Evaluating Multidimensional Liquid Chromatographic Separations … · Evaluating Multidimensional...
1
Evaluating Multidimensional Liquid Chromatographic Separations of Complicated Protein Mixtures Kevin M. Millea 2 , Ignatius J. Kass 1 , John C. Gebler 1 , Ira S. Krull 2 , Scott J. Berger 1 1 Life Sciences R&D, Waters Corporation, Milford, MA; 2 Department of Chemistry, Northeastern University, Boston, MA Overview Instrumentation Results Conclusions • Liquid phase protein separations are becoming viable alternatives to gel based methodologies for resolving sample components within a proteomic mixture. • Liquid phase chromatographic separations offer a wide variety of separation mechanisms, a high degree of automation and easy coupling with mass spectrometry • In this work, a 2-D LC separation is coupled with ESI-TOF mass spectrometry to extract component level information from a complex proteomic sample. • Automated spectral deconvolution of the LC/MS data set provided the capability to assess chromatographic behavior of individual sample components. • These approaches were combined to analyze the distribution and reproducibility of 2-D (IEX/RP) separations of E. coli cytosol using either a step or linear first dimension IEX gradient. Figure 1: Multidimensional LC/ESI-TOF MS Configurations. An Alliance ® 2D Bioseparations System (Waters) was used for online IEX/ RP separations coupled to a Micromass ® LCT™ ESI-TOF MS (Waters). Coupling the 2-D separation with online MS detection permitted the resolution of individual sample components. The second dimension RP dimension was configured using dual trapping columns, and a single analytical column. Two 10-port 2-position valves were used to direct solvent flow for trapping of analytes, column washing, and separation of sample components. A 0 to 500 mM NaCl eight-step or linear gradient was employed to develop the IEX first dimension, with a 1M column washing step following the gradient. Each second dimension reversed phase separation was effected by an water/acetonitrile gradient in a background of 0.1% formic acid/0.005% TFA. Positive mode ESI-MS data were obtained over a m/z range of 700-3000 at a 0.5 Hz acquisition rate. • A complex proteomic sample was fractionated using multidimensional chromatography combined with online MS detection of intact proteins. • Chromatographic behavior of the individual sample components was elucidated by automated time segmentation and spectral deconvolution of the protein LC/MS data set. (see poster P-8-M for details). • This methodology was applied to reveal the chromatographic differences of separating sample components using either a linear or step gradient in the first separation dimension. • The data indicate the linear IEX first dimension gradient reduced dispersion of components relative to the step gradient separation, and that adequate run-run reproducibility could be routinely achieved using multidimensional protein LC. ©2005 Waters Corporation 720001106EN Figure 2: Comprehensive 2-D (IEX/RP) Chromatographic Separation of E. coli Cytosol Employing a Linear (LEFT SIDE) or Step (RIGHT SIDE) Gradient in the First Dimension. Panel A: First and second dimension gradient profiles for the separation of E. coli cytosol. Panel B: TIC traces of three replicate injections of E. coli cytosol. Panel C: Expanded view of two reversed phase cycles demonstrating the complexity of the separation. Panel D: Analysis of the processed neutral mass data reveals individual overlapping component profiles. Segments of the data comprising ten spectral scans (20 sec) were summed, and deconvoluted using a maximum entropy based approach (See Poster P-8-M for more details on this approach). Processed data were obtained over a mass range of 7,000 to 40,000 Da with a 3 Da mass resolution. The processed data is displayed in a 2-D Gel bubble plot display (x-axis: retention time, y-axis: deconvoluted mass; bubble diameter: centroided intensity). In this display the first run is visualized in green, the second visualized in red, and overlapping components displayed in orange. A B C C D D B A Figure 3: Distribution of selected analytes demonstrating first dimension carryover. Dispersion of analyte into adjacent reversed phase cycles is characteristic of first dimension peak splitting. The distributions of two components (masses 11049 Da , and 9970 Da) are shown as the percent intensity distributed over several reversed phase cycles. Cycle n is the first cycle in which the component was detected, and n+1 constitutes the following reversed phase cycle. Triplicate analyses are shown for both gradient modes. The first component (11049 Da) illustrates splitting observed only in the step gradient mode, while the second component (9970 Da) show splitting common to both step and linear gradients. Figure 4. Distribution of selected analytes demonstrating trap column carryover. The distributions of two components (masses 11036 Da and 31065 Da) demonstrate dispersion into alternating reversed phase cycles. This is characteristic of carryover (ghosting) on the second dimension trap column . Figure 5: Distribution of selected analytes demonstrating extensive carryover into adjacent fractions. Two components (masses 9191 Da, and 10387 ) are shown to demonstrate reproducibly poor chromatographic behavior in both separation dimensions in step and linear modes. LINEAR FIRST DIMENSION IEX GRADIENT STEP FIRST DIMENSION IEX GRADIENT 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 11036 Da Linear Gradient Step Gradient 0.0 20.0 40.0 60.0 80.0 100.0 31065 Da 0.0 20.0 40.0 60.0 80.0 100.0 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 % Total Intensity % Total Intensity 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 n n+1 n+2 n+3 n+4 11036 Da Linear Gradient Step Gradient 0.0 20.0 40.0 60.0 80.0 100.0 31065 Da 0.0 20.0 40.0 60.0 80.0 100.0 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 % Total Intensity % Total Intensity Retention time (min) Retention time (min) Mass (Da) Mass (Da) Figure 6: Venn diagram describing the dispersion of the top 100 of the most intense components in E.coli cytosol. Detailed information is provided on the nature of dispersion, where: 1-D PS = first dimension peak splitting. 2-D CO = second dimension carryover. Comb. = Combined first and second dimension dispersion effects. 0.0 20.0 40.0 60.0 80.0 0.0 20.0 40.0 60.0 80.0 9191 Da 0.0 20.0 40.0 60.0 80.0 0.0 20.0 40.0 60.0 80.0 10387 Da Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 % Total Intensity % Total Intensity 0.0 20.0 40.0 60.0 80.0 0.0 20.0 40.0 60.0 80.0 n n+1 n+2 n+3 n+4 9191 Da 0.0 20.0 40.0 60.0 80.0 0.0 20.0 40.0 60.0 80.0 10387 Da Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 % Total Intensity % Total Intensity Linear Gradient 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 11409 Da 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 9970 Da % Total Intensity % Total Intensity Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 n n+1 n+2 11409 Da 0.0 20.0 40.0 60.0 80.0 100.0 0.0 20.0 40.0 60.0 80.0 100.0 9970 Da % Total Intensity % Total Intensity Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Run 1 Run 2 Run 3 Linear Gradient Step Gradient Step Gradient Top 100 Most Intense Components 49 No Peak Splitting or Carryover in Step or Linear 21 28 2 Linear Both Step Carryover Observed (n=51) 2 1-D PS 8 1-D PS (Step) 11 2-D CO (Step) 9 Comb.(Step) 8 1-D PS (Linear) 10 2-D CO (Linear) 10 Comb. (Linear) 9 1-D PS 8 2-D CO 4 Comb. Top 100 Most Intense Components 49 No Peak Splitting or Carryover in Step or Linear 21 28 2 Linear Both Step Carryover Observed (n=51) 2 1-D PS 8 1-D PS (Step) 11 2-D CO (Step) 9 Comb.(Step) 8 1-D PS (Linear) 10 2-D CO (Linear) 10 Comb. (Linear) 9 1-D PS 8 2-D CO 4 Comb. Salt Gradient %1.25M NaCl Salt Gradient %1.25M NaCl Reversed Phase Gradient %ACN Reversed Phase Gradient %ACN
Transcript of Evaluating Multidimensional Liquid Chromatographic Separations … · Evaluating Multidimensional...
Evaluating Multidimensional Liquid Chromatographic Separations of Complicated Protein Mixtures
Kevin M. Millea2, Ignatius J. Kass1, John C. Gebler1, Ira S. Krull2, Scott J. Berger1 1 Life Sciences R&D, Waters Corporation, Milford, MA; 2Department of Chemistry, Northeastern University, Boston, MA
Overview
Instrumentation
Results
Conclusions
• Liquid phase protein separations are becoming viable alternatives to gel based methodologies for resolving sample components within a proteomic mixture.
• Liquid phase chromatographic separations offer a wide variety of separation mechanisms, a high degree of automation and easy coupling with mass spectrometry
• In this work, a 2-D LC separation is coupled with ESI-TOF mass spectrometry to extract component level information from a complex proteomic sample.
• Automated spectral deconvolution of the LC/MS data set provided the capability to assess chromatographic behavior of individual sample components.
• These approaches were combined to analyze the distribution and reproducibility of 2-D (IEX/RP) separations of E. coli cytosol using either a step or linear first dimension IEX gradient.
Figure 1: Multidimensional LC/ESI-TOF MS Configurations. An Alliance® 2D Bioseparations System (Waters) was used for online IEX/RP separations coupled to a Micromass® LCT™ ESI-TOF MS (Waters). Coupling the 2-D separation with online MS detection permitted the resolution of individual sample components. The second dimension RP dimension was configured using dual trapping columns, and a single analytical column. Two 10-port 2-position valves were used to direct solvent flow for trapping of analytes, column washing, and separation of sample components. A 0 to 500 mM NaCl eight-step or linear gradient was employed to develop the IEX first dimension, with a 1M column washing step following the gradient. Each second dimension reversed phase separation was effected by an water/acetonitrile gradient in a background of 0.1% formic acid/0.005% TFA. Positive mode ESI-MS data were obtained over a m/z range of 700-3000 at a 0.5 Hz acquisition rate.
• A complex proteomic sample was fractionated using multidimensional chromatography combined with online MS detection of intact proteins.
• Chromatographic behavior of the individual sample components was elucidated by automated time segmentation and spectral deconvolution of the protein LC/MS data set. (see poster P-8-M for details).
• This methodology was applied to reveal the chromatographic differences of separating sample components using either a linear or step gradient in the first separation dimension.
• The data indicate the linear IEX first dimension gradient reduced dispersion of components relative to the step gradient separation, and that adequate run-run reproducibility could be routinely achieved using multidimensional protein LC. ©2005 Waters Corporation 720001106EN
Figure 2: Comprehensive 2-D (IEX/RP) Chromatographic Separation of E. coli Cytosol Employing a Linear (LEFT SIDE) or Step (RIGHT SIDE) Gradient in the First Dimension.
Panel A: First and second dimension gradient profiles for the separation of E. coli cytosol.
Panel B: TIC traces of three replicate injections of E. coli cytosol.
Panel C: Expanded view of two reversed phase cycles demonstrating the complexity of the separation.
Panel D: Analysis of the processed neutral mass data reveals individual overlapping component profiles. Segments of the data comprising ten spectral scans (20 sec) were summed, and deconvoluted using a maximum entropy based approach (See Poster P-8-M for more details on this approach). Processed data were obtained over a mass range of 7,000 to 40,000 Da with a 3 Da mass resolution. The processed data is displayed in a 2-D Gel bubble plot display (x-axis: retention time, y-axis: deconvoluted mass; bubble diameter: centroided intensity). In this display the first run is visualized in green, the second visualized in red, and overlapping components displayed in orange.
A
B
C C
DD
B
A
Figure 3: Distribution of selected analytes demonstrating first dimension carryover. Dispersion of analyte into adjacent reversed phase cycles is characteristic of first dimension peak splitting. The distributions of two components (masses 11049 Da , and 9970 Da) are shown as the percent intensity distributed over several reversed phase cycles. Cycle n is the first cycle in which the component was detected, and n+1 constitutes the following reversed phase cycle. Triplicate analyses are shown for both gradient modes. The first component (11049 Da) illustrates splitting observed only in the step gradient mode, while the second component (9970 Da) show splitting common to both step and linear gradients.
Figure 4. Distribution of selected analytes demonstrating trap column carryover. The distributions of two components (masses 11036 Da and 31065 Da) demonstrate dispersion into alternating reversed phase cycles. This is characteristic of carryover (ghosting) on the second dimension trap column .
Figure 5: Distribution of selected analytes demonstrating extensive carryover into adjacent fractions. Two components (masses 9191 Da, and 10387 ) are shown to demonstrate reproducibly poor chromatographic behavior in both separation dimensions in step and linear modes.
LINEAR FIRST DIMENSION IEX GRADIENT STEP FIRST DIMENSION IEX GRADIENT
0.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
nn+1n+2n+3n+4
1103
6 D
a
Linear Gradient Step Gradient
0.0
20.0
40.0
60.0
80.0
100.0
3106
5 D
a
0.0
20.0
40.0
60.0
80.0
100.0
Run 1 Run 2 Run 3
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3
% T
otal
Inte
nsity
% T
otal
Inte
nsity
0.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
nn+1n+2n+3n+4
1103
6 D
a
Linear Gradient Step Gradient
0.0
20.0
40.0
60.0
80.0
100.0
3106
5 D
a
0.0
20.0
40.0
60.0
80.0
100.0
Run 1 Run 2 Run 3
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3
% T
otal
Inte
nsity
% T
otal
Inte
nsity
Retention time (min)Retention time (min)
Mas
s (D
a)
Mas
s (D
a)
Figure 6: Venn diagram describing the dispersion of the top 100 of the most intense components in E.coli cytosol. Detailed information is provided on the nature of dispersion, where: 1-D PS = first dimension peak splitting. 2-D CO = second dimension carryover. Comb. = Combined first and second dimension dispersion effects.
0.0
20.0
40.0
60.0
80.0
0.0
20.0
40.0
60.0
80.0nn+1n+2n+3n+491
91 D
a
0.0
20.0
40.0
60.0
80.0
0.0
20.0
40.0
60.0
80.0
1038
7 D
a
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
% T
otal
Inte
nsity
% T
otal
Inte
nsity
0.0
20.0
40.0
60.0
80.0
0.0
20.0
40.0
60.0
80.0nn+1n+2n+3n+491
91 D
a
0.0
20.0
40.0
60.0
80.0
0.0
20.0
40.0
60.0
80.0
1038
7 D
a
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
% T
otal
Inte
nsity
% T
otal
Inte
nsity
Linear Gradient
0.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
n
n+1
n+2
1140
9 D
a
0.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
9970
Da
% T
otal
Inte
nsity
% T
otal
Inte
nsity
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3Run 1 Run 2 Run 30.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
n
n+1
n+2
1140
9 D
a
0.0
20.0
40.0
60.0
80.0
100.0
0.0
20.0
40.0
60.0
80.0
100.0
9970
Da
% T
otal
Inte
nsity
% T
otal
Inte
nsity
Run 1 Run 2 Run 3 Run 1 Run 2 Run 3
Run 1 Run 2 Run 3Run 1 Run 2 Run 3
Linear Gradient Step Gradient Step Gradient
Top 100 Most Intense Components
49
No Peak Splitting or Carryover
in Step or Linear 21 28 2LinearBothStep
Carryover Observed (n=51)
2 1-D PS
8 1-D PS (Step) 11 2-D CO (Step) 9 Comb.(Step)8 1-D PS (Linear) 10 2-D CO (Linear) 10 Comb. (Linear)
9 1-D PS8 2-D CO4 Comb.
Top 100 Most Intense Components
49
No Peak Splitting or Carryover
in Step or Linear 21 28 2LinearBothStep
Carryover Observed (n=51)
2 1-D PS
8 1-D PS (Step) 11 2-D CO (Step) 9 Comb.(Step)8 1-D PS (Linear) 10 2-D CO (Linear) 10 Comb. (Linear)
9 1-D PS8 2-D CO4 Comb.
Salt Gradient %1.25M NaCl Salt Gradient %1.25M NaCl
Reversed Phase Gradient %ACN Reversed Phase Gradient %ACN
![ars.els-cdn.com · Web viewGradient elution in normal phase HPLC separations separates chiral compounds within a shorter time without impairing the chromatographic resolution [3].](https://static.fdocuments.net/doc/165x107/5fdd537ec834a111a82aa411/arsels-cdncom-web-view-gradient-elution-in-normal-phase-hplc-separations-separates.jpg)
