Journal of Chromatography A, 1498 (2017) 8–21
`
`Contents lists available at ScienceDirect
`
`Journal of Chromatography A
`
`j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c h r o m a
`
`Advances in organic polymer-based monolithic column technology
`for high-resolution liquid chromatography-mass spectrometry
`profiling of antibodies, intact proteins, oligonucleotides, and peptides
`Sebastiaan Eeltink a,∗, Sam Wouters a, José Luís Dores-Sousa a, Frantisek Svec b
`
`a Vrije Universiteit Brussel, Department of Chemical Engineering, Pleinlaan 2, B-1050 Brussels, Belgium
`b Beijing Advanced Innovation Center for Soft Matter Science and Engineering, Beijing University of Chemical Technology, Beijing, China
`
`a
`
`
`
`r
`
`
`
`t
`
`i
`
`
`
`
`
`c
`
`
`
`l
`
`
`
`e
`
`
`
`i
`
`
`
`n
`
`
`
`f
`
`
`
`o
`
`a
`
`
`
`b
`
`s
`
`
`
`
`
`t
`
`
`
`r
`
`
`
`a
`
`
`
`c
`
`
`
`t
`
`This review focuses on the preparation of organic polymer-based monolithic stationary phases and
`their application in the separation of biomolecules, including antibodies, intact proteins and protein
`isoforms, oligonucleotides, and protein digests. Column and material properties, and the optimiza-
`tion of the macropore structure towards kinetic performance are also discussed. State-of-the-art liquid
`chromatography-mass spectrometry biomolecule separations are reviewed and practical aspects such
`as ion-pairing agent selection and carryover are presented. Finally, advances in comprehensive two-
`×
`dimensional LC separations using monolithic columns, in particular ion-exchange
` reversed-phase and
`×
`reversed-phase
` reversed-phase LC separations conducted at high and low pH, are shown.
`© 2017 Elsevier B.V. All rights reserved.
`
`Article history:
`Received 2 September 2016
`Received in revised form
`22 November 2016
`Accepted 2 January 2017
`Available online 4 January 2017
`
`Keywords:
`Review
`Stationary phases
`Capillary columns
`Enzymatic reactors
`Two-dimensional liquid chromatography
`Biomolecule analysis
`Spatial chromatography
`
`Contents
`
`1.
`2.
`
`Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
`Preparation and properties of polymer-monolithic capillary columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
`2.1.
`Column and material properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
`2.2.
`Tuning of the macroporous structure and performance characterization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
`2.3.
`Optimization of the surface chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
`3. High-performance monolith chromatography: practical aspects and applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
`3.1.
`LC–MS profiling of protein isoforms and antibodies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
`3.2. High-resolution separations of oligonucleotides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
`3.3.
`Analysis of protein digests and on-line enzymatic microreactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
`4. Monolithic columns for multi-dimensional LC separations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
`×
`4.1.
`Ion-exchange
` reversed-phase separations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16
`×
`4.2.
`Reversed-phase
` reversed-phase separations at high and low pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
`Concluding remarks, perspectives, and challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
`Acknowledgements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18
`References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
`
`5.
`
`Abbreviations: DVB, divinylbenzene; ESI, electrospray ionization; FA, formic acid; HFBA, heptafluorobutyric acid; HIC, hydrophobic interaction chromatography; HILIC,
`LC, two-dimensional liquid chromatography; MALDI, matrix-assisted laser desorption
`hydrophilic interaction chromatography; IEX, ion-exchange chromatography; LC
`ionization; MS, mass spectrometry; RP-LC, reversed-phase liquid chromatography; SEC, size-exclusion chromatography; S, styrene; TFA, trifluoroacetic acid; TIC, total ion
`current; TOF, time of flight; UHPLC, ultra-high-pressure liquid chromatography.
`∗ Corresponding author at: Pleinlaan 2, B-1050, Brussels, Belgium.
`E-mail address: seeltink@vub.ac.be (S. Eeltink).
`
`×
`
`http://dx.doi.org/10.1016/j.chroma.2017.01.002
`0021-9673/© 2017 Elsevier B.V. All rights reserved.
`
`MTX1073
`ModernaTX, Inc. v. CureVac AG
`IPR2017-02194
`
`1
`
`

`

`S. Eeltink et al. / J. Chromatogr. A 1498 (2017) 8–21
`
`9
`
`1. Introduction
`
`Rigid polymer-based monolithic columns have been introduced
`in the early 1990s as an alternative for packed-bed columns [1,2].
`The first monolithic entities were based on methacrylate and
`styrenic precursors and were polymerized in large i.d. columns,
`i.e., a mold, using conditions that were typically applied during
`a suspension polymerization for the preparation of porous poly-
`mer beads. The resulting column structure featured a macroporous
`interconnected structure of polymer globules. The potential of this
`novel column technology was readily demonstrated with high-
`speed separations of macromolecules, including intact proteins [3]
`and synthetic polymers [4]. The success of these columns, typically
`operated at high flow rates, in the separation of height molecular-
`weight analytes can be attributed to mass-transfer effects that
`are mainly based on convention (Cm-term). Since the polymer
`globules feature very few mesopores (pores in the size range of
`2–50 nm) stationary-phase mass-transfer (Cs) greatly affecting dis-
`persion characteristics when operating the column at high flow
`rate, are virtually absent. Although not within the scope of this
`review, it should be mentioned that the good kinetic performance
`for biomolecule separations is in sharp contrast with the effi-
`ciency typically reported for small-molecule separations [5–7].
`However, considerable efforts and progress has been made to cre-
`ate monolithic columns suitable for small-molecule separations.
`The different strategies pursued were recently reviewed by Urban
`and Svec [8,9].
`The ease of the preparation process of monolithic materials
`facilitated the development of miniaturized column formats, typi-
`cally capillaries and microfluidic chips, which are used for peptide
`and protein profiling in clinical diagnostics and life-science LC–MS
`research. Since the monolithic interconnected structure is cova-
`lently linked to the inner capillary wall, frits used in particle-packed
`columns are no longer needed and very robust column formats
`are obtained. Geiser et al. demonstrated the possibility to conduct
`almost 2200 consecutive separations of a test mixture of three
`proteins without any significant shift in either retention time or
`column pressure [10]. More recently, Urban et al. demonstrated
`the durability of monolithic capillary columns with over 10,000
`injections [11]. A landmark paper was published by the Huber
`research group in 2000, showing baseline resolved oligonucleotide
`LC–MS separations utilizing monolithic capillary columns based
`on ion-pair reversed-phase (RP) interactions [12]. The success of
`monolith chromatography is further amplified by the great vari-
`ety of functional and crosslinking monomers available, allowing to
`create monoliths carrying the desired surface chemistry to achieve
`LC separations in different modes. In this way, a large variety of
`monolithic materials have been created that were then applied for
`biomolecule separations. For example, monolithic columns were
`developed enabling high-resolution biomolecule separations based
`on ion-exchange chromatography (IEX) [13–16]. Hilder’s group
`reported macroporous polymer-monolithic stationary phases for
`hydrophobic interaction chromatography (HIC) of intact proteins
`[17].
`This review discusses advances in the preparation, characteri-
`zation, and application of polymer-monolithic column technology
`with a focus on fast and high-resolution peptide and protein
`LC–MS separations. First, material properties and the possibili-
`ties and limitations in controlling the macropore structure and
`surface chemistry are described. Next, selected examples of sep-
`arations are provided demonstrating the potential to analyze a
`wide range of biomolecules, from peptides to monoclonal antibod-
`ies, using one-dimensional and comprehensive multi-dimensional
`separation strategies. Finally, challenges and perspectives for high-
`resolution monolith chromatography are shown.
`
`2. Preparation and properties of polymer-monolithic
`capillary columns
`
`2.1. Column and material properties
`
`Fused-silica capillary column formats are typically applied
`in proteomics research because of their flow-rate compatibility
`with nanoelectrospray ionization in mass-spectrometric (ESI–MS)
`detectors. The polymerization mixture typically comprises the ini-
`tiator dissolved in a homogenous solution of monovinyl and divinyl
`monomers and inert pore-forming diluents (porogens). Typically,
`the surface of the fused-silica inner capillary wall is first func-
`tionalized with a silane spacer, such as (trimethoxysilyl)propyl
`methacrylate, where the methoxy groups react with the silanol
`groups situated at the activated fused-silica capillary surface. Pen-
`dant vinyl groups subsequently react with the monomers present
`in the polymerization mixture. A detailed investigation of mono-
`lith anchoring approaches was conducted by Courtois et al. [18].
`The best silanization approach for fused-silica used toluene as sol-
`vent for the silanization agent. Nesterenko et al. optimized the
`bonding procedure to establish a covalent bond between poly-
`mer monoliths and the wall of titanium housings [19]. Monolithic
`capillary columns prepared using this approach proved to be very
`robust and stable even at ultra-high-pressure operating conditions
`( P = 80 MPa) [20].
`Recently, procedures to chemically anchor polymer monoliths
`to the wall of polyetheretherketone tubing (PEEK) and to polyimide
`microfluidic chips have also been described [21–23].
`Two most often applied polymeric monolithic stationary phases
`for biomolecule separations are based on (meth)acrylate and
`styrenic monomers. Both methacrylate-ester-based monoliths and
`styrene-based monoliths exhibit good stability in separations using
`acidic and basic mobile phases. Whereas methacrylate monoliths
`are stable between pH 2 and 12 [24], styrene-based monolithic
`columns were stable even at a pH up to 14 [25]. Both types of
`these materials typically feature a very small volume of meso-
`pores, and as a result, those monoliths typically exhibit surface
`areas of less than 50 m2/g [26]. Whereas this may enhance the
`separation efficiency by minimizing the Cs-term contribution to
`band broadening, the smaller surface area compared to columns
`packed with fully porous particles negatively affects mass loadabil-
`␮m i.d. capillary
`ity. The loading capacity of oligonucleotides on 200
`poly(S-co-DVB) monolithic columns was investigated by Oberacher
`et al. by measuring the peak width while applying a 10 min RP-
`LC gradient [27]. The maximum mass loadability was found to be
`500 fmol (2.4 ng). Detobel et al. found that the mass loadability
`for a similar type of monolith for intact proteins strongly affected
`peak width, i.e., injecting 10 pg carbonic anhydrase resulted in 2–3 s
`wide peaks (measured at half height), which increased linearly with
`injected protein mass [28]. Asymmetric peaks, clearly indicating
`overloading, were observed in a range of 50–100 pg. The limited
`mass loadablity of polymer monolithic stationary phases should be
`taken into account when profiling low-abundant biomarkers.
`The DVB crosslinker comprises a 2:1 mixture of meta-DVB
`and para-DVB and is commercially available in two grades with
`65% and 80% purity with the rest being mostly ethylvinylben-
`zene [29]. During the course of polymerization, DVB is depleted
`earlier than S, which results in the formation of a crosslink den-
`sity gradient with the outer layer of the polymer globules being
`less crosslinked. Poly(S-co-DVB) monolithic material prepared in
`bulk quantity (∼1 g) was investigated applying thermal analysis
`techniques [30]. Experiments confirmed the presence a broad dis-
`tribution of crosslinking densities and at a temperature of 100 ◦C
`the least crosslinked outer shell of the globules in monolithic mate-
`rial devitrified. In practice however, the thermal stability of proteins
`depends on the specific protein tested and the exposure time. To
`
`2
`
`2
`
`

`

`10
`
`S. Eeltink et al. / J. Chromatogr. A 1498 (2017) 8–21
`
`prevent breaking of intra-molecular disulfide bonds and hydrolysis
`of the amide bonds, the maximum column temperature is typically
`set at 60 ◦C during RP-LC analyses. At this temperature, the mono-
`lithic polymers are completely stable [30]. The excellent long-term
`stability of poly(S-co-DVB) monoliths was confirmed by Andersen
`et al. using nano-LC gradient experiments with cytochrome c and
`lysozyme as test analytes [31].
`
`2.2. Tuning of the macroporous structure and performance
`characterization
`
`The first generation of polymer-based monolithic stationary
`phases contained relatively
`large macropores and, as a con-
`sequence, exhibited excellent column permeability [3,4]. The
`introduction of ultra-high-pressure instrumentation in 2001 [32]
`and availability of new approaches to visualize the kinetic
`performance, such as kinetic plots [33–35], have spurred the
`improvement of packed-column technology, i.e., columns packed
`with sub-2-micrometer particles. Whereas particle size affects the
`separation efficiency and permeability of packed columns, similar
`principles also apply when considering the macropore structure
`of monolithic materials. However, the total porosity of packed
`columns is fixed, while the porosity, macropore size, and globule
`size of monolithic stationary phases can be tuned independently.
`Hence, monolithic columns have the intrinsic potential to perform
`better than conventional columns packed with particles both in
`terms of speed and efficiency [36]. Another key aspect is that the
`homogeneity of the macropore structure, affecting the extent of
`eddy dispersion, must be maximized in order to reach the desired
`high column efficiencies.
`To create monolithic supports for use in affinity chromatogra-
`phy or as an enzymatic microreactor, the porous structure has to
`be optimized accordingly. On the one hand, monolithic polymer
`featuring a large surface area is preferred, while on the other hand,
`toleration of high flow rates (e.g., during enrichment) and the ability
`to analyze ‘dirty’ sample matrices are desired. This requires a trade-
`off between monoliths with relatively large flow-through pores
`and monoliths featuring a more homogeneous monolithic struc-
`ture containing both small macropores and globules to minimize
`the eddy dispersion.
`The most common mechanism used to prepare polymer-based
`monolithic stationary phases is the free-radical copolymerization
`of monovinyl and divinyl monomers (crosslinker) in the presence
`of an inert diluent (porogen) that is initiated either via heating or
`UV light. The amount of initiator, ratio and type of monovinyl and
`divinyl monomers, the composition of the porogen, and also con-
`ditions, such as temperature and time, strongly affect the phase
`separation during the polymerization reaction and the subsequent
`formation of the macroporous polymer network. In the early 2000s,
`a series of excellent papers were published studying the effects of
`reaction conditions and polymerization kinetics on pore size for
`different monolithic materials created via both free-radical and liv-
`ing polymerization techniques [37–40]. More detailed information
`about the effect of polymerization conditions can also be found in
`recent reviews [41,42].
`To advance tuning of the macroporous structure and improve
`the structural homogeneity of polymer monoliths, several novel
`approaches have been pursued recently. Fig. 1 shows scanning
`electron micrographs of monolithic polymers prepared using dif-
`ferent techniques. Mai et al. reported the preparation of monolithic
`polyamide using thermally-induced phase separation that is based
`on swelling and dissolution of the polymer at elevated tempera-
`ture, followed by precipitation achieved via cooling the solution
`below the upper critical-solution temperature [43]. This approach
`was also applied by Yoneda et al. who investigated adjustment of
`the pore structure of poly(methyl methacrylate) monolithic mate-
`
`Fig. 1. Scanning electron micrographs highlighting different approaches for tuning
`the macropore structure of organic polymer-based monolithic columns. (A) SEM of
`a monolithic cryogel that incorporates nanoparticles acting as structure-directing
`agent and enhancing the binding capacity. Adapted from [47] with permission.
`(B) SEM of aligned macroporous channels created via freeze casting; a monomer-
`contained solution was frozen using liquid nitrogen. The crystals grow into a
`continuous frozen framework and are interspersed with the non-frozen monomer
`phase. After being polymerized, the bicontinuous pore system is preserved and a
`highly porous monolith is formed. Adapted from [48] with permission. (C) Example
`of a macroporous poly(S-co-DVB) monolithic stationary phases featuring a globular
`nanostructure. Adapted from [50] with permission.
`
`3
`
`3
`
`

`

`S. Eeltink et al. / J. Chromatogr. A 1498 (2017) 8–21
`
`11
`
`rials via tuning of the molecular weight of the precursor polymer,
`cooling temperature, and porogen composition [44]. Hilder’s group
`described the preparation of composite monolithic cryopolymers
`using directional freezing in combination with photo-initiated cry-
`opolymerization [45]. They tuned the porous properties via changes
`in the monomer concentration and the immersion rate. This group
`also investigated cryopolymerization while adding positively or
`negatively charged nanoparticles to the polymerization mixture
`[46,47]. A SEM of a macroporous monolithic crypolymer embed-
`ding charged nanoparticles is depicted in Fig. 1A. Freeze casting also
`was explored by Du et al. in an attempt to control crystal growth and
`hence to create monoliths featuring aligned macroporous channels
`shown in Fig. 1B [48]. The monolith featured macropores with a size
`␮m and the three-dimensional pore structure
`of approximately 10
`exhibited high similarity with the crystal shape and crystal growth
`direction. Liu et al. reported a photoinitiated thiol-yne click poly-
`merization approach for the preparation of monolithic stationary
`phases with homogenous macropore structure [49]. In contrast to
`polymer monoliths prepared via common free-radical polymeriza-
`tion, high-efficiency separations of small molecules were achieved
`(∼90,000 N/m) indicating low A- and C-term contributions to band
`broadening (with the A- and C-term contributions being inversely
`proportional to the plate height, H). The potential of these columns
`was also demonstrated with a 60 min gradient of mobile phase used
`for the separation of a tryptic BSA digest and separation of a mixture
`of five intact proteins in a 30 min long gradient, albeit no baseline
`separations were obtained [49].
`In a recent study Vaast et al. described the development
`of nanostructured poly(S-co-DVB) monolithic capillary columns
`for high-speed gradient separations of peptides
`[50]. They
`systematically varied quantity of
`initiator, porogenic solvents
`ratio (THF/1-decanol), and crosslinker content. The morphol-
`ogy of resulting materials was visualized with scanning electron
`microscopy and the kinetic performance was assessed by measur-
`ing the plate height as a function of mobile phase flow velocity,
`providing insights in the relation between column structure and
`eddy-dispersion and mass-transfer contributions to band broaden-
`ing. In analogy to the effect of particle size on efficiency in packed
`columns, a similar effect was observed taking into account the
`effects of macropore structure of monoliths characterized by glob-
`ule and macropores size and their distributions on the peak width.
`Decreasing the macropore or globule size will generally lead to an
`increase in the separation efficiency. Using a 50 mm long mono-
`lithic capillary column featuring macropores and polymer globules
`in the 50–200 nm range (see the SEM in Fig. 1C) the peptides of the
`tryptic digest of cytochrome c were nearly baseline resolved in a
`1 min gradient UHPLC run [50]. It was noted that the performance
`of monoliths with the smallest domain size was dominated by eddy
`dispersion. In a step towards the better understanding of the effect
`of column structure on separation performance, Wouters et al.
`also investigated the morphology of the large and small domain
`size polymer monoliths using complementary physical charac-
`terization techniques and nano-LC experiments to link structure
`inhomogeneity to eddy dispersion [26]. Although small domain size
`monoliths featured a relatively narrow macropore size distribution,
`the desired homogeneity was negatively affected by the presence
`of a small number of large macropores, which induced a signifi-
`cant eddy-dispersion contribution to band broadening. In addition,
`the small domain size monolith featured a relatively steep C-term.
`These experiments again demonstrated the difficulties encoun-
`tered when attempting to create homogenous monoliths with very
`small features.
`
`2.3. Optimization of the surface chemistry
`
`The most common approach to create a monolith with the
`desired surface chemistry is applying a single-step polymerization
`using the functional monomers as a part of the polymerization mix-
`ture. The variety of surface chemistries that can be realized is almost
`unlimited since any monomer can be incorporated in the polymer
`matrix. However, a significant disadvantage is the need for re-
`optimization of the polymerization conditions for each new system
`that is required to achieve the desired macroporous structure.
`An alternative route to create monoliths with the desired
`surface chemistry is to incorporate a monomer that contains a
`reactive group in the polymer that can be used for subsequent
`modification by the parent monolith with a specific reagent. For
`example, glycidyl methacrylate monoliths are widely used to pre-
`pare ion-exchange columns. Also, a wide range of affinity media
`are obtained via conversion of the epoxy groups with ligands
`[51,52] including antibodies and antibody-binding proteins, lectins
`[53], serum albumins [54], but also chelating agent such as Ca2+
`and Fe3+ [55]. The epoxy groups of poly(glycidyl methacrylate-co-
`ethylene dimethacrylate) monoliths can also react with cysteamine
`to afford thiol functionalities that bind gold nanoparticles [56].
`Such monolithic column exhibited affinity for peptides contain-
`ing thiol functionalities. Lv et al. discussed the immobilization
`of Porcine lipase on a monolithic polymer support containing
`thiol functionalities and functionalized with gold nanoparticles
`[57]. In-situ copolymerization of 2-vinyl-4,4-dimethylazlactone
`and ethylene dimethacrylate produced monoliths designed for
`immobilized enzyme microreactors including trypsin [58] and
`pepsin [59]. The lateral amine functionalities of lysine in the pep-
`tide chain react in a single step with the azlactone functionality
`in a ring-opening reaction. In contrast, similar immobilization via
`epoxy groups using glycidyl methacrylate-based monolith requires
`several reaction steps. These enzymatic microreactors allow in-line
`digestion applied in bottom-up proteomics studies. Alternatively,
`different derivatives can also be prepared by functionalization of
`styrene-based monoliths containing chloromethylstyrene as func-
`tional monomer [60].
`The third approach affording monolithic columns with desired
`chemical functionality at the pore surface is grafting. After prepar-
`ing the generic monolith with optimized porous structure, the
`monolith is grafted with selected functional monomer using a poly-
`merization reaction. For example, one approach involves hydrogen
`abstraction from the pore surface using aromatic ketone as the
`photoinitator. The grafting reaction is then initiated with UV light
`and branched polymer architecture is grafted at the pore sur-
`face. A variety of functionalized monoliths was prepared using this
`technique including monoliths containing negatively [61] and pos-
`itively charged functionalities [62], monoliths with significantly
`reduced adsorption of proteins [63], and monoliths co-grafted
`with ionizable and hydrophobic moieties [64]. Using a UV initi-
`ation through a photomask allowed Stachowiak et al. to prepare
`a microfluidic reactor with multiple spatially localized enzymes
`[65]. Vonk et al. applied post-polymerization photografting to form
`methacrylate ester-based monolithic columns with ion-exchange
`functionalities and used them as the first-dimension column in a
`comprehensive two-dimensional IEX×RP chromatography to ana-
`lyze intact proteins and protein digests [66]. Using an experimental
`design involving variations of monomer concentration in the graft-
`ing solution and grafting time, different photografting conditions
`were tested. However, this experimental design did not allow to
`predict optimal and robust conditions.
`
`4
`
`4
`
`

`

`12
`
`S. Eeltink et al. / J. Chromatogr. A 1498 (2017) 8–21
`
`Fig. 2. The separation of intact proteins, including protein isoforms arising from
`various amino-acid modifications, employing a poly(S-co-DVB) monolithic capillary
`column in high-performance liquid chromatography coupled on-line to a time-of-
`␮m i.d.
`250 mm
`flight mass spectrometer. The separation was performed on a 200
`monolithic capillary column applying a 2 h linear gradient, yielding a maximum
`peak capacity of 680. Adapted from [70] with permission.
`
`×
`
`tion from mouse gut microbiota [74]. Walsh et al. synthesized
`poly(S-co-DVB) monolithic stationary phases using visible light
`at 470 nm for the initiation of the polymerization reaction [75].
`The potential of this stationary phase was demonstrated with
`the gradient separation of a mixture of proteins. Swelling and
`shrinking effects of organic-monolithic stationary phase in a large
`conduit during application of a solvent gradient is likely to affect
`the robustness of large i.d. monolithic columns. To reduce shrink-
`age effects, Vonk et al. developed titanium-scaffolded monolithic
`columns in narrow-bore square conduits (1.3
`1.3 mm) and suc-
`cessfully applied these columns for the UHPLC gradient analysis
`of intact proteins [76]. An additional benefit of this approach was
`minimized viscous-heating effects that have a detrimental effect
`on band broadening [77].
`The combination of monolith chromatography with high-
`accuracy TOF-MS allows to distinguish structural isoforms [70].
`Fig. 3A shows a zoom-in of the TIC chromatogram with m/z as a
`function of retention time obtained using a poly(S-co-DVB) mono-
`lithic capillary column. Based on charge deconvolution of the
`charge envelopes, peroxiredoxin I was tentatively identified and
`the mass differences of 16 Da between the Mr values experimen-
`tally found were indicative for protein oxidation. The cluster of
`peaks can be explained by the presence of single and double oxida-
`tions occurring in different sites including cysteine and methionine.
`Fig. 3B shows the density-view plot of serotransferrin, a very het-
`erogeneously glycosylated protein. Despite the high separation
`power available, only a limited number of protein isoforms could be
`identified based on deconvoluted mass spectra. Arguably, collect-
`ing this fraction and reinjecting it in a long shallow gradient with
`limited gradient span could improve the analysis. Liu et al. inves-
`tigated the possibilities of profiling glycoprotein isoforms using
`a weak-anion exchanger polymer monolithic column containing
`primary amino functionalities [78]. Peak profiles of intact protein
`standards were monitored on 50–60 cm long capillary columns
`allowing to distinguish different glycoisoforms. For example, seven
`
`×
`
`3. High-performance monolith chromatography: practical
`aspects and applications
`
`RP-LC is generally considered to be the most powerful sepa-
`ration mode and, as a result, most monolithic columns that were
`developed so far feature hydrophobic RP functionalities. Separa-
`tions of biomolecules are most often performed in gradient-elution
`mode. Whereas in isocratic mode significant peak broadening
`occurs as a function of retention time, peak broadening in gradient
`mode is minimized due to ‘peak compression’ [67,68]. When apply-
`ing a solvent gradient during the analysis, the tail of the analyte
`peak will elute in a solvent that is stronger compared to the mobile-
`phase composition at the front of the same peak. This causes the tail
`of the peak to move relatively faster, which results in peak compres-
`sion. Vaast et al. conducted a detailed study describing the effects
`of gradient steepness on peak compression in gradient separations
`of both peptides and intact proteins [69]. Significant peak focus-
`ing, positively affecting peak capacity was observed when applying
`steep gradients upon UHPLC conditions and short monolithic cap-
`illary columns.
`An ion-pairing agent is often added to the mobile phase to
`enhance the electrospray
`ionization efficiency prior to mass-
`spectrometric detection. Although
`trifluoroacetic acid
`(TFA)
`typically provides for slightly narrower peak widths and hence
`better kinetic separations, it has been demonstrated that formic
`acid (FA) significantly enhanced the ionization efficiency [70]. In
`particular, an increase in detection sensitivity for the LC-ESI-TOF-
`MS analysis of intact proteins by a factor of 3 was observed after
`using FA as ion-pairing agent instead of TFA. The use of FA also
`enhanced retention,