Comprehensive Profiling of Glycosphingolipid ... - ACS Publications

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES

Article

Comprehensive Profiling of Glycosphingolipid Glycans using a Novel Broad Specificity Endoglycoceramidase in a High-Throughput Workflow Simone Albrecht, Saulius Vainauskas, Henning Stoeckmann, Ciara McManus, Christopher H. Taron, and Pauline Mary Rudd Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b00259 • Publication Date (Web): 01 Apr 2016 Downloaded from http://pubs.acs.org on April 4, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Comprehensive Profiling of Glycosphingolipid Glycans using a Novel Broad Specificity Endoglycoceramidase in a High-Throughput Workflow

Simone Albrecht1, Saulius Vainauskas2, Henning Stöckmann1, Ciara McManus1, Christopher H. Taron2Δ, Pauline M. Rudd1Δ*

1NIBRT

GlycoScience Group, National Institute for Bioprocessing, Research and Training, Fosters Avenue, Mount Merrion, Blackrock, Dublin 4, Ireland

2New

ΔChristopher

England Biolabs, Ipswich, MA, USA

Taron and Pauline Rudd share senior authorship

*To whom correspondence should be addressed: Pauline M. Rudd, NIBRT GlycoScience Group, National Institute for Bioprocessing, Research and Training, Fosters Avenue, Mount Merrion, Blackrock, Dublin 4, Ireland. Tel.: +353 12158 142; Fax: +353 12158 116; E-mail: [email protected]

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract The biological function of glycosphingolipids (GSLs) is largely determined by their glycan head group moiety. This has placed a renewed emphasis on detailed GSL head group structural analysis. Comprehensive profiling of GSL head groups in biological samples requires the use of endoglycoceramidases with broad substrate specificity and a robust workflow that enables their high-throughput analysis. We present here the first high-throughput glyco-analytical platform for GSL head group profiling. The workflow features enzymatic release of GSL glycans with a novel broad-specificity endoglycoceramidase I (EGCase I) from Rhodococcus triatomea, selective glycan capture on hydrazide beads on a robotics platform, 2AB-fluorescent glycan labelling and analysis by UPLC-HILIC-FLD. R. triatomea EGCase I displayed a wider specificity than known EGCases and was able to efficiently hydrolyze gangliosides, globosides, (n)Lc-type GSLs and cerebrosides. Our workflow was validated on purified GSL standard lipids and was applied to the characterization of GSLs extracted from several mammalian cell lines and human serum. This study should facilitate the analytical workflow in functional glycomics studies and biomarker discovery.

Keywords:, ultra-performance hydrophilic interaction liquid chromatography (UPLC-HILIC), endoglycoceramidase, glycan profiling, glycosphingolipid, glycomics, high-throughput

Abbreviations: GSL, glycosphingolipid; MODY, maturity-onset diabetes of the young; rEGCase, recombinant

endoglycoceramidase; EGALC, endogalactosylceramidase; Gal, galactose; Cer,

ceramide; 2AB, 2-aminobenzamide; UPLC-HILIC-FLD, ultra performance liquid chromatographyhydrophilic interaction-fluorescence detection; IPTG, isopropyl-β-thiogalactopyranoside; ACN, acetonitrile; MeOH, methanol; GU, glucose units; QTOF, quadruploe time-of-flight; WAX, weak anion exchange; ABS, A.ureafaciens α(2-3/6/8)-sialidase; NAN1, S. pneumoniae α(2-3)-sialidase; BKF, bovine kidney α(1-2/4)-fucosidase; AMF, almond meal α(1-3/4)-fucosidase; BTG, bovine testes β(1-3/4)-galactosidase; SPG, S. pneumoniae β(1-4)-galactosidase; CBG, coffee bean α(1-3/4)galactosidase; JBH, jack bean β(1-2/3/4/6)-N-acetylhexosaminidase; CV, coefficient of variance; SSEA, stage specific embryonic antigen; HLB, hydrophilic-lipophilic balance; Neu5Ac/S, Nacetylneuraminic acid; Neu5Gc/Sg, N-glycolylneuraminic acid; Fuc/F, fucose; Hex, hexose; HexNAc, N-acetylhexosamine; GlcNAc, N-acetylglucosylamine.


Page 2 of 22

Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Many mammalian secretory proteins and lipids contain covalently linked carbohydrates (glycans). For these molecules, the structure and composition of their appended glycans plays a significant role in their function, distribution and physical properties1. As glycan biosynthesis is not directly a template-driven process, glycan structures on secreted glycoproteins and cell surface lipids are typically heterogeneous and complex. Additionally, glycans are subject to structural alterations over their lifetime due to both normal and pathological physiological processes, a feature that makes them attractive as potential biomarkers of disease. In the past few years, there have been major advances in high-throughput glycomics technologies for glycoprotein analysis. A low-cost, high-throughput, automated N-glycan sample preparation platform for glycoprofiling of immunoglobulins (IgG), antibodies and glycoproteins isolated from serum was recently described by our laboratory2-5. To date the analysis of N-glycans has largely been based on chromatographic profiling using ultra-performance hydrophilic interaction liquid chromatography with fluorescence detection (UPLC-HILIC-FLD). This technique permits rapid and semi-quantitative comparison of N-glycan structures across many samples. Serum N-glycan profiling has been used to identify glycan biomarkers of various diseases such as cancer (ovarian, prostate, breast, lung, pancreatic, stomach), mature onset diabetes of the young (MODY), as well as biomarkers associated with normal physiological processes like aging6-8. While this young field has made significant advances, it is desirable to expand the breadth of these profiling studies to other families of glycoproteins and other glycoconjugates, for example, glycolipids. One attractive class of molecule for glycoprofiling studies are glycosphingolipids (GSLs), lipids that possess a carbohydrate head group consisting of mono- or oligosaccharides attached to the lipids sphingosine or ceramide. More than 500 structural species that differ in their head group glycan or fatty acid composition are known9. GSL glycan head groups are associated with many cellular processes such as cell differentiation, signalling and receptor functions for viruses, antibodies or lectins10. GSLs are ubiquitous on cell membranes and also circulate in serum where they are present in a free form or in complex with proteins11. Aberrant GSL glycosylation has been repeatedly reported for different types of cancer including lung, breast, prostate and ovarian cancer as well as brain tumors, multiple sclerosis, rheumatoid arthritis and lysosomal storage diseases such as Gaucher’s and Fabry disease12-15. In contrast to N-glycan profiling, high-throughput analysis of GSLs is still in its infancy. Although methods have been described for the automated extraction of GSL from their matrix and their subsequent characterization en masse using shotgun lipidomics16 no such automated methods



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

are currently available for the preparation and chromatographic analysis of GSL glycans. Enzymatically released GSL head groups are commonly analyzed using MALDI-TOF MS17,18. Enzymes often play a critical role in glycomics workflows such as the enzymatic release of GSL glycans by endoglycoceramidases (EGCases) which is preferred over harsh chemical release methods that can result in low glycan yields19,20. However, for enzymatic release of GSL head groups to be viable in a glycan profiling scheme, an enzyme with broad substrate specificity is needed. To date very few EGCases have been tested for their ability to cleave a wide range of GSL classes, limiting their utility for GSL head group profiling. In the present study, we adapted our existing robotized N-glycan analysis platform2,5 for the quantitative high-throughput profiling of 2AB-labeled mammalian GSL head groups using UPLCHILIC-FLD. An enabling component of our GSL glycan workflow is our identification and characterization of a novel recombinant EGCase I enzyme from Rhodocococcus triatomea that exhibits a broad GSL specificity, including the release of globo-series GSLs and Gal(β1-1)Cer, important classes of GSLs that are not efficiently released by known EGCases. Finally, we demonstrate the ability to characterize GSL head groups from both mammalian cell surfaces and small volumes of blood serum. This analytical workflow will permit further exploration of the GSL head group repertoire of GSLs from a broad range of biological sources and will enable studies aiming to identify cellular or serum GSL-glycan biomarkers of disease.

Experimental Section Chemical reagents and solvents were from Sigma-Aldrich. Glycosphingolipid standards were from Sigma-Aldrich (GD1a, GD1b, GM1a, GT1b, GalCer, Sulfatide, Psychosine), Avanti (GD3, GlcCer) and Matreya (FGM1, GM3, Gb4, LacCer). Human serum was pooled from apparently healthy donors (courtesy

of

the

U.K.

Blood

Transfusion

Service).

Recombinant

Rhodococcus

equi

endoglycoceramidase (rEGCase II) was from Takara Bio Inc.

Cell culture HeLa (ATCC # CCL-2), NIH/3T3 (ATCC # CRL-1658), and HL60 (ATCC # CCL-240) cell lines were obtained from the American Type Culture Collection (ATCC). HeLa and NIH/3T3 cells were cultured in DMEM (Thermo Scientific HyClone) containing 10% (v/v) fetal bovine serum (FBS), 2 mM Lglutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin for 2 days at 37°C, then trypsinized and collected by centrifugation. Cells were washed with cold 1X PBS and frozen at –80°C. HL60 cells were cultured in Iscove's Modified Dulbecco's Medium (Thermo Scientific HyClone) containing


Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


10% (v/v) FBS, 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin for 3 days at 37°C in suspension. Collected cells were washed with cold 1X PBS buffer and frozen at –80°C.

Cloning, expression and purification of Rhodococcus triatomea EGCase I A gene encoding putative EGCase I (GenBank accession: EME18930.1) was identified in the R. triatomea genome database21 by BLASTP and TBLASTN using the amino acid sequences of EGCase I (R. equi), EGCase II (R. equi) or EGALC (R. equi) as queries. A codon optimized sequence encoding the R. triatomea EGCase I ORF with a C-terminal His-tag was synthesized by Genscript. A version of R. triatomea EGCase I (27-489 aa) lacking its signal peptide and having a C-terminal His-tag was created by PCR and cloned into the NdeI and XbaI sites of pJS11922 to produce pJS119/RhtrECI. Recombinant protein expression was performed in E. coli NEB Express cells (New England Biolabs) carrying pJS119/RhtrECI followed by purity and activity assay of the enzyme (see the Experimental Section in Supporting Information).

Extraction and purification of GSLs from cell membranes A method adapted from Smith et al.23 was used for extraction of GSLs from cell membranes. Lipids were extracted from aliquots of 2 x 105 cells. A mix of chloroform/methanol/H2O (1:2:0.75; v/v/v) was added, sonicated for 10 min and centrifuged (14,000 x g, 2 min). The supernatant was collected and the step was repeated once. Mixtures of chloroform/methanol (1:1; v/v) and chloroform/methanol (2:1; v/v) were sequentially added and sonicated for 10 min, after which samples were centrifuged and the supernatants collected and pooled after each step. Pooled supernatants were dried using a SpeedVac (Thermo Scientific). The dried extract was subjected to n-butanol/water partitioning according to Vidugiriene et al.24. The crude lipid extract was resuspended in n-butanol/water (1:1; v/v), vortexed and centrifuged at 1,000 x g for 2 min. The organic phase (upper) was collected and back-extracted with n-butanol-saturated H2O and the aqueous phase (lower) was re-extracted with H20-saturated n-butanol followed by vortexing and centrifugation at 1000 x g for 2 min. The pooled butanol-phases were dried in a SpeedVac.

Extraction and purification of GSLs from serum GSLs from serum were extracted using a quick protein precipitation method adapted from Huang et al

25.

Briefly, 180 μL of methanol were added to aliquots of 20 μL human serum followed by

vortexing and centrifugation at 14,000 x g for 30 min at 4°C. The supernatants were dried in a SpeedVac and subjected to n-butanol partitioning as described above.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Endoglycoceramidase digestion Dried GSLs, whether commercial standards or extracted and purified from cells or serum, were resuspended in sodium actete buffer (pH 5.2) containing 1 mg/mL Triton-X-100 and incubated at 37°C for 16 h using different concentrations (2 mU – 120 mU) of R. triatomea rEGCase I or R. equi 2 mU of rEGCase II in a total volume of 20-200 μL.

Automated hydrazide-mediated glycan head group cleanup Enzymatically released glycan head groups were captured and cleaned using a Hamilton Robotics StarLet liquid-handling platform in an automated method adapted from Stöckmann et al.2. (see the Experimental Section in Supporting Information).

2AB labeling and cleanup of excess label Fluorescent labeling mix (5 μL; 350 mM 2-aminobenzamide (2AB), 1 M sodium cyanoborohydride in acetic acid/dimethyl sulfoxide (30:70)) was dispensed into each sample plate well, and the plate was incubated at 65°C with agitation at 700 rpm for 120 min. Excess label was removed by paper chromatography in a 96-well plate according to Royle et al.26.

UPLC-HILIC-FLD and UPLC-HILIC-FLD-MS 2AB-labeled glycans were analyzed by UPLC-HILIC-FLD/MS using a method adapted from Albrecht et al.27 (see the Experimental Section in Supporting Information)

Safety Considerations Chloroform is toxic and needs to be handled only by experienced and well-trained personal using all safety laboratory measures possible (gloves, glasses, and fume hood).

Results and Discussion

Semi-automated GSL head group sample preparation and analysis A semi-automated sample preparation workflow was established for GSL glycan head group analysis (Figure 1). Our approach was inspired by a robust IgG N-glycan preparation workflow that was recently reported by our laboratory2. The adapted workflow for GSL head group analysis utilizes endoglycoceramidase release of GSL head group glycans followed by selective glycan capture on solid-supported hydrazide beads, high-throughput 2AB glycan labeling and cleanup by


Page 6 of 22

Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


HyperSep Diol SPE. Simultaneous preparation of 96 samples can be achieved using a liquidhandling robotic workstation ahead of GSL head group glycan analysis by UPLC-HILIC-FLD or UPLC-HILIC-FLD-MS. The amphipathic nature of GSLs requires the use of an enzyme reaction buffer containing detergent to permit the optimal release of head group glycans by endoglycoceramidase digestion. This detergent needed to be removed prior to subsequent sample processing steps and glycan analysis. Prior studies reported the use of “glycoblotting”28, a method where released glycans are transiently bound to hydrazide beads, to isolate GSL head groups after endoglycoceramidase digestion17,18. Therefore, we adapted an automated high-recovery version of glycoblotting used in our IgG Nglycan preparation workflow2 for use in our GSL glycan profiling workflow. The glycan capturing step was performed using low-cost UltraLink hydrazide resin and was conducted on a temperaturecontrolled robotic heater-shaker with vigorous agitation. Additional considerations were made for labeling GSL head group glycans with a fluorescent dye. We used the dye 2-aminobenzamide (2-AB) which is widely used for mole-based glycan derivatization allowing for sensitive and quantitative analysis of reducing glycans by HILIC-FLD chromatography. Although 2-AB labeling is broadly applicable to any reducing glycan species, different approaches are required for the removal of excess labeling dye that largely depend on the nature of the glycans being labeled. Compared to N-glycans that have a common pentasaccharide core (i.e. the chitobiose core extended by three mannoses), GSL head groups are structurally more diverse and range from monosaccharides to large sialylated oligosaccharides. 2-AB cleanup using normal-phase SPE, which is routinely used for N-glycans, resulted in insufficient recovery of small glycan structures. Therefore a more universal and gentle cleanup method based on paper chromatography was used instead. This method was previously adapted for processing in a 96-well plate format to ensure efficient sample-throughput26.

High-throughput characterization of ceramidases on defined GSL substrates Using our GSL glycan analysis workflow the substrate specificity of a novel recombinant R. triatomea EGCase I (see the Results Section and Figure S-1 in the Supporting Information) was evaluated on isolated GSL standards and experimentally compared to that of a recombinant commercial EGCase II from R. equi (Table 1). A theoretical comparison with literature findings for the specificity of the more closely related R.equi rEGCase I was included9. Aliquots of 2 nmol standard GSL substrates including cerebrosides (glucosylceramide [GlcCer], galactosylceramide [GalCer]), lactosylceramide [LacCer], gangliosides (GM3, GM1a, FGM1, GD1a, GD1b, GD3 and GT1b), globosides (Gb4), sulfatide and psychosine were incubated with either 2 mU EGCase I or EGCase II



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

at 37°C for 16 h in 50 mM sodium acetate buffer containing 0.1% Triton-X. The reaction was performed directly on a 96-well sample plate that was subsequently used for automated cleanup, minimizing the number of sample transfer steps. An internal standard (ISTD) was added for relative UPLC-HILIC-FLD glycan quantitation and multiple replicates (n = 2-4) were performed. Chromatographic peak retention times of the GSL head groups were converted into glucose unit (GU) values using a 2AB-labeled dextran ladder that enables the independent comparison of chromatographic profiles26. Completeness of digestion was tested by complementary TLC assay (see the Experimental Section in Supporting Information and Table S-2). The coefficients of variance (CV) for the relative GSL head group quantitation after endoglycoceramidase release and 2AB-labeling were ≤ 10% for each GSL standard/enzyme pair, highlighting the reproducibility of the workflow as well as the defined substrate-specificity of the endoglycoceramidases tested. Overall, R. triatomea rEGCase I showed very broad and robust GSL hydrolyzing activity. The release efficiency of rEGCase I was clearly superior to rEGCase II for the fucosylated ganglioside FGM1 (1:0.4; rEGCase I : rEGCase II), the tri-sialylated ganglioside GT1b (1 : 0.3; rEGCase I : rEGCase II), the globoside Gb4 (1 : 0.1; rEGCase I : rEGCase II) and cerebrosides. The substrate specificity of our R.triatomea rEGCase I was comparable to literature findings for R.equi rEGCase I except for Gb4, LacCer and cerebrosides9. It is noteworthy that certain globosides are not efficiently released by known EGCases 9,29. Thus, the efficient release of the head group from Gb4 by R. triatomea rEGCase I was compelling and suggested that this enzyme might be useful in the profiling of globosides, which comprise many antigens (e.g. P-, Pk-, Forssman-, SSEA-3- and SSEA-4antigen) in biological samples. Furthermore, R. triatomea rEGCase I efficiently hydrolyzed GlcCer and to a lesser extent GalCer further indicating specificity differences from other Rhodococcal EGCases that show little or no hydrolysis of cerebrosides9,29,30. The GSL psychosine that has galactose linked to a sphingosine backbone (instead of ceramide) was not released by rEGCase I. Psychosine is a component of brain GSLs. For example, it is of importance in Krabbe disease where it is thought to interfere with protein kinase C and accumulates in microdomains of the brain which might result in disruption of lipid raft architecture31. For samples of brain origin additional digestion with rEGCase II (R. equi) or EGALC (R.equi) that hydrolyze psychosine to about 14% 29 is recommended. Finally, none of the known EGCases has shown the ability to hydrolyze sulfatide30, a sulfated form of galactocerebroside, that is a highly enriched component in the central and peripheral nervous system and plays an important role in the biology of myelin-forming cells.32 Application of the workflow for profiling cellular GSL head groups The suitability of our workflow for profiling cellular GSL glycans was assessed using mammalian cell lines. In this experiment, R. triatomea rEGCase I was used to liberate glycan head groups from


Page 8 of 22

Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


GSLs extracted from mammalian NIH/3T3, HeLa and HL60 cell lines. In our study, only a small number of samples needed to be processed, so GSL extraction was performed manually. However, for larger numbers of samples, a robotic method for the automated extraction of GSLs as reported by Stahlman et al16 could be easily combined with our platform to permit the automated extraction of GSLs from a variety of biological matrices including tissues, cells and biofluids Well-resolved chromatographic profiles were obtained for aliquots of 2 x 105 cells after extraction of GSLs according to Smith23, separation of polar impurities by n-butanol partitioning24, on-plate treatment with 40 mU R. triatomea rEGCase I, followed by high-throughput sample preparation, inclusion of an ISTD for relative peak quantification and analysis by UPLC-HILIC-FLD (Figure 2, Table 2). The enzyme concentration (40 mU) was empirically chosen and was largely based on a prior report of optimal cellular GSL release using 25-50 mU R. equi EGCase I17. Peak assignments were confirmed by exoglycosidase treatment that resulted in characteristic shifts in GU values depending on the presence of terminal sugars (data not shown)33. High quality GSL glycan profiles were obtained for each cell line (Figure 2). Furthermore, the glycans observed for each profile matched GSL class biases previously reported for NIH/3T3, HeLa and HL60 cells that are rich in gangliosides, globosides and nLc-type GSLs, respectively17. These data also further illustrate the ability of R. triatomea rEGCase I to hydrolyze globosides (see Hela cell profile). Similarly, a broad range of nLc-type GSLs that included a N-glycolylneuraminic acid (Neu5Gc)-containing species were released from HL60 extracts. Members of this class of GSLs are not commercially available as standards and could not be evaluated in our experiments testing EGCase release from defined GSL substrates. However, these data illustrate that rEGCase I is able to liberate glycans from these lipids. Considered together, these data illustrate that our highthroughput workflow can be used for efficient profiling of cellular GSL glycans.

Application of the workflow for profiling GSLs from human serum Serum glycan profiling has become a useful tool for the monitoring of diseases and biomarker discovery. We have recently introduced an automated IgG glycoprofiling platform2 for the highthroughput serum profiling of N-glycans derived from glycoproteins5. Here we aimed to further adapt the platform for the profiling of GSL glycans in human serum. In contrast to cells or tissue, serum GSLs are not bound to membranes but are instead present in their free form or embedded in protein-lipid complexes11. As such, we used a simplified extraction method that included protein precipitation by methanol and centrifugation at low temperature (Figure 1) as adapted from Huang et al25. A small volume (as low as 20 μL) of serum was sufficient to obtain a high-quality fluorescence profiles within a total chromatographic run time of 40 minutes



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 22

(Figure 3). To determine the optimum enzyme concentration for GSL head group release, serum GSLs were incubated with increasing concentrations of EGCase I (6 mU, 30 mU, 60 mU, 120 mU) (Figure S-2). The best glycan yields were obtained for 60 mU EGCase I.

Our data show that human serum contains a complex mixture of GSL head groups that includes gangliosides, globosides, nLc- and fucosylated Lc-type GSLs (Figure 3). Serum also contains an abundance of free monosaccharides, however monosaccharides derived from GSLs are relatively low in healthy human serum34. Due to their excessive presence, monosaccharides were disregarded in our study although their concentration was considerably reduced by butanol-partitioning during sample

preparation

(Figure

1

and

Supporting

Information

Figure

S-3).

GSL-derived

monosaccharides are of special interest in certain diseases such as Gaucher’s disease for which increased glucosylcerebroside levels are observed34. To analyze GSL monosaccharides using our workflow, additional purification, for example, using glucose oxidation18 or hydrophilic-lipophilic balance (HLB, Oasis®) solid phase extraction35 would be needed. The integration of HLB for our GSL-serum application is shown in Supporting Information Figure S-3) and the use of commercially available 96-well HLB plates would permit automation of the cleanup step for high-throughput applications.

Our structural assignments of serum-derived GSL glycans were based on exoglycosidase glycan sequencing and MS, that were each performed after weak anion exchange (WAX) fractionation of serum GSL head groups into pools of neutral and charged structures (see Supporting Information for supporting data in Tables S-3 and S-4, and Figures S-4-10). Eighteen GSL head group structures with a relative abundance of ≥ 0.1% were identified in human serum in this study and quantified relative to an internal standard. Additional structures of minor relative abundance (< 0.1%) may be present but were not further considered. Prior studies on human serum reported high variability in total serum GSL concentration but low inter-individual variability in GSL composition and proportioning11,18. Thus, with a view toward discriminating between healthy and diseased serum GSL glycan profiles, our findings support the notion that relative glycan quantification using UPLCHILIC-FLD is sufficient over absolute glycan quantification13.

We observed GM3 and LacCer as the predominant serum GSL species (34% each), followed by Gb3 (14%), Gb4 (7%), nLc4 (3%), (α2-3)-sialylated nLc4 (S(3)-nLc4, 2% ), (α1-2/α1-4)-di-fucosylated Lc4 (diF(2,4)-Lc4,1.5%), GM2 (1%) and several structures of minor abundance (< 1%) including GA2, GM1a, GD1a, GD3, nLc6, S(3)-nLc4, S(6)-nLc6, F(2)-Lc4, F(4)-Lc4 and Hex2HexNAc2 (Table 3).


Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Several other studies have also sought to identify GSL glycan structures present in serum using classical methods like high performance thin layer chromatography (HPTLC). However, due to limits in separation capacity, the methods have mainly been used to analyze only selected ganglioside or globoside structures11,13,15,36,37. A first approach to serum GSL head group profiling using HILIC-HPLC after 2AB-labeling of enzymatically released glycans was performed by Wing in 200135. However, GSL head group profiling by HPLC is time-consuming (~180 minutes per run) and the separate analysis of neutral and charged structures was required due to the GSL extraction and glycan purification methods used. Furthermore, the ceramide glycanase (from Macrobdella decora) used in that study showed poor hydrolysis of neutral GSLs. Finally, Furukawa et al. recently reported on the quantitative glycosphingolipid-glycome analysis in the serum of 10 healthy human subjects using MALDI-TOF MS18. Of the 42 MALDI-TOF MS signals related to potential GSL glycan structures an average of 19-20 signals had an abundance of ≥ 0.1% of the total glycan pool.18 Similar cohorts and quantities of serum-GSL glycan structures were detected when compared to our study (Table 3). Although MALDI-TOF MS allows for the detection of additional trace signals from GSL-glycans, quantification by MALDI-TOF MS is difficult and correction factors may have to be introduced. Additionally, signals from components such as media, N-glycans or peptides can also complicate glycan identification and contrary to chromatographic methods it is not possible to separate glycoforms. Thus, the described UPLC-HILIC-FLD method offers an important alternative approach for profiling serum GSLs that increases speed and throughput without compromising detection and determination of relative abundance of a large repertoire of GSL glycan classes.

Conclusion In this report, we present the first high-throughputworkflow for the profiling of GSL glycan head groups and their subsequent analysis by UPLC-HILIC-FLD using a robotic platform including selective glycan capture and release on hydrazide beads in solution. The workflow features a novel R. triatomea rEGCase I that shows a broad specificity for globosides, fucosylated GSLs and cerebrosides compared to other EGCases. The workflow was successfully applied to perform systematic GSL head group profiling of human serum and will enable future profiling of GSL head groups in clinical disease research.

Conflict of interest The authors declare that they have no conflicts of interest with the contents of this article.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

Author contributions SA and SV designed, performed and analyzed the experiments of this study. HS provided technical assistance. SA, SV and CHT wrote the paper. CMM, CHT and PMR coordinated the study. All authors reviewed the results and approved the final version of the manuscript.

Supporting Information This document file contains SupportingExperimental Section, Supporting Results, Supplementary Figures S1-S9 and Tables S1-S4 and Supporting References. References

(1) Marino, K.; Bones, J.; Kattla, J. J.; Rudd, P. M. Nat. Chem Biol. 2010, 6, 713-723. (2) Stockmann, H.; Adamczyk, B.; Hayes, J.; Rudd, P. M. Anal. Chem. 2013, 85, 8841-8849. (3) AgilentTechnologies. http://www.chem.agilent.com/Library/flyers/Public/59911140EN.pdf 2012. (4) Reusch, D.; Haberger, M.; Selman, M. H.; Bulau, P.; Deelder, A. M.; Wuhrer, M.; Engler, N. Anal. Biochem. 2013, 432, 82-89. (5) Stockmann, H.; O'Flaherty, R.; Adamczyk, B.; Saldova, R.; Rudd, P. M. Integrative biology : quantitative biosciences from nano to macro 2015, 7, 1026-1032. (6) Adamczyk, B.; Tharmalingam, T.; Rudd, P. M. Biochim. Biophys. Acta 2012, 1820, 13471353. (7) Knezevic, A.; Gornik, O.; Polasek, O.; Pucic, M.; Redzic, I.; Novokmet, M.; Rudd, P. M.; Wright, A. F.; Campbell, H.; Rudan, I.; Lauc, G. Glycobiology 2010, 20, 959-969. (8) Thanabalasingham, G.; Huffman, J. E.; Kattla, J. J.; Novokmet, M.; Rudan, I.; Gloyn, A. L.; Hayward, C.; Adamczyk, B.; Reynolds, R. M.; Muzinic, A.; Hassanali, N.; Pucic, M.; Bennett, A. J.; Essafi, A.; Polasek, O.; Mughal, S. A.; Redzic, I.; Primorac, D.; Zgaga, L.; Kolcic, I.; Hansen, T.; Gasperikova, D.; Tjora, E.; Strachan, M. W. J.; Nielsen, T.; Stanik, J.; Klimes, I.; Pedersen, O. B.; Njølstad, P. R.; Wild, S. H.; Gyllensten, U.; Gornik, O.; Wilson, J. F.; Hastie, N. D.; Campbell, H.; McCarthy, M. I.; Rudd, P. M.; Owen, K. R.; Lauc, G.; Wright, A. F. Diabetes 2013, 62, 13291337. (9) Ishibashi, Y.; Kobayashi, U.; Hijikata, A.; Sakaguchi, K.; Goda, H. M.; Tamura, T.; Okino, N.; Ito, M. J. Lipid Res. 2012, 53, 2242-2251. (10) Schnaar RL; Suzuki A; P., S. In Essentials of Glycobiology, Varki A; Cummings RD; JD, E., Eds.; Cold Spring Harbor Laboratory Press New York Chapter 10, 2009. (11) Senn, H.-J.; Orth, M.; Fitzke, E.; Wieland, H.; Gerok, W. Eur. J. Biochem. 1989, 181, 657662. (12) Daniotti, J. L.; Vilcaes, A. A.; Torres Demichelis, V.; Ruggiero, F. M.; Rodriguez-Walker, M. Front Oncol. 2013, 3, 306. (13) Zaprianova, E.; Deleva, D.; Ilinov, P.; Sultanov, E.; Filchev, A.; Christova, L.; Sultanov, B. Neurochem. Res. 2001, 26, 95-100. (14) Tsukuda, Y.; Iwasaki, N.; Seito, N.; Kanayama, M.; Fujitani, N.; Shinohara, Y.; Kasahara, Y.; Onodera, T.; Suzuki, K.; Asano, T.; Minami, A.; Yamashita, T. PLoS One 2012, 7, e40136. (15) Ullman, M. D.; McCluer, R. H. J. Lipid Res. 1977, 18, 371-378. 12 ACS Paragon Plus Environment

Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(16) Ståhlman, M.; Ejsing, C. S.; Tarasov, K.; Perman, J.; Borén, J.; Ekroos, K. J. Chromatogr. B 2009, 877, 2664-2672. (17) Fujitani, N.; Takegawa, Y.; Ishibashi, Y.; Araki, K.; Furukawa, J.; Mitsutake, S.; Igarashi, Y.; Ito, M.; Shinohara, Y. J. Biol. Chem. 2011, 286, 41669-41679. (18) Furukawa, J.-i.; Sakai, S.; Yokota, I.; Okada, K.; Hanamatsu, H.; Kobayashi, T.; Yoshida, Y.; Higashino, K.; Tamura, T.; Igarashi, Y.; Shinohara, Y. J. Lipid Res. 2015. (19) Hakomori, S. I. J. Lipid Res. 1966, 7, 789-792. (20) Yowler, B. C.; Stoehr, S. A.; Schengrund, C. L. J. Lipid Res. 2001, 42, 659-662. (21) Kumar, S.; Bala, M.; Raghava, G. P.; Mayilraj, S. Genome Announc 2013, 1, e0012913. (22) Furste, J. P.; Pansegrau, W.; Frank, R.; Blocker, H.; Scholz, P.; Bagdasarian, M.; Lanka, E. Gene 1986, 48, 119-131. (23) Smith, D. F.; Prieto, P. A. In Current Protocols in Molecular Biology; John Wiley & Sons, Inc. : Cold Spring Harbor (NY) Unit 17.3, 2001. (24) Vidugiriene, J.; Menon, A. K. Methods Enzymol. 1995, 250, 513-535. (25) Huang, Q.; Zhou, X.; Liu, D.; Xin, B.; Cechner, K.; Wang, H.; Zhou, A. Anal. Biochem. 2014, 455, 26-34. (26) Royle, L.; Campbell, M. P.; Radcliffe, C. M.; White, D. M.; Harvey, D. J.; Abrahams, J. L.; Kim, Y. G.; Henry, G. W.; Shadick, N. A.; Weinblatt, M. E.; Lee, D. M.; Rudd, P. M.; Dwek, R. A. Anal. Biochem. 2008, 376, 1-12. (27) Albrecht, S.; Lane, J. A.; Marino, K.; Al Busadah, K. A.; Carrington, S. D.; Hickey, R. M.; Rudd, P. M. Br. J. Nutr 2014, 111, 1313-1328. (28) Miura, Y.; Hato, M.; Shinohara, Y.; Kuramoto, H.; Furukawa, J.; Kurogochi, M.; Shimaoka, H.; Tada, M.; Nakanishi, K.; Ozaki, M.; Todo, S.; Nishimura, S. Mol. Cell. Proteomics. 2008, 7, 370-377. (29) Ishibashi, Y.; Nakasone, T.; Kiyohara, M.; Horibata, Y.; Sakaguchi, K.; Hijikata, A.; Ichinose, S.; Omori, A.; Yasui, Y.; Imamura, A.; Ishida, H.; Kiso, M.; Okino, N.; Ito, M. J. Biol. Chem. 2007, 282, 11386-11396. (30) Ito, M.; Yamagata, T. J. Biol. Chem. 1989, 264, 9510-9519. (31) White, A. B.; Givogri, M. I.; Lopez-Rosas, A.; Cao, H.; van Breemen, R.; Thinakaran, G.; Bongarzone, E. R. J Neurosci. 2009, 29, 6068-6077. (32) Grassi, S.; Prioni, S.; Cabitta, L.; Aureli, M.; Sonnino, S.; Prinetti, A. Neurochem Res 2016, 41, 130-143. (33) Royle, L.; Radcliffe, C. M.; Dwek, R. A.; Rudd, P. M. Methods Mol Biol. 2006, 347, 125-143. (34) Muller, M. V. G.; Petry, A.; Vianna, L. P.; Breier, A. C.; Michelin-Tirelli, K.; Pires, R. F.; Trindade, V. M. T.; Coelho, J. C. Braz. J. Pharm. Sci 2010, 46, 643-649. (35) Wing, D. R.; Garner, B.; Hunnam, V.; Reinkensmeier, G.; Andersson, U.; Harvey, D. J.; Dwek, R. A.; Platt, F. M.; Butters, T. D. Anal. Biochem. 2001, 298, 207-217. (36) Gornati, R.; Bembi, B.; Tong, X.; Boscolo, R.; Bruno, B. Clin. Chim. Acta 1998, 271, 151161. (37) Kundu, S. K.; Diego, I.; Osovitz, S.; Marcus, D. M. Arch Biochem Biophys 1985, 238, 388400.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Page 14 of 22

Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Tables Table 1. Specificity of recombinant endoglycoceramidases using purified GSL substrates. Comparison of the relative abundance of glycans released from glycolipid standards using 2 mU R. triatomea rEGCase I or R. equi rEGCase II as assessed by UPLC-HILIC-FLD and comparison to literature results for R.equi rEGCase I9. For a relative comparison of the release efficiency of the two experimentally tested EGCases, the highest abundance of each glycolipid was normalized to “1”. Unless otherwise stated (*) this abundance also represented 100% head group release by a parallel TLC analysis (see Supplementary Table S-2). The coefficient of variance (CV) was < 10% for all substrate/enzyme pairs tested.

Substrate Gangliosides GM3

average rel. glycan quantities (n = 2-4) R.triatomea R.equi rEGCase I rEGCase II GU

glycolipid structure

Neu5Ac(α2-3)Gal(β1-4)Glc(β1-1)’Cer

1.0

1.0

3.15

Gal(β1-3)GalNAc(β1-4)[Neu5Ac(α2-3)] Gal(β1-4)Glc(β1-1)’Cer Fuc(α1-2)Gal(β1-3)GalNAc(β1-4) [Neu5Ac(α2-3)]Gal(β1-4)Glc(β1-1)’Cer Neu5Ac(α2-8)Neu5Ac(α2-3)Gal(β1-4) Glc(β1-1)’Cer Neu5Ac(α2-3)Gal(β1-3)GalNAc(β1-4) [Neu5Ac(α2-3)]Gal(β1-4)Glc(β1-1)’Cer Gal(β1-3)GalNAc(β1-4)[Neu5Ac(α2-8) Neu5Ac(α2-3)]Gal(β1-4)Glc(β1-1)’Cer Neu5Ac(α2-3)Gal(β1-3)GalNAc(β1-4) [Neu5Ac(α2-8)Neu5Ac(α2-3)]Gal(β14)Glc(β1-1)’Cer

1.0

0.9

4.39

1.0

0.4

4.69

1.0

1.0

4.62

Hydrolysis % R.equi rEGCase I9

100 GM1a FGM1 GD3 GD1a GD1b GT1b

Globosides Gb4

100 100 100 1.0

0.8

5.44

1.0

0.8

6.03

1.0

0.3

7.02

100 100 n.d.

GalNAc(β1-3)Gal(α1-4)Gal(β1-4)Glc(β11)’Cer

1.0

0.1

3.41

Cerebrosides GlcCer

Glc(β1-1)’Cer

1.0

0.2

0.90

GalCer*

Gal(β1-1)’Cer

0.05 - 0.1*

0

0.90

Sulfatide Psychosine

HSO3-3Gal(β1-1)’Cer Gal(β1-1)’sphingosine

0 0

0 1.0

--0.90

33.9 (100)+

6.7 (20.7)+ ‡ 0 n.d. LacCer

Gal(β1-4)Glc(β1-1)’Cer

1.0

0.9

2.00

*According to a TLC-based assay, the total release efficiency on GalCer was approx. 5-10% for R. triatomea rEGCase I.


29.9 (65.9)+


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

+GSLs (2nmol) were incubated at 37°C for 12h with 1mU (or 10mU) of recombinant R.equi EGCase I in 20µL of 50mM sodium acetate buffer, pH 5.5, containing 0.1% Triton X-100.9 ‡GSLs possessing the the β-galactosyl-Cer linkage (e.g. trigalactosylCer) were completely resistant to the hydrolysis by R.equi rEGCase I 9n.d. not determined

Table 2. Use of the workflow for profiling mammalian cellular GSLs. Shown are relative quantities of glycans released from GSLs extracted from murine NIH/3T3 and human HeLa and HL60 cell lines using 40 mU of R. triatomea rEGCase I in the high-throughput glycolipid workflow. The quantification was performed relative to an internal standard for which the area was set to “1”. rEGCase I (40 mU) NIH/3T3 HeLa HL60

GSL

GUav*

LacCer

1.96

1.6

6.8

18.8

GA2 GM3 GM2 GM1a GD1a total gangliosides

2.59 3.11 3.51 4.37 5.42

0.2 5.0 4.0 0.5 0.1 9.8

-3.3 5.5 --8.8

-0.6 ---0.6

Gb3 Gb4 total globosides

2.71 3.35

----

35.7 6.4 42.1

----

nLc3 nLc4 S(3)-nLc4 Sg(3)-nLc4 nLc6 S(3)-nLc6 total nLc

2.72 3.55 4.60 4.97 5.15 6.06

-------

0.5 0.8 0.5 ---1.8

1.1 3.2 0.5 0.4 0.1 0.2 5.5

TOTAL 11.4 59.5 *GUav: average Glucose Units. S: Neu5Ac. Sg: Neu5Gc.

24.9


Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Table 3: Relative abundance of GSL-glycans in human serum released by R. triatomea rEGCase I (60 mU) and analysed by UPLC-HILIC-FLD (Figure 3), compared to their relative abundances as determined by MALDI-TOF MS18. GU and relative abundance are average (av) values from four independent measurements (average CV ≤ 10%). Structural assignments were confirmed by exoglycosidase sequencing and mass spectrometry.

GUav

GSL

GSL head group

GSL class

relative abundanceav (%)

1.95

LacCer

Gal(β1-4)Glc

--

34.2

MALDI*18 relative abundanceav (%) 18.9a

2.60

GA2

GalNAc(β1-4)Gal(β1-4)Glc

ganglioside

0.2

0.1

2.73

Gb3

Gal(α1-4)Gal(β1-4)Glc

globoside

13.8

14.7

3.13

GM3

Neu5Ac(α2-3)Gal(β1-4)Glc

ganglioside

33.8

50.3a

3.41

Gb4

globoside

7.1

7.3b

3.55

GM2

ganglioside

1.0

0.2

3.60

nLc4

neoLc

3.3

7.3b

4.05

F(2)-Lc4

Lc

0.3

0.6c

4.17

Hex2HexNAc2

GalNAc(β1-3)Gal(α1-4)Gal(β1-4)Glc GalNAc(β1-4)[Neu5Ac(α2-3)]Gal(β14)Glc Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc Fuc(α1-2)Gal(β1-3)GlcNAc(β13)Gal(β1-4)Glc nd

Lc

0.2

< 0.1

4.38

F(4)-Lc4

0.4

0.6c

4.42

GM1a

0.6

3.1d

4.53

GD3

0.1

< 0.1

4.63

S(3)-nLc4

2.4

3.1d

4.99

S(6)-nLc4

0.8

3.1d

5.12

diF(2,4)-Lc4

1.5

0.7

5.16

nLc6

0.1

nd

5.44

Gd1a

0.1

0.3

6.09

S(3)-nLc6

0.2

0.2e

Gal(β1-3)[Fuc(α1-4)]GlcNAc(β1-3)Gal Lc Gal(β1-3)GalNAc(β1-4)[NeuAc(α2ganglioside 3)]Gal(β1-4)Glc Neu5Ac(α2-8)NeuAc(α2-3)Gal(β1ganglioside 4)Glc Neu5Ac(α2-3)Gal(β1-4)GlcNAc(β1nLc 3)Gal(β1-4)Glc Neu5Ac(α2-6)Gal(β1-4)GlcNAc(β1nLc 3)Gal(β1-4)Glc Fuc(α1-2)Gal(β1-3)[Fuc(α1Lc 4)]GlcNAc(β1-3)Gal(β1-4)Glc Gal(β1-4)GlcNAc(β1-3)Gal(β1nLc 4)GlcNAc(β1-3)Gal(β1-4)Glc NeuAc(α2-3)Gal(β1-3)GalNAc(β1-4) ganglioside [NeuAc(α2-3)]Gal(β1-4)Glc Neu5Ac(α2-3)Gal(β1-4)GlcNAc(β1nLc 3)Gal(β1-4)GlcNAc(β1-3)Gal(β1-4)Glc

nd: structural details not defined *relative abundances were calculated from average absolute GSL glycan quantities (pmol/100μL) in10 humn serum samples as determined byMALDI-TOF MS by Furukawa et al18 (Supplementary Table 4, Furukawa et al)18 adeviations in relative abundances between the two studies for LacCer and GM3 might be due to the use of correction factors for the absolute quantitation of LacCer and GM3 by MALDI-TOF MS18 bno structural discrimination was made between Gb3 and (n)Lc4 by MALDI-TOF MS due to same m/z18 cno structural discrimination was made between F-(n)Lc4 isomers by MALDI-TOF MS due to same m/z 18 dno structural discrimination was made between GM1 and S-(n)Lc4 by MALDI-TOF MS due to same m/z 18 em/z which represents (Hex4)(HexNAc2)(Neu5Ac1) was not further specified by Furukawa et al18 but corresponds to S(3)-nLc6 as identified in our study



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 22

Figure legends

Figure 1. Schematic of the high-throughput GSL glycan preparation and analysis workflow. The workflow includes on-plate endoglycoceramidase incubation of extracted GSLs from different matrices to which an internal standard has been added, automated hydrazide bead cleanup of released glycans, high-throughput 2AB-glycan labeling and cleanup of excess label followed by UPLC-HILIC-FLD analysis.

Figure 2. UPLC-HILIC profiles of mammalian cellular GSL-glycans. The glycans were released by R. triatomea rEGCase I (40 mU) (A) NIH/3T3-cells (B) HeLa-cells (C) HL60-cells. See Table 2 for relative peak quantification. *Internal Standard.

Figure 3. Representative UPLC-HILIC-FLD profile and peak assignments of GSL-glycans from human serum. Glycans were released by R. triatomea rEGCase I (60 mU). See Table 3 for relative peak quantification. GU: glucose units.


Page 19 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 2


Page 20 of 22

Page 21 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For TOC only


Page 22 of 22

Comprehensive Profiling of Glycosphingolipid ... - ACS Publications

Recommend Documents