Recent Developments in Capillary and Microchip Electroseparations of Peptides (2023–mid 2025)

ABSTRACT

This review presents a comprehensive overview of the developments and applications of high‐performance capillary and microchip electromigration methods (zone electrophoresis in a free solution or in sieving media, isotachophoresis, isoelectric focusing, affinity electrophoresis, electrokinetic chromatography, and electrochromatography) for analysis, microscale isolation, and physicochemical and biochemical characterization of peptides in the period from 2023 up to ca. the middle of 2025. Advances in the exploration of electromigration properties of peptides and various aspects of their analysis, such as sample preparation, sorption suppression, EOF regulation, and detection, are described. New developments in the particular CE methods are presented, and several types of their applications are reported. They include qualitative and quantitative analysis of synthetic or isolated peptides, determination of peptides in complex biomatrices, peptide profiling of biofluids and tissue extracts, and monitoring of chemical and enzymatic reactions and physicochemical changes of peptides. They also deal with amino acid and sequence analysis of peptides, peptide mapping of proteins, separation of stereoisomers of peptides, and their chiral analyses. In addition, micropreparative separations and physicochemical and biochemical characterizations of peptides and their interactions with other (bio)molecules by the above CE methods are described.

Abbreviations

AA

amino acid

AcOH

acetic acid

BFS

bare fused silica

CAs

carrier ampholytes

CITP

capillary ITP

EMMA

electrophoretically mediated microanalysis

FA

formic acid

FLD

fluorescence detection

FS

fused silica

GSH

glutathione

HPC

hydroxypropyl cellulose

IPA

isopropyl alcohol

LPA

linear polyacrylamide

LVSS

large volume sample stacking

MCE

microchip CE

MeOH

methanol

MP

mobile phase

OT‐CEC

open tubular CEC

PDADMAC

poly(diallyldimethylammonium chloride)

PEI

polyethylenimine

PEO

poly(ethylene oxide)

PSP

pseudostationary phase

PSS

poly(sodium styrene sulfonate)

SL

sheath liquid

SMIL

successive multiple ionic polymer layer

SP

stationary phase

t‐ITP

transient ITP

1. Introduction

Peptides are composed of amino acids connected by amide bonds and sometimes crosslinked by disulfide bridges. They represent a huge group of extremely important biomolecules. Acting as hormones, neurotransmitters, enzyme substrates and inhibitors, co‐enzymes, immunomodulators, ligands, receptors, ionophores, antibiotics, and toxins, they play a crucial role in the control and regulation of many vital processes in all living organisms. Moreover, many peptides or peptidomimetics are already used as registered medications or are being developed as potential new drugs [1, 2], peptide‐based carriers for drug delivery [3], and self‐assembled nanotheranostic agents [4]. Antimicrobial and antiviral peptides are considered as potential new antibiotics and antivirals [5, 6, 7]. Peptides function as biomarkers and biosensors for human health [8] and micro‐ and nanobiomaterials for biomedical applications [9, 10]. They are also used in the detection of metal ions [11] and for the selection of aptamers specifically binding the particular proteins and cells [12]. In the current era of proteomics, peptidomics, and metabolomics, the importance of peptides is even increasing. These methodologies are the major analytical tools for a study of the molecular bases of biological processes, as well as for finding new biomarkers and drug targets [13, 14, 15]. These –omics technologies are complex and interconnected disciplines. Both, structure and function of proteins are often investigated by means of their enzymatically generated peptide fragments in the so‐called “bottom‐up” or “shotgun” proteomics [16, 17]. These peptide fragments are separated and identified by LC, 2‐DE, and CE methods online coupled with MS detection [18, 19, 20]. In addition, for understanding both normal and pathological physiological processes, a comprehensive investigation of the whole peptide set (peptidome) of a cell, organ, or organism at a given time frame is necessary as well. This is the subject of peptidomics—a bridge between proteomics and metabolomics [14, 17, 21]. Thus, separation, analysis, isolation, purification, and characterization of peptides by capillary and microchip electromigration methods is one of the most challenging tasks of these high‐performance separation techniques.

This review presents a comprehensive overview of the recent developments of CE and microchip CE (MCE) methods and their application for analyses, characterizations, and micropreparative separations of peptides in the period 2023 – ca. mid‐2025. It is an update of the previous reviews on CE and MCE of peptides [22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34] covering the years 1997–2022. All CE and MCE methods, both the electrophoretic ones (zone electrophoresis (ZE), ITP, IEF, and ACE) and the mixed electrokinetic and electrochromatographic techniques (EKC and CEC), have been intensively developed also in the last two and a half years. Applications of these methods for analysis, micropreparation, and characterization of peptides have been further extended. Nowadays, CE methods are appreciated as recognized counterparts and/or complements of the (U)HPLC, capillary LC, and nano‐LC methods [18, 35] that represent gold standards for peptide separations both in analytical and preparative scale. In addition to the last one of the above reviews [34], the recent developments of CE separations of peptides were described in some other reviews [36, 37, 38, 39], and a book [40] dealing with particular aspects of CE and MCE of peptides and related compounds, proteins, and amino acids (AAs).

2. Electromigration Properties of Peptides and Selection of Separation Conditions

Unlike previous periods, in the last two and a half years, only a few papers were in detail focused on the investigation of electromigration properties and behavior of peptides by CE methods. The relation between ion mobility in gas and liquid phases was studied by CE coupled to trapped ion mobility mass spectrometry (CE‐TIMS‐MS) for separation and analysis of a complex peptide mixture (tryptic digest of HeLa proteins) [41]. With the nanoCEasy interface, excellent sensitivity and high peak capacities (6223–9497, n = 3, mean value 7500) was obtained using a fused silica (FS) capillary internally coated with poly(ethylene oxide) (PEO) (150 cm long, 50/365 µm id/od), 30 kV separation voltage, 1.7 kV spray voltage, and BGE composed of 1 M formic acid (FA) and 10% isopropyl alcohol (IPA) (Figure 1A). Thousands of peptides were identified within a more than 220 min‐long migration time window. The heat map of migration times and mass/charge (m/z) ratio showed a peptide separation leading to distinct, slightly curved distributions due to the distinct charges in the acidic BGE (Figure 1B). The peptide charge in the acidic BGE was approximately estimated from the number of basic AAs in peptide chains, with a few exceptions where the negative charge of neighboring acidic AAs decreased the positive charge. CE in the liquid phase had a higher separation efficiency than the ion mobility (IM) separation in the gas phase (Figure 1B,C). The combined heat map of liquid and gas phase mobility shown in Figure 1D also provided a slightly curved distribution. In addition, a large area of this heat map was covered with data points of peptides showing the different selectivity of CE and IM and demonstrating a surprisingly high orthogonality of CE and MS separations. It is obviously caused by the solvation effects, leading to different charges and sizes in the liquid phase than in the gas phase.

FIGURE 1.

CE separation (peptide map) of tryptic HeLa cells digest using a 150 cm long PEO‐coated FS capillary. (A) Base peak electropherogram (BPE; 350−2200 m/z). (B) Heat map of migration time and m/z value. (C) Heat map of gas phase mobility and m/z value. (D) Combined heat map of liquid and gas phase mobility. Source: Reprinted with permission from Schairer et al. [41].

Two complementary theoretical approaches, mean‐field analysis and molecular dynamics simulations, were employed to estimate how hydration interactions change the electrophoretic mobility of peptide, protein, and DNA molecules and the peptide–G‐quadruplex complex in an aqueous solution [42]. These interactions were found to stabilize the surrounding electric double layer (EDL), resulting in more significant localized counterion concentrations while strengthening the following electrokinetic flow. Major findings of this study were that the hydration of water molecules acts as a glue for forming a stable EDL and that the influence of hydration interactions increases with increasing particle size, surface charge density, and bulk ion concentrations. The presence of larger counterions increases the drag on the particle, thus reducing its mobility. In accordance with previous modeling, the mobility of the above molecules was dependent on size, which can be used for both analytical and preparative separations.

A neural network called CPred was developed for the prediction of accurate charge state (+1 to +7 elementary charge units) of both unmodified and modified peptides in electrospray ionization (ESI) [43]. The network was trained on a large set of training data comprising tryptic and nontryptic peptides and various fragmentation methods. The model was evaluated on independent external test data sets. The Pearson correlation coefficient has shown a high correlation (0.9997) between the predicted and acquired charge state distribution. The accurate prediction of charge states is important for the determination of the m/z ratio that is used as a critical characteristic in peptide and protein identification by MS with ESI that generates multiply charged ions.

Concerning the selection of experimental conditions for peptide separations by CE methods, first, the general rules for the BGE choice presented in the earlier reviews [44, 45] should be followed. Second, the electromigration and other specific properties of peptides, that is, their experimentally determined or calculated effective charge, effective mobility, isoelectric point (pI), hydrodynamic radius, relative molecular mass (M _r), steric structure, solubility, amphiphilicity, amphotericity, binding capabilities, chemical and thermal stability, and biological activity, have to be considered. Various aspects of the BGE selection for CE separations of peptides and proteins, such as type and concentration of the BGE constituents with respect to their buffering capacity, mobility and electric conductivity, pH and ionic strength, additives suppressing adsorption of peptides/proteins to the inner surface of the most common FS capillaries or influencing separation selectivity, EOF direction and velocity, organic modifiers, temperature, Joule's heating effects and compatibility with MS detection, were in detail described and discussed by Corradini [46]. Useful information and practical advice for CZE, CITP, and CIEF analyses of peptides and proteins can also be found in the earlier published but still valid review and tutorial articles [47, 48, 49]. Important parameter for selection of cationic or anionic mode of CZE or CITP separations of peptides is their pI, which can be found in some databases [50], or it can be calculated [51] or experimentally determined by IEF in capillary or slab gel format [49] or by CZE at pH below and above pI [49, 52].

When the effective mobilities of peptides are known or can be estimated, then, for optimization of their CE separations, the freely available powerful simulation programs, PeakMaster and Simul, can be employed [53]. Another new open‐source software for simulation of CITP and other CE separation processes of various analytes, including peptides, both in classical plateau mode and in a more sensitive peak mode, was developed by Santiago's group [54]. The basic parameters for the selection of CZE experimental conditions, such as viscosity and electric conductivity of the BGE, injected and total capillary volumes, proportion and amount of injected sample, electric field strength, and others, can be obtained by the CE calculator freely available as an Android application [55].

3. Sample Preparation

3.1. Preconcentration and Preseparation

Preconcentration and/or preseparation of target peptides (“sample clean‐up”) are necessary for their analyses by CE methods if the sensitivity of the detectors used and/or the separation power of these methods are insufficient for direct analysis of peptides present at low concentrations and/or in complex (bio)matrices, such as body fluids, cell lysates, tissue extracts, foods, and feeds.

Sample preparation procedures for preseparation and preconcentration of (bio)molecules, including peptides in complex (bio)matrices before their CE analyses, are presented in detail and evaluated in general and CE methods related recent reviews [56, 57, 58, 59]. In these reviews and other special papers, the application of various procedures for sample clean‐up are described, such as (ultra)filtration [60], solid‐phase extraction (SPE) [61], microextraction [56, 62], solid‐phase microextraction (SPME) [63, 64], pseudophase‐aided in‐line sample concentrations including sweeping, micelle collapsing, micelle to solvent stacking and micelle to cyclodextrin stacking [65], liquid–liquid extraction [66] and several electro‐driven techniques [58]. Some reviews describe the sample clean‐up techniques for analysis of peptides and proteins, such as comparison of sample preparation workflows for proteomics [67] and peptidomics [68], recent advances in enrichment and MS analysis of glycopeptides and glycoproteins in complex biomatrices [69], and online enrichment of glycoproteins and glycopeptides using boronate‐functionalized poly(glycidyl methacrylate) microparticles [70].

Some papers describe the particular sample clean‐up techniques for the analysis of peptides and proteins. Online phenylboronic acid‐based extraction of glycopeptides from enzymatic digestion of recombinant human erythropoietin increased the sensitivity of CE‐MS analysis of this protein 500 times as compared with direct CE‐MS and 200 times as compared with TiO₂‐SPE‐CE‐MS [61]. A double‐frit particle‐packed FS microcartridge (250/360 µm id/od × 7 mm) filled with phenylboronic acid sorbent was inserted at 75 mm from the injection end of the CE separation capillary. For a particular peptide or specific peptide groups (e.g., phospho‐, glyco‐, or lipopeptides) preconcentration, the binding affinity of aptamers, lectins, or molecularly imprinted polymers [71] was advantageously utilized.

In SPE and SPME, the new advanced mesoporous and nanoporous materials [72], metal‐organic frameworks (MOFs) [73], covalent organic frameworks (COFs) [74] and gold and other metal nanoparticles [75, 76] are becoming useful sorbents for capture of endogenous peptides, especially glycopeptides [61] and phosphopeptides [77], from complex biological samples. The classical offline formats of SPE and SPME are nowadays being replaced by their online setups. Both preconcentration and preseparation are achieved by a solid‐phase packed or monolithic sorbents at the inlet capillary end.

Besides the above chromatographic adsorption‐based principles, peptide preseparation and preconcentration can be obtained by online electrodriven sample‐stacking techniques [58]. They include field amplified sample stacking (FASS) [78], field enhanced or field amplified sample injection (FESI or FASI) [79], large volume sample stacking (LVSS) [80], isotachophoresis [79, 81, 82], transient‐ITP (t‐ITP) [81, 83, 84, 85], electrokinetic supercharging (EKS) [86, 87], dynamic pH barrage junction, and pH‐mediated stacking [88].

On‐line LVSS was used for peptide preconcentration in CE‐UV analysis of four analogs of growth hormone‐releasing hormone (misused by athletes) in urine samples, including two diastereomeric peptides differing by the chirality of only one AA [80]. Separation was carried out in an unusually long polybrene‐coated FS capillary (110/100 cm total/effective (to the UV detector) length × 75 µm id) at a separation voltage of −30 kV (i.e., with cathode at the injection capillary end) and electric current of 70 µA. LVSS with polarity switching, with an injected sample volume equal to 80% of the total capillary volume, was found to be more suitable than the sweeping concentration and allowed the LODs in the range of 75–200 ng/mL. In combination with C‐18 SPE, the LOD was 640 times lower than in the direct CZE. A good repeatability of migration times and peak areas was achieved by regenerating the capillary after each experiment by successive flushing with 50 mM SDS for 7 min, water for 11 min, 100 mM NaOH for 5 min, 0.2% m/v polybrene in water for 5 min, and BGE for 7 min.

A relatively simple but very useful sample preparation in peptide analysis in complex biomatrices such as blood plasma and serum is protein precipitation by acidified ACN [89]. It was used as the solitary procedure for sample pretreatment in CE‐MS monitoring of the decapeptide drug colistin in blood plasma samples.

Another important sample preparation is desalting of the samples containing high concentrations of chlorides, phosphates, nitrates, acetates, and other salts. The strategies for desalting the samples containing high concentrations of salts were described and discussed in the recent review [87].

Desalting is important, especially when CE separation is coupled with MS detection, because salts suppress ionization of peptides and other analytes and contaminate the MS analyzer. The desalting procedures involve SEC, ultrafiltration, and ion exchange to remove volatile salts.

A new type of desalting tip columns (ChocoTip) with unique morphology based on thermoplastic polymer‐coated chromatographic particles provided a high recovery for a wide variety of peptides in MS‐based bottom‐up proteomic analysis of HeLa cell lysate [90] and is also recommended for CE‐MS analyses of salted peptide samples.

In CZE‐MS peptide mapping of proteins and analysis of yeast cell lysate, the desalting of tryptic digests of proteins by SPE‐based pipette tips provided better sequence coverage of proteins, but the sensitivity of the method was comparable with that obtained with nondesalted samples [91]. Thus, desalting does not always have to be performed before CE analysis. In some cases, the presence of salts can be used for the concentration of peptides by transient ITP.

Recently, electromembrane extraction (EME) has been introduced as a new sample preparation method for preseparation and preconcentration of peptides. In a wide, comprehensive study, almost 6000 peptides differing in charge, size, and hydrophobicity were extracted from the peptide pools dissolved at a 1 µg/mL concentration in aqueous donor phase at pH 3.0 [92]. The liquid membrane was composed of 2‐nitrophenyl octyl ether and carvacrol (1:1, m/m) containing 2% (m/m) di (2‐ethylhexyl) phosphate, and the acceptor phase was 50 mM H₃PO₄, pH 1.8. After 45 min of electroextraction at 10 V, a higher peptide content in the acceptor phase than in the donor phase was found for ca. 3700 peptides. These results confirmed a wide EME applicability for a clean‐up of peptide samples. In another study of the same group, the deep eutectic solvent (combination of camphor and 1‐decanoic acid (in a 1:1 m/m ratio) with 2% di(2‐ethylhexyl) phosphate (DEHP)) was found as the best organic liquid membrane for parallel EME of 13 model peptides from the sample donor phase, pH 6, to the acceptor phase, pH 2.0 [93].

3.2. Derivatization

Sample preparation procedures also comprise the derivatization of analytes of interest, in this context, peptides. Derivatization is carried out to increase the sensitivity of their detection and/or selectivity of their separation. Peptides are mostly derivatized with fluorescent labels, such as fluorescamine, FP‐488‐maleinimide, Alexa‐488‐maleinimide, CF‐488‐maleinimide, or Fluorescein‐X5‐maleinimide, to become detectable with (laser‐induced) fluorescence (LIF) detectors. The sensitivity of these detectors is 2–3 orders higher than that of the most frequently used UV–vis spectrophotometric absorption detectors that do not need any derivatization for peptides.

Derivatization can be carried out both offline and online. In the latter mode, the derivatizing reagents can be components of the BGE or the sample solution, or they can be introduced into the capillary in their independent zone in front of or behind the sample zone [94]. For other examples of precapillary, in‐line, in‐capillary, and post‐capillary derivatization for separation and detection of variable compounds, including peptides in CE with UV, LIF, and MS detection, see the recent review [95].

4. Suppression of Peptide Adsorption and Control of Electroosmotic Flow

Adsorption of peptides, especially the hydrophobic and basic polypeptides, to the inner surface of the FS capillaries and glass, quartz, or plastic chips is one of the most serious issues in their analyses by CE methods. It significantly decreases separation efficiency and may cause partial or even complete loss of analyzed peptides. Thus, suppression of adsorption is still a challenge in CE of peptides, proteins, and other analytes. Besides the separations in strongly acidic or alkaline BGEs or in BGEs with high ionic strength, various types of capillary and microchip coatings are used. Dynamic coatings are based on reversible (dynamic) adsorption of small ions, for example, oligoamines and detergents (anionic SDS or cationic CTAB), or hydrophilic polymers and cellulose derivatives, methylcellulose [96] or methyl‐hydroxyethyl cellulose [97], added to the BGE to modify the capillary or microchannel wall. Static (permanent) modifications of the inner capillary or microchannel wall include covalent bonding or strong physical adsorption of the synthetic and natural polymers, for example, neutral linear polyacrylamide (LPA) [98], PVA [98, 99], poly(vinyl pyrrolidone) [99], PEO [100], PEG [101], fluorocarbon [96, 102], and linear and grafted cellulose derivatives (methyl‐, ethyl‐, hydroxypropyl‐) [96, 103], or charged polymers, such as cationic polybrene (hexadimethrine bromide) [80], poly(diallyldimethylammonium chloride) (PDADMAC) [101], and polyethylenimine (PEI) [104], and anionic poly(sodium styrene sulfonate) (PSS) [101] or polydopamine [105].

The latest applications of all three above types of column coatings for CE of proteins that can also be used for peptides are presented in the recent reviews [39, 106]. The permanent electrostatically immobilized bilayer or multilayer coatings, so‐called successive multiple ionic polymer layers (SMIL), relatively successfully suppress peptide and protein adsorption to the capillary. However, their efficiency strongly depends on the character of the polyelectrolyte layers creating the SMIL coating.

Various parameters of this coating, such as polyelectrolyte nature, molar mass and concentration, effect of the construction buffer pH and ionic strength, and the influence of the BGE pH and composition on the coating performance were studied by the Cottet group [101]. The influence of the chemical nature and the PEGylation of the last polycationic layer in the SMIL coatings on the previously adsorbed four layers (PDADMAC/PSS)₂ on the separation performance of a model mixture of four proteins (myoglobin, trypsin inhibitor, ribonuclease A, and lysozyme) was evaluated. From the five polycationic electrolytes tested, PDADMAC, PEI, ε‐poly(L‐lysine), α‐poly(L‐lysine), and poly(allylamine hydrochloride) (PAH), the last one was the best candidate to test the impact of PEGylation in this fifth layer. The SMIL coating with PEGylated last PAH layer generated the resulting electroneutral layer and allowed highly efficient separation of the above proteins [101]. In the subsequent study [107], the above parameters and some experimental conditions for the construction of efficient and stable SMIL coatings were tested, and the repeatability and intra‐ and intercapillary precision were evaluated. The five‐layer PDADMAC/PSS coating provided the best separation of model proteins (trypsin inhibitor, myoglobin, ribonuclease A, and lysozyme) using 2 M AcOH, pH 2.2, as a BGE and −10 to −25 kV as separation voltage in a 50 µm id FS capillary with 40/31.5 cm total/effective lengths. The high performance of this type of coating was confirmed by highly efficient CE separation of proteins, achieving the plate heights less than 5 µm [108] and the separation efficiency close to one million theoretical plates per meter [109]. PEGylated polycation (poly(allylamine hydrochloride)) was successfully used as the last layer in the polyelectrolyte multilayer (PDADMAC/PSS)₂, also to suppress the adsorption of dendrigraft poly(L‐lysine) and to control the EOF velocity [110]. These optimized coatings and experimental conditions can also be recommended for CE analyses of peptides.

Another types of FS capillary coating suppressing the adsorption of peptides, proteins, and other analytes to the inner capillary wall are the surfactant‐based coatings. The recent review [111] has summarized the application of various neutral, cationic, anionic, zwitterionic, and amphoteric surfactants (SAs) in CE of proteins. From them, especially the double chain SAs, such as didodecyldimethylammonium bromide (DDAB) and dioctadecyldimethylammonium bromide (DODAB), were found as the best constituents of the FS capillary coatings. They provide a very homogeneous surface charge, resulting in high separation efficiencies up to one million plates per meter, and can also be well applied for CE separation of peptides.

In addition to suppression of peptide adsorption to the inner capillary wall, dynamic and/or permanent capillary coatings are also utilized for regulation of the EOF velocity/mobility. It is also a relevant parameter in CE separations because the EOF velocity/mobility also influences the separation efficiency, resolution, and speed of CE analyses of peptides and other analytes. Modulation of both EOF and protein/peptide interactions was achieved by coating the FS capillary with silylated amino‐amide blocks and covalent grafting [112].

A new type of covalent polymeric cationic coating with tuneable EOF velocity composed of poly(acrylamido‐co‐(3‐acrylamidopropyl) trimethylammonium chloride (PAMAPTAC) was used for CE‐MS‐based top‐down proteomics [113], but it can also be applied for CE separation of peptides. Due to repulsion between positively charged proteins and peptides in acidic BGEs and cationic coating, the adsorption of analytes is suppressed, and a higher separation efficiency is achieved than in capillaries coated with neutral linear polyacrylamide. The variable EOF is controlled by various ratios of charged and noncharged monomers covalently attached to the silanized inner wall of the FS capillary in the same way as in [114, 115].

A new approach for capillary coating is based on chemical vapor deposition of various silanes using a lab‐made device [116]. Via capillary modification with neutral hydrophilic, neutral hydrophobic, weakly basic, or weakly acidic silanes, both cationic and anionic EOF with variable mobility from −30 × 10⁻⁹ to +70 × 10⁻⁹ m² V⁻¹ s⁻¹. Up to now, they have been applied only for CE of inorganic anions, but they can also be useful for CE analyses of peptides.

In addition to chemical ways, EOF can also be controlled physically, in particular, electrically by the external radial electric field. It was shown already several years ago [117], and this concept was also revisited recently. The EOF was induced by the radial electric field generated across the thin capillary wall due to the difference in electric potential inside the capillary and the external metal component at the outer capillary wall held at zero (ground) potential [118]. Depending on the polarity and magnitude of the separation voltage and position of the external metal plate close to the capillary external wall, the EOF could be increased, decreased, or reversed. Improved resolution and higher separation efficiency were demonstrated on the CE separation of amino acids, but it can also be achieved in the CE separation of peptides.

Peptide and protein adsorption to the inner surface of the separation compartment and EOF regulation represent serious problems also in their CE analyses in glass/quartz or plastic microchips [119, 120].

5. Separations by the Particular CE Methods

5.1. Zone Electrophoresis

ZE is the simplest, universal, and major mode of electrophoretic methods. In a free solution in rotating glass tubes of 3 mm id, it was introduced by Hjertén [121] and in the PTFE narrow bore tubes of 200 µm id by Mikkers et al. [122]. The modern era of CZE in FS capillaries with an id <100 µm was started in 1981 by Jorgenson and Lukacs [123]. It continues up to current days and is most often used the CE method for separation and analysis of the majority of charged analytes, including peptides [38, 124]. According to IUPAC terminology [125], ZE mode is usually meant when CE is presented without any further attribute of the separation mode. In this section, only a few special topics of peptide CZE separations are presented; the majority of the CZE analyses are described in Section 7, Applications.

CZE is mostly performed in a free solution, in which the separation of analytes according to their effective mobilities is controlled by the ratio of their effective charge and the hydrodynamic radius. Alternatively, CZE is carried out in sieving media of classical agarose or polyacrylamide gels or in physical gels of entangled networks of high‐molecular‐mass linear polymers, for example, LPA or polysaccharides. CZE in such media is called capillary gel electrophoresis (CGE) or capillary sieving electrophoresis (CSE), and the analytes are separated according to their hydrodynamic radius, which is directly related to their M _r.

Various aspects of CGE, including separation principles, methodologies, a short historical overview, advantages and disadvantages of CGE as well as its new developments and applications for separation of proteins and polypeptides in biopharmaceutical, biomedical, and agricultural areas, were summarized in three recent reviews [124, 126, 127]. Native CGE and SDS‐CGE in agarose gels with UV and ESI‐MS detection were applied for the separation of peptides and proteins [128]. Three peptides (bradykinin, angiotensin II, and neurotensin) and three proteins (lysozyme, ribonuclease A, and insulin) were separated by native CGE in BFS capillary (80/70 cm total/effective length, 50/365 µm id/od) using 4% formic acid (FA), 20% glycerol, pH 2.3, as BGE, and 0.6% agarose as a sieving medium.

In addition to mostly used aqueous BGEs, CZE can also be performed in organic solvent‐based BGEs. Then it is called nonaqueous CE (NACE). Because of the mostly ionogenic and polar character of peptides, it is not often used for peptides. Recently, there has been only one example. NACE coupled with QTOF‐MS detection was employed for the determination of cyclodepsipeptides‐based emerging mycotoxins, enniatins, and beauvericin, in wheat samples [129]. Using the BGE composed of 40 mM ammonium acetate in 80:20 (v/v) ACN‐methanol (MeOH) mixed solvent in an 80 cm long, 50/363 µm id/od bare FS (BFS) capillary and 30 kV separation voltage at 20°C, the analysis was accomplished within 4 min and the LOQs were in the range 4.0‐8.3 µg/kg.

Selectivity of CZE can be improved by various additives to the BGE, such as ionic liquids [130, 131], deep eutectic solvents [132], and liposomes, lipid aggregates, and lipid emulsions [133].

The throughput of CZE analyses can be significantly increased when samples are injected in the sequential injection mode [134]. In this mode, the analysis of the first introduced sample is stopped after some time (separation voltage switched off), a new sample is introduced, the separation voltage is switched on, and analysis of both samples continues. Using this approach, the throughput of analyses of important peptide drugs, triptorelin and lanreotide, in pharmaceutical and biological matrices was increased by multiple sample injection with an 80–100 s time interval between the individual sample injections [97], see Figure 2. Analyses were performed in an unusual polytetrafluoroethylene capillary of a relatively large id of 300 µm and short length of 9 cm within a nonconventional hydrodynamically closed system at special conditions of constant current of 150 or 170 µA. Using 50 mM FA as BGE with the addition of 0.05% methyl‐hydroxyethyl cellulose as the dynamic coating, suppressing the EOF, the LOD was 250 ng/mL for peptides in water and 500 ng/mL for peptides in synthetic urine.

FIGURE 2.

Electropherogram from the simultaneous CE analysis of lanreotide (L1, L2, and L3) and triptorelin (T1, T2, and T3) therapeutic peptides using the repeated injection (RI) procedure. The separation was performed in BGE composed of 50 mM formic acid (FA) with 0.5% (v/v) methyl‐hydroxyethyl cellulose (m‐HEC). The RI time interval was 100 s, and the analysis was performed with a separation current of 170 µA. The concentration of the peptides was 1 µg/mL. Source: Reprinted with permission from Stefanik et al. [97].

5.2. Isotachophoresis

ITP is, in fact, a special name for displacement electrophoresis [135, 136]. It is the first electrophoretic method implemented in glass and PTFE narrow tubes (capillaries) with id 400–200 µm [137]. ITP is performed in a discontinuous electrolyte system composed of a leading electrolyte (LE) (containing the leading ion with the highest effective mobility of the system and the buffering counterion) and a terminating electrolyte (TE) (comprising the terminating ion with the lowest effective mobility of the system and an arbitrary counterion). The solution of analytes is introduced between LE and TE, and the analytes are separated according to their effective mobilities, that is, according to charge‐to‐size ratio, in sharply separated neighboring zones between LE and TE. With the exception of bidirectional CITP [81], conventional CITP operates in cationic or anionic mode, and within one experiment, it can separate either cations or anions [124]. The history and developments of ITP in the last 50 years were summarized by Malá and Gebauer [81], and the latest developments and applications of CITP were reported in another review [82]. Theory and microfluidic applications of ITP are presented in an earlier comprehensive review [138]. A web‐based open‐source tool for ITP [54] and a fast, highly parallel simulation tool for the design of ITP electrolyte systems [139] were developed by the Santiago group. The latter one includes a searchable database of 521 commonly used electrolytes and provides the relevant quantities—analyte concentrations and effective mobilities, and pH in steady state ITP zones. Thus, this tool makes it possible to screen a wide range of experimental conditions. In the calculations, the effects of ionic strength and finite ionic radius are taken into account.

In the case of peptide analysis, CITP is more often used for the determination of low molecular mass peptide counterions, see, for example, [140, 141], than for analysis of peptides themselves. However, ITP or transient ITP, t‐ITP, is frequently used for preconcentration of peptides at low concentration levels before their separation and analysis by CZE [79, 81, 82, 85].

5.3. Isoelectric Focusing

IEF separates peptides, proteins, and other amphoteric compounds according to charge distribution in their molecules, manifested by differences in their pIs. In the capillary format, it has been performed since 1985 [142]. CIEF is recognized especially for its high resolving power and the concentrating and self‐sharpening effects. In CIEF, the BGE is composed of a complex mixture of amphoteric carrier ampholytes (CAs) with various pIs. In the electric field, they generate a pH gradient with lower (usually acidic) pH close to the anode and higher (usually alkaline) pH at the cathodic capillary end. After some time, peptides and proteins are concentrated in the self‐sharpening zones at a pH equal to their pIs. To detect them, the zones of peptides and proteins have to be hydrodynamically or chemically mobilized to reach the UV‐absorption or MS detector [143]. The detailed studies of CIEF dynamics have shown that CIEF is, in fact, a quasistationary method. After some time, the CIEF is transformed into bidirectional CITP [144, 145].

Various modes and platforms of both conventional CIEF and whole column imaged CIEF (iCIEF) systems, as well as usual issues in CIEF of mAbs (that can also occur in CIEF of peptides), are described in the recent review [146]. Valuable information on CIEF of peptides and proteins, such as choice of CAs, pI markers, sample application and mobilization, avoiding isoelectric precipitation, immobilized pH gradient, and selected applications, can also be found in the earlier published tutorial and review articles [49]. The role of pI markers and recent applications of CIEF are discussed in [147].

CIEF coupled with ESI‐MS detection via a special interface (reducing the CAs in‐line after CIEF separation) enabled baseline resolution of tryptic peptide fragments of BSA with as little as 0.02 pI difference and a peak full width at half maximum of 7.1 s [148]. The CE setup was composed of a ca. 60 cm long 100/360 µm id/od LPA‐coated FS capillary, ca. 5 cm long Nafion tubing, and ca. 20 cm long BFS capillary. The FS capillaries were sleeved by the Nafion tubing with id tightly larger than the FS capillary od. The catholyte was 0.2 M NH₄OH, the anolyte was 0.1% FA, and amino acids Lys, His, Gly, Asn, and Glu were employed as CAs. Glycerol (10% v/v) was added to catholyte, anolyte, and a mixed sample and CAs solution to suppress the adsorption of peptides to the capillary wall and to reduce the electric current and Joule heating at 20 kV separation voltage. The hybrid chemical‐ and pressure‐assisted hydrodynamic mobilization was applied. The CE device was an Agilent 7100 analyzer, and the MS detector was an Agilent 6100 series single quadrupole MSD XT using an Agilent Jet Stream ESI‐MS sprayer. The ESI spray chamber had a drying gas temperature of 280°C with a flow rate of 7 L/min and a nebulizer pressure of 15 psig. The capillary and fragmentor voltages were 4000 and 80 V, the spray voltage was 3.85 kV. For peptides, the m/z range was 300–1300, 0.2 step width, 1.99 s/cycle. The sheath liquid (SL) buffer was 0.1% FA and 50% MeOH (v/v) in water.

For the successful CIEF analysis of peptides and proteins, especially for the determination of their pIs, it is important to know the course of the pH gradient inside the capillary. Because it is impossible to directly measure pH inside the capillary, the profile of the pH gradient is evaluated indirectly using the compounds with well‐defined pIs. Their role in CIEF is discussed, and the available pI markers and their synthesis are described in the recent review [147].

No mobilization is necessary if the iCIEF is performed with a whole‐column imaging detection system, that is, with a continuous detection over a wide central section of the capillary with the diode array spectrophotometric or LIF detector. The latest innovation of the iCIEF and its applications for the analysis of proteins and peptides are presented in an earlier review [149].

iCIEF performed in a fluorocarbon coated FS capillary with 200/365 µm id/od, 50 mm long, with integrated electrolyte vessels for anolyte and catholyte, was used for separation of two closely related synthetic peptides, linear 12‐mer (Rp5‐L) and cyclic 15‐mer variant (RP5‐C) of a mimotope (epitope mimicking peptide) of the CD20 antigen that is over‐expressed in B‐cell‐related tumors [102] (see Figure 3). Semipermeable membranes between the electrolyte vessels and the separation capillary ensured electrical contact and made possible the transport of hydroxonium cations and hydroxide anions in the separation capillary in order to form the primary pH gradient. Peptides and pI markers were detected at 280 nm with a deep UV‐LED over the whole capillary length using a metal oxide semiconductor‐based detector array with images taken every 20 s. Anolyte was 80 mM H₃PO₄ with 0.1% (m/v) methylcellulose (MC), and catholyte was 100 mM NaOH with 0.1% (m/v) MC. CAs were AESlytes (HR 3‐10 or SH 6‐9) or Pharmalytes (PL3‐10 or PL 5‐6). Focusing of analytes was done in three steps with a current limit of 20 µA: 0.5 kV for 1 min, 1.0 kV for 1 min, and 3.0 kV for 16 min. The pIs of cyclic Rp5‐C and linear Rp5 were equal to 5.99 and 6.47, respectively. The method was validated, and the LODs and LOQs were in the range 0.24–0.29 µM and 0.79–0.96 µM, respectively.

FIGURE 3.

Imaged CIEF of Rp5‐L and Rp5‐C peptides with pI markers. Sample composition: 0.50% (m/v) PL 3–10; 1.0% (m/v) PL 5–6; 0.35% (m/v) MC; 10.4 µM Rp5‐L; 18.2 µM Rp5‐C;150.7 µM pI 4.65; 104.4 µM pI 5.12 and 44.0 µM pI 7.05. Focusing conditions: +0.5 kV for 1.0 min, +1.0 kV for 1.0 min, and +3.0 kV for 16.0 min. Separation was performed in a whole‐column imaging detection cartridge with fluorocarbon‐coated FS capillary (id 200 µm). The x‐axis was converted to the pI‐scale. Since only two reference markers can be assigned in the software (i.e., 5.12 and 7.05), this causes a slight deviation of the focusing position of pI marker 4.65 from its nominal pI. Source: Reprinted with permission from Bloderer et al. [102].

The iCIEF is suitable for the separation of charge variants of intact proteins and their polypeptide fragments, especially in the development of therapeutic mAbs and other protein drugs [150, 151]. Due to its concentrating and self‐sharpening focusing effects, CIEF is often used as the first concentrating step in two‐ or multidimensional separations of complex mixtures of peptides and proteins [148, 150].

5.4. Affinity Electrophoresis

Affinity electrophoresis was first performed in slab gel format already in 1979 [152]. In the capillaries, it was introduced only in 1992 [153]. ACE takes advantage of the high selectivity of (bio)affinity interactions and a high separation efficiency of CE. ACE includes several modes that can be classified into two groups. The first one is based on the CE determination of the equilibrium concentrations from the peak areas of separated interacting species and/or their complexes. This group includes the Hummel–Dreyer method [154], the vacancy peak method [155], frontal analysis [156, 157, 158], pressure‐assisted frontal analysis [103], continuous frontal analysis [159, 160], and kinetic CE [161]. The second group is based on the CE measurement of changes in migration times and/or effective mobilities of analytes due to their interactions with ligand dispersed in the BGE in the whole capillary or in the capillary segment. In the most often ACE mode, mobility shift ACE (ms‐ACE) [162, 163], the dependence of effective mobility of analyte on the concentration of ligand in the BGEs is measured to estimate the strength of the analyte‐ligand interactions, that is, to determine the binding constant K _b of the analyte‐ligand complex. If the ligand is present only in a particular capillary segment, then this ACE mode is called partial‐filling ACE (PF‐ACE) [164].

Both theoretical and experimental aspects of these methods and their applications for estimation of the binding constants have been outlined in the above reviews [162, 163]. The particular recent applications of ACE to quantitative characterizations of biopeptide interactions are presented in Section 7.3.

Special experimental conditions have to be developed for the coupling of ACE with MS detection. The BGE has to be volatile, and the coating suppressing adsorption of the analytes to the capillary wall has to be stable, preferably covalent, to prevent ion suppression and contamination of the MS detector. These conditions were met by using 300 mM N‐methylmorpholine acetate, pH 7.4, as BGE and a polydopamine‐coated FS capillary for studying the noncovalent interactions of low molecular mass organic ligands and drugs with soft and hard proteins [105].

As shown in the above reviews [162, 163], ACE is widely used for investigations of noncovalent interactions of (bio)molecules, including peptides, under mild free‐solution conditions and for the determination of the binding (association, stability, complexation, formation) constant or dissociation constant of the (bio)molecular complexes. For data processing and statistical evaluation of ACE experiments, the freely available program CEval is recommended [165].

5.5. Electrokinetic Chromatography

EKC was introduced as a combined electrokinetic and chromatographic method by Terabe et al. [166] already in 1985. The movement of mobile phase (MP) (BGE) is driven by EOF, and the separation of both charged and electroneutral analytes is based on their different distribution between MP (BGE) and pseudostationary phase (PSP) composed of micelles in micellar EKC (MEKC) [167], microemulsions in microemulsion EKC (MEEKC) [168], liposomes in liposome EKC (LEKC) [133], cyclodextrins (CDs) in CD‐EKC [169], or other mono‐ or supramolecular structures, such as lipid aggregates and lipid emulsions [133].

The recent developments and applications of capillary EKC, including those dealing with EKC of peptides, are described in general reviews on CE methods [36, 38]. The most often used mode of EKC is the MEKC with micellar PSP constituted by ionogenic detergents, typically anionic SDS or cationic CTAB or zwitterionic CHAPS. MEKC is suitable especially for the separation of noncharged peptides, that is, peptides with blocked or derivatized N‐ and C‐terminus and other ionogenic groups of peptides, and/or for the separation of peptides and proteins with identical or very close charge to size ratios but differing in the hydrophobicity [170]. MEKC and MEEKC have been applied for the characterization of very hydrophobic drugs by the octanol water partition coefficient (logP_o/w) [171].

One of the few MEKC applications for peptides shows monitoring of phosphorylation and degradation of three fluorescamine (FAM) labeled peptide substrates for protein kinase B [172]. For the evaluation of phosphorylation of peptide B‐5/I (6FAM‐GRPRAATFAEG‐NH₂) and Crosstide (peptide counterpart of glycogen synthase kinase‐3 (5FAM‐GRPRTSSFAEG)), the MP (BGE) was composed of 100 mM borate, pH 7.7, and micellar PSP consisted of 100 mM SDS. For all assays of VI−B peptide (6FAM‐GRPRAFTFA‐NH₂), the MP (BGE) was 100 mM borate, pH 11.4, and the micellar PSP was 15 mM SDS. The degradation of Crosstide was monitored in MP (BGE) containing 100 mM HEPES, pH 8.0, and the PSP was 10 mM SDS. Analyses were carried out in a 50 µm id BFS capillary with 31/21 cm total/effective length, at an electric field strength of 400 V/cm.

In another MEKC method, mixed micellar PSP composed of green sodium salts of suberin fatty acids and the cationic tensid CTAB was applied for the separation of tryptic peptide fragments of lysozyme [173]. Optimized separations were performed in MP (BGE) composed of 20 mM KH₂PO₄, pH 2.0, with PSP containing 0.75 mM (0.27 mg/mL) CTAB and 0.036 mg/mL suberin surfactant, at −20 kV separation voltage and 30 µA electric current in a BFS capillary of 50/375 µm id/od and 700/615 mm total/effective length. Samples were injected by pressure of 50 mbar for 5 s, and peptides were detected at 210 nm.

5.6. Electrochromatography

CEC is a hybrid method combining the electrokinetic and chromatographic separation principles. It benefits both from the high selectivity of variable stationary phases (SPs) for peptide separations by (U)HPLC techniques and from the flat (piston‐like) profile and low dispersion of the EOF‐driven flow of MP (BGE). Despite these benefits, CEC in general, as well as in the separation and analysis of peptides, did not meet the optimistic forecasts for this method in the 1990s and at the beginning of this century. CEC did not become a widely used method for peptide analyses.

CEC is implemented in three modes using packed, monolithic, and open tubular (OT) columns. From them, recently, mostly the OT‐CEC was preferred for further developments. The reason for this choice was that the synthetic compounds and materials for SPs can be relatively easily physically or chemically immobilized on the inner FS capillary wall without the back‐pressure problems or bubble formation in the sorbent or in the frits. Various coating materials, including organic and inorganic compounds and materials, hybrid materials, and polymeric materials, have been introduced as physically adsorbed or covalently attached SPs. They include metal organic frameworks (MOFs) [73], molecularly imprinted polymers [174, 175], covalent organic frameworks (COFs) [74, 176], gold and other nanoparticles [76], pH‐responsive block copolymers [177], polymer materials, silica‐based materials, and biomaterials [178].

CEC separations of peptides and proteins can be performed in different separation modes, with reverse‐phase, mixed‐mode, and cationic or anionic ion‐exchange SPs [179], in classical FS capillary columns as well as in microfluidic chips. OT‐CEC column with covalently attached polystyrene sulfonate as SP was prepared by in situ polymerization using 4,4′‐azobis(4‐cyanopentanoyl chloride) as polymerization initiator [179]. It was successfully applied for the separation of different types of compounds, three peptides, five alkaloids, and five sulfonamides.

OT‐CEC in PMMA microcolumns with carboxylic groups (generated by UV/O₃ plasma activation), inducing a negative charge and hydrophilic character on the microcolumn wall, was employed for the separation of four model peptides (C‐natriuretic peptide, bradykinin, Met‐enkephalin, and C‐3‐33 peptide fragment) derivatized with fluorescent ATTO 532 label [180]. Separation and apparent mobility determination were performed in a microchannel with cross‐section dimensions 50 × 100 µm (depth×width) and 50/40 mm total/effective length using 0.5× TBE (Tris‐borate‐EDTA) buffer, pH 8.3, as MP (BGE) at 20–200 V/cm electric field strength. Peptides at 10 µM concentration were detected by the LIF detection system with a 20 mW, 532 nm excitation laser.

Attractive alternatives to the classical CEC columns packed with particulate materials are the monolithic SPs prepared in the form of polymeric columns, silica columns, nanomaterials‐based columns, and hybrid columns [181]. Their advantages are the simplicity of their in situ preparation and the variety of readily available chemistries for the synthesis of monolithic materials. The advantage of CEC as compared with MEKC is that the composition of its mobile phases and the immobilized SPs are more compatible with online coupled ESI‐MS detection.

Despite its limited usage so far, according to some researchers, CEC still has potential for further methodological development and extension of the applications, especially in pharmaceutical and biomedical analysis and separation of chiral compounds and drugs [182, 183, 184].

5.7. Multidimensional Separations

Despite the high separation efficiency of CE methods, their resolution power is usually insufficient for a complete separation of complex mixtures of peptides/proteins in various biomatrices (body fluids, tissue extracts, cell lysates) and in (multi)enzymatic and chemical digests of large proteins. In these cases, a combination of two or more complementary separation principles, generating 2D or multidimensional separation systems, such as IEF and SDS‐PAGE in the classical 2D gel electrophoresis (2D‐GE) [185, 186] or in the differential or comparative 2D‐GE [187] is needed.

The important role of multidimensional separations of biopeptides, peptide fragments of proteins, and proteins in proteomic, peptidomic, and metabolomic studies was highlighted in the recent reviews [17, 39]. Two‐ or multidimensional peptide separations are based on a combination of two or more separation methods with different separation principles, for example, LC and electrophoresis [188, 189, 190, 191, 192]. If these principles are independent, the methods are classified as “orthogonal”, and the peak capacity of this 2D separation is approximately equal to the product of the individual peak capacities of each dimension [193].

There are many combinations of separation methods applicable for 2D separations of peptides and proteins, such as LC‐CE, CE‐CE, CE‐LC, where LC represents RPLC, HILIC, IEX, or SEC, and CE represents CZE, CGE, CIEF, ACE, EKC, and CEC. Various combinations are considered in an annual review on design aspects of the microchip comprehensive spatial three‐dimensional liquid phase separation methodology for proteomic and peptidomic analyses [194]. The authors concluded that implementing a combination of CIEF, SEC, and RPLC within a microdevice with more than 4000 microchannels could make it possible to achieve peak capacity greater than 30 000 within ca. 1 h. After hyphenation with MS/MS detection, these suprapowerful multidimensional separation systems could resolve and identify hundreds to a few thousand peptides/proteins in the above complex biomatrices in large‐scale proteomic and peptidomic studies.

2D format implemented by online coupling of CIEF with ESI‐MS detection [148, 150] can be considered as a current equivalent of the classical 2D‐GE separating according to pI in the first dimension and according to size in the second dimension [185, 186]. CIEF‐ESI‐MS also provides both pI and M _r, but the obtained M _r values are much more precise than those estimated from the second step of the 2D‐GE, SDS‐PAGE.

A powerful 2D‐CZE‐CZE set‐up was composed of two BFS capillaries connected via a customized 8‐port valve and coupled with tandem ESI‐MS detection [195] (Figure 4). It was applied for the separation of charge variants of intact mAbs in the first CZE dimension (in BGE composed of 380 mM ε‐aminocaproic acid, 2 mM triethylenetetramine, 0.05% HPC, pH 5.7) and for the separation of a complex mixture of peptides generated online in capillary reduction of mAb with tris(2‐carboxyethyl)phosphine (TCEP) and digestion by pepsin in the second CZE dimension (in BGE consisting of 200 mM FA). Several common modifications, such as deamidation and oxidation, were detected and localized in four different mAb molecules.

FIGURE 4.

Set‐up of CZE(EACA)‐CZE‐MS/MS with in‐capillary reduction and digestion. First dimension: 55 (45 + 10) cm FS capillary, UV detection (214 nm) at two positions (32.5 and 40.5 cm). BGE: 380 mM EACA, 2 mM TETA, 0.05% HPC, pH 5.7. Valve: 8‐port valve with multiposition actuator (customized C2M‐4358). Second dimension: 93 (35 + 48) cm FS capillary connected to an Orbitrap Fusion Lumos via the nanoCEasy interface. BGE: 0.2 M FA. Zoom at loop 3 shows the arrangement of plugs after the cut procedure (mAb peak cut and transferred in between a large plug of pepsin/TCEP (tris(2‐carboxyethyl)phosphine) solution). Source: Reprinted with permission from Schlecht et al. [195].

2D and 3D MCE system comprising CIEF in the first dimension and CGE in the second or CITP/CGE in the second and third dimensions was employed for the separation of complex mixtures of peptides and proteins in blood plasma [119, 196]. Under the conditions partially specified only for IEF step (4% CAs pI 3‐10, 20 mM H₃PO₄ as anolyte, 20 mM NH₄OH as catholyte, 500 V/cm electric field strength in 23 mm long, 150 µm‐wide microchannel) and using the inverted fluorescence stereomicroscope equipped with CCD camera for visualization of the separation process, a high‐resolution peptide fingerprinting with high peak capacity was obtained within 1 min.

6. Detection

6.1. UV‐Absorption

Thanks to relatively strong absorption of low UV radiation (185–220 nm) by the peptide bond, peptides can be advantageously detected by the spectrophotometric UV‐absorption detection. This detection still belongs to the most often used detections in CE analyses of peptides. Due to the low injected volume of analyzed samples and the short optical path of the on‐column UV detector, the concentration sensitivity of UV detection is not too high. The typical LODs of the UV detectors in 50 µm id capillary are at the micromolar level within a ca. nanoliter detection cell volume with 10‐20 Hz frequency of data acquisition. This sensitivity can be enhanced by the earlier developed special detection cells with increased optical path, such as bubble cell, Z‐shaped cell, or sleeve cell. More details on miniaturized UV–vis spectrophotometric detectors in CE can be found in the reviews on the current state of CE instruments, including their detection systems [197, 198] and in the tutorial dealing with the detection of analytes, including peptides, in capillary and microchip LC and CE methods [199]. Increased sensitivity of peptide detection is achieved at ultralow UV region below 200 nm, for example, at 195 nm in the CZE analysis of antimicrobial dipeptide β‐Ala‐Tyr [200].

6.2. Fluorescence

Classical fluorescence detection (FLD) using a Xenon‐Mercury lamp as a polychromatic light source, and especially the laser‐induced fluorescence (LIF) detection (with the well‐defined wavelength of the excitation light), belong to the most sensitive detection modes in CE analyses of peptides. In commercial CE devices with on‐capillary LIF detection, the concentration LODs are usually in the low nM range. In the special lab‐made LIF detectors, the concentration LODs are at subnanomolar levels and the substance amount LODs are in the zeptomole range [201].

Developments of FLD systems in the last 10 years and various aspects of FLD in CE and CLC methods (optical arrangements, excitation light sources, detection zones, radiation sensors, multiwavelength detection schemes, types of fluorophores, and fluorescent labels) are described in the recent review [202]. The lab‐made CE‐LIF setup was used for the study of interactions between glutathione and other biological thiols and gold nanoparticles [203]. A green light‐emitting laser (emission maximum at 532 nm, power ∼4 mW) was used as an excitation source. The analyses were performed in BGE consisting of 15 mM HEPES, pH 7.5, in a BFS capillary (75/360 µm id/od, and 50/35 cm total/effective lengths) at 18 kV separation voltage, and ambient temperature. Samples were injected by elevation of the injection capillary end to a height of 10 cm for 30 s, and the LOD was 10 nM.

The disadvantage of the LIF detection of peptides is that they must mostly be derivatized with fluorogenic labels. The native fluorescence can be utilized only for the detection of peptides containing aromatic AAs, Trp, Tyr, or Phe. However, for their excitation, deep UV‐laser systems, such as Nd:YAG laser operating at 266 nm, or multiphoton excitation, are necessary. Intrinsic fluorescence in the visible light spectrum was utilized for the detection and identification of amino acids and proteins by an in‐lab‐constructed scanning confocal microscope using a supercontinuum pulsed laser as a green light source of the excitation wavelength of 520 nm [204]. The fluorescence from the sample was filtered with a 575 nm long‐pass filter before it was detected in the red and near‐infrared light region by avalanche photodiodes. Being applicable to both aromatic and aliphatic amino acids, the system can also be used for peptides. In migratio noncovalent fluorophore labeling of proteins by propidium iodide in CGE‐SDS analysis of mAbs provided LOD at 10 nM (2 µg/mL) [205]. It can also be potentially applied for CGE separation and FLD detection of peptides.

More often, the FLD of peptides is based on their precolumn or postcolumn derivatization with a fluorescent label (see Section 3.2). The disadvantage of this approach is that due to the usual presence of multiple derivatization sites in peptides and proteins, several derivatives with different mobilities are generated, resulting in multiple peaks in electropherograms for originally single peptide or protein species.

6.3. Mass Spectrometry

Being both widely universal and selective, and providing high sensitivity and structurally rich information on the analyzed compounds, MS represents the most powerful detection mode for CE, LC, and other separation methods. The recent developments in the area of MS detection in CE methods have been described in several reviews dealing with CE‐MS in general [206, 207] and with its application to peptide, protein, peptidomic, proteomic, and metabolomic analyses, especially [20, 208, 209, 210]. The importance of MS as an analytical and structure elucidation method has significantly increased in recent years due to its crucial role in the above omics technologies [211]. Therefore, the relevance of MS hyphenation with CE separations of peptides and proteins is growing as well.

Online and offline couplings of CE methods with earlier developed ESI‐MS [212] and MALDI‐MS [213] techniques were first introduced almost three decades ago [214]. It has become a breakthrough in the CE analysis and characterization of peptides and proteins. Hyphenation of CE separation with ESI‐MS and MALDI‐MS detection allowed not only a high‐accuracy determination of M_r of CE separated peptides and proteins but also obtaining structural data on AA sequences [215], the sites of posttranslational modifications (PTMs), peptide mapping, and noncovalent interactions of peptides and proteins with other (bio)molecules. ESI and nanoESI are the most popular modes for online CE coupling with MS, especially thanks to their ability to generate multiple charged ions. It ensures that the ion m/z (mass/charge) ratios of even large polypeptides and proteins fit within the limited m/z detection range of most MS detectors.

In the beginning of their development, the CE‐ESI‐MS set‐ups have utilized for ionization of the peptides and other analytes the adapted commercial sheath‐flow interfaces employed for MS coupling with LC methods. However, in these interfaces, the eluting analytes are diluted with the SL at the exit of the capillary. This decreases the sensitivity of the MS detection. Consequently, in the next years, new types of sheathless [216] or low‐flow (nanoflow) [217] interfaces were developed.

A robust hyphenation of CZE/CGE with ESI‐MS detection was achieved by a closed‐circuit coaxial sheath flow reactor interface [128]. The postcolumn reactor placed in front of the MS orifice made it possible to use non‐MS‐friendly components (SDS) in the BGE without significant sample ion suppression and supported stable electrospray. In SDS agarose CGE, the addition of γ‐cyclodextrin to the SL removed SDS from the sample and BGE in the flow reaction space by inclusion complexation and preserved the size‐based separations of peptides and proteins with a high separation efficiency.

For the CE‐MS analyses of ultrasmall volume samples of picoliter‐ and nanoliter‐volumes, an automated spray‐capillary platform was developed for microsampling of proteins and peptides in complex biological matrices (single cells) using a commercially available CE autosampler [104]. Good function of the system was verified by CE‐ESI‐MS analyses of Escherichia coli lysate digest spiked with 10 µM angiotensin II peptide. Separations were performed in PEI‐coated FS capillary (50/360 µm id/od, 100 cm length) using 100 mM FA as BGE and −30 kV as separation voltage. Analytes were detected by an Orbitrap Exploris 240 mass spectrometer with the following parameters: the MS‐inlet capillary temperature of 275°C, ESI voltage of 3.0 kV. MS1 scans were obtained at 120 000 resolution with two microscans from 350 to 1350 m/z range. The top 10 precursor ions with charge 2–6 were selected for fragmentation with 30% of the normalized HCD (higher‐energy collisional dissociation) energy. MS2 scans were collected at 30 000 resolution with two microscans and 200 ms maximum injection time from 350 to 1350 m/z.

A robotic capillary platform was developed for a fully automated CE‐ESI‐MS setup for high‐throughput single cell proteomics with proteomes limited to less than ca. 100 nL [218]. The platform is hermetically closed and actively pressurized to inject ca. 1–250 nL of the sample into the CE separation capillary with RSDs less than about 5%. In the proof‐of‐principle experiments with the trapped ion mobility MS detector (timsTOF PRO), approximately 1800 proteins were identified and quantified from ∼20 ng of the HeLa proteome digest.

A miniaturized nanoESI‐MS interface consisting of a microfluidic nanosprayer and nanospray module was fabricated by silicon technology, suitable for cost‐effective high‐volume mass production [219]. The nanospray module made it possible to position the nanosprayer in front of the MS orifice and its coupling with the separation capillary by liquid junction. Both bottom‐up and top‐down proteomic analyses of trypsin‐digested and intact cytochrome c proved the practical applicability of this interface.

In a new robust polymetallic‐coated sheathless CE‐ESI‐MS interface with high acid and alkali resistance, the electric contact of the interface with ESI was achieved by etching one end of the FS capillary into a tapered tip by hydrofluoric acid, followed by deposition of a thin platinum layer via the physical vapor deposition method [84]. The interface generated stable ESI even at an ultralow flow rate of 12.2 nL/min. Its acid and alkali resistance was confirmed by its good function even after its long‐term (8 days) immersion in solutions of 1 M HCl or 1 M NaOH, respectively. A good function of the interface was verified by t‐ITP‐CZE‐ESI‐MS separation of five standard peptides (angiotensins I and II, bradykinin, kemptide, and neurotensin at 2 µM concentration) and tryptic peptide fragments of BSA using LE composed of 25 mM ammonium acetate adjusted to pH 4 by AcOH and BGE consisting of 1 M AcOH. The following parameters of a Q Exactive HF mass spectrometer were selected: the temperature of the transfer capillary was 320°C, full scan range was 100–1000 m/z with resolution 60 000, the automatic gain control was 1 × 10⁶, the maximum injection time was 100 ms, the ion isolation window was 2.0 m/z and the normalized collision energy of high energy collisional dissociation was 30. The separation voltage was 30 kV, and the spray voltage was set at 2.3 kV.

The second soft ionization mode, MALDI, is usually coupled with CE offline. Its advantage is that it generates preferentially single‐charged molecular ions of even large polypeptides and proteins with M_r up to 300 000. It allows the exact determination of M_r with an accuracy of ±0.1% and identification of peptides and proteins isolated by micropreparative CE methods. Recently, it has been rarely applied to peptides and proteins; one of the few examples is MALDI‐MS peptide profiling of urine samples [220].

6.4. Other Detection Modes

Among the other detection modes, the universal capacitively coupled contactless conductivity detection (C⁴D) is relatively frequently used in CE and MCE methods [221, 222, 223]. However, mostly for small ions, such as organic acids [224], AAs [225, 226], and relatively rarely for peptides [227]. One of the few examples is the application of C⁴D for the MCE determination of dipeptide L‐carnosine (β‐Ala‐L‐His) in health supplementary formulations [228]. The highest sensitivity with 2.5 µM LOD and 5 µM LOQ with linear detector response in the range 5–50 µM was obtained with an excitation sine wave of 400 kHz with an amplitude of 40%. In addition, C⁴D improved universality and partially also sensitivity in the study of drug protein interactions by CE‐FA [229]. Worth mentioning is C⁴D integrated with UV‐absorption and fluorescence into the three‐in‐one 3D detection system fabricated by 3D printing [230].

Electrochemical, NMR, and some other detection modes are less suitable and hence much less employed for the detection of peptides separated by CE methods than the above UV‐absorption, fluorescence, MS, and C⁴D detection schemes. For that reason, only a few recent reviews and papers are cited here, which deal with general developments and applications of these detectors in separation methods. They include electrochemical detection alone [199, 231] or in combination with MS [232] and direct or indirect chemiluminescence [199]. NMR spectroscopy is mostly offline, combined with CE, especially for structural characterization of analyzed compounds [233].

7. Applications

7.1. Analysis

7.1.1. Quality Control and Determination of Purity

Quality control and purity determination of peptides are required in all fields where peptides are isolated, synthesized, investigated, and applied for various purposes. These fields include biological, physiological, biochemical, biomedical, and clinical research, for example, in the study and modeling of various biomolecular interactions, such as hormone–receptor, drug–receptor, enzyme–substrate, enzyme–inhibitor, ion–ionophore, and antigen–antibody, and in the identification of antigenic determinants (epitopes) of proteins. In addition, peptide and peptidomimetic drugs, and food and feed additives belong to the relevant compounds also in biotechnology, pharmaceutical, food, and feed industry [106, 191, 234, 235]. Hence, there is an urgent need for qualitative and quantitative analysis of peptides by CE and other analytical methods. In most of the above applications of isolated or (bio)synthesized peptides, CE methods can be used as sensitive control methods providing fast and accurate qualitative and quantitative information about the purity of peptide preparations [227]. Further, some recent applications of CE methods to peptide analyses will be presented, in addition to those shown already in the previous sections.

Purity of short amphiphilic peptides prepared by batch and flow synthesis for self‐assembled nanostructures as potential theranostic agents was checked by Agilent 7100 CE system with double, spectrophotometric DAD and ESI‐MS detection system [4]. Analyses of peptides (introduced hydrodynamically by pressure of 50 mbar for 3 s at 0.5 mg/mL concentrations) were carried out in an FS capillary (75/375 µm id/od, 100/26.5 cm total/effective length) using ammonium formate BGE at an ionic strength of 20 mM, pH 3.7, at 20 kV separation voltage and 25°C. The UV DAD was set at 200 nm, and the parameters of the Agilent 1100 single quadrupole MS detector in the positive ionization mode were as follows: nebulizing nitrogen pressure, 12 psi, drying gas flow rate of 8 L/min at 250°C, spray and skimmer voltages of 3 kV and 70 V, respectively. The SL (1% FA in 80/20 MeOH/H₂O mixed solvent) was delivered to the interface at a flow rate of 4 µL/min by a 1100 isocratic pump equipped with a 1:100 splitter. Signal acquisition was done in the scan mode of 100–1400 m/z, and the MS spectrum was then extracted under each peak for identification. The peptides could be characterized by exact sequences and purity degree at a 0.1 mg/mL level without any sample pretreatment.

CE with spectrophotometric DAD was employed for the pharmaceutical quality control of two important peptide drugs, triptorelin and lanreotide [236]. Triptorelin, a synthetic decapeptide acting as a potent agonist of the human gonadotropin‐releasing hormone, is used in oncology and in the treatment of endometriosis, uterine fibroids, premature puberty, and hypersexuality. Lanreotide is a synthetic cyclic octapeptide, a long‐acting analog of somatostatin. It suppresses various hormones and neurotransmitters and is effective in controlling symptoms and growth of neuroendocrine tumors and treating conditions like acromegaly. Simultaneous CE‐DAD analyses of both drugs were performed in an Agilent 7100 analyzer, a BFS capillary (75 µm id × 36 cm length) using 250 mM FA as the BGE. The detection wavelength for triptorelin was 200 nm, and for lanreotide, 220 nm. At a 30 kV separation voltage, the analysis time was shorter than 5 min. The LOD and LOQ were found to be 0.5 µg/mL and 2 µg/mL, respectively. After validation, the method was applied for the determination of triptorelin in a real pharmaceutical matrix of the commercially available drug, 20‐fold diluted Diphereline. The throughput of analysis of these peptides was increased three times by multiple sample injection [97] (see Figure 2).

The electrophoretic homogeneity of synthetic peptide containing 21 AAs proposed as lactoferrin binding ligand by in silico modelling was confirmed by CE analysis in BFS capillary (75 µm id, 60/50 cm total/effective length) thermostated at 30°C using 10 mM Tris/HCl buffer, pH 7.2, as BGE, 20 kV as separation voltage, and 200 nm as detection wavelength [237]. The CE analysis of a mixture of this peptide with lactoferrin proved the formation of a peptide–lactoferrin complex.

Peptide drugs and peptides used in biological assays have to be characterized also by the content of their low‐molecular‐mass ionic admixtures and counterions [235]. Basic peptides frequently contain anionic counterions, for example, fluorides, chlorides, acetates, and trifluoroacetates (TFA) originating from the synthesis and/or purification of these peptides. Peptides are often isolated as TFA salts by preparative HPLC, or they may contain fluorides after their hydrogen fluoride cleavage in solid‐state peptide synthesis. However, the toxic TFA and fluoride anions have to be completely exchanged for another counterion from the peptide preparations before their biological applications to avoid potential toxicity effects. Applications of CE methods in this field are presented in the previous reviews [140, 141].

7.1.2. Determination in Biomatrices

CE methods coupled with highly sensitive MS and LIF detectors and utilizing the online sample stacking effects possess great potential also for the determination of peptides at low concentration levels in complex biomatrices, such as biological fluids, cell lysates, tissue extracts, foods, and feeds. This fact and miniature injected sample volume (nL to pL) make CE methods powerful and highly appreciated tools in biomedical research and clinical analysis [36, 37, 220, 238], where often only minute volumes from biopsy or dialysis samples are available, and in the single cell analysis [64, 239]. Several examples of the determination of peptides and amino acids in various complex mixtures, such as blood plasma, urine, bone tissues, pharmaceutical formulations, and food extracts, are presented in the recent review [227].

Detection and identification of particular peptides in complex biomatrices is of great importance also in peptidomic and proteomic analyses of biological fluids (blood, serum, plasma, cerebrospinal fluid, urine, and saliva) [220, 240, 241]. It can result in the discovery of new specific peptide biomarkers applicable in diagnosis, therapy, and prognosis of some diseases, for example, various kidney diseases [242], bloodstream infections [238], hypertension [240], and other cardiovascular diseases [243, 244].

3D printed, automated, pressure‐driven injection microfluidic MCE device with 46 µm wide, 50 µm deep, 9.8 mm long separation channel connected to two pneumatic valves (formed by membrane thickness of 5 or 10 µm) and reservoirs was developed using Gly and Phe as model compounds and subsequently applied for determination of preterm birth (PTB)‐related peptide and protein biomarkers, peptide 1, peptide 2, corticotropin releasing factor, ferritin and lactoferrin, respectively [119].

Several CE methods have been developed for the determination of reduced form of glutathione (GSH), natural tripeptide γ‐Glu‐Cys‐Gly, as well as of its dimerized oxidized form (GSSG). GSH plays an important role in several physiological processes in humans and animals. It maintains homeostasis and regulates the redox environment in the cells. Acting as an antioxidant, it prevents and reduces oxidative damage to proteins in living cells and participates in the reduction of disulfides and other molecules. The determination of GSH in body fluids is important for early diagnosis and monitoring of various diseases. CE analyses of GSH include, for example, the determination of GSH, cysteine, and homocysteine in human blood plasma and saliva [75, 203], the determination of reduced and oxidized forms of GSH in human blood samples [88], and quantification of naphthalenedicarboxaldehyde (NDA)‐labelled GSH in HepG2 cells with LOD and LOQ of 6.0 and 20.0 nmol/L, respectively, using Tris‐borate buffer, pH 9.2, as BGE [245].

Both reduced and oxidized forms of GSH and some other aminothiols in human blood plasma were determined by the CE‐UV method using BGE composed of 115 mM sodium phosphate, pH 2.3, and 7.5% (v/v) PEG 600 in BFS capillary (50 µm id, total/effective length 42/35 cm) at 25 kV separation voltage [88]. The in‐capillary concentration of the analytes was achieved with a pH gradient created via the preinjection of triethanolamine and postinjection of H₃PO₄. The LOQs varied from 1.3 µM to 5.4 µM. The accuracy was 95%–99%, and the repeatability and reproducibility were 1.8–3.8% and 1.9–5.0%, respectively. The reduced thiols were derivatized with N‐ethylmaleimide. Mean concentrations of reduced thiols in the plasma of 41 healthy volunteers were in the low to medium µM range.

In addition to the above peptide determinations in biological fluids, CE and CEC have also been applied to the analysis of peptides and proteins in complex mixtures of food and feed products [246, 247]. CE and LC analyses of biologically active milk peptides and proteins are summarized in [248], and the application of CE methods in the analysis of coffee is described in [249].

A new integrated SPE‐CZE‐ESI‐MS/MS method was able to determine zeptomole amounts of phosphopeptides in the lysate of mouse brain cells [85]. First, the phosphopeptides from the cell lysate were enriched by Ti‐IMAC (titanium immobilized metal affinity chromatography) beads and concentrated by loading on the C18 SPE bed. Peptides were eluted by the BGE (1% FA in 50% v/v ACN/H₂O) and hydrodynamically introduced to the 22 cm long separation channel of 70 µm width and 10 µm depth. After t‐ITP focusing, CZE separation was run at 500 V/cm electric field intensity between 15 kV at the anode and 2.4 kV at the cathode. Separated peptides were sprayed into an Orbitrap Astral mass spectrometer at an ESI voltage of 3.5 kV. Precursor ions (380‐980 m/z) were measured in the Orbitrap analyzer at a resolution of 120 000, requiring either 1 million charges or a 50 ms maximum injection time for each scan. The above precursors were selected for MS2 using data‐independent acquisition (DIA) at 4 Th quadrupole isolation windows (150 total windows across the mass range). Normalized collision energy was set at 27%. Fragment ions from m/z 100–1000 were measured in the Astral analyzer, requiring either 50 000 charges or a 3.5 ms maximum injection time.

Secretion of peptide hormone glucagon, activating hepatic glycogenolysis and gluconeogenesis from pancreatic α‐cells, was monitored by microfluidic (MF) CE competitive immunoassay (IA) [250]. The CE‐IA analyses were performed in a borosilicate MF device in a 15 mm long microchannel with 5 µm width and 10 µm depth, using 20 mM borate buffer, pH 9.3, as BGE, and LIF detection with a 50 mW solid state laser as the source of the 488 nm excitation light. The emission light at 550 nm was detected by a photomultiplier. Calibration curves were obtained using 200 nM Ab, 200 nM glucagon labeled by FITC, and various concentrations of unlabeled glucagon. The sensitivity of this CE‐IA with LOD of 30 nM was up to 300 times higher than that of the fluorescence anisotropy IA.

A green CE‐MS method was developed for the determination of colistin, a decapeptide antibiotic drug, in human blood plasma [89]. The analyses were performed in an Agilent 7100 CE system coupled with an Agilent 6410 Series Triple quadrupole tandem MS system. The BFS separation capillary of 50 µm id was 99 cm long. The BGE was 50 mM FA, pH 2.54, and the separation voltage was 25 kV, generating an electric current of 5‐8 µA. The SL, composed of 0.1% FA in 50% v/v MeOH/H₂O mixed solvent, was delivered by an Agilent 1260Infinity isocratic pump at a flow rate of 8 µL/min. The MS was operated in positive ion, multiple reaction monitoring (MRM) mode using characteristic precursor ion—product ion mass transition for each substance. The MS parameters were as follows: 150 ms dwell time, 5000 V spray voltage, 5 psi nebulizing gas (N₂) pressure, and 5 L/min drying gas (N₂) flow rate at 300°C. After validation, the method was employed for monitoring of colistin in clinical plasma samples with migration times less than 13 min, LOD of 18.6 ng/mL, LOQ of 37.2 ng/mL, RSD of migration times of 0.02%, and RSD of peak area of 5.48%.

Determination of insulin‐like growth factor (IGF‐1), a 70‐AAs single‐chain polypeptide used in diagnostics as a biomarker of growth hormone disorders and as a drug for growth failure in children and adolescents, in pharmaceutical preparations was carried out by CE online coupled with the above triple quadrupole ESI‐MS detector [98]. Analyses were performed in a BFS capillary of 85 cm total length and 50/365 µm id/od. The optimized BGE was composed of 500 mM FA and 5% v/v ACN, pH 1.96. CE was hyphenated with MS by a SL coaxial interface, and the SL was delivered to ESI at a ratio of 1:100 by isocratic pump. The flow rate of SL was 6 µL/min, and the SL was 0.1% FA in 50% v/v MeOH/H₂O. The other optimized MS parameters were as follows: 150 ms dwell time, 4000 V spray voltage, 4 psi (∼27.58 kPa) nebulizing gas (N2) pressure, 300°C drying gas (N2) at 8 L/min flow rate, and the protruding length of the capillary from the sprayer was 0.3 mm. With the above‐optimized parameters, the time of analysis was ca. 12 min, LOD and LOQ were 250 and 500 ng/mL, respectively, and RSDs of migration times and peak areas were 1.11% and 11.6%, respectively. The validated method was applied for the determination of IGF‐1 in injectable solutions in nutritional tablets and liquid colostrum.

CZE with UV detection at 200 nm was applied for quantification of heterogeneous polypeptide drug, glatiramer acetate (M _r 5–9 kDa), and its AA constituents (Lys, Ala, Glu, and Tyr), in pharmaceutical preparations used as the first‐line drug for the treatment of multiple sclerosis [251]. The analyses were conducted in BFS capillaries (48.5/40 cm total/effective length, 50 µm ID) at 25 kV separation voltage with currents less than 100 µA, in BGE composed of 120 mM H₃PO₄, adjusted with Tris to pH 1.9, and supplemented with 20 mM triethylamine to a final pH 2.1. With Trp as the internal standard, the LOD and LOQ were equal to 39.2 and 130.7 µg/mL, respectively, and the time of analysis was ca. 25 min.

In the area of food analysis, MCE with C⁴D was applied for the determination of the dipeptide L‐carnosine (β‐Ala‐L‐His), in health supplementary formulations without any preconcentration and derivatization steps [228]. The analyses were performed in a double‐T‐shaped PMMA chip with a separation channel of a total 8.7 cm length and 50 × 50 µm cross‐section. From the several BGEs tested (500 mM AcOH, pH 2.52; 0.75 mM iminodiacetic acid, pH 2.5; 2.4 nM tartaric acid, pH 5.8; and 1.4 mM citric acid, pH 3.0), the first one, 500 mM AcOH, provided the best results. Fast analysis with LOD and LOQ equal to 0.12 and 0.31 µM, respectively, was achieved within less than 1.5 min. The determined amounts of L‐carnosine in real samples were verified by independent CE and HPLC analyses.

Simultaneous determination of 14 toxic cyclic cyanopeptides (M _r 825.0–1068.3 kDa) produced by cyanobacteria in thermal spring water was achieved by CZE hyphenated with tandem MS detection within ca. 7.5 min [252]. Analyses were run on an Agilent 7100 system online coupled with a triple quadrupole 6495C Agilent MS detector in BFS (80 cm total length, 50/375 µm id/od) in BGE composed of 50 mM ammonium acetate, pH 10.2. The SL, a mixture of 30:69.7:0.3 (v/v/v) MeOH/H₂O/FA, was delivered with a flow rate of 10 µL/min, with a 1:100 splitter. The MS was operated in ESI+ mode under multiple reaction monitoring conditions. Capillary and nozzle voltage were set at 200 V. The other optimized ESI parameters were as follows: dry gas flow rate of 11 L/min, dry gas temperature of 150°C, nebulizing gas pressure of 12 psi, sheath gas temperature of 195°C, and sheath gas flow rate of 5 L/min. MS/MS experiments were performed by fragmenting the protonated or deprotonated molecules that were selected as the precursor ions. Collision energies were 5–75 V, and the product ions were analyzed in the range of 58.0–925.3 m/z. Before CE‐MS/MS analyses, the peptides were isolated by the salting‐out assisted liquid–liquid extraction with ACN as an extraction solvent and MgSO₄ as an auxiliary salting‐out agent. Field‐amplified sample stacking applied for sample preconcentration with injection solvent consisted of mixed hydro‐organic solvent H₂O/MeOH (95.5/4.5 v/v), provided up to five times increased sensitivity with LODs of 0.01–0.04 µg/L and LOQs of 0.02–0.12 µg/L.

7.1.3. Monitoring of Chemical and Enzymatic Reactions

In addition to analyzing and characterizing “static” peptide preparations, CE methods are often employed for monitoring and evaluating chemical and enzymatic reactions and modifications of peptides and proteins, such as hydrolysis, oxidation, reduction, deamidation, isomerization, and racemization. Besides reactions carried out in classical offline format, the reactions performed directly inside the separation capillary on the nanoliter scale can be monitored online by CE. The reactions can be performed in heterogeneous or homogeneous modes. In the former mode, one reactant or enzyme is immobilized on the solid sorbent fixed in a short section of the capillary or on a short section of the inner capillary wall, and the other reactant or substrate is injected as a short sample zone. In the latter mode, the reactants or enzyme and substrate are mixed by electromigration or by diffusion. The electromigration mixing is ensured by different electrophoretic velocities of reactants or enzyme and substrate introduced sequentially inside the capillary as narrow zones. This mode is called electrophoretically mediated microanalysis (EMMA) [253]. The second way of mixing is based on the transverse diffusion of the parabolic laminar flow profiles of hydrodynamically introduced short zones [254, 255]. Recent applications of CE in the monitoring of enzymatic reactions are described in [256].

Peptide degradation of alpha‐amidated peptides with prolonged half‐time was quantified by low pressure (689.5 Pa (0.1 psi)) assisted CE performed in PEO coated FS capillary (50 µm id, 30/20 cm total/effective length) using 25 mM ionic strength citrate buffer adjusted by NaOH to pH 3 as BGE, 20 kV as separation voltage, and 214 nm as a UV detection wavelength [100]. The amidated and the hydrolyzed carboxylated forms of 13 oligopeptides could be easily separated within ca. 5 min thanks to one elementary charge difference in their effective charges. The obtained degrees of degradation were in good agreement with those determined by HPLC‐MS.

Deamidation of three recombinant human insulins (glargine, glulisine and humuline) under different storage conditions (−80°C, −20°C, +5°C, +25°C, and in acid medium of 0.1 M HCl) was explored by CE using BGE consisting of 50 mM ammonium acetate, 20% v/v IPA, pH 9.0, in BFS capillary (50/370 µm id/od, 85/76.5 cm total/effective length) at 25 kV separation voltage and 200 nm detection wavelength [257]. At −80°C and −20°C, the deamidation was minimal, but at +5°C, onefold deamidation was found for all insulins. In acidified samples incubated for 1 month at 25°C, up to threefold deamidations were found in Humulin.

CE methods are often applied for the study of PTMs of peptides and proteins, such as phosphorylation, glycosylation, acetylation, methylation, glycation, and others [106, 258, 259]. In‐capillary digestion of charge variants of native and thermally stressed mAb by pepsin was analyzed by a 2D‐CE‐ESI‐MS setup described in the above Section 5.7. The peptide fragments were separated in an FS capillary (50 µm id, 93 cm length) physically coated with PEO using 200 mM FA as BGE, and 50% IPA, 0.2% FA as the SL with delivery flow rates of 3–8 µL/min [195] see (Figure 5). The common PTMs of mAb, oxidation, deamidation, and others were detected and localized in the polypeptide fragments.

FIGURE 5.

Base peak electropherogram (BPE) of 1D in‐capillary reduction and digestion of trastuzumab mAb. Experimental conditions: 1 mg/mL trastuzumab, 3.8 mg/mL pepsin, 40 mM tris(2‐carboxyethyl)phosphine (TCEP), 70 cm PEO‐coated FS capillary, BGE 200 mM FA, 10 min digestion time, 30 kV, 350−2000 m/z. Selected extracted ion electropherograms (EIEs): LFPPKPKDTLM (green trace, 429.578 m/z), ISRTPEVTC (red trace, 503.255 m/z), DSDGSF (blue trace, 627.226 m/z). Source: Reprinted with permission from Schlecht et al. [195].

An immobilized‐enzyme microreactor (IMER) containing pepsin immobilized on the nylon membrane between the CE separation capillary and the MS detector, that is, in a low‐flow SL ESI interface, enabled protein digestion immediately before spray into the mass spectrometer [260]. Thus, the proteolysis carried out after CE separation did not allow the separation of digested peptides but made the simultaneous entering of all generated peptides into the MS detector possible. It may help in the identification of analyzed proteins utilizing the knowledge of their migration times and effective mobilities. Sequence coverage was greater than 75% for myoglobin and carbonic anhydrase II, but much lower for proteins with disulfide bridges.

CE methods are widely used in the monitoring of enzymatic digestions or conversions of peptides with the aim to study some details of these processes, such as drug metabolism, substrate or inhibition specificity, and/or activity of enzymes and kinetics of their cleavage of peptides [256]. These studies include the determination of Michaelis‐Menten kinetic data or the identification and characterization of inhibitors or substrates.

The CE‐based enzyme assays are conducted in two modes: (1) as precapillary assays with offline incubation of enzyme and substrate (and inhibitor) followed by CE analysis of substrate or product(s) [261], and (2) in‐capillary assays, where the enzymatic reaction and analyte separation are performed in the same capillary [262]. The latter mode is performed either as EMMA [253] with mixing of enzyme, substrate, and inhibitor due to their different electromigration velocities or via their transverse diffusion of laminar flow profiles of their short in‐series introduced closely neighboring zones [254, 255].

7.1.4. Monitoring of Molecular Interactions and Physicochemical Changes

In addition to the above chemical reactions and enzymatic conversions of peptides, CE methods are often applied for studying noncovalent molecular interactions of peptides with various ligands or receptors [263] and for monitoring the formation of covalent peptide adducts with other compounds, such as peptide–DNA fragments conjugates [264] and peptide conjugates with antibodies or other carrier proteins [3]. Applications of ACE methods for the investigation of the interactions of drugs, including peptide drugs, with serum proteins can be found in [163, 265].

Interactions of insulin (INS) with taurocholic acid (TCA) for potential oral INS application in the therapy of diabetes were studied by pressure‐assisted CE‐FA [103]. The analyses were performed in an HPC‐coated FS capillary (50/360 µm id/od, 50/40 cm total/effective length). The mixtures of INS and TCA at different concentration ratios were injected in the capillary by pressure 6.90 kPa (1 psi) for 90 s at the cathodic capillary end. They were analyzed at +15 kV separation voltage, 210 nm detection wavelength, and 25°C, and at low pressure of 4.83 kPa (0.7 psi) to speed up the analysis. The plateau height of TCA (ligand) corresponded to the unbound TCA in the equilibrium mixture of INS and TCA. The interaction was monitored in the BGE composed of 15 mM NaH₂PO₄/H₃PO₄, pH 3.5, that is, 2.1 pH units above TCA pK _a (1.4), where it was totally dissociated and negatively charged. On the other hand, INS with its pI 5.4 was positively charged at this pH. When the TCA concentration was increasing in the TCA and INS mixture, the number of TCA molecules bound to the INS (receptor) was increasing as well. Thanks to mobility differences between free TCA and INS‐TCA complex, the free TCA zone could be separated from the complex zone, forming a plateau. Under the optimized conditions, the calibration curve of TCA was generated in the range 0.4–1.0 mM. The mixtures of 100 µM INS and increasing concentration of TCA were applied for the determination of the binding constant. Nonlinear regression analysis of the binding isotherm (dependence of average binding ratio on the concentration of the unbound TCA shown in Figure 6) provided the binding constant K _b equal to (1.3 ± 0.1) ×10³ L/mol and the binding stoichiometry, n = 5.2 ± 0.1, respectively. These results indicate that TCA binding to INS can increase INS absorption and enlarge the bioavailability of INS in the oral therapy of diabetes.

FIGURE 6.

(A) The standard calibration curve of taurocholic acid (TCA). (B) The binding isotherm of insulin with TCA via pressure‐assisted CE‐frontal analysis (PACE‐FA). Source: Reprinted with permission from Sun et al. [103].

A special microbead‐assisted CE (MACE) has been developed for the separation of free peptide–oligonucleotide (single‐strand DNA) conjugates (POCs) [264] and their complexes with target proteins immobilized on the microbeads. In this method, a mixture of microbeads modified with the target protein and a library of POCs is introduced inside the capillary. The migration time of the beads coupled with the target protein was significantly shorter than that of free POCs migrating with higher velocities (mobilities) in the counter EOF mode of mobility shift ACE using 100 mM borate buffer, pH 8.5, as BGE in the BFS capillary (75/365 µm id/od, 80.5/72 cm total/effective length). MACE applies more selection pressure than washing in the so‐called SELEX (systematic evolution of ligands with exponential enrichment) procedure since the aptamers are pulled off from the target protein on the beads by the force of the electric field. The concentration of POCs was 0.9 µM, bead concentration was 1 mg/mL (7–9 × 10⁸ beads/mL) with mAb concentration of 4–5 × 10⁴ molecules per bead in the sample buffer composed of 20 mM Tris‐HCl, pH 7.4, 10 mM NaCl, 1 mM MgCl₂, and the sample volume was 65 nL. In this study, the 75‐meric DNA aptamers strongly binding the mAbs (coupled to the beads) against two synthetic oligo(octa‐ and nona)peptides were successfully selected.

Interactions of in‐silico designed peptide ligand LETI‐11 with affinity to lactoferrin (LF), one of the important target proteins for diagnosing acute inflammatory processes and infections, have been studied via CE determination of concentration of free and complexed peptide in the peptide‐protein mixture at various peptide‐protein concentration ratios [237]. The analyses were performed in a BFS capillary (75 µm id, 60/50 cm total/effective length) using 10 mM Tris buffer, pH 7.2, as the BGE, 20 kV as the separation voltage at 30°C, and 200 nm as the detection wavelength. The concentration of LETI‐11 was 3.6 µM, and it was mixed with LF in the molar ratios 1:1, 1.5:1, 2:1, and 2.5:1. From the electropherograms, the interaction with the stoichiometry 1:1 was confirmed. Interactions of peptides with chiral selectors are described in Section 7.1.7.

In addition to monitoring molecular interactions, CE methods are also employed for monitoring and evaluating physicochemical processes and physical changes of peptides and proteins—aggregation, fragmentation, denaturation, and conformation changes during unfolding and folding processes [266, 267]

7.1.5. Amino Acid and Sequence Analysis

CE methods can also be applied for the characterization and identification of peptides and proteins based on their AA composition and AA sequence. Acid–base and hydrophobicity–hydrophilicity properties of peptides and proteins can be estimated from their AA composition [268], and the structural features of peptides and proteins can be predicted from AA sequences of CE separated peptides by tandem MS analyses [215] or by other AA sequencing methods [269]. In fact, peptide and protein identifications in both bottom‐up and top‐down proteomic and peptidomic analyses are based on the determination of their AA sequences [14, 19, 209].

Recent determination of nonderivatized AAs and peptides by CE with UV‐absorption or C⁴D is described in [227], and determination of branched‐chain AAs by LC and CE methods is reviewed in [270]. Special attention is given to the analysis of the less prevalent d‐forms of AAs in complex biological matrices, such as blood, urine, and saliva [271]. They can serve as biomarkers of various human diseases, for example, some neurological diseases, cancers, and kidney disorders [272]. Chiral LC and CE separations play a major role in their determination, see Section 7.1.7.

7.1.6. Peptide Mapping of Proteins and Biomatrices

High separation efficiency and resolving power make the CE methods powerful tools for peptide mapping (also called peptide fingerprinting) of proteins and polypeptides, that is, separation of peptide fragments generated by specific enzymatic and/or chemical cleavages of proteins and polypeptides. The CE or LC records of such separation represent a characteristic map (fingerprint) of the particular protein or polypeptide. Peptide mapping is one of the basic techniques for protein/polypeptide identification, sequence determination of internal parts of proteins and polypeptides, monitoring of their PTMs and microheterogeneity, as well as for elucidation of protein and polypeptide structure [60]. In fact, peptide mapping is a basis for bottom‐up proteomics [273]. Various aspects of peptide mapping of proteins by CE methods are presented and discussed in detail in the comprehensive review [274].

Due to the high complexity of peptide maps, namely of large proteins, multiprotein samples, and cell lysates, 1D‐ or 2D‐LC and CE methods hyphenated with MS detection are necessary for a maximum resolution of peptides present in these complex mixtures, see a recent review [190] and some other papers [188, 189, 275, 276]. CE‐ESI‐MS/MS peptide mapping of proteins is now often used also for characterization and analysis of therapeutic mAbs [150, 277], see above Figure 5. Peptide mapping requires the digestion of proteins in a precise way so that the specific peptides are obtained for protein identification and quantification. The classical and most often used protease for peptide mapping is trypsin, which cleaves the protein specifically behind Lys and Arg residues. Complications of protein peptide maps by autolytic peptide fragments of trypsin can be avoided by using immobilized trypsin or by chemically modified trypsin resistant to autolysis [278]. Quenching tryptic digestion is not necessary with peptides created from the protein filter traps since trypsin activity is negligible in the eluate from these preparations [279]. Suitable conditions for tryptic cleavage of proteins and complex proteomes of bulk samples or single cells can be found in [280].

Peptide mapping of recombinant human erythropoietin (rhEPO), by CE‐MS and CLC‐MS methods, was applied for a sensitive and unequivocal identification of glycopeptides of this important biopharmaceutical protein with a high glycopeptide microheterogeneity [281]. Before enzymatic digestion, the rhEPO was first reduced by DTT and alkylated by iodoacetamide. An aliquot of 50 µg of the rhEPO was dissolved in 50 µL of the digestion buffer, 50 mM ammonium bicarbonate, pH 7.9, and 1.25 µL of 0.5 M DTT in the digestion buffer was added. The mixture was incubated in a thermoshaker at 56°C for 30 min. Then, alkylation was performed by adding 3.5 µL of 0.73 M iodoacetamide to the digestion buffer. Low‐molecular‐mass reagents were removed by centrifugal filtration. Then, the modified protein was dissolved in the digestion buffer with an enzyme‐to‐protein ratio of 1:40 (m/m) and incubated at 37°C for 18 h. Digestion was stopped by heating at 100°C for 10 min. The CE‐MS separation of tryptic peptides of reduced and alkylated rhEPO was conducted in an Agilent 7100 CE system coupled to a 6546 LC/QTOF MS detector with a G1948B ESI source and a triple‐tube coaxial sheath‐flow sprayer. The SL containing 0.05% (v/v) FA in 50:50 (v/v) H₂O:IPA mixed solvent was delivered at a flow rate of 3.3 µL/min infusion pump. CE separations were performed in a BFS capillary (50/375 µm id/od. 70 cm total length) using 50 mM AcOH and 50 mM FA, pH 2.3, as BGE, 18 kV as separation voltage at 25°C, with the autosampler kept at 10°C. The digests were reconstituted in water at a concentration of 1 mg/mL, and they were injected into the capillary by 50 mbar pressure for 15 s, that is, 0.36 pmol of rhEPO in 11 nL volume.

In a methodological CE study of peptide mapping of a single protein (HSA) or a mixture of four proteins (HSA, myoglobin, lysozyme, and β‐casein) or yeast cell lysate, it was shown that desalting of the hydrolysate is not always necessary [91]. The number of identified proteins and protein groups was even increased after the addition of 18 mM NH₄HCO₂ to the original digest, as compared with the original salted digest, but even better results were obtained when 50% ACN was added to the digest.

Another type of peptide and protein mapping is the peptide and protein maps (fingerprints or profiling) of complex biomatrices, various biological fluids (blood serum and plasma, urine, saliva, cerebrospinal fluid), or tissue extracts. They are obtained as records of CE, LC, or MS separation and detection of peptides and proteins present in these complex biomatrices. They can provide a specific pattern (fingerprint or profiling) for the various healthy or pathological states of human, animal, or plant organisms, whose samples were analyzed. For example, the peptide profiling of urine is an important indicator of a health or disease of the kidney [220]. The comparison of MS, LC‐MS/MS, and CE‐MS/MS methods for profiling of clinically relevant peptides has shown that CE‐MS/MS was able to identify a larger number of peptides and proteins than the other two methods, despite the smaller sample amount and shorter analysis time.

CE‐MS/MS and LC‐MS/MS profiling of natural urinary peptides was studied as potential noninvasive biomarkers for predicting the response to blood pressure medication [240]. Altogether 227 peptides showing a significant difference between patients treated with agents blocking the renin–angiotensin system and uncontrolled/untreated patients were identified.

CE‐MS/MS and LC‐MS/MS urinary peptidomic analyses of 24 preterm and 27 term‐born controls were performed with the aim of determining gestational age‐related peptidome in preterm infants to find potential new prognostic biomarkers [282]. Based on statistical analysis of the obtained peptide sequences, in preterm infants, one of the peptides (solute carrier family 38 member 10) was found to be the most abundant compared with the term infants. This peptide could potentially become a relevant biomarker related to the gestational age.

In a comprehensive study, CE‐MS/MS analyses of 58 urinary samples from patients with autosomal recessive polycystic kidney disease (ARPKD), 662 urine samples from pediatric patients with chronic kidney disease (CKD) with various other CKD etiologies, and 45 samples from healthy children were carried out [283]. After statistical analyses of the obtained peptide sequences, a 77‐peptide signature was found for ARPKD. It allows for distinguishing ARPKD from other causes of CKD with high precision and to identify noninvasively pediatric patients with ARPKD in the children population of other CKDs.

The kidney diseases were also studied by means of CE‐MS/MS analysis of blood plasma peptidome [284]. The analyses were performed in a Beckman Coulter Proteome lab PA800 CE system hyphenated with a Bruker Daltonic micrOTOF‐Q III MS detector using 250 mM FA, 20% (v/v) ACN as BGE in a BFS capillary (50 µm id, 95 cm length). The Agilent electroionization sprayer was grounded, and the ESI potential was −4 to −4.5 kV. During the analyses of 291 plasma samples from 136 patients with end‐stage kidney failure and 20 patients with chronic kidney disease, 3920 unique plasma peptides were identified and quantified, and 661 peptides were sequenced. From them, 169 peptides were identified with different plasma abundance in pre‐ and postdialysis samples. These peptides originated from 135 parent proteins. Thus, plasma peptides are high candidates for reliable biomarkers of kidney diseases.

CE‐MS/MS analyses of cerebrospinal fluid (CSF) of patients with amyotrophic lateral sclerosis (ALS) provided a peptide fingerprint of this serious neurodegenerative disease with abnormal protein aggregation in the motor neurons [285]. From the CSF analyses of 50 ALS patients and 50 healthy controls, 33 peptides were found. From them, 14 peptides could be sequenced using a nonlytic single‐pot proteomic CE‐MS/MS method. ALS‐deregulated peptides versus peptides of healthy controls included integral membrane protein 2B, neurosecretory protein VGF, osteopontin, neuroendocrine protein 7B2, chromogranin‐A, and several other physiologically important proteins.

7.1.7. Chiral Analysis and Separation of Stereoisomers

Being constituted of chiral building blocks, l‐ and d‐AAs, the sequentially identical peptides can exist in different enantiomeric and diastereomeric configurations. This may result in their different biological activity and binding capability to chiral ligands or receptors. Consequently, the chiral analysis of peptides and separation of their stereoisomers, especially when peptide drugs are concerned, is very important [286]. Thanks to their high separation efficiency and resolution power, short analysis times, and low consumption of not only sample but also of chiral selectors and other BGE components, CE methods represent powerful tools for chiral analysis and separation of stereoisomers in general [287, 288, 289], including chiral and stereoselective separations of peptides [290]. Among the chiral selectors, native noncharged cyclodextrins (CDs) and especially the derivatized charged CDs still represent the most popular ones [291, 292].

Separation of enantiomers of two polypeptide analogs of growth hormone‐releasing hormone (CJC‐1293 and sermorelin) was achieved by using a single isomer CD derivative, 15 mM heptakis (2,6‐di‐O‐methyl)‐β‐CD, as a chiral selector in the BGE composed of 100 mM lithium phosphate buffer, pH 3.5 [80], see Figure 7. These strongly basic peptides with pI 9.99 differed by l‐ and d‐configuration of only one AA, the second l‐ or d‐alanine of the total number of 29 AAs. Separation was performed in a BFS capillary (50/375 mm id/od, 60/50 cm total/effective length) at 30 kV separation voltage, 25°C, and 200 nm detection wavelength with hydrodynamic sample injection by pressure of 34.5 kPa for 10 s.

FIGURE 7.

Separation of four analogs of growth hormone–releasing hormone (1, tesamorelin; 2, CJC‐1293; 3, CJC‐1295; 4, sermorelin) by chiral CZE in the BGE composed of 100 mM lithium phosphate, pH 3.5, and the chiral selector (dimethyl‐β‐CD) at concentration of (A) 0 mM, (B) 15 mM, (C) 20 mM, and (D) 30 mM. Separations were performed in a BFS capillary (50/375 mm id/od, 60/50 cm total/effective length) at 30 kV separation voltage, 25°C, and 200 nm detection wavelength with hydrodynamic sample injection (34.5 kPa × 10 s). Source: Reprinted with permission from Otin et al. [80].

Chiral separation is important, especially in the determination of d‐AAs in total hydrolysates of peptides and proteins, because d‐AAs can serve as specific metabolic biomarkers for detecting various pathological states and for evaluating their recovery or progress [271, 286].

Peptides are not only the subject of chiral CE separations, but they are also used as chiral selectors for the enantioseparation of other chiral compounds. The most important peptide chiral selectors are macrocyclic glycopeptide antibiotics, vancomycin, teicoplanin, ristocetin, eremomycin, and their fragments and analogs. They are used for a broad class of chiral separations, both in electromigration and chromatographic techniques [293]. For example, vancomycin covalently attached to the silica in packed capillary columns was applied for the separation of various enantiomers by CEC [184].

7.2. Micropreparative Separations

The application of CE techniques for preparative separations of peptides and other (bio)molecules is rather limited. This is caused by two issues: first, the preparative capacity of the narrow bore capillaries with id less than 100 µm is inherently low; second, adaptation of CE devices for preparative purposes is more complicated than in LC systems. The latter issue, resulting from dipping of both capillary ends and electrodes into the BGE in the electrode vessels during the separation process, was solved by special modifications of the CE devices. These modifications and some protocols elaborated for fraction collections from the capillary were reported in the earlier reviews [294, 295], while the various both capillary and noncapillary devices and techniques applied for offline preparative separation of (bio)molecules are described in the recent review [296].

For peptide preparation and purification, CE methods are used mostly indirectly. They are utilized for analysis of peptides separated and purified by other methods, mostly preparative LC, but sometimes also by free flow electrophoresis [297]. Direct application of CE methods allows only microscale isolation of a peptide or other compound of interest. In commercial apparatuses, the autosamplers are adapted to fraction collectors, and the electro‐ or pressure‐driven flows are employed for elution of the particular compounds from the capillary.

The issue of limited preparative capacity of CE (mostly less than 1 µg per run) can be partially solved by: (1) enlarging the capillary id (but compromising separation efficiency), (2) repetitive fraction collection (but prolonging the separation time), and (3) using the multicapillary systems (but requiring special adaptation of CE devices). Multiple separations in a single narrow bore capillary and pooling of the fractions with the same mobility are used for isolation of peptides before their subsequent AA analysis or MS sequence, and M _r determination and identification. Microscale isolation of peptides is conducted in the offline coupling of CE with MALDI‐MS detection. Nanoliter volume fractions are collected on the MALDI‐compatible plates or membranes and subsequently analyzed by MALDI‐MS [298].

Substantial increase in preparative capacity can be achieved if analytical CE separations are converted into the preparative FFZE format [297, 299] following the earlier developed procedures [300, 301]. Other recent applications of various electromigration methods in the free‐flow instrumental format for middle‐scale isolation of biomolecules, including peptides, are reported in [297, 302]. FFZE conducted in microfluidic devices can also be employed for microscale isolation of peptides and other biomolecules [303]. However, due to their limited preparative capacity, they are more suitable for continuous monitoring of selected analytes than for real preparative purposes.

7.3. Physicochemical Characterization

Besides numerous analytical applications, CE methods are frequently utilized for the determination of various physicochemical parameters of peptides and other (bio)molecules. These parameters include the effective electrophoretic mobilities, actual and limiting (absolute) ionic mobilities, effective charges, pIs, M _rs, Stokes radii, diffusion coefficients, partition constants/coefficients (log P _o/w), distribution constants/coefficients (log D_o/w), acidity (acid dissociation, ionization) constants (pK_a ) of ionogenic groups, binding (association, stability, formation, complexation) constants (K _b) or dissociation constants (K _d) of peptide complexes, and Gibbs energy, enthalpy, entropy, rate constants, and activation energy of their chemical and enzymatic reactions and noncovalent and covalent interactions [304].

CZE in gel or other sieving media (CGE or CSE) [126, 127] provides data on the hydrodynamic (Stokes) radius and M_r of the separated analytes. The M_r of polypeptides and proteins can be estimated from CGE of their complexes with SDS (capillary version of SDS‐PAGE) [128]. Stokes radii and effective (net) charges of peptides can be assessed from the effective mobilities measured by CZE and from the models relating peptide mobility and their charge/size ratio [305, 306]. Nevertheless, due to the approximate character of these models, only approximate values of charge, size (Stokes radius), and M_r are obtained from these measurements. Another important peptide/protein parameter related to the charge distribution in the amphoteric molecules, pI, can be determined by CIEF [49] or by CZE [49, 52, 307]. However, also in this case, one should take into account that the obtained pI values are dependent on the composition of CAs in CIEF or BGEs in CZE, and other experimental conditions.

CZE in a free solution was successfully applied for the selective separation of the first three generations G₁–G₃ of dendrigraft poly(L‐lysine) (DGL) [110]. The separations were carried out in FS capillary (50/359 µm id/od, 58.5/50 cm total/effective length) coated with multilayer (PDADMAC/PSS)₂ + last layer coating, where the last layer was poly(allylamine hydrochloride) (PAH) with different PEGylation degree and with different methyl‐PEG (mPEG) chain length, to suppress adsorption of DGL to the inner capillary wall and to control the EOF velocity. BGE composed of 125/250 mM Tris/phosphate buffer, pH 2.2, separation voltage +20 kV at 25°C, UV‐detection at 200 nm, 0.001% (m/v) imidazole as a mobility marker, and the 1 kDa mPEG chains and low grafting density were found as the best experimental conditions for the fine EOF tuning and high peak resolution. In the CE analysis of DGL G1 generation, a high resolution of polymers with polymerization degree 5–17 was achieved (Figure 8). Polydispersity of molar masses and effective mobilities providing information about the dispersion of polymer distribution was successfully determined for the first three generations of the DGL.

FIGURE 8.

Optimized CE analysis of DGL G1 analysis at pH 2.2 using (PDADMAC/PSS)2‐(PAH‐g‐(mPEG_1kDa)_0.033)₁ coated FS capillary (50/359 µm id/od, 58.5/50 cm total/effective length) in BGE composed of 125/250 mM Tris/phosphate buffer, pH 2.2. Experimental conditions: +20 kV separation voltage, temperature of 25°C, 200 nm detection wavelength, with hydrodynamic injection for 4 s at 30 mbar pressure (0.27% of total capillary volume). Concentration of analytes: DGL G1 (0.5 mg/mL in BGE). Source: Reprinted with permission from Roca et al. [110].

Various ACE modes are often applied for the determination of the binding or dissociation constants of biomolecular complexes with different types of both low‐ and high‐molecular‐mass ligands, including peptides, for a survey, see the recent [163] or the earlier reviews [263, 308]. The study of fundamental, that is, method‐independent determinants of the accuracy of the binding constants of affinity complexes revealed the critical importance of the correct selection of the concentration of the interacting species and created a theoretical basis for improving the accuracy of the binding constants determined by ACE or by any other separation method [309]. The predicted influence of concentrations on the accuracy was also confirmed experimentally. This study provides guidelines for performing the ACE experiments under correct conditions: the binding interaction has to achieve the equilibrium and has to be performed in the so‐called binding regime, that is, the total concentration of ligand (analyte) is much less than the dissociation constant of its complex with the target (receptor, selector) molecule. In the subsequent study of the same group [310], the theoretical bases and practical recommendations to maximize the accuracy of determining the dissociation constants of affinity complexes were presented. By showing the main sources of systematic errors (concentrations of two interacting species and fraction of unbound ligand) and suggesting ways of their reduction, this study provides useful instructions on how to make the binding studies more accurate.

The importance of fast kinetics of molecular interactions for the determination of the binding constants by the mobility shift ACE methods is shown in [311]. If the interaction rate is not sufficiently high, the diffusion coefficients of interacting molecules and their complex have to be taken into account when modeling the CE separation of analyte enantiomers in the BGE containing the chiral selector.

Interactions between cyclosporin A (CsA), a cyclic undecapeptide used as an immunosuppressive drug, and cyclophilin A (CpA), a highly abundant cytosolic protein‐binding cyclosporin also used as an immunosuppressant, were explored by FA‐CE hyphenated with ESI‐TOF‐MS [158]. The analyses were performed in an Agilent 7100 CE system hyphenated with a G6224A TOF‐MS detector via a G1607‐60002 coaxial SL flow interface using a BFS capillary (50 µm id, 80 cm length) and BGE composed of 20 mM ammonium acetate, pH 7.4. MS conditions were as follows: dual ESI in positive ion mode, drying gas temperature of 325°C, drying gas flow rate of 5 L/min, nebulizing gas pressure of 5 psi, fragmentation voltage of 135 V, cone voltage of 65 V, ESI voltage 3.5 kV, octupole RF voltage 750 V. The detection range of m/z was 300‐1250. SL composed of 0.1% (v/v) FA was delivered at a flow rate of 4 µL/min by an Agilent pump with a 100:1 split ratio. Before each interaction analysis, 100 µL of a 55.44 µM solution of CsA and 100 µL of a 27.67 µM solution of CpA were mixed in a 1:1 ratio, resulting in an initial concentration of ligand L (CsP), [L]₀ = 27.72 µM, and receptor protein P (CpA), [P]₀ = 13.84 µM. The L and P were mixed by gentle flicking, followed by incubation for 1 h before being injected into the separation capillary. As a control, a 27.72 µM concentration of CsA was employed. In this study, two injections were used: one with a ligand solution only at [L]₀ concentration and another one with a pre‐equilibrated mixture of ligand [L]₀ and receptor protein [P]₀. These compounds, that is, free L, free P, and complex LP, were separated based on their different mobilities by CE, resulting in the migration of the free ligand L as an initial plateau. The height of the plateau of the free ligand L (CsA), in the ligand solution (H1) and in the pre‐equilibrated mixture (H2), was recorded by extracting the ligand's mass/charge ratio (m/z) from the total ion electropherogram (TIC). The concentration of free ligand in the mixture was determined from the linear proportionality between H1 and H2. This single‐point method combines effectiveness and rigor. The resulting [L] values were then substituted into Equations (1 and 2) for calculating the dissociation constant K _d.

$$[L]=\frac{{[L]}_0\times H_2}{H_1}$$ (1)

$$K_d=\left(\frac{{[P]}_0}{{[L]}_0-[L]}-1\right)\times [L].$$ (2)

Using this method, the K _d of the CsA–CpA complex was found to be equal to 3.32 µM. It was about five times higher than the K _d determined in the same study by a new technique, biolayer interferometry. The difference can be explained by partially different experimental conditions of these two methods.

In addition to the determination of thermodynamic parameters, kinetic characteristics of reactions and interactions can also be estimated by CE methods [161]. For example, the rate constants, the half‐maximal inhibitory concentration (IC₅₀) values, and Michaelis–Menten constants of enzymatic reactions were determined by EMMA [253] and by CE‐dynamic frontal analysis [312].

8. Concluding Remarks and Future Prospects

As shown in many of the above examples, CE and MCE methods represent powerful tools for fast, highly efficient separations, highly sensitive analyses, micropreparations, and physicochemical and biochemical characterizations of peptides. Currently, CE and MCE are recognized and appreciated as valuable complements and/or counterparts of the other high‐performance separation methods, especially various modes of (U)HPLC, capillary LC, and nano‐LC. Large applicability of CE and MCE methods in chemistry and biochemistry of peptides is demonstrated by their usage not only as highly efficient and highly sensitive analytical techniques able to determine femtomole to zeptomole amounts of peptides in nano‐ to picoliter injected sample volume of complex biological matrices, but also as highly valued physicochemical methods providing important physicochemical and biochemical characteristics of peptides. Moreover, compared with LC techniques, the green, environmental‐friendly nature of CE methods is highly appreciated.

Peptides belong to the most important biomolecules with a pivotal role in several vitally relevant physiological processes. Many of them and many of their functions have been recognized, but even more have to be revealed and further explored. For a more detailed understanding of both normal and pathological biological processes, a comprehensive investigation of the peptidome, that is, the whole peptide set of a cell, organ, or organism, has to be performed. On top of that, the structure and function of proteins in complex proteomic studies are frequently investigated through their peptide fragments. Consequently, the separation, analysis, purification, and characterization of peptides will undoubtedly remain one of the most challenging issues of CE and MCE methods also in the next years. As for these methods themselves, they are supposed to be further intensively developed. In particular, they will be further improved via implementation in microfluidic devices and integration into micro‐total analytical systems (µTAS) online, coupled with highly sensitive and high‐resolution MS or LIF detection. 2D and multidimensional separations based on a combination of orthogonal principles of CE and LC methods, and their hyphenation with tandem MS detection, will be necessary for analysis and identification of peptides and proteins in their complex mixtures in comprehensive peptidomic and proteomic studies.

In addition, the weaknesses of the methods will be reduced. Relatively low‐concentration sensitivity of UV‐absorption detection will be improved by its replacement with highly sensitive MS and LIF detections. Adsorption of analytes to the inner capillary wall will be suppressed by new capillary coatings, and selectivity of the separations will be enhanced by new types of additives to the BGE in CZE and ACE or by new constituents of pseudostationary phases in EKC and CEC.

Conflicts of Interest

The author declares no conflicts of interest.

Kašička V., “Recent Developments in Capillary and Microchip Electroseparations of Peptides (2023–mid 2025).” ELECTROPHORESIS 47, no. 1 (2026): 106–136. 10.1002/elps.70052

Data Availability Statement

The data are available from the corresponding author upon reasonable request.