§ Supported by a postdoctoral fellowship from the Leukemia Society.

^¶ Supported by a postdoctoral fellowship from the Human Frontier Science Program.

To whom correspondence should be addressed. Tel.: 650-723-6169; Fax: 650-723-6783; E-mail: pfeffer@cmgm.stanford.edu.

Published, JBC Papers in Press, May 26, 2000, DOI 10.1074/jbc.M001138200

	ABBREVIATIONS

The abbreviations used are: MPR, mannose 6-phosphate receptor; CI-MPR, cation-independent MPR; CD, cation-dependent MPR; CPD, carboxypeptidase D; TGN, trans Golgi network; GST, glutathione S-transferase; LDL, low density lipoprotein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				V. Popoff, G. A. Mardones, D. Tenza, R. Rojas, C. Lamaze, J. S. Bonifacino, G. Raposo, and L. Johannes The retromer complex and clathrin define an early endosomal retrograde exit site J. Cell Sci., June 15, 2007; 120(12): 2022 - 2031. [Abstract] [Full Text] [PDF]

				D. Aivazian, R. L. Serrano, and S. Pfeffer TIP47 is a key effector for Rab9 localization J. Cell Biol., June 19, 2006; 173(6): 917 - 926. [Abstract] [Full Text] [PDF]

				H. Robenek, S. Lorkowski, M. Schnoor, and D. Troyer Spatial Integration of TIP47 and Adipophilin in Macrophage Lipid Bodies J. Biol. Chem., February 18, 2005; 280(7): 5789 - 5794. [Abstract] [Full Text] [PDF]

				I. G. Ganley, K. Carroll, L. Bittova, and S. Pfeffer Rab9 GTPase Regulates Late Endosome Size and Requires Effector Interaction for Its Stability Mol. Biol. Cell, December 1, 2004; 15(12): 5420 - 5430. [Abstract] [Full Text] [PDF]

				T. Yamaguchi, N. Omatsu, S. Matsushita, and T. Osumi CGI-58 Interacts with Perilipin and Is Localized to Lipid Droplets: POSSIBLE INVOLVEMENT OF CGI-58 MISLOCALIZATION IN CHANARIN-DORFMAN SYNDROME J. Biol. Chem., July 16, 2004; 279(29): 30490 - 30497. [Abstract] [Full Text] [PDF]

				F. B. Schapiro, T. T. Soe, W. G. Mallet, and F. R. Maxfield Role of Cytoplasmic Domain Serines in Intracellular Trafficking of Furin Mol. Biol. Cell, June 1, 2004; 15(6): 2884 - 2894. [Abstract] [Full Text] [PDF]

				C. N. Arighi, L. M. Hartnell, R. C. Aguilar, C. R. Haft, and J. S. Bonifacino Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor J. Cell Biol., April 12, 2004; 165(1): 123 - 133. [Abstract] [Full Text] [PDF]

				G. Blot, K. Janvier, S. Le Panse, R. Benarous, and C. Berlioz-Torrent Targeting of the Human Immunodeficiency Virus Type 1 Envelope to the trans-Golgi Network through Binding to TIP47 Is Required for Env Incorporation into Virions and Infectivity J. Virol., June 15, 2003; 77(12): 6931 - 6945. [Abstract] [Full Text] [PDF]

				J. Hanna, K. Carroll, and S. R. Pfeffer Identification of residues in TIP47 essential for Rab9 binding PNAS, May 28, 2002; 99(11): 7450 - 7454. [Abstract] [Full Text] [PDF]

				P. Barbero, L. Bittova, and S. R. Pfeffer Visualization of Rab9-mediated vesicle transport from endosomes to the trans-Golgi in living cells J. Cell Biol., February 4, 2002; 156(3): 511 - 518. [Abstract] [Full Text] [PDF]

				J. V. Reddy and M. N.J. Seaman Vps26p, a Component of Retromer, Directs the Interactions of Vps35p in Endosome-to-Golgi Retrieval Mol. Biol. Cell, October 1, 2001; 12(10): 3242 - 3256. [Abstract] [Full Text] [PDF]

				A. O. Johnson, M. A. Lampson, and T. E. McGraw A Di-Leucine Sequence and a Cluster of Acidic Amino Acids Are Required for Dynamic Retention in the Endosomal Recycling Compartment of Fibroblasts Mol. Biol. Cell, February 1, 2001; 12(2): 367 - 381. [Abstract] [Full Text]

				I Miwako, A Yamamoto, T Kitamura, K Nagayama, and M Ohashi Cholesterol requirement for cation-independent mannose 6-phosphate receptor exit from multivesicular late endosomes to the Golgi J. Cell Sci., January 5, 2001; 114(9): 1765 - 1776. [Abstract] [PDF]

				P. Barbero, E. Buell, S. Zulley, and S. R. Pfeffer TIP47 Is Not a Component of Lipid Droplets J. Biol. Chem., June 22, 2001; 276(26): 24348 - 24351. [Abstract] [Full Text] [PDF]

				P. Barbero, L. Bittova, and S. R. Pfeffer Visualization of Rab9-mediated vesicle transport from endosomes to the trans-Golgi in living cells J. Cell Biol., February 4, 2002; 156(3): 511 - 518. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 275/33/25188    most recent M001138200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager