electrospray ionization ion mobility mass spectrometry ......82 m. sarbu et al. / biochimie 139...

lable at ScienceDirect

Biochimie 139 (2017) 81e94

Contents lists avai

Biochimie

journal homepage: www.elsevier .com/locate/biochi

Research paper

Electrospray ionization ion mobility mass spectrometry providesnovel insights into the pattern and activity of fetal hippocampusgangliosides

Mirela Sarbu a, b, �Zeljka Vukeli�c c, David E. Clemmer d, Alina D. Zamfir a, b, *

a National Institute for Research and Development in Electrochemistry and Condensed Matter, Timisoara, Romaniab Department of the Analysis and Modeling of Biological Systems, Aurel Vlaicu University of Arad, Romaniac Department of Chemistry and Biochemistry, University of Zagreb Medical School, Zagreb, Croatiad Department of Chemistry, Indiana University, Bloomington, IN, USA

a r t i c l e i n f o

Article history:Received 21 April 2017Accepted 24 May 2017Available online 26 May 2017

Keywords:Ion mobility separation mass spectrometry(IMS MS)GangliosidesHuman fetal hippocampusCollision-induced dissociationCholinergic activity

* Corresponding author. Plautius Andronescu StRomania.

E-mail address: [email protected] (A.D. Zamfir).

http://dx.doi.org/10.1016/j.biochi.2017.05.0160300-9084/© 2017 Elsevier B.V. and Société Française

a b s t r a c t

Gangliosides (GGs), a particular class of glycosphingolipids ubiquitously found in tissues and body fluids,exhibit the highest expression in the central nervous system, especially in brain. GGs are involved incrucial processes, such as neurogenesis, synaptogenesis, synaptic transmission, cell adhesion, growth andproliferation. For these reasons, efforts are constantly invested into development and refinement ofspecific methods for GG analysis. We have recently shown that ion mobility separation (IMS) massspectrometry (MS) has the capability to provide consistent compositional and structural information onGGs at high sensitivity, resolution and mass accuracy. In the present paper, we have implemented IMSMS for the first time in the study of a highly complex native GG mixture extracted and purified fromhuman fetal hippocampus. As compared to previous studies, where no separation techniques prior to MSwere applied, IMS MS technique has not just generated valuable novel information on the GG patterncharacteristic for hippocampus in early developmental stage, but also provided data related to the GGmolecular involvement in the synaptic functions by the discovery of 25 novel structures modified byCH3COO�. The detection and identification in fetal hippocampus of a much larger number of GG speciesthan ever reported before was possible due to the ion mobility separation according to the charge state,the carbohydrate chain length and the degree of sialylation. By applying IMS in conjunction with collisioninduced dissociation (CID) tandem MS (MS/MS), novel GG species modified by CH3COO� attachment,discovered here for the first time, were sequenced and structurally investigated in details. The presentfindings, based on IMS MS, provide a more reliable insight into the expression and role of gangliosides inhuman hippocampus, with a particular emphasis on their cholinergic activity at this level.

© 2017 Elsevier B.V. and Société Française de Biochimie et Biologie Moléculaire (SFBBM). All rightsreserved.

1. Introduction

The sialic acid-containing glycosphingolipids (GSL), known asgangliosides (GGs), contribute through both their glycan and lipidmoieties to the functions of the cellular lipidome and glycome/sialome [1]. Along with glycoproteins and proteoglycans, GSL,including GGs, are involved in the formation of the glycocalyx thatcovers eukaryotic cell surfaces [2,3]. Since the discovery of

r. 1, RO-300224, Timisoara,

de Biochimie et Biologie Molécul

gangliosides [4,5] within the ganglion cells and the brain tissue of apatient diagnosed with infantile amaurotic idiocy along withNiemann-Pick disease [6,7], important findings related to theirstructure and the correlation with various disorders were reported[8e12].

Ubiquitously present in tissues and body fluids in a low con-centration, GGs exhibit a particularly high abundance in the brain,where can amount up to 6% of the lipids weight. While the hy-drophobic ceramide moiety is rooted in the outer leaflet of theplasma membranes surrounding neurons [13], the sialoglycancomponent extends in the extracellular medium [14] tending toform specific domains, i.e. rafts, with several types of molecules,such as cholesterol and sphingomyelin. Thus, GGs are involved in

aire (SFBBM). All rights reserved.

mailto:[email protected]

http://crossmark.crossref.org/dialog/?doi=10.1016/j.biochi.2017.05.016&domain=pdf

www.sciencedirect.com/science/journal/03009084

http://www.elsevier.com/locate/biochi

http://dx.doi.org/10.1016/j.biochi.2017.05.016



M. Sarbu et al. / Biochimie 139 (2017) 81e9482

crucial processes of the nervous system, i.e. neurogenesis, syn-aptogenesis, synaptic transmission, cell adhesion, growth andproliferation [2,15]. The research conducted in the field hasrevealed that the GG profile is age-dependent; as the brain de-velops, the nature, concentration and the degree of gangliosidesialylation not just increases, but is also topographically influenced[16e20].

The permanent interest in deciphering the structural andfunctional roles of GGs in cells stimulated the development of noveland superior investigation techniques. Considering the amphiphiliccharacter of GGs responsible for their partition onto the aqueouslayer during Folch separation, the investigation of gangliosides issinuous. If in the past laborious chemical analysis was employed inGG characterization, nowadays, chromatographically-basedmethods are used for their separation, while quantification canbe achieved by staining and densitometry [21e24].

Mass spectrometry (MS) represents currently the key technol-ogy not just for GG mapping, but also for complete structuralanalysis of both ceramide and sialoglycan chains. Considering thecomplexity as well as the extremely low amount of samplesextracted from the human matrices, the separation of the complexmixtures into individual classes prior to MS characterization rep-resents a prerequisite, achievable by coupling separation tech-niques with MS. Although much less studied than other classes ofbiomolecules, gangliosides in general and brain gangliosides inparticular, still benefited by modern development in MS andvarious related protocols especially designed for their analysis.Various methods and strategies based on triple time-of-flight (TOF)MS, high capacity ion trap (HCT) MS, quadrupole time-of-flight(QTOF) MS, Orbitrap MS and Fourier transform ion cyclotronresonance (FTICR) MS on-line or off-line coupled with high-performance liquid chromatography (HPLC), ultra HPLC (UPLC),thin-layer chromatography (TLC) or capillary electrophoresis (CE)were conceived and introduced for the investigation of GGsexpressed in normal fetal and adult human brain regions, such asfrontal cortex [25], neocortex [26], frontal lobe [16,20,26], occipitallobe [20,26], hippocampus [27], cerebellum [28], caudate nucleus[29], in primary brain tumors, such as astrocitoma [30], meningi-oma [31], neuroblastoma [32], gliosarcoma [33] and hemangioma[25] and also in brain metastasis of lung adenocarcinoma [34].However, the above mentioned separation techniques presentseveral drawbacks, since: i) require laborious sample preparationprocedures, which inevitably leads to sample loss; ii) are ratherexpensive and complicated in terms of instrumentation, develop-ment of appropriate interfaces for coupling to MS and protocoldevelopment for GG separation according to ceramide and glycanchains; iii) exhibit a rather limited separation efficiency for highlycomplex GG mixtures; iv) are time consuming methods, claimingfor long runs, instrument conditioning between the runs, whichimpedes their use in high throughput mode.

Ion mobility spectrometry (IMS), introduced by Cohen andKarasek [35] at the end of 60's and initially known as plasmachromatography, represents nowadays one of the most proficientanalytical separation technique, particularly suitable for highlycomplex mixtures. Originally developed for military and securityapplications [36e38], IMS evolved with the introduction of elec-trospray ionization (ESI) at the end of 80's towards volatile small orlarge, organic or inorganic molecule characterization. Within theion mobility-equipped mass spectrometer, the gas-phase ionsproduced at atmospheric pressure are pulsed in the drift tube re-gion by an electric field, which provides the ions a constant velocity.Here, the ions collide with the molecules of the neutral buffer gasthat fills the drift region; therefore, the frictional force restrains themovement of the ions, reaching an equilibrium statewhere the ionsmove with a constant speed proportional with the applied electric

field. An important aspect in ion mobility separation is that thefrictional force depends on the mass, charge state and collisioncross-section of the analyzed molecule; multiply charged ionstravel under the electric field faster than the simply charged ones[39].

The combination of IMS with MS yields a superior and complexanalytical system offering to the traditional MS technique a newseparation dimension. Although the IMS process in the gas phase ischaracterized by a lower resolution as compared with moderncondensed phase chromatographic separation technologies, IMSand, particularly, the application of IMS prior to MS represent analternative to the above mentioned separation systems since: i)provides separation of components based on their differentialmobility through a buffer gas; ii) is not a stand-alone instrument,being incorporated within the MS; iii) does not require laborioussample preparation procedures prior to mass analysis; vi) occursseveral orders of magnitudes faster e about milliseconds e thanseparations based on liquid chromatographic techniques, whichtakes minutes-to-hours; v) the experiments allow real-time sepa-rations and characterization of complex mixtures; vi) offersanalytical data as driftscope images, where information concerningthe drift time (td), m/z values and ion abundances in a color-codedstyle are supplied; vii) allows the extraction of distinct features,offering information of possible isomers, isobars and conformerspresent in the mixture that otherwise would get lost or remainoverlapped in the traditional mass spectra. Moreover, the combi-nation of this unique platform with collision-induced dissociation(CID) in the trap or transfer region of the instrument, increases theanalysis specificity by generating structural information of all ionsin one experiment. Over the years, IMS MS was widely applied inthe study of metabolomes [40], proteomes [41], lipidomes [42e45],glycomes [46e49], non-covalent interaction [50] and glycosylationdisorders [51,52].

Considering all these advantages, we have introduced in 2016for the first time IMS MS in ganglioside brain research anddemonstrated its feasibility for mapping and detailed structuralanalysis of ganglioside mixtures extracted from healthy fetal brain[16]. By IMS MS we have identified in the fetal brain a three timeshigher number of GG species than ever reported, with an unpre-dicted diversity of the ceramide chains [16]. Moreover, IMS MSrevealed for the first time the occurrence in the human brain ofspecies with an elevated degree of sialylation, the entire series frommono-up to octasialylated GGs being detected [16].

Based on our previously achieved results, in the present paperwe extent the applicability of IMS MS towards the mapping ofgangliosides in a defined brain region. We have selected forinvestigation the hippocampus, a small brain structure locatedunder the medial temporal lobe. Hippocampus represents animportant component of the limbic system, being involved in theconsolidation of long-term memory, learning and spatial memory[53]. In Alzheimer's disease or other types of dementia, the hip-pocampus is one of the first brain regions vulnerable to damages,determining slow erosion of thinking skills and memory, leading toconfusion and disorientation [54]. Therefore, the investigation ofhippocampus in terms of ganglioside profiling and structuralelucidation present particular biomedical relevance for a betterinsight into the molecular mechanisms involved at this level. Up tonow, only one study related to the ganglioside topospecificexpression in human hippocampi by mass spectrometry was re-ported [27]. By using chip-based nanoESI QTOF MS, a significantdevelopment-related difference in GG structural and/or quantita-tive expression in the investigated fetal and adult hippocampuswas revealed. However, overall only 17 GG species were found asassociated to the human hippocampus [27]. This limited number ofdetected GG structures could be related to the complexity of the

M. Sarbu et al. / Biochimie 139 (2017) 81e94 83

investigated mixture and the lack of separation methods prior toMS.

In this context and based on the previously achieved data, in thepresent paper, a Synapt G2-S platform equipped with electrosprayionization, ion mobility separation, time-of-flight analyzer and CIDprior and/or post IMS, was developed in the negative ionizationmode for the characterization of a highly complex native ganglio-side mixture extracted and purified from normal fetal hippocam-pus in the 17th gestational week (denoted FH17). IMSMS techniquehas generated valuable data on the glycopattern characteristic forhippocampus in early developmental stage, contributing to GGbiomarker database widening by detecting a much larger numberof GG species as compared to earlier studies where no separationprior MS was applied. Moreover, IMS MS and MS/MS revealed forthe first time interesting aspects possibly connected to the role ofGGs within the synaptic mechanism.

2. Material and methods

2.1. Reagents and materials

The pure ganglioside extract was dried in a SpeedVac Concen-trator, SPD 111V-230 from Thermo Electron Corporation (Ashville,NC, USA), coupled to a vacuum pump, PC 2002 Vario with CVC 2000Controller from Vaccubrand (Wertheim, Germany). For samplepreparation, the dried material of the native extract was dissolvedin analytical grade methanol purchased from Sigma (St. Louis, MO,USA).

2.2. Hippocampus sampling

The native ganglioside mixture investigated in detail in thepresent study was purified from a normal fetal hippocampus in the17th gestational week (g.w.). The brain sample characterized herewas obtained during routine pedopathological section/autopsyexamination at Clinical Hospital for Obstetrics and Gynecology“Petrova” and Department of Forensic Medicine and Criminology,Faculty of Medicine, University of Zagreb, Croatia. The fetusdeceased because of spontaneous abortion. The pedopathologicalexamination, including histopathological analysis of the tissuedetermined no signs of malformation, aberrant development, orbrain pathology. The age of 17th g.w. of the fetus was deduced from:i) the mother's menstrual history; ii) the echographic fetal biom-etry conducted during pregnancy follow-up; iii) specific measure-ments of the aborted fetus, i.e. the femoral length, the humeruslength, the biparietal perimeter and weight measurements.Consequently, the brain was considered normal for the givengestational stage. After sampling, the fetal hippocampus wasweighed and stored at �20 �C until further extraction procedures.

2.3. Gangliosides extraction and purification

The extraction and purification of FH17 GGs mixture followedthe procedures described by us before [29,55] and the methoddeveloped by Svennerholm and Fredman [56] as modified byVukeli�c et al. [57]. Briefly, the tissue was homogenized in ice-coldwater and lipids extraction was performed twice using achloroform-methanol-water mixture (1:2:0.75, by vol.). Further, aphase partition followed by repartition was carried out in order toseparate GGs from other lipids. The combined upper phases con-taining gangliosides was collected and purified. Permission forexperiments with human tissue for scientific purposes was ob-tained from the Ethical Commission of the Zagreb Medical Facultyunder project 108120 financed by the Croatian Ministry of Scienceand Technology. A stock solution of the native GG extract (~1 mg/

mL) was prepared by dissolving the dried material in pure meth-anol. For IMSMS, dilution of the stock solution in pure methanol upto a concentration of approximately 5 pmol/mL (calculated for anaverage molecular weight of 2000 Da) was used. Prior to MSanalysis, the sample solutionwas centrifuged for 1 h at 6000 rpm ina mini-centrifuge (Thermo Fisher Scientific, USA). The supernatantwas collected and submitted to (�) ESI IMS MS and MS/MS analysisby CID at low energies.

2.4. Ion mobility mass spectrometry

A Synapt G2-S mass spectrometer (Waters, Manchester, UK)equipped with ESI source was used for all ion mobility massspectrometry experiments conducted in the present study. Theganglioside sample was infused by syringe pump at a 1.2 mL/minflow rate. The instrument was operated in the negative ionizationmode, with 1 scan/s speed for MS and MS/MS acquisition. Thecapillary potential was adjusted within 1.5e2 kV, however anefficient ionization of the components was observed at 2 kV. For anefficient ionization and minimal in-source fragmentation, the conevoltage was varied within the range of 30e50 V. The other condi-tions were set as follows: source block temperature 120 �C, des-olvation gas flow rate 100 L/h, desolvation temperature 350 �C.

The Synapt G2-S instrument consists of a quadrupole mass filter,the TriWave section which includes Trap, IMS and Transfer T-wavedevices, while the last component is an orthogonal accelerationTOF analyzer. After a narrow pulse of ions reach the drift tube re-gion filled with buffer gas, the continuous series of applied DCvoltage pulses determine the travel of ions. Several factors thatdepend not just on the tridimensional structure of the molecule, i.e.size, charge, shape and apparent surface area, but also on theinstrumental parameters, i.e. amplitude and travel velocity of thewave and pressure of IMS cell gas, influence the migration of theions; thus, the high-mobility ions are traveling with the wave,passing through the ion guide faster than low-mobility ions, whichroll over the top of the wave onto the following wave, spendingmore time in the drift section. To augment the complex separation,the IMS parameters were thoroughly adjusted as follows: IMS gasflow 90 mL/min, IMS wave velocity 650 m/s, IMS wave height 40 V.Low-mass (LM) and high-mass (HM) resolution parameters wereset at 12 and 15, respectively for the MS experiments, while for MS/MS experiments they were adjusted at 10 and 15, respectively.These values provided the best separation not only for the molec-ular ions in the mixture screened in the MS mode, but also for themass-selected ions in the quadrupole and submitted further to CIDMS/MS. Argon was used as collision gas in the transfer cell, whileNitrogenwas used as IMS cell gas. After separation in the drift tube,the ions were extracted into the TOF source for mass analysis. TheTOF analyzer was operated in the V-mode with an average massresolution of 20,000. The structural characterization of gangliosidesby low energy CID was performed after mobility separation in thetransfer cell. The option for transfer cell instead of trap cell for MS/MS analysis was guided by the fact that, in the transfer cell, thediscrimination of the parent ion isomers, if present, is provided.During MS/MS experiments, collision energies between 10 and45 eVwere employed in order to generate themaximumnumber ofdiagnostic ions. Data acquisition and IMS data processing wereperformed using Waters MassLynx (version V4.1, SCN 855) andWaters Driftscope software (version V2.7).

2.5. Ganglioside abbreviation and assignment of the spectra

The system introduced by Svennerholm [58] and the recom-mendations of IUPAC-IUB Commission on Biochemical Nomencla-ture (IUPAC-IUB 1998) [59] were followed for abbreviating the


gangliosides and the precursor glycosphingolipids. For the assign-ment of the oligosaccharide backbone sequence ions, the nomen-clature introduced by Domon and Costello [60] and revised byCostello et al. [61] was applied.

3. Results and discussions

3.1. IMS MS screening

For screening experiments, 30 mL solution of 5 pmol/mL con-centration of the native ganglioside mixture extracted and purifiedfrom FH17 were loaded into the syringe and submitted to negativeion mode ESI IMS MS. The generated driftscope image (drift timeversus m/z) of the FH17 after six minutes of signal acquisition isdepicted in Fig. 1 together with the resulted total ion chromato-gram (TIC) and the full spectrum. As previously illustrated for thistype of substrates [16], IMS MS provides the separation of GGsbased not just on the charge state, but also on the carbohydratechain length and the degree of sialylation. The same separationpattern was detected in the 2D plot of FH17. Besides, the optimizedionization conditions enhanced the ionization of long chain poly-sialylated structures and allowed the simultaneous formation ofpredominantly multiply charged, as well as of singly chargedmolecules, whose trend lines were denoted with (�4), (�3), (�2)and (�1) in Fig. 1. The driftscope display depicted in Fig. 1 was usedto generate the corresponding mass spectrum for each GG class, byextracting the drift time for every narrow region indicated withwhite color in Fig. 1 and by combining the extracted ion chro-matograms (XICs). Thus, four spectra of singly charged speciescorresponding to GM3, GM2, GM1þGD3 and GD1 (Fig. 2a) weregenerated, while five spectra related to doubly charged GM1þGD3,GD2, GD1þGT3, GT1 and GQ1 were generated from the XIC(Fig. 2b). After retaining the drift time for the indicated triply andquadruply charged structures in Fig. 1, three spectra correspondingto GT1 and GQ1 triply and GQ1 quadruply charged species weregenerated (Fig. 2c). By combining direct ionization with IMS andhigh-resolution MS, no less than 131 distinct gangliosides wereidentified, a number eight times higher than the number of GG

Fig. 1. Driftscope display (drift time versus m/z) of the negative ions from FH17 sample. Ganand (�4) representing singly, doubly, triply and quadruply deprotonated species. Besides chaon the carbohydrate chain length and the degree of sialylation. Left: integrated mass spectrusample.

structures previously reported in FH following chip nanoESI QTOFMS analysis without prior separation [27].

Besides mass calculation, the previously acquired knowledge onthis type of substrates [16,27,33,34,57,62,63], as well as thebiosynthesis pathway criteria were taken into account for theganglioside structure postulation. The ions detected by (�) ESI IMSMS in FH17 are listed in Table 1, along with their putative structuralassignment. A remarkably rich molecular ion pattern, proving thepresence of a large number of glycoforms and an unpredicted di-versity of the ceramide chains for certain species, was observed. Ifin the previous study [27], the maximum degree of sialylation wasfound to be three, here, by IMS-MS, the entire series frommono- upto tetrasialylated GGs was observed. The ionization and thedetection of the low-abundance species, i.e. polysialylated GGs,were possible due to the ionmobility separation performed prior toMS analysis. As compared to previous chip MS-based assays, whereonly four GG structures presenting modifications were detected,IMS MS evidenced no less than 54 different species modified byeither carbohydrate or non-carbohydrate type of attachments.From the monosialo group encompassing 19 GG species, one O-acetylated GM1 was detected, while in the disialo group containing54 different GGs, four GD1 were found modified by O-acetylationand 13 GD1 by fucosylation. From the 47 trisialogangliosidesdetected in FH17, three GT1 and seven GT3 were fucosylated, whilethree GT1 and six GT3 were identified as being acetylated.Furthermore, IMS MS revealed in FH17 the presence of three GT1species modified by N-acetyl galactosamine attachment. The pre-sent approach disclosed also for the first time a high variability ofGG fatty acid patterns in human hippocampus. By IMS MS struc-tures having from 12 to 28 carbon atoms in their ceramide werediscovered, while in the same sample, by using no separationmethod prior to MS, were detected only species with fatty acidchains containing from 14 to 26 carbon atoms [27]. Whereas onlytwo species with a trihydroxylated sphingoid base were previouslyreported for fetal hippocampus [27], here, from the total number ofGGs identified, 115 exhibit dihydroxylated and 16 trihydroxylatedsphingoid bases. The prevalence of human ganglioside speciesencompassing trihydroxylated sphingoid bases was previously

glioside ions separate in the drift cell into charge groups labeled with (�1), (�2), (�3)rge state separation, them/z vs. drift time plot reveals ganglioside separation also basedm over the entire range of drift time distributions. Up: Total ion chromatogram of FH17

Fig. 2. Extracted mass spectra of a) singly charged GM3, GM2, GM1þGD3 and GD1 gangliosides, b) doubly charged GM1þGD3, GD2, GD1þGT3, GT1 and GQ1, c) triply charged GT1and GQ1 and quadruply charged GQ1 (down) from the corresponding areas indicated in Fig. 1.


Fig. 2. (continued).


reported for both fetal and adult human brain. The identification inseveral human fetal brain regions, such as frontal lobe in the 27th[64], 36th [26] and 37th [16] g.w. and hippocampus in the 17th [27]g.w. and in various adult human brain regions, such as cerebellum[55], sensory and motor cortex [65] of (t18:1) or (t18:0) sphigoidbases illustrates a process associated to intrauterine and extra-uterine development [16,62].

Interestingly, by IMS MS we have identified in FH17 a series ofions which, according to the exact mass calculation, correspond toganglioside structures modified by CH3COO� attachment (furtherexplanations in Section 3.2). The highest prevalence of this unusualmodification, never reported before, was observed for GD1, GT3 andGT1 species in a number of 12, 6 and 5, respectively, while only oneGM1 and GD2 structures were detected as bearing CH3COO�. Onthe other hand, no GM2, GM3, GD3 or GQ1 structures modified byCH3COO� were distinguished. Since, according to our observations,CH3COO� provides the charge state of themolecule, such structuresdetected in FH17 were designated in Table 1 by Mn� where n rep-resents the number of CH3COO� radicals attached to the ganglio-side core.

3.2. IMS CID MS/MS structural characterization

Since the MS screening of FH17 mixture has revealed novelgangliosides modified by CH3COO�, the doubly charged species atm/z 1115.025, which, according to mass calculation (Table 1), cor-responds to triply deprotonated doubly sodiated (CH3COO�)GT1(d18:1/18:0) was isolated and submitted to CID MS/MS

experiments in order to elucidate its structural composition. TheIMS CID MS/MS results are presented in Fig. 3, while the schemeillustrating the fragmentation experienced by this species underthe employed sequencing conditions is depicted in Fig. 4. Thespectrum in Fig. 3, generated by combining 360 scans acquired forsix minutes under variable collision energy within 10 eV and 35 eV,exhibits an elevated number of diagnostic fragment ions, mostly Y-and B-type generated by the cleavage of the glycosidic linkages.

Inspection of the spectrum in Fig. 3 reveals that the postulated(d18:1/18:0) ceramide is confirmed by the Y0

� fragment ion at m/z564.530. The reducing end of the molecule is further documentedby Y1

� atm/z 726.544, [Y2a/B2b]- atm/z 888.572 and [Z3a/B1b]- atm/z1364.777, corresponding to Glc-Cer, Gal-Glc-Cer and GalNAc-Gal(Neu5Ac)-Glc-Cer. The signals detected at m/z 603.144, 621.144and 625.123 corresponding to [B2b-2HþþNaþ]-, [C2b-2HþþNaþ]-

and respectively [B2b-3Hþþ2Naþ]- provide evidence on the occur-rence of the disialo element (Neu5Ac)2 attached to the oligosac-charide backbone. Together with the monodeprotonated doublysodiated Z2a� and Y2a

� fragment ions identified at m/z 1496.741 andm/z 1514.753, these ions support (Neu5Ac)2-Gal-Glc-Cer sequence,an indicative of the disialo element attachment to the innergalactose of the molecule.

On the other hand, the sequence of the non-reducing end of themolecule, as well as the attachment of another Neu5Ac residue atthe external galactose of the molecule are substantiated by thefollowing fragment ions detected at m/z 452.154, 470.159, 474.155and 492.141, as B2a

� , C2a� , [B2a-2HþþNaþ]- and [C2a-2HþþNaþ]- cor-

responding to Neu5Ac-Gal and by the fragment ions detected atm/z

Table 1Assignment of major ionic species detected in the FH17 mixture by IMS MS.

Curr. No. m/z exp. m/z theor. Mass accuracy (ppm) Proposed structure Molecular ion

1 596.768 596.774 9.2 GQ1(d18:1/16:0) [M-4Hþ]4-

2 603.775 603.782 10.8 GQ1(d18:1/18:0) [M-4Hþ]4-

3 610.781 610.790 13.3 GQ1(d18:1/20:0) [M-4Hþ]4-

4 624.291 624.301 15.6 GQ1(d18:1/24:1) [M-4Hþ]4-

5 674.857 674.870 19.3 GM2(d18:1/16:2) [M-2Hþ]2-

6 689.652 689.658 8.7 GT1(d18:1/14:0) [M-3Hþ]3-

7 698.998 699.002 5.7 GT1(d18:1/16:0) [M-3Hþ]3-

8 707.669 707.674 7.1 GT1(d18:1/18:1) [M-3Hþ]3-

9 708.340 708.346 8.5 GT1(d18:1/18:0) [M-3Hþ]3-

10 713.011 713.005 8.4 GT1(t18:1/18:1) [M-3Hþ]3-

11 713.670 713.678 10.8 GT1(t18:1/18:0) [M-3Hþ]3-

12 714.343 714.349 8.4 GT1(t18:0/18:0) [M-3Hþ]3-

13 715.666 715.673 9.8 GT1(d18:1/18:0) [M-4HþþNaþ]3-

14 717.025 717.018 9.8 GT1(d18:1/20:1) [M-3Hþ]3-

15 717.685 717.690 7.0 GT1(d18:1/20:0) [M-3Hþ]3-

16 720.886 720.894 11.1 GD3(d18:1/16:0) [M-2Hþ]2-

17 726.358 726.361 4.1 GT1(d18:1/22:1) [M-3Hþ]3-

18 727.028 727.033 6.9 GT1(d18:1/22:0) [M-3Hþ]3-

19 734.899 734.910 15.0 GD3(d18:1/18:0) [M-2Hþ]2-

20 735.670 735.679 12.2 (CH3COO�) GT1(d18:1/18:0) [M�-3HþþNaþ]3-

21 745.895 745.901 8.1 GD3(d18:1/18:0) [M-2HþþNaþ]-

22 748.935 748.926 12.0 GD3(d18:1/20:0) [M-2Hþ]2-

23 757.922 757.912 13.2 GM1(d18:1/16:1) [M-2Hþ]2-

757.932 13.2 GD3(t18:0/20:0) [M-2Hþ]2-

24 770.930 770.920 13.0 GM1(d18:1/18:1) [M-2Hþ]2-

25 771.914 771.928 18.2 GM1(d18:1/18:0) [M-2Hþ]2-

26 775.937 775.949 15.5 GD3(d18:1/24:1) [M-2Hþ]2-

27 776.943 776.957 18.0 GD3(d18:1/24:0) [M-2Hþ]2-

28 782.909 782.919 12.8 GM1(d18:1/18:0) [M-3HþþNaþ]2-

29 785.929 785.944 19.1 GM1(d18:1/20:0) [M-2Hþ]2-

30 789.951 789.965 17.7 GD3(d18:1/26:1) [M-2Hþ]2-

31 794.010 794.018 10.1 GQ1(d18:1/16:3) [M-3Hþ]3-

32 799.975 799.959 20.0 GM1(d18:0/22:0) [M-2Hþ]2-

33 803.355 803.361 7.5 GQ1(d18:1/18:3) [M-3Hþ]3-

34 805.372 805.377 6.2 GQ1(d18:1/18:0) [M-3Hþ]3-

35 811.933 811.947 17.3 GD3(d18:1/26:1) [M-4Hþþ2Naþ]2-

36 812.026 812.033 8.6 GQ1(d18:1/18:1) [M-4HþþNaþ]3-

37 812.712 812.705 8.6 GQ1(d18:1/18:0) [M-4HþþNaþ]3-

38 812.943 812.929 17.2 (CH3COO�) GM1(d18:1/18:0) [M�-2HþþNaþ]2-

39 822.041 822.049 9.7 GQ1(d18:1/20:0) [M-3Hþ]3-

40 822.448 822.434 17.0 GD2(d18:1/16:0) [M-2Hþ]2-

41 827.429 827.444 18.1 GD2(d18:1/18:0) [M-2Hþ]2- (-H2O)42 828.462 828.452 12.1 GD2(d18:0/18:0) [M-2Hþ]2- (-H2O)43 830.712 830.720 9.6 GQ1(d18:1/22:1) [M-3Hþ]3-

44 831.387 831.392 6.0 GQ1(d18:1/22:0) [M-3Hþ]3-

45 831.393 831.409 19.3 GD2(d18:1/16:2) [M-3HþþNaþ]2-

46 835.431 835.441 12.0 GD2(d18:1/18:1) [M-2Hþ]2-

47 836.438 836.449 13.2 GD2(d18:1/18:0) [M-2Hþ]2-

48 840.071 840.064 8.3 GQ1(d18:1/24:1) [M-3Hþ]3-

49 847.424 847.440 18.9 GD2(d18:1/18:0) [M-3HþþNaþ]2-

50 850.482 850.465 20.0 GD2(d18:1/20:0) [M-2Hþ]2-

51 863.459 863.474 16.8 GD2(d18:1/22:1) [M-2Hþ]2-

52 877.438 877.450 13.6 (CH3COO�) GD2(d18:1/18:0) [M�-2HþþNaþ]2-

53 889.453 889.445 9.0 GD1(d18:1/14:0) [M-2Hþ]2-

889.463 11.2 GT3(t18:0/18:0) [M-2Hþ]2-

54 903.463 903.460 3.3 GD1(d18:1/16:0) [M-2Hþ]2-

903.479 17.7 GT3(t18:0/20:0) [M-2Hþ]2-

55 914.460 914.451 9.8 GD1(d18:1/16:0) [M-3HþþNaþ]2-

914.470 10.9 GT3(t18:0/20:0) [M-3HþþNaþ]2-

914.471 12.0 O-Ac-GT3(d18:1/20:1) [M-2Hþ]2-

56 916.457 916.467 10.9 GD1(d18:1/18:1) [M-2Hþ]2-

57 917.465 917.475 10.9 GD1(d18:1/18:0) [M-2Hþ]2-

58 927.461 927.458 3.2 GD1(d18:1/18:1) [M-3HþþNaþ]2-

59 928.459 928.466 7.5 GD1(d18:1/18:0) [M-3HþþNaþ]2-

60 931.478 931.491 14.0 GD1(d18:1/20:0) [M-2Hþ]2-

931.479 1.1 O-Ac-GT3(d18:1/22:0) [M-3HþþNaþ]2- (-H2O)61 935.484 935.474 10.7 GD1(d18:1/19:0) [M-3HþþNaþ]2-

62 936.469 936.482 13.9 GD1(d18:0/19:0) [M-3HþþNaþ]2-

63 939.441 939.457 17.0 GD1(d18:1/18:0) [M-4Hþþ2Naþ]2-

64 942.469 942.482 13.8 GD1(d18:1/20:0) [M-3HþþNaþ]2-


944.480 10.5 (CH3COO�) GT3(t18:0/20:0) [M�-2HþþNaþ]2-

944.481 11.6 (CH3COO�) O-Ac-GT3(d18:1/20:1) [M�-Hþ]2-

(continued on next page)


Table 1 (continued )


66 945.501 945.507 6.3 GD1(d18:1/22:0) [M-2Hþ]2-

945.495 6.3 O-Ac-GT3(d18:1/24:0) [M-3HþþNaþ]2- (-H2O)67 946.660 946.477 11.6 (CH3COO�) GD1(d18:1/18:1) [M�-Hþ]2-

68 950.469 950.480 11.6 O-Ac-GD1(d18:1/20:2) [M-2Hþ]2-

950.468 1.1 O-Ac-GT3(d18:1/22:1) [M-4Hþþ2Naþ]2-

69 951.499 951.488 11.6 O-Ac-GD1(d18:1/20:1) [M-2Hþ]2-

70 953.513 953.504 9.4 O-Ac-GD1(d18:0/20:0) [M-2Hþ]2-

953.505 8.4 GD1(t18:1/22:0) [M-2Hþ]2-





75 968.481 968.497 16.5 GD1(d18:1/24:2) [M-3HþþNaþ]2-

76 969.457 969.467 10.3 (CH3COO�) GD1(d18:1/18:0) [M�-3Hþþ2Naþ]2-


78 976.480 976.487 7.7 (CH3COO�)2 GD1(d18:1/18:1) M2-

79 977.479 977.497 18.4 Fuc-GD1(d18:0/16:0) [M-2Hþ]2-

977.472 7.2 GD1(d18:1/22:1) [M-5Hþþ3Naþ]2-

977.486 7.2 Fuc-GT3(d18:1/20:1) [M-3HþþNaþ]2-

80 979.495 979.488 7.2 GD1(d18:1/24:2) [M-4Hþþ2Naþ]2-

979.502 7.2 Fuc-GT3(d18:1/22:2) [M-2Hþ]2-

81 980.479 980.490 11.2 (CH3COO�) O-Ac-GD1(d18:1/20:2) [M�-Hþ]2-

980.478 1.0 (CH3COO�) O-Ac-GT3(d18:1/22:1) [M�-3Hþþ2Naþ]2-

82 987.492 987.478 14.1 (CH3COO�)2 GD1(d18:1/18:1) [M2--HþþNaþ]2-


84 991.509 991.512 3.0 Fuc-GD1(d18:0/18:0) [M-2Hþ]2-

991.502 7.1 Fuc-GT3(d18:1/24:4) [M-2Hþ]2-

85 992.524 992.509 15.1 Fuc-GT3(d18:1/24:3) [M-2Hþ]2-


87 999.469 999.477 8.0 (CH3COO�)2 GD1(d18:1/18:0) [M2--2Hþþ2Naþ]2-


89 1003.508 1003.513 5.0 Fuc-GD1(d18:1/20:1) [M-2Hþ]2-

90 1007.490 1007.507 16.8 (CH3COO�) Fuc-GD1(d18:0/16:0) [M�-Hþ]2-

1007.482 7.9 (CH3COO�) GD1(d18:1/22:1) [M�-4Hþþ3Naþ]2-

1007.496 5.9 (CH3COO�) Fuc-GT3(d18:1/20:1) [M�-2HþþNaþ]2-

91 1009.501 1009.498 2.9 (CH3COO�) GD1(d18:1/24:2) [M�-3Hþþ2Naþ]2-

1009.512 10.9 (CH3COO�) Fuc-GT3(d18:1/22:2) [M�-Hþ]2-

92 1010.492 1010.500 7.9 (CH3COO�)2 O-Ac-GD1(d18:1/20:2) M2-

1010.488 3.9 (CH3COO�)2 O-Ac-GT3(d18:1/22:1) [M2--2Hþþ2Naþ]2-

93 1013.482 1013.496 13.8 Fuc-GD1(d18:1/20:2) [M-3HþþNaþ]2-

94 1018.953 1018.960 6.9 GT1(d18:1/12:2) [M-2Hþ]2-

95 1021.515 1021.522 6.8 (CH3COO�) Fuc-GD1(d18:0/18:0) [M�-Hþ]2-

1021.512 2.9 (CH3COO�) Fuc-GT3(d18:1/24:4) [M�-Hþ]2-

96 1024.517 1024.536 18.6 Fuc-GD1(d18:1/23:1) [M-2Hþ]2-

97 1025.526 1025.544 17.6 Fuc-GD1(d18:1/23:0) [M-2Hþ]2-

98 1040.504 1040.510 5.7 (CH3COO�)3 O-Ac-GD1(d18:1/20:2) [M3-þHþ]2-

1040.498 5.7 (CH3COO�)3 O-Ac-GT3(d18:1/22:1) [M3--Hþþ2Naþ]2-

99 1050.534 1050.551 16.2 Fuc-GD1(d18:1/25:0) [M-3HþþNaþ]2-

100 1051.529 1051.532 2.8 (CH3COO�)2 Fuc-GD1(d18:0/18:0) M2-

1051.522 6.6 (CH3COO�)2 Fuc-GT3(d18:1/24:4) M2-

101 1055.536 1055.554 17.0 (CH3COO�) Fuc-GD1(d18:1/23:0) [M�-Hþ]2-

102 1056.542 1056.551 8.5 Fuc-GD1(d18:1/26:1) [M-3HþþNaþ]2-

103 1059.987 1059.999 11.3 GT1(d18:1/16:0) [M-3HþþNaþ]2-

104 1062.532 1062.550 16.9 Fuc-GD1(d18:0/25:0) [M-4Hþþ2Naþ]2-

105 1063.003 1063.023 18.8 GT1(d18:1/18:0) [M-2Hþ]2-

106 1069.545 1069.559 13.1 Fuc-GD1(d18:1/28:2) [M-3HþþNaþ]2-

107 1070.516 1070.520 3.7 (CH3COO�)4 O-Ac-GD1(d18:1/20:2) [M4�þ2Hþ]2-

1070.508 7.4 (CH3COO�)4 O-Ac-GT3(d18:1/22:1) [M4�þ2Naþ]2-

108 1072.986 1073.006 18.6 GT1(d18:1/18:1) [M-3HþþNaþ]2-

109 1073.998 1074.014 14.9 GT1(d18:1/18:0) [M-3HþþNaþ]2-

110 1081.993 1082.012 17.6 O-Ac-GT1(d18:1/18:2) [M-2Hþ]2-

111 1085.017 1085.005 11.1 GT1(d18:1/18:0) [M-4Hþþ2Naþ]2-

112 1088.012 1088.031 17.5 GT1(d18:1/22:3) [M-2Hþ]2-

1088.030 16.5 GT1(d18:1/20:0) [M-3HþþNaþ]2-

113 1093.015 1093.033 16.4 (CH3COO�) GT1(d18:1/18:0) [M�-Hþ]2-

114 1095.985 1095.997 11.0 GT1(d18:1/20:3) [M-4Hþþ2Naþ]2-

115 1097.993 1098.013 18.2 GT1(d18:1/20:1) [M-4Hþþ2Naþ]2-

116 1103.007 1103.016 8.5 (CH3COO�) GT1(d18:1/18:1) [M�-2HþþNaþ]2-

117 1107.023 1107.038 13.6 GT1(d18:1/23:2) [M-3HþþNaþ]2-

118 1111.043 1111.052 8.1 GT1(t18:1/24:2) [M-2Hþ]2-

119 1114.034 1114.045 9.9 GT1(d18:1/18:1) [M-4Hþþ2Naþ]2-

120 1115.025 1115.015 8.9 (CH3COO�) GT1(d18:1/18:0) [M�-3Hþþ2Naþ]2-

121 1123.030 1123.043 11.5 (CH3COO�)2 GT1(d18:1/18:0) M2-

122 1125.031 1125.036 4.4 GT1(d18:1/24:2) [M-4Hþþ2Naþ]2-



Table 1 (continued )


124 1129.057 1129.069 10.6 GT1(d18:1/26:1) [M-3HþþNaþ]2-

1129.045 10.6 Fuc-GT1(d18:1/17:0) [M-2Hþ]2-

125 1134.052 1134.073 18.5 O-Ac-GT1(d18:0/24:0) [M-2Hþ]2-

126 1139.646 1139.666 17.6 GM3(t18:1/14:0) [M-Hþ]-

127 1148.512 1148.531 16.6 GalNAc-GT1(d18:1/16:2) [M-2Hþ]2-

128 1149.518 1149.539 18.3 GalNAc-GT1(d18:1/16:1) [M-2Hþ]2-


130 1156.009 1156.017 6.9 (CH3COO�)2 GT1(d18:1/20:3) [M2--2Hþþ2Naþ]2-

131 1164.064 1164.083 16.3 (CH3COO�) O-Ac-GT1(d18:0/24:0) [M�-Hþ]2-

132 1165.695 1165.681 12.0 GM3(t18:1/16:0) [M-Hþ]-

133 1167.038 1167.059 18.0 Fuc-GT1(d18:1/21:1) [M-3HþþNaþ]2-

134 1167.677 1167.697 17.1 GM3(t18:1/16:1) [M-Hþ]-

135 1175.054 1175.075 17.9 Fuc-GT1(d18:1/22:0) [M-3HþþNaþ]2-

136 1178.526 1178.541 12.7 (CH3COO�) GalNAc-GT1(d18:1/16:2) [M�-Hþ]2-

137 1179.714 1179.733 16.1 GM3(d18:1/18:0) [M-Hþ]-

138 1190.501 1190.523 18.5 GQ1(d18:1/14:1) [M-3HþþNaþ]2-

139 1195.751 1195.728 19.2 GM3(t18:1/18:0) [M-Hþ]-

140 1216.549 1216.568 15.6 GQ1(t18:1/18:0) [M-2Hþ]2-

141 1230.537 1230.553 13.0 GQ1(d18:1/18:0) [M-4Hþþ2Naþ]2-

142 1271.578 1271.592 11.0 GQ1(d18:1/24:1) [M-4Hþþ2Naþ]2-

143 1354.757 1354.781 17.7 GM2(d18:1/16:0) [M-Hþ]-

144 1380.770 1380.796 18.8 GM2(d18:1/18:1) [M-Hþ]-

145 1382.786 1382.812 18.8 GM2(d18:1/18:0) [M-Hþ]-

146 1464.751 1464.779 19.1 GD3(d18:1/16:0) [M-2HþþNaþ]-

147 1492.837 1492.810 18.1 GD3(d18:1/18:0) [M-2HþþNaþ]-

148 1516.806 1516.833 17.8 GM1(d18:1/16:0) [M-Hþ]-

149 1542.867 1542.848 12.3 GM1(d18:1/18:1) [M-Hþ]-

150 1544.894 1544.864 19.4 GM1(d18:1/18:0) [M-Hþ]-

151 1566.872 1566.846 16.6 GM1(d18:1/18:0) [M-2HþþNaþ]-

152 1572.917 1572.896 10.2 GM1(d18:1/20:0) [M-Hþ]-

1572.933 13.4 GD3(t18:0/24:0) [M-Hþ]-

153 1582.851 1582.841 6.3 GM1(t18:1/18:0) [M-2HþþNaþ]-

1582.842 5.7 O-Ac-GM1(d18:1/18:2) [M-Hþ]-

154 1626.884 1626.866 11.0 (CH3COO�) GM1(d18:1/18:0) [M�-HþþNaþ]-

155 1648.942 1648.924 10.9 GM1(d18:0/24:0) [M-2HþþNaþ]-

156 1829.922 1829.910 6.6 GD1(d18:1/16:0) [M-2HþþNaþ]-

157 1855.937 1855.925 6.5 GD1(d18:1/18:1) [M-2HþþNaþ]-

158 1857.963 1857.941 11.8 GD1(d18:1/18:0) [M-2HþþNaþ]-

159 1871.937 1871.957 10.7 GD1(d18:1/19:0) [M-2HþþNaþ]-

160 1873.959 1873.973 7.5 GD1(d18:0/19:0) [M-2HþþNaþ]-

161 1879.939 1879.923 8.5 GD1(d18:1/18:0) [M-3Hþþ2Naþ]-

162 1885.956 1885.973 9.0 GD1(d18:1/20:0) [M-2HþþNaþ]-

1885.974 9.5 GD1(d18:1/22:3) [M-Hþ]-

163 1939.952 1939.943 4.6 (CH3COO�) GD1(d18:1/18:0) [M�-2Hþþ2Naþ]-


655.196, 673.191, 677.163 and 695.237 as B3a� , C3a

� , [B3a-2HþþNaþ]-

and [C3a-2HþþNaþ]-, respectively, which are associated withNeu5Ac-Gal-GalNAc chain. The fragment ions corresponding toNeu5Ac-Gal-GalNAc-Gal-(Neu5Ac)2 sequence observed at m/z721.186 and 1461.208 as [B4-4Hþþ2Naþ]2- and its counterpart [C4-4Hþþ2Naþ]2- and to (Neu5Ac)3Gal2GlcGalNAc moiety at m/z802.172 as [B5-4Hþþ2Naþ]2- support not just the triple sialylationof the molecule, but also the sequence of the entire saccharidechain. Furthermore, a series of fragment ions generated byGT1b(d18:1/18:0) desialylation are also observed in the spectrumillustrated in Fig. 3. This is the case of the doubly and singly chargedfragment ions at m/z 917.475, 1835.890, 1839.856, 1857.921,1861.875 and 1879.975 corresponding to [Y4a or Y3b -2Hþ]2-, [Y4a orY3b -Hþ]-, [Z4a or Z3b -2HþþNaþ]-, [Y4a or Y3b -2HþþNaþ]-, [Z4a orZ3b -3Hþþ2Naþ]- and [Y4a or Y3b -3Hþþ2Naþ]- generated after thedetachment of one Neu5Ac residue and at m/z 772.163, 1544.784,1566.776 and 1588.799 assigned to [Y2b or Y3b/B1a -2Hþ]2-, [Y2b orY3b/B1a -Hþ]-, [Y2b or Y3b/B1a -2HþþNaþ]- and [Y2b or Y3b/B1a-3Hþþ2Naþ]- generated after the detachment of two Neu5Acresidues.

Interestingly, the most abundant signal identified in the spec-trum in Fig. 3 as a doubly charged ion at m/z 1085.025 correspondsto a 59 Da mass loss from the precursor ion, which is consistentwith CH3COO� radical. This detachment of CH3COO� radical from

the precursor ion was also observed as a singly charged ion at m/z2170.994, which, according to mass calculation, is attributable adoubly-sodiated GT1b(d18:1/18:0) composition. Of high relevancefor this concept are a number of fragment ions which occurred bythe cleavage of glycosidic linkages with the preservation ofCH3COO� attachment. The neutral loss of a Neu5Ac residue, of aNeu5Ac residue and a sodium adduct, of a Neu5Ac residue and bothsodium adducts, or of the disialo element and a sodium adduct,respectively from the precursor ion were identified as ions of fairintensities at m/z 1939.946, 1917.958, 1895.969 and 1626.750,respectively. Additionally, the neutral loss of both sodium adductsfrom the precursor ion with the preservation of the CH3COO�

attachment generated the [M�-Hþ]2- visible in Fig. 3 at m/z1093.020. Thereby, the present fragmentation pattern substantiatesa compact unite-like behavior that cannot be interpreted as indi-vidual acetyl and water adduct entities in Fig. 3. Moreover, a virtualCID MS3 for CH3COO� radical detected at m/z 59.021 by enhancingin source fragmentation in MS was conducted in order to elucidateits structure. The obtained fragmentation mass spectrum is pre-sented in Fig. 5. By applying collision energies within 30e45 eV, afragment ion at m/z 44.007 was generated indicating a neutral lossof a 15 mass unit that corresponds to the CH3 detachment from theprecursor ion. Such a fragmentation pathway confirms the acetatestructure.

Fig. 3. (�) ESI IMS CID MS/MS of the [M�-3Hþþ2Naþ]2- at m/z 1115.025 corresponding to (CH3COO�) GT1b(d18:1/18:0) species isolated and fragmented from FH17 sample. Conevoltage 40 V. Capillary voltage 2 kV. Acquisition 360 scans. CID at variable collision energy within 10e35 eV.

Fig. 4. Fragmentation scheme by IMS CID MS/MS of GT1b(d18:1/18:0) moiety.


Since no acetic acid or acetate was involved in the extraction,purification or analysis of FH17, the presence of the acetate in thesample can be explained only on the basis of the biological and

biochemical processes involving GGs that occur in vivo in the brain.The involvement of gangliosides in the synaptic function rep-

resents an aspect broadly known and studied [66e73]. During the

Fig. 5. Virtual (�) ESI IMS CID MS3 of acetate anion detected at m/z 59.021. Cone voltage 40 V. Capillary voltage 2 kV. Acquisition 300 scans. CID at variable collision energy within30e45 eV.


cholinergic transmission, the neurotransmitter acetylcholine (Ach)synthesized by choline acetyltransferase from choline and acetylcoenzyme A (Ac-CoA) is stored and transported within the synapticvesicles up to the presynapses, where, during the membrane de-polarization is released into the synaptic cleft [66]. There, part ofthe Ach binds to acetylcholine receptors, while most of the releasedAch molecules are degraded to acetate and choline, which arefurther re-used for Ach synthesis. Gangliosides, especially thecholinergic-specific antigen (Chol-1a) structures play a pivotal rolein the cholinergic synaptic transmission by stimulating the cholineassimilation by synapses and enhancing Ach synthesis [66e70].Within the treatment of synaptosomes with two neuronal mem-brane gangliosides, the GQ1b was demonstrated to be more effec-tive than GM1 in enhancing the long-term potentiation assumed tobe involved in the cellular basis of learning and memory [74]. Un-like GQ1b that improves the cognitive function by Ach release anddepolarization-evoked calcium influx stimulation, the GT1aa andGQ1ba stereoisomers belonging to the Chol-1a group acceleratethe choline uptake by synapses and enhance Ach synthesis,although have no effects on calcium ion influx [67e69]. The activityof GT1aa and GQ1ba was demonstrated both through in vitro andin vivo studies where anti-Chol-1amonoclonal antibody was addedto a synaptosomal fraction [69]. Synthetic a - and b-stereoisomersof sialylcholesterol employed as GG analogues and added sepa-rately to synaptosomes were demonstrated to mimic the action ofGQ1b and Chol-1a in synapses [66]. Although a research directionwas oriented towards the identification of endogenous GGs insynaptic function, up to now it was not possible to determine whatspecific GGs participate in any particular functions [66].

On the other side, the Ac-CoA molecule, synthesized in the

cytosol migrates via an Ac-CoA-transporter in the Golgi lumenwhere it serves as a substrate for O-acetyltransferases. Along withthe sialyltransferases activity in the Golgi apparatus, O-acetyl-transferases relocate the acetyl moiety of Ac-CoA to the C7 and/orC9 hydroxyl groups on the glycerol-like side chain of a terminal a2-8 linked Neu5Ac residue of gangliosides [75]. Apparently, the pri-mary insertion site for the O-acetyl group is at the C7 position, fromwhere, under pH variation, the O-acetyl groupmight migrate via C8to the C9 position [76], allowing additional O-acetylation reactionto take place at C7 position. Moreover, such a modification of sialicacids has several implications, since: i) decreases of GG polarity andhydrophobicity; ii) prevents subsequent sialylation of GGs [70,77],iii) prevents a fast degradation of sialylated glycoconjugates [78].

Therefore, considering all the synaptic processes enhanced bygangliosides, we believe that CH3COO� was attached during thecholinergic activity of GGs within the brain. These new findingssuggest that the 25 different GG species of which 12 belonging toGD1, 6 to GT3, 5 to GT1, 1 to GM1 and 1 to GD2 series, all foundmodified by CH3COO�, might belong to the cholinergic-specificgangliosides in fetal hippocampus.

4. Conclusions

In the present study, ion mobility separation mass spectrometrywas applied for the first time for the characterization of a highlycomplex native ganglioside mixture extracted and purified from adefined human brain region, i.e. normal fetal hippocampus in 17thgestational week. The combination of electrospray ionization, ionmobility separation and high resolution mass spectrometry in thenegative ionization mode enhanced ganglioside separation based


not just on the m/z value, but also on the charge state, the carbo-hydrate chain length and the degree of sialylation. In the driftscopeplot (drift time versus m/z) of the FH17 generated after 6 min ofsignal acquisition, 131 distinct gangliosides characterized by highvariability of the oligosaccharide core and diversity of the ceramidemoiety were identified with an average mass accuracy of 12.3 ppm.Under the ESI flow rate of about 1.2 mL/min and using an approxi-mate sample concentration of 5 pmol/mL, 6 min signal acquisitionfor IMS screening spectrum was equivalent with a consumption ofaround 36 pmol, while five and six minutes acquisition, respec-tively for CID MS/MS and virtual MS3 experiments, were equivalentto 30 and 36 pmol sample consumption. Thus, about 100 pmol ofFH17 were employed for the entire study. Ganglioside separationbased on their mobility through a buffer gas together with thepossibility to extract and retain the drift times for narrow regions ofthe 2D plot, allowed the identification of structures exhibiting a lowexpression in the mixture. Hence, by IMSMS were discovered eightfolds more species than previously found in fetal hippocampus byusing chip nanoESI MS, which is nowadays considered one of themost advanced MS technologies. Additionally, the presentapproach revealed a higher variability of fatty acid patterns thanever reported in human hippocampus; structures having from 12 to28 carbon atoms in their ceramidewere detected and assignedwithhigh mass accuracy. IMS MS also evidenced no less than 29different species modified by fucosylation, acetylation and N-acetylgalactosamine attachment. Remarkably, MS screening of FH17 byIMS MS revealed the presence of 25 ions corresponding to novelganglioside structures, modified by CH3COO� attachment. All theseGG species might be involved in cholinergic activity, an aspectwhich was here, for the first time, emphasized by MS through thediscovery of their CH3COO� modification. CID MS/MS experimentsconducted after mobility separation in the transfer cell provided adetailed structural characterization of CH3COO� modifiedGT1b(d18:1/18:0), a new and biologically relevant compound. Un-der optimized fragmentation conditions, using variable collisionenergy, a high number of fragment ions diagnostic for the proposednovel species were generated. The sequence ions supported notonly the postulated sugar core, structure and ceramide composi-tion, but also the modification by CH3COO� attachment. In view ofthe present findings, wemay conclude that IMSMS and CIDMS/MSproved their high efficiency in unequivocally detecting and char-acterizing even low abundant glycolipid species, which might playa significant biological role. Moreover, the discovery of speciesmodified by CH3COO�, raises a series of issues regarding thecholinergic activity of gangliosides, which is to be thoroughlyinvestigated in further studies.

Conflict of interest

The authors declare no conflict of interest.

Acknowledgements

This work was carried out within the collaborative EU-USAresearch project FP7 Marie Curie-PIRSES-GA-2010-269256. Thisproject was also supported by the Romanian National Authority forScientific Research, UEFISCDI, projects PN-II- PN-II-PCCA-2013-4-0191 granted to ADZ.

Abbreviations

Ac acetyl/acetylationAc-CoA acetyl coenzyme AAch acetylcholineCE capillary electrophoresis

Cer ceramideCID collision-induced dissociationESI electrospray ionizationFuc fucose/fucosylationFTICR MSFourier transform ion cyclotron resonance mass

spectrometer/spectrometryGG gangliosideGSL glycosphingolipidg.w. gestational weekHCT MS high capacity ion trap mass spectrometer/spectrometryHM high mass resolutionHPLC high-performance liquid chromatographyIMS ion mobility spectrometry/separationLM low mass resolutionMS mass spectrometer/spectrometryMS/MS tandem mass spectrometrym/z mass-to-charge rationanoESI nanoelectrospray ionizationNeu5Ac N-acetyl neuraminic acidQTOF MSquadrupole time-of-flight mass spectrometer/

spectrometryTIC total ion chromatogramTLC thin-layer chromatographyTOF time-of-flightUPLC ultra high-performance liquid chromatographyXIC extracted ion chromatogram

Abbreviations used for gangliosidesGD3 II3-a-(Neu5Ac)2-LacCerGT3 II3-a-(Neu5Ac)3-LacCerGM2 II3-a-Neu5Ac-Gg3CerGD2 II3-a-(Neu5Ac)2-Gg3CerGM1 II3-a-Neu5Ac-Gg4CerGalNAc-GM1b IV3-a-Neu5Ac-Gg5CerGD1a IV3-a-Neu5Ac,II3-a- Neu5Ac-Gg4CerGD1b II3-a-(Neu5Ac)2-Gg4CerGT1b IV3-a-Neu5Ac,II3-a-(Neu5Ac)2-Gg4CerGQ1b IV3-a-(Neu5Ac)2,II3-a-(Neu5Ac)2-Gg4Cer

References

[1] M. Cohen, A. Varki, The sialome-far more than the sum of its parts, OMICS 14(2010) 455e464.

[2] T. Kolter, Ganglioside biochemistry, ISRN Biochem. 2012 (2012) 506160.[3] Y. Horng Tan, Scanning probe microscopy for the study of interactions

involving glycoproteins and carbohydrates, in: K.J. Stine (Ed.), CarbohydrateNanotechnology, John Wiley & Sons, New Jersey, 2015, pp. 285e309.

[4] E. Klenk, Beitrage zur chemie der lipidosen, Hoppe-Seyler's Z. Physiol. Chem.262 (1939) 128e143.

[5] E. Klenk, Uber die ganglioside, eine neue gruppe von zuckerhaltigen gehirn-lipoiden, Hoppe-Seyler's Z. Physiol. Chem. 273 (1942) 76e86.

[6] E. Klenk, Die fettstoffe des gehirns bei amaurotischer idiotie und Niemann-Pickscher Krankheit, Ber. Ges. Physiol. 96 (1937) 659e660.

[7] E. Klenk, Uber die ganglioside des gehirns bei der infantile amaurotischenIdiotie vom typ Tay-Sachs, Ber. Dtsch. Chem. Ges. 75 (1942) 1632e1636.

[8] R. Kuhn, H. Wiegandt, Die konstitution der ganglio-N-tetraose und des gan-gliosides GI, Chem. Ber. 96 (1963) 866e880.

[9] L. Svennerholm, The gangliosides, J. Lipid Res. 5 (1964) 145e155.[10] H. Jatzkewitz, K. Sandhoff, On a biochemically special form of infantile

amaurotic idiocy, Biochim. Biophys. Acta 70 (1963) 354e356.[11] K. Sandhoff, Variation of b-N-acetylhexosaminidase-pattern in Tay-Sachs

disease, FEBS Lett. 4 (1969) 351e354.[12] K. Sandhoff, U. Andreae, H. Jatzkewitz, Deficient hexosaminidase activity in an

exceptional case of Tay-Sachs disease with additional storage of kidney glo-boside in visceral organs, Pathol. Eur. 3 (1968) 278e285.

[13] H.H. Hess, A. Pope, in: A. Lajtha (Ed.), Handbook of Neurochemistry: Patho-logical Chemistry of the Nervous System, Springer Science & Business Media,U.S.A, 2012.

[14] S. Sonnino, L. Mauri, V. Chigorno, A. Prinetti, Gangliosides as components oflipid membrane domains, Glycobiology 17 (2006), 1Re13R.

[15] K. Palmano, A. Rowan, R. Guillermo, J. Guan, P. Mc. Jarrow, The role of gan-gliosides in neurodevelopment, Nutrients 7 (2015) 3891e3913.

http://refhub.elsevier.com/S0300-9084(17)30140-2/sref1














































[16] M. Sarbu, A.C. Robu, R.M. Ghiulai, �Z. Vukeli�c, D.E. Clemmer, A.D. Zamfir,Electrospray ionization ion mobility mass spectrometry of human braingangliosides, Anal. Chem. 88 (2016) 5166e5178.

[17] Y. Sugiura, S. Shimma, Y. Konishi, M.K. Yamada, M. Setou, Imaging massspectrometry technology and application on ganglioside study; visualizationof age-dependent accumulation of C20-ganglioside molecular species in themouse hippocampus, PLoS One 3 (2008) e3232.

[18] S. Sonnino, V. Chigorno, Ganglioside molecular species containing C18- andC20-sphingosine in mammalian nervous tissues and neuronal cell cultures,Biochim. Biophys. Acta 1469 (2000) 63e77.

[19] P. Palestini, M. Masserini, A. Fiorilli, E. Calappi, G. Tettamanti, Age-relatedchanges in the ceramide composition of the major gangliosides present in ratbrain subcellular fractions enriched in plasma membranes of neuronal andmyelin origin, J. Neurochem. 61 (1993) 955e960.

[20] M. Sarbu, L. Dehelean, C.V.A. Munteanu, �Z. Vukeli�c, A.D. Zamfir, Assessment ofganglioside age-related and topographic specificity in human brain by Orbi-trap mass spectrometry, Anal. Biochem. 521 (2017), 40e54.

[21] M. Zarei, J. Müthing, J. Peter-Katalini�c, L. Bindila, Separation and identificationof GM1b pathway Neu5Ac- and Neu5Gc gangliosides by on-line nanoHPLC-QToF MS and tandem MS: toward glycolipidomics screening of animal celllines, Glycobiology 20 (2010) 118e126.

[22] Q. Huang, X. Zhou, D. Liu, B. Xin, K. Cechner, H. Wang, A. Zhou, A new liquidchromatography/tandem mass spectrometry method for quantification ofgangliosides in human plasma, Anal. Biochem. 455 (2014) 26e34.

[23] E.R. Sturgill, K. Aoki, P.H. Lopez, D. Colacurcio, K. Vajn, I. Lorenzini, S. Maji�c,W.H. Yang, M. Heffer, M. Tiemeyer, J.D. Marth, R.L. Schnaar, Biosynthesis of themajor brain gangliosides GD1a and GT1b, Glycobiology 22 (2012) 1289e1301.

[24] K. Nakamura, S. Handa, Coomassie brilliant blue staining of lipids on thin-layer plates, Anal. Biochem. 142 (1984) 406e410.

[25] C. Schiopu, A. Serb, F. Capitan, C. Flangea, E. Sisu, �Z. Vukeli�c, M. Przybylski,A.D. Zamfir, Determination of ganglioside composition and structure in hu-man brain hemangioma by chipbased nanoelectrospray ionization tandemmass spectrometry, Anal. Bioanal. Chem. 395 (2009) 2465e2477.

[26] A. Serb, C. Schiopu, C. Flangea, �Z. Vukeli�c, E. Sisu, L. Zagrean, A.D. Zamfir, High-throughput analysis of gangliosides in defined regions of fetal brain by fullyautomated chip-based nanoelectrospray ionization multi-stage mass spec-trometry, Eur. J. Mass Spectrom. 15 (2009) 541e553.

[27] �Z. Vukeli�c, M. Zarei, J. Peter-Katalini�c, A.D. Zamfir, Analysis of human hippo-campus gangliosides by fully-automated chip-based nanoelectrospray tan-dem mass spectrometry, J. Chromatogr. A 1130 (2006) 238e245.

[28] C. Mosoarca, R.M. Ghiulai, C.R. Novaconi, �Z. Vukeli�c, A. Chiriac, A.D. Zamfir,Application of chip-based nanoelectrospray ion trap mass spectrometry tocompositional and structural analysis of gangliosides in human fetal cere-bellum, Anal. Lett. 44 (2011) 1036e1049.

[29] A.F. Serb, E. Sisu, �Z. Vukeli�c, A.D. Zamfir, Profiling and sequencing of gangli-osides from human caudate nucleus by chip nanoelectrospray mass spec-trometry, J. Mass Spectrom. 47 (2012) 1561e1570.

[30] A.D. Zamfir, D. Fabris, F. Capitan, C. Munteanu, �Z. Vukeli�c, C. Flangea, Profilingand sequence analysis of gangliosides in human astrocytoma by high reso-lution mass spectrometry, Anal. Bioanal. Chem. 405 (2013) 7321e7335.

[31] C. Schiopu, �Z. Vukeli�c, F. Capitan, E. Sisu, A.D. Zamfir, Chip-nanoelectrosprayquadrupole time-of-flight tandem mass spectrometry of meningioma gan-gliosides: a preliminary study, Electrophoresis 33 (2012) 1778e1786.

[32] A. Robu, �Z. Vukeli�c, C. Schiopu, F. Capitan, A.D. Zamfir, Mass spectrometry ofgangliosides in extracranial tumors: application to adrenal neuroblastoma,Anal. Biochem. 509 (2016) 1e11.

[33] �Z. Vukeli�c, S. Kalanj Bognar, M. Froesch, L. Bindila, B. Radi�c, M. Allen, J. Peter-Katalini�c, A.D. Zamfir, Human gliosarcoma-associated ganglioside composi-tion is complex and highly distinctive as evidenced by high-performancemass spectrometric determination and structural characterization, Glycobi-ology 17 (2007) 504e515.

[34] A.D. Zamfir, A. Serb, �Z. Vukeli�c, C. Flangea, C. Schiopu, D. Fabris, F. Capitan,E. Sisu, Assessment of the molecular expression and structure of gangliosidesin brain metastasis of lung adenocarcinoma by an advanced approach basedon fully automated chipnanoelectrospray mass spectrometry, J. Am. Soc. MassSpectrom. 22 (2011) 2145e2159.

[35] M.J. Cohen, F.W. Karasek, Plasma chromatography TM - new dimension forgas chromatography and mass spectrometry, J. Chromatogr. Sci. 8 (1970)330e337.

[36] M.A. M€akinen, O.A. Anttalainen, M.E.T. Sillanp€a€a, Ion mobility spectrometryand its applications in detection of chemical warfare agents, Anal. Chem. 82(2010) 9594e9600.

[37] M. M€akinen, M. Nousiainen, M. Sillanp€a€a, Ion spectrometric detection tech-nologies for ultra-traces of explosives: a review, Mass Spectrom. Rev. 30(2011) 940e973.

[38] R.G. Ewing, D.A. Atkinson, G.A. Eiceman, G.J. Ewing, A critical review of ionmobility spectrometry for the detection of explosives and explosive relatedcompounds, Talanta 54 (2001) 515e529.

[39] S.L. Bernstein, D. Liu, T. Wyttenbach, M.T. Bowers, J.C. Lee, H.B. Gray,J.R. Winkler, a-Synuclein: stable compact and extended monomeric structuresand pH dependence of dimer formation, J. Am. Soc. Mass Spectrom. 15 (2004)1435e1443.

[40] R. Heydari, A new HPLC method for the simultaneous determination ofacetaminophen, phenylephrine, dextromethorphan and chlorpheniramine inpharmaceutical formulations, Anal. Lett. 41 (2008) 965e976.

[41] S. Niu, J.N. Rabuck, B.T. Ruotolo, Ion mobility-mass spectrometry of intactprotein-ligand complexes for pharmaceutical drug discovery and develop-ment, Curr. Opin. Chem. Biol. 17 (2013) 809e817.

[42] M. Kliman, J.C. May, J.A. McLean, Lipid analysis and lipidomics by structurallyselective ion mobility-mass spectrometry, Biochim. Biophys. Acta - Mol. CellBiol. Lipids 1811 (2011) 935e945.

[43] J.E. Kyle, X. Zhang, K.K. Weitz, M.E. Monroe, Y.M. Ibrahim, R.J. Moore, J. Cha,X. Sun, E.S. Lovelace, J. Wagoner, S.J. Polyak, T.O. Metz, S.K. Dey, R.D. Smith,K.E. Burnum�Johnson, E.S. Baker, Uncovering biologically significant lipidisomers with liquid chromatography, ion mobility spectrometry and massspectrometry, Analyst 141 (2016) 1649e1659.

[44] I. Milic, M. Kipping, R. Hoffmann, M. Fedorova, Separation and characteriza-tion of oxidized isomeric lipid-peptide adducts by ion mobility mass spec-trometry, J. Mass Spectrom. 50 (2015) 1386e1392.

[45] M. Groessl, S. Graf, R. Knochenmuss, High resolution ion mobility-massspectrometry for separation and identification of isomeric lipids, Analyst140 (2015) 6904e6911.

[46] K. Pagel, D.J. Harvey, Ion mobility-mass spectrometry of complex carbohy-drates: collision cross sections of sodiated N-linked glycans, Anal. Chem. 85(2013) 5138e5145.

[47] J. Hofmann, H.S. Hahm, P.H. Seeberger, K. Pagel, Identification of carbohydrateanomers using ion mobility-mass spectrometry, Nature 526 (2015) 241e244.

[48] M.M. Gaye, R. Kurulugama, D.E. Clemmer, Investigating carbohydrate isomersby IMS-CID-IMS-MS: precursor and fragment ion cross-sections, Analyst 140(2015) 6922e6932.

[49] J.A. Leary, R.L. Miller, W. Wei, R. Schwӧrer, O.V. Zubkova, P.C. Tyler,J.E. Turnbull, Composition, sequencing and ion mobility mass spectrometry ofheparan sulfate-like octasaccharide isomers differing in glucuronic andiduronic acid content, Eur. J. Mass Spectrom. 21 (2015) 245e254.

[50] D.S. Lee, C. Wu, H.H. Hill Jr., Detection of carbohydrates by electrosprayionization-ion mobility spectrometry following microbore high-performanceliquid chromatography, J. Chromatogr. A 822 (1998) 1e9.

[51] M. Sarbu, F. Zhu, J. Peter-Katalini�c, D.E. Clemmer, A.D. Zamfir, Application ofion mobility tandem mass spectrometry to compositional and structuralanalysis of glycopeptides extracted from the urine of a patient diagnosed withSchindler disease, Rapid Commun. Mass Spectrom. 29 (2015) 1929e1937.

[52] S.Y. Vakhrushev, J. Langridge, I. Campuzano, C. Hughes, J. Peter-Katalini�c,Identification of monosialylated N-glycoforms in the CDG urinome by ionmobility tandem mass spectrometry: the potential for clinical applications,Clin. Proteom 4 (2008) 47e57.

[53] N. Otmakhova, E. Duzel, A.Y. Deutch, J. Lisman, The hippocampal-VTA loop:the role of novelty and motivation in controlling the entry of information intolong-term memory, in: G. Baldassarre, M. Mirolli (Eds.), Intrinsically Moti-vated Learning in Natural and Artificial Systems, Springer Berlin Heidelberg,2013, pp. 235e254.

[54] Y. Mu, F.H. Gage, Adult hippocampal neurogenesis and its role in Alzheimer'sdisease, Mol. Neurodegener. 6 (2011) 85.

[55] A. Zamfir, �Z. Vukeli�c, L. Bindila, J. Peter-Katalini�c, R. Almeida, A. Sterling,M. Allen, Fully-automated chip-based nanoelectrospray tandem mass spec-trometry of gangliosides from human cerebellum, J. Am. Soc. Mass Spectrom.15 (2004) 1649e1657.

[56] L. Svennerholm, P. Fredman, A procedure for the quantitative isolation ofbrain gangliosides, Biochim. Biophys. Acta 617 (1980) 97e109.

[57] �Z. Vukeli�c, W. Metelmann, J. Müthing, M. Kos, J. Peter- Katalini�c, Anencephaly:structural characterization of gangliosides in defined brain regions, Biol.Chem. 382 (2001) 259e274.

[58] L. Svennerholm, Ganglioside designation, Adv. Exp. Med. Biol. 125 (1980) 11.[59] IUPAC-IUB joint commission on biochemical nomenclature, Eur. J. Biochem.

257 (1998) 293e298.[60] B. Domon, C.E. Costello, A systematic nomenclature of carbohydrate frag-

mentation in FAB-MS/MS spectra of glycoconjugates, Glycoconj. J. 5 (1988)397e409.

[61] C.E. Costello, P. Juhasz, H. Perreault, New mass spectral approaches toganglioside structure determination, Prog. Brain Res. 101 (1994), 45e61.

[62] R.M. Ghiulai, M. Sarbu, �Z. Vukeli�c, C. Ilie, A.D. Zamfir, Early stage fetalneocortex exhibits a complex ganglioside profile as revealed by high resolu-tion tandem mass spectrometry, Glycoconj. J. 31 (2014) 231e245.

[63] A.D. Zamfir, N. Lion, �Z. Vukeli�c, L. Bindila, J. Rossier, H. Girault, J. Peter-Katalini�c, Thin chip microsprayer system coupled to quadrupole time-of-flightmass spectrometer for glycoconjugate analysis, Lab. Chip 5 (2005) 298e307.

[64] R. Almeida, C. Mosoarca, M. Chirita, V. Udrescu, N. Dinca, �Z. Vukeli�c, M. Allen,A.D. Zamfir, Coupling of fully automated chip-based electrospray ionization tohigh-capacity ion trap mass spectrometer for ganglioside analysis, Anal. Bio-chem. 378 (2008) 43e52.

[65] C. Flangea, C. Fabris, �Z. Vukeli�c, A.D. Zamfir, Mass spectrometry of gangliosidesfrom human sensory and motor cortex, Aus. J. Chem. 66 (2013) 781e790.

[66] S. Ando, Neuronal dysfunction with aging and its amelioration, Proc. Jpn.Acad. 88 (2012) 266e282.

[67] S. Ando, Y. Hirabayashi, K. Kon, F. Inagaki, S. Tate, V.P. Whittaker,A trisialoganglioside containing a sialyl a2-6 N-acetylgalactosamine residue isa cholinergic-specific antigen, Chol-1a1, J. Biochem. 111 (1992) 287e290.

[68] Y. Hirabayashi, T. Nakao, F. Irie, V.P. Whittaker, K. Kon, S. Ando, Structuralcharacterization of a novel cholinergic neuron-specific ganglioside in bovinebrain, J. Biol. Chem. 267 (1992) 12973e12978.

[69] S. Ando, Y. Tanaka, S. Kobayashi, F. Fukui, M. Iwamoto, H. Waki, T. Tai,












































































































































































































































































Y. Hirabayashi, Synaptic function of cholinergic-specific Chol-1 ganglioside,Neurochem. Res. 29 (2004) 857e867.

[70] Y. Hirabayashi, A. Kanamori, K.H. Nomura, K. Nomura, The acetyl-CoA trans-porter family SLC33, Pflugers Arch. 447 (2004) 760e762.

[71] R. Buccoliero, J. Bodennec, A.H. Futerman, The role of sphingolipids inneuronal developmental: lessons from models of sphingolipid storage dis-eases, Neurochem. Res. 27 (2002) 565574.

[72] H. Rahmann, Brain gangliosides and memory formation, Behav. Brain Res. 66(1995) 105116.

[73] B. Wang, J. Brand-Miller, The role and potential of sialic acid in humannutrition, Eur. J. Clin. Nutr. 57 (2003) 13511369.

[74] S. Fujii, K. Igarashi, H. Sasaki, H. Furuse, K. Ito, K. Kaneko, H. Kato, J. Inokuti,H. Waki, S. Ando, Effects of mono- and tetrasialogangliosides GM1 and GQ1bon ATP-induced long-term potentiation in hippocampal CA1 neurons, Gly-cobiology 12 (2002) 339e344.

[75] J. Fleurence, S. Fougeray, M. Bahri, D. Cochonneau, B. Cl�emenceau, F. Paris,A. Heczey, S. Birkl�e, Targeting O-Acetyl-GD2 ganglioside for cancer immu-notherapy, J. Immunol. Res. 2017 (2017). Article ID 5604891.

[76] C. Mandal, M. Chatterjee, D. Sinha, Investigation of 9-O-acetylated sialogly-coconjugates in childhood acute lymphoblastic leukaemia, Br. J. Haematol.110 (2000) 801e812.

[77] A.M.T. Baumann, M.J.G. Bakkers, F.F.R. Buettner, M. Hartmann, M. Grove,M.A. Langereis, R.J. de Groot, M. Mühlenhoff, 9-O-Acetylation of sialic acids iscatalysed by CASD1 via a covalent acetyl-enzyme intermediate, Nat. Commun.6 (2015) 7673.

[78] R. Schauer, H. Schmid, J. Pommerencke, M. Iwersen, G. Kohla, Metabolism androle of O-acetylated sialic acids, in: Albert M. Wu (Ed.), The MolecularImmunology of Complex Carbohydrates-2, Springer Science, New York, 2001,pp. 325e342.





































electrospray ionization ion mobility mass spectrometry ......82 m. sarbu et al. / biochimie 139...

Documents