ANALYTICAL BIOCHEMISTRY 2 0 6 , 84-90 (1992)
Isoelectric Focusing by Free Solution Capillary Electrophoresis Shiaw-Min Chen a n d J o h n E. Wiktorowicz Applied Biosystems, Inc., 850 Lincoln Centre Drive, Foster City, California 94404
Received March 27, 1992
A reproducible, quantitative isoelectric focusing method using capillary electrophoresis that exhibits high resolution and linearity over a wide pH gradient w a s d e v e l o p e d . R N a s e T1 a n d R N a s e b a a r e t w o prot e i n s t h a t h a v e i s o e l e c t r i c p o i n t s (pFs) at t h e t w o e x tremes of a pH 3-10 gradient. Site-directed mutants of the former were separated from the wild-type form and p/'s determined in the same experiment. The p/'s of R N a s e T~ w i l d - t y p e , its t h r e e m u t a n t s , a n d R N a s e b a w e r e d e t e r m i n e d f o r t h e first t i m e as 2 . 9 , 3 . 1 , 3 . 1 , 3 . 3 , and 9.0, respectively. The paper describes the protocol f o r i s o e l e c t r i c f o c u s i n g b y c a p i l l a r y e l e c t r o p h o r e s i s , as well as presenting data describing the linearity, resolution, limits of mass loading, and reproducibility of the method.
© 1992 Academic Press, Inc.
Isoelectric focusing (IEF) 1 is a method for purification and analysis t hat causes charged molecules to migrate electrophoretically through a pH gradient until they experience a p H at which they exhibit zero net charge. At the time of its introduction (1), IEF was accomplished in a density-stabilized, vertical column of ampholytes. These ampholytes were small zwitterions t h a t exhibited sufficient buffering capacity in order to maintain a stable pH gradient. Sample was dissolved in the ampholyte solution, and electrophoresis was performed in free solution, i.e., without the use of gels or solids as anticonvective agents. Partly because of the attractiveness of IEF as a first dimension for two-dimensional electrophoresis (2) and the need therefore for a semisolid matrix, but mostly because convective currents generated by Joule heating limited the resoluAbbreviations used: IEF, isoelectric focusing; CE, capillary electrophoresis; CCK, cholecystikin; RNase A, ribonuclease A; RNase T1, ribonuclease T~; RNase ba, ribonuclease from Bacillus amylofasciens; DB-1, dimethyl polysiloxane; T E M E D , N,NJVJV-tetramethylethylenediamine; DB-17, dimethyldiphenylpolysiloxane. 84
tion achievable by this system, IEF was further developed using a cross-linked polyacrylamide matrix. Because of this development, standard detection strategies and gel handling mechanisms were quickly applied to IEF analysis of proteins and peptides. The high degree of resolution led to its popularity as an analytical technique for the measurement of pFs as well as another method for gauging purity and establishing identity. Commercial systems designed to partially automate the separation and detection became available (Phast systems, Pharmacia, Piscataway, NJ). However, the limitations of gel-based separations, in addition to the semiquantitative nature of current staining strategies, hinder the realization of the full potential of IEF; e.g., loss of resolution due to Joule heating, inability to accurately quantify the am ount of analyte in bands, long analysis time, labor intensive preparation, postseparation handling, and large sample mass requirements. T h e most recent developments in gel-based IEF address some of these limitations, such as the application of silver staining to decrease mass loading requirements, t hi nner gels to permit more efficient heat dissipation, or shorter gels to decrease analysis time, but all achieve their benefit through some sacrifice, either in expense, resolution, gel fragility, etc. The most recent IEF development, using capillary electrophoresis (CE-IEF) as the separation strategy (3-9), maintains and enhances all the advantages of gel-based IEF but suffers none of the disadvantages. This method rediscovers the merits of free solution electrophoresis without the detrimental effects created by convection by accomplishing separations in a 50-#m diameter fused-silica capillary. T h e high efficiency of heat dissipation of the smaller diameter tube permits high field strengths (volts per unit length) and therefore higher resolution and shorter analysis times. T he narrow diameter capillary permits analysis with minute mass requirements (pg). In addition the uv transparency of the fused silica permits online detection with precise and accurate quantification. Finally the free solution format eliminates the need for 0003-2697/92 $5.00 Copyright © 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
CAPILLARY ELECTROPHORESIS-ISOELECTRIC FOCUSING polymerization steps in preparation for electrophoresis and permits complete control over the composition of the separation buffer. In general, C E - I E F poses two challenges. Since detection is on-line, focusing must be performed in the absence of endosmotic flow; however, once established, the ampholyte gradient must be mobilized past the online detector in order to observe the focused proteins and pH gradient. Various strategies have been published (3,5,6,8,9), and the yardstick for effectiveness is (or should be) the demonstration of the linearity of the pH gradient, the accuracy and precision of the pI estimation, and the degree of resolution achieved. Indeed, the majority of the literature on C E - I E F is dedicated to the resolution of these two challenges (see Discussion for more complete comparison). This paper describes capillary electrophoresis isoelectric focusing and demonstrates its utility in the measurement of the pI of proteins at the extremes of a pH 3-10 gradient. In addition, the linearity, accuracy, precision, and resolution observed for this technique at the extremes of the pH gradient is discussed. MATERIALS AND METHODS In this study Servalyt 3-10 was purchased from Serva (Serva Fine Chemicals, Paramus, NJ), and methylcellulose (1500 cps at 2%) from Sigma (Sigma Chemicals, St. Louis, MO). Carbonic anhydrase (pI 5.9),/~-lactoglobulin A (pI 5.1), and ribonuclease A (RNase A, pI 9.45) were obtained in the highest purity from Sigma. Unsulfated cholecystikinin (CCK) flanking peptide (pI 2.75) was obtained from Peninsula Labs (Belmont, CA.). Ribonuclease T1 (RNase TO and mutants from Aspergillus oryzae and ribonuclease from Bacillus amylofasciens (RNase ba) were kind gifts of Drs. C. Nick Pace and Bret Shirley at Texas A&M University. All other reagents used were of analytical grade. Capillary electrophoresis was performed on the Applied Biosystems Model 270A-HT capillary electrophoresis system (Foster City, CA) using a dimethylpolysiloxane (DB-1)-coated capillary with 50 #m internal diameter and 0.05 #m coating thickness (J&W Scientific, Folsom, CA). The total length of the capillary used in this study was 72 cm and the length from the sample injection end to the detector was 50 cm. The output of the 270A-HT was plotted and integrated by a HewlettPackard Model 3396 integrator. The protocol used to achieve the separations requires the accurate application of precise vacuum to the capillary in order to fill it with the correct amount of electrode buffers, ampholytes, and sample. The system was programmed to first fill the capillary with 20 mM NaOH in 0.4% methylcellulose (8 min of 20 in. Hg vacuum), followed by a section of 0.5% Servalyt 3-10 in 0.4% methylcellulose (4.5 rain of 20 in. Hg vacuum). An ali-
85
quot of sample and marker (dissolved in Milli-Q water) was loaded separately into the capillary (30 s of 5 in. Hg vacuum each) and another section of 0.5% Servalyt 3-10 with 0.4% methylcellulose (0.1 min of 20 in. Hg vacuum) was added to the capillary. As illustrated in Fig. 1, at this point the interface between the NaOH solution and Servalyt solution was positioned on the sample side of the detector. High voltage, usually 30 kV, was then applied for 6 min to focus the ampholyte and proteins. After focusing, the system was programmed to apply a precisely regulated vacuum (5 in. Hg), while maintaining the high voltage. The applied vacuum caused the focused zone of proteins in the capillary to flow pass the detector while the voltage maintained the pH gradient and zone sharpness even in the presence of the distorting effects of laminar flow. The protein bands were detected at 280 nm as they passed the detector under the influence of vacuum and voltage. RESULTS
Linearity of the pHgradient. The linearity of the pH gradient was determined by running RNase A, carbonic anhydrase, ~-lactoglobulin A, and CCK-flanking peptide as markers. Figure 2 shows the electropherogram of a 20-nl injection of 100 #g/ml of each protein (2 ng each loaded). The proteins were focused and mobilized as described under Materials and Methods. After mobilization and detection at 280 nm, the relative mobility of each protein was calculated (see below) and plotted against the published isoelectric point (Fig. 2, inset). The double peak before the RNase A in Fig. 2 represents the interface between cathodic buffer and ampholyte solution. The peak after CCK-flanking peptide is the interface of anodic buffer and ampholyte solution. These two interface peaks are characteristics of this procedure and can be used to indicate the beginning and the end of the pH gradient. The times at which these peaks appear after mobilization is initiated (mobilization vacuum and voltage applied) is used to calculate the relative mobility (Rm) as follows: Rm
=
to t f - to
tp
where tp is the mobilization time of the protein peak of interest, t o (origin) is the mobilization time of the cathodic buffer interface, and tf (front) is the mobilization time of the anodic buffer interface. The linearity of the pH gradient was determined by plotting the pI vs relative mobility (Fig. 2, inset). This plot shows this method can provide linear range between pH 2.75 and 9.5 (CCK-flanking peptide and RNase A, respectively). Resolution. RNase T1 from A. oryzae (wild-type) is a protein with a calculated pI of 2.9. Because of its ex-
86
CHEN AND WIKTOROWICZ Detector
~
Hydrodynamic flow LOAD CATHOLYTE
NaO
Detector
NaOH
~
~:)7..~:~.~7~'..~:~`~.7.;:.:~.~.`~.~:~`~i..;?.:~:.('~:~.ii/~:~.~i~:~.~:.:'.~:'?~.!.~Ampholyte :..~7.~.~`~(~(~.AMPHOLYTE ::.LOAD 7i:~.~.~: Detector
NaOH
Hydrodynamic flow
~
Hydrodynamic flow
LOAD ~i:~:~i.(i.;~:~`.~``.:.:.:. ;~:i;~i:;.?i~:~.~:~i:~.:. ~i:.)~.~:~i.~:~i.~i:.:.;;i:./`i:./`~i:~:~:i};). Sample .i:~:::;i(~(:. ).I SAMPLE Detector
--
Hydrodynamic flow "INSULATE" SAMPLE
Detector
•',,IIII......................Electrical
field FOCUS
Detector
~
Hydrodynamic flow
'",llll...................... Electrical field ~fl,
NaOH H i
:::::::::::::::: ::::::::::::;::.':::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::.',::;::.~?::::= ;:::::::::::;:::: . ........... ' : : . , v : : t "............. . ' . , ' : : . , , I.:::.,'.':.'.:.~ . . . . . . . . . . . . . . :.': ! ".'.'.'.'.,'::.:. . . . . . . . ...................... ............. ".:.:.:
Isoelectric Focusing by Free Solution Capillary Electrophoresis Shiaw-Min Chen a n d J o h n E. Wiktorowicz Applied Biosystems, Inc., 850 Lincoln Centre Drive, Foster City, California 94404
Received March 27, 1992
A reproducible, quantitative isoelectric focusing method using capillary electrophoresis that exhibits high resolution and linearity over a wide pH gradient w a s d e v e l o p e d . R N a s e T1 a n d R N a s e b a a r e t w o prot e i n s t h a t h a v e i s o e l e c t r i c p o i n t s (pFs) at t h e t w o e x tremes of a pH 3-10 gradient. Site-directed mutants of the former were separated from the wild-type form and p/'s determined in the same experiment. The p/'s of R N a s e T~ w i l d - t y p e , its t h r e e m u t a n t s , a n d R N a s e b a w e r e d e t e r m i n e d f o r t h e first t i m e as 2 . 9 , 3 . 1 , 3 . 1 , 3 . 3 , and 9.0, respectively. The paper describes the protocol f o r i s o e l e c t r i c f o c u s i n g b y c a p i l l a r y e l e c t r o p h o r e s i s , as well as presenting data describing the linearity, resolution, limits of mass loading, and reproducibility of the method.
© 1992 Academic Press, Inc.
Isoelectric focusing (IEF) 1 is a method for purification and analysis t hat causes charged molecules to migrate electrophoretically through a pH gradient until they experience a p H at which they exhibit zero net charge. At the time of its introduction (1), IEF was accomplished in a density-stabilized, vertical column of ampholytes. These ampholytes were small zwitterions t h a t exhibited sufficient buffering capacity in order to maintain a stable pH gradient. Sample was dissolved in the ampholyte solution, and electrophoresis was performed in free solution, i.e., without the use of gels or solids as anticonvective agents. Partly because of the attractiveness of IEF as a first dimension for two-dimensional electrophoresis (2) and the need therefore for a semisolid matrix, but mostly because convective currents generated by Joule heating limited the resoluAbbreviations used: IEF, isoelectric focusing; CE, capillary electrophoresis; CCK, cholecystikin; RNase A, ribonuclease A; RNase T1, ribonuclease T~; RNase ba, ribonuclease from Bacillus amylofasciens; DB-1, dimethyl polysiloxane; T E M E D , N,NJVJV-tetramethylethylenediamine; DB-17, dimethyldiphenylpolysiloxane. 84
tion achievable by this system, IEF was further developed using a cross-linked polyacrylamide matrix. Because of this development, standard detection strategies and gel handling mechanisms were quickly applied to IEF analysis of proteins and peptides. The high degree of resolution led to its popularity as an analytical technique for the measurement of pFs as well as another method for gauging purity and establishing identity. Commercial systems designed to partially automate the separation and detection became available (Phast systems, Pharmacia, Piscataway, NJ). However, the limitations of gel-based separations, in addition to the semiquantitative nature of current staining strategies, hinder the realization of the full potential of IEF; e.g., loss of resolution due to Joule heating, inability to accurately quantify the am ount of analyte in bands, long analysis time, labor intensive preparation, postseparation handling, and large sample mass requirements. T h e most recent developments in gel-based IEF address some of these limitations, such as the application of silver staining to decrease mass loading requirements, t hi nner gels to permit more efficient heat dissipation, or shorter gels to decrease analysis time, but all achieve their benefit through some sacrifice, either in expense, resolution, gel fragility, etc. The most recent IEF development, using capillary electrophoresis (CE-IEF) as the separation strategy (3-9), maintains and enhances all the advantages of gel-based IEF but suffers none of the disadvantages. This method rediscovers the merits of free solution electrophoresis without the detrimental effects created by convection by accomplishing separations in a 50-#m diameter fused-silica capillary. T h e high efficiency of heat dissipation of the smaller diameter tube permits high field strengths (volts per unit length) and therefore higher resolution and shorter analysis times. T he narrow diameter capillary permits analysis with minute mass requirements (pg). In addition the uv transparency of the fused silica permits online detection with precise and accurate quantification. Finally the free solution format eliminates the need for 0003-2697/92 $5.00 Copyright © 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.
CAPILLARY ELECTROPHORESIS-ISOELECTRIC FOCUSING polymerization steps in preparation for electrophoresis and permits complete control over the composition of the separation buffer. In general, C E - I E F poses two challenges. Since detection is on-line, focusing must be performed in the absence of endosmotic flow; however, once established, the ampholyte gradient must be mobilized past the online detector in order to observe the focused proteins and pH gradient. Various strategies have been published (3,5,6,8,9), and the yardstick for effectiveness is (or should be) the demonstration of the linearity of the pH gradient, the accuracy and precision of the pI estimation, and the degree of resolution achieved. Indeed, the majority of the literature on C E - I E F is dedicated to the resolution of these two challenges (see Discussion for more complete comparison). This paper describes capillary electrophoresis isoelectric focusing and demonstrates its utility in the measurement of the pI of proteins at the extremes of a pH 3-10 gradient. In addition, the linearity, accuracy, precision, and resolution observed for this technique at the extremes of the pH gradient is discussed. MATERIALS AND METHODS In this study Servalyt 3-10 was purchased from Serva (Serva Fine Chemicals, Paramus, NJ), and methylcellulose (1500 cps at 2%) from Sigma (Sigma Chemicals, St. Louis, MO). Carbonic anhydrase (pI 5.9),/~-lactoglobulin A (pI 5.1), and ribonuclease A (RNase A, pI 9.45) were obtained in the highest purity from Sigma. Unsulfated cholecystikinin (CCK) flanking peptide (pI 2.75) was obtained from Peninsula Labs (Belmont, CA.). Ribonuclease T1 (RNase TO and mutants from Aspergillus oryzae and ribonuclease from Bacillus amylofasciens (RNase ba) were kind gifts of Drs. C. Nick Pace and Bret Shirley at Texas A&M University. All other reagents used were of analytical grade. Capillary electrophoresis was performed on the Applied Biosystems Model 270A-HT capillary electrophoresis system (Foster City, CA) using a dimethylpolysiloxane (DB-1)-coated capillary with 50 #m internal diameter and 0.05 #m coating thickness (J&W Scientific, Folsom, CA). The total length of the capillary used in this study was 72 cm and the length from the sample injection end to the detector was 50 cm. The output of the 270A-HT was plotted and integrated by a HewlettPackard Model 3396 integrator. The protocol used to achieve the separations requires the accurate application of precise vacuum to the capillary in order to fill it with the correct amount of electrode buffers, ampholytes, and sample. The system was programmed to first fill the capillary with 20 mM NaOH in 0.4% methylcellulose (8 min of 20 in. Hg vacuum), followed by a section of 0.5% Servalyt 3-10 in 0.4% methylcellulose (4.5 rain of 20 in. Hg vacuum). An ali-
85
quot of sample and marker (dissolved in Milli-Q water) was loaded separately into the capillary (30 s of 5 in. Hg vacuum each) and another section of 0.5% Servalyt 3-10 with 0.4% methylcellulose (0.1 min of 20 in. Hg vacuum) was added to the capillary. As illustrated in Fig. 1, at this point the interface between the NaOH solution and Servalyt solution was positioned on the sample side of the detector. High voltage, usually 30 kV, was then applied for 6 min to focus the ampholyte and proteins. After focusing, the system was programmed to apply a precisely regulated vacuum (5 in. Hg), while maintaining the high voltage. The applied vacuum caused the focused zone of proteins in the capillary to flow pass the detector while the voltage maintained the pH gradient and zone sharpness even in the presence of the distorting effects of laminar flow. The protein bands were detected at 280 nm as they passed the detector under the influence of vacuum and voltage. RESULTS
Linearity of the pHgradient. The linearity of the pH gradient was determined by running RNase A, carbonic anhydrase, ~-lactoglobulin A, and CCK-flanking peptide as markers. Figure 2 shows the electropherogram of a 20-nl injection of 100 #g/ml of each protein (2 ng each loaded). The proteins were focused and mobilized as described under Materials and Methods. After mobilization and detection at 280 nm, the relative mobility of each protein was calculated (see below) and plotted against the published isoelectric point (Fig. 2, inset). The double peak before the RNase A in Fig. 2 represents the interface between cathodic buffer and ampholyte solution. The peak after CCK-flanking peptide is the interface of anodic buffer and ampholyte solution. These two interface peaks are characteristics of this procedure and can be used to indicate the beginning and the end of the pH gradient. The times at which these peaks appear after mobilization is initiated (mobilization vacuum and voltage applied) is used to calculate the relative mobility (Rm) as follows: Rm
=
to t f - to
tp
where tp is the mobilization time of the protein peak of interest, t o (origin) is the mobilization time of the cathodic buffer interface, and tf (front) is the mobilization time of the anodic buffer interface. The linearity of the pH gradient was determined by plotting the pI vs relative mobility (Fig. 2, inset). This plot shows this method can provide linear range between pH 2.75 and 9.5 (CCK-flanking peptide and RNase A, respectively). Resolution. RNase T1 from A. oryzae (wild-type) is a protein with a calculated pI of 2.9. Because of its ex-
86
CHEN AND WIKTOROWICZ Detector
~
Hydrodynamic flow LOAD CATHOLYTE
NaO
Detector
NaOH
~
~:)7..~:~.~7~'..~:~`~.7.;:.:~.~.`~.~:~`~i..;?.:~:.('~:~.ii/~:~.~i~:~.~:.:'.~:'?~.!.~Ampholyte :..~7.~.~`~(~(~.AMPHOLYTE ::.LOAD 7i:~.~.~: Detector
NaOH
Hydrodynamic flow
~
Hydrodynamic flow
LOAD ~i:~:~i.(i.;~:~`.~``.:.:.:. ;~:i;~i:;.?i~:~.~:~i:~.:. ~i:.)~.~:~i.~:~i.~i:.:.;;i:./`i:./`~i:~:~:i};). Sample .i:~:::;i(~(:. ).I SAMPLE Detector
--
Hydrodynamic flow "INSULATE" SAMPLE
Detector
•',,IIII......................Electrical
field FOCUS
Detector
~
Hydrodynamic flow
'",llll...................... Electrical field ~fl,
NaOH H i
:::::::::::::::: ::::::::::::;::.':::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::.',::;::.~?::::= ;:::::::::::;:::: . ........... ' : : . , v : : t "............. . ' . , ' : : . , , I.:::.,'.':.'.:.~ . . . . . . . . . . . . . . :.': ! ".'.'.'.'.,'::.:. . . . . . . . ...................... ............. ".:.:.:
