U.S. Patent Documents:
Other References:
The present invention relates to size separation of proteins by capillary electrophoresis in sieving media, wherein one or more cationic surfactants form charged complexes with the proteins and equalize their surface charge density, making their migration in sieving media independent of their intrinsic charge and thus allowing their size separation and molecular-weight determination. Specifically, the invention is directed to capillary sieving electrophoresis of proteins in the presence of cationic surfactants at low pH.
Electrophoresis in Sieving Media
Electrophoretic sieving media are used to size separate biopolymers: nucleic acids, polysaccharides, and proteins. They provide a system of obstacles in the electrophoretic migration path so that migrating biopolymers collide with the obstacles and these collisions retard their apparent migration velocity. Larger molecules and particles are retarded in their migration more than small molecules. The first electrophoretic sieving media were starch and polyacrylamide gel. Nucleic acids are equally ionized at non-acidic pH and need not be modified to size separate during electrophoretic migration in sieving media. On the other hand, protein ionization and charge significantly vary depending on the amino acid composition. Therefore, native proteins are not size separated in sieving media in absence of ionic surfactants. However, when heated with an ionic surfactant, proteins denature and bind the surfactants, generating complexes with more or less equal surface charge density. These complexes migrate in sieving media according to their size. Ionic surfactants such as sodium dodecyl sulfate (SDS) (Shapiro, A. L. et al., 1967), cetyltrimethylammonium bromide (CTAB) Panyim, S. et al., 1977), cetylpyridinium chloride (Schick, M., 1975) have been used to equalize the surface charge density of proteins prior electrophoresis.
Slab Gel Electrophoresis
SDS electrophoresis in polyacrylamide slab gel (SDS PAGE) was the first method separating proteins according to their size (Shapiro, A. L. et al., 1967;Shapiro, A. L., Maizel, J. V., 1969; Weber, K., Osborn, M., 1969; Dunker, A. K., Rueckert, R. R., 1969). Formation of SDS-protein complexes is independent of ionic strength (Reynolds, J. A., Tanford, C., 1970). However, some proteins exhibit an anomalous migration in SDS PAGE (Shapiro, A. L. et al., 1967; Williams, J. G., Gratzer, W. B., 1971). The anomalous migration of acidic proteins in SDS PAGE was, however, normalized by esterification of carboxyl groups (Williams, J. G., Gratzer, W. B., 1971) suggesting insufficient surfactant binding. This hypothesis was corroborated by an observation that some acidic proteins, such as pepsin, papain, and glucose oxidase do not bind measurable amount of SDS (Nelson, C. A., 1971).
Shortly after the invention of SDS PAGE, a method separating proteins by polyacrylamide gel electrophoresis (PAGE) in the presence of cationic surfactants was described (Williams, J. G., Gratzer, W. B., 1971). A study based on observations of behavior of protein-cationic-surfactant-complexes followed, predicting a failure of the electrophoresis in the presence of cationic surfactants to determine molecular weights of proteins (Nozaki, Y. et al., 1974). Later, cetylpyridinium chloride (Schick, M., 1975) and cetyltrimethylammonium bromide (Akins, R. E. et al., 1992; Akins, R. E., Tuan, R. S., 1994; Akin, D. T. et al., 1985; Eley, M. H. et al., 1979; Panyim, S. et al., 1977) were used for size separations of proteins by PAGE. Several protocols have been developed to denature proteins with cetyltrimethylammonium bromide (Akins, R. E. et al., 1992; Akins, R. E., Tuan, R. S., 1994; Akin, D. T. et al., 1985; Eley, M. H. et al., 1979; Panyim, S. et al., 1977), including a protocol without any heating of the sample (Akins, R. E. et al., 1992). Capillary Electrophoresis
When electrophoresis of proteins in sieving media was transferred from slab gels into capillaries, crosslinked polyacrylamide gel was initially used as the sieving matrix (Hjerten, S., 1983; Cohen, A. S., Karger, B. L., 1987; Dolnik, V. et al., 1991). When linear hydrophilic polymers were introduced as a replaceable sieving matrix for separation of polynucleotides (Hjerten, S. et al., 1989), various polymers were utilized as a sieving matrix for electrophoretic size separation of biopolymers, including linear polyacrylamide (Ganzler, K. et al., 1992; Sudor, J. et al., 1991; Heiger, D. N. et al., 1990, Werner, W. E. et al., 1993;;Karim, M. R. et al., 1994;Tsuji, K., 1994; Hu, S. et al., 2002; Craig, D. B. et al., 1998; Hunt, G., Nashabeh, W., 1999; Salas-Solano, O. et al., 2006)), poly(ethylene oxide) (Guttman, A. et al., 1993), dextran (Ganzler, K. et al., 1992), guaran (Dolnik, V. et al., 2001), glucomannan (Izumi, T. et al., 1993), poly(vinyl alcohol) (Kleemiss, M. H. et al., 1993), poly(dimethyl acrylamide) (Madabhushi, R. S., 1998), poly(hydroxypropyl acrylamide) (Lindberg, P. et al., 1997), poly(ethoxyethyl acrylamide) (Chiari, M. et al., 1994; Chiari, M. et al., 1997), agarose (Motsch, S. R. et al., 1991), and pullulan (Nakatani, M. et al., 1996; Nakatani, M. et al., 1994). Capillary size separations of proteins were performed exclusively by SDS capillary sieving electrophoresis (CSE) and limited to a molecular-weight range between about 14,000 and 205,000. The method was also modified for size separation of proteins on microchip (Yao, S. et al., 1999; Bousse, L. et al., 2001) with poly(dimethyl acrylamide) as a sieving polymer (Bousse, L. et al., 2001). Capillary electrophoresis brought a number of advantages as compared to electrophoresis in slab gel: faster analysis, automation, higher separation efficiency, higher detection sensitivity. Nevertheless, a small size of capillaries emphasized the effect of the capillary wall: typically used fused silica capillaries contained ionized silanol groups on their internal surface, resulting in strong wall adsorption, significant electroosmotic flow, eddy migration, and consequent mediocre separation efficiency. Electroosmotic flow was eventually suppressed by applying a hydrolytically stable neutral coating on the capillary wall (U.S. Pat. No. 5,143,753). Nevertheless, in SDS CSE, SDS adsorbs on the neutral coating and generates secondary electroosmotic flow. Mediocre reproducibility and separation efficiency are the results of this deleterious effect. Currently, SDS CSE is performed in bare capillaries after extensive rinsing of the capillary between runs, significantly reducing the throughput of the analysis. Hypothetically, electroosmotic flow in SDS CSE could be also suppressed by reducing pH of the sieving medium and a consequent suppression of the silanol ionization in the capillary wall. However, SDS binding of proteins is weaker at pH<6 and SDS electrophoresis at this pH results in significantly broader peaks (Gilbert, H. F., 1995) excluding this alternative from a real world practice.
When migration of biopolymers in sieving media was studied in a greater detail, three distinct migration modes were identified, depending on the pore size of the sieving matrix and size of the polyelectrolyte: Ogston mode, reptation without stretching, and reptation with stretching (Noolandi J., 1992). In capillary electrophoresis, where sieving polymers serve as the sieving matrix, the pore size of the sieving matrix is fine tuned by changing the concentration of the sieving polymer (Ganzler, K. et al., 1992; Karim, M. R. et al., 1994). Molecular weight of the sieving polymer also plays an important role, particularly in separation of large polyelectrolytes migrating in the reptation-with-stretching mode (Bae, Y. C., Soane, D., 1993). High-molecular-weight polymers are typically better sieving matrices than those with low molecular weight (Dolnik, V. et al., 2001) although blends of polymers of high and low molecular weight were also advocated and used (Salas-Solano, O. et al., 1998). Stiffness and branching of the sieving polymers affects the sieving properties of the sieving matrix as well: branched polysaccharides such as dextran or Ficoll® exhibit rather limited sieving and are suitable more for separation of polyelectrolytes migrating in the Ogston mode. Stiff polysaccharides such as hydroxyethyl cellulose, guaran, and locust bean gum (Dolnik, V. et al., 2001) form a rigid sieving matrix at lower molecular weights and/or concentrations than easily bending synthetic polymers such as linear polyacrylamide. In practical CSE analysis, viscosity is a limiting factor as high pressure (about 900 psi) is necessary to replace a viscous sieving matrix in the capillary. The viscosity of the sieving matrix makes a practical limit when increasing the concentration and molecular weight of the sieving polymer. In DNA sequencing by capillary electrophoresis, the weight molecular weight (Mw, further molecular weight) of linear polyacrylamide serving as a sieving polymer, exceeded 10,000,000 (Goetzinger, W. et al., 1998). In CSE of proteins, most proteins were separated in sieving media with large pore size and sieving polymers typically did not exceed molecular weight of 1 million. To provide efficient sieving, the sieving polymers should have molecular weight about 100,000 and more. However, this value is rather arbitrary, as lower-molecular-weight polymers can be prepared at a higher concentration and still keep the viscosity of the solution at an acceptable level. Stiff polysaccharides, such as hydroxyethyl cellulose, guaran, or locust bean gum can be used even at a lower molecular-weight. On the other hand, crosslinked and branched polysaccharides, such as dextran, are preferably used at high molecular weight (2 million and higher).
Cationic Surfactant in Capillary Electrophoresis
Cetyltrimethylammonium bromide was used in capillary electrophoresis for two purposes: a) as a dynamic coating to reverse the direction of electroosmotic flow (Reijenga, J. C. et al., 1983; Tsuda, T., 1987; Corradini, D., 1997; Ding, W. L., Fritz, J. S., 1997; Chiari, M. et al., 1998), b) as a pseudostationary phase in capillary electrokinetic micellar chromatography (Ong, C. P. et al., 1994). None cationic surfactant has been used for size separation of proteins by capillary sieving electrophoresis. An interesting group of potential cationic surfactants are gemini surfactants (Menger, F. M., Keiper, J. S., 2000) with two amino groups connected by an aliphatic arm but they have not been used in capillary electrophoresis yet.
The present invention is suitable for a fast, quantitative, and highly reproducible size separation of proteins by means of capillary sieving electrophoresis. Disclosed herein are a composition of a separation medium and a protein denaturing solution, and a method of capillary sieving electrophoresis in the presence of a cationic surfactant for size separation of proteins with molecular weight in the range between about 14,000 and about 500,000. In the preferred embodiment, the separation medium comprises an acidic buffer that keeps pH in the range between about 3 and about 5, a hydrophilic sieving polymer with moderate viscosity, and between about 0.5 and about 30 g/L cationic surfactant.
Cationic surfactants
The cationic surfactant used in the sieving matrix should exhibit a sufficient solubility in water and, simultaneously strongly bind proteins. Primary, secondary, tertiary or quaternary amines with one or more long aliphatic chains are suitable surfactant cations for CSE of proteins. Typically longer aliphatic chains are preferred because then the surfactant binds proteins more strongly. Solubility of the cationic surfactant in water may be a limiting factor for a practical use. Cationic surfactants suitable for capillary sieving electrophoresis of proteins contain one or more of the following cations: octadecyldimethylethylammonium, cetyldimethylethylammonium, tetradecyldimethylethylammonium, dodecyldimethylethylammonium, octadecyltrimethylammonium, cetyltrimethylammonium, tetradecyltrimethylammonium, dodecyltrimethylammonium, octadecylpyridinium, tetradecylpyridinium, dodecylpyridinium, octadecylammonium, cetylammonium, tetradecylammonium, dodecylammonium, decylammonium, didodecyldimethylammonium, and a cationic gemini surfactant alkanediyl-.α.,.ω.-bis(dimethylalkylammonium), with a formula CmH2m+1(CH3)2N+(CH2)sN+(CH3)2CmH2m +1, m being 12, 14, 16, or 18 and s being 2, 3, 4, 5, 6, 7, or 8.
The counter anion in the cationic surfactant can be an inorganic anion such as chloride, bromide, sulfate, bicarbonate, etc. The nature of the counter anion plays a significant role. In capillary sieving electrophoresis of monoclonal antibodies, a broad system peak appears when cetyltrimethylammonium bromide is used as the cationic surfactant. With cetyltrimethylammonium chloride, the system peak is much smaller and such a sieving medium is more preferred.
Acidic Buffer
Size separations of proteins in capillary format based on SDS suffer from mediocre separation efficiency and lower reproducibility of qualitative and quantitative analysis. Capillary sieving electrophoresis of proteins with cationic surfactant at low pH, where ionization of silanol groups in the capillary wall is suppressed, is a solution to this problem. The separation medium contains a cationic surfactant, a sieving polymer, and an acidic buffer with pH below 5. Low pH is absolutely essential for high performance separations, because silanol groups in the fused-silica capillary wall are not ionized at low pH. This leads to a lower electroosmotic flow, which otherwise deteriorates electrophoretic separation. At low pH, the adsorption of cationic surfactants on the capillary wall, which normally leads to a significant reversed electroosmotic flow, is also suppressed.
The pH of the sieving matrix, however, requires some optimization: Below pH 3, the high-mobility H+ ion contributes significantly to the conductivity of the sieving matrix. This results in elevated Joule heat and overheating of the capillary. Above pH 5.5, the silanol ionization is not negligible and electroosmotic flow becomes a serious issue. Keeping the pH of the sieving matrix at about pH 4 is the best compromise. One possibility is to use a free weak acid, e.g., acetic acid, as the only electrolyte. Another option is to use fully ionized cation, e.g., Tris, with a buffering anion, e. g., formate, acetate, propionate, butyrate, capronate, valproate, pimelate, fumarate, maleate, succinate, glutarate, adipate, malate, tartrate, glycolate, lactate, 2-hydroxybutyrate, 2-hydroxyisobutyrate, citrate, nicotinate, glutamate, and aspartate. pH can be also kept at a proper level with a buffering cation, e.g., β-alanine, γ-aminobutyric acid (GABA), glycine, ε-aminocaproic acid, or nicotinamide and a fully ionize anion. Probably the most attractive buffering option is an equimolar mixture of a weak base and a weak acid, having close pK's, e.g., GABA (pK 4.0) and glutamic acid (pK 4.2), β-alanine (pK 3.6) and glutamic acid (pK 4.2), and β-alanine (pK 3.6) and 2-hydroxy-isobutyric acid (pK 3.9). It is essential to use dicarboxylic and tricarboxylic acids at pH, where only one carboxylic group is partly dissociated. Other factors than pK, which may be difficult to predict from the physicochemical properties of the buffers, may be also important: Buffers containing β-alanine show better separation efficiency than buffers with GABA, but exhibit some protein adsorption on the wall. This results in a minor but discernable baseline elevation when proteins migrate through the detection cell.
Sieving Polymer
The sieving matrix enables size separation of proteins. It provides obstacles in the migration path and makes proteins complexed with cationic surfactants to electrophoreticaly migrate according to their size. The sieving polymer should be (i) soluble in water, (ii) non-ionic, (iii) not significantly binding cationic surfactants, (iv) sufficient sieving (i.e., having sufficient molecular weight), (v) non UV absorbing, if UV detection is used. The sieving properties of the matrix can be fine tuned by changing the molecular weight and concentration of the sieving polymer. Higher concentration and/or higher molecular weight of the sieving polymer results in smaller pores, i.e., in more efficient sieving but also in longer migration times. Simultaneously, the viscosity rises and may prevent fast replacement of the sieving matrix in the capillary. Linear polysaccharides, such as hydroxyethyl cellulose, hydroxypropyl cellulose, scleroglucan, guaran, locust bean gum, glucomannan, and pullulan, have a stiff molecule and will be typically used at concentrations from about 4 to about 60 g/L and molecular weight from about 20,000 to about 500,000. Hydrophilic synthetic polymers, such as linear polyacrylamide, poly(dimethyl acrylamide), poly(hydroxyethyl acrylamide), poly(hydroxypropyl acrylamide), poly(ethoxyethyl acrylamide), poly(vinyl alcohol), poly(vinyl pyrrolidone), and poly(ethylene oxide) will be effective sieving polymers at concentration from about 8 to about 80 g/L and molecular weight from about 100,000 to about 1 million. Branched polysaccharides such as dextran or Ficoll® are less efficient sieving polymers and have to be used at concentration from about 100 to about 400 g/L and molecular weight about 2 million. UV absorbing polymers, such as poly(vinyl pyrrolidone) will not properly work in CSE with UV detection but may be used in CSE with laser induced fluorescence detection.
We disclose here a separation medium for capillary electrophoretic size separation of proteins, comprising
The sample denaturing solution should contain a cationic surfactant, which may but need not be identical with the cationic surfactant in the sieving matrix, a reduction agent, which can disrupt disulfide bridges (β-mercaptoethanol or dithiotreitol), and a high-mobility cation that allows a transient isotachophoresis during the electrokinetic injection and helps to focus the analytes into sharp bands.
The other role of the high-mobility cation in the sample denaturing solution is to allow the quantitative analysis with electrokinetic injection. Pressure injection, common in capillary electrophoresis, is not recommended for quantitative analysis by capillary sieving electrophoresis. The sieving matrix contains a polymer solution and exhibits an increased viscosity. As the result, the precision of the pressure injection may be compromised. If EOF in the separation capillary is suppressed, the amount of analytes injected electrokinetically is not necessarily proportional to their concentration in the sample and a non-linear calibration curve may be obtained. This holds particularly for low-conductivity samples. Nevertheless, if an electrolyte is added and the analytes do not contribute significantly to the overall conductivity of the sample, the calibration curves become linear. Moreover, polymer solution on the capillary inlet tip may affect the reproducibility of quantitative analysis; a tip wash can lead to a better reproducibility.
We disclose here a protein denaturing solution for the sample preparation prior capillary sieving electrophoresis with cationic surfactant comprising:
octadecyldimethylethylammonium, cetyldimethylethylammonium, tetradecyldimethylethylammonium, dodecyldimethylethylammonium, octadecyltrimethylammonium, cetyltrimethylammonium, tetradecyltrimethylammonium, dodecyltrimethylammonium, octadecylpyridinium, tetradecylpyridinium, dodecylpyridinium, octadecylammonium, cetylammonium, tetradecylammonium, dodecylammonium, decylammonium, didodecyldimethylammonium, and cationic gemini surfactant alkanediyl-.α.,.ω.-bis(dimethylalkylammonium), with a formula CmH2m+1(CH3)2N+(CH2)sN+(CH3)2CmH2m+1, wherein m is 12, 14, 16, or 18, and s is 2, 3, 4, 5, 6, 7, or 8;
Method of Capillary Sieving Electrophoresis with a Cationic Surfactant
Here we disclose a method for capillary sieving electrophoresis with cationic surfactant for size separation of proteins comprising steps:
The separations described in these examples were performed in 3D CE capillary electrophoresis instrument at 20° C. in a bare or coated capillary of internal diameter 75 μm and outer diameter 360 μm with UV detection at 214 nm.
Preparation and Composition of the Sieving Matrix
The separation medium for capillary sieving electrophoresis with a cationic surfactant was formulated to contain cetyldimethylethylammonium bromide (CDMEAB), or cetyltrimethylammonium bromide (CTAB), or cetyltrimethylammonium chloride (CTAC) as the cationic surfactant, polyacrylamide or poly(ethylene oxide) (PEO) as a sieving matrix, β-alanine or γ-aminobutyric acid as the buffering co-ion, and 2-hydroxyisobutyric acid or glutamic acid as the buffering counter-ion. Standard formulations contained CDMEAB; however, the formulations with CTAC were preferred for separation of monoclonal antibodies. Formulation with β-alanine were designed for high resolution separations, formulations with γ-aminobutyric acid were preferred where straight baseline was necessary. The specific formulations contained:
Composition of Sample Denaturing Solution and Method of Sample Preparation
Several compositions of the sample denaturing solution were formulated to enable protein quantitation with electrokinetic injection:
During the sample preparation, proteins were dissolved in the sample denaturing solution and incubated at 95° C. for 2 min. Some proteins, e.g., lysozyme, were resistant to the thermal denaturation with cationic surfactants and an extended incubation at 95° C. was necessary (5 min in case of lysozyme). Proteins such as BSA, on the other hand, did not require any denaturation at all prior to electrophoresis.
The Method of Capillary Sieving Electrophoresis
Capillary sieving electrophoresis with a cationic surfactant was performed in a fused silica capillary, 75 μm ID, 360 μm OD, 335 mm total length, 250 mm effective length. Bare capillaries were also used, but for high-resolution separations, capillaries with internal hydrophilic coating were preferred. After each electrophoretic run, the capillary was flushed with 100 mM citric acid at pressure of 930 mbar for 7 min to remove the sieving matrix from the previous run and wash proteins and other material potentially adsorbed on the capillary wall. In the next step, the capillary was prepared for the next run: the fresh sieving matrix was pumped into the capillary with pressure of 930 mbar for 3 min. The samples were injected either electrokinetically or by pressure. The amount of the injected sample depended on the protein concentration in the sample. The samples prepared with the sample denaturing solution containing 10 g/L CDMEAB, 100 mM KCl, and 10 g/L dithiotreitol and containing 0.1-1 g/L proteins were typically injected for 8 s at 6 kV. The separation was performed at +10 and typically took 10-12 minutes. The separation of a model protein mixture is shown in
The Method of Capillary Sieving Electrophoresis with Cationic Surfactant for Separation of Large Proteins
For the separation of native BSA oligomers, the sample containing 10 g/L BSA in 1 g/L CDMEAB was injected at 8 kV for 15 s. The capillary sieving electrophoresis (CSE) of BSA oligomers took about 12 min. and revealed eight to nine peaks (
Separation Efficiency
CSE with a cationic surfactant provided narrow peaks with high separation efficiency. Table 1 summarizes the average separation efficiency of model proteins from 7 runs. The calculation of the separation efficiency from a half-height peak width that assumed ideal Gaussian peaks provided results rather lower than the calculation based on an unrevealed algorithm used in ChemStation software (Agilent).
acalculated from half-height peak width
bobtained directly from the ChemStation software (Agilent)
Reproducibility of Migration Times
Low pH of the separation medium minimized electroosmotic flow and improved reproducibility of the electrophoretic separation of proteins. 10 overlaid consecutive electropherograms of a model mixture containing 0.8 g/L of insulin B, lysozyme, β-lactoglobulin, α-chymotrypsinogen A, ovalbumin, and BSA are shown in
Quantitative Analysis
CSE with cationic surfactant allowed quantitative analysis with electrokinetic injection. When proteins were denatured in 10 g/L CDMEAB, 100 mM KCl, and 10 g/L DTT and injected 30 s at +10 kV, the calibration lines for lysozyme, β-lactoglobulin, ovalbumin, and BSA were linear in the concentration range 0-1.0 g/L (