Original Article
Quantitative Assessment of Core Fucosylation for Congenital Disorders of Glycosylation
2024 Volume 13 Issue 1 Pages A0159
Details
2024 Volume 13 Issue 1 Pages A0159
Congenital disorders of glycosylation (CDG) include a group of diseases characterized by defects of N-glycan fucosylation. The analytical molecule of choice for the diagnosis of CDG affecting N-glycosylation is serum transferrin: approximately 10% of the glycans attached to transferrin are fucosylated via an α1,6 linkage at the innermost N-acetylglucosamine residue, termed “core fucosylation.” Isoelectric focusing (IEF) of transferrin is often used for diagnosis, but IEF is ineffective in detecting abnormal fucosylation. Here, we present mass spectrometry (MS) methods for detecting fucosylation disorders. First, the level of core fucosylation of the glycan attached to Asn630 of transferrin can be measured by the signal intensity ratio of tryptic peptide ions containing fucosylated and nonfucosylated biantennary oligosaccharides. The core fucosylation level at this glycosylation site in the 0- to 32-year-old group (n = 68) was 7.9 ± 1.7 (%, mean ± SD), and nearly null for SLC35C1-CDG caused by defects in the GDP-fucose transporter. More simply, fucosylation levels can be measured by quadrupole time-of-flight (QTOF) MS of intact transferrin. The fucosylation levels of intact transferrin measured by MS with a Q-mass analyzer, which is currently used as an instrumental standard for newborn screening for inborn errors of metabolism and has a lower resolution than the QTOF analyzer, correlated well with the values obtained by glycopeptide analysis. These methods, namely the analysis of glycopeptides or intact transferrin by Q MS, can also be used on dried blood spots and are expected to help facilitate the diagnosis of CDG affecting N-glycan fucosylation.
A large group of disorders caused by genetic abnormalities related to the synthesis of sugar chains of glycoproteins and glycolipids is called congenital disorders of glycosylation (CDG). The total number of causative genes of CDG exceeds 150, including those responsible for inherited glycosylphosphatidylinositol deficiency and dystroglycanopathies.1) In a narrow sense, CDG refers to the disorders in the synthesis of asparagine-linked oligosaccharides, or N-glycans, modifying proteins. Recently, the number of suspected cases of CDG has increased through genetic analysis to search for diseases of unknown cause,2) and the need for sugar chain analysis to confirm the diagnosis is increasing.3)
Serum transferrin is the first-line biomarker of CDG because it is easy to purify and has low heterogeneity in its sugar chain, which is composed of two biantennary oligosaccharides.3) For screening, isoelectric focusing (IEF) is still used to detect changes in the isoelectric point of transferrin resulting from a decrease in the number of N-acetylneuraminic acids due to oligosaccharide deletion or aberrant structure. On the other hand, mass spectrometry (MS) detects a change in the molecular mass due to a structural change of attached oligosaccharides. MS demonstrated its unique capabilities in 1992 by identifying oligosaccharide deficiencies in the transferrin molecule in patients with PMM2-CDG4) and is now widely used for the diagnosis of CDG.5) MS quantitatively assesses changes in oligosaccharide profiles in CDG patients.
Approximately 10% of the oligosaccharide of transferrin is fucosylated at the innermost N-acetylglucosamine residue through an α1,6 linkage.6) This type of fucose moiety is called core fucose and plays specific roles in the function and regulation of glycoproteins. Core fucosylation occurs in the Golgi apparatus and is catalyzed by the enzyme FUT8 using the donor substrate GDP-fucose.7) Additionally, genes involved in the synthesis of GDP-fucose in the cytoplasm, its transport to the Golgi apparatus, and the reutilization of fucose from lysosomes are all candidates for CDG. To date, FUT8-CDG, FCSK-CDG, SLC35C1-CDG, and GFUS-CDG have been reported to be associated with impaired fucosylation.8)
Fucose is a neutral sugar, and changes in fucosylation are not detected by IEF. Instead, liquid chromatography (LC) with spectrophotometric or mass spectrometric detection of N-glycans from whole serum glycoproteins is often used for glycoprofiling, including fucosylation levels.9) The main source of N-glycans in serum is IgG, which is largely fucosylated in healthy individuals. However, this approach is not suitable for screening because it requires multiple steps, that is, the release of glycans from glycoproteins and subsequent derivatization with chromophores. Meanwhile, electrospray ionization (ESI) MS of intact transferrin has also been used to diagnose CDG, but its application in detecting defective fucosylation has not been studied. Here, we present MS-based approaches to assess fucosylation levels of transferrin for CDG, with a particular focus on its potential for incorporation into current MS-based newborn screening (NBS) programs.
Blood or serum samples without personally identifiable information were obtained from the doctors in charge of the patients at the Osaka Women’s and Children’s Hospital (OWCH) for the diagnosis of CDG. This study has been approved by the institutional review board of OWCH (#844-3, 2018).
Immunopurification of transferrinImmunopurification was performed according to a previously reported method.10) Briefly, an affinity column was prepared using a rabbit (DAKO, Glostrup, Denmark) or a goat (Invitrogen, Thermo-Fisher Scientific, Waltham, MA, USA) polyclonal antibody against human transferrin and a ligand-coupling Sepharose column (HiTrap NHS-activated HP, GE Healthcare, Piscataway, NJ, USA), and the antibody-coupled Sepharose was then recovered from the column. A 10 μL portion of serum was mixed with a 20-μL slurry of the antibody-coupled Sepharose in 0.5mL of phosphate-buffered saline (PBS), and the solution was incubated at 4°C for 30min. After washing with PBS, transferrin was eluted from the Sepharose with 0.1M glycine–HCl buffer at pH 2.5. For carbamidomethylation, transferrin was dissolved in 0.5 mL of 6 M guanidine hydrochloride, 0.25 M Tris–HCl, pH 8.5, and reduced with 5 mg of dithiothreitol at 60°C for 30 min. Then, 10 mg of iodoacetamide was added, and the solution was incubated in the dark at room temperature for 30 min. The reagents were removed by a NAP-5 gel filtration column (GE Healthcare) equilibrated with 0.05 N HCl, and the recovered protein solution was adjusted at pH 8.5 with Tris–HCl. Digestion was performed by a mixture of trypsin (Sequencing Grade Modified Trypsin, Promega, Madison, MI, USA) and Acromobacter lysylendopeptidase (Wako, Japan) at 37°C for 12 h.
Extraction from DBS was described previously.11) Briefly, DBS was incubated in PBS with gentle agitation for 1 h, and the solution was collected and subjected to immunopurification as described above.
Mass spectrometryLC-MS of intact transferrin was carried out by an API4500 triple quadrupole (Q) mass spectrometer or a QSTAR Q time-of-flight (QTOF) mass spectrometer (Sciex, Framingham, MA, USA) connected to a C4 reversed-phase column (2 mm diameter and 10 mm length, GL Sciences, Japan). After sample injection, the column was washed with 0.1% formic acid at a flow rate of 0.2 mL/min and then eluted with 60% acetonitrile/0.1% formic acid at a flow rate of 0.05 mL/min. API4500 was operated in the positive Q1 MS mode with the optimized parameters as follows: the gas temperature was at 150°C, curtain gas pressure was 10 psi, ion source gas pressure was 16 psi, IonSpray voltage was 5.5 kV, declustering potential was 150 V, and entrance potential was 10 V. The full scan range was set from 1780 to 2000, and the scan rate was 10 Da/s. QSTAR was operated in the positive TOF MS mode. The optimized parameters were as follows: gas temperature was ambient, nebulizer gas (GS1) was 40 psi, curtain gas pressure was 50 psi, IonSpray voltage was 5.5 kV, and declustering potential was 100 V. The full scan range was set from 1500 to 3000. The pulsar frequency was 5 kHz and the pulse duration was 20 µs. The zero-charge mass spectrum was generated by the Promass protein deconvolution software (Thermo-Fisher Scientific).
LC-MS of the tryptic digest was carried out by an API4500 connected to a C8 reversed-phase column (1 mm diameter, 10 mm length, 300 Å pore size, GL Sciences). After sample injection, the column was washed with 0.1% formic acid at a flow rate of 0.06 mL/min and then eluted with a linear gradient of acetonitrile. API4500 was operated in the positive Q1 MS mode with the optimized parameters as follows: the gas temperature was at 120°C, curtain gas pressure was 10 psi, ion source gas pressure was 16 psi, IonSpray voltage was 5.5 kV, declustering potential was 90 V, and entrance potential was 10 V. The full scan range was set from 1210 to 1650, and the scan rate was 200 Da/s.
When evaluating fucosylation levels by MS, the potential loss of fucose during the ionization process must be considered. In our previous report on the dissociation profile of protonated fucosyl glycopeptide, the α1,6 linkage of core fucose was stable during ESI or MALDI processes, ensuring the estimation of fucosylation levels by the ion intensity in the mass spectrum.12) The values obtained from ESI and MALDI MS correlate quite well.
Evaluation of fucosylation levels by MS of intact transferrinESI-QTOF MS of intact transferrin is currently the gold standard method for CDG diagnosis. This is because the QTOF analyzer has sufficient resolution and mass range for multiply charged ions of transferrin (average mass 79,569), which are widely distributed in the m/z 1500 to m/z 3000 range. Diagnosis of CDG has so far been performed by detecting the appearance or increase in abnormal peaks corresponding to an N-glycan defect (−2,206 u), loss of N-acetylneuraminic acid (NeuAc, −291 u) or NeuAc + galactose (−453 u).13) Therefore, it is sufficient to have resolving power that allows the detection of peaks with a mass difference of 0.37% or more for transferrin. When fucose (C6H10O4) is added, the mass increase is 146 u, or 0.18%, which can theoretically be separated by a resolving power of 1500 based on full width at half maximum (FWHM). Indeed, the fucosylated peak is clearly detected in the ESI-QTOF mass spectrum (Fig. 1), in which the fucosylation levels can be calculated by comparing the intensity of transferrin with biantennary chains with the fucosylated counterpart. It should be noted that there is a peak approximately 98 u higher than the main peak.
MS of glycopeptides allows facile analysis of site-specific glycosylation,10) and MS of glycopeptides provides sufficient information for diagnosis of CDG.14–16) Transferrin is glycosylated at Asn432 and Asn630, and the microheterogeneity at these glycosylation sites, including fucosylation levels, differs from each other. The glycan at Asn630 has a core fucose, while the glycan at Asn432 has not only a core fucose but also a Lewis fucose that binds to the α1,3 antenna.17) As shown in the mass spectrum covering the triply protonated molecules, the overall level of fucosylation at Asn432 is very small (Fig. 2). Therefore, in this study, a glycopeptide containing Asn630 is used to measure fucosylation levels, which are determined by the signal intensity ratio of the protonated molecules of non-fucosylated and fucosylated glycopeptides. The percentage ratio in the 0- to 32-year-old group (n = 68) is 7.9 ± 1.7 (mean ± SD), and the fucosylation level is independent of age (Fig. 3).
In several countries, including Japan, NBS of inborn errors of metabolism is performed by ESI-MS, a method referred to as “Tandem MS,”18) and the analysis is performed by triple quadrupole mass spectrometers. Previously, we reported measurements using an NBS-dedicated Q-mass analyzer instrument to expand the use of MS in CDG diagnosis.13) In the deconvoluted spectrum of the intact transferrin measured by Q MS (Fig. 4A), the FWHM of the main peak is more than 60 Da, that is, the mass resolution >1500 by the peak width definition, and the +98 u peak is not well resolved from the main peak and the fucosylated peak (+146 u). The percentage ratio of non-fucosylated and fucosylated species calculated by Q MS of intact transferrin is 15.9 ± 3.5 (mean ± SD) in the same cohort (Fig. 5). The average value is twice that obtained from MS of Asn630 glycopeptides. However, the coefficients of variation are exactly the same as each other, and the scatter plot shows a good correlation (correlation coefficient 0.82). This finding suggests that the fucosylation level can also be assessed by NBS-dedicated quadrupole mass spectrometers.
In our previous report, we demonstrated that dried blood spots (DBS) could be used for MS screening of CDG.11) Considering that the mass spectrum of transferrin from DBS is virtually identical to that from serum (Fig. 4B), DBS can be used to quantify core fucosylation levels.
Prospects for application to fucosylation disordersReduced core fucosylation has been reported in FUT8-CDG, SLC35C1-CDG (CDG-IIc/LAD2), and GFUS-CDG. We previously reported that no core-fucosylated glycans were detected in the MALDI mass spectra of Asn630 glycopeptides from a patient with SLC35C1-CDG.12) The same results were subsequently obtained by the analysis of N-glycans from whole plasma.19) Similarly, a complete loss of core fucose was reported in FUT8-CDG.20) In GFUS-CDG, core fucosylated glycans were reduced to one-third of controls.21) Taken together, these findings suggest that the methods described here are useful for diagnosing CDGs with defective fucosylation.
Some types of CDG can be treated with oral supplementation of monosaccharides, and similar treatment trials for fucosylation disorders are in progress.22–24) The quantification methods, that is, MS of intact transferrin or glycopeptides, will be useful for diagnosing CDG with defective fucosylation and assessing therapeutic efficacy.
In this paper, we have described two methods to detect changes in fucosylation by MS: first, LC-MS of tryptic peptides to precisely measure the fucosylation level of glycans attached to peptides containing Asn630, and second, a simpler method by MS of intact transferrin. These approaches can be performed on mass spectrometers equipped with Q mass analyzers for advanced NBS programs and will hopefully make quantification of core fucosylation levels practical and facilitate diagnosis and therapeutic evaluation of CDG with fucosylation defects.
The spectrum data files of Figs. 1, 2, and 4 are available in J-STAGE Data.
Special thanks go to Dr. Okamoto for his efforts as the central figure in the CDG diagnostic project at OWCH. This work was supported by a Grant-in-Aid from AMED (24ek0109614).
Mass Spectrom (Tokyo) 2024; 13(1): A0159
