Phosphopeptidomics Reveals Differential Phosphorylation States and Novel SxE Phosphosite Motifs of Neuropeptides in Dense Core Secretory Vesicles

Abstract

Neuropeptides are vital for cell-cell communication and function in the regulation of the nervous and endocrine systems. They are generated by post-translational modification (PTM) steps resulting in small active peptides generated from prohormone precursors. Phosphorylation is a significant PTM for the bioactivity of neuropeptides. From the known diversity of distinct neuropeptide functions, it is hypothesized that the extent of phosphorylation varies among different neuropeptides. To assess this hypothesis, neuropeptide-containing dense core secretory vesicles from bovine adrenal medullary chromaffin cells were subjected to global phosphopeptidomics analyses by liquid chromatography (LC)-mass spectrometry (MS/MS). Phosphopeptides were identified directly by LC-MS/MS and indirectly by phosphatase treatment followed by LC-MS/MS. The data identified numerous phosphorylated peptides derived from neuropeptide precursors such as chromogranins, secretogranins, proenkephalin and pro-NPY. Phosphosite occupancies were observed at high and low levels among identified peptides and many of the high occupancy phosphopeptides represent prohormone-derived peptides with currently unknown bioactivities. Peptide sequence analyses demonstrated SxE as the most prevalent phosphorylation site motif, corresponding to phosphorylation sites of the Fam20C protein kinase known to be present in the secretory pathway. The range of high to low phosphosite occupancies for neuropeptides demonstrates cellular regulation of neuropeptide phosphorylation. Graphical Abstract ᅟ.

Keywords: Adrenal medulla; Chromogranin; Fam20C; Neuroendocrine; Neuropeptide; Phosphatase; Phosphopeptidomics; Phosphosite; Phosphosite occupany; Post-translational modification (PTM); Proenkephalin; Prohormone; Secretogranin; Secretory vesicle; VIF.

Figure 1. Workflow of DCSV peptidomics and phosphopeptidomics

A purified peptide pool from bovine adrenal cell DCSVs was split for quantitative phosphatase experiments that utilize alkaline phosphatase (+AP) and no AP (−AP) treatments and qualitative phosphopeptide enrichment experiments (IMAC). Phosphosite occupancy was directly measured by intensity comparisons of phosphopeptides and their non-phosphorylated counterparts in −AP samples. Occupancy was also inferred by the intensity differences of non-phosphorylated peptides between +AP and −AP samples.

Figure 2. In silico analysis of known prohormone and endogenous peptide phosphorylation

A.I.) iceLogo amino acid motif analysis of phosphosites on bovine prohormones. The background set consisted of all known phosphosites among all proteins in the Uniprot canonical protein sequence database for Bos taurus (UP000009136). The positive set consisted of the phosphosites on bovine prohormones. Percent-differences of residues in black are at a significance level of p < 0.1, while residues in red and blue are significant at p < 0.05. A.II.) Bar charts show the number of SxE phospho site motifs of prohormones and endogenous peptides compared to the number of known (identified) phosphosites in the Uniprot protein database for B. taurus. In B.I and B.II, parallel analyses of human prohormones and endogenous peptides from all reviewed proteins in the Uniprot canonical protein sequence database for Homo sapiens (UP000005640) are shown.

Figure 3. Phosphosite localization on DCSV peptides

A) Distribution of localized phosphosites—unambiguous identification of the phosphorylated residue on a phosphopeptide (AScore ≥ 13)—on peptides from various neuropeptide and protein precursors. The Y-axis represents the number of unique single-residue and multi-residue phosphosite localizations per precursor. Precursors in red denote known prohormones. B) The Venn diagram shows the overlap of single-residue phosphosite localization in each experiment. Results from digested and undigested IMAC-enrichments were combined into a single dataset (blue circle).

Figure 4. Peptide quantification metrics in +AP/−AP experiments

A) Log₂(+AP/−AP) intensity ratio histogram for all quantifiable undigested peptides. B) Plot of −Log₁₀(p-value) versus Log₂(+AP/−AP) for all quantifiable undigested peptides. Blue dots denote non-phosphorylated peptides; green dots denote localized phosphoserine (pS), phosphothreonine (pT), and unlocalized (UL) phosphopeptides; and red dots denote localized phosphotyrosine phosphopeptides. The dotted lines denote significance thresholds [Log₂(+AP/−AP) > +2.0 ‖ Log₂(+AP/−AP) < −2.0; p < 0.05]. C) and D) contain parallel histograms and scatter plots for tryptic peptides.

Figure 5. Direct phosphosite occupancy measurements for chromogranin B (CHGB)

A) Colored bars represent endogenous phosphopeptide IDs and their location in the prohormone sequence. The color of the bar denotes phosphosite occupancy, calculated by comparison of its intensity to that of its non-phosphorylated counterpart by direct measurements (in −AP samples). Residues in green denote verified phosphosites. Residues in purple denote di/tri-basic cleavage sites for endogenous DCSV proteases. Neuropeptides with established bioactivity originate from CHGB regions highlighted in yellow.

Figure 6. Distribution of phosphosite occupancy measurements by direct and inferred methods

A) A pie chart displays the direct only, direct and inferred, and inferred only methods utilized for phosphosite occupancy measurements for each of the 182 single-residue phosphosites that were identified and quantified in this study. Red indicates that occupancy was only measured directly, blue indicates that occupancy was only be inferred through +/− AP experiments, and green indicates site occupancies that were measured with both methods. B) A Venn diagram of single-residue phosphosite occupancies that could only be measured by inferred methods shows the overlap of site occupancies measured from intact endogenous peptides (no digest) and tryptic peptides (Trypsin/LysC).

Figure 7. PEP (peptide-averaged endogenous phosphosite) intensity and occupancy on VIF and enkelytin

Relative +AP (blue) and −AP (red) PEP intensities of verified phosphosites on A) VIF and B) Enkelytin. VIF peptide is residues 97–131 of the chromogranin A (CHGA) precursor. Enkelytin peptide is residues 233–261 of the proenkephalin (PENK) precursor. PEP intensity was normalized to largest intensity observed among all sites listed. Error bars represent standard error of the mean. An asterisk (*) indicates that the Log₂(+AP/−AP) > 2.0 and p-value < 0.05.

Figure 8. Hierarchical clustering of PEP intensity and phosphosite iceLogo analysis

A) Hierarchical clustering of single-residue Peptide-averaged Endogenous Phosphosite (PEP) intensity. Each column represents a single-residue phosphosite, and each row/color denotes the Log₂-scale mean-normalized PEP intensity within +AP and −AP experimental replicates. Sites were clustered according to the Euclidean distances between intensities from each +AP and −AP experiment. B) iceLogo analysis of all phosphosites from PEP clusters in A). The background set consisted of all annotated Uniprot Bos taurus phosphosites, as detailed in Figure 1. Percent differences of residues in black are significant at p < 0.1, while residues in red and blue are significant at p < 0.05. The percent difference represents the change in frequency of an amino acid in the dataset vs. all known bovine phosphosite motifs.