The accurate quantitation of proteins and peptides in complex biological systems is one of the most challenging areas of proteomics. Mass spectrometry-based approaches have forged significant in-roads allowing accurate and sensitive quantitation and the ability to multiplex vastly complex samples through the application of robust bioinformatic tools. These relative and absolute quantitative measures using label-free, tags, or stable isotope labelling have their own strengths and limitations. The continuous development of these methods is vital for increasing reproducibility in the rapidly expanding application of quantitative proteomics in biomarker discovery and validation. This paper provides a critical overview of the primary mass spectrometry-based quantitative approaches and the current status of quantitative proteomics in biomedical research.
Quantification in a proteomics setting relies on the ability to detect small changes in protein and peptide abundance in response to an altered state [
Quantitative proteomics comes in two forms: absolute and relative. Relative quantitation compares the levels of a specific protein in different samples with results being expressed as a relative fold change of protein abundance [
Traditional proteomic quantitation approaches rely on high-resolution protein separation by 2D gels. The use of dyes, fluorophores, or radioactivity to label proteins allows visualization of spots/bands with differential intensities [
The ultimate aim of biomarker discovery is to develop a simple differential test to be used as a clinical evaluation tool. This requires a lengthy and difficult process which involves candidate discovery, verification, validation, and translation to clinical laboratory use [
A challenge facing biomarker development is the sheer complexity and range of concentrations within the human proteome [
Disease-specific proteins, including low-mass peptides, can be low in abundance and difficult to detect amongst a diverse “sea” of proteins [
Relationship between the peptide ion content and the difficulty in obtaining sufficient MSMS information to both identify and also quantitate those peptides. Adapted from Michalski et al. [
There are a number of novel techniques that allow for the fractionation, depletion, enrichment, and equalisation of complex samples to assist in improving the proteome coverage and number of peptide ions targeted for MS/MS within an instrument’s detection range. Fractionation techniques can be applied to cut samples into subgroups of fewer proteins [
Data analysis is yet another significant challenge associated with MS-based proteomics. With the enormous volumes of proteomic data generated, expert manual analysis would be inconsistent and unfeasible [
Protein mass spectrometry is not inherently quantitative. There are many reasons as to why the amount of analyte compared to the MS signal intensity does not always show a linear relationship [
Overview of the main approaches for quantitative proteomics. Modified from Schulze and Usadel [
Method | Dynamic rangea | Coverage | Quant accuracy, (throughput) |
Associated software | Link |
---|---|---|---|---|---|
Label-free | |||||
2D gels | 1 to 4, |
Medium | Medium (low) |
PDQuest |
|
Ion intensities MS1 | 3 | Good | Medium, (medium to high) |
Progenesis LCMS |
|
Spectrum count MS2 | 3, Inaccurate for low abundance. | Good | Poor, (medium to high) |
Scaffold [ |
|
APEX,emPAI | 3 or 4 | Good | Poor, (high) |
APEX [ |
|
Metabolic labeling | |||||
15N | 1 to 2 | Medium | Precise, (low). |
Scaffold |
|
SILAC | 1 to 2 | Medium | Precise, (low). |
Scaffold |
|
Isotopic labeling | |||||
ICAT, |
1 to 2 | Poor | Precise, (low). |
Elucidator |
|
Isobaric labeling | |||||
ITRAQ, |
2 |
Medium | Medium, (low). |
ProteinPilot |
|
Targeted | |||||
MRM |
5 |
Poor1 | Precise, (high). |
Skyline [ |
|
APEX: absolute protein expression profiling. emPAI: exponentially modified protein abundance index. SILAC: stable isotope labelling by amino acids. DIGE: Difference Gel Electrophoresis. ICAT: isotope-coded affinity tags. ITRAQ: isobaric tags for absolute and relative quantitation. TMT: tandem mass tags. MRM: multiple reaction monitoring.
MS2: MSMS
Two widely used label-free quantitative methods are spectral counting and peptide peak intensity measurement. Spectral counting requires proteins to have sufficient peptides (both in number and abundance) to trigger MS/MS data for quantification and identification. The approach is based on the observation that more abundant proteins will produce more MS/MS spectra than less abundant proteins, and abundant peptides are sampled more often in fragment ion scans than are low abundance peptides. Relative quantitation by spectral count thus involves comparing the number of identified spectra from the same protein between different samples [
Relative quantitation using peptide peak intensity measurements involves comparing the MS peptide ion intensities belonging to a given protein [
Stable isotope labelling techniques are based on the introduction of a differential mass tag which affects only the mass of a protein or peptide without changing the chemical properties during chromatography or MS [
Metabolic labelling involves the introduction of stable isotopes to whole cells through the growth medium, which enables the labels to be incorporated during normal cell growth and division [
In chemical labelling, the isotope label is introduced to proteins or peptides by a chemical reaction, such as with isotope-coded affinity tags (ICAT) [
The development of isobaric mass tags such as tandem mass tag (TMT) [
Recent quantitative MS-based studies involving human samples.
Authors/year | Specimen | Quantitative approach | Sample |
Outcomes |
---|---|---|---|---|
Yang et al. 2011 [ |
Urine |
Label-free—spectral count | NIL | Quantified 265 glycoproteins. alpha-1-antitrypsin, 74% sensitivity and 80% specificity for bladder cancer patients. |
Quintana et al. 2009 [ |
Urine |
Label-free—peak peptide intensity | SCX using |
Peptides from uromodulin and kininogen significantly elevated in control compared to CAD patients. |
Hanas et al. 2008 [ |
Serum |
Label-free—peak peptide intensity | NIL | Quantified 20 low-mass serum peaks. Bootstrap analysis showed peaks could differentiate cancer from control sera with 95% accuracy. |
Xue et al. 2010 [ |
Cell lysates |
Label-free—peak peptide intensity | NIL | 145 differential proteins. Western blot and ROC curve analysis confirmed that 2 specific proteins could predict colorectal cancer metastasis. |
Besson et al. 2011 |
Colorectal cancer tissue |
Stable isotope labeling—iTRAQ | Peptide |
555 proteins with significant fold change between different cancer stages. Identified a candidate with increased abundance in adenomas and early stage colorectal cancer. |
Bondar et al. 2007 [ |
Serum |
Stable isotope labeling | NIL | Higher abundance of Zn- |
Chaerkady et al. 2008 [ |
Liver tissue |
Stable isotope labeling—iTRAQ | SCX | 59 proteins increased in abundant, 92 proteins were less abundant in HCC compared to normal tissue. 12 proteins further validated using immunohistochemical labeling. |
Dayon et al. 2008 |
Cerebrospinal fluid |
Stable isotope labeling—tandem mass tag isobaric labeling | Immunoaffinity depletion of 6 |
78 proteins more abundant in postmortem samples compared to antemortem. |
Multiple reaction monitoring (MRM) is the main current approach for highly confident protein and peptide quantification. MRM targets specific peptides in complex samples by typically using a triple quadrupole mass spectrometer or hybrid triple quadrupole/linear ion trap mass spectrometer. These instruments have two mass filters that can select a predefined peptide ion and a combination of its specific fragment ions to analyse and monitor over time for accurate quantitation [
Absolute quantitation can be achieved when MRM is incorporated with isotopically labelled synthetic peptide internal standards, which are designed to be identical to target peptides [
MRM has a greater sensitivity towards low abundance peptides and relatively good quantitative precision compared to other methods discussed [
MRM has been used to quantify major plasma proteins and target biomarkers for a range of diseases. Table
Summary of MRM quantitative analysis in blood for a variety disease types.
Authors/year | Specimen | Target | Sample preparation | MS platform | Outcomes |
---|---|---|---|---|---|
Stahl-Zeng et al. 2007 [ |
Plasma |
N-glycoproteins | Selective isolation of N-glycosites. Stable isotope 13C- and/or 15N-labelled reference peptides. | LC ESI MS/MS Hybrid triple quadrupole linear ion trap | Detection ≤ ng/mL concentration range and accurate quantification over a linear range of ~ 105. |
Anderson and Hunter 2006 [ |
Plasma |
53 plasma proteins | Top six abundant proteins depleted. Stable isotope labeled internal standards. | ESI LC-MS/MS |
Quantitative data for 47 proteins in the µg/mL level over linear range of 104 |
Keshishian et al. 2007 [ |
Plasma |
6 low abundance plasma proteins | Abundant protein depletion and SCX chromatography. Stable isotope-labeled amino acids. | ESI LC-MS/MS |
LOQ of 1–10 ng/mL range and linearity ≥ 102. LOD in high pg/mL. |
McKay et al. 2007 [ |
Plasma |
18 liver-derived proteins in plasma | Immunodepletion (Albumin and IgG removed) | ESI LC-MS/MS |
Increase in target plasma proteins during treatment. Similar trends found in MRM assays and 2-D DIGE |
Kirsch et al. 2007 [ |
Blood bank pooled serum |
2 human growth hormones (IGFBP-3, IGF-1) | NIL | ESI LC-MS/MS |
Detection ranges of 4–10 ng/µl for IGFBP-3 and 2–8 ng/µl for IGF-1. |
Kuhn et al. 2004 [ |
Serum. |
C-reactive protein | Immunodepletion of haptoglobin, IgG and HSA, then size exclusion chromatography | ESI LC-MS/MS |
Correlation between erosive RA, RA and increased CRP over healthy patients. |
Fortin et al. 2009 [ |
Serum |
PSA | Immunodepletion of albumin and mixed cation exchange peptide fractionation | ESI LC-MS/MS |
Absolute quant. of PSA to low ng/mL, with good correlation to clinical ELISA tests. |
Huillet et al. 2012 [ |
Serum |
Clinically validated cardiovascular biomarkers (LDH-B, CKMB, myoglobin, troponin I) | Immunodepletion of six highest abundant proteins and SDS-PAGE |
ESI LC-MS/MS |
Absolute quant. using Protein Standard Absolute Quantification (PSAQ) and MRM. |
Zhao et al. 2010 [ |
Serum |
Candidate biomarkers of hepatocellular carcinoma (vitronectin and clusterin) | NIL | ESI LC-MS/MS |
Stable isotope dilution-MRM using 18O-labelling method demonstrated significant downregulated in HCC compared to healthy group. Results comparable to ELISA. |
Kuhn et al. 2009 [ |
Plasma |
Troponin I, and Interleukin 33 | Immunoaffinity enrichment SISCAPA | ESI LC-MS/MS |
Linearity from1.5 to 5000 µg/L and correlated with commercial immunoassay. |
Lopez et al. 2011 [ |
Serum |
12 putative markers of Trisomy 21 | NIL | ESI LC-MS/MS |
Developed a workflow for Trisomy 21. Protein biomarkers targeted are high abundance proteins. |
Domanski et al. 2012 [ |
Plasma |
67 putative markers of cardiovascular disease | NIL | ESI LC-MS/MS |
117 from 135 peptides with attomolar LOQ for 81 peptides. |
Further instrument developments have taken advantage of the high resolution and mass accuracy of the TOF and orbitrap analysers and combined them with the selectivity of the quad analysers by replacing the third quad with either an orbitrap or a TOF analyser. These high resolution/accurate mass (HR/AM) instruments are addressing the challenge of eliminating cofiltering interfering ions, while taking advantage of the accuracy afforded by these instruments. In experiments similar to MRM called parallel reaction monitoring (PRM), it is possible to detect all product ions of a peptide in parallel rather than just few transitions per peptide. This allows an increased number of peptides to be quantitated in the one experiment. This combination of analysers firstly uses the quadrupole to select a restricted m/z range (with broad mass filtering window typically 2–100Th, rather than broad scan of around 700Th), and the MS/MS mode provides further selectivity and accuracy utilizing the orbitrap or TOF analyser to achieve higher resolution and mass accuracy in both MS and MS/MS scanning modes [
The use of multiparametric assays is becoming an increasing necessity in quantitative studies to overcome a variety of challenges associated with properties of the marker and/or the techniques including immobilisation efficiencies, detection, signal-to-noise [
Quantitative proteomic analysis has been a point of discussion for the last four decades, with comparative and once limited MS-based techniques heralding the advances that would forge the necessary connection between the dynamic biology of a system and its quantitative proteomic content. The major advances in quantitative MS proteomics have been exceptionally demonstrated over the last decade with the introduction of compatible and reliable label and label-free techniques. These advances now require further developments in bioinformatics and downstream validation, technologies that are required to make sense of complex data and enable researchers to infer more meaningful data that will transform into clinical benefit for years to come.