ANALYTICAL
BIOCHEMISTRY
204,
1033106
(19%)
Optimization and Validation of Analytical Conditions for Bovine Serum Albumin Using Capillary Electrophoresis Poonam Department
Received
G. Pande, Ranjani V. Nellore, and Hitesh of Pharmaceutics, School of Pharmacy. University
December
R. Bhagatl of Mar.yland
at Baltimore,
Baltimore,
Maryland
21201
16, 1991
A quick and reproducible capillary electropboresis method was optimized and validated for the assay of bovine serum albumin (BSA). The effects of various parameters such as pH of buffer, concentration of buffer, capillary dimensions, use of coated capillaries, and additives such as surfactants and protein solubilizers were evaluated. The capillary coatings or additives did not give any advantage in reducing the surface adsorption of BSA on the capillary walls. The optimized conditions include use of borate buffer, pH 8.5 having a concentration of 150 mM in a 27 cm capillary with an aperture window of 100 X 200 Mm for detection. The optimized method for the detection of BSA was validated. The interday and intraday coefficient of variation was not greater than 7.59% at BSA concentrations of 25-1000 pug/ml. The method developed was reproducible and accurate. ‘cl 1992 Academic Press, Inc.
Currently, rapid progress is being made in the field of biotechnology and pharmaceutical interest in that field is increasing. Analytical methods for the detection of proteins include HPLC, RPHPLC, GPC’ (PAGE and SDS-PAGE), SEC, and other tests for specific compounds. An ideal analytical technique for proteins should be quick, sensitive, accurate, and reproducible. Gel electrophoresis, which is a well-characterized technique and is widely used, has the disadvantages of being time-consuming, labor intensive, and unsuitable for quantitative analysis of a large number of samples. In the past few years, capillary zone electrophoresis (CZE) has emerged as a powerful, new method for rapid separation and analysis of charged molecules such as proteins
and peptides. In the present study, we developed a quick and reproducible capillary electrophoretic assay method for bovine serum albumin (BSA). In our laboratory, BSA is frequently used as a model protein especially during evaluation of controlled drug delivery systems such as biodegradable polymeric microspheres. The molecular weight of BSA is about 66,000 and its pK, is 4.7 (1). In CZE, charged molecules migrate in buffer solution through a capillary under the influence of an applied electrical field. The migration of these molecules depends on a number of factors, including the charge to mass ratio of the molecule, the electrical field gradient across the capillary, the pH and ionic strength of the running buffer, and the capillary dimensions (2-4). However, proteins get adsorbed onto the walls of uncoated fused silica capillaries. The silica walls have free silanol groups that have a negative charge in aqueous solutions above a pH value of 2. Proteins can have a net negative charge if the pH of the solutions is adjusted to 2 units above their isoelectric point. If the protein molecules have a net negative charge, then they should be repelled from the wall surface and their adsorption should be minimized (5). Protein adsorption can also be decreased by modifying the capillary wall characteristics with coatings (6,7) or by using additives (8) in the separation buffer or the sample. Additives can act either by getting preferentially adsorbed onto the capillary walls or by solubilizing the protein to a greater extent thereby decreasing its interaction with the capillary walls. We examined the effect of some of these parameters for the analysis of BSA and optimized the analytical conditions to be used. EXPERIMENTAL
1 To whom correspondence should be addressed. ’ Abbreviations used: GPC, gel permeation chromatography; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SEC, size exclusion chromatograph; CZE, capillary zone electrophoresis; BSA, bovine serum albumin; CV, coefficient of variation. 0003.2697/92 $5.00 Copyright 1211992 by Academic Press, All rights of reproduction in any form
SECTION
Materials
Bovine serum albumin, boric acid, sodium borate (decahydrate), dibasic sodium phosphate (heptahy103
Inc. reserved.
104
PANDE,
NELLORE,
drate), monobasic sodium phosphate (monohydrate), 0.1 N sodium hydroxide solution, urea, sodium lauryl sulfate, lecithin, sodium deoxycholate, reduced Triton X-100, triethylamine, ethylene glycol, and sodium azide. All the chemicals were of analytical grade and obtained from Sigma. Sterile water for injection was used for preparation of all solutions. Instrumentation Capillary electrophoresis was performed in an automated PACE 2100 unit (Beckman Instruments, Inc., CA) controlled by System Gold software (Beckman Instruments, Inc.). The PACE unit is fitted with a 30 kV high voltage power supply. The current limit is 250 PA. The electrophoresis can be performed at constant voltage, current, or power. A fixed wavelength detector is connected at the cathodic end of the capillary. The capillary cartridges were supplied by Beckman and fitted with a 75 /*rn i.d. fused silica column, 27 or 57 cm in length (20 or 50 cm effective length up to the detector). The aperture window for detection was 50 or 100 pm wide. Coated capillaries were supplied by J & W Scientific and were built into cartridges using a capillary cartridge rebuild kit from Beckman Instruments, Inc. Methods The factors affecting the analytical parameters such as migration time, peak shape, peak area, and peak height were studied. These included pH, concentration, capillary dimensions, and use of additives. The temperature during operation was maintained at 20 + O.l”C. Constant voltage of 12 kV was applied for the electrophoresis. Sample injection was done by using pressure for 3 s (about 15 nl). Each sample injection was done in duplicate. The detection was carried out at 214 nm. For the assay development, sample solutions of 100 pglml were freshly prepared in isotonic phosphate buffer having a pH of 7.4, unless indicated otherwise. The buffer systems used for the runs are listed in Table 1. The additives added to the sample solutions were reduced Triton X-100 (0.15%), triethylamine (l%), urea (7 M), ethylene glycol (20%), and a mixture of urea (5%) and ethylene glycol (10%). The fused silica capillaries were washed with 0.1 N sodium hydroxide, water, and running buffer for 5 min each prior to the runs. This process was repeated after every 12-16 runs (corresponding to duplicate injections of six to eight samples) and the buffer vials were changed. The coated capillaries evaluated included DB-1 (100 pm i.d., 0.2 wrn film thickness), DB-17 (50 pm i.d., 0.05 pm film thickness and 100 pm i.d., 0.1 pm film thickness) and DB-wax (100 wrn i.d., 0.1 pm film thickness). These were 27 cm in length, with 20 cm in length to the detector window. The buffers tried for the coated capil-
AND
BHAGAT TABLE Buffer
Systems
Evaluated
1 for
Electrophoresis Concentration
Buffer Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate
(additive)
PH
(mM
8.0 8.0 8.0 8.5 8.5
50 150 300
9.0 (30 mM SLS) (1% Tween 80) (0.1% lecithin) (0.6% deoxycholate) (5% urea and 10% ethylene (1% triethylamine)
8.0
glycoll
150
300 300
9.0
100 100
8.0 8.0 8.0 8.0
300 300 300 300
laries included borate buffer 150 mM at pH of 8.0, 8.5, and 9.0 and phosphate buffer 150 mM at pH 2.5. The other conditions varied included reverse polarity and “sandwich injection” (sample injected under pressure followed by buffer injected under pressure to push sample plug toward the detector and then electrophoresis). Using the optimized conditions for the analysis, the standard curves were generated with the sample concentrations ranging from 10 to 1000 pug/ml on three different days for interday variability and three runs were performed during a day for intraday variability. Peak area was plotted against the concentration. Standard solutions were freshly prepared in phosphate buffer, pH 7.4, on the day of the runs. The accuracy and reproducibility of the assay was determined by calculating the percentage CV. RESULTS
AND
Optimization
DISCUSSION
of the Analytical
Conditions
The samples were prepared in isotonic phosphate buffer, pH 7.4, containing sodium azide since our application involved the release studies of BSA from controlled drug delivery systems in that buffer. Sodium azide was subsequently deleted from the release system since it was found to interfere with the protein analysis and its omission did not give rise to any microbial contamination in the release media. The electropherograms of BSA sample in phosphate buffer and in water are shown in Fig, 1. The peak areas and shape are not significantly different. However, the migration time is delayed in the buffer due to the presence of the ions of the buffer salts that decrease the mobility of the BSA in the sample plug. The total length and the effective length of the capillary influence the migration time and the peak shape. At
CAPILLARY
ELECTROPHORETIC
ASSAY
FOR
BOVINE
SERUM
105
ALBUMIN
0 11
0 050
010
0 042
0 05
VI = c ;
; E
0 026 0018
m 9
Sample
0 034
0 01 0
I” phosphate
10 E 6 0 11 b 0 Lo 9 0
r\
‘u L
J
\\ Sample
-ooo;t‘-:7--1
05
10
07 06 05
150 mM
04
0 03
0 002
00
z 0 08 c
20
15
25
in water
0 01 i
I
30
35
300 mM
0 02 00
40
45
50
(100 pg/ml)
run
in bo-
05
10
15
FIG. 1. Electropherograms rate buffer, pH 8.0, 300 mM.
of BSA
samples
constant voltage, the longer the capillary, the smaller is the electrical field gradient (V/cm). This leads to slower migration, greater zone broadening, and wider peak. For our application, the 27 cm capillary was found suitable. Comparing capillaries with 50 X 200 Frn aperture to those with 100 X 200-pm aperture for detection, better signal-to-noise ratio was obtained with the latter one. Hence the sensitivity is greater with the bigger aperture window. The pH studies were performed at 12 kV using 300 mM buffer at pH values of 8.0, 8.5, and 9.0. The migration times were not significantly different but better sensitivity, smaller peak width, and less tailing of the peak were obtained at a pH of 8.5 (Fig. 2). The pH of the buffer affects the charges on the solute molecules and on the capillary walls. At the pH values studied, the net charge on the protein molecule is negative. However, BSA does get adsorbed on to the capillary walls indicated by the tailing peaks. This may be due to the distribution of the positive and negative charges on the mole-
pH 5 5
OOl'L
00
FIG. BSA
05
1 10
15
20
1 25 30 Time (min)
2. Effect of pH of run buffer sample (100 pg/ml in phosphate
(borate) buffer,
35
" 40
45
on electropherogram pH 7.4).
3.0
25
30
35
40
45
50
Ttme (mn)
Time (mln)
J 50
of
FIG. 3. pherogram
Effect of concentration of BSA (100 Kg/ml
of run buffer (borate) in phosphate buffer, pH
on electro7.4).
cule and their orientation in the capillary. If the positive sites on the surface of the protein molecule are exposed to the capillary, then adsorption can take place. The results indicate least adsorption of BSA at a pH of 8.5. Some conformational changes in the BSA molecule at that pH might have led to a change in the distribution and orientation of charges on the molecule, resulting in reduced adsorption and, therefore, greater sensitivity and lesser zone broadening. The effect of concentration was examined using 50, 150, and 300 mM pH 8.0 buffer. The concentration affects the mobility of the charged molecules across the capillary. The peak shape obtained with 50 and 150 mrvr buffers was the same whereas a broad peak was observed with the 300 mM buffer (Fig. 3). This indicated that greater adsorption of the protein occurred at higher concentration. This was in contrast to the results obtained by Mclaughlin et al. (3) where they observed that higher ionic strength of the buffer reduced wall interactions of the proteins. Although fastest migration of BSA was obtained at a buffer concentration of 50 mM, the peak overlapped with the solvent front. As the concentration of the buffer decreases, the resistance to the movement of ions decreases and the ionic mobility increases. This decreases the migration time. The tailing of the peak at higher ionic strength would be due to slower migration and greater zone broadening. The additives were studied in an attempt to reduce the capillary adsorption of BSA since they are reported to get preferentially adsorbed on to the capillary walls. However, the sensitivity was greatly reduced and in most cases, multiple and distorted peaks were obtained. The coated capillaries did not give any promising results. No peak for BSA was obtained although the runs were performed for 30 min indicating that BSA did not reach the detector. The preliminary conclusion is that there is some interaction occurring between the
106
PANDE.
NELLORE,
AND
BHAGAT
0.007
TABLE Intraday
0 006 m .E
0.005
: : x‘;
0.004
v1 9
Concentration (pg/ml)
and
Interday
0.002 0.001 -0.000' 0.0
0.5
1 1 0
h 2.0
1 .5
t 1 2.5 3.0 Time (min)
3.5
4.0
4.5
Coefficients
Intraday
25 50 100 250 500 1000
0.003
2 of Variation
% CV
Interday
4.75 2.39 0.73 1.29 0.99 0.30
% CV 3.81 7.59 4.24 0.44 3.90 0.86
1 5.0
ACKNOWLEDGMENTS FIG. 4. Electropherogram buffer, pH 7.4) run in borate
of BSA buffer,
sample (100 pg/ml pH 8.5, 150 mM.
in phosphate
This project was funded by a BRSG Grant. edge Ms. Patricia Albuquerque for her help.
The
authors
acknowl-
REFERENCES
coatings of BSA.
Validation
and the protein
of the Optimized
that prevents
the migration
Method
The assay technique was validated using pH 8.5 buffer having a concentration of 150 mM (Fig. 4). The migration time for BSA under these conditions was around 2.2 min. The interday and intraday assay of BSA was reproducible and fairly accurate with a CV of not more than 7.59% at concentrations of 25-1000 wg/ml (Table 2). These differences in response could be attributed to instrumental variations and to slight differences in the sample concentrations prepared for different runs.
Putnam, F. W. (1975) The Plasma Proteins, demic Press, New York. 2. Jones, H. K., and Ballou, N. E. (1990) Anal. 2490. 1.
3. Mclaughlin, G., Palmieri, niques in Protein Chemistry demic Press, New York.
R., and Anderson, (J. Villafranca,
pp. Chem
58-183,
Aca-
62,
2484-
K. (1990) in TechEd.), pp. 3-19, Aca-
4. Lambert, W. J., and Middleton, D. L. (1990) Anal. Chen. 62, 1585-1587. 5. Lauer, H. H., and McManigill, D. (1986) Anal. Chem. 58, 166170. 6. Towns, J. K., and Regnier, F. E. (1991) Anal. Chem. 63, 11261132. 7. Bruin, G. J. M., Huisden, R., Kraak, J. C., and Poppe, H. (1989) J. Chromatogr. 480,339-349. 8. Zhu, M., Rodriguez, R., Hansen, D., and Wehr, T. (1990) J. Chromatogr. 516, 123-131.
BIOCHEMISTRY
204,
1033106
(19%)
Optimization and Validation of Analytical Conditions for Bovine Serum Albumin Using Capillary Electrophoresis Poonam Department
Received
G. Pande, Ranjani V. Nellore, and Hitesh of Pharmaceutics, School of Pharmacy. University
December
R. Bhagatl of Mar.yland
at Baltimore,
Baltimore,
Maryland
21201
16, 1991
A quick and reproducible capillary electropboresis method was optimized and validated for the assay of bovine serum albumin (BSA). The effects of various parameters such as pH of buffer, concentration of buffer, capillary dimensions, use of coated capillaries, and additives such as surfactants and protein solubilizers were evaluated. The capillary coatings or additives did not give any advantage in reducing the surface adsorption of BSA on the capillary walls. The optimized conditions include use of borate buffer, pH 8.5 having a concentration of 150 mM in a 27 cm capillary with an aperture window of 100 X 200 Mm for detection. The optimized method for the detection of BSA was validated. The interday and intraday coefficient of variation was not greater than 7.59% at BSA concentrations of 25-1000 pug/ml. The method developed was reproducible and accurate. ‘cl 1992 Academic Press, Inc.
Currently, rapid progress is being made in the field of biotechnology and pharmaceutical interest in that field is increasing. Analytical methods for the detection of proteins include HPLC, RPHPLC, GPC’ (PAGE and SDS-PAGE), SEC, and other tests for specific compounds. An ideal analytical technique for proteins should be quick, sensitive, accurate, and reproducible. Gel electrophoresis, which is a well-characterized technique and is widely used, has the disadvantages of being time-consuming, labor intensive, and unsuitable for quantitative analysis of a large number of samples. In the past few years, capillary zone electrophoresis (CZE) has emerged as a powerful, new method for rapid separation and analysis of charged molecules such as proteins
and peptides. In the present study, we developed a quick and reproducible capillary electrophoretic assay method for bovine serum albumin (BSA). In our laboratory, BSA is frequently used as a model protein especially during evaluation of controlled drug delivery systems such as biodegradable polymeric microspheres. The molecular weight of BSA is about 66,000 and its pK, is 4.7 (1). In CZE, charged molecules migrate in buffer solution through a capillary under the influence of an applied electrical field. The migration of these molecules depends on a number of factors, including the charge to mass ratio of the molecule, the electrical field gradient across the capillary, the pH and ionic strength of the running buffer, and the capillary dimensions (2-4). However, proteins get adsorbed onto the walls of uncoated fused silica capillaries. The silica walls have free silanol groups that have a negative charge in aqueous solutions above a pH value of 2. Proteins can have a net negative charge if the pH of the solutions is adjusted to 2 units above their isoelectric point. If the protein molecules have a net negative charge, then they should be repelled from the wall surface and their adsorption should be minimized (5). Protein adsorption can also be decreased by modifying the capillary wall characteristics with coatings (6,7) or by using additives (8) in the separation buffer or the sample. Additives can act either by getting preferentially adsorbed onto the capillary walls or by solubilizing the protein to a greater extent thereby decreasing its interaction with the capillary walls. We examined the effect of some of these parameters for the analysis of BSA and optimized the analytical conditions to be used. EXPERIMENTAL
1 To whom correspondence should be addressed. ’ Abbreviations used: GPC, gel permeation chromatography; SDSPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; SEC, size exclusion chromatograph; CZE, capillary zone electrophoresis; BSA, bovine serum albumin; CV, coefficient of variation. 0003.2697/92 $5.00 Copyright 1211992 by Academic Press, All rights of reproduction in any form
SECTION
Materials
Bovine serum albumin, boric acid, sodium borate (decahydrate), dibasic sodium phosphate (heptahy103
Inc. reserved.
104
PANDE,
NELLORE,
drate), monobasic sodium phosphate (monohydrate), 0.1 N sodium hydroxide solution, urea, sodium lauryl sulfate, lecithin, sodium deoxycholate, reduced Triton X-100, triethylamine, ethylene glycol, and sodium azide. All the chemicals were of analytical grade and obtained from Sigma. Sterile water for injection was used for preparation of all solutions. Instrumentation Capillary electrophoresis was performed in an automated PACE 2100 unit (Beckman Instruments, Inc., CA) controlled by System Gold software (Beckman Instruments, Inc.). The PACE unit is fitted with a 30 kV high voltage power supply. The current limit is 250 PA. The electrophoresis can be performed at constant voltage, current, or power. A fixed wavelength detector is connected at the cathodic end of the capillary. The capillary cartridges were supplied by Beckman and fitted with a 75 /*rn i.d. fused silica column, 27 or 57 cm in length (20 or 50 cm effective length up to the detector). The aperture window for detection was 50 or 100 pm wide. Coated capillaries were supplied by J & W Scientific and were built into cartridges using a capillary cartridge rebuild kit from Beckman Instruments, Inc. Methods The factors affecting the analytical parameters such as migration time, peak shape, peak area, and peak height were studied. These included pH, concentration, capillary dimensions, and use of additives. The temperature during operation was maintained at 20 + O.l”C. Constant voltage of 12 kV was applied for the electrophoresis. Sample injection was done by using pressure for 3 s (about 15 nl). Each sample injection was done in duplicate. The detection was carried out at 214 nm. For the assay development, sample solutions of 100 pglml were freshly prepared in isotonic phosphate buffer having a pH of 7.4, unless indicated otherwise. The buffer systems used for the runs are listed in Table 1. The additives added to the sample solutions were reduced Triton X-100 (0.15%), triethylamine (l%), urea (7 M), ethylene glycol (20%), and a mixture of urea (5%) and ethylene glycol (10%). The fused silica capillaries were washed with 0.1 N sodium hydroxide, water, and running buffer for 5 min each prior to the runs. This process was repeated after every 12-16 runs (corresponding to duplicate injections of six to eight samples) and the buffer vials were changed. The coated capillaries evaluated included DB-1 (100 pm i.d., 0.2 wrn film thickness), DB-17 (50 pm i.d., 0.05 pm film thickness and 100 pm i.d., 0.1 pm film thickness) and DB-wax (100 wrn i.d., 0.1 pm film thickness). These were 27 cm in length, with 20 cm in length to the detector window. The buffers tried for the coated capil-
AND
BHAGAT TABLE Buffer
Systems
Evaluated
1 for
Electrophoresis Concentration
Buffer Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate Borate
(additive)
PH
(mM
8.0 8.0 8.0 8.5 8.5
50 150 300
9.0 (30 mM SLS) (1% Tween 80) (0.1% lecithin) (0.6% deoxycholate) (5% urea and 10% ethylene (1% triethylamine)
8.0
glycoll
150
300 300
9.0
100 100
8.0 8.0 8.0 8.0
300 300 300 300
laries included borate buffer 150 mM at pH of 8.0, 8.5, and 9.0 and phosphate buffer 150 mM at pH 2.5. The other conditions varied included reverse polarity and “sandwich injection” (sample injected under pressure followed by buffer injected under pressure to push sample plug toward the detector and then electrophoresis). Using the optimized conditions for the analysis, the standard curves were generated with the sample concentrations ranging from 10 to 1000 pug/ml on three different days for interday variability and three runs were performed during a day for intraday variability. Peak area was plotted against the concentration. Standard solutions were freshly prepared in phosphate buffer, pH 7.4, on the day of the runs. The accuracy and reproducibility of the assay was determined by calculating the percentage CV. RESULTS
AND
Optimization
DISCUSSION
of the Analytical
Conditions
The samples were prepared in isotonic phosphate buffer, pH 7.4, containing sodium azide since our application involved the release studies of BSA from controlled drug delivery systems in that buffer. Sodium azide was subsequently deleted from the release system since it was found to interfere with the protein analysis and its omission did not give rise to any microbial contamination in the release media. The electropherograms of BSA sample in phosphate buffer and in water are shown in Fig, 1. The peak areas and shape are not significantly different. However, the migration time is delayed in the buffer due to the presence of the ions of the buffer salts that decrease the mobility of the BSA in the sample plug. The total length and the effective length of the capillary influence the migration time and the peak shape. At
CAPILLARY
ELECTROPHORETIC
ASSAY
FOR
BOVINE
SERUM
105
ALBUMIN
0 11
0 050
010
0 042
0 05
VI = c ;
; E
0 026 0018
m 9
Sample
0 034
0 01 0
I” phosphate
10 E 6 0 11 b 0 Lo 9 0
r\
‘u L
J
\\ Sample
-ooo;t‘-:7--1
05
10
07 06 05
150 mM
04
0 03
0 002
00
z 0 08 c
20
15
25
in water
0 01 i
I
30
35
300 mM
0 02 00
40
45
50
(100 pg/ml)
run
in bo-
05
10
15
FIG. 1. Electropherograms rate buffer, pH 8.0, 300 mM.
of BSA
samples
constant voltage, the longer the capillary, the smaller is the electrical field gradient (V/cm). This leads to slower migration, greater zone broadening, and wider peak. For our application, the 27 cm capillary was found suitable. Comparing capillaries with 50 X 200 Frn aperture to those with 100 X 200-pm aperture for detection, better signal-to-noise ratio was obtained with the latter one. Hence the sensitivity is greater with the bigger aperture window. The pH studies were performed at 12 kV using 300 mM buffer at pH values of 8.0, 8.5, and 9.0. The migration times were not significantly different but better sensitivity, smaller peak width, and less tailing of the peak were obtained at a pH of 8.5 (Fig. 2). The pH of the buffer affects the charges on the solute molecules and on the capillary walls. At the pH values studied, the net charge on the protein molecule is negative. However, BSA does get adsorbed on to the capillary walls indicated by the tailing peaks. This may be due to the distribution of the positive and negative charges on the mole-
pH 5 5
OOl'L
00
FIG. BSA
05
1 10
15
20
1 25 30 Time (min)
2. Effect of pH of run buffer sample (100 pg/ml in phosphate
(borate) buffer,
35
" 40
45
on electropherogram pH 7.4).
3.0
25
30
35
40
45
50
Ttme (mn)
Time (mln)
J 50
of
FIG. 3. pherogram
Effect of concentration of BSA (100 Kg/ml
of run buffer (borate) in phosphate buffer, pH
on electro7.4).
cule and their orientation in the capillary. If the positive sites on the surface of the protein molecule are exposed to the capillary, then adsorption can take place. The results indicate least adsorption of BSA at a pH of 8.5. Some conformational changes in the BSA molecule at that pH might have led to a change in the distribution and orientation of charges on the molecule, resulting in reduced adsorption and, therefore, greater sensitivity and lesser zone broadening. The effect of concentration was examined using 50, 150, and 300 mM pH 8.0 buffer. The concentration affects the mobility of the charged molecules across the capillary. The peak shape obtained with 50 and 150 mrvr buffers was the same whereas a broad peak was observed with the 300 mM buffer (Fig. 3). This indicated that greater adsorption of the protein occurred at higher concentration. This was in contrast to the results obtained by Mclaughlin et al. (3) where they observed that higher ionic strength of the buffer reduced wall interactions of the proteins. Although fastest migration of BSA was obtained at a buffer concentration of 50 mM, the peak overlapped with the solvent front. As the concentration of the buffer decreases, the resistance to the movement of ions decreases and the ionic mobility increases. This decreases the migration time. The tailing of the peak at higher ionic strength would be due to slower migration and greater zone broadening. The additives were studied in an attempt to reduce the capillary adsorption of BSA since they are reported to get preferentially adsorbed on to the capillary walls. However, the sensitivity was greatly reduced and in most cases, multiple and distorted peaks were obtained. The coated capillaries did not give any promising results. No peak for BSA was obtained although the runs were performed for 30 min indicating that BSA did not reach the detector. The preliminary conclusion is that there is some interaction occurring between the
106
PANDE.
NELLORE,
AND
BHAGAT
0.007
TABLE Intraday
0 006 m .E
0.005
: : x‘;
0.004
v1 9
Concentration (pg/ml)
and
Interday
0.002 0.001 -0.000' 0.0
0.5
1 1 0
h 2.0
1 .5
t 1 2.5 3.0 Time (min)
3.5
4.0
4.5
Coefficients
Intraday
25 50 100 250 500 1000
0.003
2 of Variation
% CV
Interday
4.75 2.39 0.73 1.29 0.99 0.30
% CV 3.81 7.59 4.24 0.44 3.90 0.86
1 5.0
ACKNOWLEDGMENTS FIG. 4. Electropherogram buffer, pH 7.4) run in borate
of BSA buffer,
sample (100 pg/ml pH 8.5, 150 mM.
in phosphate
This project was funded by a BRSG Grant. edge Ms. Patricia Albuquerque for her help.
The
authors
acknowl-
REFERENCES
coatings of BSA.
Validation
and the protein
of the Optimized
that prevents
the migration
Method
The assay technique was validated using pH 8.5 buffer having a concentration of 150 mM (Fig. 4). The migration time for BSA under these conditions was around 2.2 min. The interday and intraday assay of BSA was reproducible and fairly accurate with a CV of not more than 7.59% at concentrations of 25-1000 wg/ml (Table 2). These differences in response could be attributed to instrumental variations and to slight differences in the sample concentrations prepared for different runs.
Putnam, F. W. (1975) The Plasma Proteins, demic Press, New York. 2. Jones, H. K., and Ballou, N. E. (1990) Anal. 2490. 1.
3. Mclaughlin, G., Palmieri, niques in Protein Chemistry demic Press, New York.
R., and Anderson, (J. Villafranca,
pp. Chem
58-183,
Aca-
62,
2484-
K. (1990) in TechEd.), pp. 3-19, Aca-
4. Lambert, W. J., and Middleton, D. L. (1990) Anal. Chen. 62, 1585-1587. 5. Lauer, H. H., and McManigill, D. (1986) Anal. Chem. 58, 166170. 6. Towns, J. K., and Regnier, F. E. (1991) Anal. Chem. 63, 11261132. 7. Bruin, G. J. M., Huisden, R., Kraak, J. C., and Poppe, H. (1989) J. Chromatogr. 480,339-349. 8. Zhu, M., Rodriguez, R., Hansen, D., and Wehr, T. (1990) J. Chromatogr. 516, 123-131.
