An atlas of O-linked glycosylation on peptide hormones reveals diverse biological roles

Abstract

Peptide hormones and neuropeptides encompass a large class of bioactive peptides that regulate physiological processes like anxiety, blood glucose, appetite, inflammation and blood pressure. Here, we execute a focused discovery strategy to provide an extensive map of O-glycans on peptide hormones. We find that almost one third of the 279 classified peptide hormones carry O-glycans. Many of the identified O-glycosites are conserved and are predicted to serve roles in proprotein processing, receptor interaction, biodistribution and biostability. We demonstrate that O-glycans positioned within the receptor binding motifs of members of the neuropeptide Y and glucagon families modulate receptor activation properties and substantially extend peptide half-lives. Our study highlights the importance of O-glycosylation in the biology of peptide hormones, and our map of O-glycosites in this large class of biomolecules serves as a discovery platform for an important class of molecules with potential opportunities for drug designs.

Fig. 1. Summary of the identification of glycosylated peptide hormones using LC–MS/MS.

a Number of glycosites identified on mature and propeptide hormone proteins. b Heatmap illustrating the identification of glycans on mature peptide hormones across different biofluids, tissues, and cell lines. Greyscale intensity illustrates the number of extraction methods that have resulted in the identification of a given glycopeptide. See Supplementary Data 4 for examples of extracted spectra of glycopeptides from each peptide hormone. c Bar graph of the total number of mature glycopeptide hormones identified in each tissue or cell line (all enzymatic digestion procedures combined). d Venn diagram illustrating the distribution of identified glycosylated peptide hormones across biosources between the different extraction methods (total protein extractions (Rapigest extractions and crude biofluids), low molecular weight extractions (LMWE: H₂O, acetone, acetone/AcOH, AcOH, and EtOH/HCl) and cell lines (supernatant and total cell lysate)). Please see Supplementary Data 1 for peptide hormone abbreviations.

Fig. 2. Glycosylation of peptide hormones is a widespread phenomenon.

Bar graph illustrating that 28 out of the 46 human families annotated in the NeuroPep database contain at least one glycosylated member (yellow). The glycosylation prediction algorithm, NetOGlyc 4.0, predicts that almost all (41 out of 46) families contain glycopeptide hormones (light gray) and five families only have members without glycosylation (dark gray). The number of identified peptide hormones that NetOGlyc 4.0 does not predict are marked by a dotted line. Please see Supplementary Data 1 for peptide hormone family abbreviations.

Fig. 3. Topology of O-glycosites in peptide hormones and their predicted functions.

a Representation of a generic peptide hormone with indications of potential functional impact of glycosylation sites. b Illustrations of glycopeptide hormone families with message-associated O-glycosites conserved or semi-conserved among the paralogous family members. The message parts of the peptide hormones are indicated by light blue circles. Glycans have been identified in conserved positions in family members in red. Faded glycosites are not shared within paralogous family members. (*) Denotes peptide hormones containing non-conserved glycosites. (#) The site over the dotted line in the insulin family is covered by ambiguously assigned glycosite(s) in the C-terminal part of IGFII. Full sequence alignments of the five presented glycopeptide hormone families are shown in Supplementary Fig. 3. Please see Supplementary Data 1 for peptide hormone abbreviations.

Fig. 4. Functional analysis of O-glycans within the receptor-binding region of peptide hormones.

a Schematic depiction of the Thr⁷ O-glycan positioned within the receptor-binding domain of the glucagon hormone family (see Supplementary Fig. 7e). b Schematic depiction of the Thr³² O-glycan positioned within the receptor-binding domain of the NPY hormone family. c Structural representation of the sialyl-T O-glycan. d Potencies (Log₁₀(EC₅₀)) of glycosylated GCG, VIP, and GLP-1 determined in cell-based receptor activation assays (n = 3 independent experiments in duplicate determinations for all groups, except non-glyc. GCG where n = 5). e Potencies of NPY and PYY in their Thr³² glycoforms determined and shown as in d (n = 3 for all groups). All error bars represent mean ± S.E.M. Two-tailed one-way ANOVA (Tukey’s post hoc test) was performed in d and e except for comparisons within NPY2R+PYY-; NPY5R+PYY-, and NPY5R+NPY-assays in e where two-tailed Student’s t-test was performed. ns not significant; *p < 0.05; **p < 0.01; ***p < 0.001 (all significant p-values are <0.0001 except d GCGR-Tn/GCGR-T p = 0.0149; e NPY2R-PYY-Non-glyc/NPY2R-PYY-Tn p = 0.0027; NPY2R-NPY/NPY2R-NPY-Tn p = 0.0002; NPY2R-NPY-Tn/NPY2R-NPY-T p = 0.0046). Source data for d and e are provided as a Source data file.

Fig. 5. Molecular and structural analysis of NPY and its precursor.

a Western blot of conditioned media from STC-1 cells stably expressing proNPY using an antibody recognizing pro- and mature NPY. The conditioned media from proNPY-transfected STC-1 cells was concentrated using TCA precipitation, and analyzed with and without neuraminidase treatment (Neu). b Western blot of conditioned media from proNPY-transfected HEK293-6E cells were analyzed with and without neuraminidase treatment (Neu), OpeRATOR treatment, or both in combination. Both western blot experiments in a and b were repeated three times with similar results. c Alpha-helical content of 20 µM NPY or its glycoforms determined by CD spectroscopy (pH 7 at 25 °C) (n = 4 independent measurements for all groups except Tn where n = 3). All error bars represent mean ± S.E.M. Two-tailed one-way ANOVA (Tukey’s post hoc test) was performed in c. ***p < 0.001, **p < 0.01 (all p-values are <0.0001 except c Non-glyc/Tn where p = 0.0018). Source data for a–c are provided as a Source data file.

Fig. 6. Stability of glycosylated peptide hormones in vitro and in vivo.

a Summary of in vitro degradation of peptide hormones using purified NEP or DPP-IV monitored by matrix assisted laser desorption/ionization-time of flight (MALDI-TOF) timecourse assay (See Supplementary Fig. 8 for spectra). Qualitative scoring ++: fully susceptible to degradation, +: reduced susceptibility, −: not susceptible to degradation. b Plasma concentrations of GLP-1 or glycoforms over time after intravenous bolus injection (20 nmol per kg) in anesthetized rats (n = 8 rats per group). GLP-1 was measured by an assay targeting the amidated C-terminus of the molecule measuring total GLP-1 (intact 7–36 amide and metabolites hereof). Half-lives (T_1/2) are indicated with 95% confidence intervals. For each glycoform, timepoints represented by n > 4 measurements were included in the analysis (n_(GLP-1) = 35, n_{(GLP-1 T)} = 38, n_{(GLP-1 ST)} = 41 measurements in total). Multiple linear regression was performed in b. ns not significant; ***p < 0.001 (all indicated p-values < 0.0001). Source data for b is provided as a Source data file.