Abstract
Oligosaccharides have multiple functions essential for health. Derived from the condensation of two to several monosaccharides, they are structurally diverse with many co‐occurring structural isomer families, which make their characterization difficult. Thanks to its ability to separate small molecules based on their mass, size, shape, and charge, ion mobility‐mass spectrometry (IM‐MS) has emerged as a powerful tool for separating glycan isomers. Here, the potential of such a technique for the rapid characterization of main human milk oligosaccharides (HMOs) was investigated. Our study focused on 18 HMO standards. The IM‐MS analysis enabled to distinguish almost all the HMOs studied, in particular thanks to the single ion mobility monitoring acquisition using the trapped ion mobility spectrometry device, providing high ion mobility resolution and enhanced ion mobility separation. Alternatively, the combination of IM‐MS separation with MS/MS experiments has proven to increase performance in identifying HMOs and especially isomers poorly separated by ion mobility alone. Finally, collision cross‐section values are provided for each species generated from the 18 HMOs standards, which can serve as an additional identifier to characterize HMOs.
Keywords: human milk oligosaccharide, ion mobility, isomer characterization, mass spectrometry, MS/MS experiments
1. INTRODUCTION
Oligosaccharides are involved in many biological processes, including cell recognition and immune response. Carbohydrate chains result from the condensation of two to up to 20 monosaccharides by glycosidic bonds. There are glycoproteins, glycosphingolipids, and free oligosaccharides such as human milk oligosaccharides (HMOs) that are very abundant in breast milk and known to play a crucial role as prebiotics, allowing the development of the infant's microbiota.1
HMOs are structurally complex glycans, each HMO being composed of a lactose unit (galactosyl‐β1–4‐glucose), condensed with one or more monosaccharides among the five basic monosaccharides including D‐glucose (Glc), D‐galactose (Gal), L‐fucose (Fuc), N‐acetylglucosamine (GlcNAc), and N‐acetylneuraminic acid or sialic acid (SA) (see Figure1). Lactose can be fucosylated or sialylated to form, for example, 2′‐fucosyllactose (2′FL) and 3‐fucosyllactose (3FL) or 3′‐sialyllactose (3′SL) and 6′‐sialyllactose (6′SL), respectively. It can also be elongated by a disaccharide unit leading for example to lacto‐N‐tetraose (LNT) and/or lacto‐N‐neotetraose (LNnT). Additional monosaccharide and/or disaccharide units can be added, generating linear (e.g., Lacto‐N‐fucopentaose I [LNFP I] and sialyllacto‐N‐neotetraose a [LST a]) or branched HMOs (e.g., lacto‐N‐hexaose [LNH] and difucosyllacto‐N‐hexaose [DFLNH]). Therefore, HMOs consist of monosaccharides linked by various types of bonds (i.e., linear or branched structures and α or β linkages via 1–2,3,4 or 2–3,6 linkages), resulting in diverse structures and numerous families of structural isomers; for example, over 200 different HMO structures have been reported in pooled human milk samples.6
FIGURE 1.
Open in a new tab
Various analytical techniques have been developed for carbohydrate analysis.7, 8 Most of them are often based on mass spectrometry (MS) allowing sensitive and specific detection.9 In addition, tandem mass spectrometry (MS/MS) provides access to structural details essential to unambiguously characterize a given structure. The coupling of MS with a hyphenated method such as liquid chromatography (LC, using hydrophilic‐interaction chromatography or porous graphitized carbon columns) or capillary electrophoresis has been widely used to analyze oligosaccharides in complex mixtures.9 However, such hyphenated techniques are not suitable for high throughput assays, because of the time need to separate complex oligosaccharide mixtures before MS detection. In addition, reduction and permethylation of oligosaccharides are often performed before the LC/MS analysis, requiring additional sample preparation steps that lengthen the overall analysis time.10
Alternatively, ion mobility (IM) technology based on the ion separation in the gas phase has proven to be a promising tool for the rapid characterization of isomeric compounds.11 Its coupling with mass spectrometry (IM‐MS) offers an additional dimension of separation without lengthening the acquisition time unlike traditional gas chromatography‐mass spectrometry (GC/MS) or LC/MS analysis because the IM separation is done on a millisecond scale. In addition to its ability to separate ions based on their size, shape, mass, and charge, IM‐MS provides a relevant intrinsic parameter, that is, the collision cross‐section (CCS), which can be used as an additional descriptor to aid in assigning the identity of a given compound. Few studies have focused on the structural characterization of HMOs by IM‐MS.12, 13, 14 In a previous work, we have evaluated the potential of direct introduction‐trapped ion mobility spectrometry time‐of‐flight mass spectrometry (TIMS‐TOF) for the rapid characterization of HMO isomers and demonstrated its ability to distinguish four pairs of HMO isomers that are 2′FL/3FL, 3′SL/6′SL, LNT/LNnT, and LNFP I/LNFP V.15 A major advantage of the TIMS‐TOF instrument is the possibility of modifying the scan mode allowing to increase the resolving power of the TIMS analyzer and therefore to improve the IM separation.
The objective of this work is to extend the previous IM‐MS based method to a rapid identification of a large number of HMOs. Eighteen HMO standards (Figure1) were analyzed using a full ion mobility scan and single ion mobility monitoring (SIM2) method to assess IM separation, providing a CCS library of HMOs for various positively and negatively charged adduct ions. The SIM2 method implies the detection of a narrow ion mobility range centered on the studied species, providing a high mobility resolving power (up to 200) that facilitates the separation and/or the distinction of very closed mobility peaks. Complementary MS/MS experiments were conducted to obtain essential data for the subsequent high throughput structural characterization of isomers.
2. MATERIAL AND METHODS
2.1. Chemicals and milk samples
Ultrapure water with a resistance of 18.2 MΩ.cm was produced by a Select HP water purification system (Purite France eau, Lormont, France). Methanol high‐performance liquid chromatography grade) was acquired from VWR Chemicals (Fontenay‐sous‐Bois, France). The calibration solution called ESI‐L Low Concentration Tuning Mix (G1969–85000) was purchased from Agilent Technologies (Santa Clara, CA, USA) for mass and ion mobility calibration of the TIMS‐TOF instrument. HMO standards (Figure1) were purchased from Carbosynth (Slovakia, 2′FL, 3FL, 3′SL, 6′SL, LNFP I and II, LST b and c, LNH, LDFT, LNDFH I, and DFLNH) and Elicityl (Crolles, France, LNT, LNnT, LST a, LNFP V and VI, and LNDFH II). HMO standards have a purity ≥ 90%, except for LDFT and LNFP V whose purity was about 80%. Among the 18 HMO standards, there are four isomer pairs, that is 2′FL/3FL, 3′SL/6′SL, LNDFH I/II, LNT/LNnT; two isomer families, that is, LST a/b/c and LNFP I/II/ V/VI; and three individual HMOs, that is, LNH, LDFT, and DFLNH.
All HMO standards were dissolved in water–methanol (1:1, v/v) at final concentrations ranging from 1 to 5 ng/μl depending on the experiment performed. For positive ionization experiments, formic acid was added at low level (0.1% in volume) to the HMO standard solutions to promote the formation of protonated ions.
2.2. Ion mobility spectrometry–mass spectrometry analysis
All IM‐MS experiments were performed on a trapped ion mobility spectrometer ‐ quadrupole ‐ time of flight (Bruker Daltonics, Bremen, Germany)16 equipped with an electrospray ionization (ESI) source using direct infusion at a flow rate of 3 mL min−1. The instrumental parameters were optimized for the detection of HMOs using the oTof control software (Bruker Daltonics). The end plate offset was set at 500 V and the electrospray voltage at −4500 V and +3500 V in positive and negative ionization, respectively. The capillary temperature was maintained at 250°C. Nitrogen was used as the spray and drift gas. The dry gas and the nebulizer gas were fixed at 3.0L min−1 and 0.3 bar, respectively. In the TIMS analyzer, the Funnel 1 RF, Funnel 2 RF, and deflection delta were set at 260 Vpp, 250 Vpp, and ±80 V, respectively. The ion charge control was kept at 1.5 Mio to avoid TIMS saturation.
Mass spectra were recorded in a range of m/z 100–1650 with a transfer time of 70 μs and a pre‐pulse storage of 5 μs. Two ion mobility detection modes were used: (i) a full ion mobility scan where ion mobility spectra are collected within a large ion mobility range, that is, reduced mobilities (1/K0) from 0.55 to 1.90V.s.cm−2 with a scan rate of 9.52 Hz, and (ii) a SIM2 method using a narrow ion mobility range detection, typically a 1/K0 window of about 0.10V.s.cm−2 with a scan rate between 2 and 5 Hz.
CCS values were obtained from the reduced mobilities measured, K0, requiring a calibration step. Therefore, external calibrations in m/z (in quadratic mode) and reduced mobility values (in linear mode) were carried out using the ESI‐L low concentration tuning mix solution.15 Note that the ΔCCS value, corresponding to the difference in CCS between two ion mobility peaks A and B, is systematically calculated (in %) in order to evaluate the separation capacity of the TIMS device. We also used ΔCCS (in percent) to compare the measurements with the values reported in the literature.
MS/MS experiments were performed after ion mobility separation within a full ion mobility scan using CID (collision‐induced dissociation) conditions. The following MS/MS conditions were used: 2 u for the isolation width of precursor ions, an activation energy between 18 and 80 eV (the exact value used is indicated on the MS/MS spectra, see FigureS1 in the supporting information) depending on the selected precursor ions in order to obtain high abundant fragment ions while maintaining sufficient precursor intensity into MS/MS spectra. Nitrogen was used as the collision gas.
3. RESULT AND DISCUSSION
3.1. Full ion mobility scan for analysis of HMOs and distinction of their isomers
IM‐MS analysis of the 18 HMO standards was first conducted by applying a full ion mobility scan in order to detect a maximum of ionic species in a wide ion mobility range where 1/K0 ranged from 0.55 to 1.90V.s.cm−2 (FigureS2 in the supporting information). All IM‐MS analyses were carried out in triplicate, and the averaged TIMSCCSN2 values for both positively and negatively charged species generated from every HMOs are reported in TableS1 in the supporting information.
In the negative ionization mode, all HMOs studied were detected as deprotonated species [M − H]−. [M + Cl]− and [M + HCOO]− adducts were also observed, except for 3′SL, 6′SL and LST a, b, and c (TableS1 in the supporting information). In positive ionization mode, many species like [M + H]+, [M + Ca]2+, [M + Na]+, [M + K]+, and/or [M − H + Ca]+ were produced. However, the [M + H]+ ions were not observed for 2′FL and 3FL, and they were weakly detected for 3′SL. It should be noted that the isobaric [M + K]+ and [M − H + Ca]+ ions are detected together under the same ion mobility spectrum extracted at a given m/z value. Indeed, their distinction is not achieved using a TOF analyzer with a resolving power of 50,000 even if their annotation can be done from accurate mass measurements and/or salt addition for unambiguous identification (FigureS3 in the supporting information). Additionally, multiple mobility peaks could be observed for a given HMO ion (Figure2), which is likely because of (i) isobaric contaminants from solvent or consumables, and/or more probably to (ii) the presence of different oligosaccharide conformations and/or ion structures (e.g., protomers).17, 18 Note that multiple mobility peaks are often observed for both protonated and deprotonated compounds, especially under high‐resolution mobility, certainly because of the existence of protomers or deprotomers.17
FIGURE 2.
Open in a new tab
Examination of CCS values recorded for all the species detected in three replicate analyses indicates good repeatability of the CCS measurements with RSD < 0.5%. This high precision is consistent with our previous work.15, 19 Some of experimental CCS values obtained in the present study were also compared with those found in the studies from Zheng et al.20 and May et al.21 that assessed cross‐platform measurement accuracy (TableS2 in the supporting information). Small deviations, that is, ΔCCS% < 2%, were observed, which are within the accepted tolerance for cross‐platform measurements.22 Thus, the TIMS‐TOF instrument provides reliable TIMSCCSN2 values that can be used for further characterization of the 18 HMOs.
The ability of TIMS‐TOF to separate isomers can be evaluated based on the ΔCCS% between the isomers considered. As previously reported,15, 19 a baseline separation between two mobility peaks is achieved in full ion mobility scan of TIMS device (i.e., using a wide ion mobility detection range from 0.55 to 1.90V.s.cm−2) for ΔCCS% > 3.5%, whereas a SIM2 mode (i.e., using a narrow ion mobility range) is necessary to baseline‐separate mobility peaks with 1.5% < ΔCCS% < 3.5%. Only partially separated peaks could be obtained with SIM2 for 1% < ΔCCS% < 1.5%. Based on this criteria, only LNT/LNnT isomer pair can be easily separated in full ion mobility scan at moderate resolving power (50 < Rp < 80) when considering the [M‐H]− or [M + H]+ species (Table1). Although the other isomer pairs could not be well separated in full ion mobility scan, the distinction of between 2′FL and 3FL or between LNDFH I and LNDFH II could be achieved for the [M + Cl]− and [M + K]+/[M − H + Ca]+ species (Figure2A). Other adduct ions could also be used to distinguish isomers like [M + Na]+ ions for 3′SL/6′SL, 2′FL/3FL, and LNDFH I/II pairs.
TABLE 1.
Optimal isomer separation (or distinction) conditions using TIMS instrument
| Acquisition mode | Isomer separation | Studied HMOs | Considered ions |
|---|---|---|---|
| Full ion mobility scan | Full separation with ΔCCS% > 3.5% | LNT/LNnT | [M − H]−, [M + H]+ |
| SIM2 mode | Full separation with 1.5% < ΔCCS% < 3.5% | 2′FL/3FL | [M + Cl]−, [M + Na]+, [M + Ca]2+, [M + K]+ |
| LNDFH I/LNDFH II | [M + Cl]−, [M + Na]+, [M + K]+ | ||
| Isomer distinction 1% < ΔCCS% < 1.5% | 3′SL/6′SL | [M − H]−, [M + Na]+ | |
| LST a/LST b/LST c | [M + Na]+, [M + K]+ | ||
| No separation ΔCCS% < 1% | LNFP I/LNFP II/LNFP V/LNFP VI | All species |
Open in a new tab
Note: SIM2, single ion mobility monitoring; HMOs, human milk oligosaccharides; CCS, collision cross‐section; TIMS, trapped ion mobility spectrometry.
Although baseline separation could not be achieved for all HMO isomers studied using full ion mobility scan, their distinction is possible for most of them, except for the four LNFP isomers that could not be individually characterized whatever the species considered. To improve the ion mobility separation and the distinction of HMO isomers, the SIM2 method was then applied (Table1).
3.2. SIM2 for enhanced isomer separation
A prior IM‐MS analysis in full ion mobility scan is necessary to determine the 1/K0 value of relevant species. Then, SIM2 was performed to monitor a specific species by narrowing the detection range of ion mobility centered on the targeted HMO ion. Indeed, considering the TIMS principle, the use of a narrower voltage range (Vramp) and a larger ramp time (Tramp) during the elution step allows enhancing the ion mobility separation with resolving power up to 200.23 Using SIM2 method with TIMS‐TOF instrument, the baseline separation of isomers with differences in CCS values smaller than 2% has been demonstrated.19
SIM2 method was applied for the analysis of all HMO standards in positive and negative ion modes in order to improve isomer separation. Thus, all species generated from the 18 HMO standards could be detected at high mobility resolving power (Rp > 100). For example, we show comparison of the extracted ion mobility spectra for the [M + Cl]− or [M + Na]+ adducts from IM‐MS analysis in full ion mobility scan versus SIM2 mode for LNDFH I and LNDFH II pair (Figure2A and D); LSTa, LSTb, and LSTc (Figure2B and E); and LNFP I, LNFP II, LNFP V, and LNFP VI (Figure2C and F). In full ion mobility scan, LNDFH I and LNDFH II could be separated as [M + Cl]− ions but with peak overlapping at 20% valley (Figure2A), whereas the three isomers of LST could not be well separated as [M + Na]+ ions. (Figure2B). Obviously, the use of SIM2 method allows obtaining high mobility resolution detection and reducing the ion mobility peak width (Figure2D, E, and F) compared to that from a full ion mobility scan (Figure2A, B, and C). Therefore, the separation between LNDFH I and LNDFH II is enhanced (Figure2D). In addition, the broad mobility peak of LSTa as well as that of LSTc are split into two peaks, and these two isomers could be well separated (Figure2E). Although the LSTb peak still overlaps the signals of LSTa and LSTc, its presence could be detected with the SIM2 method. Nevertheless, the distinction of LSTb could be achieved by considering another species such as [M + K]+ (FigureS2 in the supporting information). By increasing the ion mobility resolution, the SIM2 method improved the HMO separation, making easier the distinction of isomers with ΔCCS% > 1%.
However, the LNFP isomer family could not be well characterized whatever the species considered. Indeed, even using SIM2 method offering high ion mobility resolution, individual isomer separation could not be obtained. For the best ion mobility separation conditions, the [M + Na]+ species of LNFP isomers only allowed the separation of LNFP I and LNFP V isomers from the others, but the LNFP II and LNFP VI signals are still overlapping (Figure2E). In contrast, for all negatively charged species, the four isomers of LNFP could only be separated into two sub‐groups as exemplified in Figure3A. The extracted ion mobility spectra of the [M − H]− species generated by the four isomers of LNFP display two main mobility peaks: (i) the first one containing the mixture of LNFP V and VI was detected at 1/K0 between 1.25 and 1.30V.s.cm−2 (Figure3A), and (ii) the second that contains LNFP I, II, and VI was detected at 1/K0 between 1.30 and 1.35V.s.cm−2 (Figure3A). Therefore, complementary tandem mass spectrometry (MS/MS) experiments were performed to help in isomer structural characterization.
FIGURE 3.
Open in a new tab
3.3. Ion mobility‐tandem mass spectrometry to improve structural characterization of HMOs
Complementary to ion mobility separation, tandem mass spectrometry experiments (MS/MS) were carried out on all species detected in positive and negative ionization modes for each studied HMO in order to get structural information essential for further HMO characterization, in particular to distinguish isomers not well separated by ion mobility. As already reported,15, 24 specific fragmentation patterns of HMO isomers can be obtained from negatively charged species (FigureS1 in the supporting information), which was not the case from MS/MS experiments conducted in positive ionization mode. The nomenclature introduced by Domon and Costello3 (Figure1) was used to annotate the two main types of produced ions resulting from (i) glycosidic bond cleavage (B, Y, C, and Z) and (ii) cross‐ring fragmentation (A and X). Therefore, the detection of corresponding ions reflects the presence of a given isomer in the analyzed sample.
Despite the partial ion mobility separation in full ion mobility scan of deprotonated ions (Figure3A), the four isomers of LNFP could be distinguished thanks to the combination with the MS/MS experiments allowing the detection of the diagnostic product ions characteristic of each isomer (Figure3B and C). Decomposition of deprotonated species results in the detection of diagnostic product ions: at m/z 325 for LNFP I; m/z 348, m/z 288, and m/z 184 for LNFP II; m/z 202 for LNFP V; and m/z 179, m/z 263, and m/z 281 for LNFP VI (Figures3B and C and S1). Similarly, the characterization of the other HMO isomers was obtained thanks to diagnostic fragment ions generated from negatively charged precursor ions under IM‐MS/MS analysis, with the exception of LST set (FigureS1 in the supporting information).
4. CONCLUSION
High ion mobility resolution is essential for the separation of structurally related compounds with very close ion mobility, in particular for the characterization of isomers. In our study, all the studied HMO isomers could be distinguished by considering at least one specific ionic species generated under positive or negative ion mode thanks to the ion mobility separation and, for those poorly separated by ion mobility, complementary MS/MS experiments are able to provide diagnostic product ions. Therefore, the combination of ion mobility separation and MS/MS provides optimal performance for in‐depth characterization of HMOs and isomer distinction. Such an approach appears very useful for the rapid analysis of breastmilk samples and, more importantly, for the characterization of the HMO profiles of breastmilk samples, which varies from one individual to another and especially between individuals from different regions.
Supporting information
Table S1.TIMSCCSN2 measured from different species detected both in positive and negative ESI ionization IM‐MS analyses of HMOs. Experimental values are averaged from replicate IM‐MS analyses (N ≥ 3.) and reported as CCS(a) values for the most abundant peaks (low abundant peaks are not reported, i.e. normalized abundances<15%) using large ion mobility range detection, with the relative standard deviation (RSD) ≤ 0.5%, except for the [M + Ca]2+ ions from LNFP V and VI and for the peak of the [M + H]+ ions from LNDFH I which can reach 0.9%. CCS(b) values are only reported for additional peaks detected under high resolution conditions (SIM2 mode). Note that normalized abundances are also provided if multiple mobility signals are observed for a given HMO ion.
Table S2. Comparison of experimental CCS values with published values (DTIMSCCSN2 (a) from Zheng et. al (doi;10.1039/c7sc03464d) and DTIMSCCSN2 (b) from May et. al (doi: 10.1021/ac4038448), acquired with a drift tube ion mobility mass spectrometry (DTIMS).
Figure S1. CID spectra from negatively charged HMOs. The nomenclature introduced by Domon and Costello (Domon B, Costello CE. A systematic nomenclature for carbohydrate fragmentations in FAB‐MS/MS spectra of glycoconjugates. Glycoconj J 1988;5 (4):397–409. doi: 10.1007/BF01049915) was used to annotate the two main types of produced ions resulting from i) glycosidic bond cleavage (B, Y, C and Z) and ii) cross‐ring fragmentation (A and X). Note that an asterisk on some annotated product ions corresponds to additional water loss.
Figure S2. Ion mobility spectra of various HMO species detected under positive and negative ionizations modes using full ion mobility scan (1/K0 from 0.55 to 1.90V.s.cm−2). Note that some species are not reported here because they are not detected or have too low abundance peaks for at least one HMO isomer.
Figure S3. Ion mobility spectra extracted from annotated [M + k]+ and/or isobaric [M‐H + Ca]+ ions from LNFP VI dissolved in (A) MeOH/H2O (1:1, v/v) at 5 ng/μL final concentration and (B) after KCI or CA3(PO4)2 salt addition (10 ng/μL final concentration). In (C) the extracted mass spectra obtained for both mobility ranges around the adduct species allow to compare the experimental and theoretical isotopic patterns for accurate annotation. Full ion mobility scan (1/K0 from 0.55 to 1.90 V.s.cm−2) was used.
Click here for additional data file. (1.4MB, pdf)
ACKNOWLEDGEMENTS
The authors thank the MetaboHUB infrastructure for funding (ANR‐11‐INBS‐0010 grant). This study was also funded by the InfaDiet project (ANR‐19‐CE36‐0008).
Rathahao‐Paris E, Delvaux A, Li M, et al. Rapid structural characterization of human milk oligosaccharides and distinction of their isomers using trapped ion mobility spectrometry time‐of‐flight mass spectrometry. J Mass Spectrom. 2022;57(10):e4885. doi: 10.1002/jms.4885
Contributor Information
Estelle Rathahao‐Paris, Email: estelle.paris@inrae.fr.
Sandra Alves, Email: sandra.alves@sorbonne-universite.fr.
DATA AVAILABILITY STATEMENT
Data available in article supplementary material
REFERENCES
- 1.Wiciński M, Sawicka E, Gębalski J, Kubiak K, Malinowski B. Human milk oligosaccharides: health benefits, potential applications in infant formulas, and pharmacology. Nutrients. 2020;12(1):266. doi: 10.3390/nu12010266 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Neelamegham S, Aoki‐Kinoshita K, Bolton E, et al. Updates to the symbol nomenclature for glycans guidelines. Glycobiology. 2019;29(9):620‐624. doi: 10.1093/glycob/cwz045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Domon B, Costello CE. A systematic nomenclature for carbohydrate fragmentations in FAB‐MS/MS spectra of glycoconjugates. Glycoconj J. 1988;5(4):397‐409. doi: 10.1007/BF01049915 [DOI] [Google Scholar]
- 4.Triantis V, Bode L, van Neerven RJJ. Immunological effects of human milk oligosaccharides. Front Pediatr. 2018;6:190. doi: 10.3389/fped.2018.00190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Walsh C, Lane JA, van Sinderen D, Hickey RM. From lab bench to formulated ingredient: characterization, production, and commercialization of human milk oligosaccharides. J Funct Foods. 2020;72:104052. doi: 10.1016/j.jff.2020.104052 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ninonuevo MR, Park Y, Yin H, et al. A strategy for annotating the human milk glycome. J Agric Food Chem. 2006;54(20):7471‐7480. doi: 10.1021/jf0615810 [DOI] [PubMed] [Google Scholar]
- 7.Alley WR, Novotny MV. Structural glycomic analyses at high sensitivity: a decade of progress. Annu Rev Anal Chem. 2013;6(1):237‐265. doi: 10.1146/annurev-anchem-062012-092609 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Nielsen SS (Ed). Food analysis. Springer International Publishing; 2017. doi: 10.1007/978-3-319-45776-5. [DOI] [Google Scholar]
- 9.Kailemia MJ, Ruhaak LR, Lebrilla CB, Amster IJ. Oligosaccharide analysis by mass spectrometry: a review of recent developments. Anal Chem. 2014;86(1):196‐212. doi: 10.1021/ac403969n [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Costell CE, Contado‐Miller JM, Cipollo JF. A glycomics platform for the analysis of permethylated oligosaccharide alditols. J Am Soc Mass Spectrom. 2007;18(10):1799‐1812. doi: 10.1016/j.jasms.2007.07.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dodds JN, Baker ES. Ion mobility spectrometry: fundamental concepts, instrumentation, applications, and the road ahead. J Am Soc Mass Spectrom. 2019;30(11):2185‐2195. doi: 10.1007/s13361-019-02288-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Peterson TL, Nagy G. Toward sequencing the human milk glycome: high‐resolution cyclic ion mobility separations of core human milk oligosaccharide building blocks. Anal Chem. 2021;93(27):9397‐9407. doi: 10.1021/acs.analchem.1c00942 [DOI] [PubMed] [Google Scholar]
- 13.Struwe WB, Baldauf C, Hofmann J, Rudd PM, Pagel K. Ion mobility separation of deprotonated oligosaccharide isomers—evidence for gas‐phase charge migration. Chem Commun. 2016;52(83):12353‐12356. doi: 10.1039/C6CC06247D [DOI] [PubMed] [Google Scholar]
- 14.Wei J, Tang Y, Ridgeway ME, Park MA, Costello CE, Lin C. Accurate identification of isomeric glycans by trapped ion mobility spectrometry‐electronic excitation dissociation tandem mass spectrometry. Anal Chem. 2020;92(19):13211‐13220. doi: 10.1021/acs.analchem.0c02374 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Delvaux A, Rathahao‐Paris E, Guillon B, et al. Trapped ion mobility spectrometry time‐of‐flight mass spectrometry for high throughput and high resolution characterization of human milk oligosaccharide isomers. Anal Chim Acta. 2021;1180:338878. doi: 10.1016/j.aca.2021.338878 [DOI] [PubMed] [Google Scholar]
- 16.Ridgeway ME, Lubeck M, Jordens J, Mann M, Park MA. Trapped ion mobility spectrometry: a short review. Int J Mass Spectrom. 2018;425:22‐35. doi: 10.1016/j.ijms.2018.01.006 [DOI] [Google Scholar]
- 17.Marlton SJP, McKinnon BI, Ucur B, et al. Selecting and identifying gas‐phase protonation isomers of nicotineH+ using combined laser, ion mobility and mass spectrometry techniques. Faraday Discuss. 2019;217:453‐475. doi: 10.1039/C8FD00212F [DOI] [PubMed] [Google Scholar]
- 18.Oranzi NR, Kemperman RHJ, Wei MS, et al. Measuring the integrity of gas‐phase conformers of sodiated 25‐hydroxyvitamin D3 by drift tube, traveling wave, trapped, and high‐field asymmetric ion mobility. Anal Chem. 2019;91(6):4092‐4099. doi: 10.1021/acs.analchem.8b05723 [DOI] [PubMed] [Google Scholar]
- 19.Delvaux A, Rathahao‐Paris E, Alves S. An emerging powerful technique for distinguishing isomers: trapped ion mobility spectrometry time‐of‐flight mass spectrometry for rapid characterization of estrogen isomers. Rapid Commun Mass Spectrom. 2020;34(24):e8928. doi: 10.1002/rcm.8928 [DOI] [PubMed] [Google Scholar]
- 20.Zheng X, Aly NA, Zhou Y, et al. A structural examination and collision cross section database for over 500 metabolites and xenobiotics using drift tube ion mobility spectrometry. Chem Sci. 2017;8(11):7724‐7736. doi: 10.1039/C7SC03464D [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.May JC, Goodwin CR, Lareau NM, et al. Conformational ordering of biomolecules in the gas phase: nitrogen collision cross sections measured on a prototype high resolution drift tube ion mobility‐mass spectrometer. Anal Chem. 2014;86(4):2107‐2116. doi: 10.1021/ac4038448 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Paglia G, Williams JP, Menikarachchi L, et al. Ion mobility derived collision cross sections to support metabolomics applications. Anal Chem. 2014;86(8):3985‐3993. doi: 10.1021/ac500405x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hernandez DR, DeBord JD, Ridgeway ME, Kaplan DA, Park MA, Fernandez‐Lima F. Ion dynamics in a trapped ion mobility spectrometer. Analyst. 2014;139(8):1913‐1921. doi: 10.1039/C3AN02174B [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Mank M, Welsch P, Heck AJR, Stahl B. Label‐free targeted LC‐ESI‐MS2 analysis of human milk oligosaccharides (HMOs) and related human milk groups with enhanced structural selectivity. Anal Bioanal Chem. 2019;411(1):231‐250. doi: 10.1007/s00216-018-1434-7 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1.TIMSCCSN2 measured from different species detected both in positive and negative ESI ionization IM‐MS analyses of HMOs. Experimental values are averaged from replicate IM‐MS analyses (N ≥ 3.) and reported as CCS(a) values for the most abundant peaks (low abundant peaks are not reported, i.e. normalized abundances<15%) using large ion mobility range detection, with the relative standard deviation (RSD) ≤ 0.5%, except for the [M + Ca]2+ ions from LNFP V and VI and for the peak of the [M + H]+ ions from LNDFH I which can reach 0.9%. CCS(b) values are only reported for additional peaks detected under high resolution conditions (SIM2 mode). Note that normalized abundances are also provided if multiple mobility signals are observed for a given HMO ion.
Table S2. Comparison of experimental CCS values with published values (DTIMSCCSN2 (a) from Zheng et. al (doi;10.1039/c7sc03464d) and DTIMSCCSN2 (b) from May et. al (doi: 10.1021/ac4038448), acquired with a drift tube ion mobility mass spectrometry (DTIMS).
Figure S1. CID spectra from negatively charged HMOs. The nomenclature introduced by Domon and Costello (Domon B, Costello CE. A systematic nomenclature for carbohydrate fragmentations in FAB‐MS/MS spectra of glycoconjugates. Glycoconj J 1988;5 (4):397–409. doi: 10.1007/BF01049915) was used to annotate the two main types of produced ions resulting from i) glycosidic bond cleavage (B, Y, C and Z) and ii) cross‐ring fragmentation (A and X). Note that an asterisk on some annotated product ions corresponds to additional water loss.
Figure S2. Ion mobility spectra of various HMO species detected under positive and negative ionizations modes using full ion mobility scan (1/K0 from 0.55 to 1.90V.s.cm−2). Note that some species are not reported here because they are not detected or have too low abundance peaks for at least one HMO isomer.
Figure S3. Ion mobility spectra extracted from annotated [M + k]+ and/or isobaric [M‐H + Ca]+ ions from LNFP VI dissolved in (A) MeOH/H2O (1:1, v/v) at 5 ng/μL final concentration and (B) after KCI or CA3(PO4)2 salt addition (10 ng/μL final concentration). In (C) the extracted mass spectra obtained for both mobility ranges around the adduct species allow to compare the experimental and theoretical isotopic patterns for accurate annotation. Full ion mobility scan (1/K0 from 0.55 to 1.90 V.s.cm−2) was used.
Click here for additional data file. (1.4MB, pdf)
Data Availability Statement
Data available in article supplementary material
