Electrophoresis

The rate of migration of ions in an electrical field depends on the factors listed in Box 11.1. The equation expressing the driving force in such a system is given by

F=(X)(Q)=(EMF)(Q)d

(11.1)

where F = the force exerted on an ion; X = the current field strength (V/cm) (i.e., voltage drop per unit length of medium); Q = the net charge on the ion; EMF = the electromotive force [voltage (V) applied]; and d = the length of the electrophoretic medium (cm).

Steady acceleration of the migrating ion is counteracted by a resisting force characteristic of the solution in which migration occurs. This force, expressed by Stokes’ law, is

F′=6πrηv

(11.2)

where F′ = the counter force; π = 3.1416; r = the ionic radius of the solute; η = the viscosity of the buffer solution in which migration is occurring; and ν = the rate of migration of the solute = velocity (length [l] traveled per unit of time [cm/s]).

The force F′ counteracts the acceleration that would be produced by F if no counter force were present, and the result of the two forces is a constant velocity. Therefore, when

F=F′

(11.3)

then

6πrη=(X)(Q)

(11.4)

v/X=1×d/t×E=Q/6πrη=μ

(11.5)

where ν/X is the rate of migration (cm/s) per unit field strength (E/cm), defined as the electrophoretic mobility and expressed by the symbol μ.

Electrophoretic mobility is directly proportional to the net charge and is inversely proportional to the size of the molecule and the viscosity of the electrophoresis medium. Mobility may be positive or negative, depending on whether a protein migrates in the same or the opposite direction as the electrophoretic field (defined as extending from the anode to the cathode).

In addition to the factors listed in Box 11.1, other factors that affect electrophoretic mobility include electroendosmosis (endosmosis) and wick flow. Electroendosmosis affects mobility by causing uneven movement of water through the support medium. An electrophoretic support medium, such as a gel in contact with water, takes on a negative charge caused by the adsorption of hydroxyl ions. These ions are fixed to the surface and are immobile. Positive ions in solution cluster about the fixed negative charge sites, forming an ionic cloud of mostly positive ions. The number of negative ions in the solution increases with increasing distance from the fixed negative charge sites until eventually positive and negative ions are present in equal concentrations (Fig. 11.1).

BOX 11.1 Factors Affecting the Motility of Ions in an Electrophoretic System

Net charge of the molecules
Size and shape of the molecules
Strength of the electrical field
Support medium properties
Ionic strength of the buffer
Temperature

FIG. 11.1 Distribution of + and − ions around the surface of an electrophoretic support. Fixed on the surface of the solid is a layer of − ions. A second layer of + ions is attracted to the surface. These two layers compose the Stern potential. The large, diffuse layer containing mostly + ions is the electrokinetic or zeta potential. Extending farther from the surface of the solid is homogeneous solution. The Stern potential plus the zeta potential equals the electrochemical potential, or epsilon potential.

When current is applied to such a system, charges attached to the immobile support remain fixed, but the cloud of ions in solution moves to the electrode of opposite polarity. Because ions in solution are highly hydrated, this results in movement of the solvent as well. Movement of solvent and its solutes relative to the fixed support is referred to as endosmosis and causes preferential movement of water in one direction. Macromolecules in solution that move in the direction opposite this flow may remain immobile or may even be swept back toward the opposite pole if they are insufficiently charged. In media in which endosmosis is strong, such as conventional cellulose acetate and unpurified agarose gel, γ-globulins are swept back from the application point. Because the inner surface of a glass capillary contains many such charged groups, endosmosis is very strong and is actually the primary driving force for migration in CE systems. However, the endosmosis in CE can be manipulated to modify the magnitude of the endosmotic effect. In electrophoretic media in which surface charges are minimal (starch gel, purified agarose gel, or polyacrylamide gel), endosmosis is minimal.

Wick flow results from the movement of buffer into the support medium. During electrophoresis, heat evolved because of the passage of current through a resistive medium can cause evaporation of solvent from the electrophoretic support. This drying effect draws buffer into the support, and, if significant, the flow of buffer can affect protein migration and hence the calculated mobility.

Clinical Electrophoresis

In this section, focus will be on the electrophoresis methodology that is frequently used in the clinical laboratory. Refer to Chapter 18 for a thorough discussion on the various electrophoretic fractions present in clinical protein electrophoresis.

Suggested Wavelengths for Quantitation of Protein Zones by Direct Densitometry
Separation Type	Stain	Nominal Wavelength, nm
Serum proteins in general	Amido Black (Naphthol Blue Black)	640
	Coomassie Brilliant Blue G–250 (Brilliant Blue G)	595
	Coomassie Brilliant Blue R–250 (Brilliant Blue R)	560
	Ponceau S	520
Isoenzymes	Nitrotetrazolium Blue	570
Lipoprotein zones	Fat Red 7B (Sudan Red 7B)	540
	Oil Red O	520
	Sudan Black B	600
DNA fragments	Ethidium bromide (fluorescent)	254 (Ex) 590 (Em)
CSF proteins	Silver nitrate	—

Staining

If staining is used to visualize separated proteins, the proteins usually are fixed first by precipitating them in the gel with a chemical agent such as acetic acid or methanol. This prevents diffusion of proteins out of the gel when submersed in the stain solution. The amount of dye taken up by the sample is affected by many factors, such as the type of protein and the degree of denaturation of the proteins by fixing agents.

Table 11.1 lists dyes commonly used in electrophoresis, along with suggested wavelengths for quantification by densitometry. Most commercial methods for serum protein electrophoresis use Amido Black B or members of the Coomassie Brilliant Blue series of dyes for staining. Isoenzymes are typically visualized by incubating the gel in contact with a solution of substrate, which is linked structurally or chemically to a dye before fixing. Silver nitrate stains proteins and polypeptides with sensitivity 10- to 100-fold greater than that of conventional dyes. Selective fixing and staining of protein subclasses can also be achieved by combining a stain molecule with an antiglobulin, as is done in IFE.

Improvements in conducting sensitive measurements have been achieved by linking an enzyme, such as alkaline phosphatase or peroxidase, to a single or double antibody specific for particular proteins, such as oligoclonal immunoglobulin (Ig), or by spraying separated proteins with luminal and peroxide to develop chemiluminescence, which, in turn, exposes x-ray film to form a permanent image. Chemiluminescence has been used in this way to quantify IgE, and DNA fragments have been detected by linking with a fluorescent dye label.

In practice, a typical stain solution may be used several times before it is replaced. A good rule of thumb is that a stain solution of 100 mL may be used for a combined total of 387 cm² (60 in²) of agarose film. The stain solution may be considered faulty if leaching of stained protein zones occurs in the 5% acetic acid wash solution. Whenever protein zones appear too lightly stained, the stain or substrate reagent—in the case of isoenzymes—should always be suspected. Stain solution must be stored tightly covered to prevent evaporation.

Quantification

A densitometer is used to quantify stained zones. This instrument measures the absorbance of each fraction as the gel (or other medium) is moved past a photometric optical system and displays an electropherogram on a computer display. The software is able to automatically integrate the area under each peak and report each as a percent of total or as absolute concentration or activity computed from the total protein or activity of enzyme in the sample.

Reliable densitometric quantification requires: (1) light of an appropriate wavelength; (2) linear response from the instrument; and (3) a transparent background in the medium being scanned. Linearity may be tested with a neutral density filter designed with separated or adjacent areas of linearly increasing density. The densities are permanent and have expected absorbance values. The very small sample sizes used and the transparency of agarose gels satisfy the requirement for a clear background. Nevertheless, problems can occur with densitometry because of differences in the quantity of stain taken up by individual proteins and differences in protein zone sizes.

Essential features of a densitometer include: (1) the ability to scan gels 25 to 100 mm in length; (2) electronic adjustment of the most intense peak to full scale; (3) automatic background zeroing (peaks are not lost or “cut off”); (4) variable wavelength control over the range of 400 to 700 nm; (5) variable slits to allow adjustment of the beam size; (6) an integrating device with both automatic and manual selection of cut points between peaks; and (7) automatic indexing, a feature that advances the electrophoresis strip from one sample channel to the next.

Desirable features of a densitometer include computerized integration and printout, built-in diagnostics for instrument troubleshooting, the choice of one of several scanning speeds, and the ability to measure in the reflectance mode. Models with a separate computer for data processing permit storage and reformatting of data, if desired, and reprinting or delayed transmission to a host computer.

DNA analysis requires the ability to scan larger gels, which may contain several dozen bands of DNA fragments of different length. Modern automated electrophoresis systems also use larger gels containing 30 or more samples, which are scanned on a new generation of densitometers referred to as flatbed scanners or digital image analyzers. These instruments are capable of scanning and storing digitized light intensity readings from large areas and use ultrasensitive charge-coupled device detectors having a resolution of up to 1200 dots per inch (21 μm). Sophisticated data processing software permits manipulation of stored image information to produce conventional scans and computations or more complex outputs, such as overlaying and subtraction of patterns from two different samples.

FIG. 11.2 A schematic diagram of a typical electrophoresis apparatus showing two buffer boxes with: (1) baffle plates, (2) electrodes, (3) electrophoretic support, (4) wicks, (5) cover, and power supply.

Blotting Techniques

In 1975, Edward Southern developed a technique that is widely used to detect fragments of DNA. This technique, known as Southern blotting, first requires electrophoretic separation of DNA or DNA fragments by agarose gel electrophoresis (AGE). Next, a strip of nitrocellulose or a nylon membrane is laid over the agarose gel, and the DNA fragments are transferred or “blotted” onto it by capillary, electro, or vacuum blotting. They are then detected and identified by hybridization with a labeled, complementary nucleic acid probe. This technique is widely used in molecular biology for identifying a particular DNA sequence; determining the presence, position, and number of copies of a gene in a genome; and typing DNA.

Northern and Western blotting techniques, named by analogy to Southern blotting, were subsequently developed to separate and detect ribonucleic acids (RNAs) and proteins, respectively. Northern blotting is carried out identically to Southern blotting except that RNA species are separated by electrophoresis. Western blotting is used to separate, detect, and identify one or more proteins in a complex mixture. It involves first separating the individual proteins by polyacrylamide gel and then transferring or “blotting” onto an overlying strip of nitrocellulose or a nylon membrane by electro-blotting. The strip or membrane is then reacted with a reagent that contains an antibody raised against the protein of interest.

Instrumentation

Although modern electrophoresis equipment and systems vary considerably in form and degree of automation, the essential components common to all systems (Fig. 11.2) include: (1) two reservoirs that contain the buffer used in the process, (2) a means of delivering current from a power supply via platinum or carbon electrodes contacting the buffer, and (3) a support medium in which separation takes place connecting the two reservoirs. In some systems, wicks may connect the medium to the buffer solution or directly to the electrodes. The entire apparatus is enclosed to minimize evaporation and protect both the system and the operator. The direct current power supply sets the polarity of the electrodes and delivers current to the medium.

Power Supplies

The power supply drives the movement of ionic species in the medium and allows adjustment and control of the current or the voltage. With more sophisticated units, the power may be controlled as well, and conditions may be programmed to change during electrophoresis. Capillary systems use power supplies capable of providing voltages in the kilovolt range.

Current flowing through a medium that has resistance produces heat:

Heat=(E)(I)(t)

(11.6)

where E = electromotive force (EMF) in volts (V); I = current in amperes (A); and t = time in seconds (s).

This heat is released into the medium and increases the thermal agitation of all dissolved ions, and therefore the conductance of the system (decreases resistance). With constant-voltage power supplies, the resultant rise in current increases both protein migration and evaporation of water from the medium. Any water loss increases the ion concentration and further decreases the resistance (R). Under these circumstances, the current and therefore the migration rate will progressively increase. To minimize these effects, it is best to use a constant-current power supply. According to Ohm’s law,

E=(I)(R)

(11.7)

therefore if R decreases, the applied EMF also decreases, keeping the current constant. This in turn decreases the heat effect and stabilizes the migration rate.

Buffers

Buffer ions have a twofold purpose in electrophoresis: they carry the applied current, and they fix the pH at which electrophoresis is carried out. Thus they determine: (1) the type of electrical charge on the solute, (2) the extent of ionization of the solute, and therefore (3) the electrode toward which the solute will migrate. The buffer’s ionic strength determines the thickness of the ionic cloud (buffer and non-buffer ions) surrounding a charged molecule, the rate of its migration, and the sharpness of the electrophoretic zones. With increasing concentration of ions, the ionic cloud increases in size, and the molecule becomes more hindered in its movement.

According to Joule’s law, power produced when current flows through a resistive medium is dissipated as heat. This heat increases in direct proportion to the resistance, but also in proportion to the square of the current. The reduction in resistance caused by a high ionic strength buffer therefore leads to increased current and excessive heat. These buffers yield sharper band separations, but the benefits of sharper resolution are diminished by the Joule (heat) effect that leads to denaturation of heat-labile proteins or the degradation of other components.

Ionic strength (also denoted by the symbol μ) is computed according to the following:

μ=0.5∑cizi2

(11.8)

where c_i = ion concentration in mol/L and z_i = the charge on the ion.

The ionic strength μ of an electrolyte (buffer) composed of monovalent ions is equal to its molarity (mol/L). The ionic strength of a 1 mol/L electrolyte solution with one monovalent and one divalent ion is 3 mol/L, and for a doubly divalent electrolyte, it is 4 mol/L.

A relatively high ionic strength buffer used in high-resolution electrophoresis improves the separation of serum proteins into as many as 13 bands, with two or more bands in the α₁, α₂, and β-globulin regions and one or more additional bands seen in various pathologic conditions. Because of higher conductivity and the associated heat produced, it is necessary to reduce the temperature of the system to 10°C to 14°C. “Submarine” techniques, in which gels are submersed in circulating buffer cooled by an external cooling device, or are supported on an electrophoresis chamber cooled by circulating water or an integral Peltier plate, provide exact temperature control. Effective cooling with less precise temperature control may also be achieved using chambers designed with a sealed compartment of cooled ethylene glycol, which is in contact with the gel during running.

Because buffers used in electrophoresis are good culture media for the growth of microorganisms, they should be refrigerated when not in use. Moreover, a cold buffer is preferred in an electrophoretic run, because it improves resolution and decreases evaporation from the electrophoretic support. Buffer used in a small-volume apparatus should be discarded after each run because of pH changes resulting from the electrolysis of water that accompany electrophoresis. If volumes used are larger than 100 mL, buffer from both reservoirs may be combined, mixed, and reused up to four times.

Support Media

The support medium provides the matrix in which protein separation takes place. Various types of support media have been used in electrophoresis and range from pure buffer solutions in a capillary to insoluble gels (e.g., sheets, slabs, or columns of starch, agarose, or polyacrylamide) or membranes of cellulose acetate. Gels are cast in a solution of the same buffer to be used in the procedure and may electrophorese in a horizontal or vertical direction. In either case, maximum resolution is achieved if the sample is applied in a very fine starting zone. Separation is based on differences in charge-to-mass ratio of the proteins and, depending on the pore size of the medium, possibly molecular size.

Cellulose Acetate

Cellulose acetate, a thermoplastic resin made by treating cellulose with acetic anhydride to acetylate the hydroxyl groups, is primarily of historical interest. When dry, the membranes contain about 80% air space within the interlocking cellulose acetate fibers and are opaque, brittle films. As the film is soaked in buffer, the air spaces fill with liquid, and it becomes pliable. Samples are applied with a twin-wire applicator or the edge of a glass slide. Because of their opacity, stained membranes need to be made transparent (cleared) for densitometry by soaking in 95:5 methanol:glacial acetic acid. Cleared membranes are strong and could be stored as a permanent record, but because of the necessity for presoaking and clearing, cellulose acetate has largely been replaced by agarose in most clinical applications.

Agarose

Agarose is a linear polymer containing alternating D-galactose and 3,6-anhydro-L-galactose monomers. It is the purified, essentially neutral fraction of agar obtained by separating agarose from agaropectin, a more highly charged fraction containing acidic sulfate and carboxylic side groups. Because the pore size in agarose gels is large enough for all proteins to pass through unimpeded, separation is based only on the charge-to-mass ratio of the protein. Advantages of agarose gel include its low affinity for proteins and its native clarity after drying, which permits excellent densitometry. It is essentially free of ionizable groups and so exhibits little endosmosis.

Most routine procedures for AGE are now performed using commercially produced, prepackaged microzone gels, and the sample is applied by means of a comb or a thin plastic template, with small slots corresponding to sample application points. The template is placed on the agarose surface, and 5 to 7 μL samples are placed on each slot. The serum sample is allowed to diffuse into the agarose for 5 minutes, the excess sample is removed by blotting, and the template is removed. AGE separation for most routine serum applications requires an electrophoresis time of 20 to 30 minutes.

Polyacrylamide Gel

Polyacrylamide is a polymeric matrix consisting of linear chains of acrylamide cross-linked with bis-acrylamide. It is thermostable, transparent, strong, and relatively chemically inert, and—depending on concentration—can be made in a wide range of pore sizes. Its average pore size in a typical 7.5% gel is about 5 nm (50 Å), which is large enough to allow most serum proteins to migrate unimpeded. However, proteins with a molecular radius and/or length that exceeds critical limits will be impeded in their migration. Some of these proteins are fibrinogen, β-lipoprotein, α₂-macroglobulin, and γ-globulins; a schematic representation of serum protein electrophoresis by polyacrylamide gel electrophoresis (PAGE) is shown in Fig. 11.3. The separation is based on both charge-to-mass ratio and molecular size (a phenomenon referred to as molecular sieving), and serum proteins can be resolved into more individual fractions than with an agarose gel. Furthermore, these gels are uncharged, thus eliminating electroendosmosis. Precast minigels are available in a variety of concentrations and acrylamide-to-bis-acrylamide ratios suitable for most protein or nucleic acid separations. However, because of the known neurotoxicity of acrylamide, appropriate caution must be exercised when handling this material if the gels are prepared by hand.

FIG. 11.3 A simplified schematic drawing of a protein pattern from the serum of a subject with haptoglobin type 2-1 (separation by polyacrylamide gel electrophoresis). Some zones contain more than the one protein shown, as demonstrated by immunologic techniques. AAT, α₁-Antitrypsin; ALB, albumin; AMG, α₂-macroglobulin; BLP, β-lipoprotein; C3, complement 3; FIB, fibrinogen; Gamma, γ-globulin; HP, haptoglobin; TRF, transferrin.

Attempts to improve the hydrophilic nature of polyacrylamide have led to the development of mono- and di-substituted monomers, one of which is N-acryloyl-tris(hydroxymethyl)aminomethane, or Poly(NAT). This material is more hydrophilic than polyacrylamide, and its matrix has larger pores, thereby presenting less resistance to the passage of large molecules. It is ideally suited to the separation of DNA fragments up to 20 kilobases (kb) in size using a homogeneous (nonpulsed) electric field. Fragments that differ in size by as little as 2% can be resolved. Gels are submersed in buffer during use, allowing temperatures to be tightly controlled at values between 50°C and 60°C. Use of increased temperatures results in shorter run times and more reproducible band migration.

Automated Systems

Because of increased volume of testing, primarily for serum proteins, many laboratories are converting to automated systems for electrophoresis. Such a system is the Helena SPIFE 4000 (Helena Laboratories, Beaumont, TX), an automated electrophoresis system providing automated reagent application and a variety of gel sizes that permit analysis of 10 to 100 samples simultaneously. It also features in-line sample application, automated electrophoretic separation and staining of analytes, multiple stain ports, and positive sample identification. The Interlab Microgel system (Interlab Srl, Rome, Italy) also fully automates the process and integrates sample application, temperature-controlled electrophoresis, staining, and densitometry into a single unit with the capability of managing four gels simultaneously. Other systems that have partially automated the procedure or incorporated the ability to process sequentially multiple gels of different compositions include the Phast System (Pharmacia LKB, Gaithersburg, MD), the HITE Fractoscan (Olympus, Invicon, München, Germany), the Hydragel-Hydrasys (Sebia Inc., Durham, NC), and the High-Performance Gel Electrophoresis (HPGE)-1000 system (LabIntelligence Inc., Belmont, CA). Most CE systems (see “Capillary Electrophoresis” section) have autosampling capability for sequentially processing specimens, but the Sebia CAPILLARYS permits simultaneous processing of seven samples by using multiple capillaries. Newer microchip-based analyzers like the Agilent 2100 Bioanalyzer (Agilent Technologies Inc., Santa Clara, CA) significantly miniaturize and increase the speed of the process for separating proteins, nucleic acids, or even entire cells. These advances substantially reduce the labor component associated with electrophoresis.

Points To Remember

Capillary Zone Electrophoresis

• Uses electroosmotic flow (EOF) to mobilize the sample plug pass the detector
• The magnitude and direction of EOF can be manipulated through methodology
• Higher voltages can be applied due to more efficient dissipation of heat; this can translate to faster separations

Capillary Electrophoresis

With CE, classic techniques of are carried out in a small-bore (10- to 100-μm internal diameter) fused silica capillary tube, 20 to 200 cm in length.

Two distinct advantages of the capillary format include the ability to apply much higher voltages than in traditional electrophoresis and the ease of automation. Applications are also more extensive and include the separation of low molecular weight ions in addition to proteins and other macromolecules. Even uncharged molecules can be separated using CE in the micellar electrokinetic chromatography (MEKC) mode discussed later. CE has also proved useful for separation of inorganic ions, amino acids, organic acids, drugs, vitamins, porphyrins, carbohydrates, oligonucleotides, proteins, and DNA fragments.

General Operations

A schematic diagram of a typical instrumental configuration for CE is shown in Fig. 11.4. As indicated, the capillary serves as an electrophoretic chamber, analogous to a lane on a gel, which is connected to buffer reservoirs at both ends, which, in turn, are connected to a high-voltage power supply. It is important to note that at some point along the length of the capillary (typically close to the end), a detector is interfaced for online detection. Improved heat dissipation from the capillary (as opposed to a slab gel) permits the application of voltages in the range of 10 to 30 kV, which enhances separation efficiency and reduces separation time—in some cases to less than 1 minute. Only a few microliters of the sample are required, with injected volumes in the nanoliter range. The small sample plug volume minimizes distortions in the applied field caused by the presence of analytes or other sample species.

In contrast to the cumbersome and time-consuming tasks of conventional electrophoresis, CE is easily automated. Analogous to high-performance liquid chromatography (HPLC) technology, samples typically are stored in a temperature-controlled environment and are automatically injected into the capillary, with a variety of detector types available; the resulting electropherograms are analyzed and manipulated in much the same manner as chromatograms.

FIG. 11.4 A schematic for capillary electrophoresis instrumentation.

The capillaries used as separation columns in CE are most commonly made from fused silica (i.e., pure glass) coated with a thin exterior covering of polyimide to provide strength and flexibility. Although capillaries can be made from other materials, such as polyethylene or Teflon, such capillaries have seen limited use. The polyimide coating is usually removed from a small portion of the capillary close to the terminal end, creating a window for online optical detection. The outer diameter of the capillary tubing typically varies from 180 to 375 μm, the inner diameter from 20 to 180 μm, and the total length from 20 cm up to several meters. Noncylindrical capillary tubing suitable for CE is now available from some commercial providers. For example, rectangular capillaries (Polymicro Technologies, Phoenix, AZ) provide a flat surface that is more amenable to optical detection than that of their curved counterparts.

CE is distinguished from other forms of electrophoresis by the fact that extraordinarily high voltage (30,000 volts) has been used to obtain rapid, high-efficiency separations. The problems encountered with noncapillary platforms (e.g., slab gels) are avoided because of the effective dissipation of Joule heat by forced air convection or liquid cooling of the narrow-bore capillary.

Sample Injection

In CE, sample volumes of 1 to 50 nL are loaded into the capillary by hydrodynamic injection or electrokinetic (EK) injection. With hydrodynamic injection, an aliquot of a sample is introduced by applying positive pressure at the inlet vial or vacuum at the outlet vial. The volume of sample loaded is governed by a number of parameters, including (but not restricted to): (1) the inner diameter of the capillary; (2) buffer viscosity; (3) applied pressure; (4) temperature; and (5) time. With some earlier commercial or homemade systems, gravity was used as the source of pressure by raising the inlet vial (or lowering the outlet vial), thus allowing “siphoning” to occur for a timed interval. With EK injection, an aliquot of a sample is introduced by applying a voltage for a timed interval. The magnitude of the voltage is dependent on the analyte and buffer system used but typically involves field strengths three to five times lower than that used for separation. It is important to note that although hydrodynamic methods introduce a sample representative of the bulk specimen, EK injection favors the preferential movement of more electrokinetically mobile analytes into the capillary.

In practice, to maintain high separation efficiency, the sample plug length is usually less than 2% of the total capillary length.

Direct Detection

With CE, separated analytes are detected and measured as they migrate past a point on the capillary that is optically interrogated. Optical detection is based on classical methods, such as photometric absorbance, refractive index, and fluorescence (see Chapter 9). As with HPLC, ultraviolet-visible photometers are widely used as detectors to monitor CE separations. To interface such online detectors with the capillary, a detection window is created toward the outlet end of the capillary. This “window,” which serves as an inline cuvette, typically is formed by burning off the polyimide with a small flame and cleaning the window with ethanol. Although this configuration allows high-efficiency separation, the inner diameter of the capillary tube defines the optical path length (OPL) of the inline cuvette. Because absorbance is directly proportional to the length of the cuvette used in an optical system, the 20 to 100 μm inner diameter of the capillary limits UV-visible absorbance detection limits to concentrations of 10⁻⁸ to 10⁻⁶ molar.

More sensitive optical techniques that have been used with CE include: (1) fluorescence; (2) refractive index; (3) chemiluminescence; (4) Raman spectrophotometry; and (5) circular dichroism. The most sensitive optical detection method used in CE is laser-induced fluorescence, which is capable of detection limits in the 10⁻⁹ to 10⁻¹² molar (or better) range. This detection mode is easily accomplished with analytes that may be easily labeled with a fluorescent substrate (e.g., intercalators for double-stranded DNA) or may be naturally fluorescent (e.g., proteins or peptides containing tryptophan). CE systems have also been interfaced with mass spectrometers, and electrochemical detection methods have been developed, although such detectors must be isolated electrically from the electrophoretic voltages.

Indirect Detection

When strong chromophores are lacking in the analyte of interest, absorbance and fluorescence detection have been used in an indirect mode. In this mode, a strongly absorbing ion is added to the running electrolyte and is monitored at a wavelength that gives a constant, high background absorbance. As solute ions move into their discrete zones during the electrophoretic process, they displace the indirect detection agent through mutual repulsion, and this produces a decrease in background absorbance as the zone passes through the detector. Reagents with appropriate fluorescence properties have been used in a similar manner. Indirect detection of amino acids by CE has been demonstrated, with the potential for use in the diagnosis of amino acidurias. Investigators have demonstrated the direct extrapolation of this technique to microchip detection when UV detection is difficult, if not impossible.

Types of Electrophoresis

Capillary Zone Electrophoresis

CZE, also called open-tube or free-solution CE, is the simplest form of CE. It includes capillary ion electrophoresis, which refers to the analysis of inorganic ions by CZE, often using indirect detection. The power of the CZE mode is its ability to electrophoretically resolve charged species without a sieving matrix; this applies to a broad spectrum of analytes ranging from proteins, peptides, and amino acids to small molecules (e.g., drugs) and ions.

Capillary Gel Electrophoresis

Capillary gel electrophoresis (CGE) is directly comparable with traditional slab or tube gel electrophoresis because the separation mechanisms are identical. Size separation is achieved with a suitable polymer, which acts as a molecular sieve or sizing mechanism. As charged analytes migrate through the polymer network, they become hindered to a degree that is governed by their size (larger molecules are hindered more than smaller ones). Macromolecules, such as DNA and sodium dodecyl sulfate (SDS)-saturated proteins, cannot be separated without a gel or some other separation mechanism, because they have a mass-to-charge ratio that is size independent. The term gel in CGE is a misnomer, primarily because cross-linked “gels,” as we know them in slab format, are not routinely used in CE. A more suitable term is a sieving matrix or soluble polymer network, a linear polymeric structure that is soluble, has reasonably low viscosity, and is capable of self-entangling in a manner that forms pores through which sieving can occur. A variety of polymeric matrices have been defined for DNA (e.g., polyacrylamide, cellulosic materials) and protein analysis (e.g., dextran-base matrices), provided that pores can be formed inherently that have diameters in the range of tens to hundreds of nanometers. One of the requirements that often accompany this type of analysis is reduction of electro-osmotic flow. This is accomplished by covalently, adsorptively, or dynamically coating the surface. Cross-linked polyacrylamide was the main polymer of choice for this but recently has been supplanted by a host of polymeric matrices that not only provide effective molecular sieving but also adsorptively coat the capillary surface.

Technical Considerations

Gel

In performing electrophoretic separations, a number of technical and practical aspects need to be considered, as they affect the process.

Sampling

To achieve a proper balance between sensitive measurements and resolution, the amount of serum protein applied to an electrophoretic support must be optimum. Albumin is about 10 times more concentrated in serum than the smallest fraction, the α₁-globulins. Therefore, the amount of serum applied should prevent overloading with albumin, but should still be adequate to quantify α₁-globulin. For the separation of serum proteins using PAGE, 3 μL of serum containing approximately 210 μg of total protein is applied. For alkaline phosphatase isoenzymes, up to 25 μL of a normal serum may be applied (less may be used if activity is greatly increased). Urine specimens require 50- to 100-fold greater concentrations or extended application times for adequate sensitivity, and CSF may or may not require concentration, depending on the staining approach used.

Discontinuities in Sample Application

Discontinuities in sample application may be caused by: (1) dirty applicators; (2) uneven absorption by sample combs; or (3) inclusion of an air bubble if the sample is pipetted onto the gel. The pipette tip should be checked for air bubbles before the sample is applied to the agarose gel template.

Unequal Migration Rates

Unequal migration of samples across the width of the gel may be caused by dirty electrodes, which may cause uneven application of the electric field, or by uneven wetting of the gel. If wicks are used to connect the gel to a power supply, uneven wetting of the wicks could cause unequal migration or bowing of sample lanes at the gel edges. Gels must be kept horizontal during storage to avoid sagging and uneven thickness. Finally, gels that may have been stored too close to heat sources (e.g., in a cabinet over a light fixture) could have partially and unevenly dried areas, contributing to similar problems.

Distorted, Unusual, or Atypical Bands

Distorted protein zones may be caused by: (1) bent applicators; (2) incorporation of an air bubble during sample application; (3) over-application; or (4) inadequate blotting of the sample. Excessive drying of the electrophoretic support before or during electrophoresis may also cause distorted zones. Irregularities (other than broken zones) in the sample application probably are due to excessively wet agarose gels. Portions of applied samples may look washed out.

In most cases, unusual bands are artifacts that may be easily recognized. Hemolyzed samples are frequent causes of increased β-globulin (where free hemoglobin migrates) or an unusual band between the α₂- and β-globulins, the result of a hemoglobin–haptoglobin complex. A band occurring at the starting point of an electropherogram may be fibrinogen. The sample should be verified as being serum before this band is reported as an abnormal protein. The α- and β-lipoproteins may migrate ahead of their normal positions in some samples. Occasionally, a split albumin zone is observed in the rare, benign, genetically related condition of bis-albuminemia. However, a grossly widened albumin zone could be due to albumin-bound medication and not to faulty electrophoresis.

Capillary

Temperature and surface effects influence the separation capabilities of CE. Artifacts also have been known to arise with CE.

Temperature Effects

In most slab or tube platforms for electrophoresis, moderate electric fields (up to 1000 volts) are used, because the Joule heating that accompanies the use of higher field strengths causes nonuniform temperature gradients, local changes in viscosity, and subsequent zone broadening. CE is distinguished from other forms of electrophoresis by the fact that extraordinarily high fields (30,000 volts) are used to obtain rapid, high-efficiency separations. The problems encountered with noncapillary platforms are prevented by effective dissipation of Joule heat by forced air convection or liquid cooling of the capillary, both of which are possible because of the narrow bore of the capillary. The Joule heat produced is a function of (1) buffer type, (2) concentration, (3) voltage applied, (4) capillary inner diameter, and (5) length, and can be determined for any given system by generating an Ohm’s law plot, which allows easy determination of the maximum voltage that can be used effectively. Reducing the inner diameter of the capillary, the ionic strength of the running buffer, or the applied voltage will reduce the heat produced by the electrophoretic process. It should be noted that reducing the inner diameter will compromise the detection limit of UV measurements (smaller OPL); reducing the applied field is less desirable in that resolution is directly proportional to the applied field. Consequently, attempts should be made to alter other parameters before reducing inner diameter or the applied field.

Surface Effects

As in electrophoresis in general, the flow of fluid (electro-osmotic or electroendosmotic flow [EOF]) in CE is a consequence of surface charge on the solid support. In CE, EOF can play a significant role in the separation process. The charge on the inner surface of a fused silica capillary is determined by the ionization state of the silanol groups (SiOH) that populate it. Interaction of positively charged buffer species with bound surface anions generates a layer of mobile cations that move toward the cathode when voltage is applied. This induces a very strong EOF that mobilizes all analytes in the same direction, regardless of their charge. Separation is consequently achieved because of differences in the electrophoretic migration rates of analytes superimposed on this EOF.

Because the driving force of the flow is distributed along the wall of the capillary, the flow profile is nearly flat or plug-like, contrasting with the laminar or parabolic flow generated by a pressure-driven system caused by shear forces at the wall. A flat flow profile is beneficial because it does not contribute to the dispersion of solute zones. The magnitude and direction of the EOF are influenced by several parameters, including: (1) type of electrolyte used; (2) pH; (3) ionic strength; (4) use of additives (e.g., surfactants, organic solvents); and (5) polarity and magnitude of the applied electric field.

Although advantageous for the dissipation of Joule heat, the large surface area-to-volume ratio of the inner capillary space increases the likelihood of analyte adsorption onto the surface of its inner wall. This causes phenomena such as peak tailing and even total and irreversible adsorption of the analyte. Adsorption is typically noted between cationic solutes and the negatively charged inner wall of the capillary, primarily through ionic interactions (with deprotonated silanols), but also involves hydrophobic interactions (with siloxanes). Because of the numerous charges and hydrophobic regions, significant adsorptive effects have been noted, especially for highly cationic proteins. In practice, adsorption of substances, whether from the sample or from the buffer, to the inner surface of the capillary will alter migration times and other separation characteristics; unaddressed, the capillary eventually may become “fouled.” Buffer components and/or additives, such as surfactants, can often render permanent changes to the inner surface of the capillary (through adsorption) and may warrant dedication of specific capillaries for use with particular surfactants.

To minimize these inner wall effects, capillaries are conditioned by chemical treatment, most commonly with base, to remove adsorbates and rejuvenate the surface. A typical wash method includes flushing the chamber with 10 to 20 capillary volumes of 0.1 to 1.0 mol/L NaOH, followed by flushing with “run” buffer. To prevent exposing the capillary surface to drastic fluctuations in pH, conditioning procedures for separations at low pH may be better served by using strong acids (e.g., HNO₃), surfactants (e.g., SDS), or organic solvents, such as acetonitrile or methanol.

Serum Protein Analysis

Compared with AGE and cellulose acetate electrophoresis (CAE), CZE is more advantageous for serum protein analysis. Fig. 11.5 shows a comparison of the separation of serum proteins by CAE, AGE, and CE. The presence of the classical zones with CE is apparent, albeit in reversed order, as is the identification of serum protein abnormalities in gamma regions. Retrospective studies have shown CE to be effective for detecting monoclonal proteins, which could then be immunotyped by conventional techniques (IFE and isoelectric focusing [IEF]). Moreover, CE can do both serum protein electrophoresis and immunotyping in hundreds of samples simultaneously. These and other studies put forth the same conclusion: that CE is more sensitive than AGE in identifying abnormalities. Furthermore, CZE can be used effectively in serum protein analysis.