Quantitative Analysis of Oligosaccharides Derived from Sulfated

Quantitative Analysis of Oligosaccharides Derived from Sulfated ...pubs.acs.org/doi/full/10.1021/ac3034097?src=recsys&mobileUi=0Mucopolysaccharidoses ...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Quantitative Analysis of Oligosaccharides Derived from Sulfated Glycosaminoglycans by Nanodiamond-Based Affinity Purification and Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry Chih-Chien Hsieh,†,∥ Jiun You Guo,‡ Shain-Un Hung,‡ Rui Chen,§ Zongxiu Nie,§ Huan-Cheng Chang,†,∥ and Chih-Che Wu*,‡ †

Department of Chemistry, National Taiwan Normal University, Taipei 106, Taiwan Department of Applied Chemistry, National Chi Nan University, Puli, Nantou 545, Taiwan § Key Laboratory of Analytical Chemistry for Living Biosystems, Institute of Chemistry, The Chinese Academy of Sciences, Beijing 100190, China and Beijing National Laboratory for Molecular Sciences, Beijing 100190, China ∥ Institute of Atomic and Molecular Sciences, Academia Sinica, Taipei 106, Taiwan ‡

S Supporting Information *

ABSTRACT: Degraded fragments of sulfated glycosaminoglycans (GAGs) are key reporters for profiling the burden of mucopolysaccharidosis (MPS) disease at baseline and during therapy. Here, we present a high-throughput assay, which combines microwave-assisted degradation, solid-phase affinity purification, and matrix-assisted laser desorption/ionization mass spectrometry (MALDI MS), for quantitative analysis of sulfated oligosaccharides in biological samples. First, sulfated oligosaccharides such as chondroitin-4-sulfate (CS) were efficiently isolated from highly diluted solutions or spiked artificial cerebrospinal fluid (aCSF) using polyarginine-coated nanodiamonds (PA-coated NDs) as affinity sorbents. Next, they were degraded to disaccharides through microwave-assisted methanolysis or enzymatic digestion for subsequent MALDI-TOF MS analysis. The reaction times for GAG depolymerization were significantly reduced from a few hours to less than 7 min under the microwave irradiation. Deuterium-labeled internal standards were then mixed with the CS-derived disaccharides for quantitative analysis by MALDI-TOF MS using the N-(1-naphthyl) ethylenediamine dihydrochloride (NEDC) matrix. The new assay is facile, specific (with distinct chlorine-isotope trait markers), sensitive (with a detection limit of ∼70 pg), and potentially useful for clinical diagnosis of MPS.

M

analysis,6,7 and mass spectrometric analysis.8 Although dried blood spot (DBS)-based assay9 is currently the most widely used methodology to assess the GAG elevation, improper sample collection and handing can often result in false-positive and false-negative identification.10 Clinical diagnosis, therefore, must be combined with multiple laboratory tests to identify the disease subtype,11 which complicates the procedures and makes the analysis laborious and time-consuming. There is a need to develop a high-throughput method for facile and accurate clinical diagnosis of MPS.12 The application of liquid chromatography−tandem mass spectrometry (LC−MS/MS) to the clinical diagnosis of MPS patients, especially in the newborn screening programs, has been proven to be a significant technological advance owing to its high specificity, sensitivity, and data reproducibility.3 A wide range of sulfated oligosaccharides, derived from the degradation

ucopolysaccharidoses (MPS) are a group of lysosomal storage disorder diseases caused by inherited deficiencies of specific enzymes to degrade glycosaminoglycans (GAG).1 To date, 11 enzyme deficiencies leading to accumulation of various types of GAGs in tissues, blood, urine, and cerebral spinal fluid have been identified. These GAGs include chondroitin sulfate (CS), dermatan sulfate (DS), heparan sulfate (HS), hyaluronan (HA), keratan sulfate (KS), or their combinations. Early and accurate identification of the MPS phenotype is crucial for prompt intervention of the disorder in a primary stage with disease-specific treatment.2 For example, early treatment with hematopoietic stem cell transplantation and/or enzyme replacement therapy can improve patients’ long-term outcomes and enhance their quality of life. MPS patients usually excrete excess amounts of GAGs as compared to age-matched normal controls. Quantitative analysis of GAGs in MPS patients has always been difficult because of their high molecular complexity and heterogeneity.3 At present, various methodologies have been developed and applied to analyze GAGs, ranging from spot testing,4 colorimetric test,5 nuclear magnetic resonance spectroscopic © 2013 American Chemical Society

Received: November 24, 2012 Accepted: March 27, 2013 Published: March 27, 2013 4342

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

of GAG, has been reported as specific MPS markers for LC− MS/MS analysis.13 Despite the advances, high-throughput screening of GAG excretion by LC−MS/MS remains a formidable challenge because of their high negative-charge density, polydispersity, and heterogeneity. A combination of techniques including controlled enzymatic or chemical depolymerization, chromatographic separation, and MS detection is thus indispensable for successful identification and quantification of GAGs.14 Although enzymatic depolymerization offers the advantage of high specificity, chemical depolymerization using reagents with high chemoselectivity is advantageous in terms of speed and ease of implementation. For instance, methanolysis of GAGs allows rapid determination of GAG composition by cleaving glycosidic bonds to produce O-methyl glycosides of monosaccharides and/or of disaccharides.15 This methodology has been successfully applied for quantitative analysis of GAG excretion in the urine3 and the cerebrospinal fluid16 of MPS patients by LC−MS/MS. However, in these experiments, it takes quite a few hours to complete the sample preparation and subsequent LC−MS/MS analysis of each sample.14 It is therefore important to develop a high-throughput screening method for specific detection and quantitative determination of GAG abundance in these biological fluids. We have previously demonstrated that polyarginine-coated nanodiamonds (PA-coated NDs) are useful for selective extraction and enrichment of multiphosphorylated peptides and sulfonated peptides.17,18 It has been shown that guanidinium of arginine residues have an extremely high affinity toward phosphate and sulfate residues.19 Such a high affinity has been observed for specific binding between sulfonate groups and arginine residues in large biomolecules such as heparin and heparin sulfate.20 The pristine NDs are readily functionalized with PA to produce a polycationic polymer surface. The extraordinarily strong noncovalent interactions between PA-coated NDs and polyanionic analytes greatly facilitate the prefractionation of highly phosphorylated or sulfonated peptides prior to MS analysis.17,18 Similar to phosphorylated and sulfonated peptides, sulfated GAGs also possess complementary binding characteristics to electropositive amino acid residues including arginine, histidine, and lysine.19 Particularly for arginine, two salt bridges between quanidinium and sulfonate groups can form in a fork-like configuration, a trait absent in histidine and lysine. Furthermore, the binding affinity of sulfated GAGs to the PA-coated NDs is greatly enhanced by the formation of multiple quanidinium−sulfonate interactions. The PA-coated NDs are expected to be useful in this application to rapidly and efficiently isolate CS from a small volume (25 μL) of highly diluted sample solutions for subsequent microwave-assisted methanolysis and characterization by matrix-assisted laser desorption/ionization-time-of-flight mass spectrometry (MALDI-TOF MS). We present in this work a new method for facile isolation, characterization, and quantification of GAGs using polyarginine-based affinity purification, microwave-assisted GAGs degradation, and MALDI MS. A summary workflow of the analysis is presented in Scheme 1. We validated the method by using chondroitin sulfate (CS) and applied it to the analysis of CS extracted from a small volume of highly diluted solution or spiked artificial cerebrospinal fluid (aCSF). Deuterium-labeled methanolysates of CS were externally spiked as an internal standard for MALDI MS quantification. The method is

Scheme 1. Schematic Procedures for Identification and Quantification of Glycosaminoglycans Using NanodiamondBased Affinity Purification and MALDI-MS

expected to be useful for high-throughput screening of MPS disorders and as a means for rapid monitoring GAG excretion of patients upon clinical treatment.

■

EXPERIMENTAL SECTION Materials and Chemicals. Acetyl chloride, bovine serum albumin (BSA), chondroitin-4-sulfate (CS), chondroitinase ABC, dermatan sulfate (DS), 2,5-dihydroxybenzoic acid (DHB), 3 N methanolic HCl, heparin (HP), heparan sulfate (HS), methanol, tetradeuteromethanol (CD3OD), and N-(1naphthyl) ethylenediamine dihydrochloride (NEDC) were from Sigma Aldrich (St. Louis, MO). 3-Aminoquinoline (3AQ) was from Acros (Geel, Belgium), and acetonitrile (ACN) was from Merck (Darmstadt, Germany). Deionized water was obtained from a Simplicity Millipore system (Burlington, MA). Isolation and Extraction of CS by PA-Coated NDs. PAcoated NDs were prepared according to our previously reported procedures17 and described in details in the Supporting Information. For isolation of CS from highly diluted solution or artificial cerebrospinal fluid, the sample solution (25 μL) was diluted with a 2-fold volume of water and mixed with 20 μL of PA-coated ND suspension (1 mg/mL) for 15 min with gentle vortexing. After separation by centrifugation at 15 000 rpm for 5 min and careful removal of the supernatant by pipetting, the slurry was evaporated to dryness under a steam of nitrogen for subsequent on-probe degradation by methanolysis or by chondroitinase ABC enzyme. Microwave-Assisted Methanolysis of CS. For CS degradation by methanolysis, 200 μL of 3 N methanolic HCl solution was mixed with the particle slurry in a glass tube and heated at 65 °C in a microwave reactor for 7 min. After separation by centrifugation at 15 000 rpm for 5 min, the glycan-containing supernatant was transferred to a new tube, evaporated to dryness under a steam of nitrogen, and mixed with 1 μL of NEDC matrix solution (1 mM) for MALDI-TOF MS analysis.21 An isotope-labeled internal standard was used for CS quantification by MALDI-TOF MS. The internal standard was prepared by deuterio-methanolysis of CS. The CS internal standard (200 μg) was incubated with 200 μL of freshly prepared deuterio-methanolysis reagent for 75 min at 65 °C. 4343

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

Scheme 2. (a) Chemical Degradation of Chondroitin-4-Sulfate (CS) by Methanolysis and (b) Detected Product Ions of the Methanolysates of CS by Negative Ion MALDI-TOF MS Using the NEDC Matrix

methanolysis which cleaves the O-glycosidic bonds and produces desulfated disaccharides for MS analysis.16 After methanolysis, we obtained methylated oligosaccharides as degradation products for subsequent MS analysis. Recently, we reported a new organic matrix, NEDC, for sensitive detection of small molecules such as glucose under high salt concentration conditions.21 The ability of NEDC to promote chloride attachment of neutral analytes to form [M + Cl]− in negative ion MALDI MS significantly improved the detection sensitivity of small oligosaccharides. Figure 1 presents MALDI mass spectra of the methanolysis products of CS obtained by using NEDC, 3-AQ, and DHB as matrixes. Chondroitin-4sulfate has a structural domain consisting of segregated blocks of repeating GlcA-GalNAc(4S) disaccharides (N4S domains).

The deuterio-methanolysis reagent was prepared by dropwise addition of acetyl chloride solution (1 mL) to CD3OD (3 mL) in an ice bath under dry N2 conditions. Scheme 2 presents the chemical structure of the CS disaccharide derived from methanolysis of chondroitin-4-sulfate. Microwave-Assisted Enzymatic Degradation of CS. For CS degradation by chondroitinase ABC enzyme, 50 μL of digestion buffer containing 50 mM Tris (pH 8.0), 60 mM sodium acetate, and 0.02% BSA and 1 μL of the Chondroitinase ABC enzyme solution (5 mU/μL) were mixed with the particle slurry in a gas tube and heated at 45 °C in a microwave reactor for 5 min. After separation by centrifugation at 15 000 rpm for 15 min and careful removal of the supernatant by pipetting, the slurry was mixed with 1 μL of 3-AQ matrix solution (30 mg/ mL) for MALDI-TOF MS analysis. MALDI-TOF Mass Spectrometry and Data Analysis. MALDI-TOF MS was performed on a Bruker Microflex timeof-flight mass spectrometer (Bruker Daltonics, Bremen, Germany). All spectra were obtained in negative reflector ion mode and at an extraction voltage of 19 kV. The laser power was adjusted to a value slightly above the desorption/ionization threshold, and each mass spectrum was obtained by averaging 300 laser shots scanned across the sample surface. The spectra and the detected peak signal-to-noise (S/N) ratios were processed using the FlexAnalysis software (Bruker Daltonics).

■

RESULTS AND DISCUSSION Enhanced MALDI MS Detection Using the NEDC Matrix. MALDI MS detection of sulfated GAGs such as chondroitin sulfate (CS) is often hampered by their low ionization efficiency, high molecular complexity, and structural heterogeneity. A method to enhance the MS detection is to perform chemical depolymerization of the molecules through

Figure 1. Negative ion MALDI-TOF mass spectra of the methanolysis products of CS (5 μg) obtained using (a) NEDC, (b) 3-AQ, and (c) DHB as a matrix. Asterisks denoted the methanolysis products of CS. 4344

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

Figure 2. (a) Negative ion MALDI-TOF mass spectra of the methanolysis products of CS (25 ng) obtained at different exposure time of microwave irradiation and (b) relative peak intensity of the dimer versus the monomer for the methanolysates of CS obtained at different microwave exposure times.

intense with the 7 min setting. The experiments were performed using the optimized microwave heating time of 7 min. A back-to-back comparison of microwave and conventional heating for methanolysis of CS is shown in Figure S3 in the Supporting Information. Compared to conventional heat treatment, the microwave irradiation not only significantly reduces the reaction time of CS methanolysis from 75 to 7 min but also increases the S/N of the CS disaccharide ions by 7.3fold. The results indicate that microwave-assisted methanolysis is a rapid and efficient approach for facile degradation of GAGs prior to subsequent MS analysis. Extraction of Sulfated GAGs by PA-Coated NDs. Extraction and concentration of GAGs by PA-coated NDs prior to methonolysis and subsequent MALDI MS analysis was illustrated using CS as a model system. Figure 3 shows the experimental evaluation of the methanolysis efficiency of CS, obtained after enrichment using PA-coated NDs, at different substrate/reagent ratios. For 200 μL of a 3 N methanolic HCl solution, the optimum amount of CS for microwave-assisted on-probe methanolysis was found to be less than 250 ng of CS per 20 μg of PA-coated NDs. The primary product from the

The GAG polymer chains were desulfated and cleaved during methanolysis. The primary products of the methanolysis of CS were presumed to be the methylated disaccharides (cf. Scheme 2 for the structures). The dominant peak at m/z 460.1 corresponds to the chloride-attached ions of CS disaccharides, dimethylated at both the carboxylic acid and terminal hemiacetal function.3 The intense peak at m/z 410.1 was attributed to the deprotonated adduct of the monomethylated CS disaccharide ion due to incomplete methanolysis of CS. Compared to the mass spectra obtained using conventional matrixes such as DHB and 3-AQ, the spectrum is much cleaner and nearly free of matrix interference in the low mass region. Moreover, the detection sensitivity with NEDC is superior to that achieved with the other two matrixes. Sensitivity detection of the CS disaccharide peak at m/z 460.1 showed a signal-tonoise (S/N) ratio of 49 for 1.3 ng of CS methanolysates (Figures S1 and S2 in the Supporting Information). The detection limit is as low as 70 pg with S/N = 3. To the best our knowledge, this is the lowest detection limit ever reported for CS-derived disaccharides by MALDI-TOF MS. Microwave-Assisted Methanolysis of CS. A critical step for sensitive detection of GAGs by mass spectrometry is the chemical depolymerization, methanolysis. This chemical degradation typically carries out for a few hours using conventional heating methods.16 Microwave irradiation has emerged as a power technique to promote a variety of chemical reactions. Microwave heating can significantly decrease reaction times, increase yields, reduce side reactions, and improve data reproducibility. A previous study has shown that the CS disaccharide concentration in the methanolysates reached a maximum after a 75-min heating at 65 °C.16 Choosing the same operating temperature, the methanolysis efficiency under microwave irradiation was optimized by measuring the MALDI MS peak intensity ratio of CS disaccharides versus monosaccharides. Figure 2 shows that the dimer/monomer ratio reached its maximum at 7 min after that the monosaccharide ion peak became more dominant in the mass spectra. Compared to the value of the dimer/momer ratio derived from 7 min microwave irradiation time, most of the values in Figure 2b are below the 1% significance level except the value derived from 11 min microwave irradiation time (5% significance level). Furthermore, the mass spectra in Figure 2a show that peak intensity of CS-derived disaccharide is the most

Figure 3. Negative ion MALDI-TOF mass spectra of microwaveassisted methanolysis products of CS, obtained with affinity enrichment using 20 μg of PA-coated NDs. The amounts of CS samples used in the measurements are (a) 250 ng, (b) 50 ng, (c) 25 ng, and (d) 1.3 ng, respectively. Asterisks denoted the methanolysis products of CS. 4345

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

Figure 4. (a) Negative ion MALDI-TOF mass spectra of different amounts of methanolysis products of CS (denoted by asterisks), mixed with 50 ng of deuterio-methanolysis products of CS (peaks denoted by #). (b) A calibration curve for quantification of CS methanolysates using the disaccharide peak intensity ratios of CS and deuterated CS (CS-D).

concentrations of CS spiked aqueous solutions were sequentially enriched using PA-coated NDs (20 μg per extraction) and analyzed by microwave-assisted methanolysis and MALDI-TOF MS (Figure S8 in the Supporting Information). After sequential extraction using PA-coated NDs, the ion peak intensity of CS disaccharide gradually decreased as the number of extractions increased. Analysis of relative peak intensity ratio of CS disaccharide in the first extraction and in all the extractions revealed the effective binding capacity of 20 μg of PA-coated NDs were 471.1 ± 13.3 ng of CS. The mass ratio of the GAGs to the PA-coated NDs was thus suggested to keep a level of 1:50 to ensure sufficient GAG extraction. The extraction efficiency of CS from aqueous solution and aCSF using PAcoated NDs was estimated from the intensity ratios of CS disaccharide peaks in the mass spectra obtained using the same laser fluence (Figure S9 in the Supporting Information). The percent recovery of CS in spiked aqueous solution and aCSF was determined to be 91.6% and 76.4%, respectively. The declined extraction efficiency in aCSF was possibly due to the high salt concentration interfering with the cationic−anionic binding interaction between the PA-coated NDs and CS. However, the high-salt interference could be reduced by additional dilution steps prior to analysis. For quantitative analysis of GAGs in biological samples, additional sample dilution is recommended to minimize the effect of interference and to improve the recovery efficiency. The results demonstrate that PA-coated NDs can be used as an affinity extraction-based platform for rapid and efficient analysis of sulfated GAGs by MALDI MS. CS Quantification by MALDI-TOF MS. To perform CS quantification by MALDI-TOF MS, the deuterio-methanolysis products of CS were externally spiked into the sample matrix as an internal standard. Quantification of GAGs by MALDI-TOF MS can be an effective alternative for rapid and highthroughput screening of MPS disease or monitoring patients’ responses during clinical treatment. Figure 4a shows negative ion MALDI-TOF mass spectra of the methanolysis products of CS externally spiked with 50 ng of the deuterio-methanolysis products of CS as internal standards. Analysis of the peak intensity ratio of the CS disaccharide at m/z 460.1 relative to the deuterated CS disaccharide at m/z 466.1 and at m/z 413.1

methanolysis of CS was attributed to be GlcA-GalNAc disaccharides, dimethylated at both the carboxylic acid and terminal hemiacetal function. The dominant peak at m/z 460.1 corresponds to the chloride-attached molecular ions of dimethylated CS disaccharides. This predominant feature is a discernible molecular marker because of the distinctive 3:1 isotope pattern of chlorine atom (35Cl and 37Cl). When the amount of CS was too high, an additional peak appears at m/z 410.1, corresponding to the deprotonated adduct of the monomethylated CS disaccharide due to incomplete methylation of CS disaccharides and/or in-source ion fragmentation. The dimeric structures were confirmed by the tandem mass spectrometry experiments using Fourier transform ion cyclotron resonance mass spectrometer (FTICR MS) (Figure S5 in the Supporting Information). Collision-induced dissociation (CID) of isolated parent ions at m/z 460.1 yielded a prominent fragment ion at m/z 410.1. Further evidence was derived from the mass spectra of the products of deuterio-methanolysis of CS (Figure S6 in the Supporting Information). Figure S6 in the Supporting Information shows the negative mode electrospray ionization (ESI) mass spectra of deuterio-methanolysis products of CS. The spectra showed a shift of +6 Da for the chloride-attached dimethylated CS disaccharide ions and a shift of +3 Da for the deprotonated monomethylated CS disaccharide ions, corresponding to the deuterated analogues of methylated CS disaccharide ions. The correlation of these two dimeric structures was again confirmed by the CID experiment as shown in Figure S6b in the Supporting Information. These dimeric structures are derived from the depolymerization of CS chains that become desulfated and cleaved during methanolysis and are structurally related to those produced by enzymatic depolymerization.16 With the capability of preconcentration and isolation using PA-coated NDs as a solid-phase extraction support, the CS can be detected at a concentration level as low as 0.5 mg/L using MALDI MS (Figure S4 in the Supporting Information). A control experiment for CS extraction using pristine NDs has also been performed, but no ion signals for CS disaccharides can be found in the spectra, due to the poor affinity of NDs toward oligosaccharides. To evaluate the effective CS-binding capacity of PA-coated NDs, high 4346

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

revealed a linear relationship over the weight range of 2.5−25 ng (Figure 4b). The R2 value obtained for the plot is 0.9993. Relative standard derivation (RSD) varied from 4.2% to 18.1% over this range. It is instructive to compare the presently developed method with a previously reported MALDI-TOF MS approach based on the S/N ratio of the CS disaccharide ion peaks for quantitative analysis.22 In that study, the CS disaccharides were obtained by enzymatic digestion with chondroitin ABC. However, no apparent linear relationship between the S/N ratios and the amounts of CS on the MALDI sample plate was obtained. This is attributed to the inevitable laser power fluctuation and inhomogeneous analyte/matrix cocrystallization, which often yield peaks with poorly reproducible intensities. Clearly, the MS quantification using externally spiked isotope standards derived from deuterio-methanolysis of CS as the references offers much greater data reproducibility. Compared to other GAG quantification methods such as dimethylmethylene blue (DMB) assay or carbazole assay, the sensitivity of this new approach (70 pg) is significantly higher, roughly by more than 1 to 3 orders of magnitude (few hundred nanograms for the DMB assay5,23 and a few micrograms for the carbazole assay24). Qualitative Analysis of the Complex Mixture of GAGs. Owing to the high resolution and mass accuracy afforded by MALDI MS, an unambiguous link between masses and the glycan composition can be beneficial for analysis of complex glycosaminoglycan polysaccharide. Chondroitin-4-sulfate (CS) has a domain structure consisting of repeating GlcA-GalNAc(4S) disaccharides. Some GlcA residues are epimerized into Liduronic acid (IdoA); the resulting disaccharide is then referred to as dermatan sulfate. Dermatan sulfate (DS) has a similar structural domain consisting of repeating IdoA-GalNAc(4S) disaccharides. The CS/DS chains are polydisperse with respect to chain length, the degree and position of sulfation, and IdoA/ GlcA epimerization content. Heparin (HP) is a highly sulfated glycosaminoglycan that consists of repeating IdoA(2S)-GlcNS(6S) disaccharides and is closely related in structure to heparan sulfate (HS). Heparan sulfate (HS) has multiple domain structures with a low-sulfated repeating disaccharide region (GlcA-GlcNAc disaccharide, NA domain) and high-sulfated repeating disaccharide regions (IdoA-GlcNS disaccharide, NS domain). In HS, GlcN can be either N-sulfated or N-acetylated, whereas in HP the N-sulfated region is the major component. Figure 5 shows the direct negative mode MALDI MS analysis of methanolysis products of chondroitin sulfate (CS), dermatan sulfate (DS), heparin (HP), and heparan sulfate (HS). The methylated disaccharide derived from the methanolysis of DS behaves similarly to that derived from CS. The mass spectra of methanolysis products of CS and DS all showed a predominant peak at m/z 460.1 with a Cl isotope pattern, corresponding to chloride-attached dimethylated hexuronic acid-N-acetyl-galatctosamine (UA-GalNAc) disaccharide. The methanolysis product of HP yielded another prominent ion peak at m/z 418.1, corresponding to the chloride-attached dimethylated iduronic acid-glucosamine (IdoA-GlcN) disaccharide. For MS analysis of methanolysis product of HS, two intense peaks at m/z 460.1 and m/z 418.1 were found in spectrum (Figure 5d). These two ion peaks are attributed to the chloride-attached dimethylated disaccharide ions derived from CS-like disaccharide residues (NS domains) and HP-like disaccharide residues (NA domains) of HS.

Figure 5. Negative ion MALDI-TOF mass spectra of 250 ng of methanolysis products of (a) chondroitin sulfate (CS), (b) dermatan sulfate (DS), (c) heparin (HP), and (d) heparan sulfate (HS). The peaks at m/z 460.1 and m/z 418.1 in part d are attributed to the chloride-attached dimethylated disaccharide ions derived from CS-like residues (GlcA-GlcNAc disaccharide) and HP-like residues (IdoAGlcN disaccharide) of HS.

The distinct mass difference (42 Da) caused by the Nacetylation could be used to distinguish the N-acetylated hexosamine residues of CS/DS and the N-sulfated hexosamine residues of HP. A proof-of-concept experiment has been carried out to investigate MALDI-MS fingerprints of the complex mixture of chondroitin sulfate (CS), dermatan sulfate (DS), and heparin (HP) with different molar ratios (Figure S7 in the Supporting Information). After methanolysis, the CS/DSderived disaccharides can be distinguished from that of HPderived disaccharides by the mass shift of 42 Da. Oversulfated chondroitin sulfate (OSCS) in heparin has been reported as a potential contamination and produced acute and serious side effects during heparin therapy.25 Traditional screening tests cannot differentiate the affected and unaffected heparin. Detailed structure analysis using high-resolution NMR spectroscopy is required to clearly define the extent of heparin. Our new approach could serve as a facile and effective screening alternative to determine whether or not the heparin lots contain the OSSC contaminant. There are fundamental limitations in the current MS-based approach for analysis of CS/DS mixtures after methanolysis because the structures of these GAG-derived disaccharides share the same mass spectral features. However, the qualitative identification of the exact types of GAGs using MALDI-TOF MS still can be accomplished by incorporating specific enzymatic depolymerization of the GAGs. Current practice for MPS diagnosis occurs through a combination of clinical findings and laboratory tests.12 The laboratory tests involve quantitative and qualitative analysis of GAGs and enzyme activity analysis. Quantitative analysis of GAG typically measures the total amount of GAG excreted in the urine and is commonly used as an initial screening strategy even though it is not specific for qualitative identification of the exact types of GAGs. Further qualitative GAG analysis is required to identify the exact types of GAG elevation to provide more insights for MPS diagnosis. Our approach here is aimed to identify and quantify the disaccharides derived from methanolysis of the GAGs using MALDI MS and isotope-labeled internal standards. The approach has great promise as a first-step screening 4347

dx.doi.org/10.1021/ac3034097 | Anal. Chem. 2013, 85, 4342−4349

Analytical Chemistry

Article

speedier, drastically reducing the analysis time to be less than 1 h. It is expected that this new approach will be suitable for use as an effective alternative for rapid and sensitive profiling of the GAG accumulation level in patients’ CSF during clinical therapy. Microwave-Assisted Enzymatic Degradation of CS. Rapid and sensitive quantification of GAGs by microwaveassisted methonalysis and MALDI MS is an attractive merit of the technique for clinical diagnosis of MPS disease. However, identification of the exact phenotype of the GAGs is also imperative to provide insight into which MPS disorder is present.28 A major disadvantage of the chemical degradation is that critical structural information such as disaccharide composition and degree of sulfation is lost during methanolysis. Therefore, a workflow for performing parallel analysis with specific enzyme that degrade specific GAGs is described in Scheme 1, and an example is provided in Figure 7 using CS as a

tool for rapid profiling of the total amount of the GAGs. The next is to perform the qualitative analysis of the GAGs using various specific GAG-degrading enzymes. When both qualitative and quantitative analysis of the GAGs are performed, the risk of a false-positive finding for MPS patients will be reduced. Quantitative Analysis of CS spiked in Artificial Cerebrospinal Fluid (aCSF). One of the considerations in evaluating the effectiveness of treatments such as hematopoietic stem cell transplantation or enzyme replacement therapy for MPS is the assessment of the burden of disease at baseline and during therapy.26 New MPS therapeutic approaches using intrathecal delivery of enzyme though cerebrospinal fluid have proven to be successful in reducing total brain GAG storage.27 To evaluate the suitability of our new approach for profiling GAG abundance in cerebrospinal fluid, CS was spiked into aCSF as a mimic system. We analyzed 25 μL of aCSF samples externally spiked with different amounts (12.5−250 ng) of CS, corresponding to the concentrations of 0.5−10 mg/L. Distinct CS disaccharide ion peaks at m/z 460.1 can be observed in all the mass spectra (Figure 6), suggesting that the high-

Figure 7. Mass spectra of oligosaccharides obtained with (a) on-probe and (b) in-solution enzymatic degradation of CS (1 μg) by Chondroitinase ABC (5 mU). Asterisks denoted the CS-derived disaccharide ions detected in the negative mode. The peaks at m/z 426.4 were attributed to matrix-related ions.

Figure 6. Negative ion MALDI-TOF mass spectra of microwaveassisted methanolysis products of CS derived from spiked artificial cerebrospinal fluid (aCSF). The concentrations of CS spiked into the fluid are (a) 10 mg/L (25 μL), (b) 5 mg/L (25 μL), (c) 1 mg/L (25 μL), and (d) 0.5 mg/L (25 μL).

model system. Degradation of GAG by its corresponding enzyme cleaves specific glycosidic linkages present in acidic polysaccharides and produces low-molecular-weight oligosaccharides for subsequent structural characterization by tandem mass spectrometry. Figure 7 shows the mass spectra of sulfated disaccharides obtained with on-probe and in-solution enzymatic digestion of CS (1 μg) using chondroitinase ABC (5 mU). The most abundant ion at m/z 458.1 in the mass spectra corresponds to the molecular ion [M − H]− for CS-derived disaccharides with one sulfate group (ΔUA-β-1,3-GalNAc4S) attached to it. The location of the sulfate group in the disaccharide subunit can be further identified by tandem mass spectrometry.29 Compared with the ion intensity obtained by using conventional insolution CS digestion at 37 °C for 2 h, the corresponding features obtained with microwave-assisted digestion are more intense in peak strength. Moreover, the reaction time of the enzymatic digestion was reduced from 2 h to 5 min with the assistance of the microwave irradiation, which enables rapid and correct profiling of patients’ GAG excretion upon clinical application.

concentration inorganic salts including NaCl, KCl, CaCl2, and MgCl2 in the aCSF do not interfere the affinity extraction of CS by PA-coated NDs and subsequent MALDI MS analysis when NEDC is used as a matrix.21 Only a very small quantity (25 μL) of aCSF solution was required for the detection in this study, when compared to the large quantity (∼9 mL) of human CSF used for total GAG analysis using a colorimetric assay.27 The CSF GAG results reported a normal control value for total GAG of (