In order to comprehensively explore the diversity of O-linked glycans, a new procedure for in-gel reductive β-elimination, combined with permethylation and a rapid phase-partition method, is applied to the analysis of O-linked glycans directly released from glycoproteins resolved by SDS-PAGE and amenable to subsequent glycomic analysis by mass spectrometry.
Separation of proteins by SDS-PAGE followed by in-gel proteolytic digestion of resolved protein bands has produced high-resolution proteomic analysis of biological samples. Similar approaches, that would allow in-depth analysis of the glycans carried by glycoproteins resolved by SDS-PAGE, require special considerations in order to maximize recovery and sensitivity when using mass spectrometry (MS) as the detection method. A major hurdle to be overcome in achieving high-quality data is the removal of gel-derived contaminants that interfere with MS analysis. The sample workflow presented here is robust, efficient, and eliminates the need for in-line HPLC clean-up prior to MS. Gel pieces containing target proteins are washed in acetonitrile, water, and ethyl acetate to remove contaminants, including polymeric acrylamide fragments. O-linked glycans are released from target proteins by in-gel reductive β-elimination and recovered through robust, simple clean-up procedures. An advantage of this workflow is that it improves sensitivity for detecting and characterizing sulfated glycans. These procedures produce an efficient separation of sulfated permethylated glycans from non-sulfated (sialylated and neutral) permethylated glycans by a rapid phase-partition prior to MS analysis, and thereby enhance glycomic and sulfoglycomic analyses of glycoproteins resolved by SDS-PAGE.
Glycosylation is an essential protein post-translational modification, contributing to organismal physiology, tissue pathology, and cellular recognition 1-3. Despite major advances in analytical glycoscience, characterizing the complete diversity of glycans on a specific protein remains an extremely challenging task, especially on proteins isolated from primary biological sources. Nonetheless, the microheterogeneity of glycoprotein glycans frequently affects functional interactions with other proteins. Therefore, characterization of glycan diversity is essential for understanding the physiological significance of cellular and tissue glycosylation 4,5. In order to understand the contribution of glycoprotein glycosylation to tissue physiology and pathophysiology, robust, sensitive, and comprehensive glycomic analytical techniques have become increasingly important. In proteomic analysis, protein identifications are generally achieved by LC-MS/MS analysis of tryptic peptides 6. Protein digestion can be carried out using a purified protein or proteins resolved by SDS-PAGE following in-gel digestion with proteases such as trypsin 7-9. Pre-enrichment of the protein mixture by SDS-PAGE enhances the depth and accuracy of protein ID. The development of analogous strategies for glycomic analysis of glycoprotein glycosylation lies at the forefront of glycoscience.
The two major classes of glycans are attached to protein backbones through either N-linkage or O-linkage. N-linked glycans are attached to asparagine (Asn) residues found as part of a sequon defined as Asn-X-Ser/Thr/Cys (X is any amino acid except proline), and can be released by enzymatic digestion with peptide-N-glycanase (PNGaseF or A), either in solution, in-gel, or on-blot 10-12. O-linked glycans are mainly attached to serine (Ser) or threonine (Thr) residues. However, only one enzyme has been identified that is capable of releasing O-linked glycans from glycoprotein and it has an extremely limited glycan specificity, releasing only the simplest O-linked glycans. Chemical release strategies remain the method of choice for comprehensive release of O-linked glycans from glycoproteins. Reductive or non-reductive β-elimination, or hydrazinolysis are well-characterized chemical release techniques and are currently the most commonly used approaches for releasing O-linked glycans from glycoproteins 13,14. Although reductive β-elimination has been used to release O-linked glycans from glycoproteins separated by SDS-PAGE, previous approaches required HPLC separation for subsequent analysis 15-17.
Multidimensional MS (MSn) analysis currently provides the richest source of structural data for characterizing glycans released from glycoproteins isolated in the amounts expected from most biological sources. The depth of MS-based structural characterization is greatly facilitated by permethylating the released glycans prior to their analysis. Permethylation enhances ionization and tends to equalize molar signal responses across a broad range of glycan structures 18,19. In addition, permethylation unambiguously tags terminal and substituted monosaccharide moieties with distinctive masses, thereby enhancing structural elucidation 20-23. For example, acidic glycans are generally difficult to detect as non-permethylated species by MS. Although acidic glycans can be detected in negative ion mode by MS, it is impossible to detect both acidic and neutral glycans in the same ion mode. A major advantage of glycan permethylation is that all of the free hydroxyl groups (OH) on a glycan’s monosaccharide substituents will be capped with a methyl group (OCH3 or OMe), thus a sialylated glycan’s charges are neutralized, making them as detectable as permethylated neutral (asialo) glycans. However, the hydroxyls of sulfate moieties on sulfoglycans are resistant to permethylation, resulting in retention of anionic charge, which suppresses ionization and decreases sensitivity. This suppression currently prevents comprehensive glycomic analysis of very complex glycoproteins such as mucins, which carry a high abundance of sulfated glycans 24-26.
Recent reports on purifying sulfated glycans used charged, reverse-phase chromatography to purify and separate permethylated glycans prior to MALDI analysis. This method relies on complete separation of sulfated and non-sulfated glycans using different mobile phases for elution, which we have found to be less stringent than organic phase partitioning. Therefore, new techniques suitable for the detection and enrichment of sulfoglycans are presented here. These techniques allow for the quantitative recovery of sulfated glycans in the aqueous phase following water:DCM (dichloromethane) extraction, which is routinely performed at the end of glycan permethylation reactions 27. Importantly, this robust separation of permethylated sulfoglycans from a mixture of permethylated non-sulfated glycans concomitantly enriches for charged species while also simplifying MS2 fragmentation patterns. A comprehensive protocol for improved in-gel O-linked glycan analysis is also presented. The improved protocol enhances glycan recovery, increases the structural information obtainable through MSn analysis of permethylated glycans, and improves the sensitivity of sulfoglycomic analyses applied to essential glycoproteins isolated from biological sources.
This protocol is intended for O-linked glycan analysis of whole glycoprotein extracts or of a specific glycoprotein of interest resolved by SDS-PAGE and is composed of three experimental procedures; A) gel clean-up, B) in-gel reductive β-elimination, and C) glycan permethylation. The goal is to obtain comprehensive O-linked glycomic data for glycoproteins harvested from primary sources of biological interest (Figure 1). Glycoproteins separated by SDS-PAGE are visualized by staining and bands of interest are excised and the resulting gel band is sliced into small pieces. The gel pieces are destained and subjected to ethyl acetate washes to remove gel contaminants (Figure 2A). Glycan release is achieved by in-gel reductive β-elimination (Figure 2B) and the released glycans are permethylated. Aqueous-organic extraction following permethylation quantitatively partitions the anionic sulfated glycans away from non-sulfated neutral glycans (Figure 2C). In-gel reductive β-elimination coupled to aqueous-organic extraction enables the characterization of O-linked glycans and sulfoglycans released from small amounts of glycoprotein separated by SDS-PAGE. The strategic overview is summarized in Figure 1 and the details are shown in Figure 2. In addition, a portion of the destained and washed gel pieces can be used for protein ID by standard LC-MS/MS proteomic techniques.
NOTE: Lab Safety Concerns
In keeping with standard laboratory best practices, observe the following. Store all organic solvents in appropriate locations. Keep all waste materials in chemical waste containers with clear labeling of chemical compositions. As several reagents used in these protocols are potential carcinogens or generate volatile combustible gases, handle all reagents in a fume hood with ventilation. Wear personal protective equipment such as gloves, lab coat, and eye protection when working with organic solvents.
Figure 1. Procedure for in-gel O-glycomics. (A) Proteins of interest are resolved by SDS-PAGE, detected by appropriate staining procedures (Coomassie or silver), and excised. Excised gel pieces are sliced into small cubes, destained, and washed with ethyl acetate to reduce contaminants that interfere with subsequent MS analysis. A portion of the gel slice can be reserved for in-gel tryptic digestion and subsequent proteomic characterization by LC-MS/MS. (B) O-linked glycans are released from resolved glycoproteins by in-gel reductive β-elimination. Essential steps include desalting and borate removal by azeotrope with methanol. (C) O-linked glycans released by reductive β-elimination are permethylated and subsequently partitioned into aqueous and organic phases. Sulfoglycans are quantitatively recovered in the upper (aqueous) phase while neutral and sialylated glycans partition into the lower (DCM) layer.
Figure 2. Detail flow chart for in-gel glycomics. Each experimental step shown in Figure 1 is individually illustrated. (A) gel destaining and EtOAc extraction, (B) direct in-gel reductive β-elimination for O-linked glycan release, and (C) permethylation and phase partition. Please click here to view a larger version of the figure.
1. Gel Excision and Removal of Gel-derived Contaminants
2. In-gel O-glycan Release by Reductive β-Elimination
3. Desalting on Cation-exchange Chromatography
4. Borate Removal
5. C18 Clean-up
6. Base Preparation for Permethylation
NOTE: In order to achieve robust permethylation, prepare NaOH slurry fresh. All glassware used for permethylation reactions should be extensively cleaned.
7. Permethylation of Released Glycans
8. C18 Clean-up of Permethylated Sulfated O-glycans from the Aqueous Phase
9. Mass Spectrometry of Permethylated O-glycans
Effect of Ethyl Acetate Treatment Prior to In-gel Reductive β-Elimination
A representative mass spectrum of permethylated O-linked glycan samples released from bovine mucin using in-gel reductive β-elimination is shown in Figure 3. The EtOAc wash of the gel pieces effectively removes SDS and polyacryl contaminants which interfere with subsequent MS analysis 27.
Figure 3. Ethyl acetate wash of polyacrylamide gel piece prior to in-gel reductive β-elimination enhances detection of released glycans. (A) Without wash with ethyl acetate, the mass spectrum of permethylated O-linked glycans released by in-gel β-elimination from bovine submaxillary mucin is dominated by an abundance of a polydisperse contaminant, which almost completely obscures the MS signals associated with O-linked glycans. (B) Washing the gel piece with ethyl acetate eliminates the contaminant peaks, allowing sensitive detection of glycans. Modified from Kumagai25.
Recovery of O-linked Glycan is Greater from Small Gel Slices
O-linked glycans were released from bovine submaxillary mucin by in-gel reductive β-elimination using gel pieces that were either sliced small (~2 x 2 mm) or large (~5 x 5 mm). Gel pieces smaller than ~2 x 2 mm were not efficiently recovered through the washing steps. Following permethylation, a known amount of a permethylated external glycan standard (maltotri- and maltotetrasaccharide, Dp3 and Dp4) was added to each to facilitate quantification of glycan recovery 27. The recovery of O-linked glycans from small gel pieces was almost 10-fold greater than from large gel slices (Figure 4).
Figure 4. Glycan recovery from small gel pieces is more efficient than from larger pieces. Small pieces (~2 x 2 mm) yielded more glycan than larger cubes (~5 x 5 mm). Bars show relative signal intensities for the sum of all O-linked glycan structures normalized to the signal detected for an external standard (maltotrisaccharide Dp3, set to 100%), which was added to the released glycan before MS analysis. Values are mean standard deviation for n=3. Modified from Kumagai K25.
Enrichment of Sulfoglycans by Phase Partition
A permethylated sulfated glycan standard (sulfo-Lea) was completely recovered in the aqueous phase. Permethylated sialylated glycans released from the mucin glycoprotein partitioned into the DCM phase (Figure 5). The recovery of each permethylated glycans by aqueous-organic partition was comparable to that of the previously characterized C18 Sep-Pak clean-up method 27. The efficiency and simplicity of the phase partition method greatly facilitates sample throughput and subsequent analytic approaches.
Figure 5. Differential recovery of sulfo- and sialo-glycans by phase partition. Bovine mucin glycoprotein was spiked with a known amount of a sulfo-glycan standard (sulfo-Lea) before being subjected to reductive β-elimination. Released glycans were permethylated and the permethylation reaction was adjusted to 1:1:water:DCM. The resulting organic and aqueous phases were separated and analyzed by MS. Glycan recovery was quantified relative to permethylated external glycan standards, which were spiked into the sample before MS analysis. Glycan recoveries into the organic (DCM) and aqueous (Water) phases are shown relative to the external standard, which was set to 100%. Sulfoglycans were undetectable in the organic phase but quantitatively recovered in the aqueous phase. Results represent the mean ± S.E. of three independent experiments. Modified from Kumagai25.
Application to In-depth Proteomic and Glycomic Analyses in Biological Samples
Human saliva proteins were separated by SDS-PAGE and stained with Coomassie Brilliant Blue (G-250). A high molecular weight protein at MW ~600 kDa was excised from the gel and a portion of the gel piece was subjected to in-gel tryptic digestion and LC-MS/MS-based proteomic analysis, which identified this band as MUC5B 16,24. The remainder of the gel piece was subjected to in-gel reductive β-elimination for O-glycan analysis 27. Following permethylation and aqueous-organic phase partition (Water:DCM, 1:1), permethylated glycans in the aqueous and organic phases were analyzed by NSI-MS. Non-sulfated permethylated O-glycans were recovered from the organic phase and all of the permethylated sulfoglycans were recovered from the aqueous phase, which facilitates the identification and characterization of nearly isobaric sulfated and non-sulfated glycans, e.g., Fuc1Hex3HexNAc2GalNAc-ol (m/z=1606.8, [M+Na]+) and (SO3)1NeuAc1Fuc1Hex2HexNAc1GalNAc-ol (m/z=1606.7, [M+2Na-H]+) differ by only 0.1 mass units and would be difficult to resolve without physically separating the two species by phase partition (Figure 6).
Figure 6. Detection of isobaric complexity in neutral and sulfo-glycans separated by aqueous-organic partition. MS2 fragmentation patterns obtained from parent ions detected by total ion mapping (TIM) analysis of permethylated O-glycans released from human salivary mucin by in-gel β-elimination and water:DCM partition following permethylation. (A) MS2 from TIM analysis of DCM phase for a 2.8 mass unit window around m/z = 1608. (B) The MS2 spectrum of the same mass window for TIM analysis of the water phase. In the DCM-phase, the major fragment ions correspond to loss of Hex1-O (m/z 1370.8), Fuc1Hex3HexNAc2+Na (1331.8), loss of Fuc1Hex1-O (1196.7), loss of Hex1HexNAc1 (1143.6), loss of Fuc1Hex1HexNAc1 (969.5), loss of Fuc1Hex2HexNAc1-O (747.4), Fuc1Hex1HexNAc1+Na (660.4), and Hex1HexNAc1+Na (486.3). Based on these fragment ions, a mixture of non-sulfated structures with a composition of Fuc1Hex3HexNAc2GalNAc-ol is proposed as shown to the right of the spectrum. In contrast, the major fragment ions for the water phase are loss of SO3Na (1486.8), loss of NeuAc (1231.5), loss of Fuc1Hex1-O (1196.7), loss of combination of NeuAc and SO3Na (1111.8), and loss of NeuAc1Hex1-O (1009.4). A mixture of sulfated structures with a composition of (SO3–)1NeuAc1Fuc1Hex2HexNAc1GalNAc-ol is proposed. Without physical separation of the neutral and sulfoglycans by phase partition, interpreting MS2 spectra of such mixtures is significantly more challenging. Modified from Kumagai25. Please click here to view a larger version of the figure.
Combining in-gel reductive β-elimination with aqueous-organic extraction enhances the sensitivity and depth of structural data that can be acquired for characterizing sulfated and non-sulfated O-linked glycans harvested from small amounts of mucin-type glycoproteins resolved from other proteins by SDS-PAGE. The essential advances of the technical approaches presented in this study are: (a) facile removal of gel derived contaminants by simple washing steps; (b) quantitative recovery of permethylated sulfoglycans in the aqueous phase following a rapid water-DCM partition. The workflow described here dramatically improves the sensitivity of detection for sulfated and non-sulfated glycans released from glycoproteins expressed at physiological levels in biological samples.
While the methods described here are robust and highly reproducible when applied with careful attention to standard analytic practices, a few considerations regarding possible sources of contamination are worth a brief mention. Ensuring that all glassware is exceptionally clean and free of dust and detergent is critical. Upon noticing contaminant peaks in MS spectra, all glassware should be washed thoroughly and rinsed with MeOH prior to final drying. Another source of contamination is the use of rubber-lined screw caps in place of PTFE-lined caps. The lack of organic chemical resistance characteristic of the rubber lining results in samples contaminated with an interfering polymeric compound. Although this protocol was specifically designed with the intention of utilizing MSn analysis by NSI-MS on an ion trap instrument, other types of mass spectrometers, such as MALDI-TOF/TOF can be also applied to dissecting sulfoglycoforms in biological materials using the preparative methods described here.
In comparison to other methodologies currently in use for in-gel glycan release and analysis, the techniques described here do not require HPLC separation prior to MS analysis and are applicable to permethylated glycans, producing greater structural characterization than can be extracted from non-derivatized samples. Moreover, the methods described here are capable of separating permethylated forms of sulfated O-glycans from non-sulfated neutral and sialylated O-glycans by a simple phase partition. The optimized sample clean-up and simple phase partition methodologies minimize sample manipulation and losses that can occur through chromatography steps. Methods requiring greater sample handling may lead to an underestimate of sulfoglycan diversity due to losses during work-up. Furthermore, detection of permethylated sulfated glycans in negative mode is significantly less sensitive than detection of permethylated glycans in positive mode. Desulfation of permethylated sulfoglycans by solvolysis specifically removes sulfate groups, leaving naked hydroxyls that can be subsequently permethylated with deuterated methyl iodide, increasing MS-signal intensities approximately 5-fold and providing a unique isotopic mass tag at the site of sulfation25.
The authors have nothing to disclose.
This work was supported by the grant P01HL107151 from the NHLBI/NIH. The authors also gratefully acknowledge the support and access to instrumentation provided through grant P41GM103490 from the NIGMS/NIH.
Name of Material/ Equipment | Company | Catalog Number | Comments/Description |
Water, deionized water | Sigma-Aldrich | 270733-4L | CHROMASOLV, for HPLC |
Acetic acid, Glacial | Fisher | A38-500 | Certified ACS≥99.7 w/w % |
Methanol | Sigma-Aldrich | 34860-4L-R | CHROMASOLV, for HPLC, ≥99.9% |
Ammonium bicarbonate | Fluka | 09830-500G | BioUltra, ≥99.5% (T), Step 1 |
Acetonitrile | Sigma-Aldrich | 34998-4L | CHROMASOLV Plus, for HPLC, ≥99.9%, Step 1 |
Ethyl acetate | Fluka | 34972-1L-R | LC-MS CHROMASOLV, Step 1 |
Sodium borohydride | Aldrich | 213462-25G | ReagentPlus, 99%, Step 2 |
Iodomethane | Sigma-Aldrich | 289566-100G | ReagentPlus, 99.5%, Step 6. Store at 4 °C until use. Sit at room temperature before use. |
Sodium hydroxide solution, 50% w/w | Fisher | SS254-1 | Certified, Step 6 |
Dimethyl sulfoxide, anhydrous | Sigma-Aldrich | 276855-1L | Anhydrous, ≥99.9%, Step 6 |
Methanol, anhydrous | Sigma-Aldrich | 322415-100ML | Anhydrous, 99.8%, Step 6 |
Dichloromethane | Sigma-Aldrich | 34856-4L | CHROMASOL®, for HPLC, ≥99.8%, contains amylene as stabilizer, Step 7 |
Dowex 50WX8 hydrogen formhydrogen form 100-200 mesh | Sigma-Aldrich | 217506-500G | Step 3 |
AG 50W-X8 Resin | Bio-Rad | 142-1441 | Step 3 |
BAKERBOND spe 1 ml x 100 mg Solid Phase Extraction Column, PP, Octadecyl (C18) Reverse Phase | JT Baker | JT-7020-01 | Step 5 and 8 |
7.5% Mini-PROTEAN TGX Precast Gel | Bio-Rad | Step 1 | |
Bio-Safe Coomassie Stain | Bio-Rad | 161-0786 | G-250, Step 1 |
Silver Stain Kit for Mass Spectrometry | Pierce | 24600 | Step 1 |
Oligosaccharides Kit (Maltotriose, Dp3: R474140) | Supelco | 47265 | |
Oligosaccharides Kit (Maltotetraose, Dp4: R474135) | Supelco | 47265 | |
Disposable Pasteur Pipets, Glass, Short Tip | VWR | 14673-010 | Wash before use |
PYREX 13 x 100 mm Disposable Round Bottom Threaded Culture Tubes | Corning | 99447-13 | Wash before use |
PYREX 16 x 125 mm Disposable Round Bottom Threaded Culture Tubes | Corning | 99447-16 | Wash before use |
Phenolic Caps/Closures with PTFE-Faced Rubber Liner | Kimble | 45066C-13 | Wash before use |
Phenolic Caps/Closures with PTFE-Faced Rubber Liner | Kimble | 45066C-15 | Wash before use |
Hamilton HPLC syringe | HAMILTON | 81265 | volume 500 μL, needle size 22 ga |
Hamilton HPLC syringe | HAMILTON | 81165 | volume 250 μL, needle size 22 ga |
Hamilton HPLC syringe | HAMILTON | 81065 | volume 100 μL, needle size 22s ga |
Hamilton HPLC syringe | HAMILTON | 80965 | volume 50 μL, needle size 22s ga |
Hamilton Calibrated Syringes | HAMILTON | 80300 | volume 10 μL, needle size 26s ga |
Petri Dish Glass 100mm x 15mm | GSC INTERNATIONAL INC | 1500-4 | Wash before use, Step 1 |
Bard-Parker Surgical Blades | Fisher | 371310 | Step 1 |
Reacti-Therm Heating/Stirring Module | Pierce | 18870 | Step 1, 4 and 7 |
Heating Blocks | Fisher | 125D | Step 2 |
Pyrex fiber glass wool borosilicate pore size 8 μm | Aldrich | CLS3950 | Step 3 |
Fused Silica CutterTubing cutter | alltech | 3194 | Step 3 |
Multi-tube vortexers | VWR | 444-7063 | Step 6 |
Lyophilizer 25EL Freezemobile | Virtis | 25EL | |
Centrifuge | VWR | Clinical 50 | |
LTQ Orbitrap Discovery | Thermo Fisher Scientific | 0 |