Special Report
For reprint orders, please contact [email protected]
Quantitative proteomics using the high resolution accurate mass capabilities of the quadrupole-orbitrap mass spectrometer
High resolution/accurate mass hybrid mass spectrometers have considerably advanced shotgun proteomics and the recent introduction of fast sequencing capabilities has expanded its use for targeted approaches. More specifically, the quadrupoleorbitrap instrument has a unique configuration and its new features enable a wide range of experiments. An overview of the analytical capabilities of this instrument is presented, with a focus on its application to quantitative analyses. The high resolution, the trapping capability and the versatility of the instrument have allowed quantitative proteomic workflows to be redefined and new data acquisition schemes to be developed. The initial proteomic applications have shown an improvement of the analytical performance. However, as quantification relies on ion trapping, instead of ion beam, further refinement of the technique can be expected.
Over the past two decades, MS technology has driven the progress of the proteomics field. The routinely used MS strategies rely on the enzymatic digestion of the proteins constituting the proteome and the separation and analysis of resulting peptides by LC–MS. Two main approaches arose from this generic overall process, which are commonly referred to as discovery and targeted proteomic strategies. Discovery proteomics relies on data-dependent acquisition (DDA) and is routinely used to extensively characterize proteomes, allowing the identification of thousands of proteins in biological samples [1–3] . In such an experiment, the ions corresponding to intact peptide species (i.e., precursor ions) are automatically selected by using an initial survey scan acquired in single-stage mode using simple criteria, such as their signal intensity and charge state. The precursors are fragmented to generate MS/MS. This acquisition procedure is repeated over the full LC separation and the fragment ions along with the precursor ion information are exploited to determine the peptide amino acid sequence by database search. However, the tremendous complexity of biological samples definitely exceeds the
10.4155/BIO.14.115 © 2014 Future Science Ltd
Sebastien Gallien1 & Bruno Domon*,1 1 Luxembourg Clinical Proteomics Center (LCP), CRP-Santé, Strassen, Luxembourg *Author for correspondence: Fax: +352 26970 717 [email protected]
peak capacity and the sampling rate of current LC–MS platforms, resulting in a bias towards the most abundant species [4] . The latest developments in instrumentation, with mass spectrometers having fast-sequencing capabilities [5,6] , have only partially addressed the issue of unsystematic measurements of low abundant analytes across multiple runs. Alternatively, strategies to reduce the sample complexity prior to MS analysis can be used (e.g., extensive fractionation through multidimensional LC or use of shallow gradients applied to long LC columns), with the additional benefit of alleviating the ion suppression/enhancement effects, however, at the expense of the analytical throughput. Shotgun proteomics turned out to be very powerful in establishing the profile of proteomes but has some limitations regarding quantification, which requires high sensitivity, selectivity and reproducibility. More specifically, the occurrence of ‘missing values’ in the replication of experiments is a major bottleneck in large-scale quantitative studies. The presence of analytes at low concentrations in complex matrix and the sampling limits of data-dependent acquisition are the main limiting factors.
Bioanalysis (2014) 6(16), 2159–2170
part of
ISSN 1757-6180
2159
Special Report Gallien & Domon Targeted proteomics methods were developed to systematically quantify peptides in complex proteomes [7] . In this strategy, a subset of analytes (peptides selected as surrogates for the proteins of interest) is predefined based on an initial hypothesis. These peptides are consistently measured with high sensitivity across large sample sets by accordingly setting instrumental parameters, thus addressing the ambiguity of ‘missing values’. Selected reaction monitoring (SRM; often referred to as multiple reaction monitoring) on triple quadrupole mass spectrometers [8] has become the reference technique for targeted analysis, also called targeted data acquisition (TDA), owing to its high sensitivity, wide dynamic range and excellent reproducibility of measurements [9,10] . In a targeted acquisition method, nearly 100% of measurement time is devoted to measure the peptides of interest [11] . This directly affects the number of peptides measured in one LC–MS experiment, the time allocated to the measurement of each peptide and their interdependence with the cycle time (i.e., the time needed to acquire one set of data points for a set of co-eluting peptides). The cycle time, determined by the chromatographic characteristics, is adjusted to collect eight to ten data points across the peptide elution profiles to allow precise quantification (i.e., typically 2–2.5 s for a chromatographic peak width of 20 s). A trade-off is often required as the number of analytes is increased at the expense of the sensitivity and selectivity of the measurements (resulting from the lower acquisition time for each measurement), which in turn impacts the analytical precision of experiments, and vice versa. Accordingly, two types of experiments should be distinguished, which have specific scopes and purposes (Figure 1) . The first one (called quantification mode) is intended to measure accurately and with high sensitivity the amount of a limited number of peptides present in a sample. Their absolute quantification requires a specific set up, such as the addition of reference peptides as internal standards (isotopically labeled peptide homologs to the targeted endogenous peptides) in well-defined amounts into the sample. The second mode (called screening mode) focuses on the detection and estimation of the actual abundance, that is, the detection of relative changes between samples, for a large number of peptides. The developments of the screening modes [12,13] Key terms Quantification mode: A type of targeted experiment used to measure accurately and with high sensitivity the amount of a limited analyte present in a sample. Screening mode: A type of targeted experiment used to detect and estimate the abundance (relative changes between samples) of a large number of analytes.
2160
Bioanalysis (2014) 6(16)
have dramatically increased the coverage of a targeted experiment, but not sufficiently to substitute discovery shotgun proteomics, which remains the current reference method to generate initial hypotheses, in spite of a limited penetration into the proteome. Each proteomic strategy has unique advantages to answer a specific biological question [14] . As an example, the biomarker development pipeline [15] , shown in Figure 2, combines an initial discovery phase to generate the candidates to be further evaluated using targeted methods, that is, the screening and quantification phases. The results obtained for each type of experiment are tightly linked to the mass spectrometer and to the acquisition procedure used, and contribute to the overall analytical performance of the workflow. Targeted proteomics experiments, performed using the ‘reference’ SRM technique on triple quadrupole mass spectrometers, suffer from the low resolution of quadrupole mass filters, which is often not sufficient to overcome the selectivity limitation due to the high complexity of biological samples [16] . The latest generation of quadrupole-time of flight [5] and the new quadrupole-orbitrap mass spectrometer [6] , both capable of high-frequency sequencing, represent alternatives to perform targeted proteomics experiments, especially when high selectivity is required on the measurement of fragment ions. While both instrument types yield high resolution and accurate mass measurements, the quadrupole-orbitrap has a higher resolution power and trapping capabilities offering a distinct benefit for the analysis of low abundant components over quadrupole-time of flight instruments. This account describes an overview of the analytical capabilities of the quadrupole-orbitrap mass spectrometer, with a focus on novel quantification approaches based on high-resolution/accurate mass (HR/AM) measurements. The unique configuration of the instrument and its versatility to operate in various modes of operation has enabled the design of innovative proteomic experiments. The initial results from this technology are promising but will require additional validation as quantification is based on an instrument with different physical principles (ion trapping). In addition, the technique is expected to further mature as quantification using triple-quadrupole instruments has done over the past two decades. Quadrupole-orbitrap instrument: a configuration offering new analytical capabilities The hybrid quadrupole-orbitrap mass spectrometer (Q-Exactive™; Thermo Scientific, Bremen, Germany) combines a quadrupole mass filter and an orbitrap mass analyzer (Figure 3A) together with a C-trap
future science group
Quantitative proteomics using the quadrupole-orbitrap mass spectrometer
A
Special Report
B Sensitivity
Scale
Sensitivity
Selectivity
Scale
Selectivity
Figure 1. Trade-off in targeted proteomics experiments. The interdependence between ‘scale’ and ‘sensitivity’–’selectivity’ has prompted the development of two classes of experiments with specific scope and purpose. (A) The ‘quantification mode’ aims at measuring accurately and with high sensitivity and selectivity the amount of peptide present in the sample, at the expense of the number of analytes (limited scale). (B) The ‘screening mode’ is dedicated to assess the detection of the peptides in a complex matrix and to estimate their actual abundance for a large set of analytes (large scale), which has a detrimental effect on sensitivity and selectivity. Adapted from [9] .
and a higher energy collisional dissociation (HCD) cell for ion trapping and fragmentation. This instrument enables various modes of operation in high resolution; either in single-stage (full scan or selected ion monitoring [SIM]) or tandem mass modes. It exploits the multiplexing capabilities of the trapping devices, which allow the temporary storage of several distinct ion populations, sequentially isolated (precursors) or
Discovery
Verification (detection)
Tissue samples ~10
generated (fragments), before their simultaneous measurement. The generic acquisition for the two types of experiments is presented in Figure 3B & C . Briefly, in the single-stage mode, the peptide ions (or sets of peptide ions) are isolated by the front-end quadrupole and accumulated in the C-trap, prior to the final transfer in the orbitrap for high-resolution mass analysis (Figure 3B) . In the MS/MS mode, following their Evaluation (precise quant)
Bodily fluids >50
100s
1000s samples
Candidates 100s
(Pre)-clinical
>100
> 20
5–10 Biomarkers
Data-dependent acquisition
Targeted data acquisition Screening mode
Quantification mode
Figure 2. The development pipeline to evaluate and validate biomarkers. Shotgun proteomics strategy is used in a discovery phase to analyze tissue samples or bodily fluids, using data-dependent acquisition. Candidate biomarkers are analyzed in bodily fluids using a targeted approach. A large number of candidates are first screened to assess their detection and determine their abundance in bodily fluids for a controlled set of disease/control samples. Finally, precise quantification is performed on a limited number of candidates to assess their clinical utility.
future science group
www.future-science.com
2161
Special Report Gallien & Domon
Quadrupole precursor ion selection
A
C-trap
Collision cell (HCD)
S-lens
Orbitrap
1
2a
1
2b
3
B
3
C
Figure 3. Schematic of a quadrupole-orbitrap mass spectrometer. (A) The instrument is constituted of a back-end high-resolution orbitrap mass analyzer and a quadrupole mass filter for the precursor ion selection. (B) In single-stage mode of operation (full scan or selected ion monitoring modes), the precursor ions are selected in the quadrupole (1) to be accumulated in the C-trap (2a) before their transfer and analysis in the orbitrap (3). (C) In tandem mode, they are transferred via the C-trap to the HCD cell, where they undergo dissociation (2b). The resulting accumulated fragment ions are transferred and measured in the orbitrap (3). HCD: Higher energy collisional dissociation. Adapted from [17] .
isolation, peptide ions (precursors) are transferred to the collision cell (HCD) via the C-trap where they undergo fragmentation. The resulting fragment ions are temporarily accumulated in the collision cell. They are then transferred back to the C-trap and ultimately to the orbitrap to be analyzed (Figure 3C) . The ability of the quadrupole mass filter to operate with an isolation window of different widths (ranging from 0.5 Th to a very wide mass-to-charge ratio range) offers different types of analyses, and therefore fully exploits the trapping capabilities. More specifically, it expands the dynamic range of measurements, and therefore enables the detection of low-abundance components present in complex samples (Figure 4) . The
2162
Bioanalysis (2014) 6(16)
ability to isolate selectively narrow populations of precursor ions by using restricted mass-to-charge ratio windows (typically ≤4 Th) leads to longer fill times while operating below the capacity limit of trapping devices. This can result in a significant improvement in the sensitivity of the measurements. The performance obtained in single-stage MS using isolation windows of 700 and 2 Th was compared in the analysis of a dilution series of eight synthetic isotopologous peptides. The results showed a tenfold increase in sensitivity by using the narrow quadrupole isolation widths [17] . This improvement resulted from the increase in the fill time from 0.3 to 250 ms. Nonetheless, too narrow quadrupole isolation windows (
For reprint orders, please contact [email protected]
Quantitative proteomics using the high resolution accurate mass capabilities of the quadrupole-orbitrap mass spectrometer
High resolution/accurate mass hybrid mass spectrometers have considerably advanced shotgun proteomics and the recent introduction of fast sequencing capabilities has expanded its use for targeted approaches. More specifically, the quadrupoleorbitrap instrument has a unique configuration and its new features enable a wide range of experiments. An overview of the analytical capabilities of this instrument is presented, with a focus on its application to quantitative analyses. The high resolution, the trapping capability and the versatility of the instrument have allowed quantitative proteomic workflows to be redefined and new data acquisition schemes to be developed. The initial proteomic applications have shown an improvement of the analytical performance. However, as quantification relies on ion trapping, instead of ion beam, further refinement of the technique can be expected.
Over the past two decades, MS technology has driven the progress of the proteomics field. The routinely used MS strategies rely on the enzymatic digestion of the proteins constituting the proteome and the separation and analysis of resulting peptides by LC–MS. Two main approaches arose from this generic overall process, which are commonly referred to as discovery and targeted proteomic strategies. Discovery proteomics relies on data-dependent acquisition (DDA) and is routinely used to extensively characterize proteomes, allowing the identification of thousands of proteins in biological samples [1–3] . In such an experiment, the ions corresponding to intact peptide species (i.e., precursor ions) are automatically selected by using an initial survey scan acquired in single-stage mode using simple criteria, such as their signal intensity and charge state. The precursors are fragmented to generate MS/MS. This acquisition procedure is repeated over the full LC separation and the fragment ions along with the precursor ion information are exploited to determine the peptide amino acid sequence by database search. However, the tremendous complexity of biological samples definitely exceeds the
10.4155/BIO.14.115 © 2014 Future Science Ltd
Sebastien Gallien1 & Bruno Domon*,1 1 Luxembourg Clinical Proteomics Center (LCP), CRP-Santé, Strassen, Luxembourg *Author for correspondence: Fax: +352 26970 717 [email protected]
peak capacity and the sampling rate of current LC–MS platforms, resulting in a bias towards the most abundant species [4] . The latest developments in instrumentation, with mass spectrometers having fast-sequencing capabilities [5,6] , have only partially addressed the issue of unsystematic measurements of low abundant analytes across multiple runs. Alternatively, strategies to reduce the sample complexity prior to MS analysis can be used (e.g., extensive fractionation through multidimensional LC or use of shallow gradients applied to long LC columns), with the additional benefit of alleviating the ion suppression/enhancement effects, however, at the expense of the analytical throughput. Shotgun proteomics turned out to be very powerful in establishing the profile of proteomes but has some limitations regarding quantification, which requires high sensitivity, selectivity and reproducibility. More specifically, the occurrence of ‘missing values’ in the replication of experiments is a major bottleneck in large-scale quantitative studies. The presence of analytes at low concentrations in complex matrix and the sampling limits of data-dependent acquisition are the main limiting factors.
Bioanalysis (2014) 6(16), 2159–2170
part of
ISSN 1757-6180
2159
Special Report Gallien & Domon Targeted proteomics methods were developed to systematically quantify peptides in complex proteomes [7] . In this strategy, a subset of analytes (peptides selected as surrogates for the proteins of interest) is predefined based on an initial hypothesis. These peptides are consistently measured with high sensitivity across large sample sets by accordingly setting instrumental parameters, thus addressing the ambiguity of ‘missing values’. Selected reaction monitoring (SRM; often referred to as multiple reaction monitoring) on triple quadrupole mass spectrometers [8] has become the reference technique for targeted analysis, also called targeted data acquisition (TDA), owing to its high sensitivity, wide dynamic range and excellent reproducibility of measurements [9,10] . In a targeted acquisition method, nearly 100% of measurement time is devoted to measure the peptides of interest [11] . This directly affects the number of peptides measured in one LC–MS experiment, the time allocated to the measurement of each peptide and their interdependence with the cycle time (i.e., the time needed to acquire one set of data points for a set of co-eluting peptides). The cycle time, determined by the chromatographic characteristics, is adjusted to collect eight to ten data points across the peptide elution profiles to allow precise quantification (i.e., typically 2–2.5 s for a chromatographic peak width of 20 s). A trade-off is often required as the number of analytes is increased at the expense of the sensitivity and selectivity of the measurements (resulting from the lower acquisition time for each measurement), which in turn impacts the analytical precision of experiments, and vice versa. Accordingly, two types of experiments should be distinguished, which have specific scopes and purposes (Figure 1) . The first one (called quantification mode) is intended to measure accurately and with high sensitivity the amount of a limited number of peptides present in a sample. Their absolute quantification requires a specific set up, such as the addition of reference peptides as internal standards (isotopically labeled peptide homologs to the targeted endogenous peptides) in well-defined amounts into the sample. The second mode (called screening mode) focuses on the detection and estimation of the actual abundance, that is, the detection of relative changes between samples, for a large number of peptides. The developments of the screening modes [12,13] Key terms Quantification mode: A type of targeted experiment used to measure accurately and with high sensitivity the amount of a limited analyte present in a sample. Screening mode: A type of targeted experiment used to detect and estimate the abundance (relative changes between samples) of a large number of analytes.
2160
Bioanalysis (2014) 6(16)
have dramatically increased the coverage of a targeted experiment, but not sufficiently to substitute discovery shotgun proteomics, which remains the current reference method to generate initial hypotheses, in spite of a limited penetration into the proteome. Each proteomic strategy has unique advantages to answer a specific biological question [14] . As an example, the biomarker development pipeline [15] , shown in Figure 2, combines an initial discovery phase to generate the candidates to be further evaluated using targeted methods, that is, the screening and quantification phases. The results obtained for each type of experiment are tightly linked to the mass spectrometer and to the acquisition procedure used, and contribute to the overall analytical performance of the workflow. Targeted proteomics experiments, performed using the ‘reference’ SRM technique on triple quadrupole mass spectrometers, suffer from the low resolution of quadrupole mass filters, which is often not sufficient to overcome the selectivity limitation due to the high complexity of biological samples [16] . The latest generation of quadrupole-time of flight [5] and the new quadrupole-orbitrap mass spectrometer [6] , both capable of high-frequency sequencing, represent alternatives to perform targeted proteomics experiments, especially when high selectivity is required on the measurement of fragment ions. While both instrument types yield high resolution and accurate mass measurements, the quadrupole-orbitrap has a higher resolution power and trapping capabilities offering a distinct benefit for the analysis of low abundant components over quadrupole-time of flight instruments. This account describes an overview of the analytical capabilities of the quadrupole-orbitrap mass spectrometer, with a focus on novel quantification approaches based on high-resolution/accurate mass (HR/AM) measurements. The unique configuration of the instrument and its versatility to operate in various modes of operation has enabled the design of innovative proteomic experiments. The initial results from this technology are promising but will require additional validation as quantification is based on an instrument with different physical principles (ion trapping). In addition, the technique is expected to further mature as quantification using triple-quadrupole instruments has done over the past two decades. Quadrupole-orbitrap instrument: a configuration offering new analytical capabilities The hybrid quadrupole-orbitrap mass spectrometer (Q-Exactive™; Thermo Scientific, Bremen, Germany) combines a quadrupole mass filter and an orbitrap mass analyzer (Figure 3A) together with a C-trap
future science group
Quantitative proteomics using the quadrupole-orbitrap mass spectrometer
A
Special Report
B Sensitivity
Scale
Sensitivity
Selectivity
Scale
Selectivity
Figure 1. Trade-off in targeted proteomics experiments. The interdependence between ‘scale’ and ‘sensitivity’–’selectivity’ has prompted the development of two classes of experiments with specific scope and purpose. (A) The ‘quantification mode’ aims at measuring accurately and with high sensitivity and selectivity the amount of peptide present in the sample, at the expense of the number of analytes (limited scale). (B) The ‘screening mode’ is dedicated to assess the detection of the peptides in a complex matrix and to estimate their actual abundance for a large set of analytes (large scale), which has a detrimental effect on sensitivity and selectivity. Adapted from [9] .
and a higher energy collisional dissociation (HCD) cell for ion trapping and fragmentation. This instrument enables various modes of operation in high resolution; either in single-stage (full scan or selected ion monitoring [SIM]) or tandem mass modes. It exploits the multiplexing capabilities of the trapping devices, which allow the temporary storage of several distinct ion populations, sequentially isolated (precursors) or
Discovery
Verification (detection)
Tissue samples ~10
generated (fragments), before their simultaneous measurement. The generic acquisition for the two types of experiments is presented in Figure 3B & C . Briefly, in the single-stage mode, the peptide ions (or sets of peptide ions) are isolated by the front-end quadrupole and accumulated in the C-trap, prior to the final transfer in the orbitrap for high-resolution mass analysis (Figure 3B) . In the MS/MS mode, following their Evaluation (precise quant)
Bodily fluids >50
100s
1000s samples
Candidates 100s
(Pre)-clinical
>100
> 20
5–10 Biomarkers
Data-dependent acquisition
Targeted data acquisition Screening mode
Quantification mode
Figure 2. The development pipeline to evaluate and validate biomarkers. Shotgun proteomics strategy is used in a discovery phase to analyze tissue samples or bodily fluids, using data-dependent acquisition. Candidate biomarkers are analyzed in bodily fluids using a targeted approach. A large number of candidates are first screened to assess their detection and determine their abundance in bodily fluids for a controlled set of disease/control samples. Finally, precise quantification is performed on a limited number of candidates to assess their clinical utility.
future science group
www.future-science.com
2161
Special Report Gallien & Domon
Quadrupole precursor ion selection
A
C-trap
Collision cell (HCD)
S-lens
Orbitrap
1
2a
1
2b
3
B
3
C
Figure 3. Schematic of a quadrupole-orbitrap mass spectrometer. (A) The instrument is constituted of a back-end high-resolution orbitrap mass analyzer and a quadrupole mass filter for the precursor ion selection. (B) In single-stage mode of operation (full scan or selected ion monitoring modes), the precursor ions are selected in the quadrupole (1) to be accumulated in the C-trap (2a) before their transfer and analysis in the orbitrap (3). (C) In tandem mode, they are transferred via the C-trap to the HCD cell, where they undergo dissociation (2b). The resulting accumulated fragment ions are transferred and measured in the orbitrap (3). HCD: Higher energy collisional dissociation. Adapted from [17] .
isolation, peptide ions (precursors) are transferred to the collision cell (HCD) via the C-trap where they undergo fragmentation. The resulting fragment ions are temporarily accumulated in the collision cell. They are then transferred back to the C-trap and ultimately to the orbitrap to be analyzed (Figure 3C) . The ability of the quadrupole mass filter to operate with an isolation window of different widths (ranging from 0.5 Th to a very wide mass-to-charge ratio range) offers different types of analyses, and therefore fully exploits the trapping capabilities. More specifically, it expands the dynamic range of measurements, and therefore enables the detection of low-abundance components present in complex samples (Figure 4) . The
2162
Bioanalysis (2014) 6(16)
ability to isolate selectively narrow populations of precursor ions by using restricted mass-to-charge ratio windows (typically ≤4 Th) leads to longer fill times while operating below the capacity limit of trapping devices. This can result in a significant improvement in the sensitivity of the measurements. The performance obtained in single-stage MS using isolation windows of 700 and 2 Th was compared in the analysis of a dilution series of eight synthetic isotopologous peptides. The results showed a tenfold increase in sensitivity by using the narrow quadrupole isolation widths [17] . This improvement resulted from the increase in the fill time from 0.3 to 250 ms. Nonetheless, too narrow quadrupole isolation windows (
