The glucose-dependent insulinotropic polypeptide (GIP) is a 42-amino-acid polypeptide secreted by enteroendocrine K-cells that potentiates the glucose-dependent release of insulin from pancreatic β cells and exerts extrapancreatic glucoregulatory activities through its systemic receptors. GIP is encoded by the GIP gene and biosynthesized from the 153 amino acid precursor protein by removal of the flanking 30- and 81-amino-acid extensions at the N- and C-terminals, respectively. The C-terminal end amino acid residue of the N-terminal peptide, arginine, is thought to be removed by carboxypeptidases during peptide biosynthesis. However, the presence and exact amino acid sequence of such endogenously processed peptides in human plasma remain largely unknown because of the technical difficulties associated with plasma peptidomic analysis. Furthermore, this putative N-terminal peptide has no known biological activity.
Although human plasma represents an informative resource describing the proteome, identifying plasma bioactive peptides and disease biomarkers remains extremely difficult because these endogenous peptides exist at trace levels, while the plasma contains an extraordinarily dynamic range of high-molecular-weight proteins. Thus, direct in-depth plasma peptidomic profiling remains challenging in contemporary analytical biochemistry. We previously identified novel bioactive peptides by predicting putative endogenous peptide sequences from the bioinformatic analysis of human cDNA database information, explored biological activities of the synthesised peptides, and confirmed their immunoreactive presence in human plasma and tissues. This method identified potent bioactive peptides with physicochemical characteristics that prevented their identification by conventional methods. An in silico search for peptides with specific functional motifs was used to discover bioactive peptides. However, to complete the discovery process of putative peptide hormones, their exact native amino acid sequences in the peripheral circulation needs to be confirmed by mass spectrometry. Despite long-term endeavours to elucidate the plasma proteome, the comprehensive identification of plasma low-molecular weight native peptides has not been achieved until very recently.
In a recent study, we established improved technology that improved the high-yield plasma extraction technique, enabling the large-scale identification of plasma native peptides with mass spectrometry. We deposited the results into a newly generated human plasma native peptidomic database. We also synthesised peptides of identified sequences and validated their biological activity using cultured human cells. The present study aimed to use this plasma native peptidomic resource to identify a novel plasma native polypeptide hormone that may be involved in the pathophysiology of vascular diseases.
Our initial peptidomic sequencing data acquired from 189 analyses using liquid chromatography tandem mass spectrometry (LC–MS/MS) were searched against the SwissProt_2015_02.fasta database with two different data processing pipelines and search engines: (1) the Mascot Distiller (version 2.5.1.0, Matrix Science) deconvoluted the MS/MS spectra and performed an MS/MS ion search, and (2) the PEAKS Studio (version 7) used a PTM algorism and performed a database search based on a de novo sequencing. These analyses were conducted until the beginning of 2018 and after excluding peptides derived from the keratin protein family, they resulted in the identification of 18,552 polypeptide sequences with a peptide identification false discovery rate (FDR) of 1%. All these acquisition data were already deposited in the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD003533. As described in the Methods section, we reanalysed these resources using the PEAKS Studio (version X), and a database search was performed against the SwissProt_2020_03.fasta with five variable PTMs in addition to the carbamidomethylation without using the PTM algorithm. This process resulted in the identification of 7,959 or 11,256 distinct native peptide sequences depending on FDRs of 0% or 1%, respectively. We performed a bioinformatic analysis of this updated resource to select bioactive peptide candidates for chemical synthesis and functional validation using the criteria described in the Methods section. Synthesised peptides were tested for their high purity and liquid solubility using LC–MS/MS analysis and 135 peptides were confirmed to be appropriate for biological validation (Supplementary Table S1).
Of the 135 synthetic peptides, we completed functional screening of 129 peptides in three types of cultured human cells, including the human monocytic leukaemia cells (THP1) induced to differentiate into macrophages, aortic endothelial cells (HAoECs), and aortic smooth muscle cells (HAoSMCs). We found that a 30-amino-acid peptide derived from GIP (Fig. 1a,b), EKKEGHFSALPSLPVGSHAKVSSPQPRGPR (GIP_HUMAN[22–51], monoisotopic mass 3179.6951), caused an increase in intracellular free Ca 2+ levels ([Ca 2+]i) in the THP-1-induced macrophages. The ability of GIP_HUMAN[22–51] to bind and elicit intracellular responses was examined, using the three types of cultured human cells. These cells were incubated with the peptide labelled with 5-carboxyfluorescein at the N-terminus (FAM-GIP_HUMAN[22–51]) to determine whether GIP_HUMAN[22–51] was bound to the intact cell surface. Confocal immunofluorescence microscopy revealed the presence of FAM-GIP_HUMAN[22–51] on the surface of the HAoECs (Fig. 2a) and THP1-derived macrophages (Fig. 2b), but not in the HAoSMCs. Of the remaining 128 peptides, two suprabasin-derived peptides elicited significant cellular responses in the HAoSMCs. Three other peptides induced marginal increases in [Ca 2+]i in the HAoECs; however, their detailed biological activity has not subsequently been investigated.
