Urine oligosaccharide tests for the diagnosis of oligosaccharidoses

Introduction

Lysosomes are organelles that contain different enzymes that are capable of hydrolysing biological polymers, such as proteins, DNA, RNA, polysaccharides and lipids. Mutations in the genes that encode these catabolic enzymes are responsible for lysosomal storage diseases (LSDs), a heterogeneous group of over 50 genetic disorders, which are characterized by the accumulation of undigested material in the lysosomes of affected individuals. Depending on the affected lysosomal enzyme, different catabolic intermediates, including glycans, lipids and proteins, can accumulate in tissues and organs. This accumulation ultimately leads to cellular dysfunction, causing progressive damage to multiple tissues, including the central nervous system, lung, heart, liver, spleen, kidney, joint and bones ( Casado et al., 2014; Sewell, 2008).

Oligosaccharidoses (or glycoproteinosis) are a subgroup of LSD caused by deficient glycoprotein degradation. In this catabolic pathway, the first common step is the digestion of the protein core by proteases. Then, the carbohydrate chains linked to the amino acids are sequentially degraded by a specific glycosidase, generally beginning at the nonreducing end of the oligosaccharide. The great majority of glycosidic groups of the digested glycoproteins are N-linked to the amino acid asparagine or O-linked to the amino acids serine or threonine and are composed of six sugars linked in one or two anomeric configurations: β-N-acetylglucosamine (βGlcNAc), α/β-N-acetylgalactosamine (α/βGalNAc), α/β-galactose (α/βGal), α/β-mannose (α/βMan), α-fucose (αFuc) and α-sialic acid (αSia). Figure 1 shows the mechanism by which the lysosomal enzymes degrade an N-linked glycan. The lack of any enzyme inhibits this catabolic pathway and results in the accumulation of non-degraded oligosaccharides in the lysosomes and body fluids. The composition of these non-degraded chains depends on the specific enzyme deficiency; therefore, these chains are good biomarkers for the rapid identification of the candidate gene that causes the disease. Oligosaccharidoses and related disorders that also present abnormal urine oligosaccharide profiles are shown in Table 1. These related disorders are other LSDs that are caused by enzymatic defects in the degradation of the glycosidic groups on different biomolecules, such as gangliosides or mucolipids. Some glycogen storage diseases (GSDs) may also present altered urine oligosaccharide profiles ( Sewell, 2008).

Figure 1:

Catabolic pathway of N-linked oligosaccharides.

The biochemical diagnosis of the diseases listed in Table 1 relies on the detection of abnormal oligosaccharides in random urine samples from a patient. Usually, non-pathological urine samples contain small amounts of some oligosaccharides. The major excreted oligosaccharide is the tetrasaccharide 6-α-D-glucopyranosyl-maltotriose, which is derived from the intravascular degradation of glycogen under physiological conditions ( Sluiter et al., 2012). This tetrasaccharide is present in urine samples from healthy populations, and its excretion decreases with increasing age. The urine from breast-milk-fed children, particularly children younger than 1 month old, contains a range of other oligosaccharides derived from human breast milk ( De Leoz et al., 2013). Some infant formulas add a large variety of galacto- or fructooligosaccharides that can also be excreted in the urine. Under these conditions, differential diagnosis may be difficult. However, urine samples from patients with oligosaccharidoses and other disorders displaying oligosacchariduria present characteristic pattern of alterations, with abnormal oligosaccharides that are not present in non-pathological samples. Early clinical and biochemical recognition is crucial to patient prognosis because the window for an effective intervention, such as bone marrow transplantation and enzyme replacement therapies, may be very narrow after the onset of symptoms. Moreover, prenatal diagnosis may be of paramount importance.

The most widely used method for the biochemical screening of oligosaccharidoses is the qualitative analysis of the urinary oligosaccharides by thin-layer chromatography (TLC). However, this method is not sensitive, and it is time-consuming (three working days), making it impossible to apply it for the screening of large series of patients. For these reasons, biochemical screening methods have been improved, and in the last few years, different procedures based on diverse analytical technologies have been developed that are able to detect oligosaccharides that accumulate in urine, including high-performance liquid chromatography (HPLC) and, more recently, mass spectrometry (MS) and capillary electrophoresis (CE). Here, we will review traditional and new approaches for the screening of oligosaccharidoses and related disorders that present oligosacchariduria.

Thin-layer chromatography (TLC)

This silica gel method was described several decades ago for the screening of urine oligosaccharides from patients with aspartylglucosaminuria ( Palo & Savolainen, 1972). Three years later, a group reported data from patients with fucosidosis, mannosidosis and GM1 gangliosidosis using this method and described their pathological urinary oligosaccharide profiles ( Humbel & Collart, 1975). Further modifications were introduced to expand the range of pathological profiles and disorders detected by this approach ( Friedman et al., 1978; McLaren & Ng, 1979; Sewell, 1979).

TLC is performed on 20-cm-long silica gel plates. Urine samples are applied using a glass capillary and dried in a current of warm air. Standards, such as raffinose, lactose, sialyl lactose and glucotetrasaccharide, can be used. The plate is generally developed over 6 h in the first solvent, a freshly prepared mixture of n-butanol, glacial acetic acid and deionized water. On the next day, the plate is dried in warm air, and this development procedure is repeated twice more in the same solvent on two successive days. There are variations of this procedure that use different organic solvents for the second and third development processes, such as a mixture of nitromethane, n-propanol and deionized water ( Sewell, 2008). The plate is finally removed, dried and sprayed with a freshly prepared solution of orcinol in sulphuric acid. The oligosaccharide bands are then visualized by heating at 100°C for 10 min. In hot acid, orcinol reacts with the oligosaccharides, producing colour-stained bands that are detected by visual examination. TLC images have been reported elsewhere ( Peelen, de Jong & Wevers, 1994).

Only weak bands below the raffinose standard are usually detected in normal urine. In breast-milk-fed children, many bands are observed below the raffinose standard due to the presence of galacto- and fructooligosaccharides in human breast milk as well as in some fortified infant milk formulas. In these cases, the oligosaccharide pattern could be confused with a pathological profile, and repeated testing is recommended when the children decrease their milk intake, at ages from 6 months to 1 year. The same pattern can be observed in urine from pregnant or lactating women.

Pathological TLC patterns specific for particular diseases are well described. In 2008, Sewell described the profiles of sialidosis, GM1 gangliosidosis, α-mannosidosis, fucosidosis, GM2 gangliosidosis type Sandhoff and aspartylglucosaminuria. In patients with sialidosis, a densely stained band close to the origin and a weaker band that migrated further are the characteristic profile. Patients with infantile GM1 gangliosidosis show heavily stained bands close to the origin, corresponding to an octasaccharide, and an additional band in the pentasaccharide region. Patients with α-mannosidosis excrete a series of mannose-containing oligosaccharides (disaccharide to heptasaccharide). Patients with fucosidosis excrete a characteristic brown/pink band and densely stained bands at the origin. In Sandhoff disease, again, dense staining is observed at the origin, and urine from patients with aspartylglucosaminuria shows the presence of bands migrating close to the origin. Several types of GSDs show an abnormal profile by TLC, with a high excretion of a characteristic glucotetrasaccharide derived from accumulated glycogen. These pathological patterns on TLC have been described for GSD type 2 ( Blom et al., 1983), type 3 ( Galvin-Parton & Hommes, 1996) and type 4 ( Sewell, 1986).

TLC is the most widely used procedure for screening oligosaccharidoses, but even under the best circumstances, it has important limitations. It is not always very sensitive, so patients with juvenile forms of some oligosaccharidoses with slight phenotypes may be missed. In addition, it is not automatable, making it impossible to screen a large number of patients. Moreover, the between-run variations in this procedure may be significant.

High-performance liquid chromatography (HPLC)

HPLC technology is also used in the screening of oligosaccharidoses. In 1994, an HPLC method for analysis of oligosaccharides in urine was developed by Peelen, de Jong, and Wevers (1994). In this procedure, urine samples were previously deproteinized. Then, oligosaccharides were separated by high-pH anion exchange chromatography, using solutions of sodium hydroxide and sodium acetate as eluents. Different gradients were developed for monosaccharide or oligosaccharide separation. Separated oligosaccharides in the eluent were detected by pulsed amperometric detection and by a post-column derivatization reaction with 4-aminobenzoic acid hydrazide and posterior UV detection at 400 nm. In this reference, abnormal oligosaccharide profiles of α-mannosidosis, GM1-gangliosidosis, GM2-gangliosidosis type Sandhoff, Pompe disease and β-mannosidosis were established and compared with the TLC patterns. HPLC offers improved resolution and the possibility of automatization.

Capillary electrophoresis (CE)

CE has a high resolution efficiency and offers excellent analytical sensitivity when coupled to a laser-induced fluorescence detector (CE-LIF). Some methods based on CE-LIF have been developed to analyse N-linked oligosaccharides that are enzymatically released from glycoproteins (Chen & Evangelista, 1998; Guttman, Chen & Evangelista, 1996; Mechref, Muzikar & Novotny, 2005; Ruhaak et al. 2010a), These methods are able to resolve very closely related oligosaccharide structures and even isomers, which are indistinguishable by tandem MS (MS/MS) or matrix-assisted laser desorption/ionization (MALDI)-time of flight (TOF) procedures. Recently, a CE-LIF-based procedure was developed and validated for diagnosis of oligosaccharidoses and related disorders ( Casado et al., 2014).

In this study, analysis of the urine oligosaccharides was standardized for the first time using CE-LIF detection with a 488-nm argon ion laser module, adapting the commercially available Carbohydrate Labelling and Analysis Kit (Beckman-Coulter) (details of the procedure are available in the kit insert). Because oligosaccharides do not exhibit native fluorescence, derivatization is required before LIF detection. The derivation reagent used was 8-aminopyrene-1,3,6-trisulphonate (APTS), which reacts with the reducing glycocompounds and is commonly used in glycomic studies. In most of the MS procedures used for glycan analysis, derivatization is often necessary because oligosaccharides display a poor ionization efficiency ( Sowell & Wood, 2011). In addition to this derivatization reaction, solid-phase extraction (SPE) is frequently required in MS procedures ( Bruggink et al., 2012). Thus, the time and labour required for sample preparation and labelling for the CE procedure are similar to those of various MS procedures.

A solution of semi-hydrolysed dextran containing linear glucose polymers from monomer glucose to oligosaccharides with 15 glucose units is used as the ladder standard (Figure 2). The peaks of the ladder standard may be used as size reference markers and represent the degree of polymerization. All analyses are performed on random urine samples, if possible, the first morning urine samples. A urine volume containing 150 nmol of creatinine is concentrated by heating at 50°C in an oven for 5 h. Then, 5 µl of the concentrated urine samples or the dextran ladder standard solution are mixed with the APTS solution and sodium cyanoborohydride in tetrahydrofuran. The mixed solutions are reacted overnight in the dark on a 37°C heating block. Regarding the derivatization procedure, mild labelling conditions are used to avoid the loss of sialic acid residues from the oligosaccharide chains ( Casado et al., 2014). An overnight derivatization reaction at 37°C is chosen for this purpose. These derivatization conditions are commonly used for the reductive amination of glycans ( Chen & Evangelista, 1995). After derivatization, these solutions undergo dilution in deionized water before they are injected onto the CE apparatus for analysis. The CE experiments were performed using a Beckman P/ACE MDQ system equipped with a 488-nm argon ion laser module (Beckman-Coulter, Fullerton, CA, USA), but other CE systems are suitable for this purpose. The compounds are separated in neutrally coated N-CHO capillaries (for example, a 50-µm I.D., a 60.2-cm total length and a 50-cm distance to the detector). The capillary cartridge must be maintained at 20°C using liquid coolant to avoid changes in the retention time of the glycocompounds. All of the new capillaries are conditioned, and before each injection, the capillary is rinsed as for other CE procedures ( Casado et al., 2014). The sample is injected and a separation voltage of 30 kV is applied with reverse polarity, which results in an electrophoretic current of 14 µA. The vials of buffer are replaced with new fresh buffer after 10 runs, which avoid changes in the retention time between each run of analyses.

Figure 2:

Urinary oligosaccharide profiles by CE-LIF.

(A) Representative non-pathological profile from children ≤1 year old. (B) Representative non-pathological profile from children >1 year old. (C) Fucosidosis profile. (D) α-Mannosidosis profile. (E) GM-1 gangliosidosis profile. (F) GM-2 gangliosidosis-type Sandhoff profile. (G) GSD type 2 (Pompe disease) profile. (H) GSD type 3 profile. (I) Galactosialidosis profile. In electrophoregrams of pathological urines, the upper trace shows the urinary profile from the patient. The oligosaccharide excretion pattern for each disease is compared with an age-matched control (middle trace). Arrows show abnormal oligosaccharides. The lower trace in each electrophoregram shows the separation of the maltooligosaccharide ladder standard, and numbers indicate the polimeration degrees (number of the glucose residues) in the corresponding ladder oligosaccharide (1: glucose, 2: maltose, 3: maltotriose, …, n: Glucose_n)

The oligosaccharide profiles of healthy controls and individuals with a disease have been documented using CE-LIF ( Casado et al., 2014). Urine samples from control subjects (from 1 week to 16 years old) and from patients diagnosed with different lysosomal diseases (fucosidosis, α-mannosidosis, GM1 gangliosidosis, GM2 gangliosidosis type Sandhoff, GSD type 2, GSD type 3, aspartylglucosaminuria, Schindler disease and galactosialydosis), and samples included in the oligosaccharide educational program from the ERNDIM Quality Control Scheme are presented.

Examples of non-pathological urine samples are presented in Figure 2. Typical electropherograms display high excretion of monosaccharides and several disaccharides ( Casado et al., 2014). After these peaks, a peak corresponding to the oligosaccharide 6-α-D-glucopyranosyl-maltotriose (Glc4), which is derived from the intravascular degradation of glycogen, is observed. In very young infants (age <1 month), several peaks in the trisaccharide to pentasaccharide area may be detected in most cases (Figure 2A and Figure 2B), which disappear as the children decrease their milk intake.

Regarding the pathological samples, patients with fucosidosis, α-mannosidosis, GM1 gangliosidosis, GM2 gangliosidosis type Sandhoff, GSD type 2, GSD type 3 and galactosialydosis clearly show impaired profiles (Figure 2C–I).

Patients with fucosidosis accumulate reducing oligosaccharides, which are excreted in urine. The profile obtained by CE-LIF (Figure 2C) shows an abnormal peak with an electrophoretic mobility corresponding to the hexa-heptasaccharide in the ladder standard corresponding to the hexasaccharide, Fuc-Gal-GlcNAc-Man2-GlcNAc, which is excreted in the urine of patients with fucosidosis ( Ramsay et al., 2005; Xia et al., 2013).

Patients with α-mannosidosis accumulate high-mannose-containing oligosaccharides (Mann-GlcNAc, n≥2), which are excreted in urine ( Michalski & Klein, 1999). The pattern obtained by CE-LIF (Figure 2D) was similar to previously described patterns ( Klein et al., 1998; Xia et al., 2013) and had a peak in the trisaccharide area (Man2-GlcNAc), as several smaller peaks for Man3-5-GlcNAc.

Patients with GM1 gangliosidosis present an increase in the excretion of galactosylated oligosaccharides ( Sluiter et al., 2012) (eight oligosaccharides with 3 to 12 monosaccharide units) (Figure 2E). This profile was similar to the results obtained using MS procedures, with substantial excretion of the octasaccharide Gal2-GlcNAc2-Man3-GlcNAc ( Xia et al., 2013).

Patients with Sandhoff disease present a urine oligosaccharide pattern consisting of eight different oligosaccharides ( Sandhoff & Harzer, 2013), with migration times ranging between tetra- and hexasaccharides (Figure 2F), while in MS procedures, only three oligosaccharides have been described ( Xia et al., 2013). These differences in the excretion profiles may be due to the capacity of CE to resolve positional isomers with the same molecular weight.

Patients with GSD type 2 (Pompe disease) accumulate glycogen in the lysosomes. In this disease, the intravascular degradation of accumulated glycogen causes an increased urinary excretion of the tetrasaccharide 6-α-D-glucopyranosyl-maltotriose (Glc4). Additionally, urinary Glc4 is a biomarker for monitoring the progression and outcome of the disease after the enzyme replacement therapy ( Sluiter et al., 2012; Young et al., 2009). CE-LIF analysis showed a remarkable increment in the excretion of Glc4 (Figure 2G). Other small amounts of oligosaccharides (penta- to heptasaccharides) were also detected. Increased excretion of the tetrasaccharide Glc4 has also been described in patients with GSDs ( Sluiter et al., 2012), as we state in Figure 2H. Other greater oligosaccharides (penta- to octasaccharides) were also detected.

Galactosialidosis is a rare LSD that is caused by a combined deficiency of GM1 β-galactosidase (β-gal) and neuraminidase, which is secondary to a defect in a lysosomal enzyme/protective protein cathepsin A and a mutation in the CTSA gene. A clearly impaired profile was detected in a case with genetic confirmation of the disease (Figure 2I), corresponding to endo-beta-N-acetylglucosaminidase-cleaved products of sialylated N-glycans, O-sulphated oligosaccharide moieties and other carbohydrate moieties with reducing-end hexose residues (Bruggink et al., 2010).

Importantly, it should be noted that pathological oligosaccharides usually have branched structures, whereas the oligosaccharides in the ladder have linear structures. This difference causes slight discrepancies in their relative migration times because the tendency of linear structures to coil may result in higher migration times compared with the branched oligosaccharides ( Guttman & Pritchett, 1995).

A limitation of this CE-LIF procedure for urinary oligosaccharide analysis is that the defects that excrete non-reducing glycocompounds in urine, such as aspartylglucosaminuria and Schindler disease, cannot be detected using this procedure. The glycocomponus excreted in these two diseases cannot be derivatized by the APTS reagent and, thus, no fluorescence signal is expected ( Casado et al., 2014; Michalski & Klein, 1999).

MS techniques

Other recent technological advances in carbohydrate analysis have enabled the rapid and accurate characterization of LSDs. Many of these technological developments include MS techniques, which have evolved as a powerful tool for glycan analysis. MS is an important tool for the structural analysis of oligosaccharides and offers precise results, analytical versatility, very high sensitivity, high precision and high speed. MS is tolerant of mixtures and is a natural choice for the analysis of this class of molecules.

Oligosaccharides must be ionized for MS analysis. The most frequently applied ionization methods for oligosaccharide analyses are electrospray ionization (ESI) ( Zhang & Linhardt, 2009) and MALDI ( Xia et al., 2013), which are performed in positive and negative ionization modes. The detailed structures of the oligosaccharide molecules are obtained by MS/MS ( Kailemia et al., 2014). The use of a powerful mass analyser with MS/MS (triple quadrupole, TOF/TOF) or MSⁿ (ion trap) capabilities in conjunction with reverse-phase, normal phase, porous graphited carbon, size exclusion, ion exchange, liquid chromatographic or capillary electrophoretic separation methods has increased the use of MS in characterizing oligosaccharide structures. The main function of these coupled techniques is to reduce complexity by separating isobaric structures that are not resolved by MS.

ESI-MS/MS

In ESI-MS, the sample solution is highly charged and sprayed through a capillary into a strong electric field to form a fine mist of charged droplets. Generally, the sensitivity of ESI decreases as the mass of the oligosaccharides increases, and these analytes cause problems in MS due to their poor ionization efficiencies, but derivatization of the sample enhances MS sensitivity and supports detailed structural characterization by MS/MS. However, samples containing carbohydrate isomers with the same mass but different structural features might be not distinguished by this approach. MS is often combined with liquid chromatography to analyse oligosaccharides ( Peelen, de Jong, and Wevers 1994).

Physiological samples (urine, plasma, dried blood spot and amniotic fluid) must undergo derivatization, such as permethylation, which enables uniform ionization for acidic and basic oligosaccharides, reduces their polarity and improves the sensibility for neutral and acidic oligosaccharides ( Faid, Michalski & Morelle, 2008), to enhance the ionization of oligosaccharides in MS and facilitate the elucidation of their structures. Derivatization with 1-phenyl-3-methyl-5-pyrazolone (PMP) was used to analyse sialylated and neutral oligosaccharides. PMP improves the inherently poor ionization efficiency of oligosaccharides and their chromatographic retention and produces a common fragment with a mass-to-charge ratio m/z of 175 in all product ion scans, as shown in Figure 3. Identification of oligosaccharides using this ESI-MS/MS method could identify patterns in the following diseases: α-fucosidosis, α-mannosidosis, Pompe disease, gangliosidosis GM2 (Sandhoff disease) (these four diseases are shown in Figure 4), mucolipidosis type II and type III, I-cell disease, sialidosis, Gaucher disease, sialic acid storage disease and gangliosidoses GM1 and GM2 (Tay-Sachs disease) ( Ramsay et al., 2005; Sowell & Wood, 2011). Other different labelling strategies to enhance oligosaccharide ionization and to facilitate their subsequent quantification have been widely reported in the literature (Kapková, 2009; Ruhaak et al. 2010b; Sakaguchi et al., 2014; Volpi, 2010). Most of these derivatization reactions can be accomplished by coupling the reagents with the reducing end of the oligosaccharides. However, oligosaccharides could also be analysed without derivatization ( Verardo, Duse & Callea, 2009).

Figure 3:

ESI-MS/MS profile of a precursor ion scan of m/z 175 in positive ion mode, scan range m/z 475–1700, performed on the PMP-derivatized urine from a control.

Control samples (like many of the pathological urine samples) contain monosaccharides such as pentose, hexose (Hex), N-acetylhesoxamine, and uronic acid at m/z values 481.3 (data not shown), 511.5, 551.6, and 524.5 respectively. Various other oligosaccharides are observed, including the Hex2 to Hex4 series at m/z 672.7, 835.2, and 997.0, respectively; and sialyllactose at m/z 964.0. Many less intense signals are present in the control urines, and these varied substantially (this variation could be the result of diet, parenteral nutrition, pregnancy and drug use). MeLac served as an internal standard (m/z 686.7, in red).

Figure 4:

ESI-MS/MS profiles of precursor ion scan of m/z 175 in positive ion mode performed on the PMP-derivatized urine from patients with oligosaccharidurias. MeLac served as an internal standard (m/z 686.7, in red).

(A) Precursor 175 scan of PMP-derivatized urine from a patient with fucosidosis. The oligosaccharides present in the fucosidosis sample were fucosylated oligosaccharides at m/z 697.7 and 859.9 (in green), with composition Fuc-HexNAc and Fuc-Hex-HexNAc, respectively. (B) Precursor 175 scan of PMP-derivatized urine from a patient with α-mannosidosis. The oligosaccharides present in the α-mannosidosis sample were mannose-rich oligosaccharides of the (Man)nα-Manβ1-4GlcNAc series (m/z 713.7, 875.9, 1038.0, 1200.2, and 1363.4; in green). (C) Precursor 175 scan of PMP-derivatized urine from a patient with Pompe disease. The major oligosaccharide excretion present in the urine of a patient with Pompe disease was the glucose tetrasaccharide Glc4 (m/z 998.0 in green). (D) Precursor 175 scan of PMP-derivatized urine from a patient with Sandhoff disease. The oligosaccharides present in Sandhoff disease were oligosaccharides with terminal HexNAc residue and with the following compositions: Hex2-HexNAc2, Hex3-HexNAc2, Hex3-HexNAc3, and Hex3-HexNAc4 (m/z 1080.1, 1242.2, 1445.4, and 1647.6, respectively, in green).

During the derivatization reaction, salts, solvents and derivatization reagents are present in a large excess, and a clean-up procedure must be implemented to remove the interference and facilitate oligosaccharide detection. SPE has been widely employed for the purification of labelled oligosaccharides due to its rapid and simple operation. There are diverse stationary phases of SPE, and the two major types of stationary phases used are hydrophilic and carbon materials ( Zhang et al., 2014). Briefly, the carbon stationary phase shows good performance in the purification of underivatized oligosaccharides, but elution using trifluoroacetic acid 0.1% could cause the loss of sialic acid residues. In the hydrophilic stationary phase, the derivatized oligosaccharides are retained by hydrophilic interactions, and the excess derivatization reagents can be removed due to their lower hydrophilicity.

ESI-MS/MS is the most widely used analytical method for the determination of oligosaccharides in many newborn screening laboratories because it enables the quantification of oligosaccharides and allows the detection of oligosaccharidoses with relative ease. However, the derivatization procedures could possibly induce sample deterioration and contamination, thereby potentially decreasing its benefits. The need for improved methods to analyse oligosaccharides will serve as a driver for further developments in MS for the screening of oligosaccharidoses.

MALDI-TOF

MALDI-TOF MS (MALDI-TOF) is also contributing to the improvements in the diagnosis of oligosaccharidoses, particularly in comparison with the challenging TLC method ( Raymond & Rinaldo, 2013). Positive and negative ionization modes are used for the analysis of both native oligosaccharides (depending on the functional groups) and those present after chemical methylation (permethylation). Methylation is believed to provide uniform ionization for acidic and basic oligosaccharides, but the spectra are not easier to interpret. The mass-to-charge ratio (m/z) of the pseudomolecular ion of a specific oligosaccharide will change depending on the use of a derivatization step, the type of derivatizing agent and the type of adduct formed. Permethylation would improve the liquid chromatography analysis, but it is difficult to couple MALDI with online separation techniques. Therefore, offline purification methods are needed, mainly C18 and carbograph columns.

The samples are dissolved in an organic solvent mixed with a matrix, spotted on a MALDI target and dried. The dry mixture is irradiated using a laser, and finally, the oligosaccharides are ionized, assisted by the matrix. Several matrices are being studied because of the different effects on the type of ions generated. The majority of the ions are singly charged in negative and positive mode; however, adducts with sodium or potassium and protonated or deprotonated ions are also formed. It is necessary to consider that the energy imparted by the MALDI could fragment labile groups, such as sialic acid.

The use of the TOF as mass analyser derives an accurate mass measurement and a mass range >2000 m/z with high resolution (in contrast to quadrupole MS), and MS/MS, such as MALDI-TOF/TOF, can provide an oligosaccharide signature and contribute to the diagnosis of a specific pattern for each oligosaccharidosis.

Klein et al. (1998) described a MALDI-TOF method for analysing oligosaccharides in urine that could be used to identify oligosaccharidosis in four lysosomal defects (alpha- and beta-mannosidoses, galactosialidosis and GM1 gangliosidosis) and other oligosaccharidurias, such as Pompe disease. The authors also showed the complexity of the oligosaccharide profile, even in the urine of healthy controls, which depends on the diet (breast- or fortified-milk-fed neonates) or blood group. The oligosaccharides were derivatized with 8-aminonaphtalene-1,3,6-trisulphonic acid (ANTS), a fluorescent tag, and the obtained derivatives were analysed by electrophoresis and MALDI-TOF after purification with a SPE carbon column. 3-Aminoquinoline was used as the matrix and the ANTS-oligosaccharides were registered as [M-H]⁻ in negative ion mode, due to the sulphonic group. This method does not enable the diagnosis of aspartylglucosaminuria or Schindler disease, but in contrast to electrophoresis, it identifies the oligosaccharides by their molecular ions. In ambiguous cases, the digestion with selective exoglycosidades could help to achieve the correct identification. For alpha-mannosidosis, the major compound was the trisaccharide Hex2HexNAc, corresponding to Man2GlcNAc at m/z 911. The HexHexNAc at m/z 749 was a more important marker of beta-mannosidosis and corresponded to ManGlcNAc. The highest signals at m/z 1567 (NeuACHex3HexNAc2) and m/z 1803 (Hex5HexNAc3) were observed in galactosialidosis and GM1, respectively. Of course, these oligosaccharides are part of a profile in which other less intense ions must be observed.

Xia et al. (2013) was able to identify the oligosaccharide patterns of 11 different diseases by MALDI-TOF/TOF in a single analysis of urine samples (GM1 and GM2, alpha-mannosidosis, sialidosis, aspartylglucosaminuria, mucolipidoses II and III, galactosialidosis, fucosidosis and Pompe and Gaucher diseases). Oligosaccharides are permethylated after two-step offline purification (C18 and carbograph columns). The authors suggest that there may be a possibility of diagnosing aspartylglucosaminuria, in which glycoamino acids cannot be detected when the methods are based on the derivatization of the reducing end of the oligosaccharide. The protection of the terminal sialic acid from permethylation is also required. The authors studied control samples at seven different age ranges and found that the profiles were very similar. Finally, the authors identified 43 diagnostic profiles of 10 oligosaccharidoses and differentiated between subtypes of some of these diseases with severe or mild clinical presentations, suggesting that the method is a reliable tool for the first-tier screening of lysosomal disorders. Other diseases characterized by the excretion of oligosaccharides could be screened in the same analysis, such as glucosidase I deficiency (congenital disorder of glycosylation IIb).

Bonesso et al. (2014) analysed urine samples (10 µl) without any pretreatment to isolate oligosaccharides, taking advantage of the tolerance for salts in MALDI-TOF. It is well remarked that this method is much less laborious and shorter (30 min) than the method reported by Xia et al.

The authors were able to identify eight different oligosaccharidoses in previously diagnosed samples, which were used as training in a blind analysis series that included control samples (fucosidosis, aspartylglucosaminuria, sialidosis, GM1, Sandhoff disease, alfa-mannosidosis and mucolipidosis II and III).

The authors tried two different matrices, 2,5-dihydroxybenzoic acid and 3-aminoquinoline, and showed that the first matrix was better for the negative ion mode and the latter was better for the positive ion mode. Therefore, samples from each patient were analysed twice. Other matrices, such as 9-aminoacridine or 1,5-diaminonaphtalene, were studied and produced poor results. Neutral glycans were detected in positive mode as [M + Na]⁺ and [M + K]⁺; therefore, sodium salts were added to reduce the occurrence of sodium adducts and obtain an oligosaccharide profile that was easier to interpret. Sialylated oligosaccharides were analysed in negative ion mode as [M-H]⁻, but sodium and potassium adducts were also observed.

TOF/TOF was used to identify the more relevant peaks obtained in the MALDI-TOF spectra, bearing in mind that GlcNAc and GalNAc as well as Glc, Man, and Gal are indistinguishable and were only identified as N-acetylhexosamines and hexoses, respectively. However, asparaginyl and fucosyl residues could also be detected, contributing to the identification of the species studied. The authors suggest the use of deuterated standards and correcting for creatinine excretion to obtain quantitative data.

Conclusions

In summary, CE-LIF, ESI-MS/MS, and MALDI-TOF procedures used for the study of oligosaccharidoses and related diseases that present oligosacchariduria are powerful tools for the rapid identification and characterizations of these disorders. The characteristic patterns of urinary oligosaccharides in patients with different diseases are described. In all cases, the oligosaccharide profiles are strongly informative and showed abnormal compounds that are not present in any of the urine samples from the control subjects. These technologies are very good alternatives to the traditional screening test using TLC. These methods have great sensibility and resolution and are automatable, enabling extended screening of these diseases, both as a first step in diagnosis or for confirmation of the pathogenicity of mutations that may be detected with next generation sequencing techniques.