Molecular Genetics and Metabolism 114 (2015) 123–128
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
A straightforward, quantitative ultra-performance liquid chromatography‐tandem mass spectrometric method for heparan sulfate, dermatan sulfate and chondroitin sulfate in urine: An improved clinical screening test for the mucopolysaccharidoses☆ Haoyue Zhang a, Tim Wood b, Sarah P. Young a, David S. Millington a,⁎ a b
Duke Medicine, Pediatrics, Medical Genetics Division, NC, USA Greenwood Genetic Center, SC, USA
a r t i c l e
i n f o
Article history: Received 16 September 2014 Accepted 19 September 2014 Available online 5 October 2014 Keywords: Mucopolysaccharidoses Glycosaminoglycans Urine UPLC–MS/MS Methanolysis Screening
a b s t r a c t Mucopolysaccharidoses (MPS) are complex storage disorders that result in the accumulation of glycosaminoglycans (GAGs) in urine, blood, brain and other tissues. Symptomatic patients are typically screened for MPS by analysis of GAG in urine. Current screening methods used in clinical laboratories are based on colorimetric assays that lack the sensitivity and specificity to reliably detect mild GAG elevations that occur in some patients with MPS. We have developed a straightforward, reliable method to quantify chondroitin sulfate (CS), dermatan sulfate (DS) and heparan sulfate (HS) in urine by stable isotope dilution tandem mass spectrometry. The GAGs were methanolyzed to uronic acid-N-acetylhexosamine or iduronic acid-N-glucosamine dimers and mixed with stable isotope labeled internal standards derived from deuteriomethanolysis of GAG standards. Specific dimers derived from HS, DS and CS were separated by ultra-performance liquid chromatography and analyzed by electrospray ionization tandem mass spectrometry using selected reaction monitoring for each targeted GAG product and its corresponding internal standard. The method was robust with a mean inaccuracy from 1 to 15%, imprecision below 11%, and a lower limit of quantification of 0.4 mg/L for CS, DS and HS. We demonstrate that the method has the required sensitivity and specificity to discriminate patients with MPS III, MPS IVA and MPS VI from those with MPS I or MPS II and can detect mildly elevated GAG species relative to age-specific reference intervals. This assay may also be used for the monitoring of patients following therapeutic intervention. Patients with MPS IVB are, however, not detectable by this method. © 2014 Elsevier Inc. All rights reserved.
1. Introduction The mucopolysaccharidoses (MPS) are a group of rare inherited disorders that result from a deficiency in one or more of the lysosomal enzymes required to degrade glycosaminoglycans (GAGs) [1]. The progressive accumulation of GAG in body fluids and tissues of affected patients ultimately causes severe neurological and physical impairment [2]. New therapeutic strategies, including allogeneic hematopoietic
Abbreviations: MPS, mucopolysaccharidoses; GAGs, glycosaminoglycans; UPLC–MS/ MS, ultra-performance liquid chromatography-tandem mass spectrometry; CS, chondroitin sulfate;DS, dermatan sulfate;HS, heparan sulfate; HS-NS, N-sulfated domain of heparan sulfate; DiCS, dimer from methanolyzed CS; DiDS, dimer from methanolyzed DS; DiHS-NS, dimer from methanolyzed HS-NS; DMB, dimethylmethylene blue; KS, keratan sulfate; CSF, cerebrospinal fluid. ☆ Subject category: Glycosaminoglycans. ⁎ Corresponding author at: Biochemical Genetics Laboratory, Duke Medicine, 801-6 Capitola Drive Durham, NC 27713, USA. E-mail address: [email protected] (D.S. Millington).
http://dx.doi.org/10.1016/j.ymgme.2014.09.009 1096-7192/© 2014 Elsevier Inc. All rights reserved.
stem cell transplantation, bone marrow transplantation and enzyme replacement therapy have been at least partially successful in alleviating the symptoms in patients affected with MPS. Consequently there has been increasing interest in screening both newborns and older patients that exhibit the symptoms of MPS in the hope of subsequently making a diagnosis and applying a treatment that could improve their quality of life [3]. The method of choice to screen symptomatic patients for MPS is based on the analysis of GAG in urine, where they are readily excreted owing to their high polarity. Typically, these assays are performed in clinical laboratories in spot urine samples, using initially a spectrophotometric method that determines total GAG relative to creatinine in the patient [4–6,16–18] and compares it with an age-matched reference range. A positive total GAG screen is followed by a chromatographic separation of the intact GAG, usually by gel electrophoresis followed by band staining to identify the specific GAG that appears to be elevated. Although these methods are inexpensive and straightforward, they lack both sensitivity and specificity and are unable to detect the individual GAG in persons unaffected by the diseases, thus making it difficult to establish control ranges.
124
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Several improved methods to detect and analyze GAG that include ELISA and tandem mass spectrometry (MS/MS) have recently been reviewed [7]. Of these, the most promising for screening purposes are liquid chromatography–tandem mass spectrometry (LC–MS/MS) methods targeted to oligosaccharides derived from the principal GAG, chondroitin sulfate (CS), dermatan sulfate (DS), heparan sulfate (HS) and keratan sulfate (KS). Some of these disease-specific species are excreted naturally in small quantities in individuals unaffected with MPS [8,9], and their detection is increased considerably by controlled degradation of the GAG, either enzymatically [10] or chemically [11]. The challenge for these methods is to provide robust quantitative assays that are applicable to the technology available in clinical laboratories. Many such laboratories, including our own, are now equipped with a basic triple quadrupole MS/MS systems to quantify acylcarnitines, amino acids and other valuable disease biomarkers. We recently reported the application of such a system in a novel LC–MS/MS method for GAG analysis in liquid urine and dried urine spots on filter paper, based on the degradation of GAG by methanolysis followed by the analysis of dimeric products [11,12]. Due to limitations of the method used at that time, we were unable to establish reference ranges for HS and DS. A modified version of this method was subsequently developed to analyze GAG in both urine and cerebrospinal fluid (CSF), from patients and unaffected controls, utilizing ultra-performance liquid chromatography (UPLC) with MS/MS with the incorporation of isotope-labeled dimers derived from DS, HS and CS as pseudo-internal standards [13,14]. The use of UPLC increased the resolution of the dimers and thereby afforded a significant improvement in sensitivity, and the use of isotope-labeled internal standards greatly improved the quantitative accuracy and precision. The method was thereby able to detect and quantify CS, DS and HS in the urine and CSF of unaffected individuals. Here we report the application of this improved method to establish the urinary concentrations of CS, DS and HS in pediatric and adult control groups, and to determine the urinary levels of these biomarkers in patients with MPS I, II, III, IVA and VI as well as in other lysosomal storage disorders. This method did not produce a signal from KS that would enable detection of patients with MPS-IVB, although patients having the more common sub-type MPS-IVA were expected to have elevated CS. The main objectives of this research were to report age-specific reference intervals of specific GAG in urine and to objectively compare methanolysis UPLC–MS/MS with a standard clinical laboratory method to detect and quantify urinary GAG.
six urine samples from patients with MPS IH, MPS II, MPS IIIA, MPS IIIB, MPS IIIC, MPS IVA and MPS VI, with diagnoses confirmed by enzymatic and/or gene mutation analysis, were provided by Greenwood Genetic Center, South Carolina. Among them, five samples were collected from a single patient with MPS IIIB over a nine day period to assess short-term fluctuation of HS. Specimens were also provided from patients with other lysosomal storage disorders including sialuria, alphafucosidase deficiency, alpha-mannosidase deficiency and Pompe disease (acid maltase deficiency) to represent disease controls. 2.3. Total GAG assay Total GAG was quantified in 54 samples using the Blyscan kit (1,9-dimethyl-methylene blue staining) (Blyscan Sulfated Glycosaminoglycan Assay, Biocolor Ltd., Northern Ireland, U.K.). 2.4. Internal standards, calibrators and quality controls for UPLC–MS/MS assay Internal standards were prepared by deuteriomethanolysis of DS and HS standards as described previously [14]. Detailed procedures for the preparation and analysis of calibrators and QC samples have also been previously described [14]. 2.5. Sample preparation and analysis by UPLC–MS/MS
2. Materials and methods
The creatinine values of urine specimens were determined by spectrophotometry using the alkaline picrate (Jaffe) method [15]. The preparation and analysis of urine samples for the UPLC–MS/MS assay have been published in detail elsewhere [14]. Briefly, 25 μL urine was evaporated to dryness under nitrogen. The residue was incubated with anhydrous 3 mol/L HCl–MeOH (200 μL) at 65 °C for 75 min and again evaporated to dryness under nitrogen. The residue was dissolved in a solution containing a fixed amount of the internal standard mixture (200 μL) and an aliquot (5 μL) was injected into a Xevo-TQ™ mass spectrometer equipped with an Acquity UPLC® system (Waters Corporation, Milford, MA). Separation was effected using a UPLC® BEH Amide 1.7 μm, 2.1 × 50 mm column (Waters Corporation) with a linear gradient [14]. Data were acquired by selected reaction monitoring (SRM) using the specific transitions corresponding to dimers DiCS and DiDS derived from CS and DS and for the dimer DiHS-NS derived specifically from HS-NS, plus those corresponding to their respective deuterium-labeled internal standards.
2.1. Reagents
3. Results
Dermatan, chondroitin and heparan sulfates were obtained from Sigma Aldrich (St Louis, MO). Keratan sulfate was from Glycosyn, LLC (Boston, MA) Anhydrous 3 mol/L-HCl in methanol, deuterium (2H)labeled methanol and acetyl chloride were from Sigma Aldrich. Acetonitrile (MeCN) was from EMD Chemicals (Gibbstown, NJ), ammonium acetate was from Sigma Aldrich and deionized water was prepared in-house.
3.1. Calibration and validation of the assay
2.2. Patients and controls Infant controls were 48 patients (b 1 year of age) referred for urine organic acid testing, but in whom no diagnosis was made and urinary organic acid levels were within normal limits. These samples were de-identified and employed to establish infant control levels of urinary GAG. In addition, 121 urine samples were obtained, some from patients that were undiagnosed and others from healthy volunteers, between 1 and 70 years of age. They were subdivided into four age groups of 1–3, 4–9, 10–17 and N18 years of age respectively, with at least 29 subjects in each group. These protocols were approved by the Duke University Institutional Review Board. Twenty
The peak area ratios of the signals derived from DiCS, DiDS and DiHSNS to those of the corresponding internal standards were calculated using TargetLynx® software (Waters Corp). Calibration curves derived from these peak area ratios were linear over the calibration range 8 to 400 mg/L (r2 N 0.99). The mean slopes of these curves for CS, DS and HS were 0.66, 0.22 and 0.09 (n = 8) respectively, and the coefficient of variation (CV) of the slopes was lower than 9% for each analyte. Precision and accuracy of the method were determined by comparing measured concentrations back-calculated from the calibration curves with those of the known added concentrations in the calibrators (Table 1). CV of the measured concentrations was from 1 to 10% for CS, DS and HS calibrators (n = 8). Mean inaccuracy was 1 to 15% (n = 8) for CS, DS and HS over the concentration range 8 to 400 mg/L. For urine samples with GAG concentrations below 8 mg/L, calibrators in a lower concentration range (0.4–8 mg/L, inaccuracy: 1–20%) were used to derive calibration curves to quantify CS, DS and HS, respectively (not shown). Precision of the assay was also evaluated by replicate analysis of low, medium and high QCs (Table 2). Intra-day imprecision (CV) of the low,
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128 Table 1 Accuracy and precision of the GAG assays according to the replicate analysis of calibrators. The measured concentrations were back-calculated from the slopes of the calibration curves and compared with the nominal values. The data are compiled from 8 calibration curves performed on different days. Units are mg/mL. Sample ID
Nominal conc.
Mean conc. (n = 8)
SD
CV (%)
Inaccuracy (%)
8 20 40 80 200 400 8 20 40 80 200 400 8 20 40 80 200 400
7 21 41 82 203 394 7 21 43 85 204 389 8 21 43 83 197 397
0.5 1.1 1.7 4.4 3.9 3.7 0.5 1.2 1.2 3.5 4.7 5.4 0.7 1.6 3.4 3.3 5.9 7.1
7 5 4 5 2 1 8 6 3 4 2 1 10 8 8 4 3 2
8 5 2 3 1 2 15 6 8 6 2 3 7 3 7 4 2 1
CS Cal 1 CS Cal 2 CS Cal 3 CS Cal 4 CS Cal 5 CS Cal 6 DS Cal 1 DS Cal 2 DS Cal 3 DS Cal 4 DS Cal 5 DS Cal 6 HS Cal 1 HS Cal 2 HS Cal 3 HS Cal 4 HS Cal 5 HS Cal 6
medium and high QCs was below 4% for CS and DS, and below 8% for HS (n = 5). The inter-day imprecision (CV) of QC samples was 11% or lower for low, medium, and high QC samples of CS, DS and HS (n = 10, over an eight week period). The lower limit of quantification was 0.4 mg/L where inaccuracy and imprecision were within acceptable limits (b20%) for CS, DS and HS. The limits of detection, defined by a signal to noise ratio of 3:1, were 0.2 mg/L for CS, DS and HS. In samples from unaffected individuals with low absolute concentrations of GAG, a concentration protocol was applied that required admixture of 50 μL of urine with the normal amount of internal standard. Urine samples were stable after three freeze–thaw cycles, in storage at −20 °C for at least one year and at room temperature and 4 °C for at least seven days, as judged by the analysis of CS, DS, and HS in QC samples. 3.2. GAG concentrations in the controls As has been the standard practice for biomarker assays in urine including GAG, the measured concentrations were normalized to the urinary creatinine concentration. The upper reference limits (mean + 2SD) for CS, DS and HS were established for different age groups and are summarized in Table 3A. Also included in Table 3A are the reported concentrations of dimers from a recently published enzymatic degradation LC–MS/MS assay [19], which did not provide data for different age groups. Reference values of total GAG in urine
Table 2 Intra-day and inter-day imprecision of the GAG assay based on replicate analysis of quality control samples. Low QC was prepared by spiking 10 mg/L DS and HS, and 5 mg/L CS into control urine, medium QC was prepared by adding 36 μg/mL DS and 45 μg/mL HS to control urine (no CS added), and high QC was prepared by adding 320 mg/L DS and HS, and 200 mg/L CS to control urine. Intra-day imprecision (n = 5) Mean (mg/L) CS DS
HS
Low QC High QC Low QC Medium QC High QC Low QC Medium QC High QC
9.2 246 15.0 39.6 325 16.0 42.7 342
SD 0.3 7.3 0.6 0.3 12 1.2 3.2 25
Inter-day imprecision, (n = 10) CV (%) 3 3 4 1 4 8 7 7
Mean (mg/L) 9.1 246 14.1 38.6 315 15.5 42.7 322
SD 0.1 9.3 1.1 1.9 17 1.7 4.6 20
CV (%) 1 4 8 5 5 11 11 6
125
Table 3A Age-specific upper limits (M + 2SD) of the reference intervals for the dimers derived from urinary CS, DS and HS by the methanolysis UPLC–MS/MS method; comparison with reference intervals for the corresponding dimers generated and analyzed by the published enzymolysis LC–MS/MS method [19]. Units are g/mol creatinine. Age (years)
DiCS mean + 2SD
DiDS mean + 2SD
Methanolysis with stable isotope dilution-UPLC–MS/MS b1 44 3.4 1–3 23 2.3 4–9 10 2.3 10–17 8.2 2.5 N18 3.3 2.4 Enzymolysis with LC–MS/MSa 0 to 51 0.1–4.1 a
0.03–0.8
DiHS-NS mean + 2SD
n
4.1 3.7 1.8 1.4 1.6
48 29 30 31 31
0.01–0.4
50
Tomatsu S et al. [19].
for different age groups extracted from published spectrophotometric methods [16–18] are shown in Table 3B. Note that the sum of CS, DS and HS concentrations measured by UPLC–MS/MS was similar in each age group to the total GAG measured by spectrophotometric methods, especially by the uronic acid [16] and DMB [18] methods. The reference ranges for the dimers generated by enzymolysis [19], especially diDS and DiHS-NS, were noticeably lower than those from methanolysis, even if one assumes that most of the controls for the enzymatic method were adults. This discrepancy may be accounted for by differences in the calibration method and by the possibility that enzymes are unable to degrade GAG as completely as methanolysis. 3.3. GAG concentrations in patients with MPS In Table 4, we compared the urinary GAG CS, DS and HS in patients with various types of MPS and other storage disorders analyzed using the new UPLC–MS/MS method with the total GAG assayed using dimethylmethylene blue (DMB) spectrophotometry. The concentrations of DS and HS in urine specimens from all patients with MPS-I and MPS-II were well above the upper limits of agematched controls. The concentration of HS in patients with MPS-III ranged from 968 g/mol creatinine for a patient with MPS IIIA, to 32 g/mol creatinine for a patient with MPS IIIB. It is noteworthy that for several of these patients, the total GAG spectrophotometric assay returned values below or close to the reference range limits (Table 4, P9, P6, P10, P8, P32 and Table 5, P22B, P22C) and may have missed the diagnoses. The concentrations of both CS and DS in one of the patients with MPS VI (P5, Table 4) were markedly higher than the age-matched reference limits, while in another patient with MPS VI (P8) they were mildly though measurably higher than the agematched reference limits. The concentration of HS was not elevated in either of the patients with MPS VI. Seven patients with MPS-IVA (P28–P34) showed modest but clear elevations of CS, but not DS or HS, consistent with the expected elevation of chondroitin-6-sulfate in
Table 3B Published reference intervals or up limits for total urinary GAG determined by spectrophotometric- based methods. Units are g/mol creatinine. Age (years)
b1 1–3 4–17 N18 a b c
Total GAG Uronic acid testa
Alcian blueb
Dimethylene bluec
22–50 16–21 6.0–15 3.6
20 9.2 9.2 4.2
36 16 4.9–9.2 3.2
Stone JE [16]. Dembure PP et al. [17]. Jong JG et al. [18].
126
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Table 4 Urinary CS, DS and HS concentrations in patients with various MPS disorders. Urinary CS, DS and HS concentrations with age-matched reference limits (M + 2SD) determined by methanolysis UPLC–MS/MS in patients with MPS disorders (untreated) and comparison with the total GAG concentration measured by spectrophotometry. Units are g/mol creatinine. Values in bold represent elevations relative to the reference ranges. Disorder
Subject
Age (yrs)
UPLC–MS/MS
Spectrophotometry
CS
MPS I MPS I MPS II MPS II MPS IIIA MPS IIIA MPS IIIA MPS IIIA MPS IIIB MPS IIIB MPS IIIC MPS IIIC MPS IIIC MPS VI MPS VI MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA α-Fucosidosis α-Mannosidosis Aspartylglucosaminuria Sialuria Pompe Pompe
P3 P13 P9 P11 P1 P2 P7 P12 P22 P27 P4 P6 P10 P5 P8 P28 P29 P30 P31 P32 P33 P34
1 0.8 3 0 12 10 1 8 13 5 8 9 7 1 41 7 3 6 3 6 5 7 6.5 24 49 12 0.6 5.9
DS
HS
Total GAG
Patient
Control
Patient
Control
Patient
Control
Patient
Control
59 86 15 35 5.2 6.0 19 6.1 5.6 4.5 5.7 3.9 6.0 79 8.0 26 33 21 28 26 16 19 13 1.1 22 4.8 13 7.3
23 44 23 44 8.2 8.2 23 10 8.2 10 10 10 10 23 3.3 10 23 10 23 10 10 10 10 3.3 3.3 8.2 44 10
148 177 38 88 2.1 2.1 4.0 2.2 1.6 1.0 2.2 2.0 2.1 103 7.2 3.0 3.0 2.0 2.2 1.8 1.8 2.3 1.9 0.8 6.2 1.4 1.0 0.9
3.7 4.1 3.7 4.1 2.5 2.5 3.7 2.3 2.5 2.3 2.3 2.3 2.3 2.3 2.4 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.4 2.4 2.5 4.1 2.3
200 417 139 416 405 360 968 599 50 32 421 419 345 2.5 0.9 2.0 2.7 1.3 1.5 1.5 0.7 1.4 3.5 1.0 6.8 1.4 3.0 1.3
3.7 4.1 3.7 4.1 1.4 1.4 3.7 1.8 1.4 1.8 1.8 1.8 1.8 3.7 1.6 1.8 3.7 1.8 3.7 1.8 1.8 1.8 1.8 1.6 1.6 1.4 3.7 1.8
90 209 17 60 21 18 67 34 22 20 19 11 11 95 8.3 22 46 25 34 11 20 31 ND ND ND ND ND ND
31 31 16 31 10 12 31 12 10 16 12 12 12 31 6.5 12 16 12 16 12 16 12 N/A N/A N/A N/A N/A N/A
to a lack of sensitivity for the spectrophotometric method, which targets uronic acid moieties within the GAG polymer, some of which may be inaccessible to the reagent.
this condition. Therefore, the UPLC–MS/MS method was able to discriminate patients with MPS-III, MPS-IVA and MPS-VI from those with MPS-I and MPS-II, and from controls. Patients with other lysosomal storage conditions, including sialuria, alpha-fucosidase deficiency, alpha-mannosidase deficiency, aspartylglucosaminuria, Pompe disease (acid maltase deficiency) are included in Table 4 as disease controls. Most of these gave GAG values within the control ranges, however, the patient with aspartylglucosaminuria had elevations of all three targeted GAGs, for reasons that are not clear. To test the short-term fluctuation of HS excretion in a patient with MPS III, five urine samples (P#22–26) were collected from a 13 year-old patient with MPS IIIB over a period of nine days. The results, shown in Table 5, show that the HS concentration varied from 50 to 85 g/mol creatinine with a mean value of 71 and CV of 22% over this period and was far in excess of the upper limit of the age-matched controls (1.5 g/mol creatinine). By contrast, the variation in the total GAG spectrophotometric assay was from 9.6 to 31 g/mol creatinine with a mean value of 21 g/mol creatinine and a CV of 46%. The mean value was only twice that of the upper control limit (10 g/mol creatinine). We ascribe these differences
4. Discussion Historically, urine GAG measurement has been performed by quantitative urine GAG tests that provide a numeric representation of the total GAG level but do not differentiate between the individual GAG and qualitative tests that chromatographically separate the different GAG species but do not specifically quantify them. In many clinical laboratories, the DMB spectrophotometric method is used as a preliminary screen to quantify total GAG, where GAG sulfates including CS, DS, HS and KS form a color complex with DMB. Diagnostic laboratories are therefore typically running two separate tests on the same sample, one a total GAG assay and the other a separation method based on electrophoresis to determine which GAG species may be elevated. It is widely known that these testing strategies can have false positives and, more disconcertingly, false negatives. The analytical method for GAG
Table 5 Urinary CS, DS and HS concentrations (UPLC–MS/MS) and total GAG concentration (spectrophotometry) in five urine samples collected from a patient with MPS IIIB over nine days (subject P22, age 13 yrs). Units are g/mol creatinine. Values in bold represent elevations relative to the reference ranges. Patient
UPLC–MS/MS
Spectrophotometry
CS
P22A P22B P22C P22E P22F
DS
HS
Total GAG
Patient
Control
Patient
Control
Patient
Control
Patient
Control
5.6 4.7 6.1 5.7 6.2
8.2 8.2 8.2 8.2 8.2
1.6 1.5 1.7 1.7 1.7
2.5 2.5 2.5 2.5 2.5
50 74 85 56 66
1.4 1.4 1.4 1.4 1.4
22 13 9.6 31 29
10 10 10 10 10
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
described here clearly has much greater sensitivity and specificity than current methods employed in clinical laboratories to detect and quantify signals specifically derived from DS and HS, the principal accumulating storage materials in the lysosomes of patients with MPS I, MPS II, MPS III and MPS VI. Although this method did not produce a specific marker from KS that enabled differentiation of patients with MPS-IV from controls, all the patients with MPS-IVA included in this study had elevated CS compared with age-matched controls. In fact, UPLC–MS/MS affords greater separation between affected patients and age-matched controls than any previously published method that we are aware of. It should be noted that the CS concentration determined by this method in patients with elevated values of DS may be over-estimated because of the contribution of methanolyzed GlcA–-GalNAc dimers from DS, which consists of both GlcA–GalNAc and IdoA–GalNAc repeating disaccharide units, while CS consists of a GlcA-GalNAc repeating unit [20]. Thus, the presence of IdoA in DS distinguishes it from CS and provides a specific marker for DS. When compared with the previously reported enzymolysis-based LC–MS/MS method, methanolysis apparently yields significantly higher concentrations of HS and DS, presumably because both sulfated and non-sulfated GAG units are detected after derivatization. Unlike enzymolysis-LC–MS/MS or spectrophotometry, the method of quantification is based on stable isotopedilution, which provides a level of accuracy and precision not readily achievable by alternative analytical methods. This new method may thus afford a lower false negative rate by means of the improved sensitivity and specificity to detect and quantify DS and HS separately from CS. It is the only published method we are aware of that provides urine control ranges for all three GAGs in different age groups. The assay relies for its accuracy and precision upon the reproducibility of the methanolysis step, which must be performed with fresh reagent and with careful temperature and time control of the reaction. Further advantages of the method are its short turn-around time and its suitability for batched analysis of multiple samples. As previously stated, this method apparently lacks the ability to detect keratan sulfate, specific for MPS IVA and B (Morquio syndrome). However, patients with MPS type IVA are known to accumulate chondroitin-6sulfate(C-6S) as well as KS. Although methanolysis cannot distinguish between C-4S and C-6S, all urine samples from known MPSIVA patients that we have analyzed thus far have shown an elevation of CS without elevated DS or HS, which clearly distinguishes them from other MPS and from age-matched controls. Chondroitin-6sulfate has recently been shown to be a novel biomarker for both MPS IVA and MPS VII [21]. A specific test for KS in urine that uses LC–MS/MS to target a dimer produced by enzymatic degradation [22,23] is now available. We propose to use this assay as a second tier test either to confirm a presumptive positive screen for MPSIVA or to rule out MPS-IVB if the primary screening test is negative. Our experience suggests that patients with certain other storage diseases may have elevated urine GAG, so the test is probably not 100% specific for MPS. In summary, the method described here not only improves the sensitivity of current testing strategies, but also uses a single test to define which GAG species are increased and their relative levels. This combination should save time in diagnostic labs, allowing two tests to be combined into one, and offers a more reliable alternative to the older screening methods. The protocol for the assay is straightforward and can readily be incorporated into clinical laboratories that are equipped with MS/MS for the purpose of analyzing acylcarnitines and other disease biomarkers. It is important to understand that this test is intended for screening patients at risk for MPS. Following a presumptive positive screen by this test, further diagnostic testing by enzymology and/or molecular testing is warranted to confirm the specific diagnosis in the patient. It remains to be seen if the ability to quantify specific GAG species in urine will provide further benefit when used to monitor patients on enzyme
127
replacement or other therapies, as has already been demonstrated from the study of plasma GAG [7] and CSF GAG [13]. Conflict of interest The authors declare that they have no conflicts of interest. Acknowledgments The authors are grateful for the financial support from Genzyme Corporation, a Sanofi Company and from BioMarin Pharmaceuticals. We thank Dr. Steve Skinner and Fran Anese for their help in obtaining urine samples on known patients and also Laura Pollard and Teresa Thompson from the Greenwood Genetic Center Biochemical Genetics lab for their help in sample preparation. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.09.009. References [1] E. Neufeld, J. Muenzer, The mucopolysaccharidoses, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and Molecular Bases of Inherited Disease, 8th ed.McGraw Hill, New York, 2001, pp. 3421–3452. [2] J. Muenzer, The mucopolysaccharidoses: a heterogeneous group of disorders with variable pediatric presentations, J. Pediatr. 144 (2004) S27–S34. [3] H. Zhou, P. Fernhoff, R.F. Vogt, Newborn bloodspot screening for lysosomal storage disorders, J. Pediatr. 159 (2011) 7–13. [4] J.G. de Jong, R.A. Wevers, R. Liebrand-van Sambeek, Measuring urinary glycosaminoglycans in the presence of protein: an improved screening procedure for mucopolysaccharidoses based on dimethylmethylene blue, Clin. Chem. 38 (1992) 803–807. [5] R.W. Farndale, C.A. Sayers, A.J. Barrett, A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures, Connect. Tissue Res. 9 (1982) 247–248. [6] G. Panin, S. Naia, R. Dall'Amico, L. Chiandetti, F. Zachello, C. Catassi, L. Felici, G.V. Coppa, Simple spectrophotometric quantification of urinary excretion of glycosaminoglycan sulfates, Clin. Chem. 32 (1986) 2073–2076. [7] S. Tomatsu, T. Shimada, R.W. Mason, A.M. Montano, J. Kelly, W.A. LaMarr, F. Kubaski, R. Giugliani, A. Guha, E. Yusada, W. Mackenzie, S. Yamaguchi, Y. Suzuki, T. Orii, Establishment of glycosaminoglycan assays for mucopolysaccharidoses, Metabolites 4 (2014) 655–679. [8] S.L. Ramsay, P.J. Meikle, J.J. Hopwood, Determination of monosaccharides and disaccharides in mucopolysaccharidoses patients by electrospray ionisation mass spectrometry, Mol. Genet. Metab. 78 (2003) 193–204. [9] M. Fuller, T. Rozaklis, S.L. Ramsay, J.J. Hopwood, P.J. Meikle, Disease-specific markers for the mucopolysaccharidoses, Pediatr. Res. 56 (2004) 733–738. [10] T. Ogama, S. Tomatsu, A.M. Montano, O. Okazaki, Analytical method for the determination of disaccharides derived from keratan, heparan and dermatan sulfates in human serum and plasma by high-performance liquid chromatography/turbo ionspray ionization tandem mass spectrometry, Anal. Biochem. 368 (2007) 79–86. [11] C. Auray-Blais, P. Bherer, R. Gagnon, S.P. Young, H.H. Zhang, Y. An, J.T. Clarke, D.S. Millington, Efficient analysis of urinary glycosaminoglycans by LC–MS/MS in mucopolysaccharidoses type I, II and VI, Mol. Genet. Metab. 102 (2011) 49–56. [12] C. Auray-Blais, P. Lavoie, H. Zhang, R. Gagnon, J.T. Clarke, B. Maranda, S.P. Young, Y. An, D.S. Millington, An improved method for glycosaminoglycan analysis by LC–MS/ MS of urine samples collected on filter paper, Clin. Chim. Acta 413 (2012) 771–778. [13] H. Zhang, S.P. Young, C. Auray-Blais, P.J. Orchard, J. Tolar, D.S. Millington, Analysis of glycosaminoglycans in cerebrospinal fluid from patients with mucopolysaccharidoses by isotope-dilution ultra-performance liquid chromatography–tandem mass spectrometry, Clin. Chem. 57 (2011) 1005–1012. [14] H. Zhang, S.P. Young, D.S. Millington, Quantification of glycosaminoglycans in urine by isotope-dilution liquid chromatography–electrospray ionization tandem mass spectrometry, Curr. Protoc. Hum. Genet. 17 (2013) 12. [15] H. Husdan, A. Rapoport, Estimation of creatinine by the Jaffe reaction: a comparison of three methods, Clin. Chem. 14 (1968) 222–238. [16] J.E. Stone, Urine analysis in the diagnosis of mucopolysaccharide disorders, Ann. Clin. Biochem. 35 (1998) 207–225. [17] P.P. Dembure, J.E. Drumheller, S.M. Barr, L.J. Elsas, Selective urinary screening for mucopolysaccharidoses, Clin. Biochem. 23 (1990) 91–96. [18] J.G. de Jong, R.A. Wevers, C. Laarakkers, B.J. Poorthuis, Dimethylmethylene bluebased spectrophotometry of glycosaminoglycans in untreated urine: a rapid screening procedure for mucopolysaccharidoses, Clin. Chem. 35 (1989) 1472–1477. [19] S. Tomatsu, A.M. Montano, T. Oguma, V.C. Dung, H. Oikawa, T.G. de Carvalho, M.L. Gutierrez, S. Yamaguchi, Y. Suzuki, M. Fukushi, N. Sakura, L. Barrera, K. Kida, M.
128
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Kubota, T. Orii, Dermatan sulfate and heparan sulfate as a biomarker for mucopolysaccharidosis I, J. Inherit. Metab. Dis. 33 (2010) 141–150. [20] J.M. Trowbridge, R.L. Gallo, Dermatan sulfate: new functions from an old glycosaminoglycan, Glycobiology 12 (2002) 117R–125R. [21] T. Shimada, S. Tomatsu, E. Yasuda, R.W. Mason, W.G. Mackenzie, Y. Shibata, F. Kubaski, R. Giugliani, S. Yamaguchi, Y. Suzuki, K. Orii, T. Orii, Chondroitin 6-Sulfate as a Novel Biomarker for Mucopolysaccharidosis IVA and VII, JIMD Rep, 2014. (Epub ahead of print).
[22] L.A. Martell, R.L. Cunico, J. Ohh, W. Fulkerson, R. Furneaux, E.D. Foehr, Validation of an LC–MS/MS assay for detecting relevant disaccharides from keratan sulfate as a biomarker for Morquio A syndrome, Bioanalysis 3 (2011) 1855–1866. [23] S. Tomatsu, A.M. Montano, T. Oguma, V.C. Dung, H. Oikawa, T.G. de Carvalho, M.L. Gutierrez, S. Yamaguchi, Y. Suzuki, M. Fukushi, K. Kida, M. Kubota, L. Barrera, T. Orii, Validation of keratan sulfate level in mucopolysaccharidosis type IVA by liquid chromatography–tandem mass spectrometry, J. Inherit. Metab. Dis. 33 (2010) 35–42.
Contents lists available at ScienceDirect
Molecular Genetics and Metabolism journal homepage: www.elsevier.com/locate/ymgme
A straightforward, quantitative ultra-performance liquid chromatography‐tandem mass spectrometric method for heparan sulfate, dermatan sulfate and chondroitin sulfate in urine: An improved clinical screening test for the mucopolysaccharidoses☆ Haoyue Zhang a, Tim Wood b, Sarah P. Young a, David S. Millington a,⁎ a b
Duke Medicine, Pediatrics, Medical Genetics Division, NC, USA Greenwood Genetic Center, SC, USA
a r t i c l e
i n f o
Article history: Received 16 September 2014 Accepted 19 September 2014 Available online 5 October 2014 Keywords: Mucopolysaccharidoses Glycosaminoglycans Urine UPLC–MS/MS Methanolysis Screening
a b s t r a c t Mucopolysaccharidoses (MPS) are complex storage disorders that result in the accumulation of glycosaminoglycans (GAGs) in urine, blood, brain and other tissues. Symptomatic patients are typically screened for MPS by analysis of GAG in urine. Current screening methods used in clinical laboratories are based on colorimetric assays that lack the sensitivity and specificity to reliably detect mild GAG elevations that occur in some patients with MPS. We have developed a straightforward, reliable method to quantify chondroitin sulfate (CS), dermatan sulfate (DS) and heparan sulfate (HS) in urine by stable isotope dilution tandem mass spectrometry. The GAGs were methanolyzed to uronic acid-N-acetylhexosamine or iduronic acid-N-glucosamine dimers and mixed with stable isotope labeled internal standards derived from deuteriomethanolysis of GAG standards. Specific dimers derived from HS, DS and CS were separated by ultra-performance liquid chromatography and analyzed by electrospray ionization tandem mass spectrometry using selected reaction monitoring for each targeted GAG product and its corresponding internal standard. The method was robust with a mean inaccuracy from 1 to 15%, imprecision below 11%, and a lower limit of quantification of 0.4 mg/L for CS, DS and HS. We demonstrate that the method has the required sensitivity and specificity to discriminate patients with MPS III, MPS IVA and MPS VI from those with MPS I or MPS II and can detect mildly elevated GAG species relative to age-specific reference intervals. This assay may also be used for the monitoring of patients following therapeutic intervention. Patients with MPS IVB are, however, not detectable by this method. © 2014 Elsevier Inc. All rights reserved.
1. Introduction The mucopolysaccharidoses (MPS) are a group of rare inherited disorders that result from a deficiency in one or more of the lysosomal enzymes required to degrade glycosaminoglycans (GAGs) [1]. The progressive accumulation of GAG in body fluids and tissues of affected patients ultimately causes severe neurological and physical impairment [2]. New therapeutic strategies, including allogeneic hematopoietic
Abbreviations: MPS, mucopolysaccharidoses; GAGs, glycosaminoglycans; UPLC–MS/ MS, ultra-performance liquid chromatography-tandem mass spectrometry; CS, chondroitin sulfate;DS, dermatan sulfate;HS, heparan sulfate; HS-NS, N-sulfated domain of heparan sulfate; DiCS, dimer from methanolyzed CS; DiDS, dimer from methanolyzed DS; DiHS-NS, dimer from methanolyzed HS-NS; DMB, dimethylmethylene blue; KS, keratan sulfate; CSF, cerebrospinal fluid. ☆ Subject category: Glycosaminoglycans. ⁎ Corresponding author at: Biochemical Genetics Laboratory, Duke Medicine, 801-6 Capitola Drive Durham, NC 27713, USA. E-mail address: [email protected] (D.S. Millington).
http://dx.doi.org/10.1016/j.ymgme.2014.09.009 1096-7192/© 2014 Elsevier Inc. All rights reserved.
stem cell transplantation, bone marrow transplantation and enzyme replacement therapy have been at least partially successful in alleviating the symptoms in patients affected with MPS. Consequently there has been increasing interest in screening both newborns and older patients that exhibit the symptoms of MPS in the hope of subsequently making a diagnosis and applying a treatment that could improve their quality of life [3]. The method of choice to screen symptomatic patients for MPS is based on the analysis of GAG in urine, where they are readily excreted owing to their high polarity. Typically, these assays are performed in clinical laboratories in spot urine samples, using initially a spectrophotometric method that determines total GAG relative to creatinine in the patient [4–6,16–18] and compares it with an age-matched reference range. A positive total GAG screen is followed by a chromatographic separation of the intact GAG, usually by gel electrophoresis followed by band staining to identify the specific GAG that appears to be elevated. Although these methods are inexpensive and straightforward, they lack both sensitivity and specificity and are unable to detect the individual GAG in persons unaffected by the diseases, thus making it difficult to establish control ranges.
124
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Several improved methods to detect and analyze GAG that include ELISA and tandem mass spectrometry (MS/MS) have recently been reviewed [7]. Of these, the most promising for screening purposes are liquid chromatography–tandem mass spectrometry (LC–MS/MS) methods targeted to oligosaccharides derived from the principal GAG, chondroitin sulfate (CS), dermatan sulfate (DS), heparan sulfate (HS) and keratan sulfate (KS). Some of these disease-specific species are excreted naturally in small quantities in individuals unaffected with MPS [8,9], and their detection is increased considerably by controlled degradation of the GAG, either enzymatically [10] or chemically [11]. The challenge for these methods is to provide robust quantitative assays that are applicable to the technology available in clinical laboratories. Many such laboratories, including our own, are now equipped with a basic triple quadrupole MS/MS systems to quantify acylcarnitines, amino acids and other valuable disease biomarkers. We recently reported the application of such a system in a novel LC–MS/MS method for GAG analysis in liquid urine and dried urine spots on filter paper, based on the degradation of GAG by methanolysis followed by the analysis of dimeric products [11,12]. Due to limitations of the method used at that time, we were unable to establish reference ranges for HS and DS. A modified version of this method was subsequently developed to analyze GAG in both urine and cerebrospinal fluid (CSF), from patients and unaffected controls, utilizing ultra-performance liquid chromatography (UPLC) with MS/MS with the incorporation of isotope-labeled dimers derived from DS, HS and CS as pseudo-internal standards [13,14]. The use of UPLC increased the resolution of the dimers and thereby afforded a significant improvement in sensitivity, and the use of isotope-labeled internal standards greatly improved the quantitative accuracy and precision. The method was thereby able to detect and quantify CS, DS and HS in the urine and CSF of unaffected individuals. Here we report the application of this improved method to establish the urinary concentrations of CS, DS and HS in pediatric and adult control groups, and to determine the urinary levels of these biomarkers in patients with MPS I, II, III, IVA and VI as well as in other lysosomal storage disorders. This method did not produce a signal from KS that would enable detection of patients with MPS-IVB, although patients having the more common sub-type MPS-IVA were expected to have elevated CS. The main objectives of this research were to report age-specific reference intervals of specific GAG in urine and to objectively compare methanolysis UPLC–MS/MS with a standard clinical laboratory method to detect and quantify urinary GAG.
six urine samples from patients with MPS IH, MPS II, MPS IIIA, MPS IIIB, MPS IIIC, MPS IVA and MPS VI, with diagnoses confirmed by enzymatic and/or gene mutation analysis, were provided by Greenwood Genetic Center, South Carolina. Among them, five samples were collected from a single patient with MPS IIIB over a nine day period to assess short-term fluctuation of HS. Specimens were also provided from patients with other lysosomal storage disorders including sialuria, alphafucosidase deficiency, alpha-mannosidase deficiency and Pompe disease (acid maltase deficiency) to represent disease controls. 2.3. Total GAG assay Total GAG was quantified in 54 samples using the Blyscan kit (1,9-dimethyl-methylene blue staining) (Blyscan Sulfated Glycosaminoglycan Assay, Biocolor Ltd., Northern Ireland, U.K.). 2.4. Internal standards, calibrators and quality controls for UPLC–MS/MS assay Internal standards were prepared by deuteriomethanolysis of DS and HS standards as described previously [14]. Detailed procedures for the preparation and analysis of calibrators and QC samples have also been previously described [14]. 2.5. Sample preparation and analysis by UPLC–MS/MS
2. Materials and methods
The creatinine values of urine specimens were determined by spectrophotometry using the alkaline picrate (Jaffe) method [15]. The preparation and analysis of urine samples for the UPLC–MS/MS assay have been published in detail elsewhere [14]. Briefly, 25 μL urine was evaporated to dryness under nitrogen. The residue was incubated with anhydrous 3 mol/L HCl–MeOH (200 μL) at 65 °C for 75 min and again evaporated to dryness under nitrogen. The residue was dissolved in a solution containing a fixed amount of the internal standard mixture (200 μL) and an aliquot (5 μL) was injected into a Xevo-TQ™ mass spectrometer equipped with an Acquity UPLC® system (Waters Corporation, Milford, MA). Separation was effected using a UPLC® BEH Amide 1.7 μm, 2.1 × 50 mm column (Waters Corporation) with a linear gradient [14]. Data were acquired by selected reaction monitoring (SRM) using the specific transitions corresponding to dimers DiCS and DiDS derived from CS and DS and for the dimer DiHS-NS derived specifically from HS-NS, plus those corresponding to their respective deuterium-labeled internal standards.
2.1. Reagents
3. Results
Dermatan, chondroitin and heparan sulfates were obtained from Sigma Aldrich (St Louis, MO). Keratan sulfate was from Glycosyn, LLC (Boston, MA) Anhydrous 3 mol/L-HCl in methanol, deuterium (2H)labeled methanol and acetyl chloride were from Sigma Aldrich. Acetonitrile (MeCN) was from EMD Chemicals (Gibbstown, NJ), ammonium acetate was from Sigma Aldrich and deionized water was prepared in-house.
3.1. Calibration and validation of the assay
2.2. Patients and controls Infant controls were 48 patients (b 1 year of age) referred for urine organic acid testing, but in whom no diagnosis was made and urinary organic acid levels were within normal limits. These samples were de-identified and employed to establish infant control levels of urinary GAG. In addition, 121 urine samples were obtained, some from patients that were undiagnosed and others from healthy volunteers, between 1 and 70 years of age. They were subdivided into four age groups of 1–3, 4–9, 10–17 and N18 years of age respectively, with at least 29 subjects in each group. These protocols were approved by the Duke University Institutional Review Board. Twenty
The peak area ratios of the signals derived from DiCS, DiDS and DiHSNS to those of the corresponding internal standards were calculated using TargetLynx® software (Waters Corp). Calibration curves derived from these peak area ratios were linear over the calibration range 8 to 400 mg/L (r2 N 0.99). The mean slopes of these curves for CS, DS and HS were 0.66, 0.22 and 0.09 (n = 8) respectively, and the coefficient of variation (CV) of the slopes was lower than 9% for each analyte. Precision and accuracy of the method were determined by comparing measured concentrations back-calculated from the calibration curves with those of the known added concentrations in the calibrators (Table 1). CV of the measured concentrations was from 1 to 10% for CS, DS and HS calibrators (n = 8). Mean inaccuracy was 1 to 15% (n = 8) for CS, DS and HS over the concentration range 8 to 400 mg/L. For urine samples with GAG concentrations below 8 mg/L, calibrators in a lower concentration range (0.4–8 mg/L, inaccuracy: 1–20%) were used to derive calibration curves to quantify CS, DS and HS, respectively (not shown). Precision of the assay was also evaluated by replicate analysis of low, medium and high QCs (Table 2). Intra-day imprecision (CV) of the low,
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128 Table 1 Accuracy and precision of the GAG assays according to the replicate analysis of calibrators. The measured concentrations were back-calculated from the slopes of the calibration curves and compared with the nominal values. The data are compiled from 8 calibration curves performed on different days. Units are mg/mL. Sample ID
Nominal conc.
Mean conc. (n = 8)
SD
CV (%)
Inaccuracy (%)
8 20 40 80 200 400 8 20 40 80 200 400 8 20 40 80 200 400
7 21 41 82 203 394 7 21 43 85 204 389 8 21 43 83 197 397
0.5 1.1 1.7 4.4 3.9 3.7 0.5 1.2 1.2 3.5 4.7 5.4 0.7 1.6 3.4 3.3 5.9 7.1
7 5 4 5 2 1 8 6 3 4 2 1 10 8 8 4 3 2
8 5 2 3 1 2 15 6 8 6 2 3 7 3 7 4 2 1
CS Cal 1 CS Cal 2 CS Cal 3 CS Cal 4 CS Cal 5 CS Cal 6 DS Cal 1 DS Cal 2 DS Cal 3 DS Cal 4 DS Cal 5 DS Cal 6 HS Cal 1 HS Cal 2 HS Cal 3 HS Cal 4 HS Cal 5 HS Cal 6
medium and high QCs was below 4% for CS and DS, and below 8% for HS (n = 5). The inter-day imprecision (CV) of QC samples was 11% or lower for low, medium, and high QC samples of CS, DS and HS (n = 10, over an eight week period). The lower limit of quantification was 0.4 mg/L where inaccuracy and imprecision were within acceptable limits (b20%) for CS, DS and HS. The limits of detection, defined by a signal to noise ratio of 3:1, were 0.2 mg/L for CS, DS and HS. In samples from unaffected individuals with low absolute concentrations of GAG, a concentration protocol was applied that required admixture of 50 μL of urine with the normal amount of internal standard. Urine samples were stable after three freeze–thaw cycles, in storage at −20 °C for at least one year and at room temperature and 4 °C for at least seven days, as judged by the analysis of CS, DS, and HS in QC samples. 3.2. GAG concentrations in the controls As has been the standard practice for biomarker assays in urine including GAG, the measured concentrations were normalized to the urinary creatinine concentration. The upper reference limits (mean + 2SD) for CS, DS and HS were established for different age groups and are summarized in Table 3A. Also included in Table 3A are the reported concentrations of dimers from a recently published enzymatic degradation LC–MS/MS assay [19], which did not provide data for different age groups. Reference values of total GAG in urine
Table 2 Intra-day and inter-day imprecision of the GAG assay based on replicate analysis of quality control samples. Low QC was prepared by spiking 10 mg/L DS and HS, and 5 mg/L CS into control urine, medium QC was prepared by adding 36 μg/mL DS and 45 μg/mL HS to control urine (no CS added), and high QC was prepared by adding 320 mg/L DS and HS, and 200 mg/L CS to control urine. Intra-day imprecision (n = 5) Mean (mg/L) CS DS
HS
Low QC High QC Low QC Medium QC High QC Low QC Medium QC High QC
9.2 246 15.0 39.6 325 16.0 42.7 342
SD 0.3 7.3 0.6 0.3 12 1.2 3.2 25
Inter-day imprecision, (n = 10) CV (%) 3 3 4 1 4 8 7 7
Mean (mg/L) 9.1 246 14.1 38.6 315 15.5 42.7 322
SD 0.1 9.3 1.1 1.9 17 1.7 4.6 20
CV (%) 1 4 8 5 5 11 11 6
125
Table 3A Age-specific upper limits (M + 2SD) of the reference intervals for the dimers derived from urinary CS, DS and HS by the methanolysis UPLC–MS/MS method; comparison with reference intervals for the corresponding dimers generated and analyzed by the published enzymolysis LC–MS/MS method [19]. Units are g/mol creatinine. Age (years)
DiCS mean + 2SD
DiDS mean + 2SD
Methanolysis with stable isotope dilution-UPLC–MS/MS b1 44 3.4 1–3 23 2.3 4–9 10 2.3 10–17 8.2 2.5 N18 3.3 2.4 Enzymolysis with LC–MS/MSa 0 to 51 0.1–4.1 a
0.03–0.8
DiHS-NS mean + 2SD
n
4.1 3.7 1.8 1.4 1.6
48 29 30 31 31
0.01–0.4
50
Tomatsu S et al. [19].
for different age groups extracted from published spectrophotometric methods [16–18] are shown in Table 3B. Note that the sum of CS, DS and HS concentrations measured by UPLC–MS/MS was similar in each age group to the total GAG measured by spectrophotometric methods, especially by the uronic acid [16] and DMB [18] methods. The reference ranges for the dimers generated by enzymolysis [19], especially diDS and DiHS-NS, were noticeably lower than those from methanolysis, even if one assumes that most of the controls for the enzymatic method were adults. This discrepancy may be accounted for by differences in the calibration method and by the possibility that enzymes are unable to degrade GAG as completely as methanolysis. 3.3. GAG concentrations in patients with MPS In Table 4, we compared the urinary GAG CS, DS and HS in patients with various types of MPS and other storage disorders analyzed using the new UPLC–MS/MS method with the total GAG assayed using dimethylmethylene blue (DMB) spectrophotometry. The concentrations of DS and HS in urine specimens from all patients with MPS-I and MPS-II were well above the upper limits of agematched controls. The concentration of HS in patients with MPS-III ranged from 968 g/mol creatinine for a patient with MPS IIIA, to 32 g/mol creatinine for a patient with MPS IIIB. It is noteworthy that for several of these patients, the total GAG spectrophotometric assay returned values below or close to the reference range limits (Table 4, P9, P6, P10, P8, P32 and Table 5, P22B, P22C) and may have missed the diagnoses. The concentrations of both CS and DS in one of the patients with MPS VI (P5, Table 4) were markedly higher than the age-matched reference limits, while in another patient with MPS VI (P8) they were mildly though measurably higher than the agematched reference limits. The concentration of HS was not elevated in either of the patients with MPS VI. Seven patients with MPS-IVA (P28–P34) showed modest but clear elevations of CS, but not DS or HS, consistent with the expected elevation of chondroitin-6-sulfate in
Table 3B Published reference intervals or up limits for total urinary GAG determined by spectrophotometric- based methods. Units are g/mol creatinine. Age (years)
b1 1–3 4–17 N18 a b c
Total GAG Uronic acid testa
Alcian blueb
Dimethylene bluec
22–50 16–21 6.0–15 3.6
20 9.2 9.2 4.2
36 16 4.9–9.2 3.2
Stone JE [16]. Dembure PP et al. [17]. Jong JG et al. [18].
126
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Table 4 Urinary CS, DS and HS concentrations in patients with various MPS disorders. Urinary CS, DS and HS concentrations with age-matched reference limits (M + 2SD) determined by methanolysis UPLC–MS/MS in patients with MPS disorders (untreated) and comparison with the total GAG concentration measured by spectrophotometry. Units are g/mol creatinine. Values in bold represent elevations relative to the reference ranges. Disorder
Subject
Age (yrs)
UPLC–MS/MS
Spectrophotometry
CS
MPS I MPS I MPS II MPS II MPS IIIA MPS IIIA MPS IIIA MPS IIIA MPS IIIB MPS IIIB MPS IIIC MPS IIIC MPS IIIC MPS VI MPS VI MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA MPS IVA α-Fucosidosis α-Mannosidosis Aspartylglucosaminuria Sialuria Pompe Pompe
P3 P13 P9 P11 P1 P2 P7 P12 P22 P27 P4 P6 P10 P5 P8 P28 P29 P30 P31 P32 P33 P34
1 0.8 3 0 12 10 1 8 13 5 8 9 7 1 41 7 3 6 3 6 5 7 6.5 24 49 12 0.6 5.9
DS
HS
Total GAG
Patient
Control
Patient
Control
Patient
Control
Patient
Control
59 86 15 35 5.2 6.0 19 6.1 5.6 4.5 5.7 3.9 6.0 79 8.0 26 33 21 28 26 16 19 13 1.1 22 4.8 13 7.3
23 44 23 44 8.2 8.2 23 10 8.2 10 10 10 10 23 3.3 10 23 10 23 10 10 10 10 3.3 3.3 8.2 44 10
148 177 38 88 2.1 2.1 4.0 2.2 1.6 1.0 2.2 2.0 2.1 103 7.2 3.0 3.0 2.0 2.2 1.8 1.8 2.3 1.9 0.8 6.2 1.4 1.0 0.9
3.7 4.1 3.7 4.1 2.5 2.5 3.7 2.3 2.5 2.3 2.3 2.3 2.3 2.3 2.4 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.3 2.4 2.4 2.5 4.1 2.3
200 417 139 416 405 360 968 599 50 32 421 419 345 2.5 0.9 2.0 2.7 1.3 1.5 1.5 0.7 1.4 3.5 1.0 6.8 1.4 3.0 1.3
3.7 4.1 3.7 4.1 1.4 1.4 3.7 1.8 1.4 1.8 1.8 1.8 1.8 3.7 1.6 1.8 3.7 1.8 3.7 1.8 1.8 1.8 1.8 1.6 1.6 1.4 3.7 1.8
90 209 17 60 21 18 67 34 22 20 19 11 11 95 8.3 22 46 25 34 11 20 31 ND ND ND ND ND ND
31 31 16 31 10 12 31 12 10 16 12 12 12 31 6.5 12 16 12 16 12 16 12 N/A N/A N/A N/A N/A N/A
to a lack of sensitivity for the spectrophotometric method, which targets uronic acid moieties within the GAG polymer, some of which may be inaccessible to the reagent.
this condition. Therefore, the UPLC–MS/MS method was able to discriminate patients with MPS-III, MPS-IVA and MPS-VI from those with MPS-I and MPS-II, and from controls. Patients with other lysosomal storage conditions, including sialuria, alpha-fucosidase deficiency, alpha-mannosidase deficiency, aspartylglucosaminuria, Pompe disease (acid maltase deficiency) are included in Table 4 as disease controls. Most of these gave GAG values within the control ranges, however, the patient with aspartylglucosaminuria had elevations of all three targeted GAGs, for reasons that are not clear. To test the short-term fluctuation of HS excretion in a patient with MPS III, five urine samples (P#22–26) were collected from a 13 year-old patient with MPS IIIB over a period of nine days. The results, shown in Table 5, show that the HS concentration varied from 50 to 85 g/mol creatinine with a mean value of 71 and CV of 22% over this period and was far in excess of the upper limit of the age-matched controls (1.5 g/mol creatinine). By contrast, the variation in the total GAG spectrophotometric assay was from 9.6 to 31 g/mol creatinine with a mean value of 21 g/mol creatinine and a CV of 46%. The mean value was only twice that of the upper control limit (10 g/mol creatinine). We ascribe these differences
4. Discussion Historically, urine GAG measurement has been performed by quantitative urine GAG tests that provide a numeric representation of the total GAG level but do not differentiate between the individual GAG and qualitative tests that chromatographically separate the different GAG species but do not specifically quantify them. In many clinical laboratories, the DMB spectrophotometric method is used as a preliminary screen to quantify total GAG, where GAG sulfates including CS, DS, HS and KS form a color complex with DMB. Diagnostic laboratories are therefore typically running two separate tests on the same sample, one a total GAG assay and the other a separation method based on electrophoresis to determine which GAG species may be elevated. It is widely known that these testing strategies can have false positives and, more disconcertingly, false negatives. The analytical method for GAG
Table 5 Urinary CS, DS and HS concentrations (UPLC–MS/MS) and total GAG concentration (spectrophotometry) in five urine samples collected from a patient with MPS IIIB over nine days (subject P22, age 13 yrs). Units are g/mol creatinine. Values in bold represent elevations relative to the reference ranges. Patient
UPLC–MS/MS
Spectrophotometry
CS
P22A P22B P22C P22E P22F
DS
HS
Total GAG
Patient
Control
Patient
Control
Patient
Control
Patient
Control
5.6 4.7 6.1 5.7 6.2
8.2 8.2 8.2 8.2 8.2
1.6 1.5 1.7 1.7 1.7
2.5 2.5 2.5 2.5 2.5
50 74 85 56 66
1.4 1.4 1.4 1.4 1.4
22 13 9.6 31 29
10 10 10 10 10
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
described here clearly has much greater sensitivity and specificity than current methods employed in clinical laboratories to detect and quantify signals specifically derived from DS and HS, the principal accumulating storage materials in the lysosomes of patients with MPS I, MPS II, MPS III and MPS VI. Although this method did not produce a specific marker from KS that enabled differentiation of patients with MPS-IV from controls, all the patients with MPS-IVA included in this study had elevated CS compared with age-matched controls. In fact, UPLC–MS/MS affords greater separation between affected patients and age-matched controls than any previously published method that we are aware of. It should be noted that the CS concentration determined by this method in patients with elevated values of DS may be over-estimated because of the contribution of methanolyzed GlcA–-GalNAc dimers from DS, which consists of both GlcA–GalNAc and IdoA–GalNAc repeating disaccharide units, while CS consists of a GlcA-GalNAc repeating unit [20]. Thus, the presence of IdoA in DS distinguishes it from CS and provides a specific marker for DS. When compared with the previously reported enzymolysis-based LC–MS/MS method, methanolysis apparently yields significantly higher concentrations of HS and DS, presumably because both sulfated and non-sulfated GAG units are detected after derivatization. Unlike enzymolysis-LC–MS/MS or spectrophotometry, the method of quantification is based on stable isotopedilution, which provides a level of accuracy and precision not readily achievable by alternative analytical methods. This new method may thus afford a lower false negative rate by means of the improved sensitivity and specificity to detect and quantify DS and HS separately from CS. It is the only published method we are aware of that provides urine control ranges for all three GAGs in different age groups. The assay relies for its accuracy and precision upon the reproducibility of the methanolysis step, which must be performed with fresh reagent and with careful temperature and time control of the reaction. Further advantages of the method are its short turn-around time and its suitability for batched analysis of multiple samples. As previously stated, this method apparently lacks the ability to detect keratan sulfate, specific for MPS IVA and B (Morquio syndrome). However, patients with MPS type IVA are known to accumulate chondroitin-6sulfate(C-6S) as well as KS. Although methanolysis cannot distinguish between C-4S and C-6S, all urine samples from known MPSIVA patients that we have analyzed thus far have shown an elevation of CS without elevated DS or HS, which clearly distinguishes them from other MPS and from age-matched controls. Chondroitin-6sulfate has recently been shown to be a novel biomarker for both MPS IVA and MPS VII [21]. A specific test for KS in urine that uses LC–MS/MS to target a dimer produced by enzymatic degradation [22,23] is now available. We propose to use this assay as a second tier test either to confirm a presumptive positive screen for MPSIVA or to rule out MPS-IVB if the primary screening test is negative. Our experience suggests that patients with certain other storage diseases may have elevated urine GAG, so the test is probably not 100% specific for MPS. In summary, the method described here not only improves the sensitivity of current testing strategies, but also uses a single test to define which GAG species are increased and their relative levels. This combination should save time in diagnostic labs, allowing two tests to be combined into one, and offers a more reliable alternative to the older screening methods. The protocol for the assay is straightforward and can readily be incorporated into clinical laboratories that are equipped with MS/MS for the purpose of analyzing acylcarnitines and other disease biomarkers. It is important to understand that this test is intended for screening patients at risk for MPS. Following a presumptive positive screen by this test, further diagnostic testing by enzymology and/or molecular testing is warranted to confirm the specific diagnosis in the patient. It remains to be seen if the ability to quantify specific GAG species in urine will provide further benefit when used to monitor patients on enzyme
127
replacement or other therapies, as has already been demonstrated from the study of plasma GAG [7] and CSF GAG [13]. Conflict of interest The authors declare that they have no conflicts of interest. Acknowledgments The authors are grateful for the financial support from Genzyme Corporation, a Sanofi Company and from BioMarin Pharmaceuticals. We thank Dr. Steve Skinner and Fran Anese for their help in obtaining urine samples on known patients and also Laura Pollard and Teresa Thompson from the Greenwood Genetic Center Biochemical Genetics lab for their help in sample preparation. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.ymgme.2014.09.009. References [1] E. Neufeld, J. Muenzer, The mucopolysaccharidoses, in: C.R. Scriver, A.L. Beaudet, W.S. Sly, D. Valle (Eds.), The Metabolic and Molecular Bases of Inherited Disease, 8th ed.McGraw Hill, New York, 2001, pp. 3421–3452. [2] J. Muenzer, The mucopolysaccharidoses: a heterogeneous group of disorders with variable pediatric presentations, J. Pediatr. 144 (2004) S27–S34. [3] H. Zhou, P. Fernhoff, R.F. Vogt, Newborn bloodspot screening for lysosomal storage disorders, J. Pediatr. 159 (2011) 7–13. [4] J.G. de Jong, R.A. Wevers, R. Liebrand-van Sambeek, Measuring urinary glycosaminoglycans in the presence of protein: an improved screening procedure for mucopolysaccharidoses based on dimethylmethylene blue, Clin. Chem. 38 (1992) 803–807. [5] R.W. Farndale, C.A. Sayers, A.J. Barrett, A direct spectrophotometric microassay for sulfated glycosaminoglycans in cartilage cultures, Connect. Tissue Res. 9 (1982) 247–248. [6] G. Panin, S. Naia, R. Dall'Amico, L. Chiandetti, F. Zachello, C. Catassi, L. Felici, G.V. Coppa, Simple spectrophotometric quantification of urinary excretion of glycosaminoglycan sulfates, Clin. Chem. 32 (1986) 2073–2076. [7] S. Tomatsu, T. Shimada, R.W. Mason, A.M. Montano, J. Kelly, W.A. LaMarr, F. Kubaski, R. Giugliani, A. Guha, E. Yusada, W. Mackenzie, S. Yamaguchi, Y. Suzuki, T. Orii, Establishment of glycosaminoglycan assays for mucopolysaccharidoses, Metabolites 4 (2014) 655–679. [8] S.L. Ramsay, P.J. Meikle, J.J. Hopwood, Determination of monosaccharides and disaccharides in mucopolysaccharidoses patients by electrospray ionisation mass spectrometry, Mol. Genet. Metab. 78 (2003) 193–204. [9] M. Fuller, T. Rozaklis, S.L. Ramsay, J.J. Hopwood, P.J. Meikle, Disease-specific markers for the mucopolysaccharidoses, Pediatr. Res. 56 (2004) 733–738. [10] T. Ogama, S. Tomatsu, A.M. Montano, O. Okazaki, Analytical method for the determination of disaccharides derived from keratan, heparan and dermatan sulfates in human serum and plasma by high-performance liquid chromatography/turbo ionspray ionization tandem mass spectrometry, Anal. Biochem. 368 (2007) 79–86. [11] C. Auray-Blais, P. Bherer, R. Gagnon, S.P. Young, H.H. Zhang, Y. An, J.T. Clarke, D.S. Millington, Efficient analysis of urinary glycosaminoglycans by LC–MS/MS in mucopolysaccharidoses type I, II and VI, Mol. Genet. Metab. 102 (2011) 49–56. [12] C. Auray-Blais, P. Lavoie, H. Zhang, R. Gagnon, J.T. Clarke, B. Maranda, S.P. Young, Y. An, D.S. Millington, An improved method for glycosaminoglycan analysis by LC–MS/ MS of urine samples collected on filter paper, Clin. Chim. Acta 413 (2012) 771–778. [13] H. Zhang, S.P. Young, C. Auray-Blais, P.J. Orchard, J. Tolar, D.S. Millington, Analysis of glycosaminoglycans in cerebrospinal fluid from patients with mucopolysaccharidoses by isotope-dilution ultra-performance liquid chromatography–tandem mass spectrometry, Clin. Chem. 57 (2011) 1005–1012. [14] H. Zhang, S.P. Young, D.S. Millington, Quantification of glycosaminoglycans in urine by isotope-dilution liquid chromatography–electrospray ionization tandem mass spectrometry, Curr. Protoc. Hum. Genet. 17 (2013) 12. [15] H. Husdan, A. Rapoport, Estimation of creatinine by the Jaffe reaction: a comparison of three methods, Clin. Chem. 14 (1968) 222–238. [16] J.E. Stone, Urine analysis in the diagnosis of mucopolysaccharide disorders, Ann. Clin. Biochem. 35 (1998) 207–225. [17] P.P. Dembure, J.E. Drumheller, S.M. Barr, L.J. Elsas, Selective urinary screening for mucopolysaccharidoses, Clin. Biochem. 23 (1990) 91–96. [18] J.G. de Jong, R.A. Wevers, C. Laarakkers, B.J. Poorthuis, Dimethylmethylene bluebased spectrophotometry of glycosaminoglycans in untreated urine: a rapid screening procedure for mucopolysaccharidoses, Clin. Chem. 35 (1989) 1472–1477. [19] S. Tomatsu, A.M. Montano, T. Oguma, V.C. Dung, H. Oikawa, T.G. de Carvalho, M.L. Gutierrez, S. Yamaguchi, Y. Suzuki, M. Fukushi, N. Sakura, L. Barrera, K. Kida, M.
128
H. Zhang et al. / Molecular Genetics and Metabolism 114 (2015) 123–128
Kubota, T. Orii, Dermatan sulfate and heparan sulfate as a biomarker for mucopolysaccharidosis I, J. Inherit. Metab. Dis. 33 (2010) 141–150. [20] J.M. Trowbridge, R.L. Gallo, Dermatan sulfate: new functions from an old glycosaminoglycan, Glycobiology 12 (2002) 117R–125R. [21] T. Shimada, S. Tomatsu, E. Yasuda, R.W. Mason, W.G. Mackenzie, Y. Shibata, F. Kubaski, R. Giugliani, S. Yamaguchi, Y. Suzuki, K. Orii, T. Orii, Chondroitin 6-Sulfate as a Novel Biomarker for Mucopolysaccharidosis IVA and VII, JIMD Rep, 2014. (Epub ahead of print).
[22] L.A. Martell, R.L. Cunico, J. Ohh, W. Fulkerson, R. Furneaux, E.D. Foehr, Validation of an LC–MS/MS assay for detecting relevant disaccharides from keratan sulfate as a biomarker for Morquio A syndrome, Bioanalysis 3 (2011) 1855–1866. [23] S. Tomatsu, A.M. Montano, T. Oguma, V.C. Dung, H. Oikawa, T.G. de Carvalho, M.L. Gutierrez, S. Yamaguchi, Y. Suzuki, M. Fukushi, K. Kida, M. Kubota, L. Barrera, T. Orii, Validation of keratan sulfate level in mucopolysaccharidosis type IVA by liquid chromatography–tandem mass spectrometry, J. Inherit. Metab. Dis. 33 (2010) 35–42.
