PDF (675 KB) - ACS Publications

Subscriber access provided by UNIV OF SOUTHERN INDIANA

Article

Combining Glucose Units, m/z and Collision Cross Section values: Multiattribute data for increased accuracy in automated glycosphingolipid glycan identifications and its application in Triple Negative Breast Cancer Katherine Wongtrakul-Kish, Ian Walsh, Lyn Chiin Sim, Amelia Mak, Brian Liau, Vanessa Ding, Noor Hayati, Han Wang, Andre Boon-Hwa Choo, Pauline M. Rudd, and Terry Nguyen-Khuong Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01476 • Publication Date (Web): 09 Jun 2019 Downloaded from http://pubs.acs.org on June 10, 2019

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 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 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.

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 19 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

1

Combining Glucose Units, m/z and Collision Cross Section values:

2

Multi-attribute data for increased accuracy in automated

3

glycosphingolipid glycan identifications and its application in Triple

4

Negative Breast Cancer

5 6

Katherine Wongtrakul-Kish1*, Ian Walsh1, Lyn Chiin Sim1, Amelia Mak1, Brian Liau1, Vanessa

7

Ding2, Noor Hayati1, Han Wang3, Andre Choo2, Pauline M Rudd1, Terry Nguyen-Khuong1*

8 9 10

1. Analytics Group, Bioprocessing Technology Institute, Agency for Science, Technology and Research (A*STAR), Singapore 138668

11

2. Antibody Discovery Group, Bioprocessing Technology Institute, A*STAR, Singapore 138668

12

3. Waters Asia Pacific Pte Ltd, 1 Science Park Rd, #02-01/06 The Capricorn, Singapore Science

13

Park II, Singapore 117528

14

1 ACS Paragon Plus Environment

Analytical Chemistry 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

15

Abstract

16

Glycan head-groups attached to glycosphingolipids (GSLs) found in the cell membrane bilayer can alter

17

in response to external stimuli and disease, making them potential markers and/or targets for cellular

18

disease states. To identify such markers, comprehensive analyses of glycan structures must be

19

undertaken. Conventional analyses of fluorescently labelled glycans using hydrophilic interaction high

20

performance liquid chromatography (HILIC) coupled with mass spectrometry (MS) provides relative

21

quantitation and has the ability to perform automated glycan assignments using glucose unit (GU) and

22

mass matching. The use of ion mobility (IM) as an additional level of separation can aid the

23

characterisation of closely-related or isomeric structures through the generation of glycan Collision

24

Cross Section (CCS) identifiers. Here, we present a workflow for the analysis of procainamide-labelled

25

GSL glycans using HILIC IM-MS and a new, automated glycan identification strategy whereby multiple

26

glycan attributes are combined to increase accuracy in automated structural assignments. For glycan

27

matching and identification, an experimental reference database of GSL glycans containing GU, mass

28

and CCS values for each glycan was created. To assess the accuracy of glycan assignments, a distance-

29

based confidence metric was used. The assignment accuracy was significantly better compared to

30

conventional HILIC-MS approaches (using mass and GU only). This workflow was applied to the study

31

of two Triple Negative Breast Cancer (TNBC) cell lines and revealed potential GSL glycosylation

32

signatures characteristic of different TNBC subtypes.

33 34

Graphical Abstract:

GU

2. Automated glycan assignment

Mass

CCS

35

1. Multi-attribute data collection

3. Elucidating GSL glycan changes in TNBC biology


Page 2 of 19

Page 3 of 19 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


36

Glycosphingolipids (GSLs) are amphipathic lipid molecules comprised of a hydrophilic glycan head-

37

group attached to a hydrophobic ceramide tail. The regulation of GSL biosynthesis and metabolic

38

pathways is crucial for their proper biological function, including their roles in cell growth, signal

39

transduction1, and cell identity establishment and maintenance2, 3. Heterogeneity in both the

40

ceramide and glycan head-groups can result in a large number of GSL species, with over 500

41

characterised so far4, however much of the biological function is determined by the glycan head-

42

groups5.

43 44

GSL glycans share a high degree of compositional similarity, yet display a high degree of structural

45

heterogeneity due to differences in monosaccharide sequence, linkage, anomericity and branching.

46

Further complexity can arise through monosaccharide modification with substituents such as

47

sulphate, phosphate and acetate6. The analytical challenge in GSL glycomics lies in unearthing these

48

structural complexities to gain a more comprehensive understanding of altered GSL processing

49

pathways and the role of the glycans in cell function and disease. For the analysis of GSL glycans, liquid

50

chromatography (LC) and mass spectrometry (MS)-based glycomic workflows have recently been

51

successfully applied to GSL glycomics and offer a new, high-resolution and quantitative approach5, 7-9.

52 53

For released, fluorescently labelled glycans, one of the most widely used and established separation

54

techniques is Hydrophilic Interaction Ultra High Performance Liquid Chromatography coupled with

55

fluorescence and electrospray ionisation mass spectrometry detection (HILIC-U/HPLC-FLD ESI-MS)10.

56

In this approach, glycans are identified using their mass to charge (m/z) ratios and HILIC elution profiles

57

which are standardised using a dextran glucose homopolymer, giving each glycan peak a unique

58

glucose unit (GU) identifier11. This method provides relative quantitation information based on

59

fluorescence detection and allows users to search experimental GU values against glycan GU libraries

60

such as those contained in the GlycoStore database for automated glycan assignment12, 13.

61 62

As a consequence of high glycan heterogeneity, GU values can be highly similar for isomeric structures,

63

or multiple glycans may elute in a single chromatographic peak in complex samples. This can lead to

64

ambiguity in structural assignments. Recent advances in MS have included the incorporation of ion

65

mobility separation (IM-MS) which has the potential to separate closely related analytes such as

66

isomeric or isobaric glycans14-16. Separation of gas-phase ions occurs in a drift tube, where ions move

67

under an electric field in a buffer gas. The time taken for a glycan to travel through the IM tube drift

68

gas can be used to calculate a Collision Cross Section (CCS) value using the Mason-Schamp equation17.



69

CCS values can also be reliably utilised as glycan identifiers and, in addition to GUs, increase confidence

70

in automated glycan matching of experimental data against a reference database.

71 72

Here, we present a complete workflow for the analysis of GSL glycans from sample preparation to a

73

high resolution detection system where HILIC-UPLC-FLD is coupled with travelling-wave (TW) IM-MS

74

to obtain both relative quantitation data and qualitative data for glycan identification. Accompanying

75

this, we describe the construction of the first multi-attribute, experimentally derived library of GSL

76

glycans containing the following features: GU, mass and CCS values for all observed ion states ([M+H]+,

77

[M+2H]2+, and [M+H+Na]2+). To our knowledge, the work described here is the first to combine GU,

78

mass and CCS values to accurately annotate GSL glycan conjugates. A method for automated glycan

79

matching based on Euclidean distance geometry of these attributes is also presented and evaluated.

80

By combining multiple attributes we were able to improve the accuracy of glycan identification

81

compared to conventional approaches that use mass and GU alone. Furthermore, the automated

82

assignment of GSL glycans by library matching was benchmarked, allowing a confidence score for each

83

assignment to be calculated. The technology was used to aid glycan identification in cells and

84

successfully distinguish GSL glycosylation signatures between two TNBC subtypes.


Page 4 of 19

Page 5 of 19 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


85

Experimental Section

86

Materials

87

GSL glycans (73 standards covering ganglio-, lacto-, neolacto-, globo- and isoglobo series) standards

88

were purchased from Elicityl (Crolles, France) and LNFP1 glycan standard from Prozyme (CA, USA).

89

GM2 GSL, Procainamide hydrochloride, sodium cyanoborohydride, polyvinyl pyrrolidone and rEGCase

90

II from Rhodocococcus sp. was purchased from Sigma-Aldrich (MO, USA). PD MiniTrap G-10 SEC

91

cartridges were purchased from GE Life Sciences (IL, USA). Ammonium formate solution was

92

purchased from (Waters, (Milford, USA) and, Procainamide-labelled Dextran Homopolymer from

93

Ludger Ltd. (Oxon, UK). Immobilon-P PVDF membrane (0.45 µm), acetonitrile, DMSO, acetic acid,

94

methanol, 1-butanol, chloroform, sodium acetate and LC-MS grade water were from Merck (NJ, USA).

95

Phosphate Buffered Saline (PBS) was from Axil Scientific (Singapore, Singapore) and polypropylene

96

plates from Corning® Costar® (UT, USA). Rosewell Park Memorial Institute (RPMI) 1640 media,

97

Leibovitz’s L-15 media and penicillin-streptomycin were from Gibco (NY, USA) and HyClone™ Fetal

98

Bovine Serum was from FisherScientific (USA). MDA-MB-453, MCF-7 and BT474 cells were purchased

99

from the American Type Culture Collection ATCC (VA, USA) and BT549 cells were from the National

100

Cancer Institute NCl-60 panel (MD, USA).

101 102

Cell Culture and harvest

103

BT549 cells were cultured in RPMI 1640 media supplemented with 10 % Fetal Bovine Serum and 1 %

104

Penicillin-Streptomycin and collected at passage number 13. MCF7 cells were cultured in RPMI 1640

105

media supplemented with 10 % Fetal Bovine Serum (FBS) and collected at passage number 16. BT474

106

cells were grown in 1:1 DMEM:Ham’s F12 supplemented with 2 mM L-glutamine and 10 % FBS and

107

collected at passage number 18. The cells were grown to 80 % confluency at 37 ˚C in 5 % CO2. The

108

MDA-MB-453 cell line was cultured in Leibovitz's L-15 media supplemented with 10 % FBS and

109

collected at passage number 7. The cells were grown to 80 % confluency before passaging, at 37 ˚C in

110

at atmospheric gas composition. Cells were washed twice with PBS before scraping for collection. Cells

111

were pooled from multiple culture flasks to make a total of 3 x 108 cells per triplicate and pelleted by

112

centrifugation at 2500 g for 20 min. Pellets were stored at -80 ˚C.

113 114

Extraction of GSLs

115

GSL extraction was performed based on Anugraham et al.7 and further purified according to

116

Vidugiriene and Menon18 (see Experimental Section in Supporting Information).

117 118



119 120

Glycan release

121

For polyvinylidene difluoride (PVDF) membrane-based glycan release, 6 x 10 µg GM2 was solubilized

122

in 50 µL chloroform/methanol (2:1) and spotted onto individual membrane spots in a 96-well

123

polypropylene plate according to Anugraham et al.7. Samples were left to bind overnight before

124

blocking with 1 % PVP in 50 % methanol/water. For in-solution glycan release, dried GSL samples were

125

solubilized in 50 µL of 50 mM sodium acetate (pH 5.0). Both PVDF-bound and in-solution samples were

126

treated with 4 µL (8 mU) of rEGCase II and incubated at 37 ˚C for 18 h for one-night digestion. For two

127

night’s digestion, samples were incubated initially for 24 h followed by the addition of a further 2 µL

128

(4 mU) of rECGase II, and incubation for another 19 h. The released glycan solution was transferred to

129

fresh Eppendorf tubes containing 1 mL chloroform/methanol/water (8:4:3)7. Sample vials were

130

washed with 50 µL of DI water which was pooled in the Eppendorf tube. Tubes were vortexed and

131

centrifuged. The upper glycan-containing methanol/water layer was extracted and dried in a vacuum

132

centrifuge.

133 134

Fluorescent labelling with 2-aminobenzamide

135

Glycans were solubilised in 25 µL water and transferred to a glass vial for labelling with 2-

136

aminobenzamide (2-AB) via reductive amination. A mixture of 20 µL 0.35 M 2-AB and 1 M sodium

137

cyanoborohydride in 7:3 (v/v) DMSO/acetic acid was added to each sample and incubated at 37 ˚C for

138

16 h with agitation at 800 rpm.

139 140

Fluorescent labelling with Procainamide

141

Glycans were solubilised in 10 µL water and transferred to a glass vial for labelling with procainamide

142

via reductive amination. A 100 µL solution of 0.4 M procainamide hydrochloride and 0.9 M sodium

143

cyanoborohydride in 7:3 (v/v) DMSO/acetic acid was prepared, followed by the addition of 30 µL water

144

to result in a clear labelling mixture. Ten microliters (10 µL) of procainamide labelling mix was added

145

to each sample and incubated at 37 ˚C for 16 h.

146 147

Post-labelling clean-up

148

For removal of excess label, glycan-label mixtures were diluted with water to a total volume of 300 µL

149

and then applied to individual PD MiniTrap G-10 SEC cartridges. For breast cancer glycan samples, 10

150

% of the sample was removed for G-10 clean-up. Glycans were eluted in water and dried in a vacuum

151

centrifuge.

152


Page 6 of 19

Page 7 of 19 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


153 154

HILIC-UPLC-FLR and ESI-IM-MS

155

Labelled glycans were analysed by either HILIC-UPLC-FLD or HILIC-UPLC-FLD with IM-MS on an

156

ACQUITY UPLC H-Class (Waters Corporation, MA, USA) with a fluorescence detector (ex = 310 nm,

157

em = 370 nm for procainamide; ex = 330 nm, em = 428 nm for 2-AB). Chromatography analyses

158

were carried out based on Albrecht et al.5 (see Experimental Section in Supporting Information). IM-

159

MS measurements were made online using a Synapt G2S quadrupole/IMS/orthogonal acceleration

160

time-of-flight MS instrument (Waters, MA, USA) fitted with an ESI ion source (see Experimental

161

Section in Supporting Information for analysis details).

162 163

Data processing

164

MassLynx data were imported into the Waters UNIFI Scientific Information System for GU calculation

165

using the ‘Glycan Assay (FLR with MS Confirmation)’ processing method. GU values were calculated

166

by normalising glycan retention times against procainamide-labelled dextran ladder using a fifth order

167

polynomial distribution curve. Mobility data was processed for CCS calculation using UNIFI’s Accurate

168

Mass Screening on IMS data method. FLD peak integration was done manually for area-under-curve

169

based quantitation and all glycan peak areas within a sample were normalised to 100 % for relative

170

quantitation.

171 172

Seventy three GSL glycan standards were used to build an experimental reference library containing

173

up to five attributes for each glycan: theoretical mass and experimentally observed GU and CCS values

174

for each glycan and its detected ion states ([M+H]+, [M+2H]2+ and [M+H+Na]2+).

175 176

For glycan identification in breast cancer cell line samples, data files were processed as described

177

above. GU, FLD peak area, retention time, observed m/z and CCS data was extracted for each peak

178

and glycan identifications made using an automated in-house searching algorithm (see Experimental

179

Section in Supporting Information for further details).

180 181

Safety consideration

182

Chloroform is toxic, non-odorous and volatile. It is recommended that experienced personnel observe

183

and handle the chemical with strict safety measures and consideration.



184

RESULTS AND DISCUSSION

185

Methodology development for GSL extraction and labelling

186

GSLs were extracted from breast cancer cell lines using a modified Folch extraction procedure that has

187

been previously reported and demonstrated to be a reliable method for the enrichment of sialylated

188

gangliosides from cell cultures7, 19. The extraction of GSLs from cell membranes described by Albrecht

189

et al.5 also included a further purification step of the crude GSL fraction which involved n-

190

butanol/water partitioning. This step helps to remove polar impurities and reduce the amount of

191

contaminant monosaccharides from crude GSL extracts. With the aim of improving yield and

192

sensitivity of GSL glycans from biological samples, the inclusion of this step was tested with GSLs

193

extracted from BT474 cells. When omitting this extra purification step, the glycan profile showed

194

several contaminating peaks that did not correspond to glycan compositions (Figure S-1A & S-1B).

195

Several of these peaks were removed after partitioning, however peaks containing glycans were also

196

greatly reduced (Figure S-1C & S-1D) and therefore this purification step was omitted from the

197

processing of subsequent samples.

198 199

Optimisation of GSL glycan release conditions

200

For glycan release using rEGCase II from Rhodocococcus sp., GM2 standard was used to compare

201

release from solubilised GSLs with GSLs immobilised on PVDF membrane. PVDF membrane-based

202

release involves the immobilisation of the hydrophobic ceramide portion of GSLs to the hydrophobic

203

membrane surface, leaving the hydrophilic glycan portion exposed for enzymatic release7. In-solution

204

based digestions are a common way to perform glycan release from glycoconjugates and require

205

fewer experimental steps compared to PVDF-based digests. In-solution digest has also been shown to

206

produce glycan profiles with double the signal intensity compared to PVDF-based digest for

207

glycoprotein N-glycan release20. To ensure the release of different GSL types, enzyme amounts and

208

digestion times must also be taken into consideration. Digestion times ranging from 16 to 48 h have

209

previously been reported for GSLs derived from biological samples5, 7, 9. Therefore, in addition to

210

comparing in-solution vs. PVDF-based release, fluorescence peak areas for released GM2 glycan were

211

also assessed between one and two nights’ digest with rEGCase II. The amount of enzyme used was

212

based on Anugraham et al.1. The largest glycan yield was observed after two night’s in-solution digest

213

and this was significantly higher than any of the PVDF-based digests (p < 0.013; Figure S-2; Table S1).

214

Interestingly, two night’s digestion of PVDF-membrane bound GSLs resulted in the lowest glycan yield

215

which is contrary to the findings of Anugraham et al.7.

216 217


Page 8 of 19

Page 9 of 19 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


218

Selection of a high sensitivity label

219

To ensure optimal glycan fluorescence and MS signal, procainamide and 2-AB glycan labels were

220

compared. 2-AB is a widely used and long-established label for the fluorescent detection of released

221

glycans but performs relatively poorly in MS detection due to its poor ionisation efficiency21.

222

Procainamide was found to have 16.6 times higher fluorescence (p

PDF (675 KB) - ACS Publications

Recommend Documents