Analysis of 3-O-Sulfated Heparan Sulfate Using Isotopically Labeled Oligosaccharide Calibrants

Abstract

The 3-O-sulfated glucosamine in heparan sulfate (HS) is a low-abundance structural component, but it is a key saccharide unit for the biological activities of HS. A method to determine the level of 3-O-sulfated HS is lacking. Here, we describe a LC–MS/MS based method to analyze the structural motifs. We determined the levels of 3-O-sulfated structural motifs from pharmaceutical heparin manufactured from bovine, porcine, and ovine. We discovered that saccharide chains carrying 3-O-sulfation from enoxaparin, an FDA-approved low-molecular weight heparin, displayed a slower clearance rate than non-3-O-sulfated sugar chains in a mouse model. Lastly, we detected the 3-O-sulfated HS from human brain. Furthermore, we found that a specific 3-O-sulfated structural motif, tetra-1, is elevated in the brain HS from Alzheimer’s disease patients (n = 5, p = 0.0020). Our method offers a practical solution to measure 3-O-sulfated HS from biological sources with the sensitivity and quantitative capability.

Graphical Abstract

Heparan sulfate (HS) is present on the cell surface and in the extracellular matrix, participating in a wide range of biological functions.¹ It contains a disaccharide repeating unit of glucuronic acid (GlcA) or iduronic acid (IdoA) and glucosamine (GlcN) residues. The individual saccharide units can carry multiple sulfations, including a 2-OH position of IdoA and/or NH₂ and 6-OH of GlcN. The sulfation at the 3-OH position of GlcN is a rare modification to form a subpopulation known as 3-O-sulfated HS that is closely related to the biological functions.² For example, the 3-O-sulfation is the key structural motif for the anticoagulant activity of HS and pharmaceutical heparins, drugs to treat clotting disorders.³ Furthermore, 3-O-sulfated HSs serve as an entry receptor for herpes simplex virus-1,⁴ regulate axon guidance and growth of neurons,⁵ as well as control of the progenitor cell expansion for salivary gland development.⁶ The distinct biological functions from 3-O-sulfated HS are attributed to different saccharide sequences around the 3-O-sulfated GlcN residue.⁷

Disaccharide compositional analysis using a liquid chromatography coupled with a mass spectrometer (LC–MS) is a widely used technique for HS analysis. The process involves digestion with heparin lyases to degrade the polysaccharides into disaccharides followed by a LC–MS or a LC coupled with tandem MS (LC–MS/MS) analysis (Figure 1).^{8,9 13}C-labeled disaccharide calibrants have improved the quantitation capability;^10,11 however, the disaccharide analysis is unsuited for the analysis of 3-O-sulfated HS. The majority of the 3-O-sulfated structural motifs are tetrasaccharides after the digestion with heparin lyases, with some exceptions.^12,13 Lack of chemically stable authentic 3-O-sulfated di/tetra-saccharides has hindered the effort to develop a method for analyzing 3-O-sulfated HS. In this manuscript, we sought to develop a LC–MS/MS method for the analysis of the level of 3-O-sulfated HS from biological sources using structurally homogeneous ¹³C-labeled HS oligosaccharide standards.

Figure 1.

Schematic structures of HS polysaccharide and ¹³C-labeled 8-mer calibrants. (A) Representative structure of HS polysaccharide. Cleavage sites of heparin lyases are indicated by a pair of scissors. Three 3-O-sulfated motifs are highlighted by a box. Representative of non-3-O-sulfated disaccharides are indicated. Eight non-3-O-sulfated disaccharides are found in HS, and their abbreviated structures are shown in Table 1. (B) Structures of three ¹³C-labeled 8-mer calibrants. The structural characterization and purity analysis of the calibrants was completed as described in Figures S1–S3. For clarity, shorthand symbols are used to represent the structure of each saccharide unit. Keys for the shorthand symbols are shown on the bottom of the figure.

EXPERIMENTAL SECTION

Materials.

An enoxaparin reference standard was purchased from SANDOZ. 2-Aminoacridone (AMAC) and sodium cyanoborohydride were purchased from Sigma-Aldrich (St. Louis, MO, USA). Three ¹³C-labeled 8-mer calibrants were synthesized using the chemoenzymatic synthesis method.¹⁰ Recombinant heparan lyase I, II, and III were expressed in E. coli and purified by a Ni-agarose column. DEAE-Sepharose Fast Flow resin was purchased from GE Healthcare (Chicago, IL, USA). All reagents and chemicals were high-performance LC (HPLC) grade or LC-mass spectrum (MS) grade.

Chemoenzymatic Synthesis of ¹³C-Labeled 8-mer Calibrants.

The structures of ¹³C-labeled oligosaccharides were appropriately designed, allowing us to obtain 3-O-sulfated disaccharide or tetrasaccharide calibrant targets carrying ¹³C-labeled glucuronic acid (GlcA) or iduronic acid (IdoA2S) after heparin lyases digestion. We intended to minimize the number of ¹³C-labeled oligosaccharides to prepare 3-O-sulfated calibrants. To these ends, we synthesized three different 8-mer, and each oligosaccharide yielded corresponding ¹³C-labeled 3-O-sulfated calibrants. All 8-mer calibrants were synthesized using the chemosynthetic method established in our lab. The synthesis of unlabeled 8-mer-1, -2, and -3 counterparts with NMR structural characterization has been previously reported.^14,15 In the synthesis presented in the current manuscript, we employed identical synthetic procedures described in previous reports. The only difference was to use UDP-[¹³C]GlcA to replace UDP-GlcA, which was also reported in the synthesis of ¹³C-labeled HS oligosaccharides.¹⁶

The synthesis of calibrants 1 to 3 started from GlcA-pNP. Four elongation steps were involved to obtain GlcA-GlcNTFA-GlcA-GlcNTFA-GlcA-pNP with heparosan synthase-2 (PmHS2) from Pasteurella multocida. The GlcA* was introduced at the designed position (where GlcA* indicates universally ¹³C-labeled GlcA residue). The elongation from GlcA-pNP to GlcNTFA-GlcA-pNP was implemented by incubating 50 mg GlcA-pNP, 0.5 mM UDP-NTFA, PmHS2 (290 μg/mL) in a buffer solution containing 25 mM MOPs (pH 7.0) and 15 mM MnCl₂ at 37 °C overnight. In the elongation of the disaccharide substrate (GlcNTFA-GlcA-pNP) to GlcA-GlcNTFA-GlcA-pNP, the disaccharide substrate was incubated with 0.4 mM UDP-GlcA, PmHS2 (290 μg/mL), 25 mM Tris (pH 7.5), and 15 mM MnCl₂ in a total volume of 400 mL at 37 °C overnight. The pentasaccharide backbone was completed through addition of GlcNTFA and GlcA residues one additional time. The reaction degree was monitored by strong anion-exchange chromatography on a Pro Pac PA1 column (9 × 250 mm, Thermo Fisher Scientific) by measuring the absorbance at 310 and 260 nm. And the purification was performed using a C18 column (0.75 × 20 cm; Biotage). The de-N-trifluoroacetylation of pentasaccharide backbone was implemented by suspending in 0.1 M LiOH and incubation on ice for 30 min. And then the pH was adjusted to neutral using HCl. The N-sulfation of pentasaccharide was followed by incubating with NST (30 μ/mL), 50 mM MOPS (pH 7.0), and 0.5 mM PAPS at 37 °C overnight. The N-sulfated pentasaccharide was purified using Q-Sepharose (GE Health Care). For the 8-mer calibrant 2, the GlcNAc was introduced; for the 8-mer calibrants 1 and 3, instead of GlcNAc, another GlcNTFA was introduced. This step was completed through incubating UDP-GlcNAC/UDP-GlcNTFA, PmHS2 (290 μg/mL) in a buffer containing 25 mM MOPs (pH 7.0) and 15 mM MnCl₂ at 37 °C overnight. For the 8-mer calibrants 1 and 3, another de-N-trifluoroacetylation and N-sulfation was performed to obtain the hexasacchride (GlcNS-GlcA-GlcNS-GlcA-GlcNS-GlcA-pNP). The hexasaccharide (GlcNS-GlcA-GlcNS-GlcA-GlcNS-GlcA-pNP/GlcNAc-GlcA-GlcNS-GlcA-GlcNS-GlcA-pNP) substrate was followed by epimerization and 2-O-sulfation. The substrate was incubated with C5-epimerase (C5-epi) (3 μg/mL), 2-O-sulfotransferase (2-OST) (6.5 μg/mL), 0.2 mM PAPS, and 50 mM MOPS (pH 7.0) at 37 °C overnight. The additional GlcA was introduced to the hexasaccharide substrate. For the 8-mer calibrants 2 and 3, the 6-O-sulfation was followed by incubating with 6-OST-3 (900 μg/mL) and 0.3 mM PAPS in a buffer containing 50 mM MOPS (pH 7.0) in a total volume of 500 mL at 37 °C overnight. And then another GlcNAc was further introduced. Instead, for the 8-mer calibrant 1, the GlcNTFA was first introduced, and then de-N-trifluoroacetylation/N-sulfation was performed before the 3-O-sulfation. After the 6-O-sulfation for 8-mer calibrants 2 and 3, 3-O-sulfation was performed by incubating with 50 mM MOPS (pH 7.0), 0.2 mM PAPs, and 3-O-sulfotransferase-1 (3OST-1) (0.15 mg/mL) at 37 °C overnight to obtain 8-mer calibrants 2 and 3. For the 8-mer calibrant 1, the 3-O-sulfation was performed before the 6-O- sulfation by incubating with 50 mM MOPS (pH 7.0), 3-O-sulfotransferase-3 (3OST-3) (0.11 mg/mL), and 0.2 mM PAPs at 37 °C overnight. Then the 6-O-sulfation was finished through incubating with 6-OST-3 (900 μg/mL) and 0.3 mM PAPS in a buffer containing 50 mM MOPS (pH 7.0) at 37 °C overnight to obtain 8-mer calibrant 1.The elongation, epimerization, and sulfation steps were involved to synthesize the ¹³C-labeled calibrators using the chemosynthetic approach. The recovery yield of each step was in the range of 87–95%. From 50 mg GlcA-pNP as the start material, we obtained the 120 mg 8-mer-1 calibrant, 107 mg 8-mer-2 calibrant, and 108 mg 8-mer-3 calibrant, respectively.

Comparison of 3-O-Sulfated Structural Motifs in the Heparin from Different Sources.

The heparin samples were exhaustively depolymerized using heparin lyases. Before the heparin lyases digestion, a known amount of ¹³C-labled 3-O-sulfated 8-mer calibrants were added to heparin samples. The enzymatic solution was a mixture of 2 μL heparin (10 mg/mL), ¹³C-labled 3-O-sulfated oligosaccharides, 100 μL of enzymatic buffer [100 mM sodium acetate/2 mM calcium acetate buffer (pH 7.0) containing 0.1 g/L BSA], and 6 μL of enzyme cocktails containing 5 mg/mL each of heparin lyase I, II, and III. The reaction mixture was incubated at 37 °C overnight. After digestion, the fixed amount of eight ¹³C-labeled non-3-O-sulfated disaccharide calibrants were added to the solution, and then the solution was boiled at 100 °C for 10 min. The solution was subjected to a centrifuge, and the supernatant was dried before the AMAC label was added and LC–MS/MS analysis was performed.

Analysis of the Elimination of Enoxaparin from a Mouse Model.

All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of University of North Carolina at Chapel Hill (Chapel Hill, NC). Enoxaparin (7.5 mg/kg) in ~100 μL sterile saline was subcutaneously administered to 10-week-old, male C57BL6/J mice. Mice were sacrificed at the indicated time points for blood collection from IVC.

The plasma samples were subjected to the HS analysis as described below. The recovery of enoxaparin from plasma was performed with methanol precipitation, DEAE column purification, and 2 K filter desalt and heparin lyases digest. Before methanol precipitation, 20 ng recovery calibrant ¹³C-labeled NS K5P (N-sulfo heparosan) and 500 ng ¹³C-labeled 8-mer-2 calibrant were added to the 100 μL mice plasma. The plasma was mixed with 900 μL methanol and allowed to precipitate at room temperature for 10 min. Then the solution was centrifuged at 1000 g for 10 min, and the supernatant was discarded. The precipitation then was subjected to Pronase E digest (10 mg:1 g (w/w), Pronase E/protein) at 55 °C for 24 h. After digestion, the enzymatic solution was boiled at 100 °C for 10 min to denature the Pronase E and centrifuged at 14,000 rpm for 10 min to obtain the supernatant. The DEAE column (150 μL) was used to recover the enoxaparin. DEAE column mobile phase A contained 20 mM Tris, pH 7.5, and 50 mM NaCl and mobile phase B contained 20 mM Tris, pH 7.5, and 1 M NaCl. After loading the sample into the DEAE column, the column was washed with 1.5 mL mobile phase A followed by 1.5 mL mobile phase B to elute the enoxaparin fraction. The eluted enoxaparin was desalted using an YM-2KDa spin device and the device washed three times with deionized water. The spin device was reversed to recover the retentate from the membrane and washed three times with deionized water. The combined solution was dried before the digestion with heparin lyases. 100 μL of enzymatic buffer [100 mM sodium acetate/2 mM calcium acetate buffer (pH 7.0) containing 0.1 g/L BSA] and 6 μL of enzyme cocktails containing 5 mg/mL each of heparin lyase I, II, and III was added to digest enoxaparin at 37 °C overnight. After digestion, the known amount ¹³C-labeled non-3-O-sulfated disaccharide calibrants were added to the digestion solution. The digestion solution was boiled at 100 °C for 10 min and centrifuged at 14,000 rpm for 10 min to recover the disaccharides. The supernatant was recovered and freeze-dried before the AMAC labeling step.

Labeling Disaccharides and Tetrasaccharides with AMAC.

The AMAC derivatization were carried out with 5 μL of 0.1 M AMAC in the solution in DMSO/glacial acetic acid (17:3, v/v) and incubated at room temperature for 15 min. Then 5 μL of 1 M aqueous sodium cyanoborohydride (freshly prepared) was added. The mixture was incubated at 45 °C for an additional 2 h and then centrifuged to obtain the supernatant for the LC-MS/MS analysis.

Brain Samples from Alzheimer’s Disease and Healthy Patients.

Frozen post-mortem frontal cortex samples were provided by the Neuropathology Core brain bank at the Goizueta Alzheimer’s Disease Center at Emory University. Tissues were provided from five Alzheimer’s disease (AD) patients (mean age 77.2 years) and five age-matched normal control subjects (mean age 77.0 years). The neuropathologic diagnosis of AD was established following current diagnostic criteria.¹⁷ Control subjects consisted of individuals with no known history of neurological disease and no significant neurodegenerative pathology at autopsy.

Extraction and Quantitation Analysis of HS from the Alzheimer’s Disease (AD) Human Brain Tissue.

Five brain tissues from normal control and five from AD cases were chosen to analyze the HS. The HS extraction from brain tissue was carried out with excision, homogenization, and defat as described previously.¹⁰ The defatted tissues were dried and weighed to obtain the dry weight. Then the dried tissue was subjected to Pronase E digest (10 mg:1 g (w/w), Pronase E/tissue) at 55 °C for 24 h to degrade the proteins. Twenty nanograms recovery calibrant ¹³C-labeled NS K5P was added into the digestion solution before subjecting to DEAE column purification. DEAE column mobile phase A consisted of 20 mM Tris, pH 7.5, and 50 mM NaCl, and mobile phase B consisted of 20 mM Tris, pH 7.5, and 1 M NaCl. After loading the digested solution, the column was washed with 1.5 mL mobile phase A, followed by 1.5 mL mobile phase B to elute the HS fraction. The YM-3KDa spin column was applied to desalt the elute, and the retentate was subjected to heparin lyases digest. Before the digestion, a known amount of ¹³C-labeled 8-mer calibrant 1-3 was added to the retentate. One-hundred microliters of enzymatic buffer (100 mM sodium acetate/2 mM calcium acetate buffer (pH 7.0) containing 0.1 g/L BSA) and 20 μL of enzyme cocktails containing 5 mg/mL each of heparin lyase I, II, and III were added to degrade the retentate on the filter unit of the YM-3KDa column. The reaction solution was incubated at 37 °C overnight. Before recovering the disaccharides from the digest solution, a known amount of ¹³C-labeled non-3-O-sulfated disaccharide calibrants were added to the digestion solution. The HS disaccharides and tetrasaccharides were recovered by centrifugation, and the filter unit was washed twice with 200 μL of deionized water. The collected filtrates were freeze-dried before the AMAC derivatization. The AMAC label and LC-MS/MS analysis of the collected disaccharides and tetrasaccharides of tissues was performed as described above. The amount of tissue HS was determined by comparing the peak area of native di/tetra-saccharide to each corresponding ¹³C-labeled internal standard, and the recovery yield was calculated based on a comparison of the amount of ¹³C-labeled NSK5P disaccharide in the tissue samples and control, respectively.

LC–MS/MS Analysis.

The analysis of AMAC-labeled di/tetra-saccharides was implemented on a Vanquish Flex UHPLC System (Thermo Fisher Scientific) coupled with TSQ Fortis triple-quadrupole mass spectrometry as the detector. The ACQUITY Glycan BEH Amide column (1.7 μm, 2.1 × 150 mm; Waters, Ireland, UK) was used to separate di/tetra-saccharides at 60 °C. Mobile phase A was 50 mM ammonium formate in water, pH 4.4. Mobile phase B is acetonitrile. The elution gradient as follows: 0-20 min 83-60% B, 20-25 min 5% B, and 25-30 min 83% B. The flow rate was 0.3 mL/min. Online triple-quadrupole mass spectrometry operating in the multiple reaction monitoring (MRM) mode was used as the detector. The MRM transition of each component is shown in Table S1. The ESI-MS analysis was operated in the negative-ion mode using the following parameters: negative ion spray voltage at 3.0 kV, sheath gas at 55 Arb, aux gas 25 arb, ion transfer tube temp at 250 °C, and vaporizer temp at 400 °C. TraceFinder software was applied for data processing.

RESULTS AND DISCUSSION

Three ¹³C-labeled 8-mer calibrants were synthesized using the chemoenzymatic approach.¹⁸ The calibrants contain a 3-O-sulfated glucosamine saccharide surrounded by different saccharide sequences (Figure 1B). Each calibrant yields a unique 3-O-sulfated structural motif matching the one present HS polysaccharide after the digestion with heparin lyases (Figure 1B). Both Tetra-1 and Tetra-2 have been isolated from pharmaceutical heparin,¹² and Di3S has been isolated from mouse embryonic carcinoma cells (P19 cells).¹⁹ The ¹³C-labeled Di3S, Tetra-1, and Tetra-2, served as tracers to follow the target analytes during LC analysis. Identification of the analytes from LC allows us to perform the multiple reaction mode (MRM) analysis to gain sensitivity. Furthermore, complete digestions of 8-mer calibrants with heparin lyases were achieved in the presence of HS and heparin. Therefore, the use of the ¹³C-labeled 8-mer calibrants also added the quantitation capability for Di3S, Tetra-1, and Tetra-2, motifs in HS. Construction of more than one ¹³C-labeled sites in 8-mer calibrants 1 and 2 was used to confirm the complete digestion during the analysis.

Compositional Analysis of Heparin from Porcine, Bovine, and Ovine.

The method was used to analyze pharmaceutical heparins from bovine, ovine, and porcine sources (Figure 2 and Table 1). To this end, three ¹³C-labeled 8-mer calibrants were mixed with the heparin samples and subjected to heparin lyases digestion. After the digestion, eight ¹³C-labeled non-3-O-sulfated disaccharides, which were prepared from a previous study,¹⁰ were added to the digestion mixture prior to the LC-MS/MS analysis. All 11 components were well-resolved by LC as shown in Figure 2, allowing us to measure the levels of both 3-O-sulfated motifs and non-3-O-sulfated disaccharides in one-pot analysis. Ovine and porcine heparin have similar compositions in non-3-O-sulfated disaccharides and 3-O-sulfated motifs. Bovine heparin has a higher level of one disaccharide, ΔIIIS (ΔUA2S-GlcNS), than that from porcine heparin, consistent with previous reports.²⁰ A lower level of Tetra-1 in bovine heparin than its porcine counterpart was observed, which is also consistent with a previous report (Table 1).²¹

Figure 2.

Compositional analysis of heparin from porcine, bovine, and ovine. (Top) Workflow chart for the analysis. MRM ion chromatograms of AMAC-labeled disaccharides/tetrasaccharides are shown on the right. CID-MS/MS spectra of each component are shown in Figures S4–S14. AMAC represents 2-aminoacridone. Bar graph shows the composition of eight non-3-O-sulfated disaccharides and three 3-O-sulfated structural motifs. ΔIA to ΔIVA and ΔIS to ΔIVS represent eight individual non-3-O-sulfated disaccharides (Table 1). Percentage values for each composition are shown in Table 1.

Table 1.

Mass Percentage (%) of Non-3-O-Sulfated Disaccharides and 3-O-Sulfated Motifs in the Bovine, Ovine, and Porcine Heparin

di/tetra-saccharides		mass percentage (%)
name	structures	bovine	ovine	porcine
ΔIVA	ΔUA-GlcNAc	7.59	2.43	3.92
ΔIIIA	ΔUA2S-GlcNAc	2.68	0.40	1.93
ΔIIA	ΔUA-GlcNAc6S	2.18	3.05	5.05
ΔIA	ΔUA2S-GlcNAc6S	0.21	0.53	1.31
ΔIVS	ΔUA-GlcNS	4.23	0.51	1.50
ΔIIIS	ΔUA2S-GlcNS	19.49	3.50	4.42
ΔIIS	ΔUA-GlcNS6S	4.27	3.15	4.21
ΔIS	ΔUA2S-GlcNS6S	54.48	77.85	69.84
Di3S	ΔUA2S-GlcNS3S6S	1.58	0.96	0.40
Tetra-1	ΔUA-GlcNAc6S-GlcA-GlcNS3S6S	1.39	5.00	5.98
Tetra-2	ΔUA-GlcNS6S-GlcA-GlcNS3S6S	1.91	2.64	1.44

As an anticoagulant drug, only porcine heparin is approved for the US market by the US Food and Drug Administration (FDA), while bovine heparin is allowed in other countries.

A quantitative analytical method covering the 3-O-sulfated motifs will provide structural markers to distinguish bovine and porcine heparins to safeguard the supplies.

Analysis of the Plasma Concentration of Enoxaparin from Mice.

Next, we employed our method to analyze the plasma concentration of enoxaparin after the subcutaneous administration in a mouse model. As an FDA-approved low-molecular weight heparin drug, enoxaparin is a mixture of polysaccharides containing both anticoagulant and nonanticoagulant subpopulations. Our method enables the analysis of both anticoagulant subpopulation and total enoxaparin from a single process. Measuring the level of a 3-O-sulfated motif, Tetra-1, was used for anticoagulant subpopulation. Tetra-1 is a structural marker associated with the anticoagulant activity of enoxaparin.¹² Measurement of a total amount of enoxaparin was achieved by determining the levels of non-3-O-sulfated disaccharides. The elimination curve shows that the peak concentration (C_max) is reached at 15 min, and the drug was near full elimination by 240 min (Figure 3A), as measured by the total amount. Comparing to the ratio of Tetra-1 with total non-3-O-sulfated disaccharides, we observed a clear escalating trend as time increases (Figure 3B), suggesting that a higher anticoagulant subpopulation was present at the later time points. Our data also suggest that the anticoagulant subpopulation displays slower clearance than the nonanticoagulant subpopulation in the mouse model. The standard of care for patients to determine the plasma level of enoxaparin is based on antifactor Xa activity, which only measures the anticoagulant subpopulation.²² Our LC-MS/MS based method directly measures the saccharide markers covering both anticoagulant and nonanticoagulant subpopulations, allowing physicians to control the enoxaparin dose accurately.

Figure 3.

Analysis of the plasma concentration of enoxaparin from mice. (Top) Murine model experimental design schematic including time, dose, and route of administration of enoxaparin. Groups received enoxaparin were sacrificed at different time points as indicated. (A) Plasma concentration of enoxaparin as analyzed by the LC–MS/MS method. (B) Ratio of Tetra-1 and total disaccharides at different time points. Total disaccharide was the sum of eight non-3-O-sulfated disaccharides. Data are presented as mean ± SEM (n = 3). p value was determined by two-tailed unpaired t test, *p < 0.05.

Analysis of HS from Alzheimer’s Disease Brain (AD) and HS from Healthy Brain (CONT).

In the last example, we determined the level of 3-O-sulfated HS from brain tissues of Alzheimer’s disease (AD) patients. HS isolated from the brain of both healthy individuals (control group, n = 5) and AD patients (n = 5) were subjected to the analysis (Figure 4A and Table 2). The total amount of HS was measured to be 452.9 ± 96.6 ng/mg from AD group and 244.6 ± 57.9 ng/mg from the control group, an 85.1% increase (p = 0.0060) in the AD group. The percentage composition of Tetra-1 in the HS from the AD group was measured to be 1.31 ± 0.29% compared with 0.53 ± 0.18% of the control group, a 2.5-fold increase (p = 0.0020) (Figure 4D). Although statistical differences among five non-3-O-sulfated disaccharides were found, the relative percentage increase compared to the control group was small (Figure 4B,C and Table 2).

Figure 4.

Analysis of HS from Alzheimer’s disease brain (AD) and HS from healthy brain (CONT). (A) Total HS amount in from brain tissues. The total amount of HS was obtained by summing up each individual non-3-O-sulfated disaccharides and three 3-O-sulfated structural motifs. Data are presented as the nanograms of HS per dry weight of brain tissue. (B,C) Composition of eight individual non-3-O-sulfated disaccharides. (D) Percentage composition of the 3-O-sulfated motifs. Data are presented as mean ± SEM (n = 5). p value was determined by two-tailed unpaired t test, and the value was indicated. ns represents not statistically significant.

Table 2.

Mass percentage (%) of Disaccharides and 3-O-Sulfated Motifs and Total HS Amount in the Human Brain Tissue (Control Group vs AD’s Group, n = 5)^a

		mass percentage (%)
di/tetra-saccharides		control samples					AD’s samples
name	structures	C-0502	C-0503	C-E08	C-E13	C-E20	AD E05	AD E10	AD E14	AD E16	AD E19
ΔIVA	ΔUA-GlcNAc	69.83	67.03	64.62	65.42	64.49	59.66	59.63	58.76	62.84	66.04
ΔIIIA	ΔUA2S-GlcNAc	0.40	0.25	0.62	n.d.	0.19	0.87	0.57	0.64	0.64	0.64
ΔIIA	ΔUA-GlcNAc6S	8.38	10.82	12.45	11.23	13.57	13.97	15.29	15.09	12.58	12.27
ΔIA	ΔUA2S-GlcNAc6S	0.03	0.06	0.07	0.06	0.06	0.08	0.08	0.11	0.09	0.06
ΔIVS	ΔUA-GlcNS	9.55	7.32	7.38	7.06	6.83	8.49	8.11	7.17	6.49	7.52
ΔIIIS	ΔUA2S-GlcNS	4.85	5.11	5.59	5.46	4.88	6.03	5.99	5.39	5.50	5.75
ΔIIS	ΔUA-GlcNS6S	2.48	3.67	3.80	3.78	4.41	4.34	4.43	4.83	4.00	3.36
ΔIS	ΔUA2S-GlcNS6S	3.26	3.92	3.63	5.15	2.47	3.89	3.40	5.10	5.39	2.52
DI3S	ΔUA2S-GlcNS3S6S	0.19	0.24	0.22	0.25	0.36	0.24	0.21	0.25	0.26	0.17
Tetra-1	ΔUA-GlcNAc6S-GlcA-GlcNS3S6S	0.33	0.60	0.39	0.49	0.84	1.51	1.36	1.69	1.12	0.85
Tetra-2	ΔUA-GlcNS6S-GlcA-GlcNS3S6S	0.69	1.00	1.22	1.12	1.92	0.91	0.95	0.97	1.10	0.82
total amount of HS (ng/mg)		320.9	296.1	238.1	204.7	163.2	491.6	456.4	539.5	267.4	509.4

^an.d. indicates undetected.

3-O-sulfated HS has been implicated in the pathology of AD. One report shows significant increases in the mRNA levels of HS 3-O-sulfotransferase isoforms 2, 3, and 4 in the hippocampus from AD, enzymes involved in the biosynthesis of the 3-O-sulfated structural motifs.²³ The 3-O-sulfated HS was reportedly internalized into cells to induce abnormal phosphorylation of tau protein²³ as well as to assist the uptake of tau protein.²⁴ However, the 3-O-sulfated motifs have not been identified from the HS isolated from brain tissue so far. Here, we demonstrated, for the first time, that at least three 3-O-sulfated motifs are present in the brain. Tetra-1 is substantially elevated in the AD patients. It should be noted that Tetra-1 is perceived as the product from 3-O-sulfotransferase isoform 1 modification, not by isoforms 2, 3, and 4.⁷ A plausible explanation for this observation is that elevated mRNA levels for the isoforms are found in the hippocampus from AD patients as shown in a previous report.²³ In the current study, we performed the analysis of HS from the frontal cortex of AD patients.

CONCLUSIONS

Here, we describe a practical method to quantify low-abundance 3-O-sulfated HS from biological sources. The success is attributed to the availability of synthesized ¹³C-labeled 8-mer calibrants that contain the 3-O-sulfated structural motifs similar to those present in HS upon digestion by heparin lyases. The design and use of the ¹³C-labeled standards played a critical role in developing the method. The ¹³C-labeled 8-mer calibrants are stable, and the synthesis can be scaled up, suited for analyzing large numbers of samples in different laboratories. One alternative approach could be to prepare ¹³C-labeled Di3S, Tetra-1, and Tetra-2 standards prior to performing the analysis. However, this approach failed as Di3S, Tetra-1, and Tetra-2 were only stable transiently,¹³ ruling out the possibility of using these as standards for the quantitative analysis. To circumvent the impact of chemical instability of 3-O-sulfated motifs on the measurement accuracy, we chose to incubate ¹³C-labeled 8-mer calibrants and the HS sample together prior to the heparin lyase digestion. Under this condition, both 8-mer calibrants and HS are completely digested to yield 3-O-sulfated structural motifs. The impact from the undesired degradation is canceled out because ¹³C-labeled motifs and unlabeled counterparts should degrade at a similar rate. Understanding the number of 3-O-sulfated saccharide sequences in biologically sourced HS is needed to further improve this method. Construction of a comprehensive library of ¹³C-labeled 3-O-sulfated oligosaccharides to cover additional 3-O-sulfated structural motifs is underway. The implementation of this sensitive and quantitative LC-MS/MS method will add a new tool to assess the quality and purity of the heparin supply chain and improve the analysis plasma concentration of enoxaparin in hospital clinical laboratories. The method will also help to dissect the roles of specific 3-O-sulfated HS in the progression of AD disease and other biological processes.