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

J. Biol. Chem., Vol. 275, Issue 25, 18647-18656, June 23, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/25/18647    most recent M001273200v1 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 Byrd, J. C. Articles by MacDonald, R. G. Search for Related Content PubMed PubMed Citation Articles by Byrd, J. C. Articles by MacDonald, R. G. Social Bookmarking                      What's this?

Dimerization of the Insulin-like Growth Factor II/Mannose 6-Phosphate Receptor^*

James C. Byrd, Jung H. Y. Park^§, Beverly S. Schaffer, Farideh Garmroudi^¶, and Richard G. MacDonald

From the  Department of Biochemistry and Molecular Biology, University of Nebraska Medical Center, Omaha, Nebraska 68198-4525 and the ^§ Division of Life Sciences, Hallym University, Chunchon 200-702, South Korea

Received for publication, February 15, 2000, and in revised form, March 31, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The insulin-like growth factor II/mannose 6-phosphate receptor (IGF2R) interacts with lysosomal enzymes through two binding domains in its extracytoplasmic domain. We report in the accompanying article (Byrd, J. C., and MacDonald, R. G. (2000) J. Biol. Chem. 275, 18638-18646) that only one of the two extracytoplasmic mannose 6-phosphate (Man-6-P) binding domains is necessary for high affinity Man-6-P ligand binding, suggesting that, like the cation-dependent Man-6-P receptor, oligomerization of the IGF2R contributes to high affinity interaction with lysosomal enzymes. In the present study, we have directly characterized both naturally occurring and engineered forms of the IGF2R for their ability to form oligomeric structures. Whereas gel filtration chromatography suggested that purified bovine IGF2R species exist in a monomeric form, native gel electrophoresis allowed for the separation of dimeric and monomeric forms of the receptors with distinct phosphomannosyl ligand binding characteristics. The ability of the IGF2R to form oligomeric complexes was confirmed and localized to the extracytoplasmic domain through the use of epitope-tagged soluble IGF2R constructs bearing deletions of the transmembrane and cytoplasmic domains. Finally, chimeric receptors were engineered containing the extracytoplasmic and transmembrane domains of the IGF2R fused to the cytoplasmic domain of the epidermal growth factor receptor with which dimerization of the chimeras could be monitored by measuring autophosphorylation. Collectively, these results show that the IGF2R is capable of forming oligomeric complexes, most likely dimers, in the absence of Man-6-P ligands.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The insulin-like growth factor II/mannose 6-phosphate receptor (IGF2R)¹ is a 300-kDa type-I transmembrane glycoprotein that comprises a short NH₂-terminal signal sequence, followed by 15 homologous repeats, a transmembrane domain, and a 167-residue cytoplasmic domain (1, 2). Through its ability to bind several distinct ligands, this receptor is thought to carry out multiple functions in cellular physiology. The IGF2R binds insulin-like growth factor II (IGF-II) at the cell surface, resulting in the internalization and degradation of this mitogen in the lysosomal compartment (3-6). The receptor also binds urokinase-type plasminogen activator receptor, which may be involved in the activation of latent transforming growth factor- (7-9). Finally, the receptor interacts with proteins that bear the Man-6-P marker, resulting in sorting to the lysosomal compartment. Functional mapping studies of the extracytoplasmic domain of the IGF2R have revealed the location of two distinct binding domains for Man-6-P, localized to repeats 1-3 and 7-9 of the extracytoplasmic domain (10, 11), which may have distinct specificities for sorting lysosomal enzymes (12).

The IGF2R functions in the biogenesis of lysosomes and in the uptake and degradation of IGF-II, by transporting its cargo between three different compartments. This receptor cycles between the trans-Golgi network, the endosomal compartment, and the cell surface (for review see Ref. 13). Binding of ligands to the IGF2R results in their targeting via clathrin-coated vesicles to the endosomal compartment, where the decreased pH results in dissociation and packaging of the ligands into prelysosomal vesicles (14-16). Trafficking and recycling of the IGF2R occurs constitutively, even in the absence of ligand (4). However, the addition of exogenous Man-6-P-bearing proteins has recently been shown to increase the internalization rate of the IGF2R from the cell surface, possibly through receptor dimerization (17). In addition, we have demonstrated that receptor oligomerization may function in the binding of multivalent Man-6-P ligands, allowing for high affinity interaction with the IGF2R (54).

Studies of the oligomeric nature of the IGF2R to this point, however, have been inconsistent. Characterization of purified forms of the receptor by gel filtration and sucrose density gradient centrifugation has suggested that the receptor exists in solution in a monomeric form (17, 18), whereas chemical cross-linking studies have demonstrated that the receptor forms aggregates in the membranes of intact cells (19). Recent observations, however, suggest that the IGF2R may bind multivalent Man-6-P-bearing lysosomal enzymes by forming an oligomeric receptor complex. Addition of exogenous Man-6-P-bearing proteins influences the oligomeric nature of the receptor in vitro, as measured by gel filtration chromatography, and alters the rate of receptor trafficking in cell culture models (17). In addition, the repeat 3 and repeat 9 Man-6-P binding domains of the IGF2R demonstrate distinct preferences in their ability to sort lysosomal enzymes, which could be explained if the Man-6-P binding domains of the receptor interact in an oligomer to form two unique binding sites (12). Finally, we report in the accompanying article (54) that the IGF2R displayed negative cooperativity in binding a multivalent Man-6-P-bearing ligand, and that the repeat 3 and repeat 9 Man-6-P binding domains can form high affinity binding sites independently of each other. However, the mechanism of IGF2R oligomerization is poorly understood. It has remained unclear if binding of ligand stimulates dimerization of the IGF2R or if the presence of an existing receptor dimer allows for high affinity Man-6-P binding.

Receptor oligomerization plays an important role in the function of many other membrane proteins. Signal transduction through many of the tyrosine kinase receptors is controlled by receptor dimerization (20). In addition, for molecules such as the insulin family of receptors, constitutive receptor dimers are not only important for signal transduction, but for ligand binding as well (21). Finally, receptor oligomerization confers high affinity Man-6-P-specific lysosomal enzyme binding on the related cation-dependent mannose 6-phosphate receptor (CD-MPR) (22). These examples suggest that the oligomeric state of the IGF2R may be important in the overall function of this ubiquitously expressed receptor. In this report, we have studied the ability of the IGF2R to form dimeric complexes in the presence and absence of Man-6-P ligands using both naturally occurring forms of the IGF2R as well as human IGF2R constructs encompassing portions of the receptor's extracytoplasmic domain. The natural state of the IGF2R appears to be a constitutive dimer, as this receptor formed oligomeric complexes through its extracytoplasmic domain in a ligand-independent manner.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Oligonucleotides were purchased from Integrated DNA Technologies, Inc. (Coralville, IA). The native Y-2448 O-phosphomannan of Hansenula holstii was a gift from Dr. M. E. Slodki (Midwest Area Northern Regional Research Center, Peoria, IL, retired). D-Mannose 6-phosphate (Man-6-P) disodium salt, D-glucose 6-phosphate (Glc-6-P), and the -FLAG M2 antibody reagents were purchased from Sigma. Radiolabeled pentamannose phosphate-bovine serum albumin (PMP-BSA) and IGF-II were prepared from sources as described previously (23). The pCMV5 vector (24) was provided by Dr. David W. Russell (University of Texas Southwestern Medical Center, Dallas, TX). The 8.6-kilobase pair human IGF2R cDNA (1) was a gift of Dr. William S. Sly (St. Louis University Medical Center, St. Louis, MO). The EGFRvIII cDNA was from the laboratory of Dr. Surinder K. Batra (University of Nebraska Medical Center, Omaha, NE). Both the -Myc antibody and a cDNA encoding a carboxyl-terminal FLAG-tagged version of the kinase suppressor of Ras (pCMV5/KSRF) (25) were gifts of the laboratory of Dr. Robert E. Lewis (University of Nebraska Medical Center, Omaha, NE). Restriction endonucleases were purchased from New England Biolabs (Beverly, MA). Other reagents and supplies were obtained from sources as indicated.

Purification of IGF2R Species from Fetal Bovine Serum and Bovine Liver-- Two forms of the IGF2R were purified from bovine sources. First, the soluble IGF2R (sIGF2R) was purified from outdated lots of fetal bovine serum (FBS, HyClone Laboratories, Logan, UT) following a modified procedure of Valenzano et al. (26). Briefly, 500 ml of FBS was diluted 2-fold with buffer to a final concentration of 25 mM HEPES, pH 7.4, 150 mM NaCl, 0.1% Triton X-100, and 5 mM -glycerophosphate, and passed twice over a 5-ml pentamannose phosphate (PMP)-Sepharose affinity column, which was prepared as described previously (23). The column was then washed five times with 5 ml of column buffer containing 5 mM glucose 6-phosphate. The bound sIGF2R was then eluted from the column with 25 ml of column buffer containing 10 mM Man-6-P. Fractions containing the sIGF2R were pooled and lyophilized. Purification of the sIGF2R yielded about 7-10 mg of protein/liter of FBS, determined by Coomassie Blue staining of samples analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

The full-length IGF2R was purified from bovine liver according to a modified procedure of York et al. (17). After homogenization of 220 g of bovine liver and extraction in a buffer containing 50 mM imidazole, pH 7, 150 mM NaCl, 5 mM -glycerol phosphate, 2% Triton X-100, 0.25% sodium deoxycholate, 10 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 20 µg/ml aprotinin, 10 µg/ml antipain, 80 µg/ml benzamidine, and 10 µg/ml leupeptin, the full-length IGF2R was affinity-purified on PMP-Sepharose as described for the soluble receptor. Samples containing the receptor were lyophilized and stored at 20 °C until use. This procedure yielded about 1.5 mg of purified IGF2R, as determined by Coomassie Blue staining of material analyzed by SDS-PAGE. After affinity purification, aliquots of the lyophilized IGF2R species (approximately 0.75 mg) were dissolved in 150 mM ammonium acetate, 250 mM acetic acid, pH 4.5, and applied to a 20-ml Superose 12 column equilibrated under the same acidic conditions as described by Valenzano et al. (26). Elution of the purified receptor was monitored by both absorbance at 280 nm and with SDS-PAGE of the fractions and Coomassie Blue staining (data not shown). This acidic gel filtration step was used to remove any Man-6-P and other phosphomannosylated ligands that may have copurified with the IGF2R during the affinity chromatography. Fractions containing the receptors were lyophilized and stored at 20 °C until use.

Fast Protein Liquid Chromatography and Native Gel Analysis of Purified Bovine Receptors-- The oligomeric state of both the sIGF2R and full-length IGF2R species was analyzed with fast protein liquid chromatography (FPLC) analysis on an analytical 20-ml Superose-12 column (Amersham Pharmacia Biotech) equilibrated with column buffer (50 mM imidazole, pH 7.4, 150 mM NaCl, 5 mM sodium -glycerophosphate, and 0.05% Triton X-100). Aliquots (0.2 mg) of the lyophilized, affinity-purified receptors were resuspended in 0.2 ml of water and run on the column. Elution of the purified receptor was monitored by absorbance at 280 nm. Stokes radius measurements were made by comparing elution profiles of the receptors to those of protein standards: thyroglobulin, ferritin, catalase, aldolase, and bovine serum albumin (BSA) (Amersham Pharmacia Biotech).

Native gel electrophoresis of the receptors was carried out by a modified procedure of Kuehn et al. (27). Aliquots (1-10 µg protein) of sIGF2R or IGF2R directly eluted from the PMP-Sepharose column were electrophoresed on native PAGE composed of a 4-12% linear gradient containing 0.1% Triton X-100. After electrophoresis, protein was detected with Coomassie Blue staining. Stokes radius estimates were made by comparing the mobility of the receptor to that of native gel standards (Amersham Pharmacia Biotech). To determine the role that Man-6-P and other phosphomannosylated ligands play in the topology of the receptor's extracytoplasmic domain, aliquots (0.5 mg) of the purified soluble IGF2R were resuspended in water and dialyzed for 24 h against HBST (25 mM HEPES, pH 7.4, 150 mM NaCl, 0.05% Triton X-100) using Slide-A-Lyzer cassettes (Pierce) to remove the Man-6-P. In addition, a pseudoglycoprotein containing multiple Man-6-P moieties, PMP-BSA, was synthesized as previously reported (23). Aliquots of the purified sIGF2R (1-10 µg of protein) were incubated with a range of PMP-BSA concentrations (from 1-8000 nM) in 30 µl of HBST for 3 h at 3 °C and then subjected to native PAGE or cross-linked with 2.5 mM disuccinimidyl suberate (Pierce) at 3 °C for 15 min and resolved by 6% reducing SDS-PAGE followed by Coomassie Blue staining.

Western Ligand Blotting-- Ligand blotting was performed using a modified procedure published earlier for the detection of IGF-binding proteins (28). The purified sIGF2R and full-length IGF2R (~ 1 µg for each lane) were electrophoresed on native 4-12% gradient gels as described above, then transferred to BA85 nitrocellulose (Schleicher & Schuell). The blots were washed and blocked with 1% BSA. Affinity for IGF-II and PMP-BSA was detected by probing the blots with 1.5 × 10⁶ cpm ¹²⁵I-PMP-BSA or ¹²⁵I-IGF-II in 8 ml of blocking solution for 16 h at 3 °C. The blots were then washed and exposed to x-ray film. Intensity of ligand binding was determined using a densitometer with ImageQuant software (Molecular Dynamics, Sunnyvale, CA).

Preparation and Expression of Truncated IGF2R Constructs-- A soluble construct containing all 15 extracytoplasmic repeats of the IGF2R followed by a FLAG epitope, 15F, was engineered using the hIGF2R cDNA as described previously (23). A similar construct, 15myc, was generated using the same strategy with a COOH-terminal Myc epitope tag (MEQKLISEEDLN) (29) engineered in place of the FLAG tag. Transient expression of these truncated constructs was done by calcium phosphate precipitation following a previously described method in 293T human embryonic kidney cells (23). Coexpression of the 15F and 15myc constructs was performed by mixing equal amounts of the cDNAs (30 µg each/100-mm dish containing approximately 2.5 million cells) prior to transfection. Serum-free conditioned media (24 h) and 1% Triton X-100 cell lysates were prepared on the 5th or 6th day after transfection as described previously (30). To confirm expression of the FLAG- and Myc-tagged constructs, aliquots of both conditioned media (75 µl) or cell lysates (30 µl) were analyzed by immunoblotting with either the -FLAG M2 antibody or the -Myc 9E10 monoclonal antibody, or both, using a previously reported procedure (23).

Coimmunoprecipitation of the 15F and 15myc Constructs-- The ability of the FLAG- and Myc-tagged receptor constructs to interact was measured in a coimmunoprecipitation procedure (23). Routinely, aliquots of cell lysates (20-40 µl) containing the expressed constructs were incubated with 12 µl of packed M2 -FLAG resin in 25 mM HEPES, pH 7.4, and 150 mM NaCl (HBS) with 0.5% BSA at 3 °C for 2-4 h. Following immunoadsorption, the resin pellets were collected by centrifugation at 14,000 × g for 10 s and then washed twice with 0.75 ml of HBST. The resultant resin pellets could then be analyzed by immunoblotting to determine if the FLAG- and Myc-tagged receptor constructs were capable of coimmunoprecipitation.

In order to study the dynamics of complex formation between the 15F and 15myc constructs in vitro, each species was purified separately from transiently transfected 293T cells on PMP-Sepharose affinity columns, as has been described for the soluble IGF2R above. The constructs were eluted from the affinity columns using 10 mM Man-6-P, lyophilized, and stored at 20 °C until use. Aliquots of the purified receptor constructs were mixed and incubated for increasing lengths of time (0-8 h) at 37 °C. The amount of 15F and 15myc present in a complex was measured using the coimmunoprecipitation procedure described above.

Preparation of FLAG-tagged IGF2R/EGFR Chimeric Receptor Constructs and a Myc-tagged IGF2R-- The cytoplasmic domain of the human epidermal growth factor receptor (EGFR) was fused to the extracytoplasmic and transmembrane domains of the IGF2R to create a chimeric IGF2R/EGFR construct. A construct containing the IGF2R cDNA, lacking the EagI fragment between nt 162 and 5319 (23), was used as a template for amplification by Vent^TM polymerase using a 5' primer containing a KpnI restriction site preceding the sequence corresponding to nt 94-113 of the IGF2R cDNA and a 3' primer that represented sequence complementary to nt 7114-7134 followed by a HindIII site. The resultant product of the amplification was digested with KpnI and HindIII and subcloned into pCMV5. Next, the EagI fragment from the wild-type IGF2R cDNA was subcloned in, completing the extracytoplasmic portion of the IGF2R half of the chimera. Finally, the cytoplasmic domain of the EGFR was amplified using Pfu polymerase with the variant III human EGFR cDNA (EGFRvIII) as a template and two primers. The 5'-primer contained a HindIII site followed by sequence corresponding to nt 2206-2221 of the human EGFR cDNA (31), and the 3'-primer was complementary to nt 3802-3813 followed by the FLAG epitope, a UAG stop codon, and an XbaI site. The resultant product was digested with HindIII and XbaI and subcloned into the pCMV5 vector containing the IGF2R portion of the chimera, completing the construct.

The 11-TM/EGFR and 13-TM/EGFR chimeras were prepared utilizing a similar strategy. The region of the IGF2R extracytoplasmic domain to be included in these constructs was amplified from the full-length IGF2R cDNA using 5'-primers containing an EcoRI site followed by sequence corresponding to nt 4675-4692 for the 11-TM/EGFR or nt 5542-5560 for the 13-TM/EGFR. The 3'-primer for both constructs represented sequence complementary to nt 7114-7134 of the IGF2R cDNA followed by a HindIII site. The resultant product of the amplification was digested with EcoRI and HindIII and subcloned into CMV5RIX (30), which contains the first half of repeat 1. The cytoplasmic portion of the EGFR cDNA was isolated from the IGF2R/EGFR construct after digestion with HindIII and XbaI and subcloned into both the CMV5RIX/11-TM and CMV5RIX/13-TM plasmids, completing the 11-TM/EGFR and 13-TM/EGFR chimeras. The IGF2R portion of the 13-TM/EGFR was sequenced to confirm that its phenotype was not due to mutations occurring in the extracytoplasmic or transmembrane domains.

A Myc epitope tag was put on the carboxyl terminus of the IGF2R cDNA through a similar procedure as described for preparation of the 15myc construct. The IGF2R cDNA lacking the EagI fragment was used as a template for amplification with Vent^TM polymerase with a 5' primer containing an XhoI site preceding sequence corresponding to nt 94-113 of the receptor cDNA and 3' primer containing sequence complimentary to 7602-7620 of the IGF2R cDNA followed by 36 nt encoding the Myc epitope, a UAG stop codon, and an XbaI site. The resultant product was digested with XhoI and XbaI, subcloned into pBKCMV (Invitrogen, Carlsbad, CA), and then digested with HindIII and XbaI so that it could be cloned into the pCMV5 vector. Finally, the wild-type EagI fragment was subcloned in, completing the IGF2Rmyc construct.

Expression of Chimeras and IGF2Rmyc in 293T Cells-- To study the tyrosine phosphorylation of the chimeras, the constructs were transiently expressed in 293T cells using the modified calcium phosphate transfection procedure. In addition, cotransfection of the chimeras with IGF2Rmyc was conducted by mixing the same amount of the chimera cDNA (20 µg/100-mm dish) with various amounts of the IGF2Rmyc cDNA (0.5-40 µg/100-mm dish) prior to the transfection. Varying the amount of input IGF2Rmyc cDNA allowed for control of the levels of expression of the IGF2Rmyc in relation to the chimeras in the cell population. The transfection medium was replaced 24 h after the transfection with serum-free Dulbecco's modified Eagle's medium (DMEM) or DMEM containing 5 mM Man-6-P, 25 nM PMP-BSA, or 5 mM Glc-6-P. To ensure a consistent level of exogenous ligands, the cells were again fed with media containing 5 mM Man-6-P, 25 nM PMP-BSA, or 5 mM Glc-6-P 3 h prior to collection of cell lysates at 51 h after transfection. Cell lysates were prepared following the procedure of Meng and Lin (32). Each 100-mm dish of 293T cells was scraped, and the cells were washed with 2 ml of ice-cold HBS and then lysed in 400 µl of ice-cold lysis buffer containing phosphatase and protease inhibitors (20 mM HEPES, pH 7.5, 0.1% SDS, 0.5% deoxycholate, 1% Nonidet P-40, 150 mM NaCl, 4 mM EDTA, 10 mM NaF, 0.1 mM ZnCl₂, 10 mM Na₄P₂O₇, 2 mM sodium orthovanadate, 1 mM phenylmethylsulfonyl fluoride, 20 µg/ml aprotinin, 10 µg/ml antipain, 80 µg/ml benzamidine, and 10 µg/ml leupeptin). Aliquots (20 µl) were then subjected to immunoblot analysis with the -FLAG or -Myc antibodies and developed with ¹²⁵I-protein A to determine the relative levels of expression via PhosphorImager analysis.

Analysis of the Phosphorylation and Coimmunoprecipitation of Transiently Expressed Chimeras and IGF2Rmyc-- Equimolar amounts of the chimeric receptor constructs from cell lysates of each transfection set, or 30 µl of pCMV5-transfected control cell lysates, were immunoadsorbed to 15 µl of -FLAG M2 resin in the presence of HBS + 1% BSA for 2.5 h at 3 °C. The pellets were then collected and washed twice with 0.75 ml of ice-cold HBST. They were then electrophoresed on reducing 6% SDS-PAGE gels and transferred to BA 85 nitrocellulose. The blots were then subjected to the immunoblot procedure with either -FLAG M2 antibody, -Myc 9E10 antibody, or the -phosphotyrosine (-Tyr(P)) 4G10 antibody (Upstate Biotechnology, Lake Placid, NY) and developed by ¹²⁵I-protein A followed by autoradiography.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Analysis of Purified Bovine sIGF2R and Full-length IGF2R by Gel Filtration FPLC and Native Gel Electrophoresis-- Two naturally occurring forms of the IGF2R were purified from bovine sources so that they could be analyzed for the ability to form oligomeric structures. The soluble form of the IGF2R, which is present in the serum of several species (26, 33-35), and full-length IGF2R were purified from FBS and Triton X-100 liver extracts, respectively, by PMP-Sepharose affinity chromatography. The receptors were eluted from the PMP-Sepharose columns with 10 mM Man-6-P, and fractions were analyzed by SDS-PAGE under reducing conditions followed by Coomassie Blue staining to detect the purified product (Fig. 1A). The purified forms demonstrated a major band at the appropriate molecular weights: 250,000 for the sIGF2R and 300,000 for the full-length IGF2R. The minor bands of lower molecular weight were not observed when the gels were run under non-reducing conditions, suggesting that they are cleavage products of the receptor (data not shown). In order to estimate the Stokes radii of the two different receptor species, they were first analyzed by gel filtration chromatography on a 20-ml Superose 12 FPLC column (Fig. 1B). Despite their apparent purity, both species eluted with heterogeneous peaks that spanned wide size distributions. The major peak of the full-length IGF2R eluted from the column with a Stokes radius of approximately 66 Å, which is lower than previous reports of 79 Å (17) and 72 Å (18) for the monomeric bovine IGF2R. Surprisingly, the major peak of the sIGF2R demonstrated a much lower Stokes radius of about 49 Å.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   Gel filtration FPLC and native gel electrophoresis of purified bovine IGF2Rs. A, both the sIGF2R and the full-length IGF2R were purified from FBS and bovine liver extracts, respectively, on PMP-Sepharose affinity columns. Aliquots (20 µl) of each fraction from the column were loaded on a reducing SDS-PAGE gel and stained with Coomassie Blue. The first three fractions eluting from each PMP-Sepharose column are shown. B, the purified receptors (approximately 200 µg each) were analyzed on a Superose 12 FPLC gel filtration column. Elution of the purified receptors was monitored by UV absorbance at 280 nm. The elution profiles of protein standards: thyroglobulin, ferritin, aldolase, catalase, and BSA, are shown on the elution profile. C, aliquots (~10-20 µg) of either the full-length IGF2R eluted from the PMP-Sepharose column (lane 1), or IGF2R that underwent further purification through an acidic gel filtration column (lane 2), or soluble IGF2R eluted from the PMP-Sepharose column (lane 3), were analyzed by native gel electrophoresis on a 4-12% gradient PAGE gel, and the proteins were detected by Coomassie Blue staining.

The second method used to determine the structure of the IGF2R was native gel electrophoresis. Aliquots (20-30 µg) of the same concentrated samples used in the FPLC analysis were loaded onto a 4-12% native gel containing 0.1% Triton X-100 (Fig. 1C, lane 1). In addition, aliquots (20-30 µg) of full-length IGF2R, which had been further purified by gel filtration chromatography under acidic conditions, were similarly analyzed (Fig. 1C, lane 2). Following electrophoresis, protein was detected by Coomassie Blue staining. Like the FPLC analysis, the native gel demonstrated that the full-length IGF2R exists in multiple forms. In contrast to the FPLC results, the full-length receptor demonstrated a major band consistent with a dimeric form having a Stokes radius greater than that of thyroglobulin (85 Å), with the highest mobility band corresponding to 66 Å. Interestingly, the sIGF2R demonstrated variability in the amount of this dimeric complex (from approximately 0% to 15%) when analyzed under these conditions (Fig. 1C, lane 3). The majority of the sIGF2R appeared in a band identical to the monomeric full-length IGF2R with a Stokes radius of approximately 66 Å, which is considerably larger than the Stokes radius measured using the FPLC analysis.

To determine if the possible presence of phosphomannosylated proteins bound to the receptor may have had an impact on the amount of sIGF2R dimer detectable by the native gel analysis, aliquots (~10 µg) of purified sIGF2R that had been dialyzed to remove the Man-6-P were incubated in the presence of 1-8000 nM PMP-BSA. Following a 3-h incubation at 3 °C, the samples were either directly resolved by native gel electrophoresis or were cross-linked with 2.5 mM disuccinimidyl suberate and resolved on a 6% reducing SDS-PAGE gel, followed by Coomassie Blue staining. Approximately 10% of the sIGF2R was cross-linked as a dimer (data not shown), and the addition of increasing concentrations of PMP-BSA had no effect on the amount of dimer detected by either procedure (data not shown).

The Monomeric and Dimeric IGF2R Display Different Affinities for PMP-BSA-- Separation of the monomeric and dimeric forms of the bovine IGF2Rs using native PAGE allowed us to probe their ability to bind ¹²⁵I-PMP-BSA and ¹²⁵I-IGF-II. After native gel electrophoresis, the protein was transferred to nitrocellulose, and the blots were then incubated with either radiolabeled PMP-BSA or IGF-II. For both the soluble and full-length receptors, ¹²⁵I-IGF-II bound to all forms with the same stoichiometry as detected by Coomassie Blue staining (Fig. 2). When probed with ¹²⁵I-PMP-BSA, the ligand blot demonstrated that the full-length IGF2R dimer was the only form of the receptor that could interact with high affinity (Fig. 2A). Binding of ¹²⁵I-PMP-BSA to the monomeric form of the full-length receptor was detected, but required a 10-fold longer exposure to x-ray film (data not shown). Likewise, albeit not as marked an effect, the soluble receptor demonstrated a 6-fold higher stoichiometry of PMP-BSA binding to the dimeric form than would be predicted from the Coomassie Blue staining and ¹²⁵I-IGF-II ligand blot (Fig. 2B).

View larger version (45K): [in this window] [in a new window]   Fig. 2.   Ligand blot analysis of native IGF2R species. Aliquots (5 µg protein) of purified full-length IGF2R (A) or sIGF2R (B) were electrophoresed in either single or duplicate loadings on native gels as described under "Experimental Procedures." The gels were then stained with Coomassie Blue or transferred to nitrocellulose. The blots were probed for binding of either 125I-IGF-II or 125I-PMP-BSA, as indicated, and developed by autoradiography. The positions of the proposed monomeric and dimeric forms of both receptor species are indicated with arrows.

Epitope-tagged Soluble Receptor Constructs Are Capable of Forming Non-covalent Complexes When Coexpressed in 293T Cells-- In view of the opposing findings on the receptor size estimates obtained by gel filtration FPLC analysis and native gel electrophoresis, we were compelled to determine the oligomeric nature of the IGF2R by yet another approach. Two soluble constructs, 15F and 15myc, were engineered to contain the extracytoplasmic domain of the IGF2R followed by either a FLAG or Myc epitope tag. These constructs were expressed alone or coexpressed by mixing their cDNAs prior to calcium phosphate transfection in 293T cells. On the 5th or 6th day after addition of the cDNAs to the cell culture, cell extracts were prepared, and expression was confirmed by immunoblot using the -FLAG or the -Myc antibodies (Fig. 3A).

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Coimmunoprecipitation of 15F and 15myc IGF2R constructs. A, cell lysates (25 µl) from 293T cells transiently transfected with 15F, 15myc, or both (as indicated in the boxes) were analyzed by immunoblotting with either the -FLAG or the -Myc antibody in duplicate, as indicated, to confirm expression of the appropriate epitope-tagged receptor constructs. B, the ability of 15myc to coimmunoprecipitate with 15F was measured by immunoprecipitating equimolar amounts of the 15F construct with -FLAG resin from the cotransfected cell lysates. At the end of the immunoprecipitation, the resin pellets were collected, washed, heated with sample buffer, and analyzed by 6% reducing SDS-PAGE. The proteins were then transferred to BA85 nitrocellulose and immunoblotted with -Myc. As a control, lysates from cells transfected with both 15F and 15myc equivalent to the total protein in each immunoprecipitation were loaded directly on the gel (lane 1). In addition, cell lysates containing only the 15myc construct were subjected to the immunoprecipitation procedure (lane 2), as were cell lysates containing only the 15F construct (lane 3). Lanes 4-6 were immunoprecipitations of lysates from cells that coexpressed 15F and 15myc conducted in the presence of no additions (lane 4), 1 µM IGF-II (lane 5), or 5 mM Man-6-P (lane 6). C, mixing of purified 15F and 15myc, followed by incubation at 37 °C, resulted in an increase in the amount of 15F·15myc heterocomplex. 15F and 15myc were purified separately on PMP-Sepharose affinity columns. Aliquots of the purified constructs were mixed and incubated for the indicated times at 37 °C before being immunoprecipitated with -FLAG resin. The resin pellets were collected, washed, and analyzed by 6% reducing SDS-PAGE and subsequent immunoblot analysis with -Myc.

To determine if the coexpressed receptor constructs exist in an oligomeric complex, cell lysates containing the constructs were immunoadsorbed to -FLAG resin in the presence or absence of 1 µM IGF-II or 5 mM Man-6-P. The resin pellets were then washed to remove unbound construct, loaded onto a reducing SDS-PAGE gel, and analyzed by immunoblotting with the -Myc antibody to determine if the 15myc receptor construct interacts with the 15F construct (Fig. 3B). Neither the 15F nor the 15myc constructs were detected in the immunoblot analysis of lysates from cells transfected with either construct alone, but the cell lysates containing the coexpressed receptor constructs showed that the 15myc and 15F proteins form a non-covalent complex when they are expressed together. Surprisingly, the addition of exogenous IGF-II or Man-6-P did not affect the amount of complex formed when added during the immunoprecipitation. PhosphorImager analysis of the blot revealed that approximately 50% of the expressed 15myc coimmunoprecipitated with 15F, suggesting that the majority of the 15myc existed as either 15myc homodimers or as heterodimers with the 15F construct.

To determine if the interaction between the 15F and 15myc receptors could occur after mixing the proteins in vitro, each construct was purified separately by PMP-Sepharose affinity chromatography. Aliquots (50-100 µl) of the purified receptor constructs were mixed and incubated for increasing lengths of time (from 0 to 8 h) at 37 °C prior to immunoprecipitating with -FLAG resin, and the presence of coimmunoprecipitated 15myc protein was detected by immunoblotting with the -Myc antibody (Fig. 3C). The purified 15F and 15myc proteins did not interact without incubating at 37 °C. However, the amount of 15myc detected in a complex with 15F progressively increased as a function of time at 37 °C. It should be noted that this experiment was conducted in the presence of approximately 10 mM Man-6-P, a condition under which binding of other phosphomannosylated ligands should be minimized. Interestingly, conducting the in vitro mixing experiments in up to 30% dimethyl sulfoxide, which would suppress the dipolar nature of the solvent and would be predicted to disrupt hydrophobic contributions to receptor oligomers, did not change the rate of formation of the 15F·15myc protein complexes (data not shown). In addition, the 15F·15myc protein complexes immunoadsorbed from lysates of cells cotransfected with the two constructs did not dissociate when incubated for 30 min over a pH range of 5.0-7.5 (data not shown).

Tyrosine Phosphorylation of an IGF2R/EGFR Chimera Is Inhibited by Expression of a Myc-tagged IGF2R Construct-- To study the oligomeric nature of the IGF2R in the context of the cell membrane, we engineered several chimeric proteins composed of portions of the IGF2R extracytoplasmic and transmembrane domains fused to the cytoplasmic domain of the human EGFR followed by a FLAG-epitope tag (Fig. 4). The rationale behind these experiments was based on the observation that the extracytoplasmic domain of the IGF2R was capable of mediating receptor oligomerization in a ligand-independent manner. If this oligomerization occurred in the cell membrane, then the chimeric proteins should exhibit ligand-independent autophosphorylation of tyrosyl residues in their EGFR-derived cytoplasmic domain, due to transphosphorylation (37). The first chimera studied contained the entire IGF2R extracytoplasmic domain fused to the EGFR cytoplasmic domain. This construct, called the IGF2R/EGFR chimera, was transiently expressed in 293T cells so that its tyrosine phosphorylation level could be measured. Parallel transfections were also carried out with the chimera and a human IGF2R cDNA containing a c-Myc epitope tag at its carboxyl terminus. One day after the addition of the cDNAs, and again 3 h prior to preparation of cell lysates, the transfection medium was replaced with serum-free medium containing either Man-6-P, PMP-BSA, or Glc-6-P to test the ability of these ligands to influence the phosphotyrosine levels of the chimera.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Schematic of the IGF2R/EGFR chimeric constructs. Illustration of the fusion of the extracytoplasmic and transmembrane domains (residues 1-2329) of the human IGF2R (solid overline) to the cytoplasmic domain (residues 650-1186) of the human EGFR (broken overline) containing the tyrosine kinase domain. The 15 repeating units of the IGF2R extracytoplasmic domain are represented by rectangles. The Man-6-P binding domains, repeats 1-3 and 7-9, are shown as diagonally shaded boxes, and the repeats responsible for high affinity IGF-II binding are represented as gray boxes. Both the 11-TM/EGFR and 13-TM/EGFR constructs contain the signal sequence and the first half of repeat one preceding the truncated IGF2R extracytoplasmic domain. The truncated regions of 11-TM/EGFR and 13-TM/EGFR are illustrated by dashed lines. Additionally, the transmembrane (TM) domain of the IGF2R is labeled as well as the FLAG epitope at the carboxyl terminus.

To measure the phosphorylation of the IGF2R/EGFR chimera under the different conditions, the cell lysates were immunoadsorbed to M2 -FLAG resin and analyzed by immunoblotting with either -FLAG or -phosphotyrosine (-Tyr(P)) antibodies (Fig. 5). Lysates from cells transfected with pCMV5, the chimera, or the chimera plus the IGF2Rmyc construct were also subjected to direct immunoblot analysis using an -Myc antibody to confirm the expression of the IGF2Rmyc construct. As predicted for an oligomeric complex, the IGF2R/EGFR chimera demonstrated a strong signal with the -Tyr(P) antibody, indicating its phosphorylation on tyrosine. However, the addition of Man-6-P, Glc-6-P, and PMP-BSA had no effect on the basal level of phosphorylation. On the other hand, overexpression of IGF2Rmyc with the chimera almost completely eliminated phosphorylation of the chimera (Fig. 5).

View larger version (27K): [in this window] [in a new window]   Fig. 5.   Expression and phosphorylation of the IGF2R/EGFR chimera. The IGF2R/EGFR chimera was transiently transfected alone or cotransfected with IGF2Rmyc (as indicated) by mixing the cDNAs prior to the transfection. The empty pCMV5 vector was also transfected as a negative control. The cultures were treated at 24 h and again at 48 h after transfection with serum-free DMEM, or DMEM supplemented with 5 mM Man-6-P, 25 nM PMP-BSA, or 5 mM Glc-6-P prior to lysing the cells 51 h after transfection. Equimolar amounts of the chimera (from ~30 µl of cell extract) were immunoprecipitated in duplicate reactions with -FLAG resin and loaded onto 6% reducing SDS-PAGE gels for immunoblot analysis with either the -FLAG (top panel) or -Tyr(P) (middle panel) antibodies. Two sets of lysates cotransfected with the IGF2R/EGFR chimera and IGF2Rmyc are shown. To confirm the expression of the IGF2Rmyc construct, cell lysates (30 µl) from the cells transfected with pCMV5 only, the IGF2R/EGFR chimera, and the chimera cotransfected with IGF2Rmyc were analyzed by 6% reducing SDS-PAGE in duplicate and immunoblotted with the -Myc antibody (bottom panel).

To further investigate the relationship between heterodimer formation and phosphorylation of the chimera, another series of transfections was conducted using equal amounts of the chimera cDNA mixed with increasing amounts of IGF2Rmyc. Equal amounts of the cell lysates were analyzed by -FLAG and -Myc immunoblotting, revealing that mixing the IGF2Rmyc cDNA at increasing ratio with a constant amount of the IGF2R/EGFR chimera resulted in a gradient of expression of the IGF2Rmyc with a modest effect (30% maximal reduction) on the expression of the chimera (Fig. 6A). The ability of IGF2Rmyc to interact with the chimera, and the effect this interaction has on the phosphorylation of the chimera, was next measured by immunoprecipitating the chimera from cell lysates with the -FLAG resin and then immunoblotting with either the -FLAG, -Tyr(P), or -Myc antibodies (Fig. 6B). Even in the presence of the harsh conditions imposed by the cell lysis buffer (0.1% SDS, 0.5% sodium deoxycholate, and 1% Nonidet P-40), IGF2Rmyc coimmunoprecipitated with the chimera. In addition, the inhibition of chimera phosphorylation correlated directly with the amount of IGF2Rmyc that was complexed with the chimera.

View larger version (43K): [in this window] [in a new window]   Fig. 6.   IGF2Rmyc interacts directly with the IGF2R/EGFR chimera. A, aliquots (30 µl) of cell lysates from cells transiently transfected with the empty pCMV5 vector or the IGF2R/EGFR chimera expressed with or without increasing amounts of the IGF2Rmyc cDNA construct were loaded on 6% reducing SDS-PAGE gels. Following electrophoresis, both gels were transferred to nitrocellulose; one blot was probed with the -FLAG antibody, and the other was probed with the -Myc antibody to confirm expression of the constructs. The control lysates containing just IGF2Rmyc or the IGF2R/EGFR chimera were loaded in duplicate. B, coimmunoprecipitation and phosphorylation studies were carried out by incubating aliquots (30-50 µl) of the cell lysates containing equimolar amounts of the chimera with -FLAG resin. Immunoprecipitations of the control lysates containing IGF2R or the IGF2R/EGFR chimera were carried out in duplicate reactions. After immunoprecipitation, the resin pellets were washed and analyzed by SDS-PAGE and immunoblotting with -FLAG, -Tyr(P), or -Myc, as indicated.

Man-6-P Binding Domains Are Not Required for Dimerization of the IGF2R/EGFR Constructs-- Two IGF2R/EGFR constructs were engineered bearing deletions of the Man-6-P binding domains of the IGF2R extracytoplasmic domain (Fig. 4). The 11-TM/EGFR construct contains repeats 11-15 and the transmembrane domain of the IGF2R fused to the cytoplasmic domain of the EGFR. The other construct, called 13-TM/EGFR, contains repeats 13-15 and the transmembrane domain of the IGF2R fused to the EGFR cytoplasmic domain. These constructs were expressed either alone or with the IGF2Rmyc construct in 293T cells, as was described above for the full-length IGF2R/EGFR chimera. An unrelated FLAG-tagged construct, encoding kinase suppressor of Ras (KSRF), was also expressed with IGF2Rmyc as a negative control.

These expressed constructs were immunoprecipitated with the -FLAG resin and then subjected to immunoblot analyses with the -FLAG, -Tyr(P), and -Myc antibodies to determine if they display characteristics similar to those of the full-length IGF2R/EGFR construct (Fig. 7). Both the 11-TM/EGFR and the 13-TM/EGFR were capable of being translated into proteins with sizes appropriate for their predicted molecular weights. Like the full-length IGF2R/EGFR chimera, the 11-TM/EGFR when expressed on its own demonstrated a strong level of tyrosine phosphorylation, which was inhibited by coexpression of the IGF2Rmyc construct. The 13-TM/EGFR chimera also demonstrated high levels of tyrosine phosphorylation when expressed alone. Surprisingly, coexpression of the IGF2Rmyc did not affect the level of the 13-TM/EGFR phosphorylation. However, the 13-TM/EGFR chimera did interact with IGF2Rmyc, as demonstrated by coimmunoprecipitation analysis (Fig. 7). Based on the expression level of IGF2Rmyc in the cotransfected cell lysates, the amount of IGF2Rmyc that interacted with 13-TM/EGFR was 32 ± 11% (n = 4) lower than that bound by 11-TM/EGFR.

View larger version (37K): [in this window] [in a new window]   Fig. 7.   Expression and phosphorylation of the 11-TM/EGFR and 13-TM/EGFR constructs. 293T cells were transiently transfected with the 11-TM/EGFR and 13-TM/EGFR, by themselves or along with the IGF2Rmyc construct. KSRF was coexpressed with the IGF2Rmyc construct as a control. Equimolar amounts of the Flag-tagged constructs were immunoprecipitated from 293T cell lysates and resolved on 6% reducing SDS-PAGE gels in duplicate loadings. Immunoblot analysis was conducted with the -FLAG (first panel), -phosphotyrosine (second panel), and -Myc antibodies (third panel). In addition, equal amounts of cell lysates (60 µg of protein) were directly subjected to SDS-PAGE and immunoblot analysis with the -Myc antibody to determine the relative levels of IGF2Rmyc expression (fourth panel).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

One of the major obstacles in demonstrating the dimeric structure of the IGF2R is that, as previously reported, when measured by sucrose gradient and gel filtration techniques in the absence of bivalent phosphomannosylated ligands, detergent-solubilized IGF2R appears to exist primarily as a monomer (17, 18). Analysis of purified soluble and full-length IGF2R species by gel filtration on a Superose-12 FPLC column, reported herein, demonstrated that the full-length receptor has a Stokes radius of 66 Å, which is somewhat smaller than previous reports of 79 Å for the full-length bovine receptor (17) and 72 Å for the full-length rat receptor (18). Sucrose gradient centrifugation of these purified receptor species suggested that they were monomeric receptors with calculated molecular weights of 334,000 (17) and 290,000 (18), respectively. The large Stokes radius of the receptor in relation to its predicted molecular weight suggests that it is not globular in shape, but rather exists in a cigar-shaped conformation under these conditions. The sIGF2R, reported here, demonstrated a Stokes radius that was 26% smaller than the full-length receptor in the gel filtration analysis, even though this species is lacking only a small portion of the carboxyl terminus (35). This observation could be explained if the extracytoplasmic portion of the IGF2R alone folds into a globular protein under these experimental conditions. Thus, the ellipsoid character of detergent-solubilized IGF2R may require the presence of the transmembrane and cytoplasmic domains.

Analysis of the full-length IGF2R by native gel electrophoresis, using the same sample preparation as was loaded onto the Superose-12 column, resulted in quite different hydrodynamic behavior. Under these conditions, the monomeric form of the full-length receptor demonstrated a Stokes radius identical to that measured by the FPLC analysis, but the majority of the receptor was present as a much larger complex with a Stokes radius greater than 85 Å, suggesting a dimeric form of the receptor. Even when the receptor was further purified under acidic conditions that would be expected to promote dissociation of phosphomannosyl ligands, the dimeric form persisted, suggesting that exogenous ligands are not necessary for dimer formation. One of the major differences between the FPLC analysis and the native gel was the pH. The FPLC analysis was conducted at a pH of 7.4, whereas the native gel electrophoresis was conducted using the Laemmli method at a pH of 6.8 in the stacking gel and a pH of 8.3 in the resolving gel (38). However, it is difficult to explain these observations based solely on the differences in pH.

While this study was ongoing, York et al. (17) reported that the addition of a multivalent Man-6-P-bearing lysosomal enzyme to the gel filtration analysis resulted in an increase in size of the IGF2R, with a Stokes radius consistent with a dimeric complex of the IGF2R bound to a single ligand. In light of the conflicting results from the FPLC and native gel analyses reported herein, the observation of York et al. (17) suggests that ligand occupancy of the IGF2R may increase the stability of the dimeric complex so that it remains intact during gel filtration. However, when we incubated purified sIGF2R in the presence of 1-8000 nM PMP-BSA prior to native gel electrophoresis, we found no change in the relative ratios of dimeric to monomeric complexes (data not shown). Whereas it seems likely that ligand occupancy increased the stability of IGF2R dimers, these data suggest that the amount of preformed sIGF2R dimer is not influenced by the presence of multivalent ligand.

The discrepancy between the FPLC analysis and native gel electrophoresis compelled us to address the oligomeric nature of the IGF2R using other techniques. In the first approach, we used two truncated soluble IGF2R constructs bearing different epitope tags. Both constructs contained all 15 repeats of the extracytoplasmic domain followed by a FLAG or Myc epitope tag, called 15F and 15myc, respectively. The use of unique epitope tags allowed identification of heterooligomers by immunoprecipitation with the -FLAG antibody and immunoblotting with the -Myc antibody. When transiently coexpressed in 293T cells, these truncated receptors exist as heterocomplexes in 1% Triton X-100 cell lysates. Surprisingly, the addition of either exogenous PMP-BSA or IGF-II had no effect on the amount of complex measured using coimmunoprecipitation techniques. Nearly 50% of the 15myc that was coexpressed with 15F was coimmunoprecipitated with the anti-FLAG antibody, which would be the predicted result if all of the 15F and 15myc constructs exist as either homodimers or heterodimers with a 1:2:1 Gaussian distribution.

To determine if the 15F and 15myc constructs are capable of interacting outside of the cell, they were purified separately on PMP-Sepharose affinity columns. When these purified constructs were mixed in the presence of approximately 10 mM Man-6-P, they demonstrated no interaction after incubation for 3 h at 3 °C, but when the temperature was increased to 37 °C for 1-8 h, progressive formation of a heterocomplex was observed. There are several possible explanations for this finding. Either the rate of association of the monomers is very slow, or the purified constructs exist as homodimers and the rate of exchange between the dimers is very slow. The latter explanation seems more likely, as most of the coexpressed 15F and 15myc appear to exist in an oligomeric state, as determined by the immunoprecipitation experiments discussed above. The interaction between the extracytoplasmic domains of 15F and 15myc was not affected by changes in pH (from 7.5-5.0). This implies that the normal cycling of the IGF2R through the acidic endosomal compartment would not disrupt the oligomeric state of the receptor (data not shown). Thus, our data are in agreement with those of York et al. (17) that the IGF2R is capable of forming oligomeric structures, most likely dimers, through its extracytoplasmic domain.

To determine whether a membrane-bound form of the IGF2R is capable of forming dimeric complexes, a chimeric protein was engineered containing the extracytoplasmic and transmembrane domains of the IGF2R fused to the cytoplasmic domain of the EGFR. The experimental rationale was that if the chimera formed dimers, the EGFR tyrosine kinase domain would undergo autophosphorylation mediated by dimer-induced juxtaposition of the cytoplasmic domains, which we could monitor using an anti-phosphotyrosine antibody. A comparable, kinase-inactive, IGF2R/EGFR mutant chimera has been previously reported (39). Even though this kinase-inactive chimera lacked the cytoplasmic domain of the IGF2R, it shared a very similar subcellular distribution to the wild-type IGF2R (39). In addition, we found that the chimera bound both IGF-II and PMP-BSA with the same binding characteristics as the wild-type IGF2R (data not shown).

When expressed in 293T cells, the IGF2R/EGFR chimera demonstrated constitutive phosphorylation, consistent with the formation of oligomeric complexes. Further evidence that the phosphorylation of the chimera arises from the formation of a homodimer between two IGF2R/EGFR chimeras comes from the finding that overexpression of a Myc epitope-tagged IGF2R decreased the phosphorylation level of the IGF2R/EGFR chimera, presumably through formation of heterodimers with the chimera. The IGF2Rmyc construct coimmunoprecipitated with the IGF2R/EGFR chimera when they were coexpressed, suggesting that the cytoplasmic domain of IGF2Rmyc was not required for nor did it impede the extracytoplasmic domain's ability to form oligomeric structures. Although the finding that the chimera was capable of autophosphorylation was remarkable, it could be explained by dimerization mediated either by direct interaction between the IGF2R extracytoplasmic domains or by cross-linking mediated by binding of a bivalent Man-6-P ligand. However, like the CD-MPR, the IGF2R appears to form constitutive dimers in the cell that are not dependent upon ligand occupancy, as addition of exogenous PMP-BSA, Man-6-P, or Glc-6-P to cells transfected with the chimera had no effect on the phosphorylation level of the chimera. It is important to note that ligands added to the medium bathing the cells might have had access only to the chimeric constructs present at the cell surface. Since these receptors account for about 10% of the receptor population, the majority of the chimeric constructs might not be affected by this manipulation. Therefore, to show that the IGF2R can form dimers in the absence of bound Man-6-P ligands, we decided to test chimeras of the IGF2R containing deletions of the Man-6-P ligand binding domains.

Analysis of IGF2R/EGFR chimeras containing truncations of the IGF2R extracytoplasmic domain demonstrated that regions outside the Man-6-P binding domains are sufficient for the formation of receptor oligomers. The 11-TM/EGFR chimera, which lacks both Man-6-P binding domains, behaved identically to the full-length chimera in terms of its tyrosine phosphorylation when expressed alone or with IGF2Rmyc. Whereas the tyrosine phosphorylation of the 11-TM/EGFR chimera was inhibited by overexpression of IGF2Rmyc, the 13-TM/EGFR chimera behaved differently. The 13-TM/EGFR chimera showed high levels of autophosphorylation when expressed alone, but overexpression of IGF2Rmyc had no effect on its tyrosine phosphorylation. Surprisingly, the 13-TM/EGFR did interact with IGF2Rmyc in the coimmunoprecipitation assay, although to a lesser degree than the 11-TM/EGFR. One possible explanation for these findings could be that the 13-TM/EGFR construct has a lower affinity for the full-length IGF2Rmyc, allowing for increased dissociation and re-association. Such kinetic behavior could explain why the 13-TM/EGFR construct displays high levels of tyrosine phosphorylation even in the presence of excess IGF2Rmyc. Support for this explanation comes from our observation (54) that constructs that lack repeats 12-15 of the IGF2R extracytoplasmic domain formed fewer high affinity Man-6-P binding sites than constructs containing the entire extracytoplasmic domain. Collectively, these data suggest that residues in or near repeat 12 are important for the formation of binding-competent receptor dimers. Whereas repeat 12 likely plays a role in the formation of dimers, other regions must be involved as well, because constructs lacking repeat 12 are capable of forming high affinity Man-6-P binding sites (54).

Several interesting correlations can be drawn between what is known about the oligomeric nature of the homologous CD-MPR and the observations reported here. The CD-MPR forms oligomeric structures in solution, including monomers, dimers, and tetramers (19, 40-42, 44). But the effects of ligand occupancy on the oligomeric state of the CD-MPR remain somewhat controversial. Whereas the quaternary structure of detergent-solubilized CD-MPR appears to be regulated by the presence of multivalent ligands (45), chemical cross-linking studies in intact cells suggest that the relative ratios of dimer to tetramer of the CD-MPR remain unchanged in the face of changing pH, intracellular cycling, and ligand occupancy (42). The crystal structure of the extracytoplasmic domain of the CD-MPR has revealed that the CD-MPR binds multivalent Man-6-P ligands with high affinity by forming dimers through its extracytoplasmic domain (46). Indeed, truncated forms of the CD-MPR comprising only the extracytoplasmic domain are still capable of forming dimeric complexes and interacting with multivalent Man-6-P ligands with high affinity (46, 47). It seems likely that the homologous IGF2R has retained a similar mechanism for high affinity Man-6-P binding as the CD-MPR-through receptor oligomerization. Direct evidence for this mechanism comes from the observation, reported here, that both the dimeric sIGF2R and full-length IGF2R show higher stoichiometries of ¹²⁵I-PMP-BSA binding than the monomeric forms when separated by native gel electrophoresis.

The finding that the IGF2R exists as a dimer in the absence of ligands suggests that dimerization per se may not play a direct role in accelerated internalization of the IGF2R in response to ligand binding. Binding of a multivalent ligand to the dimeric IGF2R could, however, cause a conformational change in the receptor such that the internalization signals in the cytoplasmic domain are more optimally displayed. This could account for the apparent differences between the present work and that of York et al. (17), who found that the internalization rate of the IGF2R increases in response to the addition of a multivalent Man-6-P bearing lysosomal enzyme in intact cells. They pointed out that IGF2R dimers pre-existing in the plasma membrane that were readily cross-linked by a bivalent ligand binding at 4 °C might account for the rapid internalization of -glucuronidase when cells were rapidly warmed to 37 °C. Conformational changes occurring upon ligand binding have been observed for other dimeric membrane receptors such as the insulin receptor (48, 49). Those studies have shown that insulin binds to the insulin receptor through a multivalent interaction, which results in conformational changes in both the extracytoplasmic and cytoplasmic domains (50, 51) that are apparently necessary to activate the receptor's tyrosine kinase activity through transphosphorylation (36, 52, 53). Additional support for conformational changes of the IGF2R in response to ligand binding comes from the observation that treatment of the IGF2R with saturating concentrations of Man-6-P increases IGF-II binding in receptor preparations from cells that do not produce endogenous Man-6-P bearing ligands (43). This finding suggests that ligand occupancy of the Man-6-P binding domain, even by a low affinity monovalent ligand, can influence the conformation of this receptor.

In summary, we have previously reported that both Man-6-P binding domains in the extracytoplasmic domain are capable of forming high affinity Man-6-P binding sites independently of each other, and that receptor oligomerization may be required for development of high affinity binding. From our studies reported here, the IGF2R appears to share much more than sequence homology with the CD-MPR, as both receptors appear to undergo dimerization as a means of forming high affinity lysosomal enzyme binding sites. These observations support the hypothesis that the repeat 3 and repeat 9 binding domains are capable of sorting a specific subset of lysosomal enzymes, and provide important new understanding of the mechanisms by which the IGF2R carries out its functions in cellular physiology.

	ACKNOWLEDGEMENTS

We thank Dr. Surinder K. Batra for performing the Pfu amplification using the EGFRvIII cDNA in the construction of the IGF2R/EGFR chimera, and Dr. Robert E. Lewis and Jennifer A. Fulton for supplying the KSRF construct and for helpful discussion in using the IGF2R/EGFR chimera. We are grateful to Margaret H. Niedenthal of Lilly Research Laboratories for providing the IGFs, to Drs. William S. Sly and David W. Russell for providing the human IGF2R cDNA and pCMV5, respectively, and to Rockefeller University for permission to use the 293T cells. We thank the University of Nebraska Medical Center Molecular Biology Core Facility (Omaha, NE) for their work in sequencing the prepared constructs. We also appreciate discussion with and suggestions from Dr. C. Kirk Phares, Dr. Myron L. Toews, and Dr. Ming-Fong Lin and members of Dr. Lin's laboratory in studies on the phosphorylation levels of the IGF2R/EGFR chimera.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grant DK44212 (to R. G. M.) and by stipend support from the Emley Fellowship, the Dr. Fred W. Upson grant-in-aid, and the Kate Field grant-in-aid awards through the University of Nebraska Medical Center. DNA sequencing costs were subsidized by National Institutes of Health NCI Core Grant CA36727 and the Nebraska Research Initiative.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.

^¶ Present address: CytImmune Sciences, Inc., College Park, MD 20740.

To whom correspondence should be addressed: Dept. of Biochemistry and Molecular Biology, 984525 Nebraska Medical Center, Omaha, NE 68198-4525. Tel.: 402-559-7824; Fax: 402-559-3920; E-mail: rgmacdon@unmc.edu.

Published, JBC Papers in Press, April 11, 2000, DOI 10.1074/jbc.M001273200

	ABBREVIATIONS

The abbreviations used are: IGF2R, insulin-like growth factor II/mannose 6-phosphate receptor; IGF, insulin-like growth factor; Man-6-P, mannose 6-phosphate; PMP, pentamannose phosphate; BSA, bovine serum albumin; CD-MPR, cation-dependent mannose 6-phosphate receptor; MPR, mannose 6-phosphate receptor; EGFR, epidermal growth factor receptor; FPLC, fast protein liquid chromatography; PAGE, polyacrylamide gel electrophoresis; HBST, HEPES-buffered saline containing 0.1% Triton X-100; sIGF2R, soluble IGF2R; FBS, fetal bovine serum; nt, nucleotide(s); Glc-6-P, glucose 6-phosphate; DMEM, Dulbecco's modified Eagle's medium; TM, transmembrane.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES