Upload
hoangthuy
View
221
Download
0
Embed Size (px)
Citation preview
Development of Electrospray Mass Spectrometry-Based
Methods for Glycan Analysis in Biomedical Research
HABILITATION THESIS
Prof. Dr. Alina-Diana Zamfir
-2013-
2
Table of contents
Abstract
2
Rezumat
5
PART I. Principles and modern aspects of electrospray mass spectrometry: a brief overview 1.1.Electrospray ionization mass spectrometry
8
1.2. Microfluidics/electrospray ionization mass spectrometry 11 1.2.1. Capillary electrophoresis/electrospray ionization mass spectrometry
12
1.2.2. Chip-based electrospray ionization mass spectrometry 18 PART II. Microfluidics/electrospray ionization mass spectrometry: implementation of a novel concept in glycomics (Own reported results) 2.1. Interfacing microfluidic systems to the hybrid quadrupole time-of-flight (QTOF) mass spectrometer
22
2.1.1. Coupling of HPCE to the QTOF mass spectrometer 24 2.1.2. Coupling of the fully automated chip-based ionization to QTOF mass spectrometer
25
2.1.3. Interfacing the thin chip microsprayer system to QTOF MS
26
2.2. Interfacing microfluidic systems to Fourier transform ion cyclotron resonance (FTICR) mass spectrometer
35
2.2.1. Coupling of the fully automated chip-based ionization to FTICR mass spectrometer
38
2.2.2. Interfacing the thin chip microsprayer system to FTICR MS
40
2.3. Interfacing the fully automated chip-based ionization (NanoMate robot) to a high capacity ion trap (HCT) mass spectrometer
42
PART III. Applications of microfluidics/electrospray ionization mass spectrometry to structural analysis of glycoconjugates in biomedical research (Own reported results) 3.1. Introduction
45
3.2. Screening, sequencing and structural identification of O-glycopeptides from urine of patients suffering from Schindler disease
47
3.3. Screening, sequencing and structural identification of brain gangliosides
67
3.3.1. Analysis of gangliosides and glycolipids from normal tissues
69
3.3.2 Analysis of ganglioside expression and structure in pathological tissues
79
3.4. Structural analysis of chondroitin/dermatan sulfate glycosaminoglycan (GAG) oligosaccharides
105
3
PART IV Concluding remarks and perspectives 4.1. Concluding remarks
131
4.2. Plans for further research and career development 132 List of own publications 136 References 143
4
Abstract
In 2002 the Nobel Prize in Chemistry was awarded to three distinguished
scientists who developed analytical methods for biomolecule investigation.
Laureates were John B. Fenn (pioneer of electrospray ionization mass spectrometry)
and Koichi Tanaka (MALDI mass spectrometry) for “their development of soft
desorption ionization methods for mass spectrometric analyses of biological
macromolecules” and Kurt Wüthrich for “his development of nuclear magnetic resonance
spectroscopy for determining the three-dimensional structure of macromolecules in
solution”.
Biological macromolecules represent the basis of life whether expressed in
healthy, prosper diversity or in frightening diseases. To get an insight into biology
and medicine at their deep molecular level a continuous development of methods
for determination of bimolecule individual structure, functional characteristics and
interactions is required. As acknowledged by the Nobel Prize Committee, one of
the most efficient physicochemical techniques employed today for this purpose is
electrospray ionization mass spectrometry.
This habilitation thesis includes the original results related to the
development of novel electrospray mass spectrometry-based methods for the
analysis of glycans with potential biological and clinical relevance, extracted and
purified from biological (mostly human) matrices. These results were obtained
within 2001-2012 period of research carried out in three laboratories: Biomedical
Analysis Laboratory of the Institute for Medical Physics and Biophysics, University
of Münster, Germany (2001-2006), Mass Spectrometry Laboratory of the National
Institute for Research and Development in Electrochemistry and Condensed
Matter, Timisoara, Romania (2006-2012) and Laboratory for the Analysis and
Modelling of Biological Systems, “Aurel Vlaicu” University of Arad, Romania
(2006-2012).
The work carried out at the University of Münster within the postdoctoral
research, was included in the Habilitation Thesis entitled High Performance
Electrospray Mass Spectrometric Glycoscreening in Biomedicine, which I
5
defended at the University of Münster in December 2005 (Habilitation Diploma
with Venia Legendi in Biophysics in February 2006). Consequently, the present
habilitation thesis encompasses the results (2001-2006) already defended in
Germany, to which the results obtained in Romania, after my repatriation in
2006, were added.
The present thesis is structured in four main parts. Part I is dedicated to a brief
overview accompanied by a literature survey on the electrospray ionization process
and theories as well as the most modern aspects of method development. A
particular attention is paid to microfluidics/electrospray ionization mass
spectrometry methods: capillary electrophoresis in combination with electrospray
ionization mass spectrometry and chip-based electrospray ionization mass
spectrometry.
Part II includes original results related to the technical and methodological
developments and first implementation in glycomics of several
microfluidics/electrospray ionization mass spectrometry methods. This part
documents in details the original research conducted for: i) construction and
optimization of two different sheathless interfaces for capillary electrophoresis
coupling to mass spectrometry via electrospray ionization; ii) coupling of a fully
automated silicon chip-based nanoelectrospray system (NanoMate robot) to three
different mass spectrometers (quadrupole time-of-flight –QTOF, Fourier transform
ion cyclotron resonance-FTICR, and high capacity ion trap-HCT); iii) coupling of a
thin polymer microsprayer chip to two different mass spectrometers (QTOF and
FTICR MS); iv) optimization of all coupled systems for functioning on-line in MS
mode for screening, tandem MS (MS/MS) and multistage MS (MSn) for
fragmentation by collision-induced dissociation (CID) and in the case of HCT MS,
also by electron transfer dissociation (ETD) and alternate CID/ETD.
Part III highlights the experimental data related to the most relevant applications of
these analytical platforms, which were newly developed and introduced in
glycomics: 1. screening, sequencing and structural analysis of O-glycopeptides
expressed in the urine of patients suffering from Schindler’s disease vs. age-
matched healthy controls; 2. mapping, sequencing, structural analysis and
6
postulation of ganglioside/glycolipid biomarkers in healthy central nervous
system, neurodegenerative/neurodevelopmental diseases, primary and secondary
(metastases) brain tumors; 3. structural characterization of glycosaminoglycans
from extracellular matrix (ECM), in particular from human decorin, an ubiquitous
proteoglycan with key biological roles at the ECM level.
All these applications demonstrated not only the feasibility of the novel
microfluidics-MS methods but also their superiority in terms of analysis speed
(high throughput investigation, before inconceivable for glycoconjugates) and
sensitivity, data accuracy, reliability and experiment reproducibility.
The last part of the thesis presents concluding remarks, perspectives, ideas and
plans for further work in this exciting interdisciplinary research area, to confirm
Stanley Fields’ optimism: “Because the technology provides the tools and biology the
problems, the two should enjoy a happy marriage.”
7
Rezumat
In anul 2002 Premiul Nobel pentru Chimie a fost decernat unor cercetatori
remarcabili care au dezvoltat metode analitice pentru investigarea biomoleculelor.
Laureatii au fost John B. Fenn (pionier al ionizarii prin electrospray in
spectrometria de masa) si Koichi Tanaka (spectrometrie de masa cu ionizare prin
MALDI) pentru “dezvoltarile de metode de ionizare/desorptie pentru analizele prin
spectrometrie de masa a macromoleculelor biologice” si Kurt Wüthrich pentru
“dezvoltarea spectroscopiei de rezonanta magnetica nucleara in determinarea structurii
tridimensionale a macromoleculelor in solutie”.
Macromoleculele biologice reprezinta baza vietii fie ca sunt exprimate in
diversitatea viguroasa si prospera fie in boli inspaimantatoare. Pentru a patrunde
in biologie si medicina pana la nivelul molecular este necesara o continua
dezvoltare de metode pentru a determina structura individuala a biomoleculelor,
caracteristicile functionale si interactiunile lor. Asa cum a confirmat Comitetul
Nobel, una dintre cele mai eficiente tehnici fizico-chimice implicate in prezent in
acest scop este spectrometria de masa cu ionizare prin electrospray.
Prezenta teza de abilitare cuprinde rezultate originale privind dezvoltarea
de noi metode bazate pe spectrometria de masa cu ionizare prin electrospray
pentru analiza glicanilor (hidrati de carbon) cu posibila relevanta biologica si
clinica, extrasi si purificati din matrici biologice (in special de natura umana).
Aceste rezultate au fost obtinute in perioada 2001-2012, cercetarea desfasurandu-se
in trei laboratoare: Laboratorul de Analiza Biomedicala din cadrul Institutului de
Fizica Medicala si Biofizica, Universitatea din Münster, Germania (2001-2006),
Laboratorul de Spectrometrie de Masa, Institutul National de Cercetare-Dezvoltare
in Electrochimie si Materie Codensata, Timisoara, Romania (2006-2012) si
Laboratorul pentru Analiza si Modelarea Sistemelor Biologice, Universitatea
“Aurel Vlaicu” din Arad, Romania (2006-2012).
Cercetarea efectuata in cadrul stagiului postdoctoral de la Universitatea
din Münster, Germania, a fost inclusa in teza de abilitare intitulata High
Performance Electrospray Mass Spectrometric Glycoscreening in Biomedicine pe
8
care am sustinut-o la Universitatea din Münster, Germania, in decembrie 2005,
obtinand Diploma de Abilitare cu Venia Legendi in Biofizica in februarie 2006. In
consecinta, prezenta teza de abilitare include rezultate deja sustinute in acelasi
scop in Germania, carora li s-au adaugat rezultatele cercetarilor efectuate in
Romania de la momentul repatrierii mele (2006) si pana in prezent.
Lucrarea de fata este structurata in patru parti principale. Partea I este
dedicata unei scurte treceri in revista a procesului si teoriilor ionizarii prin
electrospray precum si a celor mai moderne aspecte ale dezvoltarii metodei. O
atentie speciala este acordata sistemelor microfluidice in cuplaj cu spectrometria de
masa cu ionizare prin electrospray: electroforeza capilara in combinatie cu
spectrometria de masa cu ionizare prin electrospray si spectrometria de masa cu
ionizare prin electrospray bazata pe chip.
Partea a II-a include rezultate originale legate de dezvoltarile tehnice si
metodologice precum si de prima implementare in glicomica a metodelor
microfluidice cuplate cu spectrometria de masa cu ionizare prin electrospray.
Aceasta parte descrie in detaliu cercetarile proprii cu privire la: i) constructia si
optimizarea a doua interfete diferite de tip sheathless, fara contact electric prin
solvent, pentru cuplajul electroforezei capilare cu spectrometria de masa cu
ionizare prin electrospray; ii) cuplajul unui sistem complet automat bazat pe chip
de siliciu-nanoelectrospray (robot NanoMate) cu trei spectrometre de masa diferite
(spectrometru cuadrupolar hibrid cu timp de zbor, QTOF, spectrometru cu
rezonanta ciclotronica si transformata Fourier, FTICR, si spectrometru de tip
capcana ionica de mare capacitate, HCT); iii) cuplajul unui chip micropulverizator
subtire din polimer cu doua spectrometre de masa diferite (QTOF si FTICR MS); iv)
optimizarea tuturor sistemelor cuplate pentru functionare on-line in regim MS
pentru screening, tandem MS (MS/MS) si multistagii MS (MSn) pentru
fragmentare prin disocieri induse prin ciocnire (collision induced dissociation,
CID), iar in cazul instrumentului HCT MS si prin disocieri prin transfer de electroni
(electron transfer dissociation, ETD) sau CID/ETD alternativ.
Partea a III-a ilustreaza datele experimentale cu privire la cele mai relevante
aplicatii ale acestor platforme analitice nou dezvoltate si introduse pentru prima
9
data in lume in glicomica: 1. screening-ul, fragmentarea si analiza structurala a O-
glicopeptidelor exprimate in urina unor pacienti suferind de boala lui Schindler in
comparatie cu subiecti sanatosi, avand aceeasi varsta, drept control; 2.
cartografierea, fragmentarea, analiza structurala si postularea biomarkerilor de tip
gangliozide/glicolipide in sistemul nervos central normal (sanatos), boli
neurodegenerative, tumori cerebrale primare si secundare (metastaze la creier); 3.
analiza structurala a glicozaminoglicanilor din matrici extracelulare (ECM) in
particular din decorin uman, un proteoglican raspandit in ECM, cu rol biologic
major la acest nivel.
Toate aceste aplicatii au demonstrat nu doar fezabilitatea noilor metode
microfluidice-MS ci in special superioritatea lor in ceea ce priveste viteza (in regim
throughput, care era anterior imposibil de imaginat pentru hidratii de carbon si
derivatii lor glicoconjugati) si sensibilitatea analizelor, precizia si consistenta
datelor precum si reproductibilitatea experimentelor.
Ultima parte a tezei prezinta concluziile, prespectivele temei de studiu, idei si
planuri pentru cercetari viitoare in acest interesant domeniu interdisciplinar,
pentru a confirma optimismul lui Stanley Fields: “Intrucat tehnologia ofera uneltele,
iar biologia problemele, cele doua ar trebui sa savureze un mariaj fericit.”
10
SECTION I
PART I
Principles and modern aspects of electrospray mass
spectrometry: a brief overview
1.1. Electrospray ionization mass spectrometry
„I gave electrospray wings to the molecular elephants“ John B. Fenn
Mass spectrometry is an instrumental approach that allows for the
determination of the molecular mass, therefore it is often called “the smallest scale
in the world”. The evolution of mass spectrometry has been marked by an ever-
increasing demand for its application to problems of major difficulty such as
biomolecule analysis, and the explosion of computer science. New developments in
the technology have created a complex and sophisticated array of instruments,
however the basic components of all mass spectrometers are the ion source, the
mass analyzer and the ion detector. The ion source ionizes the molecule of interest,
then the mass analyzer differentiates the ions according to their mass-to-charge
ratio (m/z) and finally, a detector measures the current of the ionic beam. Each of
these elements exists in many forms and is combined to produce a wide variety of
mass spectrometers with specialized characteristics.
Among all ionization techniques, electrospray (ESI) is one of the most
fascinating as it is able to generate, at atmospheric pressure, ions directly from
solution, which makes it applicable to a large class of non-volatile substrates.
Initial experiments carried out by the physicist John Zeleny in 1917 [1]
preceded the first description by Malcolm Dole and collab. in 1968 of the
electrospray principle, including the charge residue model (CRM) which has survived
as a main explanation for the controversial ESI process [2].
11
However, the well-defined breakthrough of ESI as a general ionization
method came in 1988 when John B. Fenn presented his experiments on
identification of polypeptides and proteins of 40 kDa molecular weight. Fenn
showed that a molecular-weight accuracy of 0.01% could be obtained by applying a
signal-averaging method to the multiple ions formed in the ESI process. The
findings were based on experiments started in 1984 [3] in Fenn’s laboratory at Yale,
when electrospray and mass spectrometry were successfully combined for the first
time. Fenn used his knowledge of free-jet expansion to improve Dole’s method
with a counterflow of gas for desolvation, eliminating re-solvation of formed
macromolecular ions. This discovery was closely followed by results from a
Russian research group (Aleksandrov et al.) [4].
In ESI, basically, the liquid containing the analyte of interest is pumped
through a metal capillary, which has an open end with a sharply pointed tip (Fig.
1.1.1).
Figure 1.1.1. Electrospray ionization process (Courtesy of New Objectives Inc.)
The tip is attached to a voltage supply and its end faces a counter-electrode plate.
As the voltage is increased, the liquid becomes charged and due to charge-
repulsion effect, it expands out of the capillary tip forming the so-called Taylor
cone. Since all droplets contain the same electrical charge, at the very end of the
cone, they emerge into a fine spray called ESI plume [5]. Depending on the polarity
of the applied electric field, the charges may be positive or negative. The droplets
12
are usually less than 10 micrometers across and contain both solvent and analyte
molecules. The charged droplets move across the electric field existing between
capillary and counter-electrode and, under a curtain gas flow, the solvent
molecules evaporate from the droplet. According to Dole’s CRM, as the droplet size
decreases while the total charge on the droplet is constant, the charge surface
density increases until the droplet’s surface tension is exceeded by the repulsive
electric forces. At this critical point, the droplet explodes into smaller, still highly
charged droplets. This process, called Rayleigh explosion, repeats itself until the
analyte molecule is stripped of all solvent molecules, and is left as a multiply
charged ion (Fig. 1.1.2.).
The number of charges retained by an analyte depends on such factors as the
composition and pH of the electrosprayed solvent as well as the chemical nature of
the sample. For small molecules (< 2000 Daltons) ESI typically generates singly,
doubly or triply charged ions, while for large molecules (> 2000 Daltons) the ESI
process typically generates a series of multiply-charged species and the resultant
ESI mass spectrum contains multiple peaks corresponding to the different charge
states (Fig.1.1.2).
Figure 1.1.2. Physicochemical processes of electrospray ionization [5]
This feature brings complexity to the interpretation of the ESI mass spectra but
concomitantly, as a first advantage, it adds to the information and can be used to
Capillary
Plume
Multiply charged ions
ESI spectrum
Counterelectrode
Electrospray
3000 Volt
13
improve the accuracy of the molecular-weight determination. The method of
deducing this way the molecular weight was described in the multiple charge theory
described by Fenn [6]. The theory showed that different charge states could be
interpreted as independent measurements of molecular weight and that an
averaging method based on the solution of simultaneous equations could provide
accurate molecular weight estimations for large molecules. The complex charge
pattern can simply be deconvoluted and the mass of the uncharged protein is
determined to dramatically higher accuracy than if the interpretation of data was
based on a single ion. The second advantage of multiple charging is the formation
of ions with reduced m/z ratio measurable with good resolution by almost any type
of analyzer with which ESI has been interfaced: magnetic sector, single or triple
quadrupole, time-of-flight, quadrupole ion trap, Fourier-transform ion cyclotron
resonance or hybrid quadrupole time-of-flight analyzer. All these make ESI the
method of choice for large biopolymers and molecular aggregates or complexes
that only have weak non-covalent interactions, such as protein-protein, enzyme-
substrate or protein-ligand complexes.
1.2. Microfluidics/electrospray ionization mass spectrometry
At the beginning of the 90’ the continuous refinement of the electrospray as
ion source in MS culminated with the low-flow (micro- and nanoESI)
configurations, which provide sensitivities at sub-picomolar level [7-8]. However,
in bioanalysis the major challenge for nanoESI MS was the high heterogeneity of
the biological samples in terms of number of components and the diversity of their
structure. Therefore, the combination of powerful liquid separation techniques
with high sensitive MS detection modes started to attract a great interest due to the
foreseen possibility to separate and directly identify the molecules in an on-line MS
experiment [9-11]. The advantage of ESI MS to form ions directly from solution has
established the technique as a convenient mass detector for high performance
liquid chromatography (HPLC) or capillary electrophoresis (HPCE) [OP1, 12-14].
While past attempts to couple LC or CE with mass spectrometry resulted in limited
14
success, ESI has made on-line HPLC- and CE-MS possible, adding a new
dimension to the capabilities of these techniques for biomolecule characterization.
In particular, the high separation efficiency, sensitivity and selectivity offered by
the capillary electrophoresis made the CE ESI MS coupling a method of choice in
complex mixture analysis [OP1].
In the post-genome era, MS develops continuously as one of the most
powerful analytical technique for structural elucidation of molecules originating
from biological matrices. As shown before, potentials of MS for high-sensitive
structural bioanalyses increased significantly after the introduction of ESI and
MALDI methods from one side and the possibility to sequence complex ionic
species by highly efficient dissociation techniques based on multiple stage MS
(MSn) on the other. In particular, in proteomics, glycomics and glycoproteomics,
nanoESI MSn in the positive as well as in the negative ion mode was shown to be
capable in sequencing minute amounts of biological material thus providing
straightforward information on various structural elements [15-18].
On the other hand, miniaturized analytical instrumentation is attracting
growing interest in chemical, biochemical and structural analysis. Nowadays,
massive effort is invested in MS interfacing to microfluidic-based systems as front-
end technologies for ESI [OP2, 19-21]. The generic term microfluidics refers to all
analytical tools where fluids can be driven in microstructured channels and/or
narrow capillaries. Thus, in terms of MS interfacing, microfluidics are: 1) stand-alone
specialized devices/instruments like capillary electrophoresis, micro LC etc; 2)
integrated mono-or polyfunctional micro-/nanosystems such as silicon, glass, or
polymer chips; 3) complex devices combination of automated sample delivery and
chip-based ionization such as automated chip-based robots for sample MS infusion
by ESI.
1.2.1. Capillary electrophoresis/electrospray ionization mass spectrometry
Capillary electrophoresis (CE) is an instrumental evolution of traditional
slab gel electrophoretic techniques and is based on differences in solute velocity in
15
an electric field [22-24]. In capillary electrophoresis the electromigration of the
analytes is taking place in narrow-bore capillaries, which includes CE in the
category of microfluidic devices. The narrow capillaries allow the application of a
high electric field up to 0.6 kV/cm thus enhancing a very high efficiency of the
separation.
In Fig.1.2.1. the scheme of a capillary electrophoresis setup with UV
detection is depicted. The assembly consists of a fused silica separation capillary of
20-100 m i.d. 20-100 cm length, two buffer vials A and B and another one
containing the solution of analyte, a high voltage power supply (delivering up to
30-40 kV and 200-250 A) and a detector which can be of various types: UV
detector, electrochemical detector, laser induced fluorescence (LIF) or MS.
The two ends of the capillary are immersed in the vials containing
electrolyte together with two electrodes. A high voltage source delivers the
potential difference between the electrodes necessary for the separation. The UV
detector is placed at 5-10 cm distance from the outlet capillary.
Figure 1.2.1. Basic CE setup [OP3]
The sample, usually dissolved in electrolyte is injected into the capillary by
either 1) hydrodynamic injection using pressure or vacuum application (ex. 0.5 psi)
while the injection end of the capillary is inserted in the vial containing the analyte
Voltage power supply
A B
Electrolyte vials
UV Detector
autosampler
PC
Electrodes
Externally polyimide coated CE capillary
16
solution or 2) electrokinetic injection induced by application of potential difference.
The dispersion processes limits the amount injected in the capillary. A practical
limit of injection plug length is less than 1-2% of the total capillary length meaning
nl or pl volumes.
Different types of capillaries are used for the CE separation: glass, teflon, polymer
capillaries etc. However, fused silica capillaries are the best option, as they meet all
requirements claimed by the CE technique: chemically and physically resistant,
transparent to UV radiation, able to dissipate Joule heat, narrow internal diameters
[OP1, OP3]. The silica capillaries are externally “coated” with a layer of polyimide
and internally may be uncoated, or coated with polymers suppressing the
adsorption of the analyte to the capillary walls.
CE separation is usually carried out at constant potential in direct polarity
(injection at anode and detection at cathode) or reverse polarity (injection at
cathode and detection at anode). However, constant current mode can also be
performed, especially when the temperature control of the capillary is not efficient.
Not commonly used, gradients or steps in the voltage may be useful in
simultaneous analysis of compounds having very different electrophoretic mobility
[25, 26].
In principle, CE separates the species according to their migration velocity
under the influence of a high electric field [25, 26]. The difference in solute velocity
is given by the different electrophoretic mobility as expressed by the eq. (1.2.1):
v=eE (1.2.1)
where E is the electric field vector and e-the electrophoretic mobility, constant for
a certain ion and medium and determined by the electric force balanced by the
frictional one (eq 1.2.2 and 1.2.3) :
e=q/6r (1.2.2)
a=e+eof (1.2.3)
where: q-ion charge, -viscosity, r-ion radius, a-apparent electrophoretic mobility
and eof-mobility of the electroosmotic flow.
The above relations show that the electrophoretic mobility of an analyte is
depending on several factors: charge of the analyte, pH of the solution, viscosity,
17
temperature inside the capillary, m/z ratio, the applied electric field, dimensions of
the capillary. Therefore, the choice of the electrolyte is a key step in performing an
efficient separation, imposing the optimization of all its parameters: pH, ionic
strength, chemical composition, and concentration.
An essential parameter of the CE separation is the electroosmotic flow (EOF), a
consequence of the surface charge on the inner capillary wall. Under certain
solution conditions (pH > 3.0) fused silica surface possess an excess of negative
charges due to the ionization of silanol group. The counterions, which balance the
surface charge, forming the diffuse double layer at the capillary wall, create a
potential difference [OP3]. When applying a voltage across the capillary, the
positive ions of the electrical double layer are attracted toward the cathode. Due to
solvation, the ionic movement drags the bulk flow solution creating the
electroosmotic flow under the electric field. Feof, the force of the EOF, is one order of
magnitude higher than Fe the electrophoretic force; therefore EOF causes the
movement of all species regardless the charge in the same direction (Fig.1.2.2).
Figure 1.2.2. Capillary electrophoresis mechanism [OP3]
The separation occurs under the action of different Fe [OP3] so that if the mixture
contains positive and negative ions as well as neutral species, (Fig.1.2.2) in forward
or direct polarity the first will elute the positive ions, followed by the molecules
which did not undergo ionization. The negative ions, dragged by Feof will also
migrate toward the cathode but will elute later. In reverse polarity, the EOF is
oriented against the desired direction of ion motion. It may be suppressed by
18
careful re-consideration of solution parameters (pH < 3.0). In this case the Fe
becomes concomitantly the drift and separation force giving rise to high separation
efficiency and resolution.
On-line CE ESI MS coupling
The utility of CE towards biomolecule screening can be greatly enhanced by
mass spectrometric (MS) detection and identification. By combined CE MS
technique, not only molecular masses of the separated components can be directly
measured with good accuracy but also specific fragment ions may be generated by
tandem MS (MS/MS) from individual components in a complex mixture, to deduce
the molecular structure.
Direct coupling of CE to MS has been introduced for over 10 years and
nowadays the most attractive method for coupling capillary separation technique
to MS is via ESI interface [27-29]. The on-line CE ESI MS coupling requires
however, an interface fulfilling the requirements related to an efficient transfer of
the sample from the CE capillary into MS without affecting the CE separation
efficiency. Numerous parameters are influencing the CE ESI MS analysis and are to
be taken into account for optimization [OP4]: the choice of an electrolyte
compatible with both the ionic species formation/separation by CE and
electrospray process, the interface design and configuration, range of the applied
CE and ESI potentials, performance of the formed CE MS electric circuit, fine
positioning of the sprayer with respect to the MS sampling orifice and general
solution and instrumental parameters such as buffer and sample concentration, pH,
injection time and pressure, capillary temperature etc.
Figure 1.2.3. Sheath flow interface using coaxial make-up liquid.
Adapted from ref. [30]
Sheath-liquid HV
CE separation capillaryESI capillary
Metal coating
19
The most widely used interface for CE ESI MS setup is the sheath liquid
interface which was conceived in various configurations but basically in all of them
a sheath liquid, serves to ensure the electrical contact between the CE capillary and
the ESI sprayer. In principle, the make-up liquid picks up the analyte eluting from
the CE capillary in a solvent appropriate for ESI and the whole resulting mixture is
then sprayed into the mass spectrometer (Fig 1.2.3).
In all sheath flow interfaces, the make-up liquid mixes the CE buffer and the
sample, therefore, the major drawback of this configuration is the reduced
sensitivity consequence of sample dilution.
To overcome the disadvantages of the sheath flow interfaces, Olivares et al.
[31] proposed a novel CE ESI MS setup, which eliminated the addition of a make-
up liquid and used instead a stainless steel capillary sheath for electrical contact.
This type of configuration allows for a sheathless CE ESI MS design spraying the
sample directly from the CE microspray tip into the mass spectrometer. The effects
of the use of this kind of microsprayers include higher analyte concentration, lower
spraying potential, closer positioning of the sprayer to the orifice of the mass
spectrometer and significant improvement of ion transfer into MS [OP4]. Moreover,
ionization and desolvation of the generated droplets are also improved
contributing to the significant gain in sensitivity. Thus, the best alternative design
for the sheath flow configuration is the high sensitivity sheathless interface. The
most effective sheathless configuration is based on coating the sprayer tip with a
conductive layer to provide the electrical contact needed for both CE and ESI. Very
popular sheathless interfaces are those based on applying either noble metals [32,
33], or carbon [34] on the spraying tip (Fig.1.2.4) which may be either the CE
capillary itself, etched as a microsprayer, or a spraying needle butted to the CE
column.
Despite the great number of CE ESI interface types, so far, no general
consensus has been reached regarding the best design and no approach generally
valid for the analysis of complex biomolecule mixtures has been reported.
20
Figure 1.2.4. Sheathless CE ESI MS interface with metallic
deposition on the microsprayer. Adapted from ref. [35]
Most of the on-line CE ESI MS couplings are, however, constructed and
completely functional for peptide and protein characterization but only a few of
them have been introduced for complex carbohydrate system analysis [OP4].
1.2.2. Chip-based electrospray ionization mass spectrometry
The actual trends in analytical science are toward the high-throughput
measurements based on the nanotechnology achievements in automatization and
miniaturization of devices. In bioscience integrated, fully automated micro and
nanosystems functioning on the basis of the “lab-on-a-chip” principle have been
demonstrated to provide one of the most rapid, sensitive and accurate analysis [36,
37, OP2].
The potential of the modern chip-based ESI systems considerably broadened
the area of MS applicability in life sciences. The option for miniaturized, integrated
devices for sample infusion into MS is driven by several technical, analytical and
economical advantages such as [OP5]: 1) simplification of the laborious chemical
and biochemical strategies required currently for MS research; 2) high throughput
nanoanalysis/identification of biomolecules; 3) elimination of the time-consuming
optimization procedures; 4) increase of the sensitivity by drastically reduction of
the sample and reagent consumption, sample handling and potential sample loss;
5) high reproducibility of the experiments; 6) potential to discover novel
biologically-relevant structures due to increased ionization efficiency; 7) high
signal-to-noise ratio; 8) reduced in-source fragmentation; 9) flexibility and broad
area of applicability; 10) low cost of analysis and chip production; 11) possibility for
CE separation capillaryESI capillary
HV Conductive layer
21
unattended high-throughput experiments reducing the man power and man
intervention; 12) possibility to perform several stages of sample preparation in a
single integrated unit followed by direct MS structural analysis; 13) elimination of
possible cross-contamination and carry-overs; 14) flexibility for different
configurations, upgrading and modifications; 15) minimal infrastructure
requirements for optimal functioning; 16) reduction of the ion source size
facilitating manipulation and efficient ion transfer by precise positioning towards
the MS sampling orifice.
Two types of chip-based devices are currently being investigated and used
for ESI MS studies. The first category is represented by the out-of plane devices,
where hundreds of nozzle-like nanospray emitters are integrated onto a single
silicon substrate from which electrospray is established perpendicular to the
substrate [37-38]. These devices are particularly well suited to high-throughput
sample delivery to ESI MS [OP2, OP6] and have the potential to completely replace
flow-injection analysis assays. Moreover, the technical quality of the nanosprayers
obtained by silicon microtechnology is so high, and the experiments so
reproducible, that such devices have been found in some instances to give more
robust and quantitative analyses than LC- or CE MS [OP2] and to be able to
suppress the need for LC or CE separation prior to MS analysis.
Due to very efficient ionization properties, silicon-based nanoESIchip-MS
preferentially forms multiply charged ions, and the in-source fragmentation of labile
groups attached to the main biomolecular framework is minimized. These features
further enhance the significantly increased ionization efficiency and sensitivity of
the analysis.
The second category consists of planar or thin microchips, made from glass
[39] or polymer [40, 41] material, embedding a microchannel at the end of which
electrospray is to be generated in-plane, on the edge of the microchip. In recent
years, the progress in polymer-based microsprayer systems was promoted by
development of simpler methods for accurate plastic replications and ease to create
lower-cost disposable chips. Though clearly amenable to automation for high-
throughput analysis, these designs are best suited to the integration of other
22
analytical functions prior to sample delivery to the MS system, such as sample
cleanup, analyte separation by CE, and chemical tagging [42-44]. Moreover, in
comparison with glass nanospray capillaries, these thin microsprayers were found
to provide superior stability of the spray with time, improved signal-to-noise ratios
at various flow rates and high flexibility and adaptability to different ion source
configurations, with the additional advantage of cheaper production costs
compared to silicon technologies.
Several approaches for interfacing MS to polymer chips providing flow rates
which include them in categories from nano- to microsprayers were reported [OP2,
45-46]. Different MS configurations such as single and triple quadrupole MS ion
trap and ultra high resolution Fourier-transform ion cyclotron resonance mass
spectrometry (FTICR MS) were adapted to polymer-based chip ESI and contributed
significant benefits for various studies. However, the general areas of
implementation and applications of chip technologies either silicon-based or
polymer microchips have by far been primarily genomics, drug discovery, and
proteomics. Despite the high potential and performance exhibited by chip based
MS methodologies, the glycomics field has not significantly benefited from the chip
technology. Optimization of ESI MS and chipESI MS for operation in the necessary
negative ion mode was considered a challenging task, and application to
carbohydrate analysis limited by their high structural diversity and
microheterogeneity [OP2]. Moreover, in addition to the low ionization efficiency
that ultimately leads to decreased sensitivity, each class of carbohydrates requires
particular and defined conditions to promote chip ionization and detection by MS.
These conditions are depending on the type of the labile attachments on the sugar
chain, the ionizability of the functional groups, the hydrophilic and/or
hydrophobic nature, branching of the sugar chains etc.
Modern achievements in robotics are currently also intensively introduced
in the MS field, tending to substitute the existing classical ESI which requires either
manual loading or pumping of the sample liquid through the electrospray
capillary. Automated, robotized and programmed systems for sample delivery into
23
MS are significantly increasing the analysis throughput and efficiency, allowing
minimization of sample volumes and handling.
Though of maximum efficiency and intensively used for proteomic analysis
and genomic surveys, the latest developments in robotized chip-based ESI MS have
been only to a limited extent introduced in glycomics. This can be rationalized also
by the special requirements, described above, for ionization/detection of long
saccharide chains and saccharides having labile attachments, which are more
difficult to be fulfilled by high throughput experiments. Due to the high
heterogeneity of the carbohydrate mixtures originating from biological sources,
automated chipMS analysis has often to be preceded by laborious chromatographic
or electrophoretic methods in conjunction with other biochemical tools.
Additionally, as mentioned before, for efficient ionization and/or fragmentation in
tandem MS, different instrumental conditions are required by each particular
carbohydrate category therefore the optimization stage is rather laborious. All these
attributes of carbohydrate mass spectrometry, made this biomolecule category less
amenable to chip-based and high-throughput methods [OP2]. However, in the
second part of this thesis it will be demonstrated that adequate chip MS and
automated chip ESI tandem MS strategies may lead to successful implementation
of this technology in glycomics, a step that had been done within this work.
24
PART II
Microfluidics/electrospray ionization mass spectrometry:
implementation of a novel concept in glycomics
(Own reported results)
2.1. Interfacing microfluidic systems to the hybrid quadrupole time-of-flight
(QTOF) mass spectrometer
High performance capillary electrophoresis, and two different chip ESI
systems: a fully automated chip-based nanoESI robot and a thin chip microsprayer
have been coupled each to a hybrid quadrupole time-of-flight (QTOF) mass
spectrometer. The technical feasibility of microfluidics/ESI MS interfacing and its
usefulness for structural elucidation of glycoconjugates originating from human
body fluids and tissues showed that this novel technique represents another key
milestone in the continuous progress of glycomics.
QTOF mass spectrometer
The QTOF mass spectrometer used in all these studies belongs to the first
generation of hybrid quadrupole time-of-flight instruments available with ESI and
nanoESI in Z-spray geometry, designed and produced by Micromass (Manchester,
UK). The QTOF is a hybrid MS/MS system combining the simplicity of a
quadrupole with the ultrahigh efficiency of an orthogonal acceleration TOF mass
analyzer. It exploits the TOF to achieve simultaneous detection of ions across the
full mass range, in contrast to triple quadrupole instruments that must scan over
one mass at a time. The resolving power of more than 5000 (FWHM) and inherent
stability of the reflectron TOF system combine to deliver high mass measurement
accuracy enabling compounds of similar nominal mass to be differentiated.
The QTOF incorporates a high performance quadrupole analyzer equipped
with a prefilter assembly to protect the main analyzer (time-of-flight) from
contaminating deposits, and an orthogonal acceleration time-of-flight (TOF) mass
25
spectrometer. A hexapole collision cell between the two MS analyzers is used to
induce fragmentation by collision-induced dissociation (CID) to assist the
structural investigations. Ions emerging from the second mass analyzer are
detected by a microchannel plate (MCP) detector and ion counting system. A post
acceleration photomultiplier detector situated after the orthogonal acceleration cell,
is used to detect the beam passing through the first stage of the instrument for
tuning and optimization.
In the QTOF MS, the ionic pathway is the following: the Z-spray source
leads the ions/ionized droplets perpendicular to the spray direction through the
counerelectrode (sampling cone) lets them pass the heater and further, via
turbulent motion into the RF-only hexapole ion guide. The advantage of this set-up
is that non-ionized species do not enter the analyzer region. In the MS mode the
quadrupole acts as a wide band-pass filter to transmit a wide mass range (RF-only
mode) through the hexapole collision cell into the pusher region of the TOF
analyzer, whereas in the MS/MS mode the quadrupole operates in the normal
resolving mode and is used to select precursor ions. The precursor ions are then
accelerated into the RF-only collision cell, where CID occurs using Ar as collision
gas. The ionic beam of the fragment ions is then focused into pusher by
acceleration, focus steer and tube lenses. The pusher is pulsing a section of the
beam towards a reflectron, which reflects the ions back to the detector. As the ions
are traveling from the pusher to the detector they are separated in mass according
to their flight times, with ions of the highest m/z ratio arriving latter. The pusher
may operate at repetition frequencies of up to 20 kHz resulting in a full spectrum
being recorded by the detector every 50 s. Each spectrum is summed in the
histogram memory of a time to digital converter (TDC) until the histogrammed
spectrum is transferred into a host PC running the MassLynx NT software system
which controls the instrument, acquires and processes the data.
Unlike scanning instruments, the TOF performs parallel detection of all
masses within a spectrum at high sensitivity, resolution and acquisition rates. The
latter characteristic is of very high importance when the MS is functioning coupled
to high performance CE (HPCE) or other fast chromatographic separation
26
techniques since each spectrum is representative of the sample composition at that
point of time, irrespective of how rapidly the sample composition is changing.
2.1.1. Coupling of HPCE to the QTOF mass spectrometer
The most accessible approach for ESI MS characterization of CE separated
components is the off-line collection of fractions. Fraction collection can be
performed using different techniques such calculating the time window when a
compound has migrated to the end of the capillary or using a prerun to obtain a CE
profile and estimate the migration velocities [OP1, OP7]. Unlike on-line coupling,
off-line method provides higher flexibility toward system optimization since the
CE instrument and the mass spectrometer can be adjusted independently and
optimized separately. Additionally, post-separation treatment of the samples,
prior to MS analysis, like concentration by solvent evaporation, modification of
buffer composition, dialysis, centrifugation etc. are possible [OP7]. However, using
the CE instrument as a fraction collector lowers the eluted sample concentrations
by the method principle itself, which imposes the collection of a few nanoliters into
tens of microliter volumes of electrolyte. Lack of sensitivity is therefore the specific
drawback of this approach often limiting the extension of its applicability toward
minute amount of samples deriving from biological matrices. Anyway, off-line CE
ESI MS method may be successfully used as a prerequisite for the direct on-line
coupling [OP7].
Due to its sensitivity and high reproducibility, on-line CE ESI MS coupling
gained in popularity as the most convenient approach to interfacing CE and MS
instruments [OP4]. It allows minimum sample handling, which significantly
reduces potential sample loss and uses the mass spectrometer as an extremely
selective and sensitive detector acquiring either the total (TIC) or extracted (XIC)
ion chromatograms. However, a number of already widely known problems are
inherently associated even with this method [OP7]. From the separation point of
view, a fundamental and general concern is that the best suited CE buffers are
usually incompatible with the electrospray process being non-volatile and causing
27
unstable behavior with the ESI source. Therefore, the compatibility of the CE
electrolyte as a spraying agent is one of the major difficulties encountered when
interfacing either off-line or on-line the CE to ESI MS. From this perspective,
carbohydrates either oligosaccharides or glycoconjugates provide even more
limited number of options because of the restrictive required conditions for ion
formation, separation and detection [OP1, OP8]. The necessity of high sensitivity,
high-speed MS acquisition, especially for on-line tandem MS, and optimization of
the coupling for detecting carbohydrate ions in the negative mode [OP9] represent
the second class of requirements for a successful and efficient on-line CE MS
glycoscreening.
As shown in Part I, sheathless design provides the best sensitivity, therefore
for the purpose of this work which is the structural identification of quantity-
limited glycoconjugates extracted from human tissues and body fluids, the
sheathless configuration was practically the only choice. This type of design allows
for two variants and both of them have been optimized, tested and implemented
for the first time in carbohydrate research in this study.
The first configuration was based on a one-piece CE column with its
terminus in-house modeled as a microsprayer [OP8]. The emitter at the CE column
terminus is a critical part because it has to have simultaneously microsprayer
properties and to act as an electrode as well. This means that it should have a
smooth and well-defined tip, which also maintains the electrical contact across the
CE buffer and produces a fine spray under mild electrospray voltage conditions. In
addition, it should be resistant to manipulations, durable, and suitable to certain
type of analysis and not to introduce dead volumes because this dramatically
deteriorates the CE separation efficiency.
The second interface, of improved sensitivity and ionization efficiency,
consists of two pieces, the CE column being butted to a nanosprayer needle
[OP10]. In this case special attention had to be paid to the butted region and
system assembling into the in-house made holder, a device responsible also for
delivering the high-voltage.
28
Sheathless on-line CE/microESI QTOF MS interface
In this configuration the ESI emitters were manufactured at the CE capillary
terminus so that, the interface consists of single separation column [OP8]. CE
capillaries were sharpened at one end by locally heating in a flame of
corresponding melting temperature. Meanwhile, the capillary tubing was manually
pulled apart. In order to obtain a very well defined tip orifice, the tiny long wire
obtained by pulling under flame has been removed with a tapered ceramic cutter
under visual inspection with a stereomicroscope. After this operation, the
measured outer diameter of the tip was around 100 m. The orifice size has been
calculated from d2 =d1d02/d01 (where d01 and d1 are the initial and final outer
diameters, respectively and d02 the initial inner diameter) where it was assumed
that by pulling under flame the inner and outer diameters decreased
proportionally. Thus, from a tip of 90 m o.d. an orifice diameter of about 15 m
resulted. Although a capillary tip with even smaller inner diameter would give a
better sensitivity, practical aspects such the ease of production, use and
manipulation as well as the possibility of capillary blockages determined the choice
of the tip diameter.
To further reduce the outer diameter of the tapered tips and to smooth the
edge, the tips were etched in hydrofluoric acid (HF) a procedure which turned out
to be a crucial step in manufacturing the CE microsprayer tips. Measurements
performed using a tapered capillary tip without HF etching showed that the
obtained sharp tip did not generate a stable spray. After etching, however, a good
quality steady spray under mild ESI voltages was achieved at the low
electroosmotic flows of about 50 nl/min.
The next stage of CE ESI tip producing was the metallic deposition.
Although in recent years several metallization techniques have been proposed, lack
of metallic deposition durability was one of the major drawback in almost all of
designs. In this study all CE tips were coated with copper. The coating process was
performed by smearing the surface of the tip with a thin liquid layer of copper
suspension in dimethylether. Copper was deposited as a thick layer on a length of
29
5-6 cm from the capillary tapered end and as a very thin layer in the near vicinity of
the tip.
The stability of the electrical contact at CE terminus is another crucial factor
for successful on–line CE ESI MS analysis. The stability of the CE current was
tested in the running buffers and the CE current, which indicates the electrical
performances of the circuit, was found stable in both positive and negative ion
mode for CE voltages of 15-30 kV.
Several advantages of this design for sheathless CE ESI MS applications in
glycoscreening are to be mentioned. First, a one-piece separation column provides
a better separation and a higher reproducibility since, in comparison with the
butted one, there is no the misalignment danger which could require, during the
same experiment, several interventions for finding appropriate connection points
between the tip and original column. Next, copper deposition proved to be very
resistant and stable in inevitable contact with aggressive media such as highly
alkaline buffers used for sugar CE separation in direct polarity. Finally, although
the copper coat is not everlasting, the electrical contact could be maintained under
ESI voltage around 30 hours without need for supplementary Cu deposition. Once
the electrical contact degraded, the copper coat from the surface of the tip could be
several times readjusted retaining the original emitter.
Figure 2.1.1. Electrical scheme of the sheathless on-line CE nanoESI QTOF MS [OP8]
CE
separation
voltage
UV
detector
QTOF
MS
CE capillary inletCu layer
Microsprayer tip
Counterelectrode (cone)
Clenching device
…
30
In order to realize the electrical contact to the ESI power supply, the shaped CE
column was introduced into a metallic body, home made stainless steel device for
clenching the capillary. The whole system was screwed onto the ESI high voltage
plate of the QTOF MS (Fig. 2.1.1). Exchange between the conventional source and
home-built CE MS interface did not claim for any definitive dismantling or special
mechanical modifications [OP8].
The tapered tip extended 1-2 cm over the holder and the position of CE emitter was
adjusted in the vicinity of the entrance hole of the counterelectrode sampling
cone, at a distance less than 5 mm, by the source assembly which can be
manipulated in x, y and z direction via micrometer screws, although fine
positioning of the microsprayer tip turned out not to be a critical parameter.
Sheathless on-line CE nanoESI QTOF MS interface
The highest sensitivity achieved by using the interface described above was
20 pmol/l. The low amount of glycans resulting after the usual sample
extraction/preparation protocols, the mild ESI source parameters and the high
ionization efficiency necessary for proper ionization of some carbohydrate
categories such as glycosaminoglycans, required the development of a novel
nanosprayer configuration to address all these issues. In this new setup [OP10] the
CE column was butted to a commercial (New Objectives Company) nanosprayer
needle by using a home-made joint. The resulting two-pieces-column is
incorporated into the stainless steel clenching device, designed and constructed to
allow the application of the ESI voltage onto the needle. The whole set up was
mounted by a set of screws directly onto the ESI high voltage QTOF plate. The
electrical scheme of the CE nanoESI QTOF MS setup is depicted in Fig. 2.1.2.
The spray could be initiated at values of 6-900 V applied to the nanoESI
needle and 12-30 V for the sampling cone potential without the need of nebulizer
gas. Leakages in the butted region or broadening of the total ion current (TIC)
peaks were not observed proving that no misalignment of the connection or gap
between the CE column and the nanosprayer needle were created.
31
Figure 2.1.2. Electrical scheme of the sheathless on-line CE nanoESI QTOF MS [OP10]
The sheathless CE nanosprayer system was optimized for glycoscreening in both
forward and reverse polarity and for two different buffer systems ammonium
acetate/ammonia at high pH values (strongly alkaline for forward polarity) and
formic acid/ammonia low pH values (strongly acidic for reverse polarity).
On-line CE ESI QTOF MS/MS sequencing by automatic switching from MS to
MS/MS (data-dependent analysis)
The QTOF mass spectrometer is uniquely suited for data-dependent MS to
MS/MS switching in view of its acquisition speed. It allows, contrary to scanning
type instruments (e.g. ion trap or triple quadrupole instruments), to switch to
MS/MS for a very short time still providing the production of high quality and
representative spectra.
Precursor ion selection for “classical” MS/MS is generally a pre-experiment
and operator intervention which in essence, would make MS/MS on a QTOF
instrument incompatible with an investigation of completely unknown molecules.
The automatic on-the-fly MS to MS/MS switching abilities of the QTOF offer the
possibility to automatically change over from MS to MS/MS and back again
according to the set parameters, without the need of a prerun to estimate the
ESI voltage
Fused silica externally
polyimide coated
CE capillary
375µm o.d.
(-)QTOF-MS Cone
voltage
Reverse polarity
(+)
Forward
polarity
Home- made
stainless steel clenching device
Home- made joint
Nanosprayer needle
350µm o.d
CE
CE electrode
32
sample composition. The “artificial intelligence” invested in this set of parameters
is thus able to move the precursor selection from a required on-line CE MS pre-
analysis to define precursors of interest at each retention time, to an instantaneous
ad-hoc process called upon every individual time slice.
For on-the-fly MS to MS/MS switching in the either use of direct infusion
[OP11] or CE MS [OP12] for systematic and comprehensive glycoprofiling,
significant parameters concerning MS to MS/MS switching had to be examined
and optimized. These included the spectral accumulation time and threshold in
both the MS and MS/MS mode, the collision energy values, the implementation of
the “included masses-only” option available on the QTOF MS for autoMS/MS of
well- and exclusively-predefined ions, as well as the elapsed acquisition time,
which defines the automatic switch back from MS/MS to MS.
During the on-line CE MS run, the quadrupole is initially set to transmit all
masses until an ion reaches a certain set threshold or the ion preset by the operator
eluted and has been detected. Thereupon, the quadrupole automatically switches
to the MS/MS mode, selecting the ion, which is subsequently fragmented in the
high-efficiency hexapole collision cell, thus generating product ions that are
further mass analyzed by the TOF. By limiting the TOF spectral accumulation time
in the MS/MS mode to a statistically acceptable minimum, the quadrupole almost
instantly switches back to the MS mode. Qualitative information, comprising the
complementary MS and MS/MS data (informative product ion profile), as well as
quantitative information obtained by integration of the MS extracted ion
chromatograms, could be obtained in one single acquisition experiment.
2.1.2. Coupling of the fully automated chip-based ionization to QTOF mass
spectrometer
Fully automated chip-based nanoelectrospray, NanoMateTM 100
incorporating ESI Chip technology (Advion BioSciences, Ithaca, USA), was coupled
to the QTOF MS and for the first time optimized for glycomics surveys [OP13]. The
robot and the incorporated silicon chip are presented in Fig.2.1.3.
33
NanoMate™ 100, the world's first fully automated nanoelectrospray system,
is a robotic device that provides an automated nanoelectrospray ion source for
mass spectrometers. The system is capable of infusing samples at low flow rates
(50-100nl/min) in an automated fashion. In addition to the robot itself, the key
component of the system is the ESI Chip. The robot holds a 96-well sample plate
and a 96-pipette tip tray. Automated sample analysis is achieved by loading a
disposable, conductive pipette tip on a moveable sampling probe, aspirating
sample via a syringe pump, and moving the sampling probe to engage against the
back of the ESI Chip.
Nanoelectrospray is initiated by applying a head pressure and voltage to the
sample in the pipette tip. Each nozzle and tip is used only once in order to
eliminate carryover and contamination typical to conventional autosamplers. The
ESI Chip is an array of nanoelectrospray nozzles of 10 µm internal diameter etched
in a planar silicon chip. The chip is fabricated from a monolithic silicon substrate
using deep reactive ion etching (DRIE) and other standard microfabrication
techniques.
Figure 2.1.3. NanoMateTM robot (courtesy of Advion BioSciences)
The inert coating on the surface allows a variety of acidic and organic
compositions and concentrations to be used to promote ionization without
degradation of the nozzle. As visible in Fig. 2.1.3., a channel extends from the
nozzle through the microchip. A unique feature of the ESI Chip is the incorporation
of the ESI ground potential into the spray nozzle.
Chip nozzle
Robot
34
Conventional electrospray devices define the electric field by the potential
difference between the spray device (fluid potential) and the mass spectrometer
inlet. In the ESI Chip, the electric field around the nozzle tip is formed from the
potential difference between the conductive silicon substrate and the voltage
applied to the fluid via the conductive pipette tip. The distance is only a few
microns, so the field is dense, and the distance is not variable. The distance that
defines the electric field is about 1000 times smaller than the distance of the nozzle
to the mass spectrometer. Therefore, the mass spectrometer position and voltage
though crucial for efficient ion transfer into analyzer, do not play any role in
forming the chip electrospray, thus essentially decoupling the ESI process from the
inlet of the mass spectrometer.
Figure 2.1.4. Coupling of the NanoMate robot to the QTOF mass spectrometer
The robot was mounted to the QTOF mass spectrometer (Fig. 2.1.4.) via a
special bracket allowing rough adjustment of the robot position with respect to the
sampling cone [OP13]. The fine positioning was however driven by the ChipSoft
software under WindowsXP which controls the robot. The position of the
electrospray chip was adjusted with respect to the sampling cone potential to give
raise to an optimal transfer of the ionic species into the mass spectrometer. In order
to prevent any contamination, for some experiments glass coated microtiter plates
were used. 2-5 µl aliquots of the working sample solutions were loaded onto the 96-
NanoMate
QTOF
35
well plate. The robot was in most cases programmed to aspirate the whole volume
of sample, followed by 2 µl of air into the pipette tip and then deliver the sample to
the inlet side of the microchip. Electrospray was initiated by applying voltages
within 1.45 kV to 1.9 kV and a head pressure of 0.3 to 1.3 p.s.i. The value of these
parameters turned out to be critical for generating and maintaining a long-lasting,
stable spray of carbohydrate solutions, regardless the solvent. Following sample
infusion and MS analysis, the pipette tip was ejected and a new tip and nozzle were
used for each sample, thus preventing any cross-contamination or carry-over. In
these studies, the whole coupled assembly was for the first time optimized to
function in the negative ion mode which is the ionization mode the best suited for
the investigation of complex O-glycan systems. By optimizing the NanoMate-ESI-
QTOF MS assembly for sugar ionization and sequence requirements a solid
methodology with major advantages in comparison with the capillary-based
nanoESI was developed.
The analytical applications demonstrated the feasibility of the fully-
automated chip-ESI-QTOF for high performance glycoscreening, sequencing and
identification. The NanoMate-chipESI-QTOF-tandem MS approach, introduced
here for the first time in glycomics, has shown its potential to discover novel
carbohydrate variants of potential diagnostic value in complex biological mixtures,
due to increased sensitivity, reproducibility and ionization efficiency and the
ability to generate a sustained and constant electrospray.
2.1.3. Interfacing the thin chip microsprayer system to QTOF MS
QTOF mass spectrometer was for the first time interfaced [OP5] to a
disposable polymer microchip with integrated microchannels and electrodes
conceived by DiagnoSwiss (Lausanne, Switzerland). The chip was microfabricated
by semiconductor techniques including photolitography. Basically, a photoresist is
patterned on a 75 m thick, copper-coated polyimide foil through a printed slide
acting as a mask. The photoresist is developed, and chemical etching is afterwards
used to remove the deprotected copper where microchannels are to be patterned.
36
The polyimide layer is plasma-etched to the desired depth. The final microchannels
are 120 m wide, 45 m deep (nearly “half moon” cross section), with gold-coated
microelectrodes placed at the bottom of the microchannel. A 35 m
polyethylene/polyethylene terephthalate is laminated to close the channels.
For MS coupling one end of each channel was manually cut in a tip shape,
which was visually inspected with a stereomicroscope. This way, the outlet of the
microchannel is located on the edge of the chip, providing a in-plane electrospray.
For sample dispensing, a reservoir was pasted over the inlet of the microchannel.
Figure 2.1.5. Photography of the polymer-chip functioning coupled to the QTOF MS [OP5]
The whole chip/reservoir assembly was mounted to the QTOF MS [OP5]. In
order to realize the electrical contact to the ESI power supply, the ESI QTOF
sampler was removed and the chip system was directly connected to the ESI high
voltage plate, which is a fixed part of QTOF conventional Micromass ESI source.
Exchange between the original source and chip system interface did not claim for
any definitive dismantling or special mechanical modifications to either of the
original assembly and no further modifications on the TOF/MS analyzer were
necessary. The position of chip emitter was adjusted in the vicinity of the entrance
hole of the QTOF MS counterelectrode (sampling cone) by the source assembly,
manipulated in x, y and z direction via micrometer screws. The microsprayer tip
was placed at a distance of about 5 mm.
Counterelectrode
Sprayer
chipQTOF
MS
ESI
plume
37
The electrical contact was ensured by a conductive wire with one terminal
connected to the chip electrode and the other fixed on the ESI high voltage plate.
The spray could be initiated at values of 2-3 kV, in the negative ion mode, applied
to the nanoESI plate and 80-100 V applied to the sampling cone without the need of
nebulizer gas.
In Fig. 2.1.5. a photography of the QTOF source assembly with mounted
polymer chip is presented. The photography has been taken immediately after
application of the negative voltages on both chip and counterelectrode [OP5]. The
ESI plume is clearly visible demonstrating the instant initiation of the electrospray.
For each sample a new chip was used thus any contamination was
prevented. Under the same ESI QTOF MS conditions, the in-run reproducibility of
the experimental data in terms of sensitivity, spray stability, number of detected
components/fragments, ion intensity and charge state was almost 100% while the
day-to-day reproducibility was 95-100%.
2.2. Interfacing microfluidic systems to Fourier transform ion cyclotron
resonance (FTICR) mass spectrometer
Fourier transform ion cyclotron resonance (FTICR) mass spectrometer at 9.4
T providing ultrahigh resolution and mass accuracy as well as the possibility to
perform multiple stage MS analysis by highly efficient dissociation techniques was
interfaced to the fully automated chip-based ESI robot and the thin polymer chip
microsprayer. Interfacing the FTICR MS to the polymer chip microsprayer was
reported for the first time in 2003 by Przybylski et al. [41]. The system described by
the authors was successfully implemented in proteomics. Fully automated chip-
based ESI robot was not previously reported in combination with a FTICR MS
instrument.
Both systems described below are representing currently the most advanced
mass spectrometric techniques. They have been introduced here for the first time in
carbohydrate research and employed for high performance glycoscreening and
sequencing.
38
FTICR mass spectrometer
FTICR MS has matured to become an indispensable tool in bioanalytical
studies for analysis of complex mixtures, such as those encountered in glycomics
[47-50]. The unique features of the FTICR MS in comparison to all other MS
methods are the ultra-high resolution exceeding 106 and the mass-determination
accuracy very often below 1 ppm. Additionally, FTICR MS provides the advantage
of several ion fragmentation techniques based on precursor dissociation such as
collision induced-dissociation (sustained off-resonance SORI CID), infrared laser
multiphoton (IRMPD), or electron capture dissociation (ECD) as well as the
possibility to perform multiple stage MS (MSn). For the analysis of native
glycoconjugate mixtures, nanoESI FTICR MS and multiple stage MS in the negative
ion mode was shown to be most substantial for screening and sequencing of
complex carbohydrate mixtures originating from biological sources [OP14].
All experiments which will be further described were performed on a high
performance Bruker Apex II Fourier transform ion cyclotron resonance mass
spectrometer (Bruker Daltonik, Bremen, Germany) equipped with nanoESI (Apollo
ion source), a 9.4 T superconducting actively shielded magnet (Magnex Scientific
Ltd., Oxford, UK) and a InfinityTM cell.
The most important part of the FTICR MS instrument is the analyzer cell,
which resides in a strong, homogeneous magnetic field. The analyzer cell can take
on different geometries but generally consists of a front and back trapping
electrode, two opposite excitation electrodes and two opposite detection electrodes.
The analyzer cell is in fact a low pressure (10–10 mbar) Penning trap in which ions
can be stored for extended periods of time. The timescale of the experiment is
another particular feature of FTICR MS, therefore it may be used to study slow
(and fast) ion-molecule reactions, slow conformational changes in biomolecules, the
dissociation of very large molecules with a large number of degrees of freedom,
and many more processes that require both gas-phase ions and time to complete.
The role of the analyzer cell is to determine the mass-to-charge ratio of the
ions stored in it. Each ion moving in a spatially uniform magnetic field will
39
describe a circular, so-called cyclotron, motion as a result of the Lorenz force and
the centrifugal force operating on it in opposite directions. The angular frequency
of this motion is given by:
c=qB0/m (2.2.1.)
where c is the unperturbed cyclotron frequency and is solely depending on the
magnetic field induction B0 and the mass-to-charge ratio m/q. Modern
superconducting magnets with a field strength ranging between 7 and 15 T usually
drift only several ppm per year, so the cyclotron frequency can be an extremely
accurate measure of m/q ratio.
The ions are exposed to an oscillating electric field that produces a net
outward electric force on the ions for a limited period of time. This oscillating field
is created by applying an RF potential on the two excitation electrodes and is
referred to as the excitation pulse. The ions will only experience a net continuous
outward force if the frequency of the oscillating electric field is resonant with the
cyclotron frequency of the ions. To ensure excitation of all ions trapped in the ICR
cell, an RF pulse comprising multiple frequencies is employed such that all ions of
different m/q ratios are exposed to a net outward electric force for the same amount
of time. The radius after excitation is shown to be independent of m/q as long as
the magnitude of the excitation signal is constant with frequency.
After excitation, the radius of the ion cloud increases and all ions with the
same mass-to-charge ratio move coherently in a circular orbit. This coherently
moving ensemble of charges at a radius close to the cell electrodes will induce an
oscillating differential image current in two the opposite detection electrodes. This
image current is then amplified and digitized yielding a time domain signal or
transient containing signal contributions from all excited ions in the cell.
Sustained off-resonance collision-induced dissociations (SORI-CID)
On the FTICR instrument, CID is performed by exciting an isolated ion to a
higher cyclotron radius (and, therefore, to higher kinetic energy) in the presence of
40
an increased background pressure of a neutral gas. Collisions occur as a result of
the reduced path length and increased ion velocity; this leads to a transfer of
energy to the two collision partners with mostly kinetic energy transfer to the
neutral and conversion to internal energy in the ion. This internal energy is rapidly
redistributed about the ion’s structure and, if it locally exceeds the energy required
for dissociation, the ion breaks apart.
There are multiple strategies [51, 52] for increasing the kinetic energy of the
ions but the one used in our studies is the sustained off-resonance ion (SORI)
irradiation/excitation.
SORI is a soft excitation technique enabling the operator to focus on the
lowest energy fragmentation pathways. In SORI, the precursor ion is subjected to
dipolar radiation at a frequency slightly offset from its cyclotron radius. This
results in the ion alternately increasing and decreasing in radius and kinetic energy
over the course of the SORI excitation so that collisions deposit lower internal
energies per collision (typically 0.3 V), but many more collisions occur, hundreds
per second, typically. As the internal energy accumulates, assuming that cooling
mechanisms such as infrared radiative cooling are slower than the internal energy
build-up, it is rapidly randomized throughout the ion and the lowest dissociation
pathways are sampled. The product ions formed in this way have cyclotron
frequencies separated far enough from the SORI frequency so that their continued
excitation is minimal and any subsequent collisions serve only to cool their residual
kinetic energy.
2.2.1. Coupling of the fully automated chip-based ionization to FTICR mass
spectrometer
Introduction of the fully-automated chip-based nanoelectrospray in combination
with QTOF-tandem MS for the first time in glycomics demonstrated the major
advantages of this approach for structural investigation of complex carbohydrate
systems. In this context, the NanoMate robot has been for the first time [OP15]
coupled to the FTICR MS at 9.4 T and optimized in the negative ion mode to
41
combine in one system automated sample delivery on the chip along with
maximal sensitivity, ultra-high resolution, accurate mass determination and
efficient tandem MS.
A specially designed interface conceived by Bruker Daltonik (Bremen,
Germany) was constructed in order to obtain a viable coupling of the Nanomate
system to the Bruker Apex II Apollo ion source. The coupling interface consists of a
prototype mounting bracket [OP15]. The NanoMate robot was connected to the
FTICR MS instrument as depicted in Fig. 2.2.1.
Initiation of the electrospray and efficient transfer of the ionic species into
MS have been accomplished by a fine adjustment of the NanoMate microchip
position with respect to the FTICR ion transfer capillary. The Apex II metal-coated
glass capillary was set to create a slightly attractive potential for the ESI generated
negative ions and the capillary exit voltage was set values minimizing the in-source
ion fragmentation. The ESI generated ions were accumulated in the hexapole
located after the second skimmer of the ion source and then transferred into the ion
cyclotron resonance cell. The generated ions were trapped by Sidekick trapping.
For SORI CID experiments the precursor ions were isolated by application of
a broadband excitation pulse to eliminate all ions except those of interest. The robot
was programmed to aspirate the sample solution and to submit it to (-) chipESI
FTICR MS experiments. The generation of the chip electrospray required an initial
high back pressure which, was slightly decreased to lower values during
acquisition.
Figure 2.2.1. Coupling of the NanoMate robot to the FTICR mass spectrometer [OP15]
42
The fine tuning on the x-, y- and z- axis of the chip nozzle toward the MS
inlet represented a critical step in the optimization of the ionic species transfer into
the FTICR MS inlet. In addition, a significant increase of the spray stability and
intensity of the MS signal was achieved by applying a low potential of 50 V to the
ion transfer capillary of the FTICR MS instrument. Under these well-defined
conditions a constant and stable spray, significant increase of sensitivity (1pmol/l)
and high intensity of the MS signal over long signal acquisition time were obtained.
Another major benefit of the NanoMate-FTICR MS system was the high ionization
efficiency, with formation of multiply charged glycoconjugate ions, which in
conjunction with the high sensitivity, resolution and mass accuracy allowed for the
identification of minor glycoforms previously undetectable by any other MS-
related technique.
The capability of the automated chip ESI FTICR MS approach for complete
structural elucidation by SORI CID MS2 was also evaluated and proved to provide
accurate and structural-informative fragmentation data [OP15].
2.2.2. Interfacing the thin chip microsprayer system to FTICR MS
The Apex II FTICR mass spectrometer equipped with a 9.4 T
superconducting actively shielded magnet and the Bruker Apex II Apollo ESI ion
source was interfaced to the disposable thin polymer microchip described in §2.1.3.
For the first time such a system was optimized in the negative ion mode, to
detect carbohydrate ions [OP16].
For sample loading and FTICR interfacing, the chip was sandwiched in a
home-made chip holder with an integrated reservoir (Fig. 2.2.2). The chip was
positioned into the holder with the microchannel in contact with the reservoir and
the front part extruding a few mm. The chip was grounded via a conductive wire
connected to the terminal electrode. The chip was coupled to the Bruker Apex II ion
source by an in-laboratory constructed mounting system [OP16]. The interface
consists of a metal plate mounted to the Apollo ion source by two 90o brackets. The
43
plate and the 90o brackets featured two slots for the screws, thus providing x- and
z-axis movement and some y-axis variability.
Figure 2.2.2. Schematic of the polymer chip incorporated into the sandwich holder for coupling with the FTICR MS [OP16]
The chip holder was attached to the metal plate and carefully positioned to
point toward the entrance orifice of the Apex II capillary. Mounting of the chip
interface system did not require any significant dismantling of, or irreversible
mechanical modifications to any of the original Apollo source components.
Moreover, the exchange between the original source and chip interface is quick and
simple.
The Apex II metal-coated glass transfer capillary was kept in the range 1500–
2500 V while the chip was grounded. This potential difference facilitated the ESI
driven by the electroosmotic flow. The generated ions were accumulated for 0.05 to
0.3 s in the hexapole located after the second skimmer of the ion source, and then
transferred into the ion cyclotron resonance cell.
The in-house made mounting system provided a robust and viable
interfacing of the polymer chip to the FTICR MS instrument. The
positioning/alignment of the polymer chip on the x-, y- and z- axes turned out to
be a crucial step in the initiation and long-term maintenance of the electrospray, in
particular with respect to the direct-spray configuration of the Bruker Apex II
Polymer
microchipESI plume
Sandwich
chip holder
Sample
reservoir
Conductive wire
44
instrument. However, under optimized conditions, the polymer chipESI-FTICR MS
system provided a high ionization yield, an extremely stable and long-lasting
spray, high sensitivity, and minimization of in-source fragmentation of labile
moieties such as sialic acid residues.
All these advantages along with the high mass accuracy detection provided
by the FTICR instrument made this technique a real option for achieving improved
and detailed structural characterization of oligosaccharides and glycoconjugates.
2.3. Interfacing the fully automated chip-based ionization (NanoMate robot) to a
high capacity ion trap (HCT) mass spectrometer
High capacity ion trap (HCT) mass spectrometer is currently one of the most
efficient types of ion trap instruments. Released by Bruker Daltonics company a
few year ago, HCT provides outstanding ion trap performance in terms of
sensitivity, speed and mass accuracy. HCT exhibits an up to 15 fold higher ion
storage capacity than the regular trapping instruments, which contributes to the
dramatic increase in sensitivity [53]. The instrument is actually the fastest and most
sensitive ion trap mass spectrometer. The high ion capacity, dynamic range, speed
and multistage fragmentation (CID up to MS11) make this instrument ideal for high
throughput glycomics and proteomics as well as for quantitative analyses.
The HCT ultra PTM (posttranslational modifications) instrument employed in
these studies is equipped with electron transfer dissociation (ETD) source using
fluoranthene as the reagent. In contrast to CID (previously the only available
fragmentation technique on ion traps) ETD induces specific N-Cα bond cleavages of
peptide backbone with the preservation of the post-translational modification [54-
55] and consequently with generation of ions that are diagnostic for the
modification site(s). Together, ETD and CID as well as alternating ETD/CAD
feasible on HCT MS, may significantly increase the sequence coverage and give
added confidence to protein, glycoprotein and glycopeptide identification [OP17].
45
Figure 2.3.1. NanoMate-HCT coupling and the robot-MS connection via silicon chip
In view of these advantages of HCT MS with CID and ETD MSn, in this stage of
research, it was conceived the first combination of fully automated chip-nanoESI
with HCT MS [OP18] to yield a platform on which high throughput glycomics to
be feasible.
The mass spectrometer employed in this work was a HCT Ultra PTM Discovery
from Bruker Daltonics (Bremen, Germany). The HCT MS is interfaced to a PC
running the Compass 1.2 integrated software package, which includes the Hystar
3.2.37 and Esquire 6.1.512 modules for instrument controlling and
chromatogram/spectrum acquisition as well as Data Analysis 3.4.179 portal for
storing the ion chromatograms and processing the MS data. The robot was coupled
to the HCT Ultra mass spectrometer [OP18] via an in-laboratory made mounting
46
system (Fig. 2.3.1.) For NanoMate interfacing, the conventional Bruker electrospray
ion source was detached and all HCT instrumental settings and electrical
connections were readapted to functioning in the MS-decoupled ESI regime of the
NanoMate system. The robot was set up on three O-xyz adjustable supports and
connected to the HCT MS nebulizer nitrogen supply pipeline. The position of the
electrospray chip was adjusted with respect to the HCT counterelectrode to ensure
an optimal transfer of the ionic species to the mass spectrometer.
NanoMate-HCT MS system demonstrated a high reliability and versatility as it
could be successfully applied to a broad class of biomolecules, which required
different instrumental conditions for ionization, detection, screening in MS and
sequencing by CID and ETD multistage MS. As described in the next chapter,
NanoMate-HCT coupling, implemented in Romania for the first time in the world,
was optimized and it is now completely functional for compositional and
fragmentation analysis in positive and negative ion mode of biomolecules such as
peptides and proteins [OP17], glycolipids/gangliosides [OP18-OP26], N- and O-
glycans [OP27-OP33] and small molecules as well [OP33-OP36].
47
PART III
Applications of microfluidics/electrospray ionization mass spectrometry to
structural analysis of glycoconjugates in biomedical research
(Own reported results)
3.1. Introduction
Carbohydrates represent a class of biopolymers with high degree of
structural complexity. They are polyhydroxylated aldehydic and ketonic
compounds classified as monosaccharides, oligosaccharides and polysaccharides
according to the size of the molecules and related to the number of monomeric
units connected by glycosidic bonds. Carbohydrates are present either as
oligosaccharides or as glycoconjugates in which the oligosaccharide chain is
covalently linked to an aglycon, frequently another biopolymer such as a protein
and/or a lipid. Carbohydrates occur ubiquitously in nature displayed on
macromolecules and the surface of cells being involved in basic biological
functions, such as antigen recognition machinery, cellular adhesion of bacteria and
viruses, and protein folding, stability and trafficking [56, 57]. Particular structures
were found biomarkers of severe diseases and others to play an essential role in
fertilization and embryogenesis. Due to the large number of saccharide building
blocks and variety of linkages between them, this biopolymer category has also a
high potential to carry information.
The large discrepancy between the extreme diversity of the glycoforms
found in nature, their high biological importance and mostly an infime quantity
available from biological sources, employed lately massive work in development of
sensitive and specific methods for compositional mapping of heterogeneous glycan
mixtures and the structural elucidation of their single components. The complete
structural analysis of carbohydrates includes: a) molecular weight determination;
b) identification of number and type of saccharide components; c) determination of
48
sequence and patterns of branching; d) determination of glycosidic attachment sites
and their anomeric configuration; e) identification of the type and conformation of
glycosyl ring; f) determination of their secondary structure. In the last years, ESI
MS demonstrated its potential for structural elucidation of carbohydrates being
able to provide information related to a)-f) determinants, which significantly
increased in amount and precision after introduction of nanoESI MS and tandem
MS.
In the Part III of this work it will be demonstrated the contribution of
microfluidic/ESI MS to elucidating complex issues raised in biomedical research, in
particular those related to the structural identification of glycoconjugates with
potential biomarker value. Throughout this part, glycan-related fragment ions were
assigned according to the nomenclature [58] introduced by Bruno Domon and
Catherine E. Costello (Fig. 3.1.1).
Figure 3.1.1. Types of fragment ions in tandem mass spectra of linear polysaccharides and glycoconjugates and their assignment according to the nomenclature [58]. a) ions produced by cleavage of glycosidic linkages; b) ring cleavage ions
O
OH
OH
OH
CH2OH
O
O
OH
OH
O
CH2OH
O
O
OH
OH
O
CH2OH
O
O
OH
OH
O
CH2OH
O R
O,2A12,4A2
2,5A3
1,5X12,5X2
1,4X3
1
23
4 5
6
O
OH
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O R
ZoYoY1Y2Y3 Z1Z2Z3
B1 B2B3 B4C1 C2
C3 C4
Capatul nereducator Capatul reducatorNon-reducing end Reducing end
O
OH
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O
O
OH
OH
CH2OH
O R
ZoYoY1Y2Y3 Z1Z2Z3
B1 B2B3 B4C1 C2
C3 C4
Capatul nereducator Capatul reducatorNon-reducing end Reducing end
a)
b)
49
3.2. Screening, sequencing and structural identification of O-glycopeptides from
urine of patients suffering from Schindler disease
“There is no treatment for this disease but knowledge of the mutations causing it permits molecular-
based prenatal diagnostic studies”. Robert J. Desnick and Detlev Schindler
Schindler disease is a recently recognized autosomal recessive disorder
caused by the deficient activity of -N-acetylgalactosaminidase (NAGA), a
lysosomal hydrolase previously known as -galactosidase B.
Clinically, the disease is rather heterogeneous with three different
phenotypes identified to date. The most severe form is the type I, an infantile-onset
neuroaxonal dystrophy. It has been described [59] in three related German infants:
two siblings born from consanguineous parents and a distant cousin. All three
children were born after a normal pregnancy, labor and delivery. The sibs are
currently alive, in the state which is described below, while their cousin died
unexpectedly at 18 months of life from apnea during a seizure with prolonged
convulsion.
The clinical course experienced by the siblings was characterized by three
stages: i) apparently normal development in the first 9 to 12 months; ii) a period of
developmental delay followed by rapid regression starting with the second year of
life (with the younger brother deteriorating faster); iii) progressive neurological
impairment resulting by 3 to 4 years of age in cortical blindness, deafness,
spasticity, myoclonus, decorticate posturing and profound psychomotor
retardation and little, if any, contact to the environment. At 4 and respectively 5
years of age, the affected brothers had developmental skills at the newborn level,
did not have anymore voluntary movements, any contact to the environment and
response to the stimuli. They did not appear to see or hear, were incontinent and
dependent on tube feeding. Both brothers survived episodes of pneumonias due to
the diligent nursing effort of their parents but remained to date in this vegetative
state.
A milder form, Schindler disease type II (also called Kanzaki disease) is an
adult-onset disorder characterized by angiokeratoma corporis diffusum and mild
50
intellectual impairment [60]. To date three affected adults: one Japanese and two
Spanish sibs, have been identified and are alive.
Schindler disease type III, described [61] very recently in two Dutch sibs and
one unrelated French child is an intermediate and variable form with
manifestations ranging from seizures and psychomotor retardation in infancy to a
milder autism, with speech and language delay and marked behavioral difficulties
in early childhood.
In all types of this rare inherited lysosomal storage disease, the severe
enzymatic defect (enzyme residual activity ranging from 0.5% to 2% in plasma,
lymphoblasts and fibroblasts) leads to an abnormal accumulation of sialylated and
asialo-glycopeptides and oligosaccharides with -N-acetylgalactosaminyl residues
(mucine type of O-glycosylation) in various tissues and body fluids.
In human urine, complex carbohydrates are catabolic products excreted
either as free oligosaccharides or linked to peptides, and their structures and
amounts are known to vary under different physiological and pathological
conditions. In all three types of Schindler disease, the deficient NAGA was found to
cause glycopeptiduria and the concentration of O-glycans in urine was estimated to
be 100 times higher than in healthy controls. For this reason, screening, structural
characterization and complete identification of O-GalNAc glycosylated aminoacids
and peptides extracted from patients‘ urine is of major diagnostic importance.
The developed arsenal of microfluidic/mass spectrometry methods
presented in the Part II has been employed de novo for the analysis of O-
glycosylated peptides in the urine of the two German siblings diagnosed with
Schindler disease type I.
The study was focused on:
1. Thorough screening of sialylated and asialo glycopeptide variants and free
oligosaccharides present in the urine of both affected children.
2. Detection, sequencing and identification of all structures possibly
modified/elongated by peripheral attachments such as O-Ac, Fuc, SO3 and
extended number of NeuAc residues.
51
3. Detailed structural analysis by tandem MS based on highly efficient collision-
induced dissociation techniques.
4. Determination of O-GalNAc-Ser/Thr expression in urine of a healthy age-
matched child and comparative analyses.
The complex mixtures of O-glycosylated peptides were extracted, purified,
separated and pre-fractionated at the University of Bonn. The whole preparation
procedure is described in details in ref. [62] and briefly in the original publications
[OP5, OP15, OP16].
In the first stage, the analysis was focused on the determination of the O-
glycopeptide expression in the urine of the younger child who suffered a faster
deterioration and completely reached the vegetative state at the age of 4 years. In
the case of this patient, the NAGA activity as percent of normal enzyme activity
was found as follows: plasma 0.5%, lymphoblasts 0.5%, fibroblasts 0.7%.
Figure 3.2.1. Core 1 a) and 2 b) of O-GalNAc glycosylation
Figure 3.2.1. Ser- (Thr, Thr-Pro) linked tetrasaccharide (core1)
GlcNAc1-6
Gal1-3GalNAc Gal1-3GalNAc
a) b)
NeuAc52 \ 6 GalNAc1-Ser (Thr, Thr-Pro) 3 / Gal1 3 / NeuAc2
52
CE in off-line conjunction with negative ion mode nanoESI QTOF CID
MS/MS was first developed for assessing the glycopeptide mixture heterogeneity
and identification of the components. In the collected CE fractions eleven structural
elements typical for O-glycosylation of proteins, like expression of core 1 and 2
(Fig. 3.2.1) type O-glycans with different numbers of N-acetyllactosaminyl repeats
and different degrees of sialylation, could be directly detected and identified by
optimized MS and CID MS/MS experiments.
In Fig.3.2.1 the structure of the tetrasacharide O-linked aminoacid (Ser, Thr)
or dipeptide (Thr-Pro) species corresponding to the most abundant ions detected in
the CE fractions is depicted.
A significant extension of the sensitivity limit for detection of minor
components in this mixture was achieved by a novel analytical approach based on
sheathless on-line forward polarity CE negative ESI QTOF MS and MS/MS [OP12].
Figure 3.2.2. On-line sheathless CE ESI MS TIC in negative ion mode of the mixture of O-glycosylated sialylated peptides from urine of the patient suffering from Schindler’s disease. CE potential, 30 kV; carrier, 50 mM aqueous+40% MeOH ammonium acetate/ammonium hydroxide, pH 12.0; c = 0.75 mg/mL buffer; 8 s injection by pressure; 25 nL injected; ESI potential, 1.1 kV; ESI sampling cone potential 40 V [OP12]
By implementation of the home-built sheathless CE ESI microspray interface
consisting of a one-piece copper-coated CE column etched as microsprayer, the
separation efficiency and the resolution obtained in CE UV experiments could be
53
Table 3.2.1. Assignment of the major ions from the mixture of O-glycosylated sialylated peptides from urine of patients suffering from Schindler’s disease [OP12]
reproduced with high sensitivity in on-line CE MS runs under mild ESI-negative
ion mode conditions, due to the compatibility of the reconsidered CE-QTOF MS
operating parameters and microspray tip performance. In the sheathless CE ESI-
QTOF MS TIC the electrophoretic peaks (Fig. 3.2.2) could be identified from their
derived mass spectra, which revealed more than 50 biologically-significant
sialylated and asialo-glycopeptide structures (Table 3.2.1).
54
Moreover, the method revealed structures elongated by fucosylation and/or
extended chains with higher degree of sialylation not detectable before and not
known to be present in the mixture. Detailed structural information upon the
separated species was obtained by data-dependent analysis carried out in the high
speed automated „on-the-fly“ MS-MS/MS mode switching which was for the first
time introduced as fragmentation method in glycomics [OP12].
For development of a more efficient protocol based on CE separation and mass
spectrometric screening of glycopeptide expression in the patient urine, the QTOF
MS was coupled to the sheathless nanoESI interface in-laboratory designed for such
purposes. The system was optimized for operating in the negative ion mode (MS)
and reverse CE polarity [OP37]. So far, the on-line reverse polarity CE (-)ESI-QTOF
MS (RPCE (-) nanoESI QTOF MS) was carried out under low buffer pH, low
concentration and with coated capillaries to suppress the EOF, and pressure
assistance to reduce the diffusion processes and the analysis time. However, a
major drawback of the pressure assisted sheathless RPCE (-)nanoESI QTOF MS is
the considerable decrease of separation efficiency and resolution. Therefore, such
an approach was not considered beneficial toward the separation/detection of all
components in this complex mixtures. For this reason, the development of a
sheathless RPCE (-) nanoESI QTOF MS method, based solely on the migration of
components in electrical field without assistance of pressure and coating of the
capillaries has been implemented by total reconsideration of the solution and CE
instrument parameters and operation mode [OP37].
The spectra derived from the most prominent detected TIC-peaks (Fig. 3.2.3)
clearly indicated that the mixture is dominated by the Ser-, Thr- and Thr-Pro-
linked tetrasaccharide bearing two sialic acid moieties, hexasaccharide bearing two
sialic acids and monosialo trisaccharides (Fig. 3.2.4., Table 3.2.2). These results are
in agreement with the data obtained by all previous experiments.
55
Figure 3.2.3. RPCE (-)nanoESI QTOF TIC MS of the BQ5 fraction of O-glycosylated sialylated peptides from urine of the patient suffering from Schindler’s disease. c = 0.75 mg/mL buffer (5 pmol injected); CE voltage, -25 kV; CE buffer, 0.1 mM methanol/water (6:4%v/v) formic acid, pH 2.8; CE capillary length, 130 cm [OP37]
A detailed inspection of data, revealed, however, that a larger number of minor
components, doubly and triply charged ions corresponding to molecular masses up
to 4000 Da, previously not detectable in this complex mixture were detected. The
low ionic intensity exhibited by these components can be rationalized by their low
abundance in this mixture showing a high dynamic range proportions.
Nevertheless, the potential of this approach to separate and detect with high
sensitivity even less abundant components, previously not accessible due to
overlapping of isobaric structures and/or low content in the original biological
material, is of major importance for progress in detailed identification of all
structures related to this disease. The assignment of some of these species
according to their molecular ions was conducted under the hypothesis that
modification of glycopeptides by sulfation and acetylation could be present.
56
Figure 3.2.4. (a) RPCE (-)nanoESI QTOF MS obtained by combining across the extracted ion chromatogram (XIC) of the ion at m/z 525.3 corresponding to Neu5Ac2HexHexNAc-Ser. Inset: XIC of the ion at m/z 525.3 processed from the TIC-MS in Fig. 3.2.3. (b) RPCE (-)nanoESI QTOF MS obtained by combining across the XIC of the ion at m/z 532.3 corresponding to Neu5Ac2HexHexNAc-Thr. Inset: XIC of the ion at m/z 532.3 processed from the TIC MS in Fig. 3.2.3. (c) RPCE (-) nanoESI QTOF MS obtained by combining across the XIC of the ion at m/z 580.9 corresponding to Neu5Ac2HexHexNAc- Thr-Pro. Inset: XIC of the ion at m/z 580.9 processed from the TIC MS in Fig. 3.2.3 [OP37]
57
Table 3.2.2. Assignment of the major and minor species detected by RPCE (-) nanoESI QTOF MS in the BQ5 fraction [OP37]
58
Thus at this point four ions could be attributed to the: (SO3)Neu5AcGalGalNAc-Ser
and -Thr linked, respectively; (SO3)Neu5AcGal2GlcNAcGalNAc-Ser/Thr and
(SO3)Neu5Ac2Gal2 GlcNAcGalNAc-Ser/Thr. Though of lower intensity, the ion
corresponding to the disialo element was detected accompanied by its O-Ac
counterpart, demonstrating the presence of an O-acetyl-modified sialic acid in the
patient urine.
To detect and identify all O-glycoforms present in patients urine, new
accurate methods for MS mapping and sequencing were required.
Figure 3.2.5. (–)NanoESI FTICR MS at 9.4 T of the BPy1 mixture of O-glycosylated sialylated peptides from urine of the patient suffering from Schindler’s disease. Solvent: 0.1M HCOOH/NH4OH (pH 2.8); sample concentration: 5 pmol/μL; number of scans: 10 [OP14] Therefore, a strategy for screening of these complex glycoconjugate mixtures based
on FTICR MS at 9.4 T was developed [OP14] in order to obtain accurate mass
determination, ultrahigh resolution and highly efficient dissociation of precursor
ion. In the ESI FTICR mass spectrometric analysis (Fig. 3.2.5) particular attention
was paid to original sialylation patterns, and degree because of the potential
lability of the sialic acid moiety during the desorption/ionization process.
59
Table 3.2.3 Assignment of the major sialylated glycopeptide ions from the BPy1 mixture detected by (–)nanoESI FTICR MS at 9.4 T [OP14]
Under solvent conditions enabling the sensitive MS analysis, negative ion nanoESI-
FTICR MS at 9.4 T was shown to represent a method of choice for identification of
these components and realistic assessment of mixture heterogeneity (Table 3.2.3).
By optimization of highly efficient fragmentation techniques such as sustained off-
resonance irradiation (SORI)-CID MS/MS in the negative ion mode, the type and
sequence of the sialylated glycopeptide components were determined from their
fragmentation patterns (Fig.3.2.6). Additionally, implementation of multiple stage
MS by SORI-CID MS/MS/MS provided detailed information regarding the
sialylation status.
In a more advanced phase of the research, the performance of the NanoMate
system to provide long-lasting electrospray signal, rendering reliable conditions for
high sensitivity, was of a particular usefulness for detection and sequencing of
minor glycopeptide components, previously not accessible for fragmentation from
such complex mixtures. The mixture heterogeneity was assessed further by
NanoMate robot in direct coupling first with QTOF MS and CID MS/MS via direct
infusion (Fig. 3.2.7, Fig.3.2.8 and Table 3.2.4) [OP13] and subsequently in off-line
coupling with CE [OP38].
60
Figure 3.2.6. SORI-CID MS2 of the singly charged ion at m/z 1051.360 corresponding to NeuAc2GalGalNAc-Ser. Ion isolation by correlated shots; Number of scans: 30; [OP14]
The latest experiment [OP38] resulted in an assembly of 3 instruments in
series: CE for off-line fraction collection, NanoMate for automatic chip infusion of
the CE fractions and QTOF MS for screening and sequencing. By NanoMate/QTOF
MS and off-line CE-NanoMate/QTOF MS singly and doubly charged ions derived
from tri-, to decasaccharide O-linked either to Ser, Thr or to the Thr-Pro dipeptide
expressing patterns typical for GalNAc-type O-glycosylation were identified [OP2,
OP13, OP38]. Singly charged ions assigned to free oligosaccharides like NeuAc,
NeuAcGal, and NeuAcGalGalNAc were observed in the mixture as well.
61
Figure 3.2.7. Fully automated chip-based (-) nanoESI QTOF mass spectrum of BPy O-glycopeptide mixture from urine of a patient suffering from Schindler’s disease. Substrate concentration, 3 pmol/μL in MeOH. Sampling cone potential, 30 V [OP13] Table 3.2.4. Assignment of the major sialylated glycopeptide ions from the BPy mixture detected by fully automated chip-based (–)nanoESI QTOF MS [OP13]
62
a)
b)
Figure 3.2.8. Fragmentation spectrum (a) and scheme (b) obtained by fully automated chip-based (-)nanoESI QTOF CID MS/MS derived by using as the precursor ion NeuAc2Gal3GlcNAc2GalNAc-Ser detected as a doubly charged ion at m/z 890.32. Collision energy range (25-40) eV; Sampling cone potential 30 V [OP13]
63
The reduced signal-to-noise ratio of low abundant molecular precursor ions
in complex mixtures obtained by classical nanoESI MS/MS can generally be
overcome by long signal accumulation in an off-line approach, but it is frequently
associated to the spray instability and/or signal interruptions. In the case of the
chip nanoESI MS/MS even the very low abundant ions, like the one corresponding
to a disialooctasaccharide O-linked to Ser species could be successfully fragmented.
According to the fragmentation pattern the structure of the extended backbones of
core 2 type present in the urine glycoconjugate sample was elucidated.
Fully automated chip electrospray (NanoMate robot) was for the first time in
the world coupled to FTICR mass spectrometry and the system was applied to
high-performance glycoscreening and sequencing of O-glycopeptides from urine of
Schidler’s disease patients. NanoMate/FTICR MS screening [OP15] provided a
spectrum of extremely high complexity (Fig. 3.2.9) and, besides the already known
species, revealed a high number of doubly and triply charged ions detected as
lower abundant components (Table 3.2.5) within a relative narrow m/z ranges, like
700-780, 820-870, and 1050-1100.
Figure 3.2.9. Fully automated (-) chipESI FTICR mass spectrum of the Q5 fraction from urine of a patient suffering from Schindler’s disease. Sample concentration: 5 pmol/μL in MeOH [OP15]
64
Table 3.2.4. Ions detected and identified with a mass accuracy below 12 ppm in the Q5 mixture of O-glycosylated amino acids and peptides at 5 pmol/μL [OP15]
An unambiguous structural assignment at a mass accuracy below 10 ppm could be
achieved for the structures Ser- and Thr- linked hexasaccharide and the
nonasaccharides Neu5Ac3Hex2HexNAc4-Ser/H2O and Neu5Ac3Hex2HexNAc4-
Thr/H2O. Four new components, not identified so far by any other method, were
detected by this method and assigned with a mass accuracy well below 10 ppm.
Interestingly, the method disclosed the presence of two previously unknown
undecasaccarides bearing three sialic acid moieties detected as triply dehydrated
sodiated counterpart ions of Neu5Ac3Hex3HexNAc5-Ser and
Neu5Ac3Hex3HexNAc5-Thr.
For comparative study, a mixture of glycopeptides extracted from urine of a
age-matched healthy control person was subjected to compositional and structural
analysis by (-)NanoMate/FTICR MS (Fig. 3.2.10, Table 3.2.5) and SORI CID
MS/MS [OP15], thin chip polymer microspray FTICR MS [OP16] and thin chip
polymer microspray /QTOF MS (Fig. 3.2.11, Table 3.2.6) and MS/MS [OP5] at the
same ionization/detection and sequencing conditions as those employed for the
mixtures originating from patient urine.
65
Figure 3.2.10. Fully automated (-)chipESI FTICR mass spectrum of the Ty mixture from urine of a healthy control person. Sample concentration: 7 pmol/μL in MeOH. Number of scans: 61 [OP15] The (-) nanoESIchip FTICR MS pattern of this mixture was observed to be different and the spectrum rather poor in ionic species. Table 3.2.5. Ions detected and identified with a mass accuracy below 12 ppm in the Ty mixture of O-glycosylated amino acids and peptides at 7 pmol/μL [OP15]
66
The mixture was found to contain a reduced number of species, having as the
dominant components structures of shorter chains and lower degree of sialylation
(maximum 2) such as: monosialo Ser, Thr-, and Thr-Pro-linked trisaccharide and
Thr-Pro linked disialo tetrasaccharides, monosialo Ser- and Thr linked
pentasaccharides of lower abundance, disialo Ser- and Thr- linked hexasaccharides
and Thr-Pro linked disialo hexasaccharide much less abundant. The structure of
the latter components were identified by NanoMate /FTICR SORI CID MS/MS
experiment, which showed that the molecule configuration is identical to the
hexasaccharide mono- and dipeptide found in the patient urine. Octasaccharides
were barely represented in the spectrum and only at low abundance Ser- linked
disialo octasaccharide was detected as a doubly charged ion. Longer chains and/or
species of higher sialylation degree were not found in the urine of the healthy
infant.
Figure 3.2.11. Thin polymer microchip ESI QTOF MS of the Ty mixture of O-glycosylated amino acids and peptides from normal human urine. ESI voltage 2.8 kV; sampling cone potential 100 V; signal acquisition 20 scans. Solvent: MeOH; average sample concentration: 5 pmol/μL [OP5]
67
Thin chip polymer microspray/ QTOF MS screening (Fig. 3.2.11) and CID
MS/MS sequencing (Fig. 3.2.12), revealed essentially the same mixture composition
with an additional detection of mono-and disialylated free short (di-to tetra)
oligosaccharides (Table 3.2.6).
Figure 3.2.12. Thin polymer microchip ESI QTOF CID MS/MS of the NeuAc2HexHexNAc-Ser doubly charged ion at m/z 532.08. ESI voltage 2.8 kV; sampling cone potential 100 V; solvent: MeOH; average sample concentration 5 pmol/μL; collision energy 40 eV; signal acquisition 30 scans; sample consumption 1.23 pmols [OP5]
The older German sib affected by Schindler disease type I experienced a
slower regression in development and reached the final state characterized by skills
at newborn level and lack of any contact to the environment at the age of 5 years. In
the case of this child, the NAGA activity as percent of normal enzyme activity was
found: plasma 1.1%, lymphoblasts 0.7%, fibroblasts 1.2%. An extensive, systematic
and comparative study aiming at the determination of O-GalNAc expression in the
urine of the older affected child in relation/comparison to his younger brother and
68
the healthy control (normal human urine) was conducted by (–)microsprayer
chipESI-FTICR MS [OP16].
Table 3.2.6. Compositional mapping of Ty mixture of glycopeptides from normal human urine as detected by thin polymer microchip ESI QTOF MS [OP5]
Briefly, in comparison with the data obtained for the other sib, the mixture was
found to contain a higher percentage of pentasaccharides as indicated by the
presence in the microspray chipESI FTICR mass spectrum of highly abundant ions
corresponding to NeuAcHex2HexNAc2-Ser and NeuAcHex2HexNAc2-Thr [OP16].
Additionally, the hexasaccharides linked to Ser and Thr bearing two sialic acid
moieties were visible at higher abundance as both doubly and singly charged ions.
A novel, previously not reported nonasaccharide structure bearing three sialic acid
moieties was detected [OP16] with a fair abundance as a doubly charged ion
assigned to sodiated dehydrated NeuAc3Hex2HexNAc4-Ser with a mass accuracy
of 4.6 ppm.
69
3.3. Screening, sequencing and structural identification of brain gangliosides
Gangliosides, sialylated glycosphingolipids (GSLs), consist of sialylated
(mono- to poly-) oligosaccharide chain of variable length attached to the ceramide
portion of different composition with respect to types of sphingoid base and fatty
acid residues.
Table 3.3.1. Designation and structure of the gangliosides according to [63]
LacCer, Gal4Glc1Cer; Gg3Cer, GalNAc4Gal4Glc1Cer; Gg4Cer,
Gal3GalNAc4Gal4Glc1Cer; nLc4Cer, Gal4GlcNAc3Gal4Glc1Cer; GD3, II3-
-(Neu5Ac)2-LacCer; GT3, II3--(Neu5Ac)3-LacCer; GM2, II3--Neu5Ac-Gg3Cer;
GD2, II3--(Neu5Ac)2-Gg3Cer; GM1a, II3--Neu5Ac-Gg4Cer; GM1b, IV3--
Neu5Ac-Gg4Cer; GD1a, IV3--Neu5Ac,II3--Neu5Ac-Gg4Cer; GD1b, II3--
(Neu5Ac)2-Gg4Cer; GT1b, IV3--Neu5Ac,II3--(Neu5Ac)2-Gg4Cer; GQ1b, IV3--
(Neu5Ac)2,II3--(Neu5Ac)2-Gg4Cer; 3'-nLM1 or nLM1, IV3--Neu5Ac-nLc4Cer;
nLD1, disialo-nLc4Cer (IV3--(Neu5Ac)2-nLc4Cer
This variability of molecular constitution gives rise to a high number of
species classified into oligosaccharide series according to the major oligosaccharide
core structure. In Table 3.3.1. some abbreviations using the Svenerholm system are
given. In this system, the fact that we are dealing with gangliosides is indicated by
the letter G, the number of sialic acid residues is stated by M for mono-, D for di-, T
for tri-, and Q for tetrasialoglycosphingolipids. A number is then assigned to the
individual compound, which referred initially to its migration order in a certain
chromatographic system. A typical ganglioside is shown in Fig.3.3.1 referred to as
GM1a.
The ceramide portion is embedded in the outer leaflet of the plasma
membrane, while a hydrophilic oligosaccharide chain protrudes into the
extracellular environment. Gangliosides are enriched in the microdomains,
functional membrane units, participating in cell-to-cell recognition/communication
and cell signaling, modulating or triggering various biological events. The central
70
nervous system (CNS) contains the highest content of gangliosides: neuronal
membranes contain at least several times higher concentrations of gangliosides
then the extraneural cell types, highlighting their special role in the CNS [64, 65].
Figure 3.3.1. Structure of GM1a
Ganglioside composition is species- and cell type-specific and changes
specifically during brain development, maturation, aging and disease or
neurodegeneration. For this reason, gangliosides are considered valuable tissue
stage- and/or diagnostic markers and even potential therapeutic agents [66].
In human brain, the brain region-specific differences in ganglioside
composition and quantity as well as in their distribution and cell surface expression
have been demonstrated primarily by thin-layer chromatographic (TLC),
immunochemical and immunohistochemical methods [67-69]. These observations
were based only on comparison concerning the major species due to detection
limitations of the used methods. The region-specific differences most probably
reflect the chemical basis of a high complexity of brain organization and the
functional specialization of regions. This important investigation issue is still far
from systematic characterization. As an example, cerebellum a highly specialized
part of the brain showed some characteristic differences in composition of major
ganglioside species in comparison to the cerebrum. Moreover, function, behavior
CeramideH
H
H
CH3
(CH2)m
CH
CH
CHOH
CHCH2O
NHCO
(CH2)n
CH3
Ceramide
71
and even survival of the cerebellar neurons strongly depend on the cellular
expression of certain, even less abundant, ganglioside species.
Detailed and unambiguous compositional mapping and structural
elucidation of individual ganglioside components are therefore of crucial necessity
for systematic characterization of the brain region-specific ganglioside
compositions in health and disease [OP9, OP39]. Such a study is of major
importance for correlating the composition and structure specificity with the
functional specialization of the particular region [OP19, OP26, OP40, OP41] and
pathological state respectively [OP18, OP 21, OP24, OP25, OP42].
Efficient separation and detailed MS structural characterization of
gangliosides from biological sources are basic prerequisites for the further
developed research strategies tending to elucidate specific function of each
particular structure and to use it, accordingly, as therapeutic agents in treatment of
diseases and/or as specific diagnostic markers.
3.3.1. Analysis of gangliosides and glycolipids from normal tissues
Healthy central nervous system (CNS) contains the highest amount of gangliosides:
neuronal membranes hold at least several times higher concentrations of
gangliosides/glycosphingolipids than the extraneural cell types, highlighting their
special role at the CNS level [70]. Mapping of the gangliosides expressed in
different regions of normal human brain using classical approaches based on TLC,
immunochemical and immunohistochemical methods offered a low amount of
information because of the detection limitations of these methods and their low
throughput. Therefore, this study will present the efforts to develop, optimize and
implement novel and high performance ESI MS methodologies in CNS ganglioside
and glycosphingolipid research.
For a high sensitive detection and structural analysis of ganglioside species in
complex mixtures from biological material CE and ESI MS and MS/MS were first
chosen as a new alternative. An approach based on off-line CE for separation,
nanoESI QTOF MS for the detection of single molecular species and MS/MS using
low-energy CID for their identification by sequencing was introduced [OP39]. The
72
strategy was found suitable for investigations of the complex ganglioside mixtures
under optimized CE separation and their analysis by the nanoESI QTOF MS and
MS/MS operating in the negative ion mode for their detection and fragmentation.
The application to a complex mixture of gangliosides from bovine brain
demonstrated that ganglioside molecular species could be identified according to
their carbohydrate and lipid structural characteristics from their MS/MS clear-cut
fragment ion data.
a)
b)
Figure 3.3.2. Fully automated chip (-)nanoESI QTOF MS of the G20y mixture. Sample concentration 2–3 pmol/μl in MeOH; acquisition time 3 min; sampling cone potential 45–135V. a) m/z (700 –980). b) m/z (980 –2050) [OP41]
1000 1050 1100 1150 1200 1250 1300 1350 1400 1450 1500 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 20500
100
%
1073.96
988.40
1063.42
988.921018.99
1088.46
1544.61
1088.92
1099.43
1230.931099.93
1114.961230.43
1115.43
1145.37
1179.53
1231.42
1245.41
1253.56 1492.541335.90
1271.40 1382.60
1545.58
1572.61
1573.63
1858.571857.56
1574.61
1575.641716.281690.55
1784.48
1886.57
1887.59
1984.311888.612036.89
1074.46
x2x4
710 720 730 740 750 760 770 780 790 800 810 820 830 840 850 860 870 880 890 900 910 920 930 940 950 960 9700
100
%
917.94
917.44
708.74
718.10
718.41
734.96718.76
719.08
806.57
735.47 749.47771.98
767.28788.57
836.46
813.06 822.41
823.08
888.61836.97
885.56
850.47
838.54863.58
878.61
904.63
889.63
931.96
918.48
918.98
919.48
932.47
940.49
945.47
946.49
952.47954.97
a)
x6 x4x4x1.8
73
Moreover, by such an approach, a fragmentation process of a single precursor ion
was able to reveal the presence of structural isomers containing variations in the
attachment site of the sialic acid moiety.
As a part of the efforts upon the implementation of the fully automated chip-based
mass spectrometry in the field of the complex carbohydrate analysis, a general
methodology for screening of glycosphingolipids under optimized conditions in
terms of ionization, sensitivity, automated sequencing, speed of analysis and
limitation of sample consumption, was probed [OP41].
Table 3.3.2. Composition of single components in the G20y ganglioside mixture from gray matter of normal human cerebellum as detected by a fully automated (-) chip nanoESI QTOF MS [OP41]
Type of Molecular Ion
m/z (monoisotopic)
Assigned structure
Detected Calculated
[M+2Na-4H]2- [M-H]-
611.40 1179.57
611.35 1179.74
GM3 (d18:1/18:0)
[M-H]- 1382.60 1382.82 GM2 (d18:1/18:0)
[M-2H]2-
[M+Na-2H]- 734.96 1492.78
734.91 1492.81
GD3 (d18:1/18:0)
[M-2H]2- 748.99 748.93 GD3 (d18:1/20:0)
[M-H]- 1518.51 1518.85 GM1, nLM1 and /or LM1 (d18:0/16:0)
[M-2H]2-
[M-H]- 771.98 1544.61
771.93 1544.85
GM1, nLM1 and /or LM1 (d18:1/18:0)
[M-2H]2-
[M-H]- 786.00 1572.61
785.92 1572.85
GM1, nLM1 and /or LM1 (d18:1/20:0)
[M-2H]2- 836.46 836.45 GD2 (d18:1/18:0)
[M-2H]2- 850.47 850.47 GD2 (d18:1/20:0)
[M-2H]2-
[M+Na-3H]2-
[M-H]-
[M+Na-2H]-
917.44 928.45 1835.62 1857.56
917.48 928.47 1835.96 1857.95
GD1, nLD1 and /or LD1 (d18:1/18:0)
[M-2H]2- 926.44 926.48 GD1, nLD1 and /or LD1 (t18:0/18:0)
[M-2H]2- 924.44 924.49 GD1, nLD1 and /or LD1 (d18:1/19:0)
[M-2H]2-
[M+Na-3H]2-
[M-H]-
931.46 942.44 1885.60
931.49 942.48 1885.98
GD1, nLD1 and /or LD1 (d18:1/20:0)
74
[M-2H]2-
940.49 940.50 GD1, nLD1 and /or LD1
(t18:0/20:0)
[M-2H]2- 938.44 938.50 GD1, nLD1 and /or LD1 (d18:1/21:0)
[M-2H]2- 945.47 945.51 GD1, nLD1 and /or LD1 (d18:1/22:0)
[M-2H]2- 954.46 954.51 GD1, nLD1 and /or LD1 (t18:0/22:0)
[M-2H]2- 952.47 952.52 GD1, nLD1 and /or LD1 (d18:1/23:0)
[M-2H]2- 958.46* 958.52 GD1, nLD1 and /or LD1 (d18:1/24:1)
[M-2H]2- 966.44 966.53 GD1, nLD1 and /or LD1 (d18:1/25:0) or (d20:1/23:0)
[M-2H]2- 988.40 988.49 Fuc-GD1 (d18:1/18:2)
[M-2H]2- 990.40 990.51 Fuc-GD1 (d18:1/18:0)
[M-2H]2- 999.41* 999.51 Fuc-GD1 (t18:0/18:0)
[M-2H]2- 1002.41 1002.51 Fuc-GD1 (d18:1/20:2)
[M-2H]2- 1004.42 1004.52 Fuc-GD1 (d18:1/20:0)
[M-2H]2- 1013.44* 1013.53 Fuc-GD1 (t18:0/20:0)
[M-2H]2- 1018.99 1019.02 GalNAc-GD1 (d18:1/18:0)
[M-2H]2- 1032.93* 1033.03 GalNAc-GD1 (d18:1/20:0)
[M-3H]3-
[M-2H]2-
[M+Na-3H]2-
[M+2Na-4H]2-
708.39 1062.96 1073.92 1084.93
708.35 1063.03 1074.02 1085.01
GT1 (d18:1/18:0)
[M-3H]3-
[M+Na-3H]2- 714.41 1082.92
714.35 1083.02
GT1 (t18:0/18:0)
[M-3H]3- [M-2H]2-
[M+Na-3H]2-
[M+2Na-4H]2-
717.75 1076.97 1087.95 1098.92
717.69 1077.04 1088.03 1099.02
GT1 (d18:1/20:0)
[M-3H]3-
[M+Na-3H]2- 723.75 1096.93
723.70 1097.04
GT1 (t18:0/20:0)
[M+Na-3H]2- 1094.95* 1095.04 GT1 (d18:1/21:0)
[M-3H]3-
[M+Na-3H]2- 727.11 1101.92
727.04 1102.05
GT1 (d18:1/22:0)
[M+Na-3H]2- 1108.92* 1109.06 GT1 (d18:1/23:0)
[M+Na-3H]2- 1114.96 1115.06 GT1 (d18:1/24:1)
[M-3H]3- 722.39 722.35 O-Ac-GT1 (d18:1/18:0)
[M-3H]3- 731.74 731.70 O-Ac-GT1 (d18:1/20:0)
[M-2H]2- 1128.95 1129.05 Fuc-GT1 (d18:1/17:0)
[M-2H]2- 1144.89 1145.06 Fuc-GT1 (t18:0/18:0)
[M-2H]2- 1159.89 1159.08 Fuc-GT1 (t18:0/20:0)
75
[M-3H]3-
[M+Na-4H]3-
[M+2Na-4H]2-
[M+3Na-5H]2-
805.40 812.73 1230.43 1241.43
805.38 812.71 1230.56 1241.55
GQ1 (d18:1/18:0)
[M-3H]3-
[M+Na-4H]3-
[M+2Na-4H]2-
814.74 822.07 1244.42
814.72 822.05 1244.57
GQ1 (d18:1/20:0)
[M-3H]3-
[M+Na-4H]3- 819.38* 826.73*
819.38 826.71
O-Ac-GQ1 (d18:1/18:0)
Automated chipESI QTOF MS (Fig.3.3.2, Table 3.3.2) and CID MS/MS (Fig.3.3.3)
was optimized in the negative ion mode for characterization of a complex
ganglioside mixture from normal human cerebellar tissue to demonstrate its
general feasibility for ganglioside analysis [OP41], and its advantages in
comparison with capillary-based ESI MS and MS/MS [OP6].
Figure 3.3.3. Fully automated negative mode chip nanoESI QTOF auto MS/MS of the GT1 (d18:1/18:0) species detected as a triply charged ion at m/z 709.03; cone potential 135 V; collision energy 50 eV; acquisition time 1 min. Inset: GT1(d18:1/18:0) fragmentation scheme by auto MS/MS CID [OP41]
726
581
564888
1544-GalGlcCer
NeuAc
NeuAc
NeuAc
Gal GalNAc Gal Glc Cer
916(2-)
1253-NeuAc
364
290/308
470 835290/308
726
581
564888
1544-GalGlcCer
NeuAc
NeuAc
NeuAc
Gal GalNAc Gal Glc Cer
916(2-)
1253-NeuAc
364
290/308
470 835290/308
76
The sample investigated in this study was a native mixture of gangliosides (G20y)
extracted from the gray matter of a normal adult human (20 years of age)
cerebellum, without pathological signs according to morphoanatomical and
histopathological examination, originating from a healthy subject who died in a
traffic accident.
Figure 3.3.4. Thin polymer microchip ESI QTOF MS of the GT1 fraction at 3 kV ESI voltage and 100 V sampling cone potential. (a) m/z range: (650–1175); (b) zoom out of the m/z range: (1030–1125) [OP5]
77
The automated chip nanoESI QTOF MS approach optimized for ganglioside
analysis provided a new insight into the structural diversity of ganglioside
expression in human cerebellar gray matter and complex molecular architecture of
the species. It was found that, in comparison with capillary-based ESI MS, a higher
sensitivity and closer representation upon the mixture composition could be
achieved.
By chip nanoESI MS screening, 44 glycoforms expressing high heterogeneity in the
ceramide motifs, as well as biologically relevant peripheral modifications such as
O-acetylation and fucosylation have been identified. By combining the fully
automated chipESI MS infusion with automatic selection and fragmentation of the
precursor ion, a complete set of structural data and sequence ions (Fig.3.3.3) could
be obtained for polysialylated single ganglioside species GT1(d18:1/18:0) in a
native mixture of high complexity, within short analysis time and with drastically
reduced sample consumption.
To test the feasibility and advantages of the thin chip polymer-based microsprayer
system in combination with QTOF MS/MS concerning information that could be
provided by both MS and CID MS/MS, as well as to define the corresponding
appropriate conditions for the GSL molecular class detection and structural
characterization, a rather structurally complex polysialylated ganglioside fraction,
GT1, was chosen as the testing sample [OP5].
The analyzed GT1 ganglioside fraction, showing migration properties of GT1b
species in high performance thin-layer chromatography (HPTLC), was isolated
from the total native ganglioside mixture purified from a normal adult human
cerebrum (45 years of age). By this approach a reproducible compositional
mapping of eight molecular components in the GT1 fraction mixture was obtained
from both triply and doubly charged formed molecular ions related to gangliosides
containing a number of lipid variants Fig. 3.3.4, Table 3.3.3). Furthermore, the high
sequencing efficiency of the microchip ESI QTOF MS/MS (Fig. 3.3.5) resulted in
information-rich fragmentation pattern. This feature was of particular importance
for elucidating the presence of structural isomers or isobars as distinct species as in
78
many cases these species play a particular physiological and/or pathological role
and therefore they might have a specific diagnostic relevance.
A protocol for negative ion nanoelectrospray ionization Fourier transform ion
cyclotron resonance mass spectrometry (nanoESI FTICR MS) investigation of
complex biological mixtures consisting of sialylated or sulfated glycosphingolipids
expressing high heterogeneity in the ceramide portion was further developed
[OP43]. Defined instrumental and solubilizing solvent system conditions were
explored to promote ionization efficiency of GSLs, reduce the in-source
fragmentation and consequently to enhance the detection of intact molecules.
For mass spectrometric analysis of gangliosides/GSLs specific adjustment of the
experimental parameters for both ionization/detection and sequencing were
required.
Table 3.3.3. The compositional mapping of the purified native GT1 ganglioside fraction (exhibiting HPTLC migration properties of the GT1b species) separated from the total ganglioside mixture isolated from adult human brain tissue as detected by thin polymer microchip ESI QTOF MS [OP5]
m/z (monoisotopic)
Type of Detected Molecular Ion
Putative Structure
708.18 1062.69
1073.69 1053.70
[M-3H+]3-
[M-2H+]2-
[M+Na+-3H+]2-
[M+2H+]2--H2O
GT1 (d18:1/18:0)
712.85 1070.19 1081.18
[M-3H+]3-
[M-2H+]2-
[M+Na+-3H+]2-
GT1 (d18:1/19:0)
717.50 1076.70
1087.69 1067.69
[M-3H]3- [M-2H]2-
[M+Na-3H]2-
[M+2H+]2--H2O
GT1 (d18:1/20:0)
1094.20 L [M+Na+-3H+]2- GT1 (d18:1/21:1)
722.18
1084.20 1095.20
[M-3H+]3-
[M-2H]2-
[M+Na+-3H+]2-
GT1 (d18:1/21:0)
726.85 [M-3H+]3- GT1 (d18:1/22:0)
1108.20L [M+Na+-3H+]2- GT1 (d18:1/23:1)
1109.21 L [M+Na+-3H+]2- GT1 (d18:1/23:0)
d =dihydroxy sphingoid base; Llow intensity ions.
79
Figure 3.3.5. Thin polymer microchip ESI QTOF CID MS/MS of the triply charged ion at m/z 717.50 corresponding to GT1 (d18:1/20:0). ESI voltage 3 kV. Sampling cone potential 100 V. Collision energy (40-70) eV. Inset: GT1 (d18:1/20:0) fragmentation pathway by CID [OP5]
Using the novel optimized protocol by (-)nanoESI FTICR MS a reliable and even
more detailed compositional fingerprint of the same polysialylated ganglioside
mixture (GT1) isolated from human brain was obtained.
Further on, in this study fragmentation analysis by SORI-CID MS2 was introduced
for the first time for structural elucidation of polysialylated gangliosides [OP43]and
under well-defined conditions an informative fragmentation pattern of the
trisialylated ganglioside GT1 was obtained (Fig.3.3.6).
80
Figure 3.3.6. (-)NanoESI FTICR SORI-CID MS2 of the doubly charged ion at m/z 1077.043 corresponding to GT1 (d18:1/20:0). Sample concentration: 5 pmol/μL in MeOH. Capillary exit voltage: -150 V. Number of scans: 150. Inset: Isolation by correlated shots of the doubly charged precursor ion at m/z 1077.043 [OP43]
The compositional mapping by FTICR MS (Fig. 3.3.7) of a more complex mixture of
sulfated glucuronic acid containing neolacto-series GSLs extracted from bovine
cauda equina provided hard evidence upon the presence of components described
before, and moreover upon new structures, previously not identified by any other
analytical method. Two new, however metabolically expectable, species detected
within 10 ppm accuracy could be assigned to the sulfo-GlcA-nLc8Cer molecules
with d18:1/20:0 and d18:1/24:0 type of ceramide portion, respectively.
Negative ion nanoESI FTICR MS at 9.4 T was shown to represent a valuable
method for screening and sequencing of GSLs, allowing for a high resolution and
mass accuracy detection of major and minor GSL glycoforms and identification of
previously unknown biologically relevant structures.
81
Figure 3.3.7. (-)NanoESI FTICR MS of the sulfated GSL mixture extracted from bovine Cauda equina. Sample concentration: 5 pmol/μL in MeOH. Capillary exit voltage: -300 V. Number of scans: 200. Inset: Table listing the assignment of the major detected structures [OP43]
3.3.2 Analysis of ganglioside expression and structure in pathological tissues
Neurodevelopmental and neurodegenerative disorders
Gangliosides participate in induction or development of various neurodegenerative
and neurodevelopmental diseases. Some autoimmune- induced neuropathies are
probably directly caused by antiganglioside auto-antibodies produced due to a
high immunogenicity of gangliosides [71]. In some lipidoses [72, 73] a group of
inherited metabolic disorders, the accumulation of gangliosides occurs in cell
bodies due to a blockage of their catabolic and/or maybe even anabolic pathways.
Gangliosides, especially GM1, were shown to have neuritogenic and
neuronotrophic activity and to facilitate repair of neuronal tissue after mechanical,
biochemical or toxic injuries. Continuous intracerebroventricular infusion of GM1
82
was demonstrated [74] to have a significant beneficial effect in patients with an
early onset Alzheimer disease (AD) type I.
In anencephaly [75,76], a congenital malformation of the fetal brain occurring when
the cephalic end of the neural tube fails to close, the first assessment of ganglioside
composition was reported by Cacić [77] on the basis of the evidence obtained by
immunostaining on thin layer chromatograms. In anencephaly where the process
of cell differentiation and maturation is severely disturbed, a significant change in
ganglioside pattern characterized by a marked reduction in of GM1a, GM1b and
GD1a content and a better expression of neolactoseries gangliosides was found.
Later on, by development and introduction in glycolipidomics of advanced MS
methods based on ESI and nanoESI as methods complementary to TLC and
immunochemical analyses, better insights into the ganglioside altered composition
in neurodegenerative diseases was possible.
In 2001, Vukelić et al. [78] optimized and applied for the first time nanoESI QTOF
MS and tandem MS for compositional and structural identification of native
gangliosides from anencephalic cerebral residue and cerebellum. By this approach
it was found that the total ganglioside concentrations in the anencephalic cerebral
remnant and in cerebellum were significantly lower than in the corresponding
regions of the age-matched brain used as control. In the cerebral remnant, GD3,
GM2 and GT1b, GM1b nLM1 and nLD1 were found highly expressed. Oppositely,
GD1a was found better expressed in the anencephalic cerebellum, while GQ1b was
reduced in both anencephalic regions.
In agreement with previously acquired information by immunochemical methods,
by nanoESI MS, members of the neolacto-series gangliosides were also discovered
in anencephalic brain tissues.
In this context, in the present work additional data corroborating a significant
alteration of ganglioside expression in anencephalic vs. age-matched normal brain
tissue was recently collected by the newly developed methodology based on
coupling of NanoMate robot to multistage MS on the HCT MS instrument tuned in
the negative ion mode [OP18].
83
Figure 3.3.8. Fully automated (-) chip nanoESI HCT MS of An28 native ganglioside mixture from glial islands of anencephalic fetus. Solvent: MeOH; sample concentration 5 pmol/μL; acquisition time 10 min; Chip ESI: -0.8 kV; capillary exit: -50 V [OP18] A native ganglioside mixture purified from glial islands of 28 weeks fetal
anencephalic brain tissue (An28) was investigated in comparison with the
ganglioside extract from a 27 weeks normal fetal frontal lobe (FL27).
Under identical instrumental and solution conditions, 25 distinct species in the
mixture from anencephalic tissue (Fig. 3.3.8, Table 3.3.4) vs. 44 of which 4
asialylated in the normal tissue (Fig. 3.3.9, Table 3.3.5) were for the first time
identified. These results systematized in Table 3.3.6 indicated that a high number of
ganglioside species associated to anencephaly could be ionized and discriminated
only by employing chip-based electrospray. Interestingly, GD3 (d18:1/18:0), GD2
(d18:1/18:0), GM1 (d18:1/18:0) and their neolacto or lacto-series isomers were
detected as ions of similar low abundances in both mixtures, while GT1
(d18:1/18:0) and GD1 (d18:1/18:0) were found highly expressed in ancencephalic
brain tissue. Moreover, several structures such as GT1, GQ1 and GQ2 emerged
clearly as associated to anencephaly. This prominent incidence of polysialylated
structures in anencephaly was considered an effect, possibly to be used as a
diagnostic of brain development stagnation [79], which occurs in this disease.
In view of the results obtained by MS/MS, the earlier report [78] has postulated
that GT1b is one of the disease markers; however, because of the limited
735.53
836.68
917.60
1063.72
1207.01
1471.031572.02
0.0
0.5
1.0
1.5
2.0
4x10
Intens.
800 1000 1200 1400 1600 1800 2000 m/z
931.72
1049.26
1077.73
1139.01
1179.90
1237.90 1249.95 1279.88
1259.92
1353.03
1375.03
1519.06
1544.16
1553.07
1653.21
1671.11
1756.01
1757.51
1858.32
1885.08
1918.11
MS2
84
information obtained by fragmentation analysis in a single CID stage, validation of
sialylation sites could not be accomplished. To close this gap, a nanoESI chip CID
MSn protocol [OP18] for fine investigation of the anencephaly-specific GT1
(d18:1/18:0) species was elaborated (Fig. 3.3.10). The beneficial combination of chip
infusion, high capacity of ion storage and multistage sequencing rendered ions
consistent with Neu5Ac2 localization at inner Gal, which, for the first time,
corroborated GT1b presence in the cerebral remnant of anencephalic brain. The
highlighted accomplishments in characterization of novel structures in a severe
neurodevelopmental disorder indicate that advanced chip-based ESI MS has real
perspectives to become a routine method for early diagnosis and therapy based on
discovery of ganglioside molecular fingerprints.
Table 3.3.4. Assignment of the major ions detected in An28 mixture [OP18]
m/z monoisotopic
Molecular ion Proposed structure
735.53 [M-2H]2- GD3(d18:0/18:0)
836.68 [M-2H]2- GD2(d18:1/18:0)
917.60 [M-2H]2- GD1(d18:1/18:0)
931.72 [M-2H]2- GD1(18:1/20:0)
1049.26 [M-2H]2- GT1(18:1/16:0)
1063.72 [M-2H]2- GT1(d18:1/18:0)
1077.73 [M-2H]2- GT1(d18:1/20:0)
1139.01 [M-H]- GM3(d18:1/14:0) or (d18:1/h14:0) or HexNAcHex2Cer (d18:1/22:4)
1179.90 [M-H]- GM3 (d18:1/18:0)
1207.01 [M-H]- GM3(d18:1/20:0)
1237.90 [M-H]- GM3 (d18:0/22:0)
1249.95 [M-H]- O-Ac-GM3 (d18:1/20:0) (or GM3 (18:1/23:0)
1259.92 [M-H]- GM3 (d18:1/24:2)
1279.88 [M-H]- O-Ac-GM3 (d18:0/22:0) (or GM3 (20:0/23:0)
1353.03 [M-H]- GM2 (d18:1/16:0)
1354.79 [M-H]- GM2 (d18:1/16:0)
1471.03 [M-H]- GD3(d18:1/18:0)
1472.79 [M-H]- GD3 (d18:0/18:0)
1519.06 [M-H]- GM1, nLM1 and/or LM1 (d18:0/16:0)
1544.16 [M-H]- GM1, nLM1 and/or LM1 (d18:1/18:0)
1553.07 [M-H]- GD3(d18:1/24:1)
1572.02 [M-H]- GM1, nLM1 and/or LM1 (d18:0/20:0)
1845.01 [M-H]- GT3(d18:1/24:0)
1858.32 [M+Na-2H]- GD1(d18:0/18:0)
1885.08 [M-H]- GT3(d18:1/24:1)
2256.50 [M-H]- GQ2(d18:1/18:0)
2417.45 [M-H]- GQ1(d18:1/18:0)
85
Figure 3.3.9. Fully automated (-) chip nanoESI HCT MS of FL27 native ganglioside mixture from normal fetus frontal lobe. Solvent: MeOH; sample concentration 5 pmol/μl; acquisition time 7 min; Chip ESI: -0.8 kV; capillary exit: -50 V [OP18]
Table 3.3.5. Assignment of the major ions detected in FL27 mixture [OP18]
m/z
monoisotopic Molecular
ion Proposed structure
735.12 [M-2H]2- GD3(d18:0/18:0)
836.71 [M-2H]2- GD2(d18:1/18:0)
851.60 [M-H]- GD2(d18:1/20:0)
918.08 [M-2H]2- GD1(d18:1/18:0)
931.72 [M-2H]2- GD1(18:1/20:0)
952.80 [M-H]- GD1(d18:1/23:0)
1037.60 [M-H]- HexNAcHex2Cer (d18:0/14:0) or (d16:0/16:0)
1041.60 [M-H]- GM4 (d18:1/20:2)
1049.18 [M-2H]2- GT1(18:1/16:0)
1063.33 [M-2H]2- GT1(d18:1/18:0)
1065.63 [M-H]- HexNAcHex2Cer (d18:0/16:0)
1077.71 [M-2H]2- GT1(d18:1/20:0)
1104.78 [M-2H]2- GT1(d18:1/24:0)
1139.01 [M-H]- GM3(d18:1/14:0) or (d18:1/h14:0) or
735.52 836.71
917.58
1544.20
0.0
0.5
1.0
1.5
2.0
2.5
6x10
Intens.
800 1000 1200 1400 1600 1800 m/z
1836.20
-
0
2000
4000
6000
1700 1750 1800 1850 1900 1950 2000 2050
851.60
931.72
952.80
1037.60
1041.60
1049.18
1063.33
1065.63
1077.71
1104.78
1139.01
1151.71
1165.80
1167.82
1179.74
1181.75
1206.77
1221.33
1235.81
1383.21
1354.79
1301.82
1279.81
1259.79
1471.03
1444.80
1858.20
1990.50
86
HexNAcHex2Cer (d18:1/22:4)
1151.71 [M-H]- GM3 (d18:1/16:0)
1165.80 [M-H]- HexNAcHex2Cer (t18:0/22:0) or (d18:0/h22:0) or (d18:2/24:4)
1167.82 [M-H]- GM3 (t18:1/16:0) or (d18:1/h16:0) or HexNAcHex2Cer (d18:1/24:4)
1179.74 [M-H]- GM3 (d18:1/18:0)
1181.75 [M-H]- GM3 (d18:0/18:0)
1206.77 [M-H]- GM3(d18:1/20:0)
1221.33 [M-H]- GM3(d18:1/18:0)
1235.81 [M-H]- GM3 (d18:1/22:0)
1237.81 [M-H]- GM3 (d18:0/22:0)
1249.78 [M-H]- O-Ac-GM3 (d18:1/20:0) (or GM3 (18:1/23:0)
1253.02 [M-H]- HexHexNAcHex2Cer(d18:1/18:0)
1259.79 [M-H]- GM3 (d18:1/24:2)
1261.81 [M-H]- GM3 (d18:1/24:1)
1263.83 [M-H]- GM3 (d18:1/24:0)
1265.84 [M-H]- GM3 (d18:0/24:0)
1275.80 [M-H]- O-Ac-GM3 (d18:1/22:1) (or GM3 (20:1/23:1)
1277.80 [M-H]- O-Ac-GM3 (d18:1/22:0) (or GM3 (20:1/23:0)
1279.81 [M-H]- O-Ac-GM3 (d18:0/22:0) (or GM3 (20:0/23:0)
1301.82 [M-H]- O-Ac-GM3 (d18:1/24:2)
1353.03 [M-H]- GM2 (d18:1/16:0)
1354.79 [M-H]- GM2 (d18:1/16:0)
1383.21 [M-H]- GM2 (d18:1/18:0)
1384.81 [M-H]- GM2 (d18:0/18:0)
1437.01 [M-H]- GM2 (d18:1/22:0)
1442.78 [M-H]- GD3 (d18:1/16:0)
1444.80 [M-H]- GD3 (d18:0/16:0)
1468.79 [M-H]- GD3 (d18:1/18:1)
1471.03 [M-H]- GD3(d18:1/18:0)
1472.83 [M-H]- GD3 (d18:0/18:0)
1519.10
[M-H]- GM1, nLM1 and/or LM1 (d18:0/16:0)
1544.16
[M-H]- GM1, nLM1 and/or LM1 (d18:1/18:0)
1545.20 [M-H]- GM1, nLM1 and/or LM1 (d18:1/18:0)
1805.23 [M-H]- GT3(d18:1/18:0)
1836.40 [M-H]- GD1(d18:1/18:0)
d- dihydroxy sphingoid base; t-trihydroxy sphingoid base
87
Table 3.3.6. Comparative overview upon gangliosides and asialo-gangliosides detected in An28 and FL27 mixtures[OP18]
GG species Proposed structure An28 FL27
GM1 nLM1 and/or LM1 (d18:0/16:0) + +
nLM1 and/or LM1 (d18:1/18:0) + +
nLM1 and/or LM1 (d18:0/20:0) + -
(d18:1/16:0) + +
(d18:1/18:0) - +
(d18:1/22:0) - +
GM3 (d18:1/14:0) or (d18:1/h14:0) or HexNAcHex2Cer (d18:1/22:4)
+ +
(d18:1/16:0) - +
(t18:1/16:0) or (d18:1/h16:0) or HexNAcHex2Cer(d18:1/24:4)
- +
(d18:1/18:0) + +
(d18:0/18:0) - +
(d18:1/20:0) + +
(d18:1/22:0) - +
(d18:0/22:0) + +
(d18:1/24:2) + +
(d18:0/24:0) + +
(d18:1/24:1) - +
(d18:1/24:0) - +
O-Ac-GM3 (d18:1/20:0) or GM3 (18:1/23:0)
- +
O-Ac-GM3 (d18:1/22:1) or GM3 (20:1/23:1)
- +
O-Ac-GM3 (d18:1/22:0) (or GM3 (20:1/23:0)
- +
O-Ac-GM3 (d18:0/22:0) (or GM3 (20:0/23:0)
+ +
O-Ac-GM3 (d18:1/24:2) - +
GM4 (d18:1/20:2) - +
GD1 (d18:1/18:0) + +
(d18:1/20:0) + +
(d18:1/23:0) - +
(d18:1/24:1) + -
88
(d18:0/18:0) + -
GD2
(d18:1/18:0) + +
(d18:1/18:1) + -
(d18:1/24:1) + -
(d18:1/24:0) + -
(d18:1/20:0) - +
GD3 (d18:0/18:0) + +
(d18:1/16:0) - +
(d18:0/16:0) - +
(d18:1/18:1) - +
(d18:1/18:0) + +
(d18:1/24:1) + -
GT1 (18:1/16:0) + +
(d18:1/18:0) + +
(d18:0/20:0) + -
(d18:1/20:0) + +
(d18:1/24:0) - +
GT3 (d18:1/18:0) - +
(d18:1/24:0) - +
(d18:1/24:1) + -
GQ1 (d18:1/18:0) + -
Asialo-GG species
HexNAcHex2Cer (d18:0/14:0) or (d16:0/16:0)
- +
HexNAcHex2Cer (d18:0/16:0) - +
HexNAcHex2Cer (t18:0/22:0) or (d18:0/h22:0) or (d18:2/24:4)
- +
HexHexNAcHex2Cer(d18:1/18:0) - +
d- dihydroxy sphingoid base; t-trihydroxy sphingoid base + the structure was detected - the structure was not detected
89
a)
b)
c) d)
Figure 3.3.10. Fully automated (-) chip nanoESI HCT multistage MS (CID MS2-MS4) of the doubly charged ion at m/z 1063.34 corresponding to GT1 (d18:1/18:0) ganglioside species detected in An28 mixture. a) MS4 stage using as the precursor the Y4β- ion detected at m/z 1544.87 in MS3; b) fragmentation scheme in MS2 of the [M-2H]2- ion at m/z 1063.34; c) fragmentation scheme in MS3 of ion at m/z 917.32; d) fragmentation scheme in MS4 of the ion at m/z 1544.87 [OP18]
0
5
10
15
Intens.
600 700 800 900 1000 1100 1200 1300 1400 1500 m/z
0
5
10
15
Intens.
600 700 800 900 1000 1100 1200 1300 1400 1500 m/z
564.62
888.42Y2α/B1β
870.42
Y2α/C1β
1253.81
Y3β
1544.87
[M-H]-
1526.87
[M-H]--H2O
980.32
707.63
1024.45
1212.61
1389.01
1375.83
Y0
Z1
Y1
726.22
B4
1179.48
Y2α
1346.50
1364.56
Z3α
Z3α-H2O
1501.82
[M-H]--CO2
1090.80
Y3α/B1β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
B4
Z0Y0
Y2α
Y3β
Y2α/B1β or
Y2α/C1β or
Fig3f
Z1Y1
C1β
B1β
Y3β /B2α
Z3β /B2αB2α
Z3αY3α
Y3α/B1β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
B4
Z0Y0
Y2α
Y3β
Y2α/B1β or
Y2α/C1β or
Fig3f
Z1Y1
C1β
B1β
Y3β /B2α
Z3β /B2αB2α
Z3αY3α
Y3α/B1β
Y2α/B1β
O
NeuAc
O
NeuAc
O
NeuAc
O
NeuAc
NeuAc – O – Gal – O – GalNAc – O – Gal – O – Glc – O – Cer
B1α C1α
Z4αY4α
Z4β
Y4βB2β
C2β
Y4β /B1α
Y2α /B2β
B1β
Fig3b
Y4α/B2β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
Y0
Z3β
Y3β
Z4β
Y4β
B1β
C1β
B2β
C2βZ3β/C2α or Z2α/B2β
Z4β/B2α or Y2α/C1β
Y4β/B1α or Y3α/B1β
Z3α/C2β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
O
NeuAc
O
NeuAc
Y0
Z3β
Y3β
Z4β
Y4β
B1β
C1β
B2β
C2βZ3β/C2α or Z2α/B2β
Z4β/B2α or Y2α/C1β
Y4β/B1α or Y3α/B1β
Z3α/C2β
Z2α/B2β
Z2α/C1β
Y3α/B1β
Z3α/C2β
Y2α/B2β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
Y0
Z3β
Y3β
Z4β
Y4β
B1β
C1β
B2β
C2βZ3β/C2α or Z2α/B2β
Z4β/B2α or Y2α/C1β
Y4β/B1α or Y3α/B1β
Z3α/C2β
Gal – O – GalNAc – O – Gal – O – Glc – O - Cer
O
NeuAc
O
NeuAc
O
NeuAc
O
NeuAc
Y0
Z3β
Y3β
Z4β
Y4β
B1β
C1β
B2β
C2βZ3β/C2α or Z2α/B2β
Z4β/B2α or Y2α/C1β
Y4β/B1α or Y3α/B1β
Z3α/C2β
Z2α/B2β
Z2α/C1β
Y3α/B1β
Z3α/C2β
Y2α/B2β
90
Primary brain tumors Tumorigenesis/malignant transformation is accompanied by aberrant cell surface
composition, particularly due to irregularities in glycoconjugate glycosylation
pathways. Various glycosyl epitopes constitute tumor-associated antigens [80,81].
Some of them promote invasion and metastases, while some other suppress tumor
progression [82].
Gangliosides are among the molecules bearing characteristic glycosyl epitopes
causing such effects. Glycosphingolipid-dependent cross-talk between
glycosynapses interfacing tumor cells with their host cells has been even
recognized as a basis to define tumor malignancy [83]. Structural elucidation of
individual ganglioside components in normal human brain as well as their spatial-
temporal distribution was an essential requirement for investigation of primary
brain tumors gangliosides. Specific changes of ganglioside pattern in brain tumors
vs. normal brain, correlating with tumor histopathological origin, malignancy
grade, invasiveness and progression have been observed [84]. A decrease in the
regular ganglioside profile and an increase in the structures detected only in small
amounts in normal brain tissue was found in primary brain tumors [85-86],
demonstrating a direct correlation between ganglioside composition and
histological type and grade of the tumors and an option to use this feature as
biochemical marker in early histopathological diagnosis, grading and prognosis of
tumors.
Glycoantigens and lipoantigens have been recognized as relevant and potentially
valuable diagnostic and prognostic markers and tumor molecular targets for
development/production of specific anti-tumor drugs, such as GSL-based vaccines,
but their investigation in this regard has been neglected comparing to proteins [87].
In the last years several biophysical methods have been developed for the
investigation of ganglioside expression in severe brain tumors. Ganglioside
profiling, their quantification and correlation to histomorphology and grading of
human gliomas has been studied [88] using a newly developed microbore HPLC
method. The use of infrared (IR) spectroscopy as an adjunct to histopathology in
detecting and diagnosing human brain tumors was also demonstrated [89]. In
91
another study [90] ganglioside expression in human glioblastoma was determined
by confocal microscopy of immunostained brain sections using antiganglioside
monoclonal antibodies. However, a large number of low abundant tumor-
associated species could not be detected by these conventional analytical methods.
Systematic studies of ganglioside composition in human brain tumors are still
restricted to several major components and many less abundant species with
possible biomarker values could not be structurally characterized.
Figure 3.3.11. Negative ion mode chip nanoESI QTOF MS of the native gliosarcoma ganglioside mixture. ESI voltage, 1.60 kV; sampling cone, 80 V; acquisition, 2 min; average sample consumption, 0.5 pmol [OP42]
This emphasized the need for detailed and systematic screening and structural
characterization of brain tumor glycoconjugate composition, which could
adequately be achieved only combining up-to-date, ultra-sensitive, high-resolution
methodological approaches of detection and sequencing of biomolecules, such as
92
advanced MS methods based on chip nanoESI sometimes complemented by
immunochemical and chromatographic techniques.
The first chip-based ESI MS method for ganglioside analysis from human brain
malignant alterations was introduced during the present work in 2007 [OP42]. The
ganglioside composition and structure were characterized for human brain
gliosarcoma obtained during surgical procedure, using the combination of
NanoMate robot and QTOF MS. Five microliter aliquots of the ganglioside mixture
working sample solutions were loaded and submitted for MS screening in negative
ion mode detection (Fig. 3.3.11). By chip nanoESI QTOF MS more than 25 species
dominated by GD3 and a high abundance of O-acetylated GD3 species could be
observed.
Figure 3.3.12. Negative ion mode chip nanoESI QTOF CID MS/MS of the [M-H]- ion at m/z 1540.96 corresponding to the O-Ac-GD3 (d18:1/20:0). ESI voltage, 1000–1250 V; for precursor ions isolation the LM and HM parameters were set to 3; collision energy: 25–40 eV; collision gas pressure: 5–10 psi.; acquisition time, 11 min; average sample consumption, 3.5 pmol. Inset, the fragmentation of O-Ac-GD3 [OP42]
93
High intensity ions corresponding to GM3 and GD2 species carrying different
ceramides were present as well. Several considerably abundant ions related to
GM2, GM1, and/or their isomers nLM1 and LM1, as well as to GD1 species
characterized by heterogeneity in composition of their ceramide moieties, were
found.
To provide a consistent structural identification, in the same study [OP42] several
detected species were subjected to fine analysis by tandem MS. Sequencing data
defined the composition and detailed structure of several gliosarcoma-associated
species among which GD3 (d18:1/24:1) OAc-GD3 (d18:1/20:0) GD2 (d18:1/18:0),
GM1a (d18:1/18:0), GM1b, nLM1 or LM1 (d18:1/18:0). A particular attention was
paid to O-Ac-GD3 molecule because this ganglioside could by itself be responsible
for the protection of tumor cells from apoptosis. Its sequencing pattern offered the
structural support to postulate a novel O-Ac-GD3 isomer, O-acetylated at the inner
Neu5Ac-residue. In Fig. 3.3.12, the Y3- ion at m/z 1249.84 represents the evidence for
Ac-O-Neu5Ac-Gal-Glcsequence carrying d18:1/20:0 ceramide. This feature
confirmed that in gliosarcoma O-acetylation of GD3 occurs at the inner Neu5Ac
residue.
Two years later, the research continued with the investigation of ganglioside
composition and structure in human brain hemangioma, a benign tumor, using
advanced mass spectrometry methods based on NanoMate HCT and CID MSn
[OP21]. The obtained mass spectrum revealed 29 different ganglioside species
dominated by mono- and disialylated structures. Two acetylated species, O-Ac-
GM4 (d18:0/29:0) and O-Ac-GD2 (d18:1/23:0), the last one correlated with the
reduced malignancy grade of the cerebral tumor were discovered. For fine
structural analysis of the unusual, hemangioma-asociated GT1, CID MS2 at variable
RF signal amplitudes within 0.6–1.0 V was applied. Five different fragment ions
supported a structure of GT1c-type bearing (d18:0/20:0) ceramide. To confirm this
assignment, the ion corresponding to GD1b (d18:0/20:0) was submitted to CID MS2
under identical conditions. It was found that GD1b structure has the same lipid
constitution as the previously sequenced GT1 however, an oligosaccharide core
lacking one Neu5Ac residue. The NanoMate-based system developed and
94
optimized for determination of ganglioside expression and structure in human
brain hemangioma was able to detect an elevated number of species and, most
importantly, to correlate the presence of O-Ac-GD2 with the low malignancy grade
of the investigated cerebral tumor.
In 2012 we have developed a strategy combining HPTLC, laser densitometry and
fully automated chip-based nanoelectrospray performed on a NanoMate robot
coupled to QTOF MS for mapping and structural identification of gangliosides
extracted and purified from human angioblastic meningioma [OP25]. While
HPTLC pattern indicated only six fractions migrating as GM3, GM2, GM1, GD3,
GD1a (nLD1, LD1), GD1b, and possibly GD2, due to the high sensitivity, mass
accuracy and ability to ionize minor species in complex mixtures, nanoESIchip-
QTOF MS was able to discover 34 distinct components of which two asialo, four
GM4, nine GM3, two GM2, two GD3, nine GM1 and six GD1 differing in their
ceramide compositions. All structures presented long-chain bases with 18 carbon
atoms, while the length of the fatty acid was found to vary from C11 to C25. MS
screening results indicated also that the diversity of the expressed GM1 structures
is higher than expected in view of the low proportions evidenced by densitometric
quantification. Simultaneous fragmentation analysis of meningioma-associated
GM1 (d18:1/24:1) and GM1 (d18:1/24:2) by MS/MS using CID confirmed the
structure of the ceramide moieties and provide data on the glycan core, which
document for the first time that both GM1a and GM1b isomers are expressed in
meningioma tissue.
Brain metastases
In 2011 this research was oriented towards the first optimization and
application of chip-based nanoelectrospray (NanoMate robot) MS for the
investigation of gangliosides in secondary brain tumors [OP24]. A native
ganglioside mixture extracted and purified from brain metastasis of lung
adenocarcinoma (male patient 73-y-old) was screened by NanoMate robot coupled
to a quadrupole time-of-flight MS (Fig.3.3.13) vs. a native ganglioside mixture from
an age matched healthy brain tissue (subject deceased in a traffic accident),
95
a)
b)
Figure 3.2.13 (-) Chip-nanoESI QTOF MS of the native ganglioside mixture isolated from brain metastasis of lung adenocarcinoma. Solvent: MeOH; sample concentration 2.5 pmol/μL; acquisition time 1min; Chip ESI: 1.5 kV; Cone voltage: 45 V. Zoomed area: a) m/z (800-1350); b) m/z (1400-2100) [OP24]
800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350m/z0
100
%
800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350m/z0
100
%
x2 1166.00834.37
806.40
808.44
832.40
1150.07
862.40
835.42
860.39
884.33
863.40
885.35
886.33
933.23
949.23
1167.05
1168.05
1250.02
1248.03
1194.05
1234.02
1195.03
1278.01
1279.02
850.38
822.38
875.37
901.33
1122.03
1138.05
947.16
963.20
965.21
981.14
983.17
856.36
1178.04
1180.08
1182.01
1184.03
1222.02
1206.03
1260.03
1262.03
1264.00
1275.99
1289.94
1291.92
1293.91
1295.94
1400 1450 1500 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 2050m/z0
100
%
1400 1450 1500 1550 1600 1650 1700 1750 1800 1850 1900 1950 2000 2050m/z0
100
%
1514.63
1421.68
1405.71
1434.71
1440.69
1491.73
1626.56
1624.57
1515.64
1598.59
1570.59
1516.69
1542.60
1627.56
1676.43
1660.48
1628.56 1678.41
1786.37
1711.44
1767.38
1749.37 1879.15
1787.32
1861.24
1834.28
1991.03
1880.17
1989.04
1959.14
1911.01
2005.13
2019.17
2050.96
1505.67
1479.62
1469.28
1528.63
1613.54
96
sampled and analyzed under identical conditions.
This comparative assay highlighted a considerable difference in the number and
type of ganglioside components expressed in brain metastasis (Table 3.3.7) vs.
healthy brain tissue (Table 3.3.8).
Table 3.3.7. Ganglioside and asialo-ganglioside species from brain metastasis of lung adenocarcinoma detected by (-) chip nanoESI QTOF MS analysis of complex native ganglioside mixture [OP24] m/z (monoisotopic) theoretical
m/z (monoisotopic) experimental
Mass accuracy (ppm)
Molecular ion
Proposed structure
875.19 874.91 33 [M-H]- LacCer(d18:1/17:0)
933.31 932.99 35 [M-H]- LacCer(d18:0/21:0)
947.34 947.19 16 [M-H]- LacCer(d18:0/22:0)
949.22 949.24 21 [M+2Na-3H]- LacCer(d18:0/19:0)
964.24 963.90 35 [M-H]- GM4(d18:0/14:0)
982.19 981.94 25 [M+Na-2H]- GM4(d18:1/14:1)
984.21 983.87 34 [M+Na-2H]-
GM4(d18:1/14:0) or GM4(d18:0/14:1)
1122.48 1122.23 22 [M-H]- GA2(d18:0/20:0)
1138.44 1138.48
1138.15 25 29
[M-H]-
Fuc-GM4(d18:0/16:0) GA2(t18:0/20:0)
1150.49 1150.40
1150.17 28 20
[M-H]-
[M-H]- GA2(d18:1/21:0) GM3(d18:1/16:1)
1168.42 1168.01 35 [M-H]- GM3(t18:0/16:0)
1178.46 1178.14 27 [M-H]- GM3(d18:1/18:1)
1179.74 1180.10 30 [M-H]- GM3(d18:1/18:0)
1182.49 1182.21 24 [M-H]- GM3(d18:0/18:0)
1184.37 1184.08 24 [M-H]- O-Ac-GA1(d18:1/10:0)
1194.50 1194.15 29 [M-H]-
GM3(d18:1/19:0) or GM3(d18:0/19:1)
1206.51 1206.64
1206.33 15 26
[M-H]-
[M-H]- GM3(d18:1/20:1) GA2(d18:0/26:0)
1222.51 1222.55
1222.19 26 29
[M-H]-
[M-H]- O-Ac-GM3(d18:1/18:0) GM3(d18:0/21:1) or GM3(d18:1/21:0)
1234.56 1234.52
1234.22 27 24
[M-H]-
GM3(d18:1/22:1) O-Ac- GM3(d18:1/19:1)
1248.55 1248.59
1248.18 33 30
[M-H]-
O-Ac-GM3(d18:1/20:1) GM3(d18:1/23:1)
1248.59 1249.02 34 [M-H]- GM3(d18:1/23:0)
1260.60 1260.33 21 [M-H]- GM3(d18:1/24:2)
1262.62 1262.35 21 [M-H]- GM3(d18:1/24:1)
97
1264.63 1264.19 35 [M-H]- GM3(d18:1/24:0)
1276.61 1276.64 1276.69
1277.01 31 29 25
[M-H]-
[M-H]-(-H2O)
O-Ac-GM3(d18:1/22:0) GM3(d20:1/23:1) GM3(d18:0/26:0)
1278.66 1278.61
1278.21 35 31
[M-H]-
[M-H]- GM3(d20:1/23:0) O-Ac-GM3(d18:1/22:0) or O-Ac-GM3(d18:0/22:1)
1288.67 1288.67
1289. 04 29 29
[M-H]-
[M-H]-
GM3(d18:1/26:2) or GM3(d18:2/26:1) GM3(d20:1/24:2)
1292.68 1292.23 35 [M-H]- GM3(d18:1/26:0) or GM3(d18:0/26:1)
1296.54 1296.59 1296.61
1296.24 23 27 28
[M-H]-
[M-H]-
[M-H]-
Fuc-GM3(d18:1/16:1) O-Ac-GA1(d18:1/18:0) GA1(d18:0/21:0) or GA1(d18:0/21:0)
1405.65 1405.21 31 [M+Na-2H]- GM2(d18:1/18:0)
1420.68 1420.80 8 [M-H]- O-Ac-GM2(d18:2/18:2)
1435.59
1435.21 26 [M+Na-2H]-
GD3(d18:1/14:1) or GD3(d18:0/14:2) or GD3(d18:2/14:0)
1441.66 1441.19 33 [M-H]- GD3(d18:1/16:1) or GD3(d18:0/16:2) or GD3(d18:2/16:0)
1471.73 1471.28 31 [M-H]- GD3(d18:1/18:0)
1493.71 1493.23 32 [M+Na-2H]- GD3(d18:1/18:0)
1515.69 1515.74 1515.78
1515.29
26 30 32
[M+2Na-3H]-
[M-H]- [M-H]-
GD3(d18:1/18:0) or GD3(d18:0/18:1) GM1(d18:1/16:1) or GM1(d18:0/16:2) or GM1(d18:2/16:0) O-Ac-GD3(d18:0/18:0)
1515.71 1515.75
1516.01 20 17
[M+Na-2H]-
[M-H]-
GD3(d18:1/20:2) or GD3(d18:0/20:3) or GD3(d18:2/20:1) GM1(d18:2/16:0) or GM1(d18:1/16:1) GM1(d18:0/16:2
1527.83 1528. 16 22 [M-H]- GD3(d18:0/22:0)
1541.79 1542. 19 26 [M-H]- GM1(d18:1/18:2) or GM1(d18:2/18:1) or GM1(d18:0/18:3)
1569.78 1569.83
1570.29 32 29
[M+2Na-3H]-
[M-H]-
GD3(d18:1/22:0) or GD3(d18:0/22:1) GM1(d18:1/20:1) or GM1(d18:0/20:2) or
98
1569.77
33
[M+Na-2H]-
GM1(d18:2/20:0) GD3(d18:0/24:2) or GD3(d18:1/24:1) or GD3(d18:2/24:0)
1597.88 1598.09 13 [M-H]- GM1(d18:0/22:2) or GM1(d18:1/22:1) or GM1(d18:2/22:0)
1611.77 1612.17 25 [M+2Na-3H]- GM1(d18:1/20:2)
1625.89 1624.92
1625.40
30 30
[M+2Na-3H]- [M-H]-
GD3(d18:1/26:1) or GD3(d18:0/26:2) or GD3(d18:2/26:0) GM1(d18:1/24:2)
1627.90 1626.93
1627.41 30 29
[M+2Na-3H]-
[M-H]-
GD3(d18:0/26:1) or GD3(d18:1/26:0) GM1(d18:0/24:2) or GM1(d18:1/24:1) or GM1(d18:2/24:0)
1629.92 1628.94
1629.42 31 29
[M-H]- [M-H]-
GM1(d18:0/24:1) or GM1(d18:1/24:0) di-O-Ac-GM1(d18:1/18:0)
1659.79 1660.18 23 [M+3Na-4H]- GM1(d18:1/22:3) or GM1(d18:0/22:4) or GM1(d18:2/22:2)
1674.87 1675.23 21 [M+Na-2H]-
(-H2O) GD2 (d18:1/18:2)
1748.97 1749.39 24 [M+Na-2H]- GD2 (d18:1/22:1)
1766.97 1767.28 18 [M-H]- (-H2O) GT3 (d18:1/20:1)
1785.07
1785.37 17 [M-H]-
O-Ac-GD2(d18:1/23:0) or O-Ac-GD2(d18:0/23:1)
1833.81 1833.07
1833.28 29 11
[M-H]- [M-H]-
GT3(d18:0/23:0) O-Ac-GT3 (d18:0/20:0)
1861.12 1861.12
1861.24 6 6
[M-H]-
[M-H] (-H2O) O-Ac-GT3-lactone(d18:0/22:0) O-Ac-GT3(d18:0/22:0)
1879.09 1879.10 1879.99
1879.39 16 15 32
[M+Na-2H]- [M-H]-(-H2O) [M-H]-
O-Ac-GT3 (d18:2/22:1) Fuc-GT3(d18:0/17:0) GT2(d18:1/12:1) or GT2(d18:2/12:0)
1909.16 1909.03 7 [M-H]- GD1 (d18:1/22:0)
1960.21 1960.12
1959.84 19 14
[M-H]-(-2H2O) [M-H]-
GT2(d18:0/20:0) GT2(d18:1/18:3) or GT2(d18:2/18:2)
1990.17 1990.19
1989.78 20 21
[M+Na-2H]- [M-H]-
GT2(d18:0/18:0) GT2(d18:0/20:3) or
99
GT2(d18:1/20:2) or GT2(d18:2/20:1)
1990.19 1990.83 32 [M-H]-
GT2(d18:1/20:1) or GT2(d18:0/20:2) or GT2(d18:2/20:0)
2005.20 2006.19
2005.63 21 28
[M-H]- Fuc-GD1(d18:1/20:2) O-Ac-GT2(d18:1/18:1)
2048.23 2048.10
2048.80 28 34
[M-H]-
[M-H]-(-H2O)
di-O-Ac-GT2(d18:0/18:0) GT1(d18:2/14:2) or GT1(d18:3/14:1)
Table 3.3.8. Ganglioside and asialo-ganglioside species from healthy brain detected by (-) chip nanoESI QTOF MS analysis of complex native ganglioside mixture [OP24]
m/z (monoisotopic) theoretical
m/z (monoisotopic) experimental
Mass accuracy (ppm)
Molecular ion Proposed structure
708.35 708.38 4 [M-3H]3- GT1(d18:1/18:0)
714.42 714.40 3 [M-3H]3-
[M+Na-4H]3- GT1(t18:0/18:0) GT1(d18:1/18:2)
717.58 717.54 5 [M-3H]3- GT1(d18:1/20:0) or GT1(d18:0/20:1)
734.91 735.12 29 [M-2H]2- GD3(d18:1/18:0) or GD3(d18:0/18:1)
756.38 756.30 11 [M-2H]2- O-Ac-GD3(d18:1/18:0)
771.95 771.93 3 [M-2H]2- GM1(d18:0/18:1) or GM1(d18:1/18:0)
822.05 822.06 1 [M+Na-4H]3- GQ1(d18:1/20:0) or GQ1(d18:0/20:1)
835.95 835.69 31 [M-2H] 2- GD2(d18:1/18:1) or GD2(d18:0/18:2) or GD2(d18:2/18:0)
844.96 844.69 32 [M-2H]2- O-Ac-GD2(d18:0/16:0)
850.47 850.22 29 [M-2H]2- GD2 (d18:1/20:0)
863.51 863.00 862.99
863.21 35 24 26
[M-2H]2-
[M-2H]2-
[M+Na-3H]2-
Fuc-GM1(d18:1/22:2) or Fuc-GM1(d18:0/22:3) or Fuc-GM1(d18:2/22:1) GD2(d18:0/22:3) or GD2(d18:1/22:2) or GD2(d18:2/22:1) GD2(d18:0/20:0)
878.03 877.88 17 [M-2H]2 – GD2(d18:1/24:1)
886.03 885.78 28 [M-2H]2 - O-Ac-
100
GD2(d18:1/22:0) or O-Ac-GD2(d18:0/22:1)
890.97 890.76 24 [M+Na-3H]2- GT3(d18:1/18:1)
905.01 905.11 11 [M+Na-3H]2- GT3(d18:1/20:0)
917.48 917.44 4 [M-2H]2- GD1(d18:1/18:0) or GD1(d18:0/18:1)
924.49 924.53
924.76 29 25
[M-2H]2 –
[M+2Na-4H]2- GD1(d18:1/19:0) or O-Ac-GT3(t18:1/20:0) or O-Ac-GT3(d18:0/20:1)
931.49 931.45 4 [M-2H]2- GD1(d18:1/20:0) or GD1(d18:0/20:1)
940.50 940.19
940.46
4 29
[M-2H]2 –
[M+2Na-4H]2-
GD1(t18:0/20:0) GD1 (d18:1/18:0) or GD1(d18:0/18:1)
945.51 945.50 1 [M-2H]2- GD1(d18:1/22:0)
952.52 952.50 2 [M-2H]2 - O-Ac-GD1 (d18:1/20:0) or O-Ac-GD1 (d18:0/20:1)
968.03 968.34 32 [M-3H]3- GH2(d18:1/24:0) or GH2(d18:0/24:1)
991.56 991.27 29 [M+Na-3H]2- GT2(d18:1/18:2) or GT2(d18:0/18:3) or GT2(d18:2/18:3)
1005.58 1005.28 30 [M+2Na-4H]2- GT2 (d18:0/18:0)
1019.02 1019.36 33 [M-2H]2– GalNAc-GD1(d18:0/18:0)
1024.62 1024.57
1024.68 6 11
[M-2H]2–
[M+2Na-4H]2- di-O-Ac-GT2 (d18:1/18:0) O-Ac-GT2 (d18:1/18:1)
1034.24
1033.94 29 [M+2Na-3H]-
GM4 (d18:1/16:0) or GM4 (d18:0/16:1)
1042.60 1042.66
1042.51 9 14
[M-2H]2 GT1 (t18:1/14:1) or GalNAc-GD1(t18:0/20:0)
1046.59 1046.46 12 [M+Na -3H]2- GT1(d18:1/14:0) or GT1(d18:0/14:1)
1049.62 1049.51 10 [M-2H]2- GT1(d18:1/16:0) or GT1(d18:0/16:1)
1059.61 1059.28 31 [M+Na-3H]2- GT1(d18:1/16:1)
1063.03 1063. 35 30 [M-2H]2- GT1(d18:1/18:0) or GT1(d18:0/18:1)
1074.02 1074.05 3 [M+Na-3H]2- GT1(d18:1/18:0)
1077.04 1077.37 31 [M-2H]2- GT1\(d18:1/20:0)
1097.18
1096.81 34
M-2H]2-
O-Ac-GT1(d18:1/20:1) or O-Ac-GT1(d18:0/20:2)
101
1097.04 21 [M+Na-3H]2- or O-Ac-GT1(d18:2/20:0) GT1(t18:0/20:0)
1110.70 1110.36 31 [M-2H]2- O-Ac-GT1(d18:1/22:2)
1118.49 1118.56 6 [M-H]- GM4(t18:1/24:0)
1180.47 1180.09 32 [M-H]- GM3(d18:1/18:0) or GM3(d18:0/18:1)
1228.51 1228.49
1228.61 8 10
[M-H]-
[M+Na-2H] - GA1(d18:0/16:0) GM3(d18:1/20:1) or GM3(d18:0/20:2)
1232.55 1232.52
1232.12 35 32
[M-H]-
[M+Na-2H] -
GM3(d18:1/22:2) or GM3(d18:0/22:3) or GM3(d18:2/22:1) GM3(d18:0/20:0)
1241.29 1240.86 35 [M+Na-3H]2- O-Ac-GQ1(d18:1/18:0)
1252.58 1252.60
1252.19 31 33
[M-H]-
O-Ac-GM3(d18:0/20:0) GM3 (d18:0/23:0)
1264.54 1264.12 33 [M-H]- di-O-Ac-GM3(d18:1/18:0)
1284.60 1284.36 19 [M+Na-2H] - GM3(d18:1/24:1)
1382.82 1382.87 4 [M-H]- GM2(d18:1/18:0) or GM2(d18:0/18:1)
1409.70 1410.19 35 [M-H]-
GM2(d18:1/20:0) or GM2(d18:0/20:1)
1467.69 1467.67
1467.86 12 13
[M-H]-
[M+Na-2H]- GD3(d18:1/18:2) GD3(d18:0/16:0)
1492.81 1492.89 5 [M+Na-2H]- GD3(d18:1/18:0)
1509.73 1509.79
1509.71 1 5
[M-H]-
O-Ac-GD3(d18:1/18:1) Fuc-GM2-lactone (d18:1/18:1)
1513.76 1513.29 31 [M-H]- O-Ac-GD3(d18:1/18:0)
1516.84 1517.71 1517.70
1517.30 30 27 26
[M-H]-
[M+Na-2H]-
[M+Na-2H]-
GM1(d18:1/16:0) GD3(d18:2/20:2) Fuc-GM2(d18:2/16:2)
1537.72 1537.26 30 [M+Na-2H]- GM1(d18:1/16:1)
1541.73 1541.21 33 [M+2Na-3H]- GD3(d18:1/20:0)
1544.87 1544.92 3 [M-H]- GM1(d18:1/18:0) or GM1(d18:0/18:1)
1561.80 1561.36 28 [M-H]- O-Ac-GM1(d18:0/16:0)
1566.85 1566.68 11 [M+Na-2H]- GM1(d18:1/18:0) or GM1(d18:0/18:1)
1572.90 1572.92 1 [M-H]- GM1(d18:1/20:0) or GM1(d18:0/20:1)
1589.83 1589.28 35 [M-H]- Fuc-GD3(d18:1/16:0)
102
1593.82 1594.16 21 [M+Na-2H]- GM1(d18:1/20:0)
1599.89 1600.23 21 [M-H]- GM1(d18:0/22:0)
1629.88 1629.32 34 [M-H]- di-O-Ac-GM1(d18:1/18:0)
1648.88 1648.32 34 [M-H]- GD2(d18:0/16:0)
1656.90 1656.88 1 [M-H]- GD2-lactone (d18:1/18:0)
1662.82 1663.16 20 [M+Na-2H]- GD2(d18:2/16:2)
1674.92 1674.33 35 [M-H]- GD2(d18:1/18:0) or GD2(d18:0/18:1)
1690.93 1690.95 1 [M-H]- Fuc-GM1(d18:1/18:0)
1700.96 1701.42 27 [M-H]- GD2(d18:1/20:0) or GD2(d18:0/20:1)
1708.88 1709.36 27 [M+Na-2H]- O-Ac-GD2 (d18:1/16:1)
1716.94 1716.91 2 [M-H]- Fuc-GM1 (d18:1/20:1)
1717.98 1718.48 29 [M-H]- Fuc-GM1 (d18:1/20:0)
1729.01 1729.62 35 [M-H]- GD2(d18:1/22:0) or GD2(d18:0/22:1)
1741.87 1741.27 34 [M+Na-2H]- Fuc-GM1 (d18:0/20:0)
1746.92 1746.38 31 [M+2Na-3H]- GD2(d18:0/20:0)
1775.04 1775.57 30 [M+Na-2H]- GD2(d18:2/24:3)
1787.98 1787.76 12 [M+2Na-3H]- Fuc-GM1(d18:1/22:1)
1796.07 1796.50 24 [M+Na-2H]- Fuc-GM1(d18:1/24:0)
1803.02 1803.64 34 [M-H]- O-Ac-GT3(d18:1/18:1)
1835.96 1835.91 3 [M-H]- GD1(d18:1/18:0) or GD1(d18:0/18:1)
1857.02 1857.59 31 [M+Na-2H]- GD1(d18:0/18:0)
1863.10 1863.61 27 [M-H]- GD1(d18:1/20:0)
1873.07 1872.68 21 [M+Na-2H]- GD1(d18:1/19:0)
1879.93 1879.91 1 [M+2Na-3H]- GD1(d18:1/18:0)
1887.79 1886.98 5 [M-H]- Fuc-GT3-lactone(d18:1/18:2) Fuc-GT3-lactone(d18:0/18:3) Fuc-GT3-lactone(d18:2/18:1)
1895.96 1896.06 5 [M+2Na-3H]- GD1(d18:0/19:0)
1901.05 1901.71 35 [M+2Na-3H]- O-Ac-GT3(d18:1/22:1)
1910.07 1909.43 34 [M-H]- GT2(d18:0/14:0) or GT2(d18:0/14:1)
1915.17 1915.78 31 [M-H]- GD1(d18:1/24:2)
1925.14 1925.77 33 [M+Na-2H]- GD1(d18:1/23:1)
1937.15 1937.80 34 [M+Na-2H]- GD1(d18:1/24:2) or GD1(d18:0/24:3) or GD1(d18:2/24:1)
1964.16 1964.84 35 [M-H]- GT2(d18:0/18:0)
1983.20 1982.55 33 [M-H]- Fuc-GD1(d18:1/18:0)
2005.20 2004.61 30 [M-H]- Fuc-GD1(d18:1/20:2)
103
2010.16
2010.78 31 [M+Na-2H]-
GT2(d18:1/20:2) or GT2(d18:0/20:3) or GT2(d18:2/20:1)
2032.23 2032.14
2032.63 20 24
[M-H]- [M+2Na-3H]-
O-Ac-GT2(d18:1/20:1) or O-Ac-GT2(d18:0/20:2) or O-Ac-GT2(d18:2/20:0) GT2(d18:1/20:2) or GT2(d18:0/20:3) or GT2(d18:2/20:1)
2050.24 2049.75 24 [M-H]- di-O-Ac-GT2 (d18:1/18:0)
2059.27 2059.78 [M+Na-2H]- Fuc-GD1(d18:1/22:0)
2076.24 2076.85 29 [M-H]- GQ3(d18:1/20:2) or GQ3(d18:0/20:3) or GQ3(d18:2/20:1)
2106.27 2105.78 23 [M-H]- (-H2O) GT1(d18:1/18:2)
2166.32 2165.03 2166.29
2165.56 35 24 34
[M-H]- [M+Na-2H]-
O-Ac-GT1(d18:1/18:2) or O-Ac-GT1(d18:0/18:3) or O-Ac-GT1(d18:2/18:1) GT1(t18:1/18:0) O-Ac-GT1(d18:1/16:0)
2172.19 2172.04 7 [M+2Na-3H]- O-Ac-GT1(t18:1/14:1)
2188.38 2188.39
2187.62 35 35
[M+Na-2H]-
(H2O) [M+Na-2H]-
GT1(d18:1/22:0) GT1(d18:1/21:1)
2198.31 2198.08 10 [M+2Na-3H]- GT1(d18:1/20:0)
2214.25 2215.78 24 [M+Na-2H]- O-Ac-GT1(d18:1/20:2)
Healthy cerebellar tissue was found to contain a higher variety of structures
differing in their sialylation degree, from short, monosialylated (GM) to large,
polysialylated carbohydrate chains (GH) and also ganglioside chains modified by
O-acetyl (O-Ac) and fucosyl (Fuc) attachments. GM1 (d18:1/18:0) or (d18:0/18:1),
GM1 (d18:1/20:0) or (d18:0/20:1) and Fuc-GM1(18:1/18:0) were detected as
abundant singly charged ions at m/z 1544.92, 1572.92 and 1690.95 respectively.
Beside these species, highly abundant doubly charged ions at m/z 917.44 and 931.45
assigned to disialylated GD1 components, with (d18:1/18:0) or (d18:0/18:1) and
(d18:1/20:0) or (d18:0/20:1), respectively, were identified. Healthy brain sample is
dominated by mono-, di- and trisialylated structures. 28 distinct m/z signals
correspond to 44 possible GM-type species, 44 m/z signals correspond to 63 possible
GD- type species, and 32 m/z signals are attributable to 59 GT- type species. Most of
104
these structures have a tetrasaccharide sugar core and exhibit high heterogeneity in
their ceramide composition. Additionally, 6 possible tetrasialylated structures (GQ)
and only one asialo species (GA) could be detected. Notable is the presence of a
hexasialylated GH2 species having (d18:0/24:1) or (d18:1/24:0) Cer constitution.
This species was detected as [M-3H]3- at m/z 968.34, and was not found in the
pathological brain sample. 18 possible GG species modified by fucosylation as well
as 30 possible O-acetylated GG variants were also identified. Most of the
fucosylated components are of GM1 and GD1-type with different fatty acid and/or
sphingoid base compositions in the Cer moiety. Unlike fucosylation, O-acetylation
was found for a higher variety of glycoforms such as GM3, GM1, GD3, GD2, GD1,
GT3, GT2, GT1 and GQ1 which differ not only in oligosaccharide chain
composition but also in their sialylation status. Interestingly, 4 possible di-O-Ac GG
variants of GT2, GM1 and GM3 were detected as well.
In contrast to the healthy cerebellar tissue, the ganglioside mixture extracted from
brain metastasis of lung adenocarcinoma exhibited mostly species of short
oligosaccharide chains and reduced overall sialic acid content. More than a half,
from the total of 59 different ions detected and corresponding to 125 possible
structures in brain metastatic tissue, represented monosialylated species of GM1,
GM2, GM3 and GM4-type. Besides the large number of monosialylated
components, 8 asialo species of GA1 and GA2-type bearing ceramides of variable
constitution were discovered. GD1, GD2 and GD3 as well as GT1, GT2 and GT3
with short carbohydrate chains, expressing different ceramide portions were also
identified in the mixture. Ganglioside components modified by Fuc or O-Ac could
also be detected, but in a different pattern than in healthy brain; most O-acetylated
gangliosides are monosialo species of GM3, as well as short GT3- and GT2- type,
while fucosylated components are represented by monosialo species of GM3 and
GM4 structure, di- and trisialylated GD1 and GT3 exhibiting high heterogeneity in
their ceramide motifs.
The most abundant singly charged ions at m/z 1150.17, 1168.01, 1515.29 and 1627.41
were assigned to GA2 (d18:0/22:0) or GM3 (d18:1/16:1); GM3 (t18:0/16:0); sodiated
GD3 (d18:1/18:0) or (d18:0/18:1) or GM1 (d18:1/16:1) or (d18:0/16:2) or GM1
105
(d18:2/16:0) or O-Ac-GD3 (d18:0/18:0) and sodiated GD3 (d18:1/26:0) or
(d18:0/26:1) or GM1 (d18:0/24:2) or (d18:1/24:1) or (d18:2/24:0).
MS data indicated the presence in the metastatic tissue of several unusual
monosialylated species modified by fucosylation or O-acetylation such as Fuc-
GM4, Fuc-GM3, di-O-Ac-GM3, O-Ac-GM3. These species were previously reported
as fetal brain-associated GGs i.e. developmentally regulated antigens, which are
only minor components of the normal brain [OP18].
Figure 3.2.14. Chip nanoESI QTOF CID MS/MS of the singly charged ion at m/z 1471.29 corresponding to GD3 (d18:1/18:0) from brain metastasis of lung adenocarcinoma. Acquisition time 1 min. Insets: fragmentation schemes of the oligosaccharide core and ceramide moiety [OP24]
106
GD3 (d18:1/18:0) was reported to enhance tumor cell proliferation, invasion and
metastasis in a variety of brain tumor cells, especially in glioma and neuroblastoma
[OP42]. GD3 influence tumor angiogenesis and metastasis by stimulating VEGF
release from tumor cells, hence its structural characterization is of high biological
importance. By tandem MS using CID, the oligosaccharide core of the brain
metastasis-associated GD3 (d18:1/18:0) species was structurally elucidated
(Fig.3.2.14). At the same time, a number of Cer-derived fragment ions allowed also
the postulation of the lipid moiety composition.
Optimized MS/MS conditions enabled also the structural assessment of Fuc-GM1
(d18:1/18:0) detected in healthy brain. It was found that the identified Fuc-GM1 is
an atypical isomer bearing the labile Fuc residue at the inner Gal molecule together
with one Neu5Ac attached at the same monosaccharide.
From the methodological point of view it is noteworthy to mention that chip-
nanoESI QTOF MS and CID MS/MS were able to provide compositional and
structural characterization of native ganglioside mixtures from secondary brain
tumors with a remarkable analysis pace and sensitivity. In view of the flow rate
delivered by the chip-nanoESI, which under the applied conditions was around 100
nL/min, 1 min acquisition time at a sample concentration of only 2.5 pmol/μL
corresponds to 250 fmol biological extract consumption. Thus a MS screening
followed by CID MS/MS required only 500 fmols of material. For all these reasons,
the bioanalytical platform demonstrated here for determination of glycolipid
molecular markers in brain tumors has real perspectives of development into a
routine, ultrafast and sensitive method applicable to other types of cancer and
molecular markers.
107
3.4. Structural analysis of chondroitin/dermatan sulfate glycosaminoglycan
(GAG) oligosaccharides
“Let’s finish the GAG paper and put it into the folder of the nice things that had been done.”
Hans Kresse†
Proteoglycans (PGs) are widely distributed in connective tissue and on the
cell surface of mammalian tissues and are functional materials influencing cell
growth, differentiation and morphogenesis. PGs (Fig. 3.4.1) encompass a core
protein linked to glycosaminoglycan (GAG) chains, which interact with a number
of growth factors and important functional proteins.
Figure 3.4.1. Schematic of proteoglycan structure [OP44]
According to the structural type of the disaccharide repeating unit (Fig.
3.4.2), GAG chains are categorized into chondroitin sulfate (CS), dermatan sulfate
(DS), heparan sulfate (HS), heparin, hyaluronic acid (HA), and keratan sulfate (KS).
Among glycosaminoglycans, dermatan sulfates (DS) are carbohydrate species
SERINE SERINE
(THREONINE) (THREONINE)
RESIDUERESIDUE
||
O = CO = C
||
--OO--CHCH22--CHCH
||
NHNH
||
GLYCOSAMINOGLYCANGLYCOSAMINOGLYCAN
n
PROTEIN COREPROTEIN CORE
LINKAGE REGIONLINKAGE REGION
BIOLOGICAL MEMBRANEBIOLOGICAL MEMBRANE
Galactose
N-Acetylated sugar (N-Acetyl
Galactosamine or N-Acetyl Glucosamine)
Xylose
Uronic Acid (Iduronic Acid
or Glucuronic Acid) — C — CH — NH —
II I
O CH2
I
C = O
I
NH
I
ASPARAGINE ASPARAGINE
RESIDUERESIDUE
Fucose
Manose
N-Glycan
108
present in particular in the fibrous connective tissue and on the cell surface.
Structurally, DS are similar to CS and were previously called chondroitin sulfates
B. In the CS case, the hexuronic acid within the repeating disaccharide unit is the D-
glucuronic acid (GlcA) whereas that in DS is either L-iduronic (IdoA) or D-
glucuronic acid. In DS two types of dissacharide units are present: -4GlcA1-
3GalNAc1- and -4IdoA1-3GalNAc1-.
Figure 3.4.2. Detailed structures of repeating disaccharide units in various CS types
[OP44]
The standard disaccharide moiety is modified by sulfation in the GalNAc
moiety. The GalNAc unit may be sulfated in position 4 or 6, while a minor
proportion of the uronic acid may be sulfated in position 2, if an additional sulfate
is present. The GlcA residues are to be found in disaccharides containing either
GalNAc-4-O-sulfate, GalNAc-6-O-sulfate or GalNAc, while the IdoA is exclusively
attached to the GalNAc-4-O-sulfate. If sulfated, the IdoA moiety endows the GAG
chain with additional negative charge content and even more, conformational
studies revealed that IdoA being more flexible is prone to exhibit energetically
more favorable conformations, which explains the implication of the DS
oligosaccharides in biologically active complexes.
109
In view of GAG oligosaccharide importance in many biological systems,
accompanied by their compositional diversity, systematic studies were conducted
for the complete characterization on one hand and for correlating the GAG
saccharide sequence to their complex biological functions on the other.
Additionally, the high importance of CS/DS forms and sulfation patterns in
mediation of biological activities focused many studies on their analysis. For
structural investigation of GAG oligosaccharides further development of specific
methods was required, among which, mass spectrometry contributed lately an
essential progress [91-93, OP44-OP48].
Difficulties encountered in the ionization of CS/DS mixtures limited for long
time mass spectrometry potentials in structural elucidation of GAG chains. The
most severe problems are related to the difficulty to obtain high ionization yield for
long GAG chains, to hinder the in-source decay of the labile sulfate group, to
generate the multiply charged ions of high molecular weight GAG species and
finally to distinguish the isobaric structures. Moreover, in the case of ESI MS, the
feature of the spectrum is more complex even for a single component because it
contains ions bearing different charge state and in normal case also one or more
alkali counter ions, which generate signals of very different m/z values [OP44]. For
complex mixtures, the spectrum is difficult to be interpreted because of the isobaric
peak overlapping. For these reasons, the screening of the complex GAG mixtures
required the combination of MS with an efficient separation technique such as CE
properly optimized for this type of analysis.
In this research, an analytical approach based on CE in conjunction with
negative ESI-quadrupole time-of-flight tandem mass spectrometry (QTOF MS/MS)
has been for the first time developed for providing the basis to obtain new insights
into the domain structure of the chondroitin/dermatan sulfates [OP49]. The
feasibility and performance of the off-line CE ESI QTOF MS approach in GAG
oligosaccharide analysis were assessed by screening CS/DS oligosaccharide
mixture obtained from bovine aorta by enzymatic depolymerization by chondroitin
B lyase.
110
Several new aspects offering another dimension to the MS applicability in
glycosaminoglycan characterization were revealed by this study. First, that
determination of molecular characteristics of GAGs, which is an essential
prerequisite in understanding their biological functions may be precisely done by
CE ESI MS. Secondly, a full study of glycosaminoglycan molecular structure must
include primarily the determination of molecular ions of all components in
oligosaccharide mixtures obtained after detachment from protein by chemical or
enzymatic means followed by size-exclusion chromatography. By sequencing
single components in such mixtures, the presence of regular or irregular units can
be clearly detected.
An interesting methodological novelty is that CE UV could provide signals
of quite high intensity for 10 GAG components, although the detection by UV
absorption is not the best suited method for carbohydrates.
Fig. 3.4.3. Negative ion mode nanoESI QTOF MS of the CE fraction collected within the first 5 min after injection. CE carrier: 50 mM ammonium acetate/ammonia, pH 12.0. CE separation voltage 25 kV, 8 s injection by pressure; ESI capillary potential 650 V; sampling cone potential 15 V [OP49]
Another important achievement was that by combining CE and ESI MS,
GAG species, which were not detectable previously by mass spectrometry alone,
111
could be identified. Besides of enhancing the MS detection of minor components,
CE separation eliminated the widely known possibility of misinterpreting the GAG
composition due to the overlapping of the isobaric MS peaks. By using optimized
ionization conditions in the nanoESI MS screening of the CE separated fractions,
the in-source desulfation of the molecules could be avoided and the formation of
multiply charged ions was favored. Both aspects provided a significant
contribution to the successful MS detection of fully sulfated octa- and
decasaccharides as well as of oversulfated hexasaccharides from the CE fractions
(Fig. 3.4.3).
The last stage of the methodology development included the use of tandem
mass spectrometry to provide the elucidation of monosaccharide building block
sequences, information on the repeating GlcA-GalNAc, GlcA-GalNAc(S) units, as
well as data on the glycosidic linkages. The fully sulfated octasaccharide detected
in the first CE fraction and the sulfated disaccharide from the second CE fractions
were the precursor ions in the MS/MS experiments. The most important outcome
of the fragmentation process of both species was the clear indication of the sulfate
group substitution pattern along the GAG chain.
According to these results, CE nanoESI MS and tandem MS methodology
appeared as a practical alternative, which overcomes part of the limitations and
barriers experienced in structural analysis of glycosaminoglycans. Moreover, these
data suggested that the domain structure of biologically active DS chains may be
elucidated by this approach.
In a subsequent study [OP50] dedicated to glycosaminoglycan structural
analysis by MS means, hybrid CS/DS glycosaminoglycan chains derived from
decorin, secreted by human skin fibroblasts were shown to interact with FGF-2, as
did oligosaccharides derived therefrom by chondroitin B lyase digestion. The
structure of decorin, this ubiquitous leucine-rich proteoglycan, is depicted in Fig.
3.4.4.
For identification of the biologically active sequence an improved CE MS
protocol for structural analysis of enzyme-resistant oligosaccharides larger than
standard trisulfated hexasaccharides was reported [OP50]. The method was based
112
also on CE for separating oversulfated species in off-line combination with nanoESI
QTOF MS/MS in the negative ion mode.
Figure 3.4.4. Structure of decorin
The heterogeneity of the oligosaccharide mixture was first demonstrated by
CE with UV detection (Fig.3.4.5).
Fig. 3.4.5. CE UV profile of CS/DS oligosaccharides from human decorin. Electrolyte: ammonium acetate/ammonia, pH 12.0; CE separation voltage: 25 kV; 3s injection by pressure; 12 nl injected volume; detection at 214 nm [OP50]
DECORIN
CC
CC CC
C
Leucine-Rich Repeat
Cysteine
Chondroitin/
Dermatan Sulfate
N-Glycan
O
H
H
H
OH
H OH
COOH
O O
O
H
HH
OH
H NHCOCH3
CH2OSO3H
O
n
CHONDROITIN 6-SULFATE
O
H
HCOOH
H
OH
H OH
O O
O
H
HH
SO3H
H NHCOCH 3
CH2OH
O
n
DERMATAN SULFATE
H
1
3
2
4 5
6 7
8 9 10
0.0
Ab
sorb
ance
-0.368
1.471
10-3
11
min. 3.00
1.0 2.0 3.0 4.0 5.0 6.0
CE fraction I
CE fraction II
time (min)
113
By nanoESI QTOF MS analysis of the CE fractions, up to 12-mer
oligosaccharides with different degrees of sulfation were identified (Fig. 3.4.6,
Table 3.4.1). A novel tandem MS protocol of collision-induced dissociation at
variable energy (CID-VE) was applied to elucidate the structure of a previously
undescribed pentasulfated CS/DS hexasaccharide (Fig. 3.4.7, Table 3.4.2.).
a) b)
Figure 3.4.6. Negative ion mode nanoESI QTOF MS, m/z range a) (450-750); b) (750-950) of the CE fraction collected within the first 3 min after the application of the separation voltage. Electrolyte: 50 mM ammonium acetate/ammonia; pH 12.0; CE separation voltage 25 kV; 6 s injection by pressure; 20 nl injected volume; ESI capillary potential 700 V; sampling cone potential 15 V [OP50] It was effectively demonstrated that this 3-stage method based on CE, ESI MS and -
MS/MS is a powerful tool for structural elucidation of GAG chains of decorin,
prepared from conditioned media of human skin fibroblasts. The success of the
method required developing new conditions for each of the analytical steps
involved, the CE separation, the ESI MS screening and the sequencing the GAG
species in tandem MS experiments by employing a new approach of CID-VE at
variable acceleration energy of the precursor ion. Thus, the CE separation
electrolyte ammonium acetate/ammonia, pH 12.0, has been adapted to the
requirements for ESI MS. By CE UV monitoring, the heterogeneity of the GAG
mixture was assessed to a reasonable extent.
Fig.4
460 480 500 520 540 560 580 600 620 640 660 680 700m/z0
100
%
x4458.02
585.03
517.38
511.38459.16
502.90
552.66
517.70
546.90
552.91
572.63
611.28
585.35
687.38
619.20
619.55
647.40
647.91
679.37
687.87
711.89
712.92
533.82
611.604-
3-
3-
5-
4-
4-
3-
3-
4-
2-
3-2-
(450-750) u
664.89
3-634.89
2-
760 780 800 820 840 860 880 900 920m/z0
100
%
x2917.13
877.25
776.35764.03
759.87
776.84
777.36
777.86
819.84
877.74
899.86878.26
900.81
908.80
917.47
917.79
3-2-
2-
3-
Fig.4 cont
(750-950) u
760 780 800 820 840 860 880 900 920m/z0
100
%
x2917.13
877.25
776.35764.03
759.87
776.84
777.36
777.86
819.84
877.74
899.86878.26
900.81
908.80
917.47
917.79
3-2-
2-
3-
Fig.4 cont
(750-950) u
114
Table 3.4.1. Molecular ions of CS/DS oligosaccharide species obtained from the human skin fibroblasts decorin detected by negative ion mode nanoESI QTOF MS in the first CE fraction collected within the first 3 min after the application of the separation voltage [OP50]
*regularly sulfated species; **oversulfated species (nS) denotes the number of sulfate groups in the molecule
Figure 3.4.7. (-) nanoESI QTOF MS/MS of the pentasulfated hexasaccharide detected as a triply charged ion at m/z 511.38 in MS of the first CE fraction. CID-VE of 10-30 eV [OP50]
m/z Charge state Structure
458.02 4 IdoAGalNAc[GlcAGalNAc]3(4S)*
511.38 3 IdoAGalNAc[GlcAGalNAc]2(5S)**
517.38 3 IdoAGalNAc[GlcAGalNAc]2(5S)**
533.88 5 IdoAGalNAc[GlcAGalNAc]5 (5S)
552.66 4 IdoAGalNAc[GlcAGalNAc]4 (4S)
572.63 4 IdoAGalNAc[GlcAGalNAc]4 (5S)*
585.03 3 IdoAGalNAc[GlcAGalNAc]3 (3S)
611.28 3 IdoAGalNAc[GlcAGalNAc]3(4S)*
647.40 4 IdoAGalNAc[GlcAGalNAc]5 (4S)
664.89 3 IdoAGalNAc[GlcAGalNAc]3(6S)**
687.38 2 IdoAGalNAc[GlcAGalNAc]2(3S)*
711.89 3 IdoAGalNAc[GlcAGalNAc]4 (3S)*
764.03 3 IdoAGalNAc[GlcAGalNAc]4(5S)*
877.25 2 IdoAGalNAc[GlcAGalNAc]3 (3S)
917.13 3 IdoAGalNAc[GlcAGalNAc]5(6S)*
160 170 180 190 200 210 220 230 240 250 2600
100
%
x18175.30
157.27
193.33
202.38237.56
237.04255.34
238.09
280 300 320 340 360 380 400 420 440 460 480m/z0
100
%
x4 476.08300.14
282.14
379.13
370.01301.14
324.00 360.05
339.05
440.05
396.38
379.6
427.06
418.14
458.15
477.11
449.08
466.2
(280-490) u
Fig.5
316.98
397.08
280 300 320 340 360 380 400 420 440 460 480m/z0
100
%
x4 476.08300.14
282.14
379.13
370.01301.14
324.00 360.05
339.05
440.05
396.38
379.6
427.06
418.14
458.15
477.11
449.08
466.2
(280-490) u
Fig.5
316.98
397.08
(490-680) u
500 520 540 560 580 600 620 640 660 680m/z0
100
%
x6647.47
568.44
506.62
507.08
520.03
554.94
607.60
568.98
586.09
599.0
608.50
608.96
634.45
648.08
648.62
678.57649.11498.01 559.4
616.94
656.45665.58
536.04
537.99
528.96
546.11638.62
700 750 800 850 900 950 1000 1050 1100
m/z0
100
%
x54x10 x124899.15819.09687.38
727.94
687.87
688.37728.45
775.14
734.05
739.24
776.13
793.12
837.06
855.05
856.05
857.04
901.27
1013.31
915.16
933.11
979.08
935.12
936.22
939.21
995.00
1093.11
1014.28
1015.34
1018.27
1094.12
696.62
(685-1100) u
Fig.5 cont
700 750 800 850 900 950 1000 1050 1100
m/z0
100
%
x54x10 x124899.15819.09687.38
727.94
687.87
688.37728.45
775.14
734.05
739.24
776.13
793.12
837.06
855.05
856.05
857.04
901.27
1013.31
915.16
933.11
979.08
935.12
936.22
939.21
995.00
1093.11
1014.28
1015.34
1018.27
1094.12
696.62
(685-1100) u
Fig.5 cont
115
Table 3.4.2. m/z values of fragment ions obtained by tandem MS experiment depicted in Fig. 3.4.7 and their structure assignment to the pentasulfated CS/DS hexasaccharide, used as a precursor ion [OP50]
m/z charge state structure type of ion
157.27 1 IdoA B1
175.30 1 IdoA C1
193.33 1 GlcA Y6/C3 or Y4/C5
237.56 2 [GlcAGalNAc] (1S) Y2*
255.34 1 IdoA(1S) C1**
282.14 1 GalNAc (1S) Z1*
300.14 1 GalNAc(1S) Y1*
316.98 2 [IdoAGalNAcGlcA] (1S) C3*
339.05 2 [GalNAcGlcAGalNAc](1S) Z3
370.01 2 [GalNAcGlcAGalNAc](2S) Z3*
378.11 1 IdoAGalNAc B2
379.13 2 [GalNAcGlcAGalNAc] (2S) Y3*
396.38 1 IdoAGalNAc C2
397.08 2 [IdoAGalNAcGlcA] (3S) C3**
418.14 2 [IdoAGalNAcGlcAGalNAc] (1S) B4
427.06 2 [IdoAGalNAcGlcAGalNAc] (1S) C4
440.05 1 [IdoAGalNAc] (1S) B2
449.08 2 [IdoAGalNAcGlcAGalNAc] (2S) B4*
458.15 1 [IdoAGalNAc] (1S) C2*
466.21 2 [GlcAGalNAcGlcAGalNAc] (2S) Y4*
476.08 1 [GlcAGalNAc] (1S) Y2*
498.01 2 [IdoAGalNAcGlcAGalNAc] (3S) C4**
506.62 2 {IdoA[GlcAGalNAc]2} (1S) B5
520.03 2 [IdoAGalNAcGlcAGalNAc] (4S) B4**-H2O
528.96 2 [IdoAGalNAcGlcAGalNAc] (4S) B4**
536.04 1 IdoAGalNAcGlcA B3
546.11 2 {IdoA[GlcAGalNAc]2}(2S) B5*
554.94 2 {IdoA[GlcAGalNAc]2}(2S) C5*
559.42 2 {[GlcAGalNAc]2GalNAc}(2S) Z5
568.44 2 {[GlcAGalNAc]2GalNAc}(2S) Y5
586.09 2 {IdoA[GlcAGalNAc]2}(3S) B5**
589.99 2 {[GlcAGalNAc]2GalNAc}(3S) C5*
595.04 2 {IdoA[GlcAGalNAc]2}(3S) C5**
599.04 2 {[GlcAGalNAc]2GalNAc}(3S) Z5*
607.60 2 {IdoAGalNAc[GlcAGalNAc]2}(1S) B6
616.94 2 {IdoAGalNAc[GlcAGalNAc]2}(1S) C6
634.45 2 {IdoA[GlcAGalNAc]2}(4S) C5**
638.62 2 {IdoAGalNAc[GlcAGalNAc]2}(2S) B6
647.47 2 {IdoAGalNAc[GlcAGalNAc]2}(2S) C6
656.45 2 [GlcAGalNAc]3 (2S) Y6
678.57 2 {IdoAGalNAc[GlcAGalNAc]2}(3S) B6*
687.38 2 {IdoAGalNAc[GlcAGalNAc]2}(3S) B6*
696.62 2 {IdoAGalNAc[GlcAGalNAc]2}(3S) C6*
727.94 2 {IdoAGalNAc[GlcAGalNAc]2}(4S) Y6**
116
739.24 1 [IdoAGalNAcGlcAGalNAc] B4 -H2O
775.14 1 [IdoAGalNAcGlcAGalNAc] C4
793.12 1 [GlcAGalNAc]2 Y4
819.09 1 [IdoAGalNAcGlcAGalNAc](1S) B4 -H2O
837.06 1 [IdoAGalNAcGlcAGalNAc](1S) C4
855.05 1 [GlcAGalNAc]2 (1S) Y4
899.15 1 [IdoAGalNAcGlcAGalNAc](2S) B4* -H2O
915.16 1 IdoA[GlcAGalNAc]2 B5
933.11 1 IdoA[GlcAGalNAc]2 C5
979.08 1 [IdoAGalNAcGlcAGalNAc](3S) B4**
995.00 1 {IdoA[GlcAGalNAc]2}(1S) B5- H2O
1013.31 1 {IdoA[GlcAGalNAc]2}(1S) B5
1093.11 1 {IdoA[GlcAGalNAc]2}(2S) B5*
*regular sequence ions **oversulfated fragment ions (nS) denotes the number of sulfate groups in fragment ions.
In comparison, using HPAEC-PAD for oligosaccharide separation and
detection, nine components in the oligosaccharide mixture were visualized,
whereas using CE UV detection, however, eleven GAG components were traced,
demonstrating a superior separation efficiency under the given conditions. Another
interesting information of the MS screening of the CE fractions is that the species
with high molar sulfate content could be clearly separated from the non-sulfated
ones, present in the GAG mixture released by -elimination. For strict
determination of the degree of sulfation in single GAG species and delimiting the
real under- and nonsulfated species from the possible artifacts induced by the in-
source decay of the sulfate groups in the MS mode, this aspect was crucial.
In analogy to the previously reports about ESI MS methods for GAG
oligosaccharide analysis it was observed that in the negative ESI MS, the in-source
desulfation may be reduced by acquiring the spectra under mild values of the
sampling cone potential. Using this protocol, it was possible to detect up to fully
sulfated dodecasaccharide and some intact oversulfated species.
Detailed structural characterization was achieved by CID-VE fragmentation
of the novel DS-containing hexasaccharide, 4,5--IdoAGalNAc[GlcAGalNAc]2(5S),
for determination of the sulfation pattern along the carbohydrate chain. According
to the MS/MS data three sulfates are distributed in the IdoAGalNAcGlcA moiety,
offering two structural variants: one containing the sulfated IdoA and the
117
disulfation of GalNAc moiety, and the other with the both UroA moieties and the
GalNAc each monosulfated (Fig. 3.4.8).
Figure 3.4.8. Two alternative structure proposals for the pentasulfated CS/DS hexasaccharide according to the data from the MS/MS depicted in Fig. 3.4.7. The upper structure proposal is the more probable one [OP50]
The sequence data for the novel DS hexasaccharide IdoAGalNAc
[GlcAGalNAc]2(5S) confirmed the presence of a tetrasulfated tetrasaccharide partial
sequence assigned either to the IdoA(S)GalNAc(S)GlcA(S)GalNAc(S) or to the
IdoA(S)GalNAc(2S)GlcAGalNAc(S) moiety.
The introduction of a 3-step-analysis by combining the CE separation with
ESI MS and a novel approach for CID-VE fragmentation provided a solid platform
for investigation of fine structure in group of CS/DS oligosaccharides, which have
been so far the less investigated among GAGs.
To extend the CE MS applicability to longer CS/DS oligosaccharide chains
species, the work was further focused on the development of a novel approach in
glycosaminoglycomics based on sheathless on-line CE nanoESI QTOF MS [OP10].
The methodology required the construction of the new nanosprayer sheathless CE
nanoESI QTOF MS configuration described in Part II, its implementation and
optimization for the high sensitivity analysis of CS/DS oligosaccharide mixtures
from conditioned culture medium of human embryonic kidney fibroblasts (HEKF).
Under newly established sheathless on-line CE (-) nanoESI conditions for GAG
ionization and MS detection, single CS/DS oligosaccharide components of
IdoA-GalNAc-GlcA-GalNAc-GlcA-GalNAc
SO3-SO3
- SO3- SO3
-SO3-
IdoA-GalNAc-GlcA-GalNAc-GlcA-GalNAc
SO3-SO3-SO3
-SO3- SO3
-SO3- SO3
-SO3-SO3
-SO3-
IdoA-GalNAc-GlcA-GalNAc-GlcA-GalNAc
SO3-SO3
-
SO3-
SO3-SO3
-
IdoA-GalNAc-GlcA-GalNAc-GlcA-GalNAc
SO3-SO3-SO3
-SO3-
SO3-
SO3-SO3-SO3
-SO3-
118
extended chain length and increased sulfation degree were identified. The spectra
generated by combining in progress across the TIC MS (Fig. 3.4.9) peaks showed a
pentacharged molecular ion corresponding to the unsaturated oversulfated
tetradecasaccharide assigned to the composition of 4,5 IdoA-GalNAc {GlcA-
GalNAc}6(9S), a hexacharged ion to be assigned to the saturated eicosasaccharide
IdoA-GalNAc {GlcA-GalNAc}9(11S), a species detected as [M-5H]5- ion
corresponding to a composition of an unsaturated octadecasaccharide 4,5 IdoA-
GalNAc {GlcA-GalNAc}8(10S) and an abundant [M-5H]5- ion assigned to the
unsaturated eicosasaccharide bearing eleven sulfate groups having the structure of
4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S) (Fig. 3.4.10).
Figure 3.4.9. Sheathless on-line CE (-)nanoESI QTOF MS total ion chromatogram of the glycosaminoglycan mixture from human kidney fibroblast decorin. CE buffer: 40mM ammonium acetate/ammonia pH 11.8. CE separation voltage 30 kV direct polarity, 6 s injection by pressure. CE column length 100 cm. Nanosprayer potential 700V, sampling cone potential 15V. ESI MS signal acquisition 15 min after injection [OP10].
These data demonstrated the complexity of the sample, which is related to
the high length of the GAG oligosaccharide chains present, as well as to their type
and high level of sulfation. The molecular ions obtained in this experiment were
0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00Time0
100
%
0.7539780
0.6735670
0.5328673
0.2614588
1.4171933
1.3267933
0.9348
841
1.7588856
1
2
3
4
56
7
8
t-elution time 15 min after injection
n- number of scans
m/z of the most abundant ion eluted at the moment t
119
shown to carry different numbers of sulfate groups per disaccharide unit and were
all expressing oversulfation of the molecules. High percentage of species separated
and detected in the spectra were assigned to carry one double bond, originating
from the specific eliminative action of chondroitin B lyase on GalNAc-IdoA
linkages, which is the characteristic of oligosaccharides representing defined
hybrid molecules bearing a single DS disaccharide unit at the non-reducing end,
linked to a variable number of CS disaccharide units at the reducing terminal.
a) b)
c)
Figure 3.4.10. Sheathless on-line CE (-)nanoESI QTOF mass spectra combined from a) the 2-nd b) 4-5th and c) 7-th TIC-MS peaks [OP10].
600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250 1300 1350 1400 14500
100
%
673. 61
587.21
650.02 766.12
673.82
674.01
674.21
m/z
[M-5H]5-
0
100
%
673.61
673.82
674.01
674.21
674.41
m/z
600 650 700 750 800 850 900 950 1000 1050 1100 1150 1200 1250 13000
100
%
780.40
780.71
841.47
780.87
841.67
841.87
780.56
[M-6H]6-
[M-5H]5-
m/z0
100
%
780. 40
780.56 780.71
780.87
0
100
%
841.47
841.28
841.67
841.87
842.07
m/z
575 600 625 650 675 700 725 750 775 800 825 850 875 900 925 950 975 10001025 1050 1075
0
100
%
933.12
727.84
933.32
933.52
933.72
933.92
[M-5H]5-
0
100
%
933.12
933.32
933.52
933.72
933.92
934.13
934.33
120
The detected and identified oversulfated eicosasaccharides, represent the
category of the largest sulfated glycosaminoglycan-derived oligosaccharides
evidenced ever by mass spectrometry. By on-line CE ESI tandem MS in data
dependent analysis mode the oversulfated eicosasaccharide species could be
sequenced and the biologically-relevant localization of the additional sulfate group
along the chain could be determined.
Figure 3.4.11. Sheathless on-line CE (-)nanoESI QTOF auto MS/MS of the
oversulfated unsaturated eicosasaccharide 4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S) detected as pentacharged ion at m/z 933.12 in the seventh TIC peak at min 15.41 after injection. ESI potential 700 V; Sampling cone potential 15 V. Collision energy 35 eV [OP10].
m/z (1400-1700) u.
1450 1500 1550 1600 1650 1700m/z0
100
%
1502.53
1434.841470.54
1435.18
1529.16
1502.8
1503.2
1529.49
1529.82
1605.74
1534.18
1535.181565.69
1596.75
1606.24 1693.761653.80
1694.26
3-3-
[M-3H]3-
2-
2-
2-
2- 2-
Z19(10S)
B19(9S)
4,5 IdoA-GalNAc{GlcA-GalNAc}9(11S)
Z19 (11S)
Y19 (11S)
3-
3-
[M-3H-SO3]3- [M-3H-2SO3]
3-
3-
Y19(10S)
C19(9S)
C14 (7S)B15 (6S)
C15 (6S)B14 (7S) 1491.14
1497.13
1- B7-H2O
C14(6S)
[M-3H-2SO3]3-
[M-3H-SO3]3-
m/z (1400-1700) u.
1450 1500 1550 1600 1650 1700m/z0
100
%
1502.53
1434.841470.54
1435.18
1529.16
1502.8
1503.2
1529.49
1529.82
1605.74
1534.18
1535.181565.69
1596.75
1606.24 1693.761653.80
1694.26
3-3-
[M-3H]3-
2-
2-
2-
2- 2-
Z19(10S)
B19(9S)
4,5 IdoA-GalNAc{GlcA-GalNAc}9(11S)
Z19 (11S)
Y19 (11S)
3-
3-
[M-3H-SO3]3- [M-3H-2SO3]
3-
3-
Y19(10S)
C19(9S)
C14 (7S)B15 (6S)
C15 (6S)B14 (7S) 1491.14
1497.13
1- B7-H2O
C14(6S)
[M-3H-2SO3]3-
[M-3H-SO3]3-
1240 1260 1280 1300 1320 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520m/z0
100
%
x21249.74
1402.81
1250.07
1308.451250.40
1250.72
1258.70
1356.71
1308.77
1326.75
1337.80
1376.17
1376.49
1376.83
1403.48
1403.831479.36
1454.771424.73 1522.75
1480.34
1498.74
3-
Y16(9S)
B14-H2O
2-
Y18(9S)
3-
Z18(9S)
3-
Z18(9S)-H2O
3-
Y18(10S)
3-
C13(6S)
2-
B12(5S)
2-
B12Na(5S)
1-
B7(2S)Y7(2S)
2-
1-
1240 1260 1280 1300 1320 1340 1360 1380 1400 1420 1440 1460 1480 1500 1520m/z0
100
%
x21249.74
1402.81
1250.07
1308.451250.40
1250.72
1258.70
1356.71
1308.77
1326.75
1337.80
1376.17
1376.49
1376.83
1403.48
1403.831479.36
1454.771424.73 1522.75
1480.34
1498.74
3-
Y16(9S)
B14-H2O
2-
Y18(9S)
3-
Z18(9S)
3-
Z18(9S)-H2O
3-
Y18(10S)
3-
C13(6S)
2-
B12(5S)
2-
B12Na(5S)
1-
B7(2S)Y7(2S)
2-
1-
4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S)
m/z (1240-1530) u.
121
Additional structural information upon the oversulfated unsaturated
eicosasaccharide 4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S) were acquired by data-
dependent MS to MS/MS mode switching available on the QTOF instrument,
which can be introduced as a supplementary on-line fragmentation analysis for
identification of single carbohydrate molecular species separated by on-line CE ESI
QTOF MS, as shown previously.
In the MS to MS/MS switching approach (Fig. 3.4.11) applied to the
pentacharged 4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S) even under the restrictive
acquisition time conditions the product ion analysis generated fair fingerprint ion
set for identification of the molecular structure in terms of epimerization and
sulfation pattern.
Figure 3.4.12. Structural proposal for the oversulfated unsaturated eicosasaccharide
4,5 IdoA-GalNAc {GlcA-GalNAc}9(11S) according to the data obtained by on-line CE QTOF autoMS/MS [OP10].
The reduced proportion of undersulfated sequence ions that occurred under
the autoMS/MS conditions chosen for this experiment showed that the formation
of ions resulting from the cleavage of glycosidic bonds was favored and the loss of
SO3- groups was kept to a limited extent. This feature is attributable also to the fact
that, by submitting a highly charged eicosasaccharide species to autoMS/MS at
rather low collision energy, the formation of multiply charged fragment ions which
are less prone to sulfate cleavage was enhanced and the localization of the
additional biologically-relevant sulfate group could be unambiguously determined
(Fig.3.4.12).
4,5-IdoA-O-GalNAc- GlcA-GalNAc -GlcA-O-GalNAc-OH
S S S
B1
Y19
C1
Z19
B19 C19
Y1Z1
8
S
4,5-IdoA-O-GalNAc- GlcA-GalNAc -GlcA-O-GalNAc-OH
S S S
B1
Y19
C1
Z19
B19 C19
Y1Z1
8
S
122
A challenge of MS based CS/DS analysis is related to the mutually exclusive CID
sequencing principles, which are required for reliable determination of sulfation
site(s) i.e., cleavage of the glycosidic bond while keeping the SO3 attached [OP44,
OP45]. Additionally, Zaia’s group [94] noticed that the interplay of uronic acids
and sulfates, determine the product ion patterns in CID experiments on CS. Thus,
unsulfated chondroitin dissociates to form C-type ions almost exclusively, while CS
produces abundant B- and Y-type ions from glycosidic bond cleavage with C- and
Z-type of ions present only in low abundances. These observations were explained
in terms of competing proton transfer reactions that occur during the collisional
heating process.
Another interesting aspect of CID MS/MS of CS is that product ion abundances
reflect sulfation position at GalNAc residues [95, 96] and epimerization of HexA
residues [97].
Based on these findings CID MS/MS was used [98] to determine positions of
sulfation and epimerization by comparing abundances of the fragment
ions formed from unknown DS oligosaccharides with those produced by CS/DS
standards with known epimerization and sulfate positions at GalNAc.
In low-energy CID experiments, product ion abundances correlate with the lability
of the cleaved covalent bonds. Thus, the sulfation and epimerization positions
influence the lability of certain bonds in the oligosaccharide ions that are reflected
by the observed ion abundances. Recently, in an application of this quantitative
method based on CID MS/MS the percent composition of CS A-like (4GlcAβ1-
3GalNAc4Sβ1-), CS B-like (4IdoAα1-3GalNAc4Sβ1-), CS C-like (4GlcAβ1-
3GalNAc6Sβ1-) isomers for cartilage, ligament, muscle, tendon, and synovium
samples could be calculated [99]. CS A, CS B, and CS C standards were used to
compare the relative amounts characteristic of each biological sample. Therefore, it
is basically proven that this analytical platform can be used to identify CS and DS
in human samples.
However, as inferable from these considerations, while multistage mass
spectrometric investigation of regular and under-sulfated regions was performed
successfully [94-102], so far the analysis of over-sulfated CS/DS domains by
123
MS/MS [OP48, OP50] resulted in rather reserved conclusions. Since sulfate esters
are more labile than glycosidic linkages under most CID conditions, the
unequivocal identification of excess sulfation sites has not been achieved in a single
dissociation stage. Therefore, in this work it was introduced a novel and
straightforward method to analyze atypical sulfation patterns in CS/DS chains
derived from human fibroblast decorin using the strategy depicted in Fig. 3.4.13
[OP51].
Figure 3.4.13. Schematic of the strategy for GAG compositional and structural analysis based on the recognition specificity of chondroitin lyases and multistage mass spectrometry (MSn) [OP51]
ESI MSn analysis encompasses: i) determination of molecular ion masses (MS1)
giving their sizes and overall extents of sulfation followed by possible correlation
of unusual sulfation content with HexA-epimerization and ii) identification of the
positions of sulfate groups along the chain from the masses of the fragment ions
generated by stepwise ion dissociation in multiple sequencing events (CID MS2-
MS4).
Decorin -elimination GAG chain
Depolymerization
A B
Mixture of
variable length
chains
Profiling/
Fractionation
Collection of
CS disaccharides
Collection of
CS/DS hexasaccharides
Purification
Size-exclusion
chromatography
ESI MSn
GalNAcI
IdoAI
GalNAcI
GlcAI
GalNAcI
I
A
B
I
IdoAI
GalNAcI
GlcAI
GalNAcI
I
A
B
Chondroitin
B lyase
Chondroitin
AC lyase
1
2
3
124
To establish the location of sulfate groups within over-sulfated GlcA- and IdoA-
rich domains, the triply charged ion at of the GlcA-rich pentasulfated
hexasaccharide [4,5 ΔIdoAGalNAc-(GlcAGalNAc)2](5S) detected by MS1 has been
chosen as the primary target for multistage MS analysis (Figs. 3.4.14-3.1.15).
Figure 3.1.14. Multiple stage ESI HCT CID MS structural analysis of pentasulfated
[4,5--IdoAGalNAc(GlcAGalNAc)2]. MS2 of the triply deprotonated ion at m/z 511.20; MS3 of the doubly deprotonated fragment ion at m/z 538.05 [OP51]
125
Figure 3.1.15. Multiple stage ESI HCT CID MS structural analysis of pentasulfated
[4,5--IdoAGalNAc(GlcAGalNAc)2]. MS4 of the doubly deprotonated fragment ion at m/z 387.23. Insets: Proposed structures, their fragmentation pathways and observed product ions. *over-sulfated, #regularly sulfated and ¥under-sulfated fragment ions [OP51]
As previously demonstrated [OP50] by CID a high coverage of structurally
informative sequence ions generated rather by glycosidic bond cleavage than SO3
loss can be obtained if variable collision energy in the low eV range is employed. In
the present case of successive CID in a single experiment, the fragmentation
amplitude, its ramping interval, and time were adjusted in each step and for each
re-sequenced fragment ion. Under such conditions, all spectra displayed a high
proportion of over-sulfated fragment ions useful for localization of sulfate groups
within the [4,5-Δ- IdoAGalNAc(GlcAGalNAc)2](5S) hexasaccharides.
In the next stage of analysis, the tetra-sulfated hexasaccharide originating from
IdoA-rich domains was detected by MS1 as a rather abundant triply charged ion at
m/z 484.67 which, according to mass calculation, had the composition [4,5-Δ-
GlcAGalNAc(IdoAGalNAc)2](4S). The sulfate groups were localized by multiple
stage MS, which included CID MS2 and MS3 (Fig. 3.1.16). MS2 yielded three over-
sulfated ions * diagnostic for GlcA sulfation and the additional sulfation to the first
IdoA from the non-reducing end or di-sulfation of GalNAc [OP51].
Y1Z2/B2
*#
250 300 350 400 450 500 550 600 650 700
m/z
237.33
282.20
300.21
387.23
458.19
538.09
695.60
[M-2H]2-
B1
C2
M-SO3
C2-SO3
B2-SO3
440.17
C1
255.35
Z2/B2
Y2/B2
-IdoA-O-GalNAc-O-GlcA
SO3 SO3 SO3
-IdoA-O-GalNAc-O-GlcA
SO3 SO3 SO3
-
--
-
-
-
-
-
B1 C1 C2
519.92
B2
B2
-
273.39
Y1
-
*
*
*
*
*#
*
*
#
*
*
*
#
#
250 300 350 400 450 500 550 600 650 700
m/z
237.33
282.20
300.21
387.23
458.19
538.09
695.60
[M-2H]2-
B1
C2
M-SO3
C2-SO3
B2-SO3
440.17
C1
255.35
Z2/B2
Y2/B2
-IdoA-O-GalNAc-O-GlcA
SO3 SO3 SO3
-IdoA-O-GalNAc-O-GlcA
SO3 SO3 SO3
-
--
-
-
-
-
-
B1 C1 C2
519.92
B2
B2
-
273.39
Y1
-
*
*
*
*
*#
*
*
#
*
*
*
#
#
126
a)
b)
Figure 3.1.16. Multiple stage ESI HCT CID MS structural analysis of tetrasulfated
[4,5--GlcAGalNAc(IdoAGalNAc)2]. a) MS2 of the triply deprotonated ion at m/z 484.69; b) MS3 of the doubly deprotonated fragment ion at m/z 489.16; Insets: Proposed structures, their fragmentation pathways and observed product ions. *over-sulfated, #regularly sulfated and ¥under-sulfated fragment ions [OP51]
237.36
B1
-
282.26
300.28
307.21
370.18
440.14
458.10
484.69
489.16
576.19
536.21
609.03
648.53 688.54
527.20352.13
Z1
Y1
-
-
B3-SO3
2-
Z3
2-
B2-SO3
-
Z4
2-
[M-3H]3-
B4
2-
MS3
B5-SO3
B5
2-
2-
Z3
2--H2O
Y5
2-
M-2SO3
2-
M-SO3
2-
250 300 350 400 450 500 550 600 650 700
m/z
#
#
#
#
#
#
# #
#
*
*
*
*
*237.36
B1
-B1
-
282.26
300.28
307.21
370.18
440.14
458.10
484.69
489.16
576.19
536.21
609.03
648.53 688.54
527.20352.13
Z1
Y1
-
-
B3-SO3
2-
Z3
2-
B2-SO3
-
Z4
2-
[M-3H]3-
B4
2-
MS3
B5-SO3
B5
2-B5
2-
2-
Z3
2--H2O
Y5
2-
M-2SO3
2-
M-SO3
2-
250 300 350 400 450 500 550 600 650 700
m/z
#
#
#
#
#
#
# #
#
*
*
*
*
*
-GlcA-O-GalNAc-O-IdoA-O-GalNAc-O-IdoA-O-GalNAc
SO3 SO3 SO3 SO3
B1B3
B5
Z4Y5Y1 Z1
B4B2
Y3 Z3
* * *
##
#
# #
#
##
Y5-2SO32-#
250 300 350 400 450 500 550 600 650
237.32
B1
-B1
-
300.24
Y1
-
282.19
Z1
-
B3
2-B3
2-
346.41
Z3
2-Z3
2-
370.15
379.10
Y3
2-Y3
2-
409.24
M-2SO3
440.16
B2-SO3
-
449.25
2-
M-SO3
2-
458.12
Z2
[M-3H]3-
B2
-B2
-
520.11
C2
-C2
-
538.12
-
B3-SO3
615.42
-
C3-SO3
633.38
-
m/z
*
* *
*
#
#
#
#
#
#
#
#
#
¥
-GlcA-O-GalNAc-O-IdoA-O-GalNAc
SO3 SO3 SO3
B1
B3
B2 C2
C3
Y3 Z2 Z1Y1Z3
* * *
*
##
# # #
#
127
To elucidate the hexasaccharide structure and the additional sulfate group position,
the tri-sulfated *B4- fragment ion [4,5-Δ-GlcAGalNAc(IdoAGalNAc)](3S) was
further fragmented by CID MS3 (Fig. 3.1.16b). Remarkably, all of the three possible
*B- sequence ions were generated. These over-sulfated fragments, together with the
regularly sulfated #Y- and #Z- ions documenting the sequences from the reducing
end substantiate a structure in which GlcA residue bears the fourth sulfate ester
group.
Unlike CS-rich hexasaccharide fraction, DS-rich fraction was found to contain an
under-sulfated hexamer detected in MS1 as a triply deprotonated molecule at m/z
404.78. This ion was assigned according to mass calculation a composition of [4,5-
Δ-GlcAGalNAc(IdoAGalNAc)3](1S). To analyze the structure of this mono-sulfated
hexamer not described previously and to determine the position of the sulfate
group, the ion at m/z 404.78 was isolated within an isolation window of 2 u and
submitted to CID MS2. The fragmentation spectrum and the pathway as deduced
from the ion assignment are shown in Fig. 3.1.17.
Fig.3.1.17. ESI HCT CID MS2 structural analysis of monosulfated [4,5--GlcAGalNAc(IdoAGalNAc)2] detected as a triply charged ion at m/z 404.76 by MS screening. Inset: Proposed structure, its fragmentation pathway and observed product ions. #regularly sulfated and ¥under-sulfated fragment ions [OP51]
m/z
157.14
175.21
202.18
200 300 400 500 600 700 800
220.15
378.27
404.76
440.22
497.33
606.66793.19B1
-
C1-
[M-3H]3-
Y1-
Z1-
#B2-
Z2-
C52-
[M-2H]2-458.24
418.93
B42-¥
#C2-
Y4-¥
¥
#
#
¥
¥
¥
Y52-¥
527.29
306.37
B3
2-#
m/z
157.14
175.21
202.18
200 300 400 500 600 700 800
220.15
378.27
404.76
440.22
497.33
606.66793.19B1
-
C1-
[M-3H]3-
Y1-
Z1-
#B2-
Z2-
C52-
[M-2H]2-458.24
418.93
B42-¥
#C2-
Y4-¥
¥
#
#
¥
¥
¥
Y52-¥
527.29
306.37
B3
2-B3
2-#
-GlcA-O-GalNAc-O-IdoA-O-GalNAc-O-IdoA-O-GalNAc
SO3
B1 C1B2 B4
# ¥C5
Z1Y1
Y4 Z2
C2# ¥
# #
¥ ¥ ¥¥Y5¥
# B3
128
As can be seen, CID generated a ¥Y4- ion which is diagnostic for a non-sulfated
tetrasaccharide sequence and a ¥Y5- ion diagnostic for a mono-sulfated
[GalNAc(IdoAGalNAc)2] motif from the reducing end. While #B1- and #C1-
fragment ions which characterize the non-reducing end provide evidence that 4,5-
Δ-GlcA is not sulfated, #B2- and #C2- are consistent with a mono-sulfated [4,5-Δ-
GlcAGalNAc] composition. These sequence ions indicate unequivocally that the
sulfate group is located at the first GalNAc unit from the non-reducing end, a
concept supported by all fragment ions generated from the reducing and non-
reducing end (inset Fig. 3.1.17).
To increase the experiment throughput, sensitivity and spray stability necessary for
efficient screening and sequencing of longer CS/DS chains and to reduce the in-
source loss of the labile sulfate groups often reported as the main downside of the
ESI MS method, the work was further conducted towards the introduction in GAG
research of chip-based nanoESI MS. Thus, an analytical platform that combines
size-exclusion chromatography (SEC) for fractionation and fully automated chip-
based nanoESI (NanoMate robot) coupled to QTOF MS and CID MS/MS was
developed and optimized for GAG analysis [OP52].
Figure 3.1.18. Strategy for decorin CS/DS extraction, purification, separation by SEC and chip nanoESI QTOF MS structural analysis [OP52]
Human
fibroblasts
Extraction and
purification
DECORIN
β-elimination
Free CS/DS chain
Purification
•DEAE-anion exchange
chromatography
•Ethanol precipitation
•HNO2 digestion
Pure CS/DS chains
IdoA-GalNAc-GlcA-GalNAc
Depolymerization with
ACI Lyase
Mixture of variable
length chains
Profiling,
fractionation
SIZE EXCLUSION
CHROMATOGRAPHYFraction collection
Screening by (-) chip-nanoESI QTOF
MS and sequencing by CID MS/MS
m/z
INTERPRETATION OF MASS SPECTRA
STRUCTURE DETERMINATION
Human
fibroblasts
Extraction and
purification
DECORIN
β-elimination
Free CS/DS chain
Purification
•DEAE-anion exchange
chromatography
•Ethanol precipitation
•HNO2 digestion
Pure CS/DS chainsPure CS/DS chains
IdoA-GalNAc-GlcA-GalNAc
Depolymerization with
ACI Lyase
Mixture of variable
length chains
Profiling,
fractionation
SIZE EXCLUSION
CHROMATOGRAPHY
SIZE EXCLUSION
CHROMATOGRAPHYFraction collection
Screening by (-) chip-nanoESI QTOF
MS and sequencing by CID MS/MS
m/zm/z
INTERPRETATION OF MASS SPECTRA
STRUCTURE DETERMINATION
INTERPRETATION OF MASS SPECTRA
STRUCTURE DETERMINATION
129
The strategy presented in Fig. 3.1.18 was applied to decorin-released CS/DS hexa-,
octa- and decasaccharides from human skin fibroblasts [OP52].
CS/DS chain of decorin from human skin fibroblasts was released by reductive β-
elimination reaction and digested with chondroitin AC I lyase. Enzymatic
hydrolysis mixture of CS/DS chains was separated by SEC. Collected
octasaccharide fraction was subjected to fully automated chip-based nanoESI
QTOF MS (Fig. 3.1.19, Table 3.4.3) and tandem MS (MS/MS).
Figure 3.1.19. Fully automated chip-based (-) nanoESI QTOF MS screening of the SEC fraction containing hexa-, octa- and decasaccharides obtained after depolymerization with AC I lyase of CS/DS released from human skin decorin. Solvent: MeOH/H2O (3:2 v/v). Chip-nanoESI voltage: 1.3–1.5 kV; cone voltage 20–30 V [OP52] MS of human skin fibroblasts decorin CS/DS displayed a high complexity due to
the large variety of glycoforms, which under chipnanoESI MS readily ionized to
form multiply charged ions. Except for the regularly tetrasulfated octasaccharide,
the investigated fraction contained four additional octasaccharides of atypical
sulfation status. Two new oversulfated glycoforms and two undersulfated species
were identified. Remarkably, the series of decasaccharides discovered in the same
SEC pool was found to encompass a trisulfated and a novel hexasulfated [4,5-Δ-
GlcAGalNAc(IdoAGalNAc)4] species. MS/MS by collision-induced dissociation
400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550m/z0
100
%
x2x3 418.08
404.78
405.11
405.44
405.77
418.32
438.12
418.57
425.99
426.18
438.36
438.61
442.81
452.59
511.30
458.84
498.01459.09
463.33
484.68
498.26
498.51
498.76
511.63
537.24
511.96
512.28
512.63
537.50
546.25
546.51
548.27
474.02
478.37
458.15
478.12
400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550m/z0
100
%
x2x3 418.08
404.78
405.11
405.44
405.77
418.32
438.12
418.57
425.99
426.18
438.36
438.61
442.81
452.59
511.30
458.84
498.01459.09
463.33
484.68
498.26
498.51
498.76
511.63
537.24
511.96
512.28
512.63
537.50
546.25
546.51
548.27
474.02
478.37
458.15
478.12
453.59
404.78
418.08
425.99
438.12
458.15
474.02
478.12
484.68
498.01
511.30
537.24
546.25
442.81
463.33
400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550m/z0
100
%
x2x3 418.08
404.78
405.11
405.44
405.77
418.32
438.12
418.57
425.99
426.18
438.36
438.61
442.81
452.59
511.30
458.84
498.01459.09
463.33
484.68
498.26
498.51
498.76
511.63
537.24
511.96
512.28
512.63
537.50
546.25
546.51
548.27
474.02
478.37
458.15
478.12
400 410 420 430 440 450 460 470 480 490 500 510 520 530 540 550m/z0
100
%
x2x3 418.08
404.78
405.11
405.44
405.77
418.32
438.12
418.57
425.99
426.18
438.36
438.61
442.81
452.59
511.30
458.84
498.01459.09
463.33
484.68
498.26
498.51
498.76
511.63
537.24
511.96
512.28
512.63
537.50
546.25
546.51
548.27
474.02
478.37
458.15
478.12
453.59
404.78
418.08
425.99
438.12
458.15
474.02
478.12
484.68
498.01
511.30
537.24
546.25
442.81
463.33
130
Table 3.4.3. Assignment of the molecular ions corresponding to CS/DS species detected by chip nanoESI QTOF MS in the human skin decorin [OP52]
The assignment indicates DS species of repeating (IdoAGalNAc) unit with terminal CS disaccharide (GlcAGalNAc). The CS disaccharide is bearing a double bond (4,5-Δ) induced by the specific eliminative action of AC lyase. # = undersulfated species; & = regularly sulfated species (1 SO3/disaccharide unit); * = oversulfated species; n.a = not assigned; nS = nSO3.
(CID) on the [M-4H]4- ion corresponding to the previously not reported [4,5-Δ-
GlcAGalNAc(IdoAGalNAc)3](5S) corroborated for a novel motif in which three
GalNAc moieties are monosulfated, 4,5-ΔGlcA and the first IdoA from the non-
reducing end bear one sulfate group each, while the second N-acetylgalactosamine
from the reducing end is unsulfated (Figure 3.1.20).
Basically, the combination of SEC and chip-nanoESI QTOF MS and CID MS/MS
leaded to the identification of up to decamers and provided a comprehensive view
upon the structural characteristics of a novel, previously undetected and
unreported oversulfated DCN CS/DS octasaccharide species.
The elevated heterogeneity of the structural motifs and atypical
ionization/fragmentation conditions, which are more difficult to be fulfilled by
high-throughput experiments with automatic infusion, made so far CS/DS a class
m/z Type of ion Composition
404.78 # [M-3H]3- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)2](1S)
418.08 # [M-4H]4- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)3](2S)
425.99 # [M-5H]5- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)4](3S)
438.12 # [M-4H]4- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)3](3S)
442.81 [M-4H]4- n.a
453.59 & [M-4H]4- [GlcAGalNAc(IdoAGalNAc)3](4S)
458.15 & [M-4H]4- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)3](4S)
463.33 [M-4H]4- n.a
474.02 * [M-5H]5- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)4](6S)
478.12 * [M-4H]4- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)3](5S)
484.68 * [M-3H]3- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)2](4S)
498.01 * [M-4H]4- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)3](6S)
511.30 * [M-3H]3- [4,5-Δ-GlcAGalNAc(IdoAGalNAc)2](5S)
537.24 # [M-4H]4- [GlcAGalNAc(IdoAGalNAc)4](3S)
546.25 [M-4H]4- n.a
131
Figure 3.1.20. Proposed structure of the pentasulfated octasaccharide according to the CID MS/MS data. The fragment ions diagnostic for SO3 localization are highlighted. # = undersulfated fragment ions; & = regularly sulfated fragment ions (1 SO3/disaccharide unit); * = oversulfated fragment ions; nS = nSO3 [OP52] of glycans less amenable to modern automated chip-ESI systems. In this context, it
was demonstrated here that optimized screening and sequencing procedures may
lead to successful implementation of this technology also in CS/DS field, with
superior results not only in terms of sensitivity, reproducibility and speed of
analysis but also of obtained structural information.
Remarkably, though the feasibility in GAG analysis of chip-based nanoESI could be
demonstrated before, within the same research only on shorter CS/DS chains, by
the present protocol, spectra of elevated signal-to-noise ratio could be produced
even for long CS/DS chains, within only a few minutes of signal acquisition and
with considerable reduction in ample consumption. Thus, in these experiments, the
analysis sensitivity was situated in the low picomole range.
Such a fast, accurate and sensitive analytical method appears ideal for GAG
structural elucidation as it may compensate not only the time invested in the rather
laborious sample preparation and SEC procedures but also the limited separation
efficiency exhibited by SEC and the inevitable loss of material during the
purification steps. Additionally, the spray stability and the particularly efficient
C2
GlcA-O-GalNAc-O-IdoA-O-GalNAc-O-IdoA-O-GalNAc-O-IdoA-O-GalNAc
SO3SO3 SO3SO3SO3 SO3SO3
B1 C1 B2 B3 B4 C4 B5 C5 B6 B7C7
Y1 Z1Y1 Z1Y5 Z5Y6Y6
Y7 Z7
237.52 255.34 268.84
2-1059.13
282.39
(1S) (1S) (2S) (4S)
(1S)
300.42
(1S)
300.42
(1S)
1139.32
(2S)
1139.32
(2S)
553.06
(4S)
553.06
(4S)1119.31
(2S)
1119.31
(2S)
837.27
(4S)
837.27
(4S)
2-3-
SO3
387.41
2-
(3S)
1-1-
1-1- 1-
1-1-
1-
273.35
(1S)
(Y6/Z5)
#
# & && &
*
***** *
(3S)
#
#
& & & *
806.89
2-
(4S)
&
132
ionization attainable under mild values of the ESI source parameters enhanced the
formation of multiply charged ions, and prevented the in-source loss of sulfate
groups.
These aspects, together with the method sensitivity, had beneficial consequences
upon the detection of species having different number of sulfates and/or close
chain lengths, which, because of the limitations exhibited by SEC method, were
collected together with the regularly sulfated octasaccharide. It is, therefore,
possible to positively summarize that in the field of CS/DS, this methodology
represents a viable alternative to classical capillary-based ESI MS protocols. It is
also obvious that the microfluidics-MS methodology has real perspectives to be
introduced in the near future as a routine analytical method in glycomics as
demonstrated also by the other original studies carried out within this work [OP53-
OP62].
By all technical and technological achievements described above, electrospray mass
spectrometry definitively crossed today the border to biophysics, biochemistry,
molecular biology and medicine. The general ways that it may offer structural
determination, identification, sequencing and trace level analysis find daily novel
applications in these fields.
133
SECTION II
PART IV
Concluding remarks and perspectives “Because the technology provides the tools and biology the problems, the two should enjoy a happy
marriage.” Stanley Fields
4.1. Concluding remarks
In the post-genome era, the trend in all analytical sciences is toward
miniaturization of devices and high-throughput analyses. In bioscience, integrated,
fully automated micro and nanosystems have been demonstrated to provide one of
the most rapid, sensitive and accurate analysis. Mass spectrometry has the potential
to revolutionize the bioanalytical field in general and the glycoanalytical one in
particular and consequently help in understanding many essential biological
phenomena and events.
Due to the above two observations, the implementation of the modern
microfluidic devices and chip-based technology is the purpose of the current
research in the field and massive efforts are invested for the routinely introduction
of the “lab-on-a-chip” principle in MS. The high potential of these systems to
discover novel structures of biological importance makes them ideal for
identification of unknown, minor components in complex mixtures. Furthermore,
the capability of structural elucidation of biomolecules possibly indicative of
pathological states gives this method clear perspectives for use in clinical
diagnostics.
Biological microfluidics/electrospray mass spectrometry, though at its very
beginning, is a nice example of the technology/biology happy “honeymoon” and
there is no reason for which, a long happy “marriage” cannot be predicted.
134
4.2. Plans for further research and career development
On medium term, the plans for research are correlated to the objectives of
the following three projects, of which I am the principal investigator (project
director or leader of the partner group):
CNCS-UEFISCDI-PCE-Ideas-2011-Nr. 0047, CNCS-UEFISCDI-PCE-Ideas-
2011-Nr. 0047; DEVELOPMENT OF CHIP-BASED NANOELECTROSPRAY
IN COMBINATION WITH ELECTRON TRANSFER DISSOCIATION MASS
SPECTROMETRY FOR TOP-DOWN GLYCOPROTEOMICS (2011-2014)
ANCS-UEFISCDI-PN-II-PT-PCCA-2011-3.1-0187; FUNCTIONALIZED
POLYSACCHARIDES FOR APPLICATIONS IN BIOMEDICINE AND
BIOTECHNOLOGY (2012-2015)
FP7 MARIE CURIE-PIRSES-GA-2010-269256; INTEGRATING HIGH
PERFORMANCE MASS SPECTROMETRY TOOLS WITH APPLICATIONS
IN LIFE SCIENCE (2012-2015)
I. Technical objectives/targets
In the next years the scientific activity will conducted towards:
a) first world merging of robotics and microfluidics (chip) as front-end technology
for nanoESI with robotic infusion, electron transfer dissociation technique and
automatic fragmentation by ETD/CID switching;
b) development of innovative protocols for high performance microfluidics/MS in
combination with ETD top-down protein and glycoprotein fragmentation will be
for the first time designed, tested and implemented in proteomics to allow rapid
and accurate protein identification and structural characterization in a single run at
superior sensitivity, accuracy, reproducibility of the experiments, spectral data
reliability and confidence due to a high sequence coverage by diagnostic ions;
c) novel glycoscreening and sequencing protocols (based on ETD and proton
transfer reaction-PTR) by microfluidics-MS techniques for natural and
functionalized oligo- and polisaccharides;
135
d) high performance liquid chromatography (HPLC) coupling on-line to MS via
NanoMate robot for separation of the complex glycan mixtures followed by on-line
chip-based nanoESI detection and fragmentation by CID, ETD and PTR MSn of
individual components;
e) development of bioinformatics platforms (computer software) to allow
automatic interpretation of ESI MS, CID MSn and ETD and PTR MSn data.
Currently, my research group conceived within our projects a database and a
computer software for automatic interpretation and assignment of
ganglioside/glycosphigolipid mass spectra. For the next few years we have
planned the extention of the software capabilities by introduction of portals for
assignment of the screening and sequencing mass spectra of glycosaminoglycans,
O- and N-glycans as well as O- and N-glycopeptides;
e) optimization of high-throughput glycan screening and sequencing with
computer assisted data analysis.
II. Scientific objectives/targets
The novel analytical platforms will be applied to:
1. structural identification of peptides, glycopeptides, proteins and glycoproteins
which will be infused by robotized chip-nanoESI MS, screened by MS and
fragmented by multistage ETD and automatic alternate ETD/CID and PTR to
collect in a single experiment of high sensitivity and speed of analysis the full set of
data upon the structure in terms of sequence and post-translational modification
structure and site(s).
2. Validation and performance of the developed methods for routine applicability
in clinical and biomedical research will be performed for complex mixtures from
blood and urine of patients suffering from lisosomal storage diseses (Fabry,
Gaucher, Pompe, Schindler etc.) and congenital disorder of glycosylation (CDG)
characterized by abnormal accumulation of O-glycans in various tissues and body
fluids because of deficient activity of specific enzymes. As no efficient therapeutic
schemes and diagnostic protocols are currently available for these diseses, accurate
136
determination of glycan expression and structure in the multicomponent samples
extracted from patients’ blood and urine vs. healthy age-matched controls will be
essential for understanding of the molecular bases of the disease and elaboration of
adequate treatment. The topic importance for the field of Life Sciences is
represented by the integration of analytical technologies of highest performance
and their subsequent conversion into a diagnostically-operative system of routine
use in molecular medicine. The progress beyond the state-of-art in the field this
research will bring out resides in the following main concepts and ideas to be
elaborated, translated into research and implemented: i) first top-down protein and
glycoprotein analysis by ETD and first top-down in high throughput mode; ii) first
combination of fully automated chip-based MS with ETD and ETD alternating with
CID; iii) first protein and protein mixture and glycoprotein screening and
sequencing in intact form and without prior separation by direct infusion via fully
automated chip-based MS ETD and ETD/PTR/CID; iv) first characterization of
post-translational modifications (in particular glycosylation) by direct infusion via
fully automated chip-based MS and fragmentation in multistage ETD and
ETD/PTR/CID.
3. to introduce for the first time microfluidics-MS technology for the analysis of
long chain polysaccharides and functionalized glycans; the newly developed
methods will be also applied to polydisperse glycans functionalized at the
preparative scale with oligo- and polypeptides, chromophores and aromatic
amines, which represent derivatives with amphiphilic properties, enabling
biochemical or biological reactions at the polymer surface;
4. based on the previously acquired knowledge on the biomarker role in CNS
afflictions played by gangliosides and glycolipids [OP18, OP21, OP24, OP25, OP42]
these advanced microfluidics-MS methods will be further implemented and
optimized for the determination of ganglioside expression and structure, with the
characterization of the relevant biomarker species, in other brain diseases such as
astrocytoma, neuroblastoma etc. in comparison with tumor surrounding tissue and
healthy brain tissue;
137
5. in the following years the research will also continue the previous work on the
development of novel glycomics and glycoproteomics methods based on high
performance mass spectrometry and related hyphenated robotics and microfluidics
(chip) technology for biomarker discovery in human blood (serum) and
cerebrospinal fluid (CSF), at an early stage of brain cancer. Cerebrospinal fluid,
which was even less explored for glyco-antigens than sera, represents a
considerably better diagnostically/prognostically relevant system for brain tumour
diseases, since all metabolic changes occurring in brain are directly reflected in the
CSF. Therefore the earliest occurring tumour-specific markers relevant for an early
diagnosis of brain tumours are first expected in the CSF. The followed markers will
be gangliosides and glycosaminoglycans, the latter ones, on the basis of our
previous investigation, which leaded to their first MS-based discovery in central
nervous system [OP29].
138
SECTION III
List of own publications
OP1. A.D. Zamfir, J. Peter-Katalinić, Capillary electrophoresis-mass spectrometry
for glycoscreening in biomedical research, Electrophoresis 25, 1949-1963, 2004.
OP2. A.D. Zamfir, L. Bindila, N. Lion, M. Allen, H. H. Girault, J. Peter-Katalinic,
Chip electrospray mass spectrometry for carbohydrate analysis, Electrophoresis 26,
3650-3673, 2005.
OP3. A.D. Zamfir, C. Flangea, A.Serb, A.-M. Zagrean, A. Rizzi, E. Sisu, Separation
and identification of glycoforms by capillary electrophoresis with electrospray
ionization mass spectrometric detection; in Mass Spectrometry of Glycoproteins:
Methods and Protocols, Methods Mol. Biol. 951, 145-169, 2013.
OP4. A.D. Zamfir, Recent advances in sheathless interfacing of capillary
electrophoresis and electrospray ionization mass spectrometry, J. Chromatogr. A
1159, 2–13, 2007.
OP5. A.D. Zamfir, N. Lion, Ž. Vukelic, L. Bindila, J. Rossier, H.Girault, J.Peter-
Katalinić, Thin chip microsprayer system coupled to quadrupole time-of-flight
mass spectrometer for glycoconjugate analysis, Lab. Chip 5, 298-307, 2005.
OP6. C. Flangea, A. Serb, E. Sisu, A.D. Zamfir, Chip-based mass spectrometry of
brain gangliosides, Biochim. Biophys. Acta (Molec & Cell Biol. of Lipids) 1811, 513–535,
2011.
OP7. A.D. Zamfir, C.Flangea, F.Altmann, A. M. Rizzi, Glycosylation analysis of
proteins, proteoglycans and glycolipids by CE-MS, Adv. Chromatogr. 49, 135-186,
2011.
OP8. A.D. Zamfir, N. Dinca, E.Sisu, J.Peter-Katalinić, Copper-coated microsprayer
interface for on-line sheathless capillary electrophoresis electrospray mass
spectrometry of carbohydrates,J. Sep. Science 29, 414-422, 2006.
OP9. E. Sisu, C. Flangea, A. Serb, A. Rizzi, A. D. Zamfir, High-performance
separation techniques hyphenated to mass spectrometry for ganglioside analysis,
Electrophoresis 32, 1591-1609, 2011.
139
OP10. A.D. Zamfir, D. Seidler, E. Schonherr, H. Kresse, J. Peter-Katalinić, On-line
sheathless capillary electrophoresis/nanoelectrospray ionization-tandem mass
spectrometry for the analysis of glycosaminoglycan oligosaccharides, Electrophoresis
25, 2010-2016, 2004.
OP11. S. Y. Vakhrushev, A. D. Zamfir, J. Peter-Katalinić, 0,2An cross-ring cleavage
as a general diagnostic tool for glycan assignment in glycoconjugate mixtures, J.
Am. Soc. Mass Spectrom. 15, 1863-1868, 2004.
OP12. A.D. Zamfir, J. Peter-Katalinić, Glycoscreening by sheathless on-line capillary
electrophoresis/electrospray quadrupole time-of-flight tandem mass spectrometry,
Electrophoresis 22, 2448-2457, 2001.
OP13. A.D. Zamfir, S. Vakhrushev, A. Sterling, H. Niebel, M. Allen, J. Peter-
Katalinić, Fully automated chip-based mass spectrometry for complex
carbohydrate system analysis, Anal. Chem. 76, 2046-2054, 2004.
OP14. M. Froesch, L. Bindila, A.D. Zamfir, J. Peter-Katalinić, Sialylation analysis of
O-glycosylated sialylated peptides from urine of patients suffering from Schindler's
disease by Fourier transform ion cyclotron resonance mass spectrometry and
sustained off-resonance irradiation collision-induced dissociation, Rapid Commun.
Mass Spectrom. 17, 2822-2832, 2003.
OP15. M. Froesch, L. Bindila, G. Baykut, M. Allen, J. Peter-Katalinić, A.D. Zamfir,
Coupling of fully automated chip electrospray to Fourier transform ion cyclotron
resonance mass spectrometry for high-performance glycoscreening and
sequencing, Rapid Commun. Mass Spectrom. 18, 3084-3092, 2004.
OP16. L. Bindila, M. Froesch, N. Lion, Ž. Vukelic, J. Rossier, H. Girault, J. Peter-
Katalinić, A.D. Zamfir, A thin chip microsprayer system coupled to Fourier
transform ion cyclotron resonance mass spectrometry for glycopeptide screening,
Rapid Commun. Mass Spectrom. 18, 2913-2920, 2004.
OP17. C. Flangea, C. Schiopu, F. Capitan, C. Mosoarca, M. Manea, E.Sisu, A.D.
Zamfir, Fully automated chip-based nanoelectrospray combined with electron
transfer dissociation for high throughput top-down proteomics, Cent. Eur. J. Chem.
11, 25-34, 2013.
140
OP18. R. Almeida, C. Mosoarca, M. Chirita, V. Udrescu, N. Dinca, Ž. Vukelić, M.
Allen, A.D. Zamfir, Coupling of fully automated chip-based electrospray
ionization to high capacity ion trap mass spectrometer for ganglioside analysis,
Anal. Biochem. 378, 43–52, 2008.
OP19. A. Serb, C. Schiopu, C. Flangea, Ž. Vukelić, E. Sisu, L. Zagrean, A.D. Zamfir,
High-throughput analysis of gangliosides in defined regions of fetal brain by fully
automated chip-based nanoelectrospray ionization multistage mass spectrometry,
Eur. J. Mass Spectrom. 15, 541-553, 2009.
OP20. A.Serb, C. Schiopu, C. Flangea, E. Sisu, A.D. Zamfir, Top-down
glycolipidomics: fragmentation analysis of ganglioside oligosaccharide core and
ceramide moiety by chip-nanoelectrospray collision-induced dissociation MS2-MS6,
J. Mass Spectrom. 44, 1434–1442, 2009.
OP21. C. Schiopu, A. Serb, F. Capitan, C. Flangea, E. Sisu, Z. Vukelic, M.
Przybylski, A. D. Zamfir, Determination of ganglioside composition and structure
in human brain hemangioma by chip-based nanoelectrospray ionization tandem
mass spectrometry, Anal. Bioanal. Chem. 395, 2465-2477, 2009.
OP22. A. D. Zamfir, Ž.Vukelić, A. Schneider, E. Sisu, N. Dinca, A. Ingendoh, A
novel approach for ganglioside structural analysis based on electrospray multiple
stage mass spectrometry, J. Biomolec. Techn. 18, 188–193, 2007.
OP23. C. Mosoarca, R.M. Ghiulai, C. R. Novaconi, Ž. Vukelić, A.Chiriac, A.D.
Zamfir, Application of chip-based nanoelectrospray ion trap mass spectrometry
to compositional and structural analysis of gangliosides in human fetal
cerebellum, Anal. Lett. 44, 1036-1049, 2011.
OP24. A. D. Zamfir, A. Serb, Ž. Vukelić, C. Flangea, C. Schiopu, D. Fabris, F.
Capitan, E. Sisu, Assessment of the molecular expression and structure of
gangliosides in brain metastasis of lung adenocarcinoma by an advanced approach
based on fully automated chip-nanoelectrospray mass spectrometry, J. American
Soc. Mass Spectrom. 22, 2145-2159, 2011.
OP25. C. Schiopu, Ž. Vukelić, F. Capitan, E. Sisu, A.D. Zamfir, Chip-
nanoelectrospray quadrupole time-of-flight tandem mass spectrometry of
meningioma gangliosides: A preliminary study, Electrophoresis, 33, 1778-1786, 2012.
141
OP26. A.F. Serb, E. Sisu, Z. Vukelić, A.D. Zamfir, Profiling and sequencing of
gangliosides from human caudate nucleus by chip-nanoelectrospray mass
spectrometry. J. Mass Spectrom. 47, 1561-70, 2012.
OP27. T.Visnapuu, A.D. Zamfir, C. Mosoarca, M. D. Stanescu, T.Alamäe, Fully
automated chip-based negative mode nanoelectrospray mass spectrometry of
fructooligosaccharides produced by heterologously expressed levansucrase from
Pseudomonas syringae pv. tomato DC3000, Rapid Commun. Mass Spectrom. 23, 1337–
1346, 2009.
OP28. C. Flangea, A. Serb, C. Schiopu, S. Tudor, E. Sisu, D. G. Seidler, A.D. Zamfir,
Discrimination of GalNAc (4S/6S) sulfation sites in chondroitin sulfate
disaccharides by chip-based nanoelectrospray multistage mass spectrometry, Cent.
Eur. J. Chem. 7, 752–759, 2009.
OP29. C. Flangea, C. Schiopu, E. Sisu, A.Serb, M. Przybylski, D. G. Seidler, A. D.
Zamfir, Determination of sulfation pattern in brain glycosaminoglycans by chip-
based electrospray ionization ion trap mass spectrometry, Anal. Bioanal. Chem. 395,
2489-2498, 2009.
OP30. T. Visnapuu, K. Mardo, C. Mosoarca, A.D. Zamfir, A. Vigants, T. Alamäe,
Levansucrases from Pseudomonas syringae pv. tomato and P. chlororaphis subsp.
aurantiaca: Substrate specificity, polymerizing properties and usage of different
acceptors for fructosylation, J. Biotechnol. 155, 338-349, 2011.
OP31. C. Flangea, E. Sisu, D.G. Seidler, A.D. Zamfir, Analysis of oversulfation in
biglycan chondroitin/dermatan sulfate oligosaccharides by chip-based
nanoelectrospray ionization multistage mass spectrometry, Anal. Biochem. 420, 155–
162, 2012.
OP32. C. Herzog, I. Lippmann, K. Grobe, A.D. Zamfir, F. Echtermeyer, D.G.
Seidler, The amino acid tryptophan prevents the biosynthesis of dermatan sulfate,
Mol. Biosyst. 7, 2872-2881, 2011.
OP33. I.M.C Ienascu, A.X. Lupea, I.M. Popescu, M.A. Padure, A.D. Zamfir, The
synthesis and characterization of some novel 5-chloro-2-(substituted alkoxy)-N-
phenylbenzamide derivatives, J. Chem. Serb. Soc. 74, 847-855, 2009.
142
OP34. D. Condrat, C. Mosoarca, A.D. Zamfir, F. Crişan, M. Szabo, A. Lupea,
Qualitative and quantitative analysis of gallic acid in Alchemilla vulgaris, Allium
ursinum, Acorus calamus and Solidago virga-aurea by chip-electrospray ionization
mass spectrometry and high performance liquid chromatography, Cent. Eur. J.
Chem. 8, 530–535, 2010.
OP35. M.D. Stanescu, F. Harja, C. Mosoarca, A.D. Zamfir, Biogenic amines
fingerprints evidenced by performant MS analysis, Rev. Roum. Chim. 55, 1053-1059,
2010.
OP36. M. Stefanut, A. Cata, R. Pop, C. Mosoarca, A.D. Zamfir, Anthocyanins
HPLC-DAD and MS characterization, total phenolic and antioxidant activity of
some berries extractsm Anal. Lett. 44, 2843-2855, 2011.
OP37. L. Bindila, J. Peter-Katalinić, A.D. Zamfir, Sheathless reverse polarity
capillary electrophoresis/electrospray mass spectrometry for the analysis of
underivatized glycans, Electrophoresis 26, 1488-1499, 2005.
OP38. L. Bindila, R. Almeida, A. Sterling, M. Allen, J. Peter-Katalinić, A.D. Zamfir,
Off-line capillary electrophoresis/fully automated chip-based electrospray
ionization quadrupole time-of-flight mass spectrometry and tandem mass
spectrometry for glycoconjugate analysis, J. Mass Spectrom. 39, 1190-1201, 2004.
OP39. A.D. Zamfir, Ž. Vukelic, J. Peter-Katalinić, A capillary electrophoresis and
off-line capillary electrophoresis/electrospray ionization quadrupole time-of-flight
tandem mass spectrometry approach for ganglioside analysis, Electrophoresis 23,
2894-2903, 2002.
OP40. Ž.Vukelic, M.Zarei, J.Peter-Katalinić, A.D. Zamfir, Analysis of human
hippocampus gangliosides by fully-automated chip-based nanoelectrospray
tandem mass spectrometry, J. Chromatogr. A. 1130, 238-245, 2006.
OP41. A.D. Zamfir, Z. Vukelic, L. Bindila, R. Almeida, A. Sterling, M. Allen, J.
Peter-Katalinić, Fully automated chip-based nanoelectrospray tandem mass
spectrometry of gangliosides from human cerebellum, J. American Soc. Mass
Spectrom. 15, 1649-1657, 2004.
OP42. Ž. Vukelić, S. Kalanj Bognar, M.Froesch, L. Bindila, B. Radić, M. Allen, J.
Peter-Katalinić, A.D. Zamfir, Human gliosarcoma-associated ganglioside
143
composition is complex and highly distinctive as evidenced by high-performance
mass spectrometric determination and structural characterization, Glycobiology 17,
504-515, 2007.
OP43. Ž. Vukelic*, A.D. Zamfir *, L. Bindila, M. Froesch, S. Usuki, R.Yu, J. Peter-
Katalinić (*equal contribution), Screening and sequencing of complex sialylated
and sulfated glycosphingolipid mixtures by negative ion electrospray Fourier
transform ion cyclotron resonance mass spectrometry, J. American Soc. Mass
Spectrom. 16, 571-580, 2005.
OP44. E. Sisu, C. Flangea, A.Serb, A. D. Zamfir, Modern developments in mass
spectrometry of chondroitin and dermatan sulfate glycosaminoglycans, Amino
Acids, 41, 235-256, 2011.
OP45. S. Amon, A.D. Zamfir, A. Rizzi, Glycosylation analysis of glycoproteins and
proteoglycans using CE–MS strategies, Electrophoresis, 29, 2485-2507, 2008.
OP46. A. D. Zamfir, C. Flangea, A.Serb, E. Sisu, L. Zagrean, A. Rizzi, D.G. Seidler,
Brain chondroitin/dermatan sulfate: from cerebral tissue to fine structure.
Extraction, preparation and fully automated chip-electrospray mass spectrometric
analysis in: Proteoglycans: Methods and Protocols, Methods Mol. Biol. 836, 145-159,
2012.
OP47. D. G. Seidler, J. Peter-Katalinić, A.D. Zamfir, Galactosaminoglycan function
and oligosaccharide structure determination, Scient. World J. 19, 233-241, 2007.
OP48. M. Morman*, A. D. Zamfir*, D.G.Seidler, H.Kresse, J. Peter-Katalinić (*equal
contribution), Analysis of oversulfation in a dermatan sulfate oligosaccharide
fraction from bovine aorta by nanoelectrospray ionization quadrupole time-of-
flight and Fourier-transform ion cyclotron resonance mass spectrometry, J.
American. Soc. Mass Spectrom. 18, 179-187, 2007.
OP49. A. D. Zamfir, D. Seidler, H. Kresse, J. Peter-Katalinić, Structural
characterization of chondroitin/dermatan sulfate oligosaccharides from bovine
aorta by capillary electrophoresis and electrospray ionization quadrupole time-of-
flight tandem mass spectrometry, Rapid Commun. Mass Spectrom. 16, 2015-2024,
2002.
144
OP50. A.D. Zamfir, D. Seidler, H. Kresse, J. Peter-Katalinić, Structural investigation
of chondroitin/dermatan sulfate oligosaccharides from human skin fibroblast
decorin, Glycobiology 11, 733-742, 2003.
OP51. A.D. Zamfir, C. Flangea, A. Serb, E.Sisu, N. Dinca, P. Bruckner, D. G. Seidler,
Analysis of novel over- and undersulfated glycosaminoglycan sequences by
enzyme cleavage and multiple stage mass spectrometry, Proteomics 9, 3435-3444,
2009.
OP52. A. D. Zamfir, C. Flangea, D. G. Seidler, E. Sisu, J. Peter-Katalinić, Combining
size-exclusion chromatography and fully automated chip-based nanoelectrospray
quadrupole time-of-flight tandem mass spectrometry for structural analysis of
chondroitin/dermatan sulfate in human decorin, Electrophoresis 32, 1639-1646, 2011.
OP53. L. Bindila, A.D. Zamfir, J. Peter-Katalinić, Characterization of peptides by
capillary zone electrophoresis and electrospray ionization quadrupole time-of-
flight tandem mass spectrometry, J. Sep. Science 15, 1101-1111, 2002.
OP54. B. Balen, A.D. Zamfir, S. Vakhrushev, M. Krsnik-Rasol, J. Peter-Katalinić,
Determination of Mammillaria gracillis N-glycan patterns by ESI Q-TOF mass
spectrometry, Croat. Chem. Acta 78, 463-477, 2005.
OP55. E. Sisu, Wouter T.E. Bosker, Wilhem Norde, Teddy M. Slaghek, Jan W.
Timmermans, J. Peter-Katalinić, M. A. Cohen-Stuart, A.D. Zamfir, Electrospray
ionization quadrupole time-of-flight tandem mass spectrometric analysis of
hexamethilene diamine-modified maltodextrin and dextran, Rapid. Commun. Mass
Spectrom. 20, 209-218, 2006.
OP56. B. Balen, M. Krsnik-Rasol, A.D. Zamfir, J. Milosevic, S.Y. Vakhrushev, J.
Peter-Katalinić, Glycoproteomic survey of Mammillaria gracillis tissues grown in
vitro, J. Proteome Res. 5, 1658-1666, 2006.
OP57. F.H. Cederkvist*, A.D. Zamfir *, S.Bahrke, V.G. Eijsink, M.Sorlie, J.Peter-
Katalinić, M.G. Peter; (*equal contribution), Identification of a high-affinity-binding
oligosaccharide by (+) nanoelectrospray quadrupole time-of-flight tandem mass
spectrometry of a noncovalent enzyme-ligand complex, Angew. Chem. Int. Ed. Engl.
45, 2429-34, 2006.
145
OP58. B. Balen, M. Krsnik-Rasol, A.D. Zamfir, I. Zadro, S. Vakhrushev, J. Peter-
Katalinić, N-glycan heterogeneity of cactus glycoproteins induced by tissue culture
conditions, J. Biomolec. Techn. 18, 150-160, 2007.
OP59. I. Perdivara, E. Sisu, I. Sisu, N. Dinca, K.B. Tomer, M. Przybylski, A.D.
Zamfir, Enhanced electrospray ionization Fourier transform ion cyclotron
resonance mass spectrometry of long-chain polysaccharides, Rapid Commun. Mass
Spectrom. 22, 773-782, 2008.
OP60. I. Sisu, V. Udrescu, C. Flangea, S. Tudor, N. Dinca, L. Rusnac, A.D. Zamfir,
E.Sisu, Synthesis and structural characterization of amino-functionalized
polysaccharides, Cent. Eur. J. Chem. 7, 66-73, 2009
OP61. I. Shin, A. D. Zamfir, Bin Ye, Protein carbohydrate analysis: gel-based
staining, liquid chromatography, mass spectrometry, and microarray screening in:
Tissue Proteomics-Pathways, Biomarkers, and Drug Discovery, Methods Mol. Biol.
44, 19-39, 2008.
OP62. T. Alamae, T. Visnapuu, K. Mardo, A. Mae, A. D. Zamfir, Levansucrases of
Pseudomonas bacteria: novel approaches for protein expression, assay of enzymes,
fructooligosaccharides and heterooligofructans, Carbohydr. Chem. 38, 176–191, 2012.
References
[1] Zeleny, J. Phys. Rev. 1917, 10, 1.
[2] Dole, M., Mack, L., Hines, R., Mobley, R., Ferguson, L., Alice, M. J. Chem. Phys.
1968, 49, 2240.
[3] Yamashita, M., Fenn, J.B. J. Phys. Chem. 1984, 88, 4451.
[4] Aleksandrov, M.L., Gall, L.N., Krasnov, V.N., Nikolaev, V., Pavlenko, V.A.,
Shkurov, V. A. Dokl. Akad. Nauk SSSR 1984, 277, 379.
[5] Masselon, C., Pasa-Tolić, L., Tolić, N., Anderson, G.A., Bogdanov, B., Vilkov,
A.N., Shen, Y., Zhao, R., Qian, W.J., Lipton, M.S. Anal. Chem. 2005, 77, 400.
[6] Fenn, J.B. Angew. Chem. Int. Ed. Engl. 2003, 42, 3871.
[7] Wilm, M., Mann, M. Anal. Chem. 1996, 68, 1.
[8] Hanisch, F.G., Jovanović, M., Peter-Katalinić, J. Anal. Biochem. 2001, 290, 47.
146
[9] Smith, R. D., Olivares, J.A., Nguyen, N.T., Udseth, H.R. Anal. Chem. 1988, 60,
436.
[10] Olivares, J.A., Nguyen, N.T., Yonker, C.R., Smith, R. D. Anal. Chem. 1987, 59,
1230.
[11] Kriger, M.S., Cook, K. D., Ramsey, R. S. Anal. Chem. 1995, 67, 385.
[12] Ramsey, R. S., McLuckey, S.A. J. Microcol. Sep. 1995, 7, 461.
[13] Chang, Y.Z., Her, G.R. Anal Chem. 2000, 72, 626.
[14] Thorsen, T. Maerkl, S.J., Quake, S.R. Science 2002, 298, 580.
[15] Groisman, A. Enzelberger, M., Quake, S.R. Science 2003, 300, 955.
[16] Raju, B., Ramesh, M., Srinivas, R., Chandrasekhar, S., Kiranmai, N., Sarma,
V.U. J. Am. Soc. Mass Spectrom. 2011, 22, 703.
[17] Wang, H., Wong, C.H., Chin, A., Taguchi, A., Taylor, A., Hanash, S., Sekiya, S.,
Takahashi, H., Murase, M., Kajihara, S., Iwamoto, S., Tanaka, K. Nat Protoc. 2011, 6,
253.
[18] Zehl, M., Pittenauer, E., Jirovetz, L., Bandhari, P., Singh, B., Kaul, V.K., Rizzi,
A., Allmaier, G. Anal Chem. 2007, 79, 8214.
[19] Dethy, J.M., Ackermann, B.L., Delatour, C., Henion, J.D., Schultz, G.A. Anal.
Chem. 2003, 75, 805.
[20] Kim, S., Rodgers, R.P., Blakney, G.T, Hendrickson, C.L., Marshall, A.G. J. Am.
Soc. Mass Spectrom. 2009, 20, 8.
[21] Sainiemi, L., Nissilä, T., Kostiainen, R., Ketola, R.A., Franssila, S. Lab Chip. 2011,
11, 3011.
[22] Klepárník, K. Electrophoresis. 2013, 34, 70.
[23] Kohler, I., Schappler, J., Rudaz, S. Anal. Bioanal. Chem. 2013, 405,125.
[24] Ramautar, R., Somsen, G.W., de Jong, G.J. Electrophoresis. 2013, 34, 86.
[25] Bao, J., Krylov, S.N. Anal. Chem. 2012, 84, 6944.
[26] Kubáň, P., Kobrin, E.G., Kaljurand, M. J. Chromatogr A. 2012, 1267, 239.
[27] Bonvin, G., Schappler, J., Rudaz, S. J. Chromatogr. A. 2012, 1267, 17.
[28] Hirayama, A., Tomita, M., Soga, T. Analyst. 2012, 137, 5026.
[29] Huang, J.L., Hsu, R.Y., Her, G.R. J. Chromatogr. A. 2012, 1267, 131.
147
[30] Smith, R. D., Olivares, J.A., Nguyen, N.T., Udseth, H.R. Anal. Chem. 1988, 60,
436.
[31] ] Olivares, J.A., Nguyen, N.T., Yonker, C.R., Smith, R. D. Anal. Chem. 1987, 59,
1230.
[32] Nilsson, S., Svedberg, T.M., Pettersson, J., Bjorefors, T.F., Markides, K.,
Nyholm, L., Anal. Chem. 2001, 73, 4607.
[33] Kriger, M.S., Cook, K.D., Ramsey, R.S. Anal. Chem. 1995, 67, 385.
[34] Chen, Y.R., Her, G.R. Rapid Commun. Mass Spectrom. 2003, 17, 437.
[35] Kriger, M.S., Cook, K. D., Ramsey, R. S. Anal. Chem. 1995, 67, 385.
[36] Rohner, T.C., Rossier, J.S., Girault, H.H. Anal. Chem. 2001, 73, 5353.
[37] Gobry, V., van Oostrum, J., Martinelli, M., Rohner, T., Rossier, J.S., Girault,
H.H. Proteomics 2002, 2, 405.
[38] Lion, N., Gellon, J. O., Girault, H.H. Rapid Commun. Mass Spectrom. 2004, 18,
1614.
[39] Deng, Y., Henion, J., Li, J., Thibault, P., Wang, C., Harrison, D. J. Anal. Chem.
2001, 73, 639.
[40] Roussel, C., Dayon, L., Lion, N., Rohner, T.C., Josserand, J., Rossier, J.S., Jensen,
H., Girault, H. H. J. Am. Soc. Mass Spectrom. 2004, 15, 1767.
[41] Rossier, J.S., Youhnovski, N., Lion, N., Damoc, E., Reymond, F., Girault, H.H.,
Przybylski, M. Angew. Chem. Int. Edn. Engl. 2003, 42, 53.
[42] Bedair, M., Oleschuk, R., Anal. Chem. 2006, 78, 1130.
[43] Ramos-Payán, M.D., Jensen, H., Petersen, N.J., Hansen, S.H., Pedersen-
Bjergaard, S. Anal. Chim. Acta. 2012, 20, 735, 46.
[44] Petersen, N.J., Pedersen, J.S., Poulsen, N.N., Jensen, H., Skonberg, C., Hansen,
S.H., Pedersen-Bjergaard, S. Analyst. 2012, 137, 3321.
[45] Hua, Y., Jemere, A.B., Harrison, D.J. J. Chromatogr. A. 2011, 218, 4039.
[46] Vázquez, M., Paull, B. Anal. Chim. Acta. 2010, 668, 100.
[47] Bereman, M.S., Young, D.D., Deiters, A., Muddiman, D.C. J. Proteome Res. 2009,
8, 3764.
[48] Irungu, J., Go, E.P., Zhang, Y., Dalpathado, D.S., Liao, H.X., Haynes, B.F.,
Desaire, H. J. Am. Soc. Mass Spectrom. 2008, 19, 1209.
148
[49] Lourette, N., Smallwood, H., Wu, S., Robinson, E.W., Squier, T.C., Smith, R.D.,
Pasa-Tolić L. J. Am. Soc. Mass Spectrom. 2010, 21, 930.
[50] Nwosu, C.C., Strum, J.S., An, H.J., Lebrilla, CB. Anal. Chem. 2010, 82, 9654.
[51] Fukui, K., Takahashi, K. Anal. Chem. 2012, 84, 2188.
[52] Capriotti, A.L., Cavaliere, C., Foglia, P., Samperi, R., Laganà, A. J. Chromatogr.
A, 2011, 1218, 8760.
[53] Jmeian, Y., Hammad, L.A., Mechref Y. Anal. Chem. 2012, 84, 8790.
[54] Zauner G., Kozak, R.P., Gardner, R.A., Fernandes, D.L., Deelder, A.M.,
Wuhrer, M. Biol. Chem. 2012, 393, 687.
[55] Guthals, A., Bandeira, N. Mol. Cell. Proteomics. 2012, 11, 550.
[56] Yu, R.K., Nakatani, Y., Yanagisawa, M. J. Lipid Res. 2009, 50, 440.
[57] Hakomori, S.I. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 225.
[58] Domon, B., Costello, C.E. Glycoconj. J. 1988, 5, 397.
[59] Wang, A. M., Schindler, D., Desnick, R. J. J. Clin. Invest. 1990, 86, 1752.
[60] Desnick, R. J., Schindler, D. in: The metabolic and molecular bases of inherited
diseases, 8th edn. McGraw-Hill, New York, 2001, pp. 3483; D. Scriver C.R. Beaudet,
A.L., Sly, W.S., Valle, D. (eds.)
[61] Sakuraba, H., Matsuzawa, F., Aikawa, S., Doi, H., Kotani, M., Nakada, H.,
Fukushige, T., Kanzaki, T. J. Hum. Genet. 2004, 49, 1.
[62] Williger, K. Doctoral Thesis, 1994, University of Bonn.
[63] Svennerholm, L. Prog. Brain Res. 1994, 101, XI-XIV.
[64] Fuller, M. Lipids Health Dis. 2010, 9, 113.
[65] Inoue, S., Sato, C., Kitajima, K. Glycobiology 2010, 20, 752.
[66] Salminen, A., Kaarniranta, K. J. Mol. Med. 2009, 87, 697.
[67] Muthing, J. Methods Mol. Biol. 1998, 76, 183.
[68] Kasperzyk, J. L., El-Abbadi, M. M., Hauser, E. C., d’Azzo, A., Platt, F. M.,
Seyfried, T. N. J. Neurochem. 2004, 89, 645.
[69] Kusunoki, S., Kaida, K., Ueda, M. Biochim. Biophys. Acta 2008, 1780, 441.
[70] Yu, R. K., Yanagisawa, M., Ariga, T., in: Kamerling, J. P. (Ed.), Comprehensive
Glycoscience, Elsevier, Amsterdam, 2007.
[71] Brunetti-Pierri, N., Scaglia, F. Mol. Genet. Metab. 2008, 94, 391.
149
[72] Kusunoki, S., Kaida, K. J. Neurochem. 2011, 116, 828.
[73] Futerman, A. H., van Meer, G., Nat. Rev. Mol. Cell. Biol. 2004, 5, 554.
[74] Ariga, T., McDonald, M. P., Yu, R. K. J. Lipid Res. 2008, 49, 1157.
[75] Copp, A.J., Greene, N.D. J. Pathol. 2010, 220, 217.
[76] Obeidi, N., Russell, N., Higgins, J.R., O'Donoghue, K. Prenat. Diagn. 2010, 30,
357.
[77] Cacić, M. Neuroreport 1995, 26, 389.
[78] Vukelic, Z., Metelman, W., Muthing, J., Kos, M., Peter-Katalinic, J. Biol. Chem.
2001, 382, 259.
[79] Saito, M., Sugiyama K. Biochim. Biophys. Acta 2000, 1474, 88.
[80] Fine, H.A. Clin Oncol. 2004, 22, 1.
[81] Shukla, G.S., Krag, D.N. Expert Opin. Biol. Ther. 2006, 6, 39.
[82] Chapman, P.B. Curr. Opin. Investig. Drugs. 2003, 4, 710.
[83] Hakomori, S., Handa, K. FEBS Lett. 2002, 531, 88.
[84] Becker, R., Rohlfs, J., Jennemann, R., Wiegandt, H., Mennel, H.D., Bauer, B.L.
Clin. Neuropathol. 2000, 19, 119.
[85] Eberlin, L.S., Dill, A.L., Golby, A.J., Ligon, K.L., Wiseman, J.M., Cooks, R.G.,
Agar, N.Y.R. Angew. Chem. Int. Ed. 2010, 49, 1.
[86] Yin, J., Miyazaki, K., Shaner, R.L., Merrill Jr., A.H., Kannagi, R. FEBS Lett. 2010,
584, 1872.
[87] Wakabayashi, M., Okada, T., Kozutsumi, Y., Matsuzaki, K., Biochem. Biophys.
Res. Commun. 2005, 328, 1019.
[88] Wagener, R., Rohn, G., Schillinger, G., Schroder, R., Kobbe, B., Ernestus, R.I.
Acta Neurochir. 1999, 141, 1339.
[89] Steiner, G., Shaw, A., Choo-Smith, L.P., Abuid, M.H., Schackert, G., Sobottka,
S., Steller, W., Salzer, R., Mantsch, H.H. Biopolymers 2003, 72, 464.
[90] Hedberg, K.M., Dellheden, B., Wikstrand, C.J., Fredman, P. Glycoconj. J. 2000,
17, 717.
[91] Leach 3rd, F. E., Wolff, J. J., Xiao, Z., Ly, M., Laremore, T. N., Arungundram, S.,
Al-Mafraji, K., Venot, A., Boons, G. J., Linhardt, R. J., Amster, I.J. Eur. J. Mass
Spectrom. 2011, 17, 167.
150
[92] Seo, Y., Andaya, A., Leary, J. A. Anal. Chem. 2012, 84, 2416.
[93] Staples, G. O., Zaia, J. Curr. Proteomics 2011, 8, 325.
[94] Zaia, J., Miller, M. J. C., Seymour, J. L., Costello, C. E., J. Am. Soc. Mass Spectrom.
2007, 18, 952.
[95] Zaia, J. OMICS 2010, 14, 401.
[96] Bielik, A. M., Zaia, J. Methods Mol. Biol. 2010, 600, 215.
[97] Zaia, J. Mass Spectrom. Rev. 2009, 28, 254.
[98] Miller, M. J. C., Costello, C. E., Malmstrom, A., Zaia, J. Glycobiology 2006, 16,
502.
[99] Hitchcock, A. M., Yates, K. E., Costello, C. E., Zaia, J. Proteomics 2008, 8, 1384.
[100] Volpi, N., Maccari, F., Linhardt, R. J. Electrophoresis 2008, 29, 3095.
[101] Wolff, J. J., Laremore, T. N., Busch, A. M., Linhardt, R. J., Amster, I. J. J. Am.
Soc. Mass Spectrom. 2008, 19, 790.
[102] Laremore, T. N., Leach 3rd, F. E., Solakyildirim, K., Amster, I. J., Linhardt, R.
J. Methods Enzymol. 2010, 478, 79.