High-affinity ligand binding by wild-type/mutant heteromeric complexes of the mannose 6-phosphate/insulin-like growth factor II receptor

Abstract

The mannose 6-phosphate/insulin-like growth factor II receptor has diverse ligand-binding properties contributing to its roles in lysosome biogenesis and growth suppression. Optimal receptor binding and internalization of mannose 6-phosphate (Man-6-P)-bearing ligands requires a dimeric structure leading to bivalent high-affinity binding, presumably mediated by cooperation between sites on both subunits. Insulin-like growth factor II (IGF-II) binds to a single site on each monomer. It is hypothesized that IGF-II binding to cognate sites on each monomer occurs independently, but bivalent Man-6-P ligand binding requires cooperative contributions from sites on both monomers. To test this hypothesis, we co-immunoprecipitated differentially epitope-tagged soluble mini-receptors and assessed ligand binding. Pairing of wild-type and point-mutated IGF-II binding sites between two dimerized mini-receptors had no effect on the function of the contralateral binding site, indicating IGF-II binding to each side of the dimer is independent and manifests no intersubunit effects. As expected, heterodimeric receptors composed of a wild-type monomer and a mutant bearing two Man-6-P-binding knockout mutations form functional IGF-II binding sites. By contrast to prediction, such heterodimeric receptors also bind Man-6-P-based ligands with high affinity, and the amount of binding can be attributed entirely to the immunoprecipitated wild-type receptors. Anchoring of both C-terminal ends of the heterodimer produces optimal binding of both IGF-II and Man-6-P ligands. Thus, IGF-II binds independently to both subunits of the dimeric mannose 6-phosphate/insulin-like growth factor II receptor. Although wild-type/mutant hetero-oligomers form readily when mixed, it appears that multivalent Man-6-P ligands bind preferentially to wild-type sites, possibly by cross-bridging receptors within clusters of immobilized receptors.

Fig. 1. Schematic diagram and ligand blot analysis of FLAG and Myc epitope-tagged M6P/IGF2R mini-receptors

A, the receptor constructs are shown in linear format from amino terminus to carboxyl terminus, with repeats of the ectodomain illustrated as rectangles. The shaded rectangles indicate repeats 3 and 9, to which the main determinants of Man-6-P binding have been mapped. The stippled rectangles represent repeat 11 containing the principal residues responsible for IGF-II binding, and the asterisk denotes the → T mutation at residue 1572 (I/T), which abrogates IGF-II binding. The black rectangles represent the FLAG or Myc epitope tags on the carboxyl terminus. B and C, equimolar amounts of the transfected cell lysates were electrophoresed on a 6% non-reducing SDS-PAGE gels. The proteins were transferred to BA85 nitrocellulose, processed for ligand blotting, and probed for binding of either ¹²⁵I-PMP-BSA (B) or ¹²⁵I-IGF-II (C) and developed by autoradiography. The autoradiograms of representative blots are shown.

Fig. 2. Analysis of ligand binding to soluble 1–15 and 1–15 I/T mutant FLAG epitope-tagged receptors immunoprecipitated with α-FLAG resin

Cell lysates, containing equimolar amounts of expressed soluble receptors, were immunoprecipitated with M2 α-FLAG affinity resin and assayed for binding of ¹²⁵I-PMP-BSA (A) or ¹²⁵I-IGF-II (B). The lines on each graph indicate the amount of binding predicted if the wild-type and mutant receptors are binding ligand independently. The triangles indicate a progressive shift in the ratio of wild-type to mutant receptor cDNA transfected into cells. Values represent the mean ± SD of three replicate measurements for each condition. These data represent the means of four independent experiments.

Fig. 3. Co-immunoprecipitation and ligand binding by FLAG and Myc epitope-tagged asymmetric dimeric soluble receptors immunoprecipitated with M2 α-FLAG resin

The ability of 1–15Myc to co-immunoprecipitate with 1–15F was measured by immunoprecipitating equimolar amounts of the 1–15F soluble receptor with M2 α-FLAG affinity resin from 293T lysates of the M6P/IGF2R mini-receptors. After immunoprecipitation, the resin pellets were collected, washed, heated with sample buffer, and analyzed by 6% reducing SDS-PAGE. The proteins were transferred to BA85 nitrocellulose, immunoblotted with α-FLAG M2 (A and B) or α-Myc 9E10 (C and D) antibodies and developed with ¹²⁵I-protein A. As a control, cell lysate in the amount that was used during the immunoprecipitation was directly loaded onto the gel (A and C). Cell lysates, containing equimolar amounts of expressed FLAG-tagged soluble receptors, were immunoprecipitated with M2 α-FLAG affinity resin and then incubated in the presence of 1 nM ¹²⁵I-PMP-BSA (E) or 2 nM ¹²⁵I-IGF-II (F) for 3 h at 4°C. Bound ligand was determined by centrifuging the resin pellets, washing, and counting the pellets in a γ-counter. Radioactivity retained in the presence of either 5 mM Man-6-P or 1 μM IGF-II was subtracted from each binding reaction to determine the specific binding for ¹²⁵I-PMP-BSA and ¹²⁵I-IGF-II, respectively. The lines on each graph (E and F) indicate the amount of binding predicted if the wild-type and mutant receptors are binding ligand independently. The tables indicate the amounts of the various cDNAs transfected into cells for each condition and apply to the data shown both above and below the table. Values represent the mean ± SD of three replicate measurements for each condition. These data represent the means of four independent experiments.

Fig. 4. Co-immunoprecipitation and ligand binding of FLAG and Myc epitope-tagged asymmetric dimeric soluble receptors immunoprecipitated with protein G-Sepharose

The ability of 1–15F to co-immunoprecipitate with 1–15Myc was measured by immunoprecipitating equimolar amounts of the 1–15Myc soluble receptor in 293T lysates of the M6P/IGF2R mini-receptors with protein G-Sepharose previously incubated with α-Myc 9E10 antibody. After immunoprecipitation, the resin pellets were collected, washed, heated with sample buffer, and analyzed by 6% reducing SDS-PAGE. The proteins were transferred to BA85 nitrocellulose and immunoblotted with α-Myc 9E10 (A and B) or α-FLAG M2 (C and D) antibodies. As a control, cell lysate in the amount that was used during the immunoprecipitation was directly loaded onto the gel (A and C). Cell lysates, containing equimolar amounts of expressed Myc-tagged soluble receptors, were immunoprecipitated with protein G-Sepharose previously incubated with α-Myc 9E10 antibody and assayed for binding of ¹²⁵I-PMP-BSA (E) or ¹²⁵I-IGF-II (F). The lines on each graph (E and F) indicate the amount of binding predicted if the wild-type and mutant receptors are binding ligand independently. The tables indicate the amounts of the various cDNAs transfected into cells for each condition and apply to the data shown both above and below the table. The values represent the mean ± SD of three representative measurements for each condition. These data represent the means of four independent experiments.

Fig. 5. Schematic diagram and ligand binding analysis of soluble 1–15 and 1–15R2A mutant FLAG epitope-tagged receptors immunoprecipitated withα α-FLAG resin

A, the receptor constructs are shown in linear format from amino terminus to carboxyl terminus, with repeats of the ectodomain illustrated as rectangles. The stippled rectangles represent repeat 11 containing the principal residues responsible for IGF-II binding. The shaded rectangles indicated repeats 3 and 9, to which the main determinants of Man-6-P binding have been mapped and the asterisk denotes the R→ A mutations at residues 426 and 1325 (R2A), which abrogates Man-6-P binding. The black rectangles represent the FLAG epitope tags on the carboxyl terminus. B and C, Cell lysates, containing equimolar amounts of expressed soluble receptors, were immunoprecipitated with M2 α-FLAG affinity resin and assayed for binding of ¹²⁵I-IGF-II (B) or ¹²⁵I-PMP-BSA (C). The lines on each graph indicate the amount of binding predicted if the wild-type and mutant receptors are binding ligand independently. The triangles indicate a progressive shift in the ratio of wild-type to mutant receptor cDNA transfected into cells. Values represent the mean ± SD of three replicate measurements for each condition. These data represent the means of four independent experiments.

Fig. 6. Schematic diagram of soluble 1–15 and 1–15 R2A mutant Myc epitope-tagged receptors, co-immunoprecipitation, and ligand binding analysis of soluble 1–15 and 1–15R2A mutant FLAG and Myc epitope-tagged asymmetric soluble heterodimeric receptors

A, the receptor constructs are shown in linear format from amino terminus to carboxyl terminus, with repeats of the ectodomain illustrated as rectangles. The stippled rectangles represent repeat 11 containing the principal residues responsible for IGF-II binding. The shaded rectangles indicated repeats 3 and 9, to which the main determinants of Man-6-P binding have been mapped and the asterisk denotes the R→ A mutations at residues 426 and 1325 (R2A), which abrogates Man-6-P binding. The black rectangles represent the Myc epitope tags on the carboxyl terminus. B-E, the ability of 1–15Myc to co-immunoprecipitate with equimolar amounts of 1–15F soluble receptor with M2 α-FLAG affinity resin from 293T cell lysates of M6P/IGF2R mini-receptors. After immunoprecipitation, the resin pellets were collected, washed, and immunoblotted with α-FLAG M2 (B and C) or α-Myc 9E10 (C and D) antibodies. As a control, cell lysate in the amount that was used during the immunoprecipitation was directly loaded on the gel (B and D). F and G, cell lysates, containing equimolar amounts of expressed FLAG-tagged soluble receptors, were immunoprecipitated with M2 α-FLAG affinity resin and assayed for binding of 2 nM ¹²⁵I-IGF-II (F) or 1 nM ¹²⁵I-PMP-BSA (G) for 3 h at 4°C. The lines on each graph (F and G) indicate the amount of binding predicted if the wild-type and mutant receptors are binding ligand independently. The tables indicate the amounts of the various cDNAs transfected into cells for each condition and apply to the data shown both above and below the table. Values represent the mean ± SD of three replicative measurements for each condition. These data represent the means of four independent experiments.