Lysosomal enzyme binding to the cation-independent mannose 6-phosphate receptor is regulated allosterically by insulin-like growth factor 2

Abstract

The cation-independent mannose 6-phosphate receptor (CI-MPR) is clinically significant in the treatment of patients with lysosomal storage diseases because it functions in the biogenesis of lysosomes by transporting mannose 6-phosphate (M6P)-containing lysosomal enzymes to endosomal compartments. CI-MPR is multifunctional and modulates embryonic growth and fetal size by downregulating circulating levels of the peptide hormone insulin-like growth factor 2 (IGF2). The extracellular region of CI-MPR comprises 15 homologous domains with binding sites for M6P-containing ligands located in domains 3, 5, 9, and 15, whereas IGF2 interacts with residues in domain 11. How a particular ligand affects the receptor's conformation or its ability to bind other ligands remains poorly understood. To address these questions, we purified a soluble form of the receptor from newborn calf serum, carried out glycoproteomics to define the N-glycans at its 19 potential glycosylation sites, probed its ability to bind lysosomal enzymes in the presence and absence of IGF2 using surface plasmon resonance, and assessed its conformation in the presence and absence of IGF2 by negative-staining electron microscopy and hydroxyl radical protein footprinting studies. Together, our findings support the hypothesis that IGF2 acts as an allosteric inhibitor of lysosomal enzyme binding by inducing global conformational changes of CI-MPR.

Fig. 1

Purification of sCI-MPR from newborn calf serum. (A) Schematic diagram of the trafficking of CI-MPR and lysosomal enzymes between the trans-Golgi network, endosomes, and cell surface. CI-MPR and lysosomal enzymes are synthesized in the endoplasmic reticulum (ER) and undergo co-translational N-glycosylation, and their N-glycans are modified in the Golgi. Phosphodiester-containing glycans are converted, or ‘uncovered’, forming phosphomonoesters. (B) Schematic diagram of the ~ 300 kDa full-length CI-MPR, a type I transmembrane protein containing 15 homologous domains in its extracellular region. The carbohydrate recognition domains (CRDs) are highlighted in green, showing domains specific for phosphomonoesters (M6P) and/or phosphodiesters (M6P-GlcNAc). The IGF2 binding site (blue) maps to domain 11 while domain 1 contains residues critical for interacting with plasminogen (Plg) or urokinase-type plasminogen activator receptor (uPAR). The 19 potential N-linked glycosylation sites are shown with red dots. (C) Schematic diagram of the soluble form of the receptor, sCI-MPR, found in extracellular fluids following its release at the plasma membrane by a metalloproteinase. (D) The soluble receptor was isolated from newborn calf serum by affinity chromatography. Purified sCI-MPR (1 μg) was resolved by SDS-PAGE on a 4–20% gradient polyacrylamide gel and visualized by staining with Coomassie blue G250. The migration of molecular weight markers is indicated. Created with https://www.biorender.com/.

Fig. 2

Analysis of sCI-MPR by mass spectrometry. The affinity-purified sCI-MPR protein isolated from newborn calf serum was digested with trypsin in solution as described in Methods. Results from three separate purifications using two different lots of serum are shown and the identified peptides are highlighted in green, aqua, or yellow in the amino acid sequence of the full-length bovine protein. The 44-residue N-terminal signal sequence is underlined, the triplet sequences identifying potential N-linked glycosylation sites are highlighted in red, the conserved 13-residue sequence at the C-terminus of each of the 15 domains is underlined, and the predicted transmembrane region is boxed.

Fig. 3

Site-specific heterogeneity of N-linked glycosylation on sCI-MPR. (A) A schematic representation of the mass spectrometry (MS) workflow performed for the site-specific N-glycosylation analysis of sCI-MPR using the method of Cao, et al.. Briefly, tryptic digests from sCI-MPR were further treated sequentially with Endo F1 followed by PNGase F/¹⁸O-water to produce unique mass signatures to identify and quantify the types of N-glycans and the occupancy at each of the 19 potential N-linked sites. (B) Pie charts illustrate relative percentages of unoccupied (gray) or occupied by high-mannose/hybrid type without core fucosylation (green) or complex-type (purple) N-glycans at each N-linked glycosylation site of sCI-MPR. The MS workflow presented in panel A was used to characterize 17 of the 19 N-linked sites, whereas tryptic digests were used to characterize glycopeptides at the remaining two sites at N409 and N2094 (shown with asterisks).

Fig. 4

Binding affinity at pH 7.4 of IGF2 and the lysosomal enzymes GAA and PPT1 to sCI-MPR as assessed by SPR. (A) Biotinylated IGF2 (GroPep, Inc.) was immobilized on the surface of a streptavidin-coupled SA sensor chip. Representative sensorgrams (red) of sCI-MPR protein (0.5, 1, 2, 5, 10, and 20 nM) flowed over the IGF2 surface are shown for a single experiment. The resulting curves from three independent experiments were fit to a 1:1 binding model (Biacore S200 evaluation software). The fitted curves are displayed in black, and the resulting kinetic constants are shown ± SE. Human lysosomal enzymes (B) GAA and (C) PPT1 were immobilized onto the surface of a CM5 sensor chip by amine-coupling. Representative sensorgrams of sCI-MPR (0.2, 0.5, 1, 2, 5, 10, 20, 40, 80, and 120 nM) flowed over each surface are shown. The binding levels (response units, RU) were determined as the average response 4 s before the end of the association phase (vertical gray bar in panels B,C) of each injection using the Biacore S200 evaluation software (Cytiva, Inc.). The binding levels (RU) for the SPR experiments were plotted against the concentration of receptor (inset graph, panels B,C) and fitted to a one-site specific binding equation as described in Methods. K_d ± SEM is shown for 10 independent SPR experiments representing four purifications of sCI-MPR from two lots of newborn calf serum. Illustrations created using https://www.biorender.com/.

Fig. 5

SPR analysis at pH 7.4 of sCI-MPR incubated with IGF2 prior to interaction with lysosomal enzymes. Samples of sCI-MPR at 20 nM were incubated with IGF2 (0, 0.01, 0.05, 0.1, 0.5, 1, 2, 5, 7.5, 10, 15, 20 and 100 nM) or IGF1 (100 nM) for 1 h at room temperature before flowing the mixture over the GAA- (A,B) or PPT1- (C,D) immobilized surfaces. (A,C) Representative sensorgrams are shown, with insets comparing the sensorgrams of sCI-MPR with IGF2 (0 and 100 nM, red) and IGF1 (100 nM, black). (B,D) Plotted are the maximal response units (RU, vertical gray bar in panels A,C) as the mean of the percentage of maximal RU ± SD for three independent SPR experiments, with 0 nM IGF2 set at 100% maximal binding. Data are fit to a 4-parameter dose response curve using GraphPad Prism v 10.2.0. LogIC₅₀ values ± SEM are shown. Illustrations created using https://www.biorender.com/.

Fig. 6

SPR analysis at pH 7.4 of sCI-MPR dissociation from lysosomal enzymes in the presence and absence of IGF2. Samples of 20 nM sCI-MPR were flowed over the GAA- or PPT1-immobilized sensor chip surface for 3 min followed by a second injection (initiation of dissociation phase) of buffer without receptor but containing increasing concentrations of IGF2 (0, 0.05, 0.1, 0.5, 1, 2, 5, 7.5, 10, 12.5, 15, 20, and 100 nM) or IGF1 (100 nM) for 2 min, and buffer alone was then flowed over the surface for the final 5 min. (A,D) Representative sensorgrams are shown, with insets comparing the sensorgrams of sCI-MPR with IGF2 (0 and 100 nM, red) and IGF1 (100 nM, black). (B,E) Plotted are the response units (RU) measured 4 s before the end of the IGF2 injection (vertical gray bar in panels (A,D)). Percent maximal response was calculated by dividing the response by the response at 0 nM IGF2 multiplied by 100. The normalized responses ± SD for three independent SPR experiments were plotted against the log of the IGF2 concentration. Data are fit to a 4-parameter dose response curve using GraphPad Prism v 10.2.0. LogIC₅₀ values ± SEM are shown. (C,F) Plotted are the k_off rates during the 2 min injection of buffer without receptor but containing increasing concentrations of IGF2 (0, 0.05, 0.1, 0.5, 1, 2, 5, 7.5, 10, 12.5, 15, 20, and 100 nM) or IGF1 (100 nM). The k_off rates ± SD for three independent SPR experiments were plotted against the log of the IGF2 concentration. Statistical analyses (ordinary one-way ANOVA, GraphPad Prism v 10.2.0) were performed comparing each concentration of IGF2 (red) or IGF1 (black) to 0 nM IGF2. ns, not significant (P ≥ 0.05), *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7

Negative-staining electron microscopy (EM) of sCI-MPR with and without IGF2. Negative-staining EM studies of sCI-MPR were performed using 1.4 μM of the receptor prepared in the absence and presence of 200 nM IGF2 at pH 6.8. Survey micrograph of sCI-MPR in the absence (A) and presence (D) of IGF2. Twenty-five representative images of sCI-MPR particles without (B) and with (E) IGF2. Ten representative reference-free class averaged images of the particles without (C) and with (F) IGF2. (G) The dimensions of 576 individual particles from each condition were measured. The length ratio of each particle was calculated and plotted as a function of the percentage of the total population of particles. This graphical representation highlights the conformational differences of sCI-MPR in the absence and presence of IGF2.

Fig. 8

HRPF analysis of topographical changes of sCI-MPR (1 μM) upon binding IGF2 (1 μM) at pH 7.4. Peptides showing increases in exposure (green spheres) and protection (red spheres) upon IGF2 binding in at least two of the three independent oxidation experiments are mapped onto the cryo-EM structures of CI-MPR. Because there is insufficient resolution of the N-terminal 3 domains of the receptor bound to IGF2 in the cryo-EM structure at pH 7.4 (PDB 6UM2), peptides located in domains 1–3 were mapped onto the receptor in the ligand unbound state using the cryo-EM structure at pH 4.5 (PDB 6UM1). Bound IGF2 is shown in blue spheres. Created using PyMOL Molecular Graphics System, Version 2.5.4 Schrödinger, LLC.