Advances in enzyme substrate analysis with capillary electrophoresis

Abstract

Capillary electrophoresis provides a rapid, cost-effective platform for enzyme and substrate characterization. The high resolution achievable by capillary electrophoresis enables the analysis of substrates and products that are indistinguishable by spectroscopic techniques alone, while the small volume requirement enables analysis of enzymes or substrates in limited supply. Furthermore, the compatibility of capillary electrophoresis with various detectors makes it suitable for K_M determinations ranging from nanomolar to millimolar concentrations. Capillary electrophoresis fundamentals are discussed with an emphasis on the separation mechanisms relevant to evaluate sets of substrate and product that are charged, neutral, and even chiral. The basic principles of Michaelis-Menten determinations are reviewed and the process of translating capillary electrophoresis electropherograms into a Michaelis-Menten curve is outlined. The conditions that must be optimized in order to couple off-line and on-line enzyme reactions with capillary electrophoresis separations, such as incubation time, buffer pH and ionic strength, and temperature, are examined to provide insight into how the techniques can be best utilized. The application of capillary electrophoresis to quantify enzyme inhibition, in the form of K_I or IC₅₀ is detailed. The concept and implementation of the immobilized enzyme reactor is described as a means to increase enzyme stability and reusability, as well as a powerful tool for screening enzyme substrates and inhibitors. Emerging techniques focused on applying capillary electrophoresis as a rapid assay to obtain structural identification or sequence information about a substrate and in-line digestions of peptides and proteins coupled to mass spectrometry analyses are highlighted.

1. Introduction

The annual global market of industrial enzymes is reported to be billions of USD [1–7], and impacts commercial sectors that include energy, animal feed, household products, food processing and pharmaceuticals [1,2,5,6]. Enzymatic processing is even recognized as a critical component for sustainable chemical manufacturing [8]. For this reason significant effort is made to discover new enzymes [9,10] as well as to improve the stability, specificity, or efficiency of existing enzymes [7,11]. These endeavors require analytical approaches to characterize the enzyme performance in order to advance manufacturing.

Current approaches of monitoring enzymes require analytical tools to quantify the amount of product generated by incubating the enzyme with a particular substrate. A barrier to most approaches is that the product and substrate are indistinguishable by optical spectroscopy. This has led to the development of a limited set of substrates that undergo significant change in the spectroscopic profile upon conversion to product. An alternative approach is the use of separation-based assays that provide a means to sort the product and substrate in space and then detect them individually.

Capillary electrophoresis separations offer many advantages over chromatographic separations that make the method an attractive technique for enzyme analysis. Capillary electrophoresis consumes nano-to picoliter sample volumes for each run. Electrophoresis runs are rapid, requiring minutes to complete. The high separation efficiency of capillary electrophoresis enables the separation of complex samples. Commercially available instruments are equipped with robotics, which automate the sample analyses. The method is amenable to total miniaturization and can ultimately be translated into portable microfluidic systems. These benefits make capillary electrophoresis an excellent platform to study enzyme kinetics, specificity, and inhibition, as well as to expand the possibilities to use enzymes as tools in analytical techniques.

The purpose of this paper is to shed light on the capabilities of capillary electrophoresis for the evaluation of enzyme performance. In order to understand how the technique is adapted to different enzyme and substrate systems, fundamental principles of the method are described. The process of converting capillary electrophoresis separations into Michaelis-Menten constants (K_M) is also delineated. Applications reported from 2012 to 2017 are summarized. Areas of focus include the calculation of K_M, inhibition studies, optimization of enzyme turnover and approaches to screen or compare different enzymes. Different strategies for enzyme immobilization for in-capillary analyses are described. Future directions in this field, especially in strategies for in-capillary sequencing as well as structural identification are addressed.

2. Background

2.1. Fundamental principles of capillary electrophoresis

Separation in capillary electrophoresis is based on the charge-to-size ratio of analytes in an electric field. A simple schematic of a capillary electrophoresis system is provided in Fig. 1. Separations take place in a silica capillary filled with a background electrolyte. For methods employing optical detection, the polyimide coating on the outside of the capillary is removed to create a detection window which allows light to pass through to the detector. A high voltage power supply that can deliver up to 30,000 V is used to apply an electric field across the capillary during the separation.

Fig. 1.

Schematic of capillary electrophoresis setup. It consists of capillary, buffer reservoir containing background electrolyte and a detector. Analytes are separated in capillary under the influence of the electric field supplied by high voltage power supply. A color version of this figure is available on-line.

There are two modes of transport in capillary electrophoresis, electroosmotic flow and electrophoretic mobility. Electroosmotic flow, which is depicted by the thin black vector in Fig. 2, is the bulk flow of liquid in the capillary. It occurs in the presence of the electric field as a consequence of the surface charge on the inner wall of the fused silica capillary. The surface of the fused silica is negatively charged at pH levels above 4, so cations accumulate at the surface of the capillary. This results in a bulk flow of the liquid towards the cathode upon application of the electric field. Electrophoretic velocity, which is denoted by a thin grey vector in Fig. 2, is based on the charge-to-size ratio of an analyte. A small cation with the same charge as a larger cation will migrate toward the cathode faster than the larger cation. The net velocity of an analyte, represented by the large open arrow in Fig. 2, is based on the sum of the vectors for electroosmotic flow and electrophoretic velocity. Electroosmotic flow is typically greater than electrophoretic velocity; therefore, cations, neutrals, and anions can be analyzed in a single run. For cations, the analytes with high charge-to-size ratio migrate to the detection window before those with a low charge-to-size ratio. However, the reverse is true for anions since the electrophoretic velocity is in the opposite direction. This is emphasized in the electropherogram in Fig. 2, where the resulting order of migration is high charge-to-size ratio cations, low charge-to-size ratio cations, neutrals, low charge-to-size ratio anions, and high charge-to-size ratio anions.

Fig. 2.

Conceptual electropherogram showing the separation of a hypothetical mixture of five analytes using capillary electrophoresis. The mixture is separated by the charge-to-size ratio of the analytes. The charge and size of the analytes depend on the molecular pK_a value and the hydrodynamic radius, respectively. The molecular weights (MW) of the analytes are proportional to the hydrodynamic volumes. The vectors represent the electroosmotic flow (EOF), and the electrophoretic (EPH) and net velocities of the analytes. In this conceptual depiction the pH of the background electrolyte maintains positive charge of cation 1 and cation 2. Because it is larger, cation 2 has a slower migration time. Although anion 1 and anion 2 are both negatively charged, anion 2 has a slower migration because it is smaller and migrates against the bulk EOF. A color version of this figure is available on-line.

2.2. Enzyme analysis using capillary electrophoresis

2.2.1 Determining K_M values

Enzyme kinetics are commonly modeled based on the Michaelis-Menten equation, a rate equation originally proposed by Henri and applied by Michaelis and Menten [12]. The catalytic efficiency is commonly defined by k_cat/K_M; however, in order to define the k_cat value the true enzyme concentration must be known. The K_M value is substrate dependent, which provides a better means to gauge enzyme performance. This is important when the enzyme reaction is applied to a specific substrate. A small K_M value means high turnover is achieved. The enzyme assays can be performed off-capillary or in-capillary for the determination of kinetic parameters.

The process requires that the substrate is incubated with enzyme for a specified time. As depicted in Fig. 3, the reaction mixture is separated with capillary electrophoresis. The concentration of product is quantified, typically using an external calibration curve, and divided by the incubation time to yield rate. This change in the concentration of the generated product is evaluated for different substrate concentrations, all of which are incubated with the same concentration of enzyme. The rate of product formation is then plotted against the substrate concentration (Fig. 3). It is useful to collect a minimum of five substrate concentration to fit a curve, as shown in Fig. 3. The non-linear curve is fit using software to determine the K_M value from the Michaelis-Menten equation. This is shown in Eq. (1)

$$\upsilon =({\mathrm{V}}_{max}\cdot [\text{Substrate}])/({\mathrm{K}}_{\mathrm{M}}+[\text{Substrate}])$$ (1)

where υ is the rate/velocity of the reaction, V_max is the maximum velocity at which substrate reaches saturation, and K_M is the substrate concentration at which enzyme performs at half of the maximum velocity.

Fig. 3.

Conceptual diagrams demonstrating K_M analysis using capillary electrophoresis. Electropherograms in inset show five different substrate concentrations and the products generated after the enzyme reactions. The generated products were zoomed to emphasize the product area increases as the initial substrate concentration increases. The curve on the right depicts the Michaelis-Menten curve is generated by plotting the rate of product formation versus the substrate concentration. A color version of this figure is available on-line.

2.2.2 Constraints of the assay

Before determining enzyme kinetics, there are some basic recommendations that should be implemented. The assumption of steady state, which refers to the condition under which the rate of formation and depletion of the enzyme-substrate complex are equal, requires that the analysis be performed when there is not a high accumulation of product. Initial rates in which the product formation or substrate consumption does not exceed more than 10% are used to avoid measuring the rate when the product concentration is too high, which will make the reversible reaction more favorable in accordance with Le Chatelier’s principle. Furthermore, rate of enzyme turnover can decrease due to product accumulation.

3. Adapting the separation to determine K_M values

From 2012 to 2017 approximately fifty K_M determinations were reported in the literature that utilized capillary electrophoresis. These reports, summarized in Table 1, were predominantly studies of hydrolases or oxoreductases, although transferases, lyases, and isomerases were also investigated. Separations were based on differences in the charge-to-size ratio of the substrate and product for most reports. Several reports evaluated enzyme specificity for enantiomers and as a result, additives that separated chiral substrates were included in the background electrolyte. The primary method of detection was UV–visible absorbance detection, which is a universal detection method applicable to most analytes. In addition, the linear range of absorbance detection, typically between 50 and 500 μM, is appropriate for the reported K_M values. Enzyme assays were performed both off-line and on-line, depending upon the conditions required of the enzyme reaction and the constraints of the assay. These aspects of enzyme analyses are addressed in greater detail in the sections that follow.

Table 1.

Michaelis-Menten constants obtained with capillary electrophoresis.

Enzyme	Substratea	KM(μM)b	Notesc	Refs
Hydrolase
Acetyl cholinesterase	acetylthiocholine	660	UV–vis, on-line	[56]
Alkaline phosphatase	p-nitrophenyl-phosphate	1515	UV–vis, off-line	[65]
Aminopeptidase N	L-leucine-p-nitroanilide	340	UV–vis, off-line	[66]
Camel chymosin	K-casein	290	UV–vis, off-line	[49]
Capsule biosynthesis protein Cap D	UDP D-GlcNac	3700	UV–vis, off-line	[15]
Capsule biosynthesis protein Cap E	UDP D-GlcNac	460	UV–vis, off-line	[15]
α-Chymotrypsin	n-benzyl-L-tyrosine ethyl ester	3530	UV–vis, on-line	[59]
β-Galactosidase	para-nitrophenyl-β-D-galactopyrnoside	430	UV–vis, on-line	[44]
β-Galactosidase	4-nitro-phenyl-D-galactopyranoside	560	UV–vis, on-line	[43]
β-Galactosidase	4-nitro-phenyl-D-galactopyranoside	600	UV–vis, off-line	[43]
β-Galactosidase	4-nitro-phenyl-D-galactopyranoside	700	UV–vis, on-line	[43]
α-Glucosidase	4-nitrophenyl-D-glucopyranoside	610	UV–vis, off-line	[38]
α-Glucosidase	4-nitrophenyl-D-glucopyranoside	810	UV–vis, off-line	[42]
Hyaluronidase	hyaluronic acid	290	UV–vis, off-line	[16]
Hyaluronidase	hyaluronic acid	560	UV–vis, off-line	[21]
Hyaluronidase	hyaluronic acid	600	MS, off-line	[16]
Leucine aminopeptidase	leucine-proline	20,700	electrochemi-luminescence, off-line	[51]
Myrosinase	glucosinalbin	72	conductivity, off-line	[19]
Myrosinase	glucotropaeolin	130	conductivity, off-line	[19]
Myrosinase	gluconapin	210	conductivity, off-line	[19]
Myrosinase	glucoraphasatin	310	conductivity, off-line	[19]
Myrosinase	glucoiberin	460	conductivity, off-line	[19]
Myrosinase	glucomoringin	2600	conductivity, off-line	[19]
Myrosinase	naphthylmethyl glucosinolate	10,000	conductivity, off-line	[19]
Myrosinase	carbazolylmethyl glucosinolate	13,000	conductivity, off-line	[19]
Myrosinase	orthosinalbin	16,000	conductivity, off-line	[19]
Myrosinase	phenanthrenylmethyl glucosinolate	17,000	conductivity, off-line	[19]
α2-3 Neuraminidase	3′-sialyllactose	3000	UV–vis, on-line	[26]
α2-3,6,8 Neuraminidase	MUNANA	126	UV–vis, on-line	[63]
α2-3,6,8 Neuraminidase	4-MuNeu5Ac	140	UV–vis, on-line	[57]
α2-3,6,8 Neuraminidase	60-sialyllactose	2000	UV–vis, on-line	[26]
α2-3,6,8 Neuraminidase	30-sialyllactose	3300	UV–vis, on-line	[26]
Neutrophil elastase	5-Fam-Ala-Ala-Ala-Phe-Tyr-Asp-OH	70	LIF, on-line	[41]
Neutrophil elastase	5-Fam-Arg-Glu-Ala-Val-Val-Tyr-OH	70	UV–vis, on-line	[41]
Neutrophil elastase	5-Fam-Arg-Glu-Ala-Val-Val-Tyr-OH	70	LIF, on-line	[41]
Neutrophil elastase	5-Fam-Ala-Ala-Ala-Phe-Tyr-Asp-OH	80	LIF, on-line	[41]
Neutrophil elastase	N-Meo-Suc-Ala-Ala-Pro-Val-pNA	130	UV–vis, on-line	[41]
Nucleotide pyrophosphatase 1	p-nitrophenyl 5′-thyminidine monophosphate	280	UV–vis, off-line	[55]
Nucleotide pyrophosphatase 3	p-nitrophenyl 5′-thyminidine monophosphate	130	UV–vis, off-line	[55]
Protein tyrosine phosphatase 2	pTS13	0.29	LIF, off-line	[39]
Protein tyrosine phosphatase 1	pTS13	0.8	LIF, off-line	[39]
Sirtuin 1	7-Moc-Lys(Ac)-NH2	80	UV–vis, off-line	[14]
Thrombin	S-2366	310	UV–vis, on-line	[58]
Trypsin	benzoyl L-arginine ethyl ester hydrochloride	240	UV–vis, off-line	[17]
Trypsin	benzoyl L-arginine ethyl ester hydrochloride	500	UV–vis, on-line	[18]
Trypsin	benzoyl L-arginine ethyl ester hydrochloride	1230	UV–vis, on-line	[13]
Oxidoreductases
Alcohol dehydrogenase	acetaldehyde	1150	UV–vis, on-line	[60]
Cytochrome p450 2D6	S-fluoxetine	30	UV–vis, on-line	[32]
Cytochrome p450 2D6	R-fluoxetine	39	UV–vis, on-line	[32]
Cytochrome p450 CYP3A12	R-ketamine	190	UV–vis, off-line	[29]
Cytochrome p450 CYP3A12	S-ketamine	380	UV–vis, off-line	[29]
Cytochrome p450 CYP3A2	R-verapamil	47	UV–vis, on-line	[33]
Cytochrome p450 CYP3A3	S-verapamil	51	UV–vis, on-line	[33]
Cytochrome p450 CYP3A4	R-ketamine	68	UV–vis, on-line	[31]
Cytochrome p450 CYP3A4	R-ketamine	84	UV–vis, on-line	[31]
Cytochrome p450 CYP3A4	S-ketamine	97	UV–vis, on-line	[31]
Cytochrome p450 CYP3A4	R-ketamine	107.5	UV–vis, on-line	[30]
Cytochrome p450 CYP3A4	S-ketamine	110	UV–vis, on-line	[31]
Cytochrome p450 CYP3A4	S-ketamine	122.3	UV–vis, on-line	[30]
D-amino acid oxidase	D-methionine	1100	UV–vis, on-line	[61]
D-amino acid oxidase	D-methionine	2400	UV–vis, on-line	[46]
Glucose 6-phosphate dehydrogenase	NAD+	400	UV–vis, on-line	[23]
Glucose 6-phosphate dehydrogenase	NAD+	450	UV–vis, on-line	[22]
Human methionine sulfoxide reductase A	diastereomeric pentapeptide sub (Ac-KIFM(o)K-Dnp)	135	UV–vis, off-line	[53]
Human methionine sulfoxide reductase A	Fmoc L-methionine sulfoxide	290	UV–vis, on-line	[34]
Human methionine sulfoxide reductase B	F-moc L-methionine sulfoxide	120	UV–vis, on-line	[34]
Human methionine sulfoxide reductase B2	diastereomeric pentapeptide sub (Ac-KIFM(o)K-Dnp)	100	UV–vis, on-line	[53]
Lactate dehydrogenase	pyruvate	450	UV–vis, on-line	[60]
Lactate dehydrogenase	D-lactate	3900	UV–vis, on-line	[62]
Lactate dehydrogenase	L-lactate	4200	UV–vis, on-line	[62]
Lignin peroxidase	manganese (II)	630	UV–vis, on-line	[67]
Lignin peroxidase	veratryl alcohol	2000	UV–vis, on-line	[67]
Manganese peroxidase	manganese (II)	220	UV–vis, on-line	[67]
Manganese peroxidase	veratryl alcohol	220	UV–vis, on-line	[67]
Manganese peroxidase 1	fructose 6-phospahte	810	UV–vis, off-line	[37]
Manganese peroxidase 1	CTP	910	UV–vis, off-line	[37]
Manganese peroxidase 1	dCTP	1560	UV–vis, off-line	[37]
Manganese peroxidase 2	CDP-d-fructose	340	UV–vis, on-line	[37]
Manganese peroxidase 2	NADPH	1380	UV–vis, off-line	[37]
Manganese peroxidase 2	NADH	2040	UV–vis, off-line	[37]
NADH oxidase	NADH	500	UV–vis, on-line	[20]
Tyrosinase	L-tyrosine	636	UV–vis, off-line	[28]
Tyrosinase	L-3,4-dihydroxy-phenylalanine	932	UV–vis, on-line	[45]
Tyrosinase	L-3,4-dihydroxy-phenylalanine	1347	UV–vis, on-line	[45]
Tyrosinase	L-3,4-dihydroxy-phenylalanine	1780	UV–vis, on-line	[68]
UDP-glucose dehydrogenase	UDP-galactose	30	UV–vis, off-line	[20]
UDP-glucose dehydrogenase	NAD+	1100	UV–vis, off-line	[20]
Xanthine oxidase	6-mercaptopurine	44	UV–vis, off-line	[25]
Transferases
Acetyl Co-A carboxylase	D-xylose-5-Phosphate	1020	UV–vis, off-line	[36]
Acetyl Co-A carboxylase	CTP	1760	UV–vis, off-line	[36]
Acetyl Co-A carboxylase	dCTP	2850	UV–vis, off-line	[36]
Acetyl Co-B carboxylase	NADPH	1100	UV–vis, off-line	[36]
Acetyl Co-B carboxylase	CDP-D-xylose	1530	UV–vis, off-line	[36]
Acetyl Co-B carboxylase	NADH	1900	UV–vis, off-line	[36]
Aminotransferase	L-valine	500	MS, on-line	[24]
Aminotransferase	L-aspartate	1100	MS, off-line	[24]
Aminotransferase	D-glutamate	5000	MS, on-line	[24]
Bovine K-casein	K casein	210	UV–vis, off-line	[49]
Galactokinase E.Coli	galactose	200	UV–vis, on-line	[20]
Galactokinase E.Coli	adenosine triphosphate	300	UV–vis, on-line	[20]
β1,4 Galactosyl-transferase 1	6-sulfo-GlcNac-Mu	90	UV–vis, off-line	[69]
β1,4 Galactosyl-transferase 1	UDP-gal	230	UV–vis, off-line	[69]
UDP sugar pyrophosphorylase	Gal-1-P	3300	UV–vis, off-line	[20]
UDP sugar pyrophosphorylase	UTP	5000	UV–vis, off-line	[20]
Lyase
bovine carbonic anhydrase	p-nitrophenyl acetate	1100	UV–vis, off-line	[40]
diaminopimelate decarboxylase	diaminopimelate	1600	UV–vis, off-line	[54]
Isomerase
UDP-glucose-4-epimerase (KfoA)	UDP D-GlucNac	2080	UV–vis, off-line	[70]
UDP-glucose-4-epimerase (KfoA)	UDP-Glucose	4210	UV–vis, off-line	[70]
UDP-glucose-4-epimerase (K4)	UDP D-GlucNac	4260	UV–vis, off-line	[70]

^aSubstrate abbreviations are as follows: CDP-D-Fucosem cytidine diphosphate-D-fucose; CDP-D-Xylose, cytidine diphosphate-D-xylose; CTP, cytidine triphosphate; dCTP, deoxycytidine triphosphate; Fmoc L methionine sulfoxide, fluorenylmethyloxycarbonyl-L methionine sulfoxide; Gal-1-P, galactose 1-phosphate; GlucNac, N-acetylglucosamine; MUNANA, 2′-(4-Methylumbelliferyl)-α-D-N-acetylneuraminic acid; NADH, nicotinamide adenine dinucleotide; NADPH, nicotinamide adenine dinucleotide phosphate; pTS13, Glu-Glu-Leu-Glu-Asp-Asp-pTyr-Glu-Asp-Asp-Nle-Glu-Glu-amide; S-2366, L-pyroglutamyl-L-prolyl-L-arginine p-Nitroaniline hydrochloride; UDP-galactose, uridine diphosphate galactose; UDP-glucose-4-epimerase, uridine diphosphate-glucose-4-epimerase; UDP-glucose dehydrogenase, uridine diphosphate-glucose dehydrogenase; UDP-D-GlucNac, uridine diphosphate N-acetylglucosamine; UDP sugar pyrophosphorylase, uridine diphosphate sugar pyrophosphorylase; UTP, uridine-5′-triphosphate; 4-MuNeu5Ac, 4-methylumbelliferyl-N-acetylneuraminic acid; 6-sulfo-GlcNac, 6-sulfate modified N-acetylglucosamine; 6-sulfo-GlcNac-Mu, 4-methylumbelliferyl-6-sulfo-N-acetyl-β-D-glucosaminide; 7-Moc-Lys(Ac)-NH2, N-acetyl-N-(4-methyl-7-methoxycoumarin)lysine amide.

^bAnalytes detected using conductivity, electrochemiliuminescence, laser induced fluorescence (LIF), mass spectrometry (MS), UV–visible absorbance detection (UV–vis).

^cWhen error is reported the K_M value are modified to contain the appropriate number of significant figures as delineated by the uncertainty.

3.1. Modes of capillary electrophoresis separations to isolate product from substrate

3.1.1. Charge-to-size ratio

The mode of separations employed in the capillary electrophoresis analysis is strongly dependent on the analyte of interest, and the predominant separation mode for many different substrates and products is based on charge-to-size ratio. In order to determine if the substrate and product can be separated electrophoretically, the charge-to-size ratio of a molecule is estimated using the pK_a to determine the charge and molecular weight is used to estimate the size. To maintain the analyte charge, the pH of the separation buffer must be chosen based on the pK_a of the analytes. Enzyme reactions can be monitored using capillary electrophoresis when the substrate has a different charge-to-size ratio from the product. Hydrolase, transferase, and lyase enzymes perform reactions that inherently change the size of the molecules [13–21], so the substrates can be separated from the products. This is depicted in the separation of deacetylated product shown in Fig. 4. The charge-to-size ratio can also be altered through a change in the charge of the molecules, as is the case for glucose-6-phosphate dehydrogenase conversion of NAD+ to NADH [22,23], aminotransferase transfer of an amine group from L-valine to α-ketogluterate [24], xanthine oxidase oxidation of 6-mercaptopurine to 6-thioric acid [25], and neuraminidase removal of sialic acids from glycans [26], including those found in tumor tissues and serum samples [27]. While the change in charge changes the charge-to-size ratio of NADH and NAD+ to enable separation, researchers can still utilize the difference in the spectroscopic properties of these molecules to quantify only NADH at 340 nm [22,23].

Fig. 4.

Electropherogram of standards relevant to the deacetylation of the substrate, (2S)-6-(acetylamino)-2-[[(7-methoxy-2-oxo-2H-1-benzopyran-4-yl) methyl]amino]-hexanamide, (peak 4) with human SIRT1 enzyme. The product, (2S)-6-amino-2-[[(7-methoxy-2-oxo-2H-1-benzopyran-4-yl)methyl]amino]-hexanamide, (peak 1), migrates faster than the substrate. Peaks 2 and 3 are the co-product nicotinamide, and the internal standard 4-(aminomethyl)benzoic acid, respectively. From H. Abromeit, S. Kannan, W. Sippl, G.K.E. Scriba, A new nonpeptide substrate of human sirtuin in a capillary electrophoresis-based assay. Investigation of the binding mode by docking experiments, Electrophoresis 33(11) (2012) 1652–1659. Copyright © 2012 by John Wiley Sons, Inc. Reprinted by permission of John Wiley & Sons, Inc.

3.1.2. Secondary equilibria with cyclodextrin host:guest complexes

Capillary electrophoresis techniques have also been developed that employ secondary equilibria in order to separate substrates and products with charge-to-size ratios that do not significantly differ. Chiral compounds have been resolved with cyclodextrin [28], highly sulfated cyclodextrins [29–33], or micelles [34]. In these techniques, analytes are separated based on differences in their interaction with the secondary selector either due to the hydrophobicity or the electrostatic charge. One technique utilized sulfated beta-cyclodextrins in a capillary electrophoresis separation to distinguish diastereomeric peptides, which enabled the researchers to determine the methionine sulfoxide reductase A preference for turnover of different stereoisomers. This technique was further applied to determine the impact of mutations in the methionine sulfoxide reductase A on substrate conversion [35].

3.1.3. Alternative secondary equilibria and size-based separation

Boric acid is also used as a complexing agent that aids in the separation of substrates and products [36–45] based on the formation of a complex between boric acid and cis-diols. In addition to borate and highly sulfated cyclodextrins, ionic liquids are used for the separation of labelled amino acid enantiomers [46]. Size-based separations are another mechanism through which enzymes are studied and are applied to biopolymers, such as DNA, for which the charge-to-size ratio of different fragments is identical regardless of size. These separations are achieved through the incorporation of a gel in the capillary that impeded the migration of charged analytes. Smaller analytes arrive at the detector first, while larger analytes are retarded by the gel and migrate slower to the detection window. For negatively charged analyte, the electroosmotic flow is entirely suppressed. During the period of this review, size based separations utilizing capillary electrophoresis were almost exclusively used for the analysis of the products of enzymes, such as polymers [47] and DNA [48].

3.1.4. Modification of the capillary surface

The surface charge on the separation capillary contributes to the electroosmotic flow, which allows cations, neutrals, and anions to be analyzed in a single run. Reversed polarity separations provide better separations for samples composed only of anions; however, the bulk electroosmotic flow must be eliminated or reversed. An additional role for surface coating is to decrease the analyte interaction with the fused silica walls. To facilitate this specialized coatings are applied to the capillary wall in order to suppress the electroosmotic flow, in the case of polyvinyl acrylamide [49] and phospholipids [26,50], or to reverse the electroosmotic flow, in the case of polybrene [15]. The use of coatings may be expanded to analytes that are not anionic by derivatizing the analytes. Labelling with a chromophore or fluorophore is one method that is commonly used to circumvent both issues with charge and with surface interaction with the capillary wall since most commercially available fluorescent labels are negatively charged.

3.2. Modes of detection of capillary electrophoresis

A practical consideration in selecting the type of detector for K_M determinations is matching the linear range of the detection mode with the concentration range that must be analyzed to obtain a K_M curve. The enzymes in Table 1 emphasize the dynamic range of concentrations (nM to mM) over which K_M analyses occur. While it is important for the range of the substrate concentration to bracket the concentration of the K_M, it is the concentration of product generated that must fall inside the linear range of the detector. Incubation times can be extended for low substrate concentrations to generate sufficient product for detection as long as the generation of product is low enough that the rate is not impacted by the presence of product. Ultraviolet absorbance is utilized to measure the K_M values in the μM to mM range, as shown in Table 1 for K_M values in the range of 30 μM [20,32] to 4.6 mM [20]. While most analytes inherently absorb light in the ultraviolet region and do not require modification for detection, labeling with a chromophore can be utilized for analytes like carbohydrates that do not absorb light. Laser induced fluorescence detection can be utilized when the K_M is in the nM to low μM range since it provides low detection limits. However, fluorescence detection requires labeling [39,41]. Conductivity [19], electrochemiluminescence [51] and mass spectrometry [16,24] offer alternative detection techniques for analytes that exhibit poor absorbance, and labeling is not practical. Mass spectrometry is especially useful when more than one product is possible and more structural information can be used to elucidate which product is being generated [16].

3.3. Off-line and on-line coupling the incubation with the capillary electrophoresis

3.3.1. Compatibility of the enzyme reaction and the separation

The potential to integrate an enzyme reaction within a separation capillary depends upon many factors. Four relevant parameters to be discussed are pH, ionic strength, time of incubation, and the amount of enzyme and substrate available for reaction. The pH is often important to support the enzyme reaction; however, the pH of the capillary electrophoresis buffer plays a role in the effectiveness of the separation. It influences the electroosmotic flow and depending upon the pK_a of the analyte, the pH will also affect the electrophoretic mobility. In addition, a difference in ionic strength of the enzyme reaction and background electrolyte must be considered, because higher ionic strength electrolytes generate higher separation currents and lead to Joule heating, which decreases the separation efficiency [52]. Often high ionic strength buffers support enzyme stability, but may not affect the rate of the enzyme reaction. For on-line reactions some processing may be required to reduce the ionic strength of the enzyme to a lower ionic strength that matches that of the background electrolyte used for the separation. If the enzyme reaction must be completed at higher temperatures, off-line reactions are used; whereas on-line reactions are performed at lower temperatures to control Joule heating. With an automated instrument it is possible to program one temperature for the enzyme reaction and a lower temperature for the separation [34]. The amount of enzyme or substrate available may also dictate the mode of analysis. Off-line reactions can accommodate larger volumes of both enzyme and substrate. The volume of a typical separation capillary is approximately 1 μl, consequently on-line enzyme incubations consume only nanoliter volumes of enzyme and even smaller volumes of substrate for each run.

3.3.2. Off-line incubations

For pre-capillary assays or off-line analysis, enzyme reactions are performed on the benchtop. An off-line enzyme incubation is preferred if the working pH range of the enzyme reaction does not support the separation of product from substrate. These reactions typically use reaction volumes of 100 μL or greater [29,38,39,49,53–55], although smaller volumes have been reported [36,37]. Off-line enzyme reactions can be quenched to allow more flexibility in the separation of each sample; for example, when reactions are run in parallel. The enzyme reactions are typically quenched by heating [38], perchloric acid [54], hydrochloric acid [39] or freezing [40] prior to analysis. Off-line kinetic studies have been performed with several enzymes, including those used in biosynthesis pathways [15,36,37] and those with enantiomers or diasteriomers as substrates [29,34,46].

3.3.3. In-line incubations: overview

For in-capillary assays or on-line analysis, the enzyme reaction and mixing occur within the capillary. Generally on-line analyses require small sample volumes of substrate and enzyme with each run typically consuming only nanoliters of sample. Incubations for enzyme reactions on-line are typically less than 10 min and the reactions do not require quenching since the substrate and product are generally migrated away from the enzyme following the incubation period.

3.3.4. Role of pH with on-line enzyme incubations

Most of the on-line incubations described in the paper either use the same pH [13,18,23,24,34,43,56–62], or a pH close to enzyme reaction [26,32,33,41,55]. In some instances a wide difference of pH existed between the enzyme incubation buffer and the separation buffer [31,63]. The limitation of buffer incompatibility with on-line enzyme incubations can be overcome by using the method of partial filling or by using diffusion based mixing for electrophoretically mediated microanalysis.

3.3.5. Partial filling for discontinuous reaction and separation pH

In the method of partial filling plugs are injected as follows: (1) incubation buffer, (2) enzyme, (3) substrate, (4) enzyme and (5) incubation buffer. Different plug lengths may be used and each plug is pressure-injected to avoid interference in the reaction from the background electrolyte [32–34]. In some cases, the enzyme plug is injected only once after the substrate injection. In this method, the substrate and enzyme are still bracketed within the incubation buffer plugs on both sides. A brief application of voltage drives substrate through enzyme plug and allows for the enzyme incubation by electrophoretic mixing or by simple diffusion.

3.3.6. Transverse diffusion for discontinuous reaction and separation pH

Another method for performing enzyme reactions when buffers are incompatible is diffusion based mixing or transverse diffusion of laminar flow profiles. In this approach, on-line mixing was achieved by injecting reactants of different pH values into the capillary using a series of small hydrodynamic injections. With this approach, an enzyme buffered at pH of 7.4 was combined with a separation medium buffered at pH of 2.5 [30,31]. The parabolic shape of consecutive injections, which was a result of the laminar flow, caused each injection band to blend with the surrounding injection bands. The mixing of reactants occurs by transverse diffusion, where reactants diffuse towards the wall of the capillary. As a result, this injection mode can be utilized to mix multiple injected reactants for enzymatic reactions where the pH of the enzyme reaction was different than that of the background electrolyte [30,31,41].

3.3.7. Mixing in capillary

The reaction time must be known and often is adjusted to obtain quantifiable product at different substrate concentrations. For static incubations in-capillary the analyte will experience extensive diffusional band broadening at longer incubation times, which will adversely impact the resolution of the product from other peaks. Analyte mixing is preferable for on-line K_M determinations. Without mixing, a change in rate will be observed over time. This is because once the substrate adjacent to the enzyme is depleted the reaction is diffusion-limited. Mixing must be uniform in order to generate consistent velocities derived from areas that increase linearly with incubation time. Mixing through polarity cycling has been reported [26,64]. This is accomplished by creating a fixed zone of enzyme through the use of a phospholipid nanogel and a buffer at the pH of the isoelectric point of the enzyme in the reaction zone. The substrate is electrophoretically driven back and forth through the enzyme zone multiple times to achieve specific incubation periods as depicted in Fig. 5. The success of this approach requires precise knowledge of the position of the enzyme zone before implementing the method.

Fig. 5.

Conceptual diagram of electrophoretic mixing in the fixed enzyme zone. The mixing duration was determined by the total number of times the substrate passes through the fixed enzyme zone and the formation of product increases with the mixing time. The results of the electrophoretic mixing are shown in the adjacent electropherogram where the rate of product formation are obtained. Reprinted with permission from S. Gattu, C.L. Crihfield, L.A. Holland, Microscale Measurements of Michaelis–Menten Constants of Neuraminidase with Nanogel Capillary Electrophoresis for the Determination of the Sialic Acid Linkage, Analytical Chemistry 89(1) (2017) 929–936. Copyright 2017 American Chemical Society. Further permissions related to the material excerpted should be directed to the ACS. <http://pubs.acs.org/doi/abs/10.1021/acs.analchem.6b04074>.

4. Evaluating enzyme inhibitors using capillary electrophoresis

4.1. Background

4.1.1 Determining K_I for competitive inhibition

The presence of inhibitors during an enzymatic reaction reduces the reaction rate and product formation. The inhibition mechanisms, including competitive, uncompetitive, non-competitive, and mixed inhibition, and the degree to which a molecule inhibits enzyme performance is evaluated by K_I. The K_I is determined by incubating the substrate and the inhibitor with the enzyme for a specified time. The reaction mixture is then separated with capillary electrophoresis. As depicted in Fig. 6, the addition of the inhibitor decreases the amount of product generated (grey trace) as compared to the enzyme reaction performed in the absence of inhibitor (black trace). The changes in the product concentration in the presence of a fixed concentration of inhibitor are recorded while varying the concentration of substrate (Fig. 6). The reaction rate of competitive inhibition is modeled using Eq. (2).

$$\upsilon =({\mathrm{V}}_{max}\cdot [\text{Substrate}])/({\mathrm{\alpha }\mathrm{K}}_{\mathrm{M}}+[\text{Substrate}])$$ (2)

where α is the indicator for how much the Michaelis-Menten constant changes in the presence of an inhibitor. The value of α·K_M is called the apparent K_M value.

$$\mathrm{\alpha }=1+[\mathrm{I}]/{\mathrm{K}}_{\mathrm{I}}$$ (3)

$${\mathrm{K}}_{\mathrm{I}}=[\mathrm{E}][\mathrm{I}]/[\text{EI}]$$ (4)

Fig. 6.

Conceptual diagrams demonstrating K_I analysis using capillary electrophoresis. Traces in A depict electropherograms generated in the absence and presence of inhibitor showing decrease in product area when inhibitor was present. Traces in B depict the product generated in the absence and presence of inhibitor at various substrate concentrations with the concentration of inhibitor being same. The traces in grey with inhibitor shows the decrease in product area and were offset in time for clearer representation. The graph in C is a hypothetical Michaelis-Menten curve generated by plotting the substrate concentration versus the rate of product formation in the absence (− inhibitor) and the presence (+ inhibitor) of the inhibitors. A color version of this figure is available on-line.

The value for a is calculated using Eq. (3), where I represents the inhibitor. The inhibition constant, K_I, is calculated using Eq. (4) and is independent of the substrate concentration.

4.2. Inhibition analyses by capillary electrophoresis

Capillary electrophoresis methods have gained wide application for inhibitor studies. Similar to K_M studies, inhibition studies for enzymes have been conducted as both off-line and on-line analyses. Off-line analysis for enzymatic inhibition studies are performed by incubation of enzyme and substrate before the capillary electrophoresis analysis. The enzymatic reactions are then quenched and injected into the capillary for separation [14,15,20,21,28,29,38,40,55,65,71,72]. The substrates and products are separated and quantified for kinetic analysis of enzyme reaction and elucidation of the inhibitor mechanism and effectiveness. On the other hand, the on-line approach initiates enzymatic reaction through mixing in-capillary. The on-line mixing of enzyme, substrate, and inhibitor has been accomplished through electrophoretically mediated microanalysis [58,59,62,63,66,73], transverse diffusion of laminar flow profiles [21,31,41,74,75], and pressure mediated microanalysis [45]. On-line analysis with the use of immobilized enzyme for inhibitor studies have also been developed and high throughput analysis was achieved [18,22,23,56,68,76–78].

4.2.1. K_I Determination: competitive inhibition

Competitive inhibition was the most prevalent mechanism of inhibition studied using capillary electrophoresis [13,20,22,23,25,31,79]. An example of such a determination is reported for allopurinol, an inhibitor of the enzymatic reaction of xanthine oxidase, which converted 6-mercaptopurine to 6-thioric acid [25]. In the absence of inhibitor, capillary electrophoresis can be used to monitor an increase in the product as time progressed. When the reaction was performed in the presence of inhibitor allopurinol competed with 6-mercaptopurine resulting in xanthine oxidase catalyzing allopurinol to oxipurinol first, and then to 6-mercaptopurine. The signal of 6-thioric acid, which appeared after 2.6 h, was reduced when allopurinol was present.

4.2.2. Mechanisms of inhibition

Other mechanisms beyond competitive inhibition have also been interrogated by capillary electrophoresis. In those cases, the reaction rate in the presence of inhibitor is modeled differently for each inhibition mechanism [80]. Calculating the K_I value requires knowledge of the Michaelis-Menten constant so that the appropriate equation is used in the determination. During the time frame of this review, capillary electrophoresis has been used to calculate K_I for inhibitors that were competitive [13,20,22,23,25,31,79], mixed [56,77], and noncompetitive [72] inhibition. When the goal is to identify the mechanism of inhibition, Eq. (4) can be linearized so that the data can easily be visualized using a Lineweaver-Burk plot, as demonstrated by several reports published during the time period of this review [13,22,25,56,72,77].

Capillary electrophoresis can be used to rapidly identify an inhibitor without calculating the K_I. This is achieved by determining the IC₅₀ value, which is defined as the inhibitor concentration necessary to obtain 50% of the original response in the absence of inhibitor for a given substrate concentration. The IC₅₀ is obtained by monitoring the product formation while varying the concentration of inhibitor [13,21–23,29,31,39,46,56,68,71,74,75,81,82]. The lower the IC₅₀ value is, the more effective the inhibitor. Alternatively, the percent differences of the product peak area in the absence and presence of inhibitor can also be used to estimate the inhibition efficiency of the inhibitor [13,21,56,68,73–77,82,83].

4.2.3. Determining K_D to evaluate K_I

The binding affinity between enzyme and inhibitor is an additional method used to determine the effectiveness of the inhibitor. The experiment is conducted by incubating the enzyme and the inhibitor without the use of substrate. The dissociation constant, K_D, is determined by monitoring the peak area of the complex of enzyme and inhibitors [72] since the K_D is equal to the product of the enzyme and inhibitor concentrations divided by the complex concentration (K_D = [E][I]/[EI]). Alternatively, the migration velocity can also be used to determine the binding affinity with the method of affinity capillary electrophoresis [84]. This method focuses on the mobility shift upon complex formation. In order to perform this analysis, the capillary is first filled with the inhibitor of interest and the complex is formed after injecting the enzyme plug. The average velocity of the inhibitor would be reduced due to the partially formed complex. The observed migration time is then used to establish the binding affinity. One study applied this method to determine that the binding site between an enzyme and inhibitor was at the metal center due to 4 orders of magnitude difference in the binding affinity after removing the metal ions (Zn²⁺) [84].

4.2.4. Screening for enzyme inhibitors

Enzyme inhibitors have potential as therapeutics and play an important role in disease treatment. The mechanism used to prevent cancer progression is to reduce the enzyme activity on the pathway that promotes cancer cell growth. In order to find a potent inhibitor, on-line analysis is generally adapted to screen the selected compounds, such as natural extracts [23,56,63,76] or traditional Chinese medicines [18,57,68,73]. In the process of screening the inhibition, the selected inhibitor is evaluated as the relative difference in the product peak area in the presence and in the absence of the inhibitor. Furthermore, statistical parameters [85], have been used to validate capillary electrophoresis of inhibitor screening [56,74,76] by evaluating the precision in the method based on mean response and error of the generated product in the presence and absence of inhibitor.

5. Enzyme immobilization techniques

5.1. Attributes of enzyme immobilization

Enzyme immobilization for the formation of enzyme reactors provides several advantages, such as added stability, minimal sample consumption, and reusability, for the development of screening assays. Immobilization of enzymes enables rapid on-line screening of substrates and inhibitors by creating a stationary zone of enzyme. Substrate, or in some instances a combination of substrate and inhibitor, is driven into the enzyme zone and converted to product prior to the separation. This is critical for developing assays that are high throughput [86,87] and can be automated [86] while still consuming small volumes of reagents for the assay. Immobilized enzymes allow for enzyme preparations to be reused [13] and may exhibit higher stability than the enzymes that are free in solution [83,87,88]. Immobilized enzyme reactors enable rapid analysis times through the use of high concentrations of enzymes in small volumes [89], which cannot be achieved with benchtop reactions due to practical reasons, such as limitations in manually handling nanoliter volumes. This enables the reaction to be independent of diffusion and substrate concentration.

5.1.1. Open tubular immobilization design and application

The advantages and drawbacks of each support for immobilization must be considered in order to develop immobilized enzyme reactors that find a balance between complexity of incorporating the enzyme reactor and enzyme density. Immobilized enzyme reactors based on an open tubular format are advantageous in that the approach does not require particle packing; however, the trade-off lies in the low surface area for enzyme immobilization leading to low enzyme density and the requirement for substrate to diffuse to the wall from the center of the capillary in order to be converted. The desired longevity of the immobilization is an important consideration when using the open tubular format.

Noncovalent coatings that employ electrostatic [22,76,77] interactions through the use of a cationic coating or hydrophobic [18] interactions through the use of a neutral coating, are simple to employ. However, they are less stable than coatings based on covalent immobilization to the silica wall, such as silanol [57] and glutaraldehyde chemistry [78]. For example, one report utilizing electrostatic interactions noted that high voltage can lead to enzyme desorption from the wall [22]. Despite this, immobilized enzyme reactors relying on electrostatic interactions have been reported to maintain up to 73% of the original enzyme activity after 50 assays [22]. A separate approach using graphene oxide to immobilize cationic trypsin enabled the incubation time for a trypsin digest to be reduced from 12 h for the free solution to 30 min with the immobilized enzyme [17].

5.1.2. Particle immobilization design and application

The use of particles for enzyme immobilization overcomes the limitation of substrate diffusion because of the higher enzyme densities in the capillary. An additional advantage is the ability to replace the capillary without repeating the immobilization procedure. Particle immobilization requires fabrication of the modified capillary, which is a difficult task that must be performed with skill since inconsistencies in preparations lead to poor reproducibility. Packed beds have a higher back pressure than wall-coated designs. Depending on the surface charge and packing homogeneity, the particles may interfere with the electroosmotic flow. Commercially available, porous particles provide a high surface area per volume leading to a high enzyme density in the reaction zone [13]. In wall-coated capillaries this has also been achieve using nanoparticles to increase the enzyme density at the wall [90]. For a packed capillary the particles are held in place using a frit on both sides of the reaction zone, which can be created using a single porous particle with a diameter slightly smaller than the inner diameter of the capillary [13]. However, physical frits such as this make the method susceptible to clogging. Magnetic particles circumvent the need for a frit by utilizing a magnetic field for immobilization [60,91,92]. This presents the additional advantage of being able to control the location of the particles and position the reaction zone in the thermostable region of the capillary [91]. Furthermore, when the procedure is optimized, this format is ideal for expensive or hard to isolate enzyme preparations since it allows the immobilized enzyme to be reused several times. One study reported the use of a single immobilized enzyme reactor for 100 consecutive runs [13].

5.1.3. Monolith immobilization design and application

Monoliths, which are continuous beds of a porous material, also offer high surface area, but since the monoliths are formed through polymerization reactions in-capillary, no packing is required and the material properties critical to the separation and immobilization, such as pore size, can be altered through modifications in the polymerization procedure. The procedure to make monoliths can be time consuming, however, with incubation steps lasting up to 24 h. Once a capillary is clogged, the immobilization procedure must be repeated. Several materials have been utilized for monoliths, including silica. In one example, fabrication of the silica monolith shown in Fig. 7 was achieved through polymerization of tetramethoxysilane [93]. The silica monolith was modified with 3-aminopropyl-trimethoxysilane to conjugate enzyme to the monolith via reductive amination. The synthesized capillary monolith bioreactor was coupled off-line with MALDI-TOF-MS and used for digestion of DNA [93]. Other materials have been utilized for monoliths, such as silica modified with graphene oxide [94], and polymers [86,95]. Avoiding protein adsorption is a driving force behind using materials, such as acrylamide. The desire to add a chiral selector may lead to the use of silica modified with graphene oxide [94]. A monolith-based immobilized enzyme reactor was recently employed to address the reaction time and labor required for deglycosylation of glycoproteins, a bottleneck in the workflow of glycoprotein analyses [86]. The report focused on the use of oriented surface chemistry to ensure the enzyme was more accessible to the substrate. While free solution PNGase F required hours for complete deglycosylation, the non-oriented immobilization reduced the required reaction time to minutes and oriented immobilized PNGase F further reduced the time to several seconds. This provided a technique capable of automating the glycoprotein workflow and greatly reducing the time required.

Fig. 7.

Scanning electron microscopic image of silica monolith. The image shows a rigid skeleton and even distribution of macropores in synthesized porus silica monolith within the capillary bioreactor. Reprinted with permission from C. Zhao, R. Yin, J. Yin, D. Zhang, H. Wang, Capillary Monolithic Bioreactor of Immobilized Snake Venom Phosphodiesterase for Mass Spectrometry Based Oligodeoxynucleotide Sequencing, Anal. Chem. 84(2) (2012) 1157–1164. Copyright 2012 American Chemical Society.

5.1.4. Microfluidic immobilization design and application

Microfluidic chips developed from poly(dimethylsiloxane) are similar to the open tubular format. The poly(dimethylsiloxane) already presents a surface for the enzyme to be immobilized on through hydrophobic interactions and hydrogen bonds, so no additional coating is required. Adsorption of the enzyme to the wall is not robust, however, which results in the enzyme desorbing from the wall. Without further modifications, microfluidic chips still result in a lower surface area than monoliths and particles, but the immobilization procedure requires short incubation times and since the material is inexpensive, microfluidic devices are an ideal substrate for disposable assays. Recently, a microfluidic device for trypsin digests was developed as an off-line immobilized enzyme reactor prior to capillary electrophoresis mass spectrometry analysis. The device, while only maintaining the original activity for 2 h, exhibited comparable results to the bench top reaction with only 50 s of reaction time required for the immobilized enzyme reaction [96].

6. Emerging techniques and future directions

6.1. Enzymatic capillary electrophoresis assays to identify substrate structure

6.1.1. Structural identification of glycans

Enzymatic reactions in combination with capillary electrophoresis analyses have been developed for the structural identification of oligosaccharides. Hydrolase enzymes are used to sequence biopolymers based on enzyme specificity for a specific motif of the terminal monomer. When the specificity of the enzyme matches the terminal monomer, the substrate is cleaved, resulting in a product of lower mass. This is observed as a shift in the substrate peak as compared to the separation performed in the absence of the enzyme. The use of multiple exoglycosidase has enabled the determination of terminal glycan identities [27,97]. Szigeti and Guttman successfully performed analyses using automated sample preparation prior to the capillary electrophoresis analysis [97]. Alternatively, on-line enzymatic reactions require nanoliter volumes of substrates and enzymes, and incubation times as low as 40 s have been reported [26]. Electroosmotic flow is typically suppressed through neutral coatings, such as poly(dimethylsiloxane) [98] or phospholipid [26,50], and background electrolytes with a higher viscosity than water. These viscosities were achieved through the use of hydroxypropylcellulose [98] or a phospholipid nanogel [26,50]. After optimization of several on-line reaction parameters, such as the length of the enzyme plug, enzyme concentration, and incubation time, the on-line sequencing of the terminal residues of complex N-glycans was achieved [26,50,98].

6.1.2. Structural determination integrated with mass spectrometry

Enzymatic protein digestion is fundamental to protein identification through tandem mass spectrometry with on-line protein databases. On-line enzymatic digestion of proteins or peptides followed by capillary electrophoresis is a powerful tool for proteomics research. The use of an on-line enzymatic reaction enables consumption of small volumes of the enzyme and substrates, significant reduction in reaction time, and reduces sample handling. The digested peptides from a proteolytic reaction are complex. Direct infusion of a complex digested product would lead to data that is difficult to interpret. Peptides with post translational modification (i.e. glycopeptides or phosphopeptides) are usually suppressed by other non-modified peptides and are difficult to detect. Incorporating capillary electrophoresis with mass spectrometry allows the separation of the enzymatic digested product and alleviates competitive ionization between peptides. As a result, this approach improves sequence recovery and facilitates protein identification.

The coupling of capillary electrophoresis and mass spectrometry requires a well-designed interface, which maintains a complete electrical circuit during electrophoresis for the separation, generates a stable electrospray, and minimizes dilution of the analyte peak. These conditions must be met so that the excellent separation resolution from capillary electrophoresis and high detection sensitivity from mass spectrometry are preserved. A recently reported design interface based on an electrokinetic pump sheath-flow nanospray was successfully used on-line [99]. The capillary electrophoresis separation current is decoupled from the electrospray current through the sheath liquid reservoir [89,95,99,100], as seen in Fig. 8A. The design minimizes dilution associated with a sheath liquid by using a taper glass nanospray emitter.

Fig. 8.

Schematic of capillary zone electrophoresis coupled with mass spectrometry showing on-line protein digests. The schematic in A is of strong cation exchange based monolith based microreactor coupled with capillary zone electrophoresis-mass spectrometry. The mass spectrum in B is of an on-line analysis of Xenopus laevis protein digests achieved by the set up used in A. Reprinted with permission from Z. Zhang, L. Sun, G. Zhu, O.F. Cox, P.W. Huber, N.J. Dovichi, Nearly 1000 Protein Identifications from 50 ng of Xenopus laevis Zygote Homogenate Using Online Sample Preparation on a Strong Cation Exchange Monolith Based Microreactor Coupled with Capillary Zone Electrophoresis, Analytical Chemistry 88(1) (2016) 877–882. Copyright 2016 American Chemical Society.

An on-line enzyme reactor successfully integrated with capillary electrophoresis and mass spectrometry utilized a solid support for enzyme [89,95] or protein [100] entrapment. The device is based on a monolithic microreactor positioned in-capillary that is made of acrylamide [89,95] or sulfonated-silica [100]. The polyacrylamide-based microreactor proved to be a compatible platform for enzyme immobilization while maintaining enzyme activity [89,95]. These microreactors were employed for two different enzymes. Immobilized alkaline phosphatase facilitated the determination of protein phosphorylation [95]. A trypsin monolith was developed for on-line protein digestion in a cell lysate. The integrated system was compatible with a 300 pg sample size. The in-line monolith microreactor decreased incubation time to minutes [89] as compare to off-line protein digestion, which required 24 h when performed in free solution [100].

This concept was expanded by using the in-capillary monolith to trap the targeted proteins rather than the trypsin [100]. The approach minimized the deactivation of the trypsin activity through immobilization to the capillary surface while preserving the benefits of on-line digestion. Additionally, the amount of digested proteins was not limited by the length of the immobilized enzymes zone in the capillary. This monolithic material contained sulfonate groups to facilitate electrostatic trapping and exhibited high surface area. Under acidic conditions, proteins were easily extracted from the crude sample and subject to multiple processing steps. The microreactor was used to reduce, alkylate, and digest entrapped proteins. Immobilizing the protein enables the use of larger concentrations of trypsin. Digestion of proteins derived from a cell lysate was completed in 10 min. The excellent efficiency of the capillary electrophoresis was well-suited to separate the complex peptide digest as shown in Fig. 8B. The median separation efficiency of the peaks in the electropherogram was 240,000 theoretical plates. Using a 100 cm long capillary and a Q executive HF mass spectrometer, 975 proteins were identified from Xenopus laevis zygote homogenate.

6.2. Future directions

Capillary electrophoresis is an enabling tool and will continue to evolve and adapt to meet the challenges of enzyme analyses. Innovative applications, such as enzyme characterization in whole cells [44], will progress to routine screening technology for disease diagnosis and drug discovery. Capillary electrophoresis approaches have even been developed for enzymatic analyses of single cells [39] using peptide reporters in order to shed light on the impact of environmental toxins on enzymes. Newly reported strategies that clarify the suitability of reporter molecules across species [101] may lead to more designer reporters to interrogate metabolism at the cellular level. Finally, recent reports of multiplexed enzyme analyses with capillary electrophoresis [20,24] point to new applications to assess larger libraries of enzymes and substrates. As multi-capillary commercial systems, previously designed for DNA analyses, are adapted to other separation-based assays [102,103], an unprecedented increase in the throughput of enzyme screening will be a reality in the very near future.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.ymeth.2018.02.005.