A clean-up strategy for identification of circulating endogenous short peptides in human plasma by zwitterionic hydrophilic liquid chromatography and untargeted peptidomics identification

Highlights

• Development of an efficient clean-up strategy for short peptides in plasma.

• Comparison of 4 different clean-up protocols (advantages and disadvantages).

• Untargeted peptidomic approach for short peptide identification in plasma.

• Identification of 91 short peptides, the largest number never identified before.

• Development of a rapid and simple method for clinical applications.

Abstract

Short peptides, namely di- tri- and tetra peptides, have been proven to play an important diagnostic role in several diseases. Therefore, the development of an analytical approach for their detection and identification is nowadays an important research goal. This paper describes an analytical procedure able to overcome the issues of short peptide isolation, clean-up and identification in plasma samples. Four different protocols were compared and tested to maximize both recovery and total number of identifications of short circulating plasma endogenous peptides. The purified peptides, coming from the four different tested protocols, were separated by zwitterionic hydrophilic liquid chromatography coupled to high-resolution mass spectrometry with the purpose of accomplishing an untargeted investigation based on suspect screening for short peptides in plasma. In particular, the use of Phree™ Phospholipid removal cartridge in combination with a purification step by solid phase extraction on a graphitized carbon black sorbent allowed the identification of the largest number of amino acid sequences (91 short peptides). The clean-up procedure allowed to tackle the issue of the low abundance of such peptides and their suppression during mass-spectrometric analysis. The results indicated that sample preparation is therefore fundamental for short peptide analysis in plasma samples.

Introduction

Nowadays, peptidomic analysis has become an independent approach in the omics science in several research fields, such as therapeutics [1] diagnostics [2] and nutraceuticals [3] especially for the biological significance of peptides. Small peptides, namely those possessing 2–4 amino acids in their sequence, have been specifically addressed in food and related for the discovery of new bioactive peptides [4], [5], [6], [7], [8]. Biofluids have currently been considerably less investigated, nevertheless diagnostics has been emerging as a considerable research field due to the role that circulating peptides can play in different diseases, such as pancreatic neuroendocrine tumor [9], papillary thyroid cancer [10], acute leukemia [11], amyotrophic lateral sclerosis [12] and Alzheimer's disease [13,14]. The identification of peptides is therefore important because they can mediate specific functions and enable peripheral detection and they play a crucial role in the progression and spread of the disease, being involved in the processes of cancer cell growth, angiogenesis, and metastasis [15]. Finally, amino acid sequences can be considered biomarkers for several diseases also in personalized medicine approaches [16].

Nevertheless, all peptidomics studies were focused on medium size peptide analysis, since detection and identification of short peptides is still an analytical challenge due to the wide range of polarity of these peptides and their low concentration in complex biological samples [17]. It is well known that small peptides, are nowadays of great importance because of their contribution in various biological processes, including protein synthesis, fertility, neurotransmission, inflammation process, pathogenic microorganism activities and other functions. These functions have made these peptides vital molecules in drug development and health care. Besides, these peptides are also being used as biological markers in biological systems [18].

Despite this, only metabolomics studies indicated some short peptides, together with other molecules, as possible biomarkers [18,19]; only three recent peptidomics papers were specifically meant for 2–4 amino acid long native peptides, one in plasma for diagnosis and prognosis of epithelial ovarian cancer [20], one in serum, for the development of an analytical platform for short peptides detection [21], and one in urine, for the selective enrichment of very short peptides [22]. As far as plasma or serum was concerned, in both papers a simple protein precipitation followed by chromatographic separation (either by C18 revesed phase (RP) or on porous graphitic carbon) was carried out, leading to the identification of a low number of peptides, 15 and 32, respectively. The low number of identifications was probably a consequence of the low electrospray ionization (ESI) efficiency, low concentration and quick degradation of such peptides in the peripheral blood. In fact, a loss of information can occur due to the suppression of short peptides caused by high abundant species, such as proteins and lipids. To overcome those issues, a pretreatment of the sample is mandatory and several strategies have been proposed for the selective removal of high abundance proteins and enrichment of low abundance ones [23].

There are some analytical issues related to the analysis of short amino acidic sequences: their wide range of polarity and concentration make their detection difficult if dedicated protocols for enrichment and clean-up are not performed. Moreover, the bioinformatic identification is not possible with the common databases employed in metabolomics or proteomics, since metabolomics databases currently do not cover all the 168,400 possible combinations of proteinogenic amino acids in di-, tri- and tetrapeptides and proteomic software furnish a too low identification score for short peptide sequences [8,24].

In this context, an analytical platform for the extraction and clean-up of circulating endogenous short peptides in plasma was developed. Four different analytical clean-up methodologies was tested and compared in order to maximize the results in terms of small peptide identification. The best results were achieved by the means of simultaneous protein precipitation and phospholipid removal using the commercial Phree™ Phospholipid removal cartridge, followed by a clean-up step of short proteinogenic peptides by graphitized carbon black (GCB) solid phase extraction (SPE). The chromatographic separation of both hydrophilic and hydrophobic short peptides was carried out by ultra-high performance liquid chromatography (UHPLC) by zwitterionic hydrophilic interaction chromatography (HILIC) coupled tohigh resolution MS (HRMS).Short peptide identification was carried out by an untargeted investigation based on suspect screening. More specifically, a database with all combinations of the 20 natural amino acids within di-, tri- and tetrapeptides was compiled and fragment ions spectra were matched to in silico fragmentation to identify the precursor ions.