Abstract
Native mass spectrometry is a
rapidly emerging technique for fast and sensitive structural analysis of
protein constructs, maintaining the protein higher order structure. The
coupling with electromigrative separation techniques under native
conditions enables the characterization of proteoforms and highly
complex protein mixtures. In this review, we present an overview of
current native CE-MS technology. First, the status of native separation
conditions is described for capillary zone electrophoresis (CZE),
affinity capillary electrophoresis (ACE), and capillary isoelectric
focusing (CIEF), as well as their chip-based formats, including
essential parameters such as electrolyte composition and capillary
coatings. Further, conditions required for native ESI-MS of (large)
protein constructs, including instrumental parameters of QTOF and
Orbitrap systems, as well as requirements for native CE-MS interfacing
are presented. On this basis, methods and applications of the different
modes of native CE-MS are summarized and discussed in the context of
biological, medical, and biopharmaceutical questions. Finally, key
achievements are highlighted and concluded, while remaining challenges
are pointed out.
Introduction
Variability in a protein’s function arising from a single gene is driven
by genetic variation, alternative splicing of RNA transcripts, and PTMs,
forming distinct molecular entities, commonly designated as
“proteoforms” . In addition, biological functions of proteins are
highly associated with their tertiary structures, including protein
folding and conformation. Furthermore, proteins regularly do not operate
alone but form homomeric (proteoforms of the same base polypeptide
sequence) or heteromeric (different protein subunits) protein complexes,
dictating quaternary structures. The nature of proteoform structure as
well as the composition of a protein complex thoroughly impact
functionality and regulation of these biological constructs. A large
number of human diseases result for example from structural protein
misfolding, protein aggregation, and dysregulations in phosphorylation
or glycosylation states, such as Alzheimer disease and Parkinson
disease, type 2 diabetes, and cancer . Apart from endogenous proteins,
therapeutic, in most cases recombinant, proteins also play a central
role in human health as their safety and efficacy strongly relies on
constitution and conformation of recombinant protein products. For
instance, aggregate formation or protein degradation can drastically
reduce the efficacy or may increase immunogenicity of a drug.
Consequently, the characterization of proteins, protein complexes, and
associated interactions is indispensable to not only better understand
protein function and biological processes at the molecular level but
also to better evaluate biotherapeutic protein products. Native mass
spectrometry (nMS) is gaining more and more traction as an asset for
proteoform and protein complex analysis close to their functional
conformational state. Native MS generally enables the analysis of
protein complexes by preserving non-covalent interactions with ligands,
metals, substrates, and drugs . In comparison with more traditional
techniques applied in structural biology, such as NMR spectroscopy,
X-ray crystallography, or electron microscopy, nMS does not provide
direct molecular structural information . However, it has some major
advantages such as speed, selectivity, and high sensitivity, thus,
requiring lower amounts of material. Additionally, nMS enables the
analysis of complex protein mixtures, especially if coupled to upfront
separation techniques. Native MS is typically conducted using solutions
close-to-neutral pH to preserve non-covalent interactions . Comparing
mass spectra between proteins in their native state to traditional
denaturing conditions, several differences can be observed . Due to the
preservation of the native protein structure, distinct parts of the
polypeptide sequence are not solvent accessible, limiting the number of
protonation events resulting in lower and fewer charge states. The
concentration onto fewer number of different signals can result in
increased sensitivity, particularly for larger proteins. In addition,
the higher m /z range accessible by native protein
distributions, due to lower charge states, results in higher
peak-to-peak resolution which facilitates the detection of smaller mass
differences between proteins with similar mass or structure.
Nevertheless, nMS is still a technique requiring specific experimental
expertise as well as dedicated mass spectrometers for efficient ion
transfer and ion separation at the elevated m /z -values .
Unfortunately, ion suppression and overlapping MS signals are often
observed in a direct infusion nMS experiment, especially considering
complex protein mixtures . In such cases, separation prior to MS
analysis can be tremendously helpful to reduce sample complexity.
Furthermore, native top-down (TD) protein analysis is becoming more and
more popular as it bridges the gap between proteomics and structural
biology for in-depth proteoform and protein complex characterization .
Stoichiometry and precise subunit composition can be quickly determined
and PTMs can be located and characterized using top-down MS (TDMS) under
native conditions . In addition, the time-consuming protein digestion
step required for bottom-up methods is eliminated. Due to the complexity
of proteome samples, high-resolution native separations of samples are
indispensable in TDMS for better proteome coverage . Several techniques
for separation of proteins under native conditions have been described
in the literature . Chromatographic separation approaches compatible
with native conditions include SEC , ion-exchange chromatography ,
hydrophobic interaction chromatography , and affinity chromatography .
Electromigrative techniques such as capillary zone electrophoresis (CZE)
, microchip electrophoresis (MCE) , mobility capillary electrophoresis
(MoCE) , affinity capillary electrophoresis (ACE) , and
capillary isoelectric focusing (CIEF) are also well suited for native
protein separation. Capillary electrophoresis (CE) exhibits high
separation efficiency and resolution and allows due to the low sample
consumption the detection of trace levels of analyte . In principle, all
these techniques can be performed under native conditions preserving the
protein natural conformation. Stand-alone native separations are limited
in terms of compositional or proteoforms information. The hyphenation of
native separation techniques with MS is beneficial for obtaining
information on e.g., protein identity, conformation, metal
factors, or stoichiometry also for protein mixtures which has been well
summarized in recent reviews . The use of electromigrative techniques to
separate analytes under native conditions is well established since the
development of blue native PAGE in 1991 . Later, capillary-based
techniques like CZE, ACE, and CIEF were used to separate proteins in
electrolyte solutions which preserve their higher order structure and
allow online native MS coupling.
In this review, we summarize and discuss requirements and opportunities
of various native electromigration techniques coupled to native mass
spectrometry both in the context of methodologies and applications. In
many studies, the native state of proteins and/or proteoforms might have
been preserved during separation, even though, this has not been called
“native” in most cases (and cannot be verified as such) due to
denaturing conditions in the ESI-MS used in the majority of all CE-MS
measurements of proteins. Here, we focus on studies where a native state
of the protein is preserved also in mass spectrometry. For
simplification, we use the term “native” even though
“non-denaturing” or “native-like” conditions might be more
appropriate. In the first part, adaptations of electromigrative methods
to preserve the higher order structure (HOS) of proteins are described
covering CZE, ACE, and CIEF separations and their respective microchip
variants. Major parameters, such as type and concentration of background
electrolyte (BGE), pH, or capillary coating are discussed. After
presentation of native ESI-MS conditions, interfacing of
electromigrative methods in the context of native protein analysis is
discussed. In the following part, methods and applications of CZE-MS,
ACE-MS, and CIEF-MS are summarized, finalized by conclusions and
perspectives of these emerging techniques. The general concept for these
methods is illustrated in Figure 1.
Native conditions in
capillary- and microchip-based electromigrative separation techniques
Capillary zone
electrophoresis (CZE)
The most common mode of separation in capillary electrophoresis is
capillary zone electrophoresis (CZE). After applying an electric field,
analytes are separated in fused silica capillaries filled with a
conductive solution (the BGE), according to their electrophoretic
mobilities. The electrophoretic mobility is a unique characteristic of
an ion in a defined system and is proportional to the ion’s charge while
being inverse proportional to its hydrodynamic radius. Contrary to a
denatured protein, in a native, folded state, both the effective surface
charge and the hydrodynamic radius of a protein heavily depend on the
HOS of a protein . Therefore, determining the effective charge of
analytes under certain conditions in order to selectively adjust
separation conditions and increase separation performance becomes even
more challenging.
In CZE, BGE composition and the type of salt used impact separation
performance. Therefore, BGE components and concentration need to be
adjusted in order to obtain the required selectivity and best separation
efficiency. In the context of native separation, the BGE also needs to
maintain an analyte’s HOS in solution. Therefore, BGE composition and pH
need to be adjusted in approximation to the preferred physiological
environment of the protein of interest . This can also include the
addition of specific salts which regulate some protein-protein,
protein-ligand, or protein-cofactor interactions . To maintain a native
conformation under near physiological conditions, aqueous BGEs like
ammonium acetate (AmAc) or formate (AmF), ammonium bicarbonate
(AmBicarb), Tris buffer, but also inorganic phosphate and borate buffers
are frequently used . Ideally, like BGE components and pH, the ionic
strength should mimic the natural protein environment as close as
possible. However, physiological ionic conditions are considered to be
within 100-200 mM which would lead to high electric currents in a CE
system. High currents produce Joule heat inside the capillary causing
analyte zone dispersion which reduces separation efficiency. Beyond
considerations regarding resolution, excessive heat generation could
potentially induce conformational changes or unfolding and therefore
needs to be avoided . Separation temperature in general must be
considered carefully especially when studying protein interactions as
complex stability and integrity are temperature dependent. Typically,
the separation capillary is cooled using air or liquids. Additionally,
in contrast to optical detectors (e.g. , ultraviolet (UV) or
light-induced-fluorescence (LIF)), online hyphenation to mass
spectrometers is often further limited regarding the current, due to the
risk of potential electrical arcing, depending on the interface (see
section 3.2). Nevertheless, while offering increased sensitivity mass
spectrometry can provide proteoform resolved assessment and charge
variant characterization. ESI-MS coupling requires volatile salts such
as AmAc to avoid interference with analyte ionization during the ESI
process. However, potential effects on analyte stability and complex
integrity due to exchanging BGEs should be monitored.
The electrophoretic mobility of an analyte reflects on its
charge-to-size ratio. Therefore, unless introducing modifications to the
analyte CZE cannot be used to determine charge and size independently.
In 2019, a liquid-phase analogue of drift tube ion mobility spectrometry
termed mobility capillary electrophoresis (MoCE) was proposed to enable
ion separation as well as hydrodynamic radius and effective charge
measurements in a single experiment . This can be achieved by applying
an electric field for separation in a laminar Poiseuille flow to
incorporate Tailor dispersion analysis and electrokinetic dispersion
theory . Being able to determine protein charge states in solution is a
great advantage as accurate hydrodynamic radii calculations rely on it.
Furthermore, protein charge states reflect on protein activity and
solubility and potentially give insight into protein conformation. For
native-like mobility and effective charge state measurements the
electrolyte components (e.g. , NaCl or AmAc ) need to facilitate
and stabilize the folded state of the protein. Analyte separation in
MoCE follows the principle of CZE. Therefore, the same ionic strength
restrictions due to Joule heating apply to capillary-based applications.
In addition, avoiding perturbations of the laminar flow due to
temperature gradients is an even more critical factor since these
effects can compromise the determination of hydrodynamic radii. Adding
nMS to MoCE applications can have the unique advantage to gain
low-resolution three-dimensional structural information . In a native
mass spectrum, the average charge state of globular proteins correlates
with its solvent assessable surface area in solution. Therefore, by
combining ellipsoid approximations gathered from MoCE in solution with
nMS data information regarding the three-dimensional shape and radius of
a protein could be obtained.
Affinity capillary
electrophoresis (ACE)
Since its first introduction, the number and versatility of ACE-based
methods continuously grow and, along with it, the variety of
applications . Affinity studies rely on native interactions between
binding partners and to maintain native conformations often phosphate or
Tris buffers are used as BGEs . ACE can be applied for the assessment in
the free solution and the immobilized mode. However, immobilization of
one binding partner introduces the risk of compromising protein
structure, thereby interfering with native interactions. Different modes
of free solution ACE can be distinguished : the receptor or ligand is
either added in varied concentrations to the BGE (i.e. , mobility
shift ACE, msACE) or to the sample (i.e. , pre-equilibrium and
kinetic CE) depending on the underlying binding kinetics. The most
extensively used mode is msACE, where the mobility shift which occurs
upon receptor and ligand complex formation is used to determine affinity
constants (Kd). In pre-equilibrium ACE, an already
equilibrated mixture of interacting species is injected into the
capillary, free ligand and complex are separated and both amounts are
determined and compared to a standard curve . In kinetic CE species
interact under non-equilibrium conditions during separation. Binding
affinities are temperature-dependant therefore precise
Kd measurements require a well-controlled capillary
temperature and low Joule heating, as discussed for CZE previously.
Commonly optical (e.g. , UV and LIF) or electrochemical detectors
are applied in ACE. Considering that in msACE affinity constants between
specific binding partners are determined, ACE does not benefit as
evidently from MS detection compared to other electromigrative
techniques. Nevertheless, due to the inherent heterogenous nature of
biomolecules and their complexes, even msACE benefits from native MS
detection offering protein variant resolved assessment and determination
of binding stoichiometry .
As already discussed for CZE, MS coupling requires the use of volatile
salts like AmAc to avoid ion suppression. However, altering the
electrolyte may influence binding kinetics and therefore should be
monitored . Fortunately, ammonium acetate or bicarbonate, at medium pH
do not seem to significantly alter protein−ligand interactions . Signal
suppression of the complex cannot entirely be avoided when the ligand is
continuously fed to the MS (i.e. , for dynamic approaches).
However, Kd may still be accurately measured without
seriously affecting signal intensity when applying ligand concentration
in the micromolar range or lower .
Capillary isoelectric focusing (CIEF)
In CIEF, the capillary is filled with different zwitterionic ampholytes
(i.e. , carrier ampholytes) to form a pH gradient upon introducing
an acidic anolyte and a basic catholyte from opposite sides into the
capillary. After the analytes are focused by applying an electric field
the sample is mobilized commonly either hydrodynamically or chemically
using an acid or base to disrupt the pH gradient. In order to circumvent
the mobilization step and the resulting loss in separation efficiency,
imaged CIEF (iCIEF) using optical detection along the entire capillary
is increasingly used . CIEF applications allow to fill the entire
capillary with the sample, typically dissolved in the carrier ampholytes
and further additives. This increases sample injection volume compared
to CZE. For pI determination, pI markers are included in this mixture as
well as additives such as methyl or hydroxypropyl cellulose which
decrease sample adsorption to the capillary wall, thereby improving peak
shape and separation resolution . These mixtures prepared for CIEF
applications are often higher in viscosity compared to BGEs typically
applied for CZE. This leads to enhanced separation performance by
creating a sieving effect, as well as decreased current and reduced
Joule heating. This is beneficial when studying labile proteins and
protein complexes under native CIEF (nCIEF) conditions. Fonslow et
al. further assessed the influence of separation temperature on complex
stability and observed an increase in stability for their complex of
interest by lowering the nCIEF separation temperature to 15°C . They
also raised the question of pH-stability of protein complexes in nCIEF
which might be critical regarding the pH of the ampholyte mixture as
well as during mobilization. In general, protein stability and the
potential risk of precipitation is a major challenge in IEF. To avoid
protein precipitation at their pI, solubility enhancers such as urea are
typically added to the carrier ampholyte mixture. In this way, clogging,
unstable current, and other undesired effects impairing the analysis are
reduced or negated . Furthermore, the presence of urea also reduces
interactions between proteins and the capillary wall . For native
measurements it is even more critical to avoid protein precipitation
which consequently results in denaturation of the protein’s structure.
Unfortunately, urea is a chaotropic salt and in high concentrations
induces protein unfolding and denaturation . Moreover, the denaturing
effect of urea depends on the size of the proteins, structure protection
by covalent bonds (e.g. , disulfide bridges) as well as the type
of underlying interactions forming their HOS. Along with sucrose as a
common alternative to urea there are other less chaotropic additives
available for CIEF including glycerol and formamide. Both glycerol and
formamide are also better compatible with ESI due to less ion
suppression compared to urea . To further decrease ion suppression
methyl or hydroxypropyl cellulose additives need to be avoided as well.
Since ampholytes are essential for IEF the amount of ampholytes can only
be minimized. Coupling native CIEF separation to mass spectrometry
allows simultaneous pI determination and heterogeneity assessment.
Especially the study of protein interactions benefits from the obtained
pI information reflecting on surface electrostatic properties of protein
complexes .
Microchip electrophoresis (MCE)
Microfluidic chips have gained interest as electrophoretic separation
platforms, referred to as microchip capillary electrophoresis (MCE), due
to fast and efficient analysis. The use of narrow flat channels with a
large width-to-height ratio in microchip electrophoresis has the benefit
of more efficiently supporting heat dissipation, thereby reducing Joule
heating . As a result, chip-based applications allow higher salt
concentrations and increased electrical field strength. Furthermore, the
production process of the chip allows a high degree of flexibility in
designing and segmenting the glass or organic polymer surface , enabling
the integration of additional workflow steps onto the chip . One of the
most prominent commercially available microchip devices coupled via ESI
to MS is the ZipChip (908 Devices). The manufacturer offers a variety of
chips designed for different applications including the high resolution
native (HRN) and high speed native (HSN) chip for native protein
separations . Additionally, kits for native applications including a
generic native antibody kit containing a sample diluent and the
background analyte are available. The native antibody BGE consists of
water, isopropyl alcohol, AmAc, acetic acid and DL-histidine at a pH of
5.5 .
Due to their clinical relevance for pathogen identification,
personalized medicine, and biomarker detection pre-equilibrium ACE
applications (e.g. , CE-immunoassays) were translated to microchip
setups following the lab-on-a-chip concept. Over the years, different
reviews have been published which cover on-chip immunoassays development
and innovations . Primarily LIF, chemiluminescence, and various
electrochemical detectors are used for these applications due to easier
miniaturization and lower limits of detection compared to UV/Vis
detection . While microchip-based immunoaffinity CE is the most
interesting ACE mode for MS coupling, to our knowledge there have not
yet reports been published where chip-based ACE applications were
directly hyphenated with MS detection.
With the intent to benefit from the same advantages of microfluidic
separations discussed for CZE different approaches to microchip-based
CIEF and iCIEF have been developed . However, to our knowledge no
reports on chip-based CIEF or iCIEF applications specifically designed
for native separations have been published so far. For ESI-MS coupling
Intabio (now Sciex) has introduced the “Blast” CIEF (iCIEF) microchip
in 2019 . The device has not yet been released but native
microchip-based CIEF-MS applications are expected in the
future.
Capillary and microchip coatings for native separations
With proteins adsorbing to hydrophilic fused silica surfaces commonly
all capillary-based electromigrative protein separations require coating
of the silica capillary. Especially large and basic proteins are prone
to be adsorbed on the silica surface, which decreases, broadens, or even
eliminates analyte signals. Another reason for surface coating is the
control of the pH-depended electroosmotic flow (EOF) superimposing
analyte migration. Pivotal for the magnitude of EOF generated are acidic
silanol groups at the contact surface between BGE and silica material
(schematically depicted in Figure 2A). It has been thoroughly discussed
in literature that the acidic properties of these silanol-groups are
highly inhomogeneous . For the three types of silanol groups present on
quartz surfaces it has been simulated that the pKaranges from 4.5 – 5.6 and 8.4-9.3 to >11.3 . Consequently,
for native measurements a strong EOF is generated at a physiological
near neutral pH.
Uncharged neutral coating agents suppress or even closely eliminate the
EOF (Figure 2B) while cationic agents can reduce but also revert the EOF
(Figure 2C). Cationic agents are often applied as successive multiple
ionic polymer layer (SMIL) to increase the overall coating stability .
It should be mentioned that even though both approaches are applicable
to native applications the use of coatings for native separations has
not yet been as thoroughly evaluated as for denatured proteins.
Generally, users can select between dynamic coating reagents such as
hydroxypropyl cellulose (often used in CIEF) added to the BGE or applied
as a precoating or static covalently bound coating agents. Dynamic
coatings can easily be added to the BGE and are commonly used when
optical detection like UV/Vis or LIF is employed. Since dynamic coating
leaches into the mass spectrometer causing ion suppression in the ESI
process as well as contaminating the instrument, only physically or
chemically precoated capillaries or chips can typically be used if MS
detection is involved.
Covalently bound neutral coating agents such as polyvinyl-alcohol (PVA)
or linear polyacrylamide (LPA) are mostly applied for MS-coupled native
CZE, ACE, and CIEF applications. Coated capillaries or microchips are
commercially available but can also be prepared in-house. For MCE a
chemical vapor deposition method was developed performing very well for
protein separation under native conditions, while also providing stable
coatings fully compatible with ESI-MS .
Native conditions for CE-MS
Native ESI-MS
Native separation techniques are coupled to MS using ESI, which enables
ionization of proteins in the liquid phase by the application of a high
voltage . The electric field leads to the formation of charged droplets.
Due to solvent evaporation the droplets undergo coulomb fission once the
repulsion between charges overcomes the surface tension. This process
finally results in the formation of free gas-phase protein ions. In
classical ESI-MS experiments, the protein is sprayed from a solution
containing organic solvents, such as ACN and low percentages of acid
(positive mode) . The organic solvent ensures a fast evaporation of
solvent by reducing the surface tension and the acid or base can promote
charge transfer to the protein molecules. The redox processes during ESI
can significantly affect the pH of the droplets . In the positive ion
mode, oxidation reactions such as 2 H2O 4
H+ + 4e- + O2 cause acidification of
the solution.
As a liquid phase ionization process, ESI can preserve non-covalent
protein interactions. However, this entails several challenges since
organic solvents, especially ACN and MeOH, as well as a high or low pH
during the ESI process lead to protein denaturation. In addition,
non-volatile salts, often used to stabilize proteins in solution (see
section 2.1) should be avoided in ESI . They reduce sensitivity and mass
accuracy of the instrument, caused by adduct formation, ion suppression
and increase of chemical noise by salt cluster ions . Moreover,
non-volatile buffers can lead to the formation of salt deposits in the
instrument, which can result in a loss of ion transmission . Adduct
formation of salt and analyte splits the signal over multiple species,
increasing the spectral complexity and reducing the
[M+nH]n+ intensity, especially for large proteins
. If adduct ions of large protein complexes are not resolved, mass
accuracy can be dramatically reduced as well as the ability to identify
proteoforms with similar masses. Volatile, aqueous buffer systems with
physiological pH like AmAc, AmF or AmBicarb can be applied for
preserving the native form of proteins as well as for subsequent ESI of
analytes . Ammonium acts as a protonating agent and is removed via
NH3 evaporation, thereby avoiding significant adduct
formation . Acetate and formate evaporate as acid . Especially for
direct infusion experiments, an extensive buffer-exchange to volatile
salt solutions prior to ESI is required for proteins stored in
non-volatile buffers . Neutral AmAc solutions are most commonly used for
native ESI-MS studies. However, AmAc has buffering properties around pH
4.75 and 9.25 but not at neutral pH, expecting a decrease of the pH down
to pH around 4.75 during the ESI process (positive ion mode) .
Nevertheless, for many proteins non-covalent interactions are still
maintained at pH 4.75 and in unbuffered solutions the pH can decrease
from the near neutral range even down to pH 1.4 . AmBicarb or AmF
buffers can also be applied for native ESI-MS. AmBicarb provides a
better buffering capacity at pH 7, however, it may cause denaturation
due to CO2 outgassing . Hedges et al.demonstrated that solutions containing bicarbonate exhibit intense
foaming when exposed to higher temperatures during the ESI process,
leading to unfolding of proteins at the surface of CO2bubbles prior to ion formation . Thus, the formation of higher charge
states is often observed when using bicarbonate buffer systems, however,
also depending on the electrospray (ES) potential . The supercharging
effect of bicarbonate can also be beneficial, e.g. , for coupling
studies about protein structure in native solutions to mass
spectrometers with limited m/z range or to improve MS/MS
performance of native protein ions AmF provides buffering around pH 3.75
in the positive ion mode and is thereby even less suitable for buffering
in the neutral pH range compared to AmAc . In addition, there is an
increased risk of protein denaturation, probably due to its chaotropic
nature . In general, ideal ESI conditions keeping the native state
depend on the protein and need to be tested. Some proteins and protein
complexes also require specific non-volatile salts and the removal of
these salts can significantly impact their stability, making the native
analysis of these molecules by ESI-MS challenging .
NanoESI is more sensitive and more tolerant towards nonvolatile salts.
In nanoESI, flow rates are typically well below 1000 nl/min using
emitter with inner diameters of typically <30µm . Thereby,
smaller initial droplets are generated, which undergo less solvent
evaporation and Coulomb fission . Thus, there is less enrichment of
salt, lowering the salt/protein ratio. This significantly decreases the
formation of salt cluster ions and salt-adducted proteins. It has been
demonstrated that this desalting effect is even more enhanced when using
emitters with tip diameters ≤1 µm, enabling direct nanoESI from
solutions containing non-volatile salts, such as NaCl, NaCl and Tris-HCl
or KCl and Tris-HCl .
As in CE low flow rates (nl/min) occur, it is, in principal, well suited
for nanoESI-MS coupling . The different options for interfacing CE with
ESI-MS will be discussed in the following section.
CE-MS interfaces for
native ESI
CE can be coupled to MS using a sheath liquid (SL) or sheathless
interface . In SL techniques, additional liquid closes the electric
circuits of both the CE and the ES at the end of the capillary and
stabilizes the ES independent of the magnitude of the CE-flow that
depends on the EOF . Conventionally, sheath liquids consist of
water-organic solvent mixtures with a small amount of volatile acid or
base . Organic solvents decrease the surface tension of the SL and
increase its volatility. The flow rate of the SL is usually much higher
than the eluent from the CE, leading to unwanted dilution, however, the
composition of the SL can be utilized to modify the ES chemistry and
increase the ionization efficiency . At least two types of SL interfaces
can be distinguished: the coaxial SL interface and the nanoflow SL
(nanoSL) interface. The coaxial SL interface is traditionally used for
the coupling of CE with MS (commercially available from Agilent
Technologies). In this configuration, the CE capillary is placed inside
a metal needle and SL is coaxially delivered inside this needle. At the
sprayer tip, the CE outflow mixes with the SL. The flow rates of the SL
are around 1-10 µl/min, requiring typically an additional nebulizer gas
and causing high dilution of the analyte, impairing the sensitivity . In
addition, the nebulizer gas might cause a suction effect at the
capillary tip, potentially decreasing separation efficiency. These
disadvantages can be reduced by using a nanoSL interface, which employs
reduced SL flow rates (below 1000 nl/min), omitting a nebulizer gas and
resulting in improved sensitivity . NanoSL interfaces are characterized
by emitters with small ID which are placed close to the inlet of the MS
to enhance the transmission efficiency. Since its introduction by Hsiehet al. , the Dovichi group improved the nanoSL design , which has
later been commercialized by CMP Scientific. Further nanoSL variations
have been designed by different laboratory groups , aiming to enhance
sensitivity, reproducibility and robustness as well as an easy handling,
such as the recently introduced nanoCEasy interface . Unlike the SL
interface, sheathless interfaces do not require additional liquid and
only use the BGE for spray generation . The most widespread concept is
based on a porous capillary tip allowing for electrical contact . The
main benefit of this nanoESI technique comprises its sensitivity. MCE
can be coupled online to MS using conductive emitter coatings or
ion-permeable membranes (analogous to the sheathless method for
CE-ESI-MS) or make-up-flow channels (analogous to the SL approach for
CE-ESI-MS) . The latter is used in the ZipChip (908 Devices) and the
Blast (Intabio/Sciex), respectively.
To transfer proteins and protein complexes in its native structure from
the liquid to gas phase by ESI the above-mentioned interfaces can be
applied when certain aspects are considered: Organic solvents, such as
ACN and MeOH used in SL interfaces often cause protein denaturation .
However, even the often used 1:1 mixture of isopropanol and water may
keep proteins in their native state, as shown for the antibody
infliximab and the glycoprotein antithrombin . Also, the addition of
0.1% acetic acid to the SL can further increase the signal intensity
without denaturing proteins such as mAbs . In general, a reduced flow of
SL is beneficial to preserve native features of proteins and for
minimizing the dilution of the samples . Higher flow rates might favour
protein denaturation depending on the composition of the SL, whereas low
flow rates cause ES instabilities . The optimal flow rate of the SL also
depends on the protein. Marie et al. obtained the highest signal
intensity for antithrombin using 2 µl/min in a coaxial SL interface ,
while Le-Minh et al. selected a flow rate of 10 µl/min as
compromise between signal intensity and structure preservation of mAbs .
Depending on the EOF, an ionic boundary at the capillary outlet may be
formed in SL interfaces due to the migration of ions from the SL into
the capillary, potentially promoting protein denaturation .
NanoSL interfaces are well suited for native ESI-MS and allow the
application of purely aqueous SL such as aqueous AmAc solutions . Shen
and Kou et al. used 25 mM NH4Ac (pH 6.9) as SL
for native SEC-CZE-MS/MS analysis of the Escherichia coli(E. coli ) proteome , whereas 10 mM NH4Ac have
been applied for CZE-MS analysis of the SigmaMAb and proteins
compromising the 70S ribosome of E. coli . The difficulties of
protein denaturation and dilution by the SL can be completely bypassed
using a sheathless interface. However, spray conditions are dependent on
the BGE and ES conditions cannot be improved by additional liquid . In
addition, high currents that often result from high concentrations of
salts applied in native CE should be avoided as they can damage the
porous tip . Furthermore, salt based BGEs can crystalize at the
sheathless spray tip, especially at low flow rates, resulting in a
reduction of the ES stability. To solve this problem, either the ES
voltage should be reduced or the spray tip should be moved further away
from the MS entrance with increasing AmAc concentrations .
The sheathless interface has been applied for native CZE-MS of various
native proteins and protein complexes such as mAbs, the ribosomal
proteome of E. coli and endogenous nucleosomes using the
commercially available emitters from SCIEX (CESI 800) . Jooß et
al. have developed a nCZE-top-down-MS method using sheathless emitters
(CESI, SCIEX) for the native analysis of a mixture of proteins and
protein complexes from 30-800 kDa .
For the hyphenation of native MCE with MS, 908 Devices offers microchips
and BGEs for native antibody analysis . An electroosmotic pump delivers
to the end of the separation channel a make-up liquid at composition
similar to CE-MS SL-interface, forming native protein ions .
Parameters for native MS
The analysis of native proteins
and protein complexes by MS often necessitates specific instrument
conditions and adaptations. Efficient transfer of the analyte through
the ion optics is crucial while maintaining its native structure and
inter- and intramolecular interactions. In the past, standard
high-resolution MS instruments designed for small molecules have been
modified for native analysis of proteins and protein complexes .
Throughout the progression of native MS, commercial TOF, FT-ICR, and
Orbitrap instruments with extended m/z range have been developed,
such as the Q Exactive™ UHMR Hybrid Quadrupole Orbitrap™ Mass
Spectrometer (Thermo Fisher Scientific) and the 6545XT AdvanceBio
LC/Q-ToF MS (Agilent).
One important finding in native mass spectrometry instrumentation has
been that it is beneficial to increase the pressure in the first vacuum
stage of the device to analyze large protein complexes . A higher
pressure in the front end of the instrument is required for ion cooling
through collisions with inert gas molecules. When large proteins enter
the ion guides at low pressures, their oscillations result in large
deviations from the central ion axis, leading to losses of ions.
However, collisions with small gas molecules decelerate large ions
(e.g. , protein complexes) and, in this way, their transmission is
improved . An increased pressure in the instruments front end also
promotes desolvation of analytes via collisions with gas molecules . A
higher pressure in the ion transfer region can be achieved by reducing
pumping efficiency, leaking additional gas into the source region,
incorporating a sleeve between the ion guide and the pump orifice to
restrict the flow, using heavier background gas, etc. .
Desolvation can also be promoted by heating up the ion source or
accelerating the molecules between the nozzle and the skimmer .
Efficient desolvation is necessary to achieve high mass accuracy and to
diminish heterogeneity caused by non-specific adduct formation,
increasing overall sensitivity . However, excessive pressure (and,
therefore, excessive collisions) can lead to a decrease in sensitivity,
resulting from complex dissociation and the displacement of ions from
the central axis .
The ion optics’ electric fields should be finely tuned to maintain the
native protein structure and to transfer large constructs that exhibit
high m /z values. In this context, radio frequencies of the
ion optics are typically reduced to enhance large-ion focusing . In
addition, similar to pressure values described in the previous
paragraph, excessive acceleration voltage can lead to protein complex
dissociation .
In Orbitrap instruments, extended trapping is often required to
stabilize larger protein species in the ion routing multipole, however,
too much trapping voltage leads to a decline in sensitivity . Fortet al. have demonstrated that in-source trapping can enhance
desolvation as well as ion transmission . In this scheme, ions are first
trapped in an injection flatapole in the front-end of an Orbitrap
instrument. The trapped ions undergo desolvation and lose their initial
momentum due to collisions with gas molecules, resulting in improved
transmission and mass resolution. Similarly, McGee et al. created
an in-source-trapping-like scheme on Orbitrap Tribrid instruments that
additionally filters low-mass contaminants via a so-called “voltage
rollercoaster” . Low-mass contaminants can suppress the spectral
quality of protein analytes via space charge effects, and their removal
improves the analysis of high m /z species.
In the context of Q-TOF instruments, Shen et al. investigated the
fragmentor and skimmer voltage as well as the collision energy in the
CID cell for native analysis of mAb SigmaMAb . They increased the
fragmentor and skimmer voltage for enhanced salt and water declustering
resulting in improved sensitivity. In general, antibodies tolerate a
higher collision energy compared to smaller molecules, however, an
increase of the collision energy decreased mAb signal intensities in
this case, likely due to analyte fragmentation from excess energy.
Similar to the pressure within the MS system, the amount of energy
analytes are exposed to must be finely balanced to achieve efficient ion
transmission. There are no broadly optimal gas pressures or radio
frequencies since these values are very instrument and analyte dependent
, but parameter settings reported in previous studies can serve as
guidance for an appropriate range of values, especially if similar
instrument setups are utilized. Nevertheless, optimal MS parameters can
vary significantly between analytes in native mode compared to the more
general applicable settings commonly encountered using denaturing
conditions.
Methods and applications of native electromigrative techniques
coupled to nMS
Methods of nCZE-nMS
The growing number of publications in recent years has demonstrated the
efficiency of CZE-MS for the analysis of protein constructs in their
native conformations. In addition, a series of CZE separations might
have been conducted preserving native features of proteins without being
recognized by the user. Potential examples are the analysis of
erythropoietin (EPO) or recombinant human growth hormone by CZE-MS.
However, these features were not reflected in the resulting mass
spectrum due to the applied denaturing conditions during the ESI process
. Before 2010, only the coaxial SL interface from Agilent was
commercially available and thus, utilized by a majority of laboratories
as CE-MS interface. This setup relies on diluting down the original CE
flow drastically in a so-called sheath liquid, which typically contains
a high amount of organic solvent, resulting in denaturing conditions. In
addition, the number of mass spectrometers adapted for the transfer of
higher m/z ions was comparably limited. Here, we will focus on
methods and applications where both, CE and MS were performed under
native conditions.
A general advantage of CZE compared to a more traditional native direct
infusion experiment is that non-volatile salts can be separated from
proteins using CZE prior to MS analysis, thereby circumventing an
extensive buffer exchange of proteins stored in non-volatile buffers .
For example, sodium tends to migrate very early in the electropherogram
being detected as sodium acetate clusters. In addition, the separation
of proteins, protein complexes, and aggregates in native CZE more
closely reflects their solution based conformations and aggregates
compared to solely relying on direct infusion which can be prone to
artefact formation during the ESI process and bias towards certain
species based on ion transfer settings .
Publications describing methods and applications of native protein
analysis by CZE-MS are summarized in Table 1. Most commonly, BGEs based
on aqueous AmAc have been applied, typically containing 10 – 50 mM AmAc
at pH 6 – 8. For instance, Le-Minh et al. concluded for
infliximab that a BGE of 40 mM AmAc and pH 6.0 is the best compromise
between signal intensity and preservation of the native mAb conformation
. In addition, aqueous AmBicarb-based solutions have been applied by
Bertoletti et al. for the analysis of
Beta2-microglobulin . To optimize separation conditions,
different BGE’s as well as different salt concentrations in the BGE
should be compared and evaluated, as demonstrated in Figure 3 for (A) a
mAb and (B) a mixture of four model proteins. Comparing 10 mM AmAc and
AmF (Figure 3 A), AmAc showed slightly better separation regarding the
peaks labelled with a black box. 50 mM AmAc resulted in a longer
migration time and wider peaks compared to 10 mM AmAc. In Figure 3 B,
separation improved with increasing AmAc concentration. However, it
should be kept in mind that high salt concentrations can lead to high CE
currents which can be detrimental for the interface set up.
In the coaxial SL interface a SL containing water and organic solvent,
such as MeOH and IPA has been applied . Le-Minh et al. compared
the influence of different SLs, ACN, MeOH and IPA mixed with water in a
ratio of 50:50 v/v for the mAb infliximab and observed that ACN and MeOH
induced denaturation. For IPA, denaturation, and the associated shift in
the charge state distribution, was drastically reduced. Sometimes
formic, acetic acid or AmAc is further added to the SL. For nanoSL
interfaces from CMP, aqueous AmAc solutions were applied . Considering
sheathless CE-MS interfaces, mainly the porous tip interface was
employed . Regarding capillary coatings, neutral coatings are most
commonly applied for nCE-MS to reduce protein adsorption to the inner CE
capillary wall and to lower the EOF and thereby, increase the resolution
between proteins (see above). Usually LPA and PAM , but also PVA
coatings have been utilized . Some publications also mention the
successful application of bare-fused silica capillaries for nCZE-MS . In
this context, Belov et al. compared the performance of bare-fused
silica and PAM coated capillaries and concluded, that the PAM coated
capillary demonstrated a better reproducibility and higher efficiency
for more complex samples . Furthermore, SMIL coatings are a promising
alternative to the more traditional single layer coatings. For instance,
Le-Minh et al. used a cationic polybrene-dextran
sulfate-polybrene (PB-DS-PB) coating and applied reversed polarity to
create a strong EOF towards the mass spectrometer. However, under these
conditions, H+-ions from the SL can migrate into the
separation capillary leading to the formation of an ionic boundary .
In 2021, Jooß et al. provided standard operation procedures for
the analysis of model proteins and protein complexes ranging from 30 to
800 kDa by nCZE-TDMS for different Orbitrap instruments using a porous
tip interface. Various parameters for nCZE-TDMS are discussed in detail,
including the influence of BGE compositions, separation voltage,
supplemental pressure, and MS parameters on the quality of the spectra
as well as optimized fragmentation conditions for native top-down
analysis of intact complexes. This protocol can be used as a guide to
optimize separation and MS parameters for the analysis of native
proteins and protein complexes .
Applications of nCZE-MS
Denaturation, misfolding, dimerization, or aggregation of antibodies can
impact the efficacy and safety of therapeutic antibodies and may
increase their immunogenicity, making an extensive characterization of
therapeutic drugs essential before patient administration. Thus, the
characterization of biopharmaceuticals such as therapeutic mAbs is an
important field of application for native CZE-MS. Belov et al.analyzed a mAb in 2017 and 2018 under native conditions and detected the
monomer and also traces of a dimeric species of the respective mAb .
However, the dimer could have also been formed during the ESI process as
the dimeric species co-migrated with the monomeric form of the mAb. Le
Minh et al. successfully analyzed the mAb infliximab by nCZE-MS
under native conditions and detected simultaneously different
conformational states including native and unfolded monomers as well as
dimers of the mAb in a storage stressed sample (6 months at 4 °C) . To
provide additional evidence that the dimers observed by nCZE-MS are not
formed artificially during ESI, they confirmed their experiments by
atomic force microscopy. In 2021, Shen et al. developed a native
CIEF-assisted CZE-MS method for the analysis of mAbs . After injection
of the samples, they first applied native CIEF stacking to focus the mAb
and increase overall sensitivity. After completion of the focusing step
the antibody was further separated under native conditions by CZE. In
this way, it was possible to increase the sample loading capacity
without loss of separation resolution. They successfully characterized
glycoforms, variants, and aggregates of two different mAbs.
Apart from mAbs, the characterization of other proteins and protein
complexes plays a pivotal role in biopharmacy, either to characterize
the quality of a protein drug or to get a better understanding of
biological processes in cells. The plasma glycoprotein antithrombin is
used as a drug for patients with hereditary antithrombin deficiency .
Only the native form of antithrombin is active and the presence of
latent or dimeric forms decreases the quality of therapeutic
antithrombin products making a fast and reliable quantification of
different protein forms necessary. Marie et al. successfully
differentiated between the native, latent, and dimeric forms of
antithrombin using CZE-MS.
Besides analyzing single proteins and complexes, whole cellular
proteomes can be analyzed by native CZE-MS/MS. Shen et al.prefractionated lysed E. coli cells using size-exclusion
chromatography and then analyzed the collected fractions by CZE-MS/MS,
which resulted in the identification of 144 proteins, 672 proteoforms
and 23 protein complexes . Though of note is that so far, there is no
efficient way to carry out elaborate data base searches for non-targeted
native top-down proteomics. In 2020, Mehaffey et al.characterized the E. coli 70 S ribosome at various
Mg2+ concentrations by CZE-MS/MS without previous
fractionation. The magnesium cation was required to maintain the 3D
structure of the ribosome, highlighting the ability to perform native
CZE-MS using such additives. 500 µM magnesium acetate resulted in the
observation of the intact 30S and 50S subunits while a reduction of the
magnesium concentration and the removal of rRNA led to observation of
subcomplexes and single proteins of the ribosome (see Figure 4). They
identified 48 of the 55 E. coli ribosomal proteins as 84
proteoforms including 22 protein−metal complexes and 10 protein−protein
complexes .
Jooß et al. generated a platform for the separation and
characterization of whole nucleosomes as well as their histone subunits
and PTMs using nCZE-TDMS . Their platform allowed a direct injection of
nucleosome samples in complex buffers into the CZE system without
previous sample preparation. Therefore, they developed a “standard”
method and a “high-resolution” method by decreasing the supplemental
pressure and thereby increasing the run time to maximize the resolution
if necessary. This resulted in significantly improved resolution and
better separation of the tetra-acetylated (H3K4,9,14,18a) and
ubiquitinated (H2BK120ub) Nucleosomes, as shown in Figure 5.
nMCE-nMS
In native MCE-MS applications,
the microfluidic CE separation was carried out on a ZipChip platform
using the HRN microchip, the BGE provided in the ZipChip Native Antibody
kit and the ZipChip interface from 908 Devices (see Table 1). The BGE is
a proprietary formulation based on AmAc at pH 5.5, which is modified
with 4% DMSO before usage . The microchip channels are modified with a
covalent coating to minimize adsorption of proteins and suppress the EOF
. Until now, native MCE-MS using the ZipChip platform was only applied
for the characterization of mAbs, especially for the analysis of native
intact mAb charge variants , as summarized in Table 1. For instance,
Carillo et al. analyzed the three widely used IgG1 mAbs
rituximab, bevacizumab, and trastuzumab and identified 52 proteoforms of
trastuzumab as well as fragments and sialylated N -glycans of
rituximab . The analysis of mAbs charge variants is an essential part of
critical quality attributes (CQA) to ensure efficacy and safety of a
therapeutic product. Especially for this purpose, the ZipChip MCE-MS
platform offers a fast and simple application platform requiring minimal
sample preparation and method optimization . The application of this set
up to other classes of proteins and protein complexes is promising, but
still needs to be explored in the future.
ACE-nMS
All ACE-nMS approaches were based on mobility shift ACE (see Table 1).
To adapt ACE separations to MS detection different BGEs were selected in
accordance with the pI and stability of the analyte as well as their
binding affinity. Sufficient separation and ionization of the respective
analytes was observed using BGEs containing 25 – 50 mM AmAc at pH 6-8 .
Domínguez-Vega et al. applied a triple layer SMIL coating of
polybrene−dextran sulfate−polybrene (PB−DS−PB) and used a coaxial SL
interface for their measurements [42]. They added cortisone as a
neutral EOF marker to determine µeff. The presence of an
EOF made it also necessary to add 25 mM AmAc (pH 8.0) to the isopropyl
alcohol−water SL to avoid ion depletion at the end of the capillary.
Gstöttner et al. used a porous tip sheathless interface equipped with
OptiMS neutrally coated capillaries (Sciex) . Domínguez Vega et
al. observed slightly higher Kd values in ACE-MS
compared to ACE-UV. They attributed this to an increase in temperature
due to the nature of the coupling leaving a large part of the capillary
without adequate temperature control, required for precise
Kd determination. With the exposure of the capillary to
the (warm) ion source, the section closest to the MS is especially prone
to temperature increases.
In their initial work , Domínguez-Vega et al. developed and
evaluated an msACE-nMS method for simultaneous heterogeneity and
affinity assessment based on trypsinogen variants and their aprotinin
complexes. They separated and singly, doubly, and triply deamidated
trypsinogen charge variants at pH 8. Furthermore, MS detection enabled
binding stoichiometry assessment as well as the analysis of comigrating
compounds. For quantitative assessment of the trypsinogen−aprotinin
interaction, Kd values were determined based on the
extracted ion electropherograms. Accuracy of the obtained values was
evaluated by comparison with ACE-UV measurements using the same BGE. In
principle, Kd values can also be obtained based on the
abundance of the complex relative to the signal of the free ligand in
the mass spectrum offering mass and Kd determination for
comigrating substances. However, higher Kd values with
increased standard deviation were observed, compared to those based on
electrophoretic mobility shift. By comparing the MS-based results to
direct infusion experiments they confirmed that the discrepancy can be
attributed to the MS detection. The authors concluded that different
ionization efficiencies of the free protein and the complex as well as
varying complex stabilities between the liquid and the gas phase could
be the reason. This observation highlights again the importance of
having a native separation system, which can reflect interactions more
adequately taking place in solution compared to direct nMS analysis. To
demonstrate the use of their method for fast affinity screening of
heterogeneous proteins they also analyzed trypsin and α-chymotrypsin in
the absence and presence of aprotinin.
In 2021, Gstöttner et al. published the first msACE-nMS approach
for functional studies on mAbs. Due to their importance for antibody
therapeutic pharmacokinetics, they focused on interactions between the
neonatal Fc receptor (FcRn) and antibodies as a modal system. In this
work, different engineered antibodies and oxidized samples were
analyzed. The hyphenation to nMS allowed simultaneous monitoring of
antibody as well as FcRn heterogeneity, individual binding assessment
and complex stoichiometry determination. The mobility shift of the
H2O2-stressed and the reference mAb due
to the presence of FcRn is illustrated in Figure 6A. The data indicated
a decrease of the FcRn affinity with antibody oxidation and a
glycosylation effect, with slightly higher affinities for galactosylated
glycoform (Figure 6B). No significant effect of the type of FcRn
glycoform on complex formation could be observed, suggesting no
influence of the FcRn glycosylation on antibody binding. MS
investigation on binding stoichiometry revealed 1:1 and 1:2 mAb/FcRn
complexes under the applied conditions. The use of differently
engineered Fc constructs further enabled the differentiation between
symmetric and asymmetric binding. In 2022, Gstöttner et al.further expanded their work and published a detailed study on glycoform
binding assessment with msACE-nMS to study the binding of different
antibody glycoforms to the FcγRIIa receptor . Clear differences in
binding between doubly-, hemi-glycosylated, and non-glycosylated
antibodies, as well as for mutated IgG1 antibodies silenced for Fcγ
binding, were observed. Moreover, they reported for the first time that
high mannose glycoforms show a decrease in affinity for FcγRIIa.
nCIEF-nMS
Native CIEF Methods and applications involving nMS are summarized in
Table 2 . To limit ion suppression ampholyte concentrations are
generally reduced to 1 – 1.5% (v/v). Furthermore, any additives are
omitted except for glycerol. Partial filling of the capillary with
catholyte/proteins-ampholyte mixtures enables direct focusing and
subsequent mobilization by providing electrical continuity throughout
the analysis . 50 mM or 0.1% formic acid as a anolyte and 100 mM
ammonia or 10 mM AmBicarb as catholyte were used. Small quantities of
glutamic acid and arginine were added to detect the pH gradient limits .
To enhance analyte solubility up to 40% glycerol were added to the
water or AmAc (10 mM) based carrier ampholyte mixture depending on the
sample. The high viscosity of the glycerol solution has the additional
benefit of an anticonvective effect. This decreases analyte adsorption,
and also results in a reduced EOF. The addition of glycerol allowed
Mokaddem et al. to use even bare fused silica instead of LPA
coated capillaries, compared to other research groups. While this could
be attributed to the lack of more stable coatings at that time, it
substantiates the advantage of glycerol as a solubility enhancer for
hydrophobic and basic proteins.
A remaining challenge is mobilizing analytes without sacrificing
separation efficiency, for example due to peak broadening resulting from
hydrodynamic mobilization. Therefore, all authors maintained the
focusing voltage during pressure mobilization. Stable spray conditions
and efficient ionization were achieved using different sheath liquids
either composed of 10 mM AmAc (pH 5) , and AmAc (pH 5) with the addition
of 10% (v/v) ACN , or by using a mixture of equal amounts of methanol
and water and 1% (v/v) acetic acid . Xu et al. observed broader
distributions and slightly higher charge states compared to nCZE-MS
results and attributed this to a supercharging effect caused by
glycerol. They also described effective desalting of holomyoglobin with
the addition of 15% glycerol resulting in comparably clean mass
spectra.
Mokaddem et al. developed and evaluated a nCIEF-nMS method for a
mixture of standard proteins commonly used as CIEF pI markers . In 2015
Przybylski et al. applied nCIEF-nMS for the characterization of
the highly basic cytokine human interferon-gamma (IFN-γ) as well as its
active non-covalent homodimeric form . Accurate focusing of protein
variants with close pI values was achieved as well as pI determination
of the homodimer. Furthermore, a strong contribution of the two
regulatory C-terminal clusters of basic amino acids (i.e. , D1 and
D2) to the positive electrostatic potential of the protein dipole was
confirmed. Comparing single and deletion mutations within these regions
enabled the investigation of the impact of these domains on the pI
values which might reflect on different expositions of the domains and
the flexibility of the C-terminal domain. They proposed that their
results could correlate with the proposed mechanism of the interaction
and binding of D1 and D2 to heparan sulfate.
In 2022 Xu et al. published a nCIEF-nMS method for
microheterogeneity assessment of streptavidin and the carbonic
anhydrase−zinc complex, in comparison to nCZE-nMS. The group achieved
liquid-phase separation and characterization of seven different forms of
a streptavidin homotetramer with different PTMs (i.e., N-terminal
methionine removal, acetylation, and formylation) (Figure 7). In
addition, various carbonic anhydrase-zinc complex proteoforms could be
detected in a separate experiment (succinimide, deamidation, etc.) and
characterized regarding their individual pI-values. Moreover, the
applied their method was applied to an interchain cysteine-linked
antibody-drug conjugate (ADC), partially resolving drug-to antibody
ratio species.
Conclusions and Perspectives
Native CE-MS has gained quite
some attraction recently. With the different approaches available,
efficient separation can be combined with detailed mass spectrometric
information for the structural analysis of proteins and protein
complexes in biological, medical, and biopharmaceutical context. In
contrast to most LC-MS strategies, such as RPLC, CE-MS can be performed
under native conditions with comparably minor efforts and allows a more
efficient separation with narrower peaks. Furthermore, the general
benefits of CE, such as low sample requirement and fast separation,
apply and can be advantageous. More importantly, the selectivity of
electromigration techniques allows to separate proteoforms or protein
complexes with minor differences in charge and/or size (CZE-MS),
according to small differences in pI (CIEF-MS), or
protein-target-affinity (ACE-MS). ACE-MS even allows to determine
binding constants accurately.
With the advent of more efficient CE-MS interfaces as well as more
sensitive mass spectrometers over the last two decades, intact protein
analysis by CE-MS technology became generally more accessible, which
includes the more recent native CE-MS developments. Previously, native
separations might have been already performed in many cases, however,
not detected as such, due to denaturing ESI-MS conditions (e.g. ,
sheath liquid composition). Additionally, the development of various
capillary coatings has been substantially contributing to nCE-MS methods
and applications. Moreover, nanoESI interfaces, either using a SL or
performed sheathless, played a pivotal role in the progress of nCE-MS,
due to increased sensitivity, increased matrix tolerance, and
ease-of-use. Furthermore, our understanding of the properties of the
“native” ESI process increased regarding mechanism and efficient
analysis of large proteins and protein complexes. Still, there is no
general setup that works for every protein or protein complex of
interest, and major parameters in both the separation and the ESI-MS
need to be carefully adjusted according to stability of the HOS of
interest.
In this way, new fields of applications have been opened for the various
kind of native capillary electromigration techniques coupled to native
ESI-MS. CZE-MS has been used in several applications to analyze
proteoforms of therapeutic proteins, especially mAbs. Besides the
success of these targeted approaches, a truly non-targeted native
CE-TDMS approach to analyze proteoforms of native constructs in a
proteome-wide fashion is of major interest and will be possible when
adequate software tools for data generation, database search and
analysis are available. This will advance CZE-MS further towards native
TD proteomics. Native CIEF-MS has also been developed with a focus on
biopharmaceutical targets. This mode of CE separation adds another layer
to native protein construct analysis by providing accurate determination
of the respective pI-values. ACE is a major topic since decades and isper se applied under native conditions. ACE-MS has been shown to
give highly important information on proteins as it combines the
characterization of proteoforms with its biological and/or therapeutic
effect.
In the future, the addition of mobility in the gas phase using modern
ion mobility-mass spectrometry (IM-MS) techniques may add valuable
information, especially considering the analysis of different
conformational states. So far, IM-MS has been used to characterize
native proteins and protein complexes , whereas the online coupling of
mobility in liquid phase(CE) and in gas phase (ion mobility) was only
realized just recently, focusing on carbohydrates and small molecules .
Over the coming years, we expect a significant increase in methods and
applications of native CE-MS to tackle the various challenges of
in-depth protein and protein complex characterization in the context of
various different fields, such as biopharmaceuticals, structural
biology, and many more.