| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 280, Issue 22, 21067-21077, June 3, 2005
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
From the Department of Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha, Nebraska 68198-5870
Received for publication, November 16, 2004 , and in revised form, February 24, 2005.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-FLAG or -Myc-based immunoprecipitation assays showed that all of the minireceptors tested were able to associate with a full-length, Myc-tagged M6P/IGF2R (WT-M). In the -FLAG but not -Myc immunoprecipitation assays, the degree of association of a series of C-terminally truncated minireceptors with WT-M showed a positive trend with length of the minireceptor. In contrast, length did not seem to affect the association of the N-terminally truncated minireceptors with WT-M, except that the 12th extracytoplasmic repeat appeared exceptionally important in dimerization in the -FLAG assays. The presence of mutations in the ligand-binding sites of the minireceptors had no effect on their ability to associate with WT-M. Thus, association within the heterodimers was not dependent on the presence of functional ligand binding domains. Heterodimers formed between WT-M and the minireceptors demonstrated high affinity IGF-II and Man-6-P-ligand binding, suggesting a functional association. We conclude that there is no finite M6P/IGF2R dimerization domain, but rather that interactions between dimer partners occur all along the extracytoplasmic region of the receptor. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
300 kDa (1, 2). This receptor comprises a large extracytoplasmic (EC) domain, a single membrane-spanning region, and a short cytoplasmic tail. The EC domain is the principal ligand-binding region of the receptor, consisting of 15 homologous repeats of 145 amino acids each (1, 3, 4). The M6P/IGF2R has been shown to bind at least two classes of ligands, the Man-6-P-containing and the non-Man-6-P-containing polypeptide ligands, all of which bind to sites within the EC region (5–7). Newly synthesized Man-6-P-containing ligands such as lysosomal acid hydrolases bind to the M6P/IGF2R in the trans-Golgi network through Man-6-P residues on their N-linked oligosaccharides, whereas other Man-6-P-containing ligands, such as latent transforming growth factor-
1 (TGF-1), proliferin, and granzyme B, bind at the cell surface (1, 8). Through binding of this large class of ligands, the M6P/IGF2R mediates several important cellular functions, such as the endocytosis and/or targeting of acid hydrolases to lysosomes (9), the proteolytic activation of latent TGF-1 (10–12), mediation of the migration and angiogenesis induced by proliferin (13), and the internalization of granzyme B (14). Two distinct high affinity binding sites and one, recently discovered, low affinity binding site for the Man-6-P-containing ligands map to specific residues that are common to repeats 3 and 9 and repeat 5 of the EC domain, respectively (15–19). The two high affinity binding sites are not functionally equivalent with respect to ligand preference, having distinct dissociation constants for the multivalent Man-6-P-ligand -glucuronidase (2.0 versus 4.3 nM for repeats 3 and 9, respectively) (20). The pH optimum for carbohydrate binding is also more acidic for repeat 9 than repeat 3 (pH 6.4 versus pH 6.9, respectively), and the two sites differ in their ability to recognize distinctive modifications found on Dictyostelium discoideum glycoproteins, such as mannose 6-sulfate and Man-6-P methyl esters (21). Additionally, repeat 9 alone can fold into a high affinity ligand binding domain, whereas repeat 3 depends on residues in adjacent repeats 1 and/or 2 for optimal ligand binding (22). Although it exhibits significant sequence homology with repeats 3 and 9, as well as sharing four conserved residues key for Man-6-P binding, repeat 5 has an 300-fold lower affinity for Man-6-P than repeat 9 or repeats 1–3, possibly due to the absence of two half-cystines that form a stabilizing disulfide bond in repeats 3 and 9 (19).
The non-Man-6-P-containing class of ligands includes the polypeptide mitogen, insulin-like growth factor II (IGF-II). The IGF-II-binding site has been mapped to repeat 11 of the EC region, with high affinity binding being conferred by residues contributed by the 13th repeat (23–26). Repeat 13 is thought to act as an enhancer of IGF-II affinity by slowing the rate of IGF-II dissociation (27). Structural analyses of repeat 11 identified the putative IGF-II-binding site in a hydrophobic pocket at the end of a -barrel structure (28). Another member of this class is retinoic acid, a unique ligand for the M6P/IGF2R in that it binds the cytoplasmic region and is thought to function by altering intracellular trafficking of the M6P/IGF2R and its cargo (29). The other members of the non-Man-6-P-containing ligands are urokinase-type plasminogen activator receptor (uPAR) and plasminogen, whose binding sites have been mapped to a peptide region within EC repeat 1 (30–32). The proposed function for the interactions between the M6P/IGF2R and these ligands is involvement in the complex responsible for the activation of TGF-1 at the cell surface, as well as endocytosis and targeting of uPAR for degradation (31, 32). The uPAR-M6P/IGF2R interaction appears to be weak, of low affinity, and confined to a small subpopulation of uPAR molecules (33), which calls into question the physiological relevance of this interaction.
Recent crystal structure data have given insight into structural features of the M6P/IGF2R (28, 34). The crystal structures for repeat 11 by Brown et al. (28) and repeats 1–3 by Olson et al. (34, 35) have allowed these groups to propose different models for the overall structure of the EC domain of the M6P/IGF2R. The EC domain of the receptor shows considerable homology among repeats and the cation-dependent Man-6-P-receptor (16–38% identity) (4). This high level of sequence identity accounts for structural similarities among domains, including conserved disulfide bond organization, random coil linker regions connecting the domains, and an overall core flattened -barrel structure. The 1–3 triple-repeat crystal revealed a structure in which repeat 3 sits on top of repeats 1 and 2 (34). Olson et al. (34) have proposed that the M6P/IGF2R forms distinct structural units for every three repeats of the EC region, producing five tri-repeat units that stack in a back-to-front manner. In this model, the IGF-II-binding site is located on the opposite face of the structure relative to the Man-6-P-binding sites.
Traditionally thought to function as a monomer (36), the M6P/IGF2R is now considered to operate optimally in the membrane as an oligomer for high affinity Man-6-P binding and efficient internalization of ligands (37–39). Intermolecular cross-linking of two M6P/IGF2R partners was shown to occur upon binding of the multivalent ligand, -glucuronidase, resulting in increased rate of ligand internalization (37). The initial rate of internalization of -glucuronidase was faster than for the monovalent ligand, IGF-II, which showed that multivalent ligands enhance the rate of receptor movement, likely due to clustering of the M6P/IGF2R for improved interaction with the endocytic machinery in the formation of clathrin-coated pits (37). Further studies demonstrated that alignment of the Man-6-P binding domains of monomeric partners of a receptor dimer is responsible for bivalent, high affinity binding, also supporting the importance of receptor oligomerization (38).
In order to determine the interrelationship between dimer formation and the function of the ligand binding domains of the receptor, we co-expressed full-length receptors with truncated receptors from which either the N- or C-terminal EC repeats were deleted and that lacked functional ligand binding domains and/or transmembrane and cytoplasmic domains. One of the main goals of this project was to map the dimerization domain(s) of the M6P/IGF2R. We hypothesized that one or more dimer interaction domains would be located at or near the ligand binding domains in the EC region of the receptor, and that these regions would contribute preferentially to receptor dimerization. A panel of M6P/IGF2R minireceptors was constructed to test for association with a full-length version of the receptor. It was observed that all of the truncated receptors were able to associate with the full-length receptor, suggesting that dimerization domains or contacts occur all along the EC region of the receptor, not just near regions of ligand interaction. However, repeat 12 seems to be particularly important to association, as N-terminally truncated minireceptors lacking this repeat gave distinctive results in the immunoprecipitation assays. We conclude that a distinct dimerization domain for the M6P/IGF2R does not exist per se, but instead, interactions between monomeric receptor partners apparently occur all along the EC region of the receptor with special contribution made by repeat 12.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
)-FLAG M2 antibody, -FLAG M2-agarose affinity gel, and the bicinchoninic acid kit for protein determination were purchased from Sigma. The -Myc 9E10 antibody was purchased from Upstate Biotechnology, Inc., or the University of Nebraska Medical Center Monoclonal Antibody Facility (Omaha, NE). The polyclonal -13D antibody (referred to as -M6P/IGF2R throughout this report) that recognizes a peptide domain in repeat 4 of the M6P/IGF2R has been described previously (40). Rabbit -mouse IgG was from Dako (Carpinteria, CA). Carrier-free Na125I and 125I-protein A were from PerkinElmer Life Sciences. Recombinant human IGFs were provided by M. H. Niedenthal (Lilly). Radiolabeled IGF-II and unlabeled and radiolabeled pentamannose phosphate-bovine serum albumin (PMP-BSA) were prepared as described previously (41). The pCMV5 vector was provided by Dr. David W. Russell (University of Texas Southwestern Medical Center, Dallas, TX) (42). The 8.6-bp human M6P/IGF2R cDNA and affinity-purified human -glucuronidase (hGUS) were provided by Dr. William S. Sly (St. Louis University Medical Center, St. Louis, MO) (3). Radiolabeled hGUS was prepared by iodination using precoated IODO-GEN tubes (Pierce) according to the manufacturer's specifications to a specific activity of 26–40 Ci/g. Other reagents and supplies were obtained from sources as indicated.
Preparation, Expression, and Analysis of Epitope-tagged Minireceptors
The truncated M6P/IGF2R minireceptors 1-8F, 1-9F, 1-9F-R/A, 1-11F, and 1-15F were tagged with the eight-residue FLAG epitope (DYKDDDDK) followed by a stop codon and an XbaI restriction site at the C terminus and cloned into the pCMV5 vector as described previously (Fig. 1A) (38). By using the full-length M6P/IGF2R cDNA as the template, the following minireceptors, containing the EC repeats starting with the repeat indicated in the name of the receptor and ending with the 15th repeat, followed by the transmembrane and cytoplasmic regions of the M6P/IGF2R, were synthesized by amplification with VentTM polymerase (New England Biolabs, Beverly, MA): 10-15CF (nt 4239–7620), 11-15CF (nt 4675–7620), 12-15CF (nt 5092–7620), and 13-15CF (nt 5542–7620) (Fig. 5A). To ensure consistent translation, the signal sequence containing the N-terminal 71 residues of repeat 1 was fused to the beginning of each construct as described previously (43). All of these minireceptors were C-terminally tagged with the FLAG epitope followed by a stop codon and XbaI restriction site for cloning purposes.
The cDNA plasmids in the vector pCMV5R1X encoding the 10-15CF-I/T and 11-15CF-I/T minireceptors, bearing the isoleucine to threonine mutation that has been shown previously to prevent IGF-II binding to the M6P/IGF2R, were synthesized from the 10-15CF and 11-15CF constructs as well as a 1-15F construct containing the I1572T mutation (1-15F-I/T) (39). Briefly, the 10-15CF or 11-15CF and 1-15F-I/T cDNAs were digested with BstEII (M6P/IGF2R nt 4698) and BstBI (M6P/IGF2R nt 5507) serially, and the resulting fragments of interest (809-bp fragment for 1-15F-I/T, corresponding to the region containing the Ile 
Thr mutation, and the larger fragment (>7000 bp) for 10-15CF and 11-15CF, which also encompasses the pCMV5 plasmid), were gel-purified (Qiagen, Valencia, CA). The purified fragments were then ligated together using T4 DNA Ligase (Invitrogen) and used to transform XL-10 Gold competent cells (Stratagene, La Jolla, CA).
The full-length Myc-tagged M6P/IGF2R was prepared as follows: full-length M6P/IGF2R cDNA that had been digested with EagI and re-ligated (lacking nt 162–5319) was used as template for amplification with a 5'-primer containing an XhoI restriction site preceding the sequence corresponding to nt 94–113 of the receptor cDNA and a 3'-primer with sequence complementary to nt 7602–7620 at the C-terminal end of the expressed M6P/IGF2R followed by the 36-nt sequence encoding the Myc epitope, MEQKLISEEDLN (44), followed by two stop codons and an XbaI site. The products from these amplifications were digested with XbaI and XhoI and subcloned into pBKCMV (Invitrogen). These plasmids were then digested with HindIII and XbaI and subcloned into the target vector, pCMV5. Finally, wild-type EagI fragments were subcloned into the construct, reconstituting the completed M6P/IGF2R-Myc (WT-M) cDNA construct.
A full-length, Myc-tagged M6P/IGF2R triple mutant (R2AxI/T-M) for residues in the repeats 3 and 9 Man-6-P binding domains (R426A and R1325A, respectively) and the repeat 11 IGF-II binding domain (I1572T) was prepared as follows: nt 100–3242 of the M6P/IGF2R cDNA served as template for amplification using a 5'-primer containing an EcoRI restriction site preceding the sequence corresponding to nt 1112–1129 of the receptor cDNA and a 3'-primer with sequence complementary to nt 2474–2487 of the M6P/IGF2R cDNA, followed by an XhoI site. The products from these amplifications were digested with EcoRI and XhoI and subcloned into pBluescript SK II+ (pBSKII+) (Stratagene). This construct was subjected to two rounds of amplification with primers designed to incorporate the R426A mutation responsible for altering the EC repeat 3 Man-6-P-binding site using the Megaprimer approach (45). The first round of amplification involved producing the mutation, by amplifying from that site (nt 1425) to the 3' end of the minireceptor (nt 2487). This "megaprimer" was then used in a second round of amplification with the 5'-primer used above. The repeat 3 mutant amplification product was digested with EcoRI and XhoI and subcloned back into pBSKII+. An 1-kb BsmI-BsmI fragment (sites at M6P/IGF2R nt 1408 and 2449) containing the mutation was removed from the megaprimer and subcloned into the corresponding positions of pBSKII+/Kpn, which contained a 2.2-kb KpnI-KpnI fragment derived from M6P/IGF2R nt 100–3242, creating a pBSKII+/Kpn-R426A minireceptor with the 3rd repeat Man-6-P-binding mutation. The KpnI-KpnI fragment from this construct was then subcloned into pCMV5/R1325A-M, a full-length Myc-tagged receptor containing the 9th repeat Man-6-P-binding mutation synthesized as described previously (39), that had also been digested with KpnI and XmnI, creating a full-length M6P/IGF2R bearing Man-6-P-binding site mutations at both repeats 3 and 9 (R2A-M). The R2A-M cDNA was digested with AflII (M6P/IGF2R nt 4740) and MluI (pCMV5 nt 933), and this 5-kb fragment was subcloned into pCMV5/I1572T-M, a full-length Myc-tagged receptor containing the 11th repeat IGF-II-binding mutation at nt 4862 (24), which had also been digested with AflII and MluI. The final product resulted in the binding-defective triple mutant, R2AxI/T-M, in the vector pCMV5.
Transient expression of the minireceptors by calcium phosphate-mediated transfection into 293T human embryonic kidney cells and immunoblot analysis of cell lysates to measure expression of the truncated and full-length receptors were performed as described previously (39, 46). Additionally, expression was tested by the use of the -M6P/IGF2R polyclonal antibody (40). Aliquots (25 µl) of Triton X-100 extracts were resolved by electrophoresis on 6% reducing SDS-polyacrylamide gels in sample buffer (50 mM Tris-HCl, pH 6.8, 2.5% SDS, 5% sucrose, and 0.01% bromphenol blue) plus 50 mM dithiothreitol (DTT) and then transferred to BA85 nitrocellulose paper (Schleicher & Schuell) at 0.35 A for 3 h at 22 °C in immunoblotting buffer (0.4 M Tris, pH 7.5, 3 M glycine, 20% methanol). The blots were incubated with blocking buffer (4% nonfat dry milk in 50 mM HEPES, pH 7.6, 150 mM NaCl, 0.1% Tween 20, 0.02% sodium azide) for 1 h at 22 °C and probed with the -M6P/IGF2R antibody (1:500 dilution). The blots were then developed with 125I-protein A and detected by autoradiography.
Western Ligand Blot Analysis
Aliquots (25–50 µl) of 293T whole-cell lysates were electrophoresed on 6% SDS-PAGE in sample buffer and transferred to BA85 nitrocellulose paper at 0.35 A for 3 h at 20 °C in ligand transfer buffer (15 mM Tris-HCl, pH 8.3, 120 mM glycine, 20% methanol), according to the method of Hossenlopp et al. (47). The blots were probed overnight with 1 x 106 cpm of 125I-IGF-II or 125I-PMP-BSA or 0.5 x 106 cpm of 125I-hGUS. The blots were subsequently washed three times for 10 min at 4 °C and developed by autoradiography.
Dimer Formation Assay
-FLAG Immunoprecipitation—Equal volumes (20 or 50 µl) of whole-cell lysates prepared from 293T cells co-transfected with the various cDNAs for the FLAG-tagged minireceptors and the Myc-tagged M6P/IGF2R were incubated with 8 µl of packed M2 resin in 25 mM HEPES, pH 7.4, 150 mM NaCl (HBS) plus 1.0% BSA at 4 °C for 3 h. The resin pellets were collected by centrifugation at 13,000 x g for 30 s and washed twice with 1 ml of HBS plus 0.05% Triton X-100 (HBST). Immunoblot analysis was performed by subjecting the resin pellets to treatment with sample buffer plus DTT, electrophoresis on 6% reducing SDS-polyacrylamide gels, and transfer to BA85 nitrocellulose paper. The blots were incubated in blocking buffer and probed with either -FLAG or -Myc monoclonal antibodies (1:2000 or 1:500 dilution, respectively). The blots were then incubated with rabbit -mouse IgG secondary antibody (1:1000), developed with 125I-protein A, and detected by autoradiography. Levels of FLAG- and Myc-tagged proteins immunoprecipitated with the M2 resin were quantified using Storm or Typhoon PhosphorImager (American Biosciences) analysis of the immunoblots.
-Myc Immunoprecipitation—Equal volumes (20 µl) of whole-cell lysates prepared from 293T cells co-transfected with the various cDNAs for the FLAG-tagged minireceptors and the Myc-tagged M6P/IGF2R were incubated with 1 µl of 9E10 -Myc monoclonal antibody in 79 µlof HBST at 4 °C for 16 h. Protein G-Sepharose aliquots (stored as a 50% (v/v) slurry in 20% EtOH buffer) were washed four times with excess HBS plus 1.0% BSA to remove EtOH and to block the resin finally aspirated to a 50% slurry (resin/buffer). Aliquots (25 µl) of resin slurry were added to the overnight incubations along with 75 µl of HBS plus 1.0% BSA and 5 mM Man-6-P (200 µl total reaction volume) and then incubated at 4 °C for 5 h. The resin pellets were collected by centrifugation at 13,000 x g for 30 s and washed three times with 1 ml of HBST. Immunoblot analysis and quantification were performed as described above for the -FLAG immunoprecipitation assays.
Radioligand Binding Assays
Four aliquots of whole-cell lysates prepared from 293T cells co-transfected with cDNAs encoding the various FLAG-tagged minireceptors and the Myc-tagged M6P/IGF2Rs were incubated with 8 µl of packed M2 affinity resin in HBS + 1% BSA buffer at 4 °C for 6 h in a total volume of 200 µl. Lysate volumes were calculated by quantification of immunoblots of these lysates by Typhoon PhosphorImager using ImageQuant 5.0 software to achieve equivalent loading of WT-M. Addition of 5 mM Man-6-P at this point prevented the co-precipitation of endogenous Man-6-P-containing ligands. The resin pellets were collected by centrifugation at 13,000 x g for 30 s, washed three times with 1 ml of HBST, and aspirated to a final volume of 100 µl after the final wash. The ability of the immunoprecipitated receptors to bind 125I-IGF-II was measured by incubating the resin pellets with 2 nM 125I-IGF-II plus 100 nM unlabeled IGF-I in HBST for 16 h at 4 °C. The addition of IGF-I to the binding reaction prevented interference from IGF-binding proteins in the cell lysates. The resin pellets were washed twice with 1 ml of HBST to remove unbound ligand, collected by centrifugation, and counted in a Wizard 1000 
-counter (PerkinElmer Life Sciences). Three of the replicate aliquots were used to assess total binding, and specific 125I-IGF-II binding was determined by subtracting counts/min radioligand bound in the fourth replicate, which was incubated in the presence of 1 µM IGF-II. Binding of 125I-PMP-BSA was measured under similar conditions using 1 nM 125I-PMP-BSA in the presence or absence of 5 mM Man-6-P to compensate for nonspecific binding. The data were graphed and analyzed using GraphPad PrismTM software.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
-FLAG and -Myc antibodies was employed to quantify expression of the proteins (Fig. 1B).
|
|
-FLAG affinity resin, any full-length Myc-tagged receptors retained on the resin would indicate an association between the receptors. Equal volumes of each singly transfected or co-transfected lysate were immunoadsorbed to aliquots of -FLAG M2 resin. An -Myc immunoblot probed with a monoclonal Myc antibody was used to assess which C-terminally truncated minireceptors were able to form stable interactions with full-length WT-M receptor. All of the FLAG-tagged minireceptors were immunoadsorbed to the resin (Fig. 3A, upper panel, lanes 2–6). Lysates from cells transfected with the WT-M expression plasmid alone or in a 2:1 transfection with empty pCMV5 vector were used as controls, to establish that WT-M would not be immunoadsorbed to the -FLAG affinity resin in the absence of a FLAG-tagged partner (Fig. 3A, upper and lower panels, lanes 1 and 7). However, the WT-M receptor was able to immunoprecipitate along with all of the C-terminally truncated minireceptors tested (Fig. 3A, lower panel, lanes 2–6). In other experiments (data not shown), even smaller C-terminally truncated minireceptors were able to immunoprecipitate WT-M, including 1F and 1-3F, as well as the C- and N-terminally truncated minireceptor 7-9F, constructs that contain only one or three EC repeats, as indicated by their names.
|
-FLAG affinity resin, immunoprecipitation experiments were also performed using the -Myc antibody followed by incubation with protein G-Sepharose. In this reciprocal assay, retention of FLAG-tagged minireceptors with WT-M in the -Myc immunoprecipitation would be indicative of association between the receptor species. Equal lysate volumes were incubated with aliquots of -Myc antibody and subsequently precipitated by protein G-Sepharose. An -FLAG immunoblot was used to assess association of the C-terminally truncated minireceptors with WT-M. WT-M was immunoadsorbed to the -Myc matrix as expected (Fig. 3C, lower panel, lanes 1–7). Additionally, all of the FLAG-tagged minireceptors were co-immunoprecipitated with WT-M (Fig. 3C, upper panel, lanes 2–6). Lysates from cells transfected with empty pCMV5 vector or 1-15F alone were used as negative controls, to verify that FLAG-tagged minireceptors were not immunoprecipitated by protein G-Sepharose in the absence of a Myc-tagged partner (data not shown).
Semi-quantitative analysis of the interactions between WT-M and the C-terminally truncated minireceptors revealed a trend that indicated a dependence of association on the length of the truncated receptor in the -FLAG association assays. In these -FLAG immunoprecipitation assays, when adjusted for the amount of immunoprecipitated FLAG-tagged receptor, the minireceptors comprising a greater number of EC repeats retained progressively more WT-M receptor molecules on the -FLAG affinity resin. The resulting immunoblots from the immunoprecipitation experiments summarized above were quantified using Typhoon PhosphorImager and ImageQuant software. Values for co-immunoprecipitation of WT-M were adjusted for the amount of FLAG-tagged receptor immunoprecipitated in each assay based on PhosphorImager data (Fig. 3B). The analysis revealed a trend in the amount of WT-M interaction with the C-terminally truncated receptors as a function of increasing minireceptor length, with 1-8F being the weakest partner and 1-15F being the strongest. The only break in this trend occurred in the assay of WT-M co-immunoprecipitation with the 1-9F-R1325A (R/A) mutant minireceptor relative to its wild-type counterpart, 1-9F (Fig. 3B, compare lanes 4 and 3). These data suggest that the interaction with WT-M is not dependent on function of the Man-6-P ligand binding domains.
|
-Myc immunoprecipitation experiments (Fig. 3D). These data were adjusted for the amount of WT-M receptor immunoprecipitated and quantified as above. The length dependence trend was not evident for these assays (compare Fig. 3, B and D). Furthermore, the 1-15F minireceptor was not as efficiently co-immunoprecipitated with WT-M as the other C-terminally truncated minireceptors, as 30% of 1-15F was immunoprecipitated in comparison to 70–80% immunoprecipitation of the other minireceptors (data not shown); therefore, quantification showed a significant decrease in WT-M association with 1-15F in the -Myc immunoprecipitation experiments (Fig. 3D, last bar).
Transient Expression of N-terminally Truncated M6P/IGF2R Minireceptors—Another panel of minireceptors, which lacked repeats from the N terminus of the EC domain but contained the transmembrane and cytoplasmic regions, was designed to provide further insight into the dimerization domain of the Man-6-P/IG2R (Fig. 4A). None of the minireceptors in this panel contains any Man-6-P-binding sites, and four of the minireceptors lack a functional IGF-II-binding site as well (10-15CF-I/T, 11-15CF-I/T, 12-15CF, and 13-15CF). In order to test whether the full-length M6P/IGF2R was able to associate with these receptors, 293T cells were co-transfected with the N-terminally truncated minireceptors plus the full-length WT-M receptor. Additionally, 1-15F was used as a positive control because it contains the full EC domain with all of the ligand binding domains intact. Whole-cell lysates were prepared and tested for protein expression by -FLAG and -Myc immunoblotting (Fig. 4B). All of the transfected proteins were expressed and appeared at the expected molecular masses.
|
The M6P/IGF2R Interacts with the N-terminally Truncated Minireceptors in a Length-independent Manner—To investigate further the potential dimerization sites in the EC domain of M6P/IGF2R, immunoprecipitation experiments were performed using the N-terminally truncated M6P/IGF2R minireceptors. As with the C-terminal truncation panel, immunoprecipitation/immunoblotting experiments with the -FLAG affinity resin followed by expression analysis with -Myc antibody indicated association of the truncated receptors with the Myc-tagged full-length receptor, WT-M. Equal volumes of transfected whole-cell lysates were again immunoadsorbed to aliquots of -FLAG M2 resin. All of the FLAG-tagged truncated receptors were able to immunoadsorb to the -FLAG resin (Fig. 6A, upper panel, lanes 2–8). Immunoblotting with -Myc revealed that all of the N-terminally truncated minireceptors tested were able to associate with WT-M (Fig. 6A, lower panel, lanes 2–8). As before, lysates from cells transfected with WT-M and the vector pCMV5 showed no immunoprecipitation to the -FLAG resin on either the FLAG or Myc immunoblots due to the lack of a FLAG-tagged partner (Fig. 6A, upper and lower panels, lane 1).
|
This panel of minireceptors was also tested for association with the WT-M receptor by a reciprocal immunoprecipitation strategy using -Myc antibody (Fig. 6, C and D). These experiments again showed that WT-M was immunoprecipitated by -Myc antibody/protein G-Sepharose (Fig. 6C, lower panel). Furthermore, all of the N-terminally truncated minireceptors were able to co-immunoprecipitate with WT-M (Fig. 6C, upper panel, lanes 3–8). As before, vector alone or singly transfected 1-15F was not immunoprecipitated (data not shown). Quantification was performed as above with normalization for the amount of immunoprecipitated Myc-tagged receptors (Fig. 6D). As with the -FLAG immunoprecipitation experiments, no length dependence was evident for the N-terminally truncated minireceptors associating with WT-M. However, the sharp drop-off of association between 13-15CF and WT-M observed in the -FLAG immunoprecipitation assays was reversed to a substantial increase in association in the -Myc immunoprecipitation experiments (Fig. 6, B and D, compare last bars).
Association of M6P/IGF2R Minireceptors with a Ligand-binding Mutant Receptor—Both the -FLAG and -Myc immunoprecipitation experiments using the C-terminally and N-terminally truncated minireceptors were repeated with the triple-mutant form of the M6P/IGF2R, R2AxI/T-M, which is unable to bind IGF-II or Man-6-P-containing glycoproteins. All of the minireceptors were able to co-immunoprecipitate R2AxI/T (data not shown). In addition, the pattern of association for both the C- and N-terminally truncated receptors with the mutated M6P/IGF2R was similar to that seen with the wild-type receptor as in Fig. 3, B and D, and Fig. 6, B and D, respectively (data not shown).
|
In this case, the pattern of IGF-II binding was highly dependent on the IGF-II binding capability of the FLAG-tagged N-terminally truncated minireceptors. The 1-15F/WT-M, 10-15CF/WT-M, and 11-15CF/WT-M pairs were able to bind labeled IGF-II to a greater extent than the other minireceptor/WT-M combinations tested (Fig. 7A, combinations labeled with asterisk). These three groups have the only FLAG-tagged receptors tested in this panel that contain a wild-type IGF-II-binding site. The contribution of WT-M to IGF-II binding in the combinations is evident, however, when the combinations containing FLAG-tagged partners that are unable to bind IGF-II are compared with the CMV5/WT-M and WT-M alone (not shown) controls, in which the WT-M receptor fails to be immunoprecipitated by the M2 resin because of the absence of any FLAG-tagged partners (Fig. 7A).
PMP-BSA binding assays were done to determine whether the WT-M partners could bind Man-6-P-containing ligands in dimeric structures with the N-terminally truncated minireceptors that have no Man-6-P binding-competent domains. The 1-15F minireceptor used as a control in these experiments contains both complete Man-6-P binding domains. As expected, the 1-15F/WT-M combination was able to bind PMP-BSA to a greater extent than any of the others tested (Fig. 7B, labeled with asterisk). The CMV5/WT-M and WT-M controls (not shown) had no detectable PMP-BSA binding because WT-M is not immunoadsorbed to the M2 resin in these combinations (Fig. 7B, control, C). Most interestingly, all of the N-terminally truncated minireceptors associated with WT-M displayed measurable PMP-BSA binding (Fig. 7B, unlabeled combinations). This observation provides evidence that N-terminally truncated minireceptors and the full-length M6P/IGF2R can form functional dimers, even though the only PMP-BSA binding-competent species in these combinations is WT-M.
R2AxI/T-M Reduces the Amount of Radioligand Bound by Heterodimeric M6P/IGF2Rs—To establish the validity of the radioligand binding detected using the co-immunoprecipitated N-terminally truncated minireceptors and WT-M, further ligand binding studies were performed using the panel of N-terminally truncated minireceptors with the mutant M6P/IGF2R, R2AxI/T-M. N-terminally truncated minireceptors co-expressed with R2AxI/T-M were assayed as above for binding of 125I-IGF-II or 125I-PMP-BSA. This series showed IGF-II binding to combinations in which the partner minireceptor contained an intact IGF-II binding domain (Fig. 7A, labeled with asterisk), but there was little to no detectable binding to the combinations in which FLAG-tagged receptors did not contain an intact IGF-II binding domain (Fig. 7A, unlabeled combinations). This was a marked decrease from the amount of IGF-II binding demonstrated with WT-M (Fig. 7A). For PMP-BSA, none of the N-terminally truncated minireceptors were able to bind 125I-PMP-BSA when associated with R2AxI/T-M (Fig. 7B, unlabeled combinations). Only the control 1-15F minireceptor demonstrated PMP-BSA binding in combination with R2AxI/T-M, which is not surprising because this is the only minireceptor in this panel that still contains Man-6-P binding-competent sites (Fig. 7B, labeled with asterisk). PMP-BSA binding to the N-terminally truncated minireceptors was reduced from that observed in combination with WT-M (Fig. 7B, compare WT-M and R2AxI/T-M). Thus, studies using this triple mutant, ligand-binding defective M6P/IGF2R further support the notion that the associations formed between full-length M6P/IGF2R and the minireceptors are functional with respect to contributions from both partners in the oligomeric structure.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
For the panel of C-terminally truncated receptors, the -FLAG-based association assays revealed overall that longer truncated receptors immunoprecipitated WT-M to a greater degree than the shorter receptors. This increasing avidity for interaction with WT-M does not seem to be related to the presence or absence of ligand binding domains, but rather depends simply on the length of the truncated receptors themselves. Minireceptors containing even a single repeat (1F) or those containing both N- and C-terminal truncations (7-9F) were still able to associate with WT-M. The possibility that increasing association with WT-M depends simply on the length of the EC region of the minireceptor implies an organization of the receptor dimer in which the intermolecular association of the receptors is mediated by a number of contact points between receptors all along the M6P/IGF2R EC domain, rather than contact points being made only between EC repeats that are involved in providing functionality for the M6P/IGF2R, such as repeats involved in ligand binding. All of the N-terminally truncated minireceptors tested were able to interact with WT-M in both -FLAG association assays. However, the degree of association between the N-terminally truncated minireceptors and the M6P/IGF2R in the immunoprecipitation was not dependent on the length of the minireceptor molecules. For the -FLAG immunoprecipitations, 13-15CF was the exception to this finding in that this minireceptor did not associate with WT-M as well as the other constructs, suggesting special importance to the presence of the 12th repeat in the EC region.
In contrast, -Myc association-based assays with these same panels did not show a trend toward length dependence for either C- or N-terminally truncated minireceptors. The major differences between the data sets from the two assays were observed in the degree of association between 1-15F and WT-M and between 13-15CF and WT-M. These apparently conflicting data may be reconciled if we consider that the association assays measure the net outcome of several competing processes that affect the following: 1) binding affinity between Myc- and FLAG-tagged partners in heterodimer formation; 2) the tendency for formation of homodimers, which likely affects longer minireceptors to a greater degree than shorter minireceptors; and 3) the physical demands of immunoprecipitation between different sized partners. If we accept the seemingly logical premise that binding affinity increases with a greater number of EC repeats, then the disparities between the FLAG- and Myc-based immunoprecipitation assays, which are expressed most markedly in the largest minireceptor 1-15F and the smallest minireceptor 13-15CF, can be resolved. If by virtue of its length and physical constraints in the immunoprecipitation assay, 1-15F forms homodimers with a preference over heterodimer formation with WT-M, this would cause underestimation of its ability to interact with WT-M in the -Myc assay. By contrast, due to its small size and the absence of repeat 12, all dimers formed between 13-15CF and its partners would be of relatively low affinity. This would reduce its effectiveness in immunoprecipitations where 13-15CF acts as the bait and WT-M is the prey, i.e. in the -FLAG immunoprecipitations. In contrast, 13-15CF makes a much better prey when WT-M serves as the bait in the -Myc assays, due to increased degrees of freedom and decreased physical constraints on the complex, which would lead to overestimation of its association with WT-M. Although neither assay presents a complete representation of the association between partners in heterodimer formation, the composite value of using both -FLAG and -Myc assays leads to important insights into dimer interactions of the M6P/IGF2R.
The presence of repeat 12 seems to be important for dimer formation in the -FLAG assays, as can be observed in the differences between 12-15CF versus 13-15CF and also 1-11F versus 1-15F, as well as by some of the -Myc results. If repeat 12 were important for receptor complex formation, it would be important in both homo- and heterodimer formation, not overly affecting the results of the -FLAG assays but affecting receptors containing repeat 12 for the -Myc assays because of the tendency of the repeat 12-containing minireceptors to homodimerize. In the C-terminally truncated minireceptor panel, 1-15F is the only minireceptor containing repeat 12, and its ability to associate with WT-M is likely underrepresented in the -Myc assays. In the N-terminally truncated minireceptor series, 13-15CF is the only minireceptor lacking repeat 12, and its ability to associate with WT-M is likely over-represented in the -Myc assay. The -FLAG immunoprecipitation data in which 13-15CF did not associate with WT-M to the same extent as the other N-terminally truncated minireceptors are further supported by previous experiments from Byrd et al. (39) that used a M6P/IGF2R-epidermal growth factor receptor (EGFR) chimera containing EC repeats 11–15 or 13–15 plus the transmembrane region of the M6P/IGF2R and the cytoplasmic domain of the EGFR. Immunoprecipitation and phosphorylation experiments showed that, although the 13–15 chimeric receptor was able to immunoprecipitate with a full-length M6P/IGF2R-EGFR chimera and participate in autophosphorylation, it did so at 30% of the activity of the 11–15 chimera (39). These assays also used -FLAG immunoprecipitation as the method of receptor immobilization, but in these assays both receptor species were FLAG epitope-tagged. Repeat 12 may contribute to stabilization of the M6P/IGF2R dimer as follows: 1) by serving as the anchoring repeat to stabilize the structure of a tri-repeat mini-domain as proposed by the recent report of the repeat 1–3 crystal structure; 2) by serving as a stabilizer of an important functional feature of the receptor, such as cooperation in IGF-II binding between repeats 11 and 13; or 3) by direct intermolecular association with the corresponding repeat 12 from its interacting dimeric partner.
Studies with these panels of truncated receptors also permitted analysis of the potential contributions of the distal Man-6-P- and IGF-II binding domains to receptor association. The minireceptors in these panels containing the 9th repeat Arg/Ala substitution that eliminates Man-6-P binding or the 11th repeat Ile/Thr mutation that eliminates IGF-II binding did not show any differences in immunoprecipitation with WT-M from that of their wild-type counterparts. Even when using the triple ligand binding mutant, R2AxI/T-M, in conjunction with these mutated minireceptors, we still did not observe an effect on the ability of the receptors to form heterodimeric complexes, supporting the overall notion that ligand-binding sites do not seem to affect oligomeric complex formation by the M6P/IGF2R.
Based on the finding that the minireceptors were capable of binding to full-length M6P/IGF2R molecules, we sought to determine whether these intermolecular associations were functional with respect to ligand binding. Binding assays showed that the N-terminally truncated minireceptors are able to form functional oligomers with the full-length M6P/IGF2R molecules. In this assay, the FLAG-tagged minireceptors, having no functional Man-6-P binding domains, mediated immunoprecipitation of either WT-M or R2AxI/T-M. To our surprise, WT-M associated with ligand binding-deficient N-terminally truncated minireceptors showed substantial high affinity PMP-BSA binding. Previous data have shown that a single Man-6-P-binding site is not capable of high affinity Man-6-P binding, but rather that two competent sites are required (38). The high affinity binding observed in these experiments could be explained by three possible mechanisms. First, intramolecular binding between the Man-6-P-binding sites on WT-M may cooperate to bind Man-6-P ligands, which agrees with the traditional view of how the M6P/IGF2R functions (6, 37). Second, pairs of dimers attached to the -FLAG affinity resin could be coming into close proximity to one another and forcing the formation of higher ordered oligomeric structures. These larger structures may force two wild-type M6P/IGF2R molecules on different heterodimers to approach closely enough to facilitate high affinity, bivalent Man-6-P binding. Finally, the most intriguing possibility is that the receptor molecules are not naturally forming just dimers but rather oligomers of higher-order structure or mixed clusters. This model would allow for the intermolecular interaction of two Man-6-P-binding sites between WT-M receptors and explain the high affinity binding observed in combinations involving the N-terminally truncated minireceptors with WT-M. Definitive evidence that this binding was contributed by the WT-M partners in such heterodimers was provided by experiments in which PMP-BSA binding was not detectable in dimers formed from N-terminally truncated minireceptors in combination with the triple ligand-binding mutant, R2AxI/T-M.
Similarly, the N-terminally truncated panel transfected with WT-M showed binding to labeled IGF-II in the presence or absence of intact IGF-II-binding sites on the minireceptors. However, co-expression with R2AxI/T-M eliminated IGF-II binding for all combinations involving minireceptors not containing a competent IGF-II-binding site. We also attempted this same series of experiments with the C-terminally truncated receptors, but the results indicated that binding of the radioligands was dependent solely on the ligand binding ability of the FLAG-tagged minireceptors with little binding contribution made by WT-M. The large disparity in length of the C-terminal ends of the C-terminally truncated minireceptors and WT-M may have caused strain on full-length WT-M, possibly forcing WT-M to fold back on itself in these interactions to alleviate interference by the resin beads. This may have led to blocking of or failure to form competent binding sites in our assays.
Our results raise the question whether the M6P/IGF2R dimerizes naturally or as a result of ligand binding. York et al. (37) showed by gel filtration and sucrose gradient sedimentation that the M6P/IGF2R in the presence of the multivalent Man-6-P ligand, -glucuronidase, resulted in a Stokes radius and sedimentation coefficient consistent with the presence of two M6P/IGF2Rs and one -glucuronidase molecule, whereas upon incubation with IGF-II, the receptor appeared to exist as a monomer. Furthermore, experiments testing the internalization of 125I-IGF-II in the presence of -glucuronidase revealed that this ligand accelerated the rate of IGF-II uptake, suggesting that intermolecular cross-linking of receptor molecules could affect internalization (37). On the other hand, Byrd et al. (39) reported, by mutational analysis, that receptors with one functional Man-6-P binding domain were capable of forming high affinity interactions with Man-6-P ligands, suggesting that oligomerization of the M6P/IGF2R contributes to high affinity binding. Furthermore, native gel electrophoresis experiments demonstrated that the M6P/IGF2R could be separated into dimeric and monomeric forms in the presence and/or absence of Man-6-P ligands, and that the two forms of the receptor displayed differing Man-6-P ligand binding characteristics (39). Tong et al. (48) showed that only 1 mol of an oligosaccharide with two phosphomonoesters bound per mol of M6P/IGF2R monomer at saturation, providing the first evidence that the M6P/IGF2R has two high affinity Man-6-P-binding sites. The structural data of Olson et al. (34) have also provided evidence relating to the bivalent binding of Man-6-P-containing ligands by the M6P/IGF2R. Their overall proposed structure does not allow for a diphosphorylated oligosaccharide to bind to repeats 3 and 9 within a single receptor because of steric considerations, but rather that high affinity Man-6-P binding must be due to the diphosphorylated oligosaccharide spanning binding sites contributed by two adjacent M6P/IGF2R molecules (34).
The ligand binding data from our N-terminally truncated minireceptors favor a mechanism of M6P/IGF2R association in which the receptor can form oligomeric structures in the absence of ligands but that oligomerization does promote high affinity ligand binding (39). In our studies, the N-terminally truncated minireceptors were able to associate with full-length receptors in the absence of ligands according to immunoprecipitation assays. These associated receptors were able to bind PMP-BSA, even though none of the N-terminally truncated minireceptors contain a Man-6-P binding domain, suggesting that the receptors did not require the presence of Man-6-P ligands to associate. This ligand binding was not evident in associations between the N-terminally truncated minireceptors and the triple-mutant R2AxI/T-M receptor molecules. Our immunoprecipitation data with the R2AxI/T-M receptor suggest that the M6P/IGF2R forms oligomeric complexes regardless of the ligand binding state of the receptor. The oligomeric state of the receptor likely influences the affinity of bivalent Man-6-P-containing ligands for the receptor to a greater degree than bivalent Man-6-P binding promotes oligomerization.
In conclusion, we report, for the first time, that severely truncated minireceptors can associate with the full-length M6P/IGF2R regardless of the presence of ligand binding domains. Additionally, we have shown that associations between the N-terminally truncated minireceptors and WT-M are functional by the criterion of ligand binding. Future studies should focus on how mutated and/or truncated M6P/IGF2Rs influence the receptor concerning the possibility of how "oligomer interference" would affect functional properties of the receptor. Additionally, further studies are needed to address definitively the oligomerization state of the receptor, i.e. to determine whether the receptor forms dimers or high order structures and to yield a better understanding of the mechanisms by which the oligomeric state of the M6P/IGF2R contributes to its functional properties.
|
FOOTNOTES |
|---|
Recipient of pre-doctoral stipend support provided by Graduate Studies; Bukey, McDonald, Emley, and Widaman fellowships; the Dr. Fred W. Upson grant-in-aid award through the University of Nebraska Medical Center, and the NASA Space Grant Scholarship.
Present address: Washington University School of Medicine, Rheumatology Division, 4921 Parkview Place, Campus Box 8045, St. Louis, MO 63110.
¶ To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, University of Nebraska Medical Center, 985870 Nebraska Medical Center, Omaha, NE 68198-5870. Tel.: 402-559-7824; Fax: 402-559-6650; E-mail: rgmacdon{at}unmc.edu.
1 The abbreviations used are: M6P/IGF2R, mannose 6-phosphate/insulin-like growth factor II receptor; Man-6-P, mannose 6-phosphate; EC, extracytoplasmic; TGF-1, transforming growth factor-1; IGF-II, insulin-like growth factor II; uPAR, urokinase-type plasminogen activator receptor; -, anti-; PMP-BSA, pentamannose phosphate-bovine serum albumin; hGUS, human -glucuronidase; WT-M, full-length wild-type M6P/IGF2R-Myc; R2AxI/T-M, full-length triple-mutant M6P/IGF2R-Myc; nt, nucleotide(s); pBSKII+, pBluescript SK II+; DTT, dithiothreitol; HBS, HEPES-buffered saline; EGFR, epidermal growth factor receptor.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
