The sieving properties of polyacrylamide gels are established by the three-dimensional network of fibers and pores that is formed as the bifunctional bisacrylamide cross-links adjacent polyacrylamide chains. Polymerization usually is catalyzed by a free radical–generating system composed of ammonium persulfate and N,N,N¢,N¢-tetramethylethylenediamine (TEMED).

As the acrylamide concentration of a gel increases, its effective pore size decreases. The effective pore size of a gel is operationally defined by its sieving properties, that is, by the resistance it imparts to the migration of macromolecules. There are limits on the acrylamide concentrations that can be used. As the pore size of a gel decreases, the migration rate of a protein through the gel decreases. By adjusting the pore size of a gel through manipulating the acrylamide concentration, analysts can optimize the resolution of the method for a given protein product. Thus, a given gel is physically characterized by its respective composition of acrylamide and bisacrylamide.

In addition to the composition of the gel, the state of the protein is an important component of electrophoretic mobility. In the case of proteins, the electrophoretic mobility depends on the pK value of the charged groups and the size of the molecule. It is influenced by the type, the concentration, and the pH of the buffer; by the temperature and the field strength; and by the nature of the support material.2S (USP38)

The method cited as an example is limited to the analysis of monomeric polypeptides with a mass range of 14,000–100,000 Da. It is possible to extend this mass range by various techniques (e.g., gradient gels and particular buffer system). For instance, tricine–sodium dodecyl sulfate (SDS) gels, using tricine instead of glycine (in the method described here) as the trailing ion, can separate very small proteins and peptides under 10,000–15,000 Da.

Denaturing polyacrylamide gel electrophoresis using glycine sodium dodecyl sulfate (SDS-PAGE) is the most common mode of electrophoresis used in assessing the pharmaceutical quality of protein products and is the focus of the example method. Typically, analytical electrophoresis of proteins is carried out in polyacrylamide gels under conditions that ensure dissociation of the proteins into their individual polypeptide subunits and that minimize aggregation. Most commonly, the strongly anionic detergent sodium dodecyl sulfate (SDS) is used in combination with heat to dissociate the proteins before they are loaded on the gel. The denatured polypeptides bind to SDS, become negatively charged, and exhibit a consistent charge-to-mass ratio regardless of protein type. Because the amount of SDS bound is almost always proportional to the molecular mass of the polypeptide and is independent of its sequence, SDS–polypeptide complexes migrate through polyacrylamide gels with mobilities dependent on the size of the polypeptide.

The electrophoretic mobilities of the resultant detergent–polypeptide complexes all assume the same functional relationship to their molecular masses. SDS complexes migrate toward the anode in a predictable manner; low-molecular-mass complexes migrate faster than larger ones. The molecular mass of a protein therefore can be estimated from its relative mobility in calibrated SDS-PAGE, and the intensity of a single band relative to other undesired bands in such a gel can be a measure of purity.

Modifications to the polypeptide backbone, such as N- or O-linked glycosylation, can change the apparent molecular mass of a protein, because SDS does not bind to a carbohydrate moiety in a manner similar to that of a polypeptide; therefore, a consistent charge-to-mass ratio is not maintained.

Depending on the extent of glycosylation and other posttranslational modifications, the apparent molecular mass of proteins may not be a true reflection of the mass of the polypeptide chain.

The most popular electrophoretic method for the characterization of complex mixtures of proteins uses a discontinuous buffer system involving two contiguous but distinct gels: a resolving or separating (lower) gel and a stacking (upper) gel. The two gels are cast with different porosities, pH, and ionic strengths. In addition, different mobile ions are used in the gel and electrode buffers. The buffer discontinuity acts to concentrate large-volume samples in the stacking gel, resulting in improved resolution. When power is applied, a voltage drop develops across the sample solution and drives the proteins into the stacking gel. Glycinate ions from the electrode buffer follow the proteins into the stacking gel. A moving boundary region is rapidly formed, with the highly mobile chloride ions in the front and the relatively slow glycinate ions in the rear. A localized high-voltage gradient forms between the leading and trailing ion fronts, causing the SDS–protein complexes to form into a thin zone (stack) and to migrate between the chloride and glycinate phases. Within broad limits, regardless of the height of the applied sample, all SDS–proteins condense into a very narrow region and enter the resolving gel as a well-defined, thin zone of high protein density. The large-pore stacking gel does not retard the migration of most proteins and serves mainly as an anticonvective medium. At the interface of the stacking and resolving gels, the proteins experience a sharp increase in retardation due to the restrictive pore size of the resolving gel and the buffer discontinuity, which also contributes to unstacking of the proteins. Once in the resolving gel, proteins continue to be slowed by the sieving of the matrix. The glycinate ions overtake the proteins, which then move in a space of uniform pH formed by the tris(hydroxymethyl)aminomethane (Tris) and glycine. Molecular sieving causes the SDS–polypeptide complexes to separate on the basis of their molecular masses.2S (USP38)

This section describes the preparation of gels using particular instrumentation. This does not apply to precast gels. For precast gels or any other commercially available equipment, the manufacturer's instructions should be used for guidance.

The use of commercial reagents that have been purified in solution is recommended. When this is not the case and when the purity of the reagents used is not sufficient, a pretreatment is applied. For instance, any solution sufficiently impure to require filtration must also be deionized with a mixed-bed (anion–cation exchange) resin to remove acrylic acid and other charged degradation products. Unopened, gas-sparged (with argon or nitrogen) acrylamide–bisacrylamide solutions and persulfate solid that is kept dry in a desiccator or a sealed bottle containing silica gel are stable for long periods. Fresh ammonium persulfate solutions are prepared daily.

Gradient gels (resolving gels) are prepared with an increasing concentration of acrylamide from the top to the bottom. Preparation of gradient gels requires a gradient-forming apparatus. Ready-to-use gradient gels are commercially available with specific recommended protocols.

Gradient gels offer some advantages over fixed-concentration gels. Some proteins that co-migrate on fixed-concentration gels can be resolved within gradient gels. During electrophoresis the proteins migrate until the pore size stops further progress, and therefore a stacking effect occurs, resulting in sharper bands. According to Table 3, gradient gels also allow separation of a wider range of protein molecular masses than do single, fixed-concentration gels.

Table 3 gives suggested compositions of the linear gradient, relating the range of acrylamide concentrations to the appropriate protein molecular ranges. Note that other gradient shapes (e.g., concave) can be prepared for specific applications.

Gradient gels also are used for molecular mass determination and protein purity determination.2S (USP38)

Coomassie and silver staining are the most common protein staining methods and are described in more detail below. Several other commercial stains, detection methods, and commercial kits are available. For example, fluorescent stains are visualized using a fluorescent imager and often provide a linear response over a wide range of protein concentrations—often several orders of magnitude, depending on the protein.

Coomassie staining has a protein detection level of approximately 1–10 µg of protein per band. Silver staining is the most sensitive method for staining proteins in gels, and a band containing 10–100 ng can be detected. These figures are considered robust in the context of these gels. Improved sensitivity of one or two orders of magnitude has been reported in the literature.

Coomassie staining responds in a more linear manner than silver staining, but the response and range depend on the protein and development time. Both Coomassie and silver staining can be less reproducible if staining is stopped in a subjective manner, i.e., when the analyst deems the staining satisfactory. Wide dynamic ranges of reference proteins are important to use because they help assess the intra-experimental sensitivity and linearity. All gel-staining steps are done while wearing gloves, at room temperature, with gentle shaking (e.g., on an orbital shaker platform), and using any convenient container.

Gels are photographed or scanned while they are still wet or after an appropriate drying procedure. Currently, gel-scanning systems with data analysis software are commercially available to photograph and analyze the wet gel immediately.

Drying of stained SDS polyacrylamide gels produces permanent documentation. This method, however, frequently results in gel cracking during drying between cellulose films, rendering the gel unsuitable for any kind of densitometry analyses later.

Depending on the staining method used, gels are treated in a slightly different way. For Coomassie staining, after the destaining step, allow the gel to stand in a 100 g/L solution of glycerol for at least 2 h (overnight incubation is possible). For silver staining, add to the final rinsing a step of 5 min in a 20 g/L solution of glycerol.

Immerse two sheets of porous cellulose film in water, and incubate for 5–10 min.

Place one of the sheets on a drying frame. Carefully lift the gel, and place it on the cellulose film. Remove any trapped air bubbles, and pour a few mL of water around the edges of the gel. Place the second sheet on top, and remove any trapped air bubbles. Complete the assembly of the drying frame. Place in an oven or leave at room temperature until dry.2S (USP38)

Molecular masses of proteins are determined by comparison of their mobilities with those of several marker proteins of known molecular weight. Mixtures of prestained and unstained proteins with precisely known molecular masses blended for uniform staining are available for calibrating gels. They are available in various molecular mass ranges. Concentrated stock solutions of proteins of known molecular mass are diluted in the appropriate sample buffer and are loaded on the same gel as the protein sample to be studied.

Immediately after the gel has been run, mark the position of the bromophenol blue tracking dye to identify the leading edge of the electrophoretic ion front. This can be done by cutting notches in the edges of the gel or by inserting a needle soaked in India ink into the gel at the dye front. After staining, measure the migration distances of each protein band (markers and unknowns) from the top of the resolving gel. Divide the migration distance of each protein by the distance traveled by the tracking dye. The normalized migration distances are referred to as the relative mobilities of the proteins (relative to the dye front), or RF. Construct a plot of the logarithm of the relative molecular masses (Mr) of the protein standards as a function of the RF values. Unknown molecular masses can be estimated by linear regression analysis (more accurately, by nonlinear regression analysis) or interpolation from the curves of log Mr against RF if the values obtained for the unknown samples are positioned along the approximately linear part of the graph.2S (USP38)

The test is not valid unless the proteins of the molecular mass marker are distributed along 80% of the length of the gel and over the required separation range covering the relevant protein bands (e.g., the product and the dimer or the product and its related impurities). The separation obtained for the expected proteins must show a linear relationship between the logarithm of the molecular mass and the RF. If the plot has a sigmoidal shape, then only data from the linear region of the curve can be used in the calculations. Additional validation requirements with respect to the test sample may be specified in individual monographs.

Sensitivity also must be validated. A reference protein control corresponding to the desired concentration limit that is run in parallel with the test samples can serve as a system suitability check of the experiment.2S (USP38)

When impurities are quantitated by normalization to the main band using an integrating densitometer or image analysis, the responses must be validated for linearity. Note that depending on the detection method and protein, as described in the introduction of the section Detection of Proteins in Gels, the linear range can vary but can be assessed within each run by using one or more control samples containing an appropriate range of protein concentrations.

When the impurity limit is specified in the individual monograph, analysts should prepare a reference solution corresponding to that level of impurity by diluting the test solution. For example, when the limit is 5%, a reference solution would be a 1:20 dilution of the test solution. No impurity (any band other than the main band) in the electropherogram obtained with the test solution may be more intense than the main band obtained with the reference solution.

Under validated conditions, impurities can be quantified by normalization to the main band, using an integrating densitometer, or by image analysis.2S (USP38)

As the acrylamide concentration of a gel increases, its effective pore size decreases. The effective pore size of a gel is operationally defined by its sieving properties, that is, by the resistance it imparts to the migration of macromolecules. There are limits on the acrylamide concentrations that can be used. At high acrylamide concentrations, gels break much more easily and are difficult to handle. As the pore size of a gel decreases, the migration rate of a protein through the gel decreases. By adjusting the pore size of a gel through manipulating the acrylamide concentration, analysts can optimize the resolution of the method for a given protein product. Thus, a given gel is physically characterized by its respective composition of acrylamide and bisacrylamide.

In addition to the composition of the gel, the state of the protein is an important component of electrophoretic mobility. In the case of proteins, the electrophoretic mobility depends on the pK value of the charged groups and the size of the molecule. It is influenced by the type, the concentration, and the pH of the buffer; by the temperature and the field strength; and by the nature of the support material.

Denaturing polyacrylamide gel electrophoresis using glycine SDS (SDS-PAGE) is the most common mode of electrophoresis used in assessing the pharmaceutical quality of protein products and is the focus of the example method. Typically, analytical electrophoresis of proteins is carried out in polyacrylamide gels under conditions that ensure dissociation of the proteins into their individual polypeptide subunits and that minimize aggregation. Most commonly, the strongly anionic detergent SDS is used in combination with heat to dissociate the proteins before they are loaded on the gel. The denatured polypeptides bind to SDS, become negatively charged, and exhibit a consistent charge-to-mass ratio regardless of protein type. Because the amount of SDS bound is almost always proportional to the molecular mass of the polypeptide and is independent of its sequence, SDS–polypeptide complexes migrate through polyacrylamide gels with mobilities dependent on the size of the polypeptide.

The electrophoretic mobilities of the resultant detergent–polypeptide complexes all assume the same functional relationship to their molecular masses. SDS complexes migrate toward the anode in a predictable manner; low-molecular-mass complexes migrate faster than larger ones. The molecular mass of a protein therefore can be estimated from its relative mobility in calibrated SDS-PAGE, and the intensity of a single band relative to other undesired bands in such a gel can be a measure of purity.

Modifications to the polypeptide backbone, such as N- or O-linked glycosylation, can change the apparent molecular mass of a protein, because SDS does not bind to a carbohydrate moiety in a manner similar to that of a polypeptide; therefore, a consistent charge-to-mass ratio is not maintained.

Depending on the extent of glycosylation and other posttranslational modifications, the apparent molecular mass of proteins may not be a true reflection of the mass of the polypeptide chain.

The most popular electrophoretic method for the characterization of complex mixtures of proteins uses a discontinuous buffer system involving two contiguous but distinct gels: a resolving or separating (lower) gel and a stacking (upper) gel. The two gels are cast with different porosities, pH, and ionic strengths. In addition, different mobile ions are used in the gel and electrode buffers. The buffer discontinuity acts to concentrate large-volume samples in the stacking gel, resulting in improved resolution. When power is applied, a voltage drop develops across the sample solution and drives the proteins into the stacking gel. Glycinate ions from the electrode buffer follow the proteins into the stacking gel. A moving boundary region is rapidly formed, with the highly mobile chloride ions in the front and the relatively slow glycinate ions in the rear. A localized high-voltage gradient forms between the leading and trailing ion fronts, causing the SDS–protein complexes to form into a thin zone (stack) and to migrate between the chloride and glycinate phases. Within broad limits, regardless of the height of the applied sample, all SDS–proteins condense into a very narrow region and enter the resolving gel as a well-defined, thin zone of high protein density. The large-pore stacking gel does not retard the migration of most proteins and serves mainly as an anticonvective medium. At the interface of the stacking and resolving gels, the proteins undergo a sharp increase in retardation due to the restrictive pore size of the resolving gel and the buffer discontinuity, which also contributes to focusing of the proteins. Once in the resolving gel, proteins continue to be slowed by the sieving of the matrix. The glycinate ions overtake the proteins, which then move in a space of uniform pH formed by the tris(hydroxymethyl)aminomethane (Tris) and glycine. Molecular sieving causes the SDS–polypeptide complexes to separate on the basis of their molecular masses.

The use of commercial reagents that have been purified in solution is recommended. When this is not the case and when the purity of the reagents used is not sufficient, a pretreatment is applied. For instance, any solution sufficiently impure to require filtration must also be deionized with a mixed-bed (anion–cation exchange) resin to remove acrylic acid and other charged degradation products. When stored according to recommendations, acrylamide/bisacrylamide solutions and solid persulfate are stable for long periods.

Gradient gels offer some advantages over fixed-concentration gels. Some proteins that co-migrate on fixed-concentration gels can be resolved within gradient gels. During electrophoresis the proteins migrate until the pore size stops further progress, and therefore a stacking effect occurs, resulting in sharper bands. According to Table 3, gradient gels also allow separation of a wider range of protein molecular masses than do single, fixed-concentration gels.

Table 3 gives suggested compositions of the linear gradient, relating the range of acrylamide concentrations to the appropriate protein molecular ranges. Note that other gradient shapes (e.g., concave) can be prepared for specific applications.

Gradient gels also are used for molecular mass determination and protein purity determination.

Coomassie staining has a protein detection level of approximately 1–10 µg of protein per band. Silver staining is the most sensitive method for staining proteins in gels, and a band containing 10–100 ng can be detected. These figures are considered robust in the context of these gels. Improved sensitivity of one or two orders of magnitude has been reported in the literature.

Coomassie staining responds in a more linear manner than silver staining, but the response and range depend on the protein and development time. Both Coomassie and silver staining can be less reproducible if staining is stopped in a subjective manner, i.e., when the analyst deems the staining satisfactory. Wide dynamic ranges of reference proteins are important to use because they help assess the intra-experimental sensitivity and linearity. All gel-staining steps are done while wearing gloves, at room temperature, with gentle shaking (e.g., on an orbital shaker platform), and using any convenient container.

Depending on the staining method used, gels are treated in a slightly different way. For Coomassie staining, after the destaining step, allow the gel to stand in a 100 g/L solution of glycerol for at least 2 h (overnight incubation is possible). For silver staining, add to the final rinsing a step of 5 min in a 20 g/L solution of glycerol.

Drying of stained SDS polyacrylamide gels is one of the methods to have permanent documentation. This method frequently results in gel cracking during drying between cellulose films.

Immerse two sheets of porous cellulose film in water, and incubate for 5–10 min.

Place one of the sheets on a drying frame. Carefully lift the gel, and place it on the cellulose film. Remove any trapped air bubbles, and pour a few mL of water around the edges of the gel. Place the second sheet on top, and remove any trapped air bubbles. Complete the assembly of the drying frame. Place in an oven or leave at room temperature until dry.

Immediately after the gel has been run, mark the position of the bromophenol blue tracking dye to identify the leading edge of the electrophoretic ion front. This can be done by cutting notches in the edges of the gel or by inserting a needle soaked in India ink into the gel at the dye front. After staining, measure the migration distances of each protein band (markers and unknowns) from the top of the resolving gel. Divide the migration distance of each protein by the distance traveled by the tracking dye. The normalized migration distances are referred to as the relative mobilities of the proteins (relative to the dye front), or RF. Construct a plot of the logarithm of the relative molecular masses (Mr) of the protein standards as a function of the RF values. Unknown molecular masses can be estimated by linear regression analysis (more accurately, by nonlinear regression analysis) or interpolation from the curves of log Mr against RF if the values obtained for the unknown samples are positioned along the approximately linear part of the graph.

Sensitivity also must be validated. A reference protein control corresponding to the desired concentration limit that is run in parallel with the test samples can serve as a system suitability check of the experiment.

When the impurity limit is specified in the individual monograph, analysts should prepare a reference solution corresponding to that level of impurity by diluting the test solution. For example, when the limit is 5%, a reference solution would be a 1:20 dilution of the test solution. No impurity (any band other than the main band) in the electropherogram obtained with the test solution may be more intense than the main band obtained with the reference solution.

Under validated conditions, impurities can be quantified by normalization to the main band, using an integrating densitometer, or by image analysis.2S (USP38)