SINGLE STAGE TANDEM MASS SPECTROMETRY ASSIGNMENT OF THE C-5 URONIC ACID STEREOCHEMISTRY IN HEPARAN SULFATE TETRASACCHARIDES USING ELECTRON DETACHMENT DISSOCIATION

Abstract

The analysis of heparan sulfate (HS) glycosaminoglycans presents many challenges, due to the high degree of structural heterogeneity arising from their non-template biosynthesis. Complete structural elucidation of glycosaminoglycans necessitates the unambiguous assignments of sulfo modifications and the C-5 uronic acid stereochemistry. Efforts to develop tandem mass spectrometric-based methods for the structural analysis of glycosaminoglycans have focused on the assignment of sulfo positions. The present work focuses on the assignment of the C-5 stereochemistry of the uronic acid that lies closest to the reducing end. Prior work with electron-based tandem mass spectrometry methods, specifically, electron detachment dissociation (EDD), have shown great promise in providing stereo-specific product ions, such as the B₃′–CO_2, which has been found to distinguish glucuronic acid (GlcA) from iduronic acid (IdoA) in some HS tetrasaccharides. The previously observed diagnostic ion are generally not observed with 2-O-sulfo uronic acids or for more highly sulfated heparan sulfate tetrasaccharides. A recent study using electron detachment dissociation and principal component analysis revealed a series of ions that correlate with GlcA versus IdoA for a set of 2-O-sulfo HS tetrasaccharide standards. This present work comprehensively investigates the efficacy of these ions for assigning the C-5 stereochemistry of the reducing end uronic acid in thirty-three HS tetrasaccharides. A diagnostic ratio can be computed from the sum of the ions that correlate to GlcA to those that correlate to IdoA.

Graphical Abstract

INTRODUCTION

A large portion of the extracellular matrix and basement membranes are composed of proteoglycans, composed of proteins covalently attached to the class of carbohydrates called glycosaminoglycans (GAGs) [1–3]. GAGs are linear negatively charged biopolymers whose basic building blocks consist of a repeating disaccharide sequence of an amino sugar and a uronic acid or galactose [4, 5]. They are categorized as either keratan sulfate (KS), chondroitin sulfate (CS), dermatan sulfate (DS), hyaluronan (HA) or heparan sulfate (HS) depending on their disaccharide repeating unit [1–6]. Among these classes of GAGs, HS is the most structurally complex [6, 7]. They are initially synthesized in the Golgi apparatus as alternating disaccharide units of D-glucuronic acid and N-acetylated glucosamine [6, 8]. C5-epimerization of glucuronic acid (GlcA) to iduronic acid (IdoA) occurs during the biosynthesis of HS followed by a series of sulfo modifications [7, 8]. Sulfo modifications may occur at the 2-O position of the uronic acid, and the N-, 3-O and 6-O positions of the glucosamine unit [8]. These structural modifications often do not go to completion, producing HS chains with varying sequences of sulfation, acetylation, and IdoA/GlcA content [6, 9]. Despite these varying structural modifications, specific structural motifs on HS chains have been reported to bind target proteins with high specificity. These categories of proteins, called heparan sulfate binding proteins (HSBPs), include chemokines, cytokines, blood coagulation factors such as serine proteases, cell adhesion proteins, growth factors and morphogenetic factors [7, 10, 11]. Documented physiologic processes influenced by HS interactions with proteins include growth and development, cancer, inflammation, viral infectivity and blood coagulation [12–15]. The numerous biological functions of this bio-molecule continue to inspire research into structure-function relationships of HS oligomers. However, this research has been hampered by their enormous micro-heterogeneity and limited availability requiring very sensitive and robust analytical methods for their analysis.

Advanced analytical methods like nuclear magnetic resonance (NMR) spectroscopy have been used to determine sulfo modifications and the C-5 stereochemistry of the uronic acid in GAGs [16, 17]. However, the quantity and purity of GAGs extracted from natural sources are often not suitable for NMR analysis[18]. These drawbacks make mass spectrometry an excellent alternative for GAG analysis. Negative electrospray ionization mass spectrometry offers an excellent platform for GAG analysis offering high sensitivity, throughput, and accuracy [19, 20]. However, tandem mass spectrometry of GAGs especially heparan sulfate is often challenging due to variations in oligomer length, hexuronic acid stereochemistry, sulfation heterogeneity[6, 21]. ESI-MS is able to determine the length, degree of sulfation of GAGs, and other features affecting the elemental composition. To determine sites of sulfation, N-acetylation, and hexuronic acid stereochemistry, more advanced methods are required. Recent advances in tandem mass spectrometric applications to GAGs using collision-induced dissociation (CID) [22, 23], infrared multiphoton dissociation (IRMPD) [24], electron induced dissociation (EID) [25], electron detachment dissociation(EDD) [26–28] and negative electron transfer dissociation (NETD) [29] have addressed some of the challenges encountered during their structural analysis. Inherent sulfo decomposition of the labile sulfate half ester group present in GAGs hinder structural characterization. This phenomenon occurs mostly during the ionization and ion activation stages of the experiment. Chemical derivatization [30], deprotonation of the acidic groups [23, 31, 32] and metal cation exchange [22, 23] have been reported to effectively reduce sulfo decomposition depending on the degree of sulfation. Recent CID MS/MS reports on Arixtra and highly sulfated heparan sulfate GAGs showed one can obtain very rich and structurally informative product ion coverage by adding dilute NaOH to the spray solution [33, 34]. Electron-based activation methods, especially EDD, have shown great potential in providing highly informative cross-ring products as well as the corresponding glycosidic cleavages which are essential for localization of sulfo positions [25–27]. Although mass spectrometry methodologies continue to gain ground in assigning sites of sulfo modifications, a remaining challenge has been the inability to discriminate diastereomers that differ by the chirality of the uronic acid C-5 center. Zaia and coworkers were the first to address this challenge. Using collision induced dissociation, they were able to differentiate chondroitin sulfate (CS) from dermatan sulfate (DS) in mammalian extracellular matrix, and quantitatively assign the amount of the diastereomers present in mixtures. [35]. The ability to assign the C-5 stereochemistry in uronic acid residues of 4-O-sulfo chondroitin sulfate epimers with varying oligomer lengths (dp4–dp10) based on diagnostic cross-ring ions ^2,4A_n, and ^0,2X_n in CID mass spectra has also been reported by Kailemia et al [36]. Compared to heparan sulfates, chondroitin sulfates have well-defined sulfation pattern, hence the former requires a more sensitive activation method for stereochemistry assignments. EDD results for HS tetrasaccharides reported by Wolff et al. showed the possibility of obtaining a stereospecific ion B₃′-CO₂ for assigning the C-5 stereochemistry for moderately sulfated tetramers (0–0.25 sulfates per disaccharide) [18]. Gas phase separation of epimeric mixtures of HS tetrasaccharides using field asymmetric ion mobility spectrometry (FAIMS) followed by EDD fragmentation confirmed the presence of the B₃′-CO₂ ion for the GlcA containing epimer [37]. Recent EDD reports, however, showed that the presence of a sulfo group at the 2-O position of the uronic acid hinders production of the B₃′–CO₂ ion in GlcA containing epimers [38]. More recent work on the assignment of the C-5 stereochemistry for 2-O-sulfated HS tetramers (0.5–2.5 sulfates per disaccharide) revealed the possibility of assigning the C-5 stereochemistry of HS tetrasaccharides using a ratio combination of selected ions obtain from EDD-PCA experiments [38]. Such analyses are useful when epimeric compounds are available. The scarcity of naturally occurring epimeric HS samples has motivated the development of a technique that assigns the stereochemistry of these tetramers without reference to their isomers. Here we present for the first time a more general approach in assigning the C-5 hexuronic stereochemistry for thirty-three HS tetrasaccharide standards from a single stage EDD tandem mass spectrum.

EXPERIMENTAL

Sample Preparation

Thirty-three heparan sulfate tetrasaccharides standards were synthesized using a modular approach [39]. All the compounds examined had their compositions confirmed using FTICR MS accurate mass measurement and had their structures confirmed by ¹H NMR, HSQC, and COSY. Supplemental Figures 1–4 show the chemical structures of all thirty-three compounds.

Mass Spectrometry Analysis

EDD experiments were performed on a 9.4T Bruker Apex Ultra QeFTMS (Billerica, MA) with a hollow cathode (HeatWave, Watsonville, CA) which serves as the source of electrons for EDD. 0.1 mg/mL of each standard were injected at a rate of 120μL/h in 50:50 methanol:H₂O and ionized by a metal electrospray capillary (Agilent Technologies, Santa Clara, CA, #G2427A). Where necessary, 0.1mM NaOH was added to the spray solvent to enhance the intensity of preferred sodium adducted precursor ions for analysis. All the HS tetrasaccharides were analyzed in the negative ion mode. Each EDD experiment was repeated three times with almost similar results for each HS standard examined.

For the EDD experiment, multiply charged precursor ions were isolated in the external quadruple and accumulated for 1–4s before injection into the FT-ICR cell. Precursor ion selections were refined using in-cell isolation with a coherent excitation frequency (CHEF) event. These ions were then irradiated with 19eV electrons for a second. The extraction lens was set to −18.5 ± 0.5V with the cathode heater at 1.5A. 24 acquisitions were signal averaged per spectrum. 512 K points were acquired for each spectrum, padded with one zero fill, and apodized using a sine bell window. Internal calibration was achieved using confidently assigned glycosidic product ions as internal calibrants, providing mass accuracy of <1 ppm. We report peaks with S/N> 10 due to the large number of low-intensity product ions formed by EDD. All cross-ring and glycosidic product ions generated from the EDD experiment were assigned using accurate mass measurement and GlycoWorkbench [40]. These ions are reported using the Domon and Costello nomenclature [41].

RESULTS AND DISCUSSION

Determination of Sulfo positions

Electron detachment dissociation has been shown to provide excellent product ion coverage for the structural analysis of glycosaminoglycans. Figure 1 shows abundant glycosidic and cross ring product ions for the epimeric di-sulfated HS standards epimers, labeled (2b, 2f, 2d, and 2g). EDD fragmentation of the [M-3H]³⁻ precursor ion produced almost identical fragmentation for these compounds as one would expect for these epimers (Figure 1). These compounds differ only in the C-5 hexuronic acid stereochemistry and have both their glucosamine unit sulfated at the 6-O position. We are able to unambiguously assign the sites of sulfation on these epimers with a combination of cross-ring and glycosidic product ions. The mass difference between product ions C₁ and ^3,5A₂, and C₃ and ^3,5A₄ confidently identifies the two 6-O sulfo groups on the second and fourth glucosamine units towards the reducing end for HS standards 2b, 2d, and 2g (Figure 1). A similar assignment of the 6-O sulfo group on the second glucosamine unit from the non-reducing end can be made using the C₁ and ^3,5A₂ for compound 2b; however, the sulfate group on the reducing end sugar is assigned using the mass difference between cross-ring ions product ions ^2,4A₄ and ^0,2A_4. We also show in Figure 2, efficient EDD fragmentation for tri-sulfated epimers 3b-(IdoA-GlcNS6S-GlcA-GlcNS-(CH₂)₅NH₂ and 3h-(GlcA-GlcNS6S-IdoA-GlcNS-(CH₂)₅NH₂) for the [M-4H+Na]³⁻ precursor ion. The N and 6-O sulfo groups on the second residue near the non-reducing ends can be assigned using the mass difference between ^0,2A₂ and B₂ and ^2,4A₂ and ^0,2A₂ product ions respectively. The N-sulfo group on the reducing end glucosamine is assigned using cross-ring product ion ^0,2X0. We have included a comprehensive mass list (m/z and intensity) showing similar fragmentation efficiency for triplicate EDD experiment for all the thirty-three standards in the accompanying supplemental material.

Figure 1.

EDD mass spectra with structural annotations for four epimeric synthetic di-sulfated HS tetrasaccharides: (a) 2b-(IdoA-GlcNAc6S-GlcA-GlcNAc6S-(CH₂)₅NH₂), (b) 2f-(GlcA-GlcNAc6S-GlcA-GlcNAc6S-(CH₂)₅NH₂), (c) 2d-(GlcA-GlcNAc6S-IdoA-GlcNAc6S-(CH₂)₅NH₂) and (d) 2g-(IdoA-GlcNAc6S-IdoA-GlcNAc6S-(CH₂)₅NH₂)

Figure 2.

EDD mass spectra and annotated structures for the tri-sulfated HS tetrasaccharides: (a) 3b-(IdoA-GlcNS6S-GlcA-GlcNS-(CH₂)₅NH₂ and (b) 3h-(GlcA-GlcNS6S-IdoA-GlcNS-(CH₂)₅NH₂) for the [M-4H+Na] ³⁻ precursor ion.

Assignment of the reducing end hexuronic acid stereochemistry

Electrospray ionization mass spectra for all HS standards considered for this work produced abundant multiply-charged ions for the EDD experiment. Figure 3 shows the general structure for the synthetic heparan sulfate tetrasaccharides epimers with an aminopentyl linker on the anomeric carbon. This modification provides a mass shift for the resulting reducing end fragments ions that enable confident assignment of product ions that would otherwise exhibit identical masses for reducing end and non-reducing end products. In this figure, arrows highlight the C-5 carbon of the uronic acid residue that is the focus of this investigation, namely the acidic sugar closest to the reducing end.

Figure 3.

General structure for the synthetic heparan sulfate tetrasaccharide epimers with an anomeric aminopentyl linker, where R₁= H, SO₃H and R₂= H, SO₃H, Ac

ESI-MS of the mono-sulfated synthetic heparan sulfate standards yield abundant doubly charged ions molecular ions, [M-2H]²⁻. With this charge state, two of the three acidic protons are removed during ionization. Close examination of the EDD spectra for the GlcA containing standards 1a and 1b, Figure 4a, reveals results consistent with previous EDD reports on modestly sulfated HS tetrasaccharides [35, 40]. Diagnostic ion B₃′-CO₂ indicative of glucuronic acid in less sulfated HS GAGs (0–0.25 sulfates per disaccharide) reported for naturally extracted and synthetically produced HS tetrasaccharides [18, 21] are observed for only samples 1a (IdoA2S-GlcNAc-GlcA-GlcNAc-(CH₂)₅NH₂₎ and 1b (IdoA-GlcNAc6S-GlcA-GlcNAc-(CH₂)₅NH₂₎ Figure 4a, occurring at m/z 589.0972. An expanded mass spectra region for the stereospecific ion B₃′-CO₂ is shown in Figure 4a. An intense B₃′ ion relative to its original B₃ ion is a feature also reported for being diagnostic for GlcA containing HS [18]. These ions for standards 1a and 1b stand in stark contrast to standards 1d (GlcA-GlcNAc-IdoA-GlcNAc6S-(CH₂)₅NH₂₎ and 1f (IdoA-GlcNAc-IdoA-GlcNAc6S-(CH₂)₅NH₂), with the B₃′ ion with a lower intensity relative to its B₃ ion (Figure 4b). The absence of the B₃′ ion for standard 1c (GlcA-GlcNAc-IdoA2S-GlcNAc-(CH₂)₅NH₂₎ and 1e (GlcA-GlcNAc-GlcA2S-GlcNAc-(CH₂)₅NH₂₎ further confirmed the influence of the 2-O sulfation on the uronic acid in the formation of the B₃′-CO₂ for 1e as reported earlier [38]. The non-specificity of this previously reported ion over a wide range of sulfo modifications has been the main motivation for this work. Principal component analysis of EDD results from our most recent work revealed several ions that could be used to differentiate GlcA2S from IdoA2S. These ions included, B₃, Y₁, C₂ and Z₂ fragments, found to be diagnostic for GlcA2S, while Y₂ and ^1,5X₂ were diagnostic for IdoA2S. We combine these ions into a diagnostic ratio formula below (equation 1) capable of assigning the C-5 stereochemistry of HS tetrasaccharide standards.

Figure 4.

Expanded EDD spectra for the [M - 2H]²⁻ precursor ion for the mono-sulfated HS (a) for compounds 1a, 1b, 1c and standards showing the region for the stereospecific ion B₃′-CO₂ 1e m/z 570–640, and (b) compounds 1d and 1f m/z 509–580.

$$\text{DR}=log(1/3((\mathrm{\sum }(\text{GlcA})/\mathrm{\sum }(\text{IdoA}))))=log(1/3((\mathrm{\sum }({\mathrm{B}}_3,{\mathrm{Y}}_1,{\mathrm{C}}_2,{\mathrm{Z}}_2)/(\mathrm{\sum }({\mathrm{Y}}_2,{{}^{1,5}\mathrm{X}}_2)))$$ (1)

Equation 1 computes the ratio of the sum of intensities of ions that are statistically validated as diagnostic for the presence of GlcA to the sum of intensities of ions diagnostic of IdoA. The factor of one-third applied to each sum was determined empirically to produce a positive value for the diagnostic ratio (DR) when GlcA is present at the second residue from the reducing end, and negative values when IdoA is present in the same position. The DR values computed for all the standards include fragments with sulfo losses from the selected ions. The discussion below examines the results for a comprehensive set of HS standards with 1 to 4 sulfo modifications.

Diagnostic ratio results for mono-sulfated HS standards

Diagnostic ratio (DR) results for all six HS standards for the mono-sulfated tetrasaccharides labeled 1a to 1f are shown in Figure 5. For a common precursor ion selection [M-2H]²⁻, at m/z 469.6327 and subsequent electron irradiation, we are able to confidently resolve the C-5 hexuronic acid stereochemistry for the residue closest to the reducing end, for singly sulfated standards. All the GlcA containing standards showed positive diagnostic ratio results while those for IdoA where negative. The lowest DR results obtained for these set of HS tetramers containing GlcA was 0.61 ± 0.07 (GlcA-GlcNAc-GlcA2S-GlcNAc-(CH₂)₅NH₂) and the highest for mono-sulfated standards containing IdoA was −0.38 ± 0.09 (IdoA-GlcNAc-IdoA-GlcNAc6S-(CH₂)₅NH₂). Epimeric pair 1c (GlcA-GlcNAc-IdoA2S-GlcNAc-(CH₂)₅NH₂) and 1e (GlcA-GlcNAc-GlcA2S-GlcNAc-(CH₂)₅NH₂) with 2-O sulfo modification on the central uronic unit are clearly differentiated as shown in Figure 5. We also note the non-reducing end stereochemistry has a negligible impact on the DR results for the mono-sulfated set of compounds. This is evident comparing the diagnostic ratio results which was −0.44 ± 0.03 for standards 1d (GlcA-GlcNAc-IdoA-GlcNAc6S-(CH₂)₅NH₂) and −0.38 ± 0.09 for 1f (IdoA-GlcNAc-IdoA-GlcNAc6S-(CH₂)₅NH₂).

Figure 5.

EDD diagnostic ratio results for the mono-sulfated HS tetrasaccharides standards (1a–1f), for the [M-2H]²⁻ precursor ion

Diagnostic ratio results for di-sulfated HS standards

ESI-MS of the di-sulfated HS standards produced abundant [M-3H]3⁻ ions at m/z 339.4050 and 325.4015 depending on the disaccharide repeating sequence. These tetrasaccharides have four ionizable protons, two on the sulfo groups and two on the carboxyl groups. Figure 6 shows the diagnostic ratio results obtained for all the eight HS standards examined upon EDD fragmentation of the [M-3H]3⁻ precursor ion. Again, we show the standards containing GlcA residues towards the reducing end have positive diagnostic ratio values relative to those containing IdoA residues. The broad selection of standards allowed for the applicability of diagnostic ratio formula to discriminate epimers. Standards 2b (IdoA-GlcNAc6S-GlcA-GlcNAc6S-(CH₂)₅NH₂) and 2d (GlcA-GlcNAc6S-IdoA-GlcNAc6S-(CH₂)₅NH₂) differing in the C-5 stereochemistry for the non-reducing end and central uronic acid residues were unambiguously resolved with diagnostic ratio values 0.64 ± 0.01 and −0.22 ± 0.01 respectively. The diagnostic ratio results for the remaining epimer pairs, [2c (GlcA-GlcNAc-IdoA2S-GlcNAc6S-(CH₂)₅NH₂: −0.33 ± 0.05, and 2e (GlcA-GlcNAc-GlcA2S-GlcNAc6S-(CH₂)₅NH₂: 0.83 ± 0.11] and [2f (GlcA-GlcNAc6S-GlcA-GlcNAc6S-(CH₂)₅NH₂ 0.60 ± 0.06 and, 2g (IdoA-GlcNAc6S-IdoA-GlcNAc6S-(CH₂)₅NH₂: −0.11 ± 0.03] further established the usefulness of the formula to discriminate epimers. The overall contribution of the non-reducing end uronic acid stereochemistry is again observed to have very minimal impact on the diagnostic ratio values as we compare standards 2b and 2f (Figure 6). Standard 2h IdoA-GlcNAc-IdoA2S-GlcNS-(CH₂)₅NH₂), which has a different disaccharide repeating sequence compared to the seven di-sulfated standards produced results consistent with the IdoA containing tetramers. The lowest diagnostic ratio value recorded for the GlcA containing standards was 0.11 ± 0.04 for 2a (GlcA2S-GlcNAc6S-GlcA-GlcNAc-(CH₂)₅NH₂) and the highest for with IdoA was −0.11 ±0.03 recorded for 2g (IdoA-GlcNAc6S-IdoA-GlcNAc6S-(CH₂)₅NH₂).

Figure 6.

EDD diagnostic ratio results for the di-sulfated HS tetrasaccharides standards (2a–2h) for the [M-3H]³⁻ precursor ion.

Diagnostic ratio results for tri-sulfated HS standards

Previous EDD and CID reports have shown the presence of sodium counter ions can provide a means to control stereochemical dependent fragmentation [21, 36]. We found that to be very useful for the tri and tetra-sulfated HS standards. For the tri-sulfated HS standards, the [M-4H+Na]3⁻ precursor ion at m/z 345.3775 was selected for the EDD experiment. This ionized state allows for all the sulfo groups and a carboxylic group to be ionized. Figure 7 shows diagnostic ratio results for ten HS tetrasaccharides standards. The C-5 stereochemistry of the central uronic acid is clearly resolved using the respective diagnostic ratios for the isomers. The lowest diagnostic ratio recorded for the GlcA containing isomers was 0.05 ± 0.01 for standard 3a (IdoA2S-GlcNS-GlcA-GlcNS-(CH₂)₅NH₂, and the closest to zero for those with IdoA was −0.12 ± 0.01, recorded for sample 3i (IdoA-GlcNS-IdoA-GlcNS6S-(CH₂)₅NH₂. The ability to resolve epimers for the tri-sulfated HS standards using the diagnostic ratio formula is tested with standards 3b (IdoA-GlcNS6S-GlcA-GlcNS-(CH₂)₅NH₂) and 3h (GlcA-GlcNS6S-IdoA-GlcNS-(CH₂)₅NH₂) and, 3c (GlcA-GlcNS-IdoA2S-GlcNS-(CH₂)₅NH₂) and 3e (GlcA-GlcNS-GlcA2S-GlcNS-(CH₂)₅NH₂) as shown on Figure 7. Even though standards 3f (GlcA-GlcNAc6S-IdoA2S-GlcNAc6S-(CH₂)₅NH₂), 3g (IdoA2S-GlcNAc6S-GlcA-GlcNAc6S-(CH₂)₅NH₂ and 3j (IdoA-GlcNAc6S-IdoA2S-GlcNAc6S-(CH₂)₅NH₂), have different repeating disaccharide units compared to the other seven isomers, EDD fragmentation of the [M-4H+Na]3⁻ precursor ion (m/z 373.3846) for 3f, 3g and 3j produced diagnostic ratio results consistent for assigning their hexuronic acid stereochemistry. We observe a minor diagnostic ratio difference between standards 3f (−0.12 ± 0.01) and 3j (−0.24 ± 0.01) which differ only at the non-reducing end uronic acid. However, this minor difference had no impact on the assignment of the stereochemistry of the central uronic acid.

Figure 7.

EDD diagnostic ratio results for the tri-sulfated HS tetrasaccharides standards (3a–3j), for the [M-4H+Na]³⁻ precursor ion.

Diagnostic ratio results for the tetra-sulfated HS standards

Highly sulfated HS GAGs are difficult to characterize due to loss of the labile sulfo groups. However, we show in Figure 8, reproducible diagnostic ratio results for these highly sulfated tetramers for the [M-5H+2Na]³⁻ precursor for all seven isomers (standards 4a–4g) and [M-5H+Na]⁴⁻ for 4h and 4i. The selected precursor ion for the diagnostic ratio analysis ensured at least all sulfo groups are ionized. The diagnostic ratio values obtained for this degree of sulfation allowed for confident assignment of their respective C-5 uronic acid stereochemistry. The lowest diagnostic ratio recorded for the GlcA containing tetra-sulfated standards was 0.066 ± 0.006 and the closest to zero for the IdoA standards was −0.06 ± 0.01, recorded for 4a (IdoA2S-GlcNS6S-GlcA-GlcNS-(CH₂)₅NH₂) and 4f (IdoA-GlcNS6S-IdoA-GlcNS6S-(CH₂)₅NH₂) respectively. Epimeric pairs 4c (GlcA-GlcNS-IdoA2S-GlcNS6S-(CH₂)₅NH₂) and 4d (GlcA-GlcNS-GlcA2S-GlcNS6S-(CH₂)₅NH₂) are clearly resolved as shown in Figure 8, with diagnostic ratio values −0.40 ± 0.03 and 0.109 ± 0.007 respectively. Epimeric standards 4b (IdoA-GlcNS6S-GlcA-GlcNS6S-(CH₂)₅NH₂) and 4f (IdoA-GlcNS6S-IdoA-GlcNS6S-(CH₂)₅NH₂) are also differentiated unambiguously based on the diagnostic ratio values 0.35 ± 0.01 and −0.06 ± 0.01 respectively. We again compare the diagnostic ratio results for standards 4b (IdoA-GlcNS6S-GlcA-GlcNS6S) and 4e (GlcA-GlcNS6S-GlcA-GlcNS6S) differing only at the non-reducing end uronic acid. Their respective diagnostic values 0.35 ± 0.01, and 0.36 ± 0.02 again indicated the non-reducing end uronic acid stereochemistry had little or negligible impact on the diagnostic ratio results. Standards 4c (GlcA-GlcNS-IdoA2S-GlcNS6S-(CH₂)₅NH₂) and 4g (IdoA-GlcNS-IdoA2S-GlcNS6S-(CH₂)₅NH₂) also differing at the non-reducing end uronic acid residue produced almost similar diagnostic ratio values as shown in Figure 8. A mass list (mass (m/z) – intensity table) for the ions used for the DR calculations for all thirty-three HS standards are included in the supplemental material.

Figure 8.

EDD diagnostic ratio results for the tetra-sulfated HS tetrasaccharides standards 4a–4g for the [M-5H+2Na]³⁻ precursor ion and [M-5H+Na]⁴⁻ for standards 4h and 4i.

CONCLUSIONS

In this study, we have demonstrated the capability to assign the C-5 hexuronic acid stereochemistry from a single stage tandem mass spectrum using EDD. The diagnostic ratio provides the means to assign the stereochemistry of the uronic acid near the reducing end for HS tetrasaccharides. For all 33 tetramers that were examined, the diagnostic ratio was positive for GlcA near the reducing end, while those having IdoA residues near the reducing end had negative DR values. The smallest absolute values of DR for an IdoA containing tetrasaccharide standard was −0.06 ± 0.01 and for a GlcA was 0.05 ± 0.01. These data show that the diagnostic ratio clearly distinguishes the uronic acid stereochemistry, not by comparison to a standard, but with a number derived directly from the data.

The applicability of this approach to typical analytical problems faced by glycosaminoglycan researchers remains to be demonstrated. Generally speaking, researchers are confronted with mixtures of GAG oligomers, and these would need to be resolved before the type of analysis presented in this paper could be performed, as the diagnostic ratio only has significance for single component samples. Secondly, the diagnostic ratio presented here have been demonstrated on tetramers that are alkylated at the reducing end. This modification breaks the symmetry of the structure, and allows one to easily distinguish reducing end from non-reducing end fragments. This derivatization can be performed on real-world samples, but will require an extra step in the work-up procedure. Finally, this approach has been demonstrated only for assigning the stereochemistry of the uronic acid residue closest to the reducing end. Future work will focus on extending this approach to assign the C-5 hexuronic acid stereochemistry of additional residues in longer chain HS GAGs using EDD.

Supplementary Material

13361_2017_1643_MOESM1_ESM

Supplemental Figure 1. Chemical structures for mono-sulfated HS standards (1a–1f)

Supplemental Figure 2. Chemical structures for di-sulfated HS standards (2a–2h)

Supplemental Figure 3. Chemical structures for tri-sulfated HS standards (3a–3j)

Supplemental Figure 4. Chemical structures for tetra-sulfated HS standards (4a–4i)

Supplemental Table 1. Diagnostic ratio results for EDD diagnostic ratio results for 1a, for [M-2H]²⁻ precursor ion

Supplemental Table 2. Diagnostic ratio results for EDD fragmentation of [M-2H]²⁻ precursor ion for 1b

Supplemental Table 3. Diagnostic ratio results for EDD fragmentation of [M-2H]²⁻ precursor ion for 1c

Supplemental Table 4. Diagnostic ratio results for EDD fragmentation of [M-2H]²⁻ precursor ion for 1d

Supplemental Table 5. Diagnostic ratio results for EDD fragmentation of [M-2H]²⁻ precursor ion for 1e

Supplemental Table 6. Diagnostic ratio results for EDD fragmentation of [M-2H]²⁻ precursor ion for 1f

Supplemental Table 7. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2a

Supplemental Table 8. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2b

Supplemental Table 9. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2c

Supplemental Table 10. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2d

Supplemental Table 11. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2e

Supplemental Table 12. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2f

Supplemental Table 13. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2g

Supplemental Table 14. Diagnostic ratio results for EDD fragmentation of [M-3H]³⁻ precursor ion for 2h

Supplemental Table 15. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3a

Supplemental Table 16. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3b

Supplemental Table 17. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3c

Supplemental Table 18. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3d

Supplemental Table 19. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3e

Supplemental Table 20. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3f

Supplemental Table 21. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3g

Supplemental Table 22. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3h

Supplemental Table 23. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3i

Supplemental Table 24. Diagnostic ratio results for EDD fragmentation of [M-4H+Na]³⁻ precursor ion for 3j

Supplemental Table 25. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4a

Supplemental Table 26. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4b

Supplemental Table 27. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4c

Supplemental Table 28. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4d

Supplemental Table 29. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4e

Supplemental Table 30. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4f

Supplemental Table 31. Diagnostic ratio results for EDD fragmentation of [M-5H+2Na]³⁻ precursor ion for 4g

Supplemental Table 32. Diagnostic ratio results for EDD fragmentation of [M-5H+Na]⁴⁻ precursor ion for 4h

Supplemental Table 33. Diagnostic ratio results for EDD fragmentation of [M-5H+Na]⁴⁻ precursor ion for 4i Mass List for all tetrasaccharides are arranged according to product ion types for easy identification of diagnostic ions

Supplemental Table 34. Mass list for 1a, [M-2H]²⁻ precursor ion

Supplemental Table 35. Mass list for 1b, [M-2H]²⁻ precursor ion

Supplemental Table 36. Mass list for 1c, [M-2H]²⁻ precursor ion

Supplemental Table 37. Mass list for 1d, [M-2H]²⁻ precursor ion

Supplemental Table 38. Mass list for 1e, [M-2H]²⁻ precursor ion

Supplemental Table 39. Mass list for 1f, [M-2H]²⁻ precursor ion

Supplemental Table 40. Mass list for 2a, [M-3H]³⁻ precursor ion

Supplemental Table 41. Mass list for 2b, [M-3H]³⁻ precursor ion

Supplemental Table 42. Mass list for 2c, [M-3H]³⁻ precursor ion

Supplemental Table 43. Mass list for 2d, [M-3H]³⁻ precursor ion

Supplemental Table 44. Mass list for 2e, [M-3H]³⁻ precursor ion

Supplemental Table 45. Mass list for 2f, [M-3H]³⁻ precursor ion

Supplemental Table 46. Mass list for 2g, [M-3H]³⁻ precursor ion

Supplemental Table 47. Mass list for 2h, [M-3H]³⁻ precursor ion

Supplemental Table 48. Mass list for 3a, [M-4H+Na]³⁻ precursor ion

Supplemental Table 49. Mass list for 3b, [M-4H+Na]³⁻ precursor ion

Supplemental Table 50. Mass list for 3c, [M-4H+Na]³⁻ precursor ion

Supplemental Table 51. Mass list for 3d, [M-4H+Na]³⁻ precursor ion

Supplemental Table 52. Mass list for 3e, [M-4H+Na]³⁻ precursor ion

Supplemental Table 53. Mass list for 3f, [M-4H+Na]³⁻ precursor ion

Supplemental Table 54. Mass list for 3g, [M-4H+Na]³⁻ precursor ion

Supplemental Table 55. Mass list for 3h, [M-4H+Na]³⁻ precursor ion

Supplemental Table 56. Mass list for 3i, [M-4H+Na]³⁻ precursor ion

Supplemental Table 57. Mass list for sample 3j, [M-4H+Na]³⁻ precursor ion

Supplemental Table 58. Mass list for 4a, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 59. Mass list for 4b, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 60. Mass list for 4c, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 61. Mass list for 4d, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 62. Mass list for 4e, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 63. Mass list for 4f, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 64. Mass list for 4g, [M-5H+2Na]³⁻ precursor ion

Supplemental Table 65. Mass list for 4h, [M-5H+Na]⁴⁻ precursor ion

Supplemental Table 66. Mass list for 4i, [M-5H+Na]⁴⁻ precursor ion