Glycan-specific precipitation of glycopeptides in high organic content sample solvents used in HILIC

Abstract

The composition of a sample solvent has a crucial impact on separations in hydrophilic interaction liquid chromatography (HILIC). In this short communication, we studied the effect of an organic modifier in the sample solvent on the solubility of different tryptic glycopeptides of hemopexin and haptoglobin proteins. The results showed that the solubility of glycopeptides in solvents with a high acetonitrile content depends on the type of attached N-glycan. We observed lower solubility in larger glycans attached to the same peptide backbone, and we demonstrated that glycopeptides containing sialic acids precipitate more readily than those without sialic acid. Therefore, the sample solvent composition in HILIC must be carefully optimized for accurate quantitative data collection and for adequate separation.

1. Introduction

Protein glycosylation is one of the most frequent post-translational modifications involved in many biological and pathological processes [1]. The function of the attached glycans depends on their structure, on the site of attachment to the protein, on their structural variability at this site and on the extent to which they occupy the site [2, 3]. A standard technique of glycoproteomics analysis is liquid chromatography coupled to mass spectrometry (LC-MS), although the ionization efficiency of glycopeptides for mass spectrometry (MS) is lower than that of bare peptides, and the fragmentation of both glycan and peptide moieties is often incomplete [4]. Although reversed-phase (RP) chromatography is the method of choice for proteomics experiments, different glycoforms of a peptide are usually inadequately resolved when using this method [6]. However, Ji et al. have recently reported the separation of N- and O-linked isomeric sialylated glycopeptides using high-temperature RP chromatography [7]. HILIC provides an alternative LC mode for the separation of glycopeptides with higher selectivity than the RP mode. HILIC methods were successfully used in glycoproteomics for sample preparation (enrichment, desalting) [8–11] and for the separation of complex glycan/glycopeptide mixtures by LC [12–18]. Moreover, its potential to separate isomeric structures of glycans [14] and glycopeptides has been demonstrated already [13, 15, 17].

In our previous study [13], we observed that some glycopeptides are less abundant in HILIC than in RP-LC. This observation led us to hypothesize that some peptide glycoforms had limited solubility. Thus, the aim of this short communication was to investigate the effect of the high organic content of a sample solvent on the solubility of glycopeptides. This information is crucial for HILIC because the increased organic content of a sample solvent improves peak shape and renders retention times reproducible. HILIC uses mobile phases composed of aqueous and organic components with the organic component in an excess of 60 percent volume [19]. Due to the HILIC principle, acetonitrile is the most popular organic component of the mobile phase in typical glycopeptide analyses. This aprotic solvent does not interfere with the aqueous layer formed on the stationary phase and thus has no adverse effect on chromatography [20]. The sample solvent composition usually corresponds to the starting point of the HILIC gradient (80% or higher organic solvent concentration). We tested the effects of acetonitrile content in the sample solvent (90, 80, 70, and 60 %) and temperature (4 and 20 °C) on the solubility of hemopexin and haptoglobin glycopeptides. To the best of our knowledge, this is the first study to systematically investigate these effects.

2. Materials and methods

2.1. Chemicals

Acetonitrile (LC-MS grade), water (LC-MS grade), formic acid (LC-MS grade), ammonium bicarbonate (purity ≥ 99%), iodoacetamide (purity ≥ 99%), dithiothreitol (purity ≥ 99%), and SOLu-Trypsin were supplied by Sigma-Aldrich (St. Louis, MO). α2-3,6,8,9 neuraminidase, with its appropriate glycobuffer, was purchased from New England BioLabs (Ipswich, MA). Hemopexin and haptoglobin from human plasma were supplied by Athens Research and Technology (Athens, Georgia,).

2.2. Sample preparation

Hemopexin and haptoglobin glycopeptides were prepared by tryptic digestion according to the previously published protocol [6]. An aliquot of the tryptic digest was desialylated with α2-3,6,8,9 neuraminidase at 37 °C overnight. Next, desialylated and sialylated glycopeptides of one protein were mixed in a 1:1 ratio, dried using a vacuum concentrator (Labconco, Kansas City, MO) and reconstituted in a solution of 0.1% formic acid in 2% acetonitrile. Glycopeptides should be fully dissolved under these conditions and this mixture was used as a reference. An appropriate portion of this solution was diluted in acetonitrile to prepare samples with a solvent composition of 90, 80, 70, and 60% acetonitrile and with protein concentration of 0.05 μg/μL. One set of samples was kept at room temperature for 10 minutes and the other one at 4 °C for 24 hours (a common temperature in autosamplers) except for the sample in 90% acetonitrile, which was tested only at room temperature. Samples were then spun down, diluted five times with 0.1% formic acid in 2% acetonitrile, and transferred into an LC vial.

2.3. Instrumentation and experimental conditions

Chromatographic measurements were performed on a nanoAcquity UPLC system with a binary pump (Waters, Milford, MA) coupled with a 6600 TripleTOF mass spectrometer (Sciex, Framingham, MA). Glycopeptides were analyzed by RP chromatography on an Acquity UPLC M-class Symmetry C18 Trap Column (5 μm, 180 μm × 20 mm) and an Acquity UPLC Peptide BEH C18 nano column (1.7 μm, 75 μm × 150 mm) both from Waters (Milford, MA). The mobile phase consisted of 0.1% formic acid in 2% acetonitrile (Solvent A) and 0.1% formic acid in 100% acetonitrile (Solvent B). The flow rate of the mobile phase was maintained at 0.4 μL/min, and the column temperature was 40 °C. The optimized gradient program included a 5-min trapping step at 1% B at 15 μL/min followed by the following [(min)/% B]: 0/1, 40/40, 45/95, 50/95, 57/1, and 60/1. The injection volume was 1 μL, and the samples were kept at 15 °C between analyses. An information-dependent acquisition workflow was used for the assignment of N-glycopeptides. The identification of N-glycopeptides was performed manually based on precursor masses and characteristic fragments in line with the settings described in our previous paper [6].

3. Results and discussion

For comparison purposes, we selected two hemopexin (SWPAVGNCSSALR; ALPQPQNVTSLLGCTH) and two haptoglobin peptides (VVLHPNYSQVDIGLIK; MVSHHNLTTGATLINEQWLLTTAK) carrying various glycoforms, as described below. Each sample was measured in triplicate. The results are expressed as median relative peak area for glycopeptides of hemopexin (Figure 1) and haptoglobin (Figure 2). The relative peak areas are calculated as the peak area of a specific glycopeptide glycoform dissolved in a solvent with high acetonitrile content divided by corresponding peak area in the reference and multiplied by 100. Figure 1A shows high precipitation of all studied glycoforms of ALPQPQNVTSLLGCTH peptide in 90% acetonitrile. More than 93 % of their original content precipitated. The decrease in the acetonitrile content of the sample solvent increased the solubility of all glycoforms, but precipitation still occurred to some extent in 80% acetonitrile. Moreover, we observed elevated precipitation during the storage of the samples at 4 °C for 24 hours, which is caused by the lower solubility of glycopeptides at a lower temperature. The increase in the number of neutral saccharide units increases the hydrophilicity of the glycopeptide and decreases its solubility in 80% acetonitrile. This behavior is observed in bi-antennary, fucosylated bi-antennary, and tri-antennary glycoforms. Additionally, the presence of sialic acid in the glycan chain results in a substantial decrease in the solubility of glycopeptides in sample solvent of higher acetonitrile content, resulting in higher precipitation, most likely because sialic acid (log P = −3.5) is more polar than neutral monosaccharides, e.g., galactose (log P = −2.6) or fucose (log P = −2.4) (https://www.ncbi.nlm.nih.gov/pccompound). Fully solubilized ALPQPQNVTSLLGCTH glycopeptides were observed in 70% acetonitrile for neutral glycans and in 60% acetonitrile for glycopeptides with sialic acids. We observed the following trends: 1. significant precipitation of all studied glycopeptides in 90% acetonitrile; 2. higher precipitation of glycopeptides with larger glycans in 80% acetonitrile; 3. Lower solubility of glycopeptides at lower temperature; and 4. strong effect of sialic acids in the structure of glycopeptides on their solubility.

Figure 1.

The effect of acetonitrile content in the sample solvent on the precipitation of studied glycoforms of the A) ALPQPQNVTSLLGCTH and B) SWPAVGDCSSALR peptides of hemopexin. Symbols: blue square, N-acetylglucosamine (GlcNAc); yellow circle, galactose (Gal); red right-pointing triangle, fucose (Fuc); green circle, mannose (Man); purple diamond, sialic acid (SA).

Figure 2.

The effect of acetonitrile content in the sample solvent on the precipitation of studied glycoforms of the A) VVLHPNYSQVDIGLIK and B) MVSHHNLTTGATLINEQWLLTTAK peptides of haptoglobin (for symbols and legend see Fig. 1.).

The same trends were observed for all glycopeptides examined. However, sialylated hemopexin SWPAVGNCSSALR (Figure 1B) and haptoglobin VVLHPNYSQVDIGLIK (Figure 2A) glycopeptides precipitated in a higher degree than sialylated hemopexin ALPQPQNVTSLLGCTH glycopeptides, possibly because these two peptide backbones show higher overall polarity than ALPQPQNVTSLLGCTH backbone. In turn, sialylated glycoforms of the haptoglobin peptide MVSHHNLTTGATLINEQWLLTTAK show higher solubility (Figure 2B). More than 65 % of the original sialylated glycopeptide content in standard was found in the sample solvent composed of 80% acetonitrile. We also observed that all of its studied glycoforms in the solvents with higher acetonitrile content were more soluble than the aforementioned glycopeptides. We assume that this behavior is caused by the lower polarity of this peptide backbone and by the higher solubility in organic solvents, that is, acetonitrile. Based on the aforementioned observations, glycopeptide solubility in the high content of acetonitrile in the sample solvent depends on the type of attached N-glycan and on the polarity of the peptide backbone. In addition, the composition of the sample solvent may affect not only the solubility of glycopeptides but also the HILIC chromatography, which we previously showed in the HILIC analysis of hemopexin glycopeptides [13]. Reduction of acetonitrile content in sample solvent from 80% to 70% caused retention time shift of all studied glycoforms, and reduction to 60%caused peak splitting [13]. On the other hand, we did not observe any influence of acetonitrile content in the sample solvent on the peak shape of studied glycopeptides in RP chromatography. The extracted ion chromatograms ([M+3H]³⁺ or [M+4H]⁴⁺ ions) of all studied glycopeptide standards obtained under RP-LC-MS conditions are shown in Supporting Information Figures S1–S4.

4. Conclusion

In this short communication, we assessed the effect of acetonitrile in the sample solvent on the solubility of different glycoforms of tryptic peptides of hemopexin and haptoglobin proteins. We showed that glycopeptides precipitate in acetonitrile content higher than 70 percent volume according to the type of the attached N-glycan. Specifically, sialoglycopeptides precipitate more readily than glycopeptides without sialic acid. This observation is very important for HILIC chromatography where a higher content of acetonitrile in the sample solvent is highly advantageous but may lead to losses of analytes due to precipitation. This glycan-specific glycopeptide precipitation will adversely affect the quantitative analysis, for example, in glycopeptide-biomarker research. Sample solvent composition in HILIC must therefore be carefully considered for reliable quantitative data collection.

Highlights.

The effect of acetonitrile in the sample solvent on solubility of glycopeptides

Glycopeptides precipitate according to the type of the attached N-glycan

Sialoglycopeptides precipitate more readily compared to those without sialic acid.