Article pubs.acs.org/ac
Butanolysis Derivatization: Improved Sensitivity in LC-MS/MS Quantitation of Heparan Sulfate in Urine from Mucopolysaccharidosis Patients Paul J. Trim, John J. Hopwood, and Marten F. Snel* Lysosomal Diseases Research Unit, South Australian Health and Medical Research Institute (SAHMRI), North Terrace, Adelaide, South Australia 5000, Australia S Supporting Information *
ABSTRACT: Heparan sulfate (HS) is a complex oligosaccharide that is a marker of a number of diseases, most notably several of the mucopolysaccharidoses (MPS). It is a very heterogeneous compound and its quantification at physiological concentrations in patient samples is challenging. Here, we demonstrate novel derivatization chemistry for depolymerization/desulfation and alkylation of HS based on butanolysis. The resultant alkylated disaccharides are quantifiable by LCMS/MS. This new method is at least 70-fold more sensitive than a previously published methanolysis method. Disaccharide yield over time is compared for methanolysis, ethanolysis, and butanolysis. Maximum disaccharide concentration was observed after 2 h with butanolysis and 18 h with ethanolysis whereas a maximum was not reached over the 24 h of the experiment with methanolysis. The sensitivity of the new technique is illustrated by the quantification of HS in 5 μL urine samples from MPS patients and healthy controls. HS was quantifiable in all samples including controls. Disaccharide reaction products were further characterized using exact mass MS/MS.
T
relationship. This uncertainty leads to a dilemma in predicting disease course for the individual, as recently highlighted with a newborn screening program for Krabbe disease.3 Measurement of the storage substrate concentrations in connection with genotyping may be a predictor of age of onset and rate of likely disease progression. GAGs are linear polysaccharides composed of disaccharide subunits, often with complex sulfation or acetylation patterns. For example HS, is composed of uronic acid-(1 → 4)-Dglucosamine disaccharide subunits. The uronic acid can either be β-D-glucuronic acid or α-L-iduronic acid, which may be 2-Osulfated. The D-glucosamine residues can be either N-acetylated or N-sulfated, additionally, they may be 6-O-sulfated and in the case of the N-sulfated species 3-O-sulfated (see Scheme 1).4 This degree of heterogeneity makes analysis and quantification of HS very challenging. A number of mass spectrometry (MS)-based assays for measuring GAG in patient samples have been reported.5−13 The strategy most commonly employed for quantification is to reduce the complexity of the analytical problem of measuring a heterogeneous oligosaccharide. This is achieved by desulfation and depolymerization of oligosaccharides to disaccharides, which can be accomplished either chemically, for example, by
he need to quantify the amount of glycosaminoglycans (GAGs) in readily available patient samples, such as urine, blood or CSF is important for the diagnosis and prognosis of a group of inherited lysosomal storage disorders known as the mucopolysaccharidoses (MPS).1 These disorders can be linked to genetic defects in specific lysosomal enzymes involved in GAG degradation. Over time, this enzymatic dysfunction leads to GAG accumulation in the lysosomes of affected individuals. Depending on the enzyme affected, different substrates are stored: MPS I and II accumulate heparan sulfate (HS) and dermatan sulfate consequent to deficient α-L-iduronidase and iduronate-2-sulfatase, respectively; MPS IIIA-D store HS because of deficient heparan sulfamidase, α-N-acetylglucosaminidase, heparan-α-glucosaminide N-acetyltransferase, and Nacetylglucosamine 6-sulfatase, respectively; keratan sulfate and chondroitin 6-sulfate accumulate in MPS IVA consequent to deficient galactose-6-sulfate sulfatase; keratan sulfate accumulation in MPS IVB is due to deficient β-galactosidase; deficient N-acetylgalactosamine-4-sulfatase leads to the primary storage of dermatan sulfate in MPS VI; and in MPS VII, storage of HS, dermatan sulfate and chondroitin 6-sulfate result from deficient β-glucuronidase.1,2 As genetic testing at early ages becomes more common in individuals with minimal symptoms (and in some cases asymptomatic) previously unknown mutations in these MPS genes are being identified. Often these mutations or combinations do not have an established genotype/phenotype © 2015 American Chemical Society
Received: May 7, 2015 Accepted: August 21, 2015 Published: August 24, 2015 9243
DOI: 10.1021/acs.analchem.5b01743 Anal. Chem. 2015, 87, 9243−9250
Article
Analytical Chemistry Scheme 1. Proposed Reaction Scheme Illustrating Disaccharide Generation by Butanolysis of Native HS
acid hydrolysis8−10,14−17 or enzymatic cleavage, for example, using heparinases.11,12 Although in both of these processes information about the native state of the analyte is lost, the molar amount of disaccharide is many times higher than the molar amount of starting material, the number of different chemical species is greatly reduced and the size of the molecule is much smaller. The size of the molecule is important because, generally, small molecules, for example, disaccharides, are more amenable to standard reversed phase separation and subsequent MS analysis. All of these factors combine to improve the limits of detection relative to measuring intact GAG species. Hence, methods involving depolymerization are often preferable for determining the total amount of GAG when sample amount is the limiting factor, which is usually the case with patient samples. Analysis of naturally occurring GAG fragments or enzymatically/chemically derived GAG products have been used successfully for GAG quantification. However, all of these approaches have limitations. Detection of GAG fragments is complex and does not provide an estimate of total GAG.7 Enzymatic GAG degradation followed by LC-MS/MS is a sensitive technique that has been demonstrated to work on a variety of samples and different GAGs.11,12 Multiple enzymes are required, which are often expensive or difficult to source and do not always provide a complete digestion of the total GAG population. On the other hand, acid hydrolysis or the closely related acid catalyzed transesterification (e.g., methanolysis) reactions use inexpensive reagents and these methods may be applicable to more than one GAG. The reaction scheme for such a reaction is shown in Scheme 1 using the example of butanolysis of HS. As can be seen the reaction product is a desulfated/deactylated dialkylated (in this case butylated) disaccharide. Recent work has shown that LC-MS/MS-based quantification of disaccharides produced by methanolysis depolymerization and desulfation can be used to analyze a
wide range of biological samples, for example, tissue,15,16 urine,9,10 and cerebrospinal fluid.8,16 Despite the success of the methanolysis method there is room to further develop transesterification-based sample preparation. Here, we compare methanol with ethanol, propan-2-ol and butan-1-ol (butanol) to determine which alcohol is most suitable for derivatization of HS, a GAG that accumulates in MPS types I, II, IIIA−D, and VII. In particular, we investigate differences in signal intensity/sensitivity and changes in retention times in reverse phase chromatography. Furthermore, we perform time course experiments to investigate the disaccharide yield over time of methanolysis, ethanolysis and butanolysis to determine the point at which all HS oligomers have been depolymerized, i.e. the reaction has gone to completion. This information is critical for accurate quantification of total HS. We characterize and identify the most prominent methanolysis and butanolysis disaccharide reaction products using exact mass MS/MS. Through these experiments, we have determined that butanolysis has favorable characteristics over methanolysis and ethanolysis. We observe increased assay sensitivity, much faster reaction times and better chromatographic characteristics in reverse phase separations. The newly developed sample preparation method was tested on MPS I, II, IIIA, IIIB, VI (negative control), and control urine samples.
■
MATERIALS Reagents. HS sodium salt was obtained from Celsus Laboratories (Cincinnati, OH, USA). HCl 3 M in methanol and butanol, 1.25 M HCl in ethanol and propan-2-ol, ammonium acetate, acetyl chloride, formic acid (Fluka brand), and 2,2-dimethoxypropane were purchased from Sigma-Aldrich (Castle Hill, NSW, Australia). Butanol-d9 was purchased from SciVac Pty. Ltd. (Hornsby, NSW, Australia). 9244
DOI: 10.1021/acs.analchem.5b01743 Anal. Chem. 2015, 87, 9243−9250
Article
Analytical Chemistry
Sample Preparation: Time Course Experiment. HS standard was added to 63 glass tubes (10 μL of 0.1 μg μL−1); the samples were dried in a rotational vacuum concentrator overnight. To all samples 50 μL of 2,2-dimethoxypropane was added, followed by either 1000 μL of 3 M HCl in methanol or butanol or 1.25 M HCl in ethanol. Samples were then heated to 65, 70, or 100 °C for 2 min, and then sealed and incubated at 65, 70, or 100 °C for methanolic, ethanolic or butanolic HCl, respectively. Samples were incubated for either 1, 2, 4, 6, 8, 18, or 24 h. Three samples were prepared and analyzed at each time point for each alcohol. After incubation, the samples were dried under nitrogen at 45 °C for 45 min. All samples were stored dry until the final time point had been dried; all samples were then reconstituted in 110 μL of deuterated methanolic HS IS solution to allow direct comparison of both sample groups. Sample Preparation for Total HS Determination in Urine. A 5 μL aliquot of each urine sample was placed in a glass culture tube and dried overnight in a rotational vacuum concentrator. To generate a calibration curve, HS standards covering a range from 1 pg to 1 μg starting material were also included. When dry, the samples were digested by adding 50 μL of 2,2-dimethoxypropane followed by 1000 μL of 3 M HCl in butanol. Samples were then heated to 100 °C for 2 min and then were sealed and incubated at 100 °C for 2 h. After incubation, the samples were dried under nitrogen at 45 °C for 45 min and then reconstituted in 200 μL of deuterated butanolic HS IS solution. Sample Preparation−Standard Curve. Methanolic and butanolic standard curves of HS were prepared in glass culture tubes to give final HS concentrations of 1 pg, 10 pg, 100 pg, 1 ng, 10 ng, 100 ng, 1 μg, 10 μg, or 100 μg. All standards were dried in a rotational vacuum concentrator overnight. To all standards 50 μL of 2,2-dimethoxypropane was added, followed by either 1000 μL of 3 M HCl in methanol or butanol. Samples were then heated to 65 °C (methanol) or 100 °C (butanol) for 2 min, and then sealed and incubated at 65 or 100 °C for methanolic or butanolic HCl, respectively. Samples were incubated for 2 h. Three samples were prepared and analyzed at each concentration for each alcohol. After incubation the samples were dried under nitrogen at 45 °C for 45 min, and then reconstituted in 110 μL of deuterated methanolic or butanolic HS IS solution. Creatinine Analysis. Creatinine analysis was performed by Adelaide Pathology Partners (Adelaide, South Australia) using an automated enzymatic colorimetric method using a Cobas Integra 800 (Roche Diagnostics Ltd., Rotkreuz, CH) and Roche reagents (Roche Diagnostics Ltd., Rotkreuz, CH). Chromatography. Chromatographic separation prior to both qualitative and quantitative MS analyses was by means of an Acquity UPLC system equipped with a BEH C18 analytical column (50 mm × 2.1 mm, 1.7 μm particle size) (Waters, Milford, MA, USA). Mobile phase A consisted of 999:1 water/ formic acid (v/v); mobile phase B consisted of 999:1 acetonitrile/formic acid (v/v). Strong and weak needle wash were acetonitrile and water, respectively. Two different gradients were used one for the evaluation of different alcohols and the other for the analysis of urine, digestion time course and standard curve shown in Table 2. For all analysis the flow rate was 0.35 mL min−1 an injection volume of 7.5 μL was used. Quantitative MS. All mass spectrometric HS quantitation experiments were performed using an API 4000 QTrap hybrid triple quadrupole/linear ion trap mass spectrometer (ABSciex,
Optima LC/MS grade acetonitrile was obtained from Fisher Scientific (Scoresby, Vic, Australia), and ultrapure water was obtained directly from a Milli-Q Advantage A10 water purification system (Millipore, Bayswater,VIC, Australia). Patient Samples. Deidentified urine samples from 14 MPS patients (5 MPS I, 3 MPS II, 2 MPS IIIA, 2 MPS IIIB, and 2 MPS VI) and 14 from healthy controls were analyzed. The study was conducted with the approval of the Women’s and Children’s Hospital Health Network Human Research Ethics Committee. Samples had been stored at −20 °C until analyzed (summarized in Table 1). Control samples were age- and Table 1. Patient and Healthy Control Demographics age/years
control
>20 10−20 5−10 1−5
