2.2. Literature review of Saccharomyces cerevisiae proteomic analysis
2.2.2. Proteomics as a tool for the identification and quantitation of S. cerevisiae
2.2.2.1. Introduction
Since protein identification is also a key part of the quantitation procedure (for shotgun proteomics in particular), this chapter aims to focus on quantitative methods more than identification methods. To date, traditional quantitative proteomics has been carried out in the main by 2-DE (two dimensional gel electrophoresis). Recently, various techniques for protein quantitation by mass spectrometry have been developed, ranging from label-free quantitation [6, 7] to label-based quantitation [8-11]. Proteomic quantitation can be performed either in vivo (metabolic labeling) or in vitro (protein and peptide labeling).
Basically, the label-based proteomic quantitation can be performed at 3 levels: cell, protein, and peptide levels. The cell labeling approach is usually known as a metabolic labeling, and includes: metabolic labeling (15N, or 13C), and stable isotope labeling of amino acids in culture (SILAC) [8, 9]. The labeling approach at the protein and peptide levels are mostly based on employing stable isotopic tags, such as trypsin digestion of protein in [18O-water]
[10], isotope-coded affinity tags (ICAT) [4], or isobaric tags for relative and absolute quantitation (iTRAQ) [11] (see Figure 2.2). The majority of the proteomic quantitation work carried out to date has been relative quantitation. Although, clearly, absolute
quantitation is highly desirable ultimately for a deeper understanding, and several emerging techniques are meeting with success in this regard, such as absolute quantitation (AQUA) [12, 13] and QConCat [14, 15] workflows (see Table 2.1 for details). The advantages of the labeling techniques are the ability to simultaneously identify and quantity simultaneously (and hopefully automatically) targets from complex protein mixtures [9].
Identification
Quantitation 2D-gels Shot-
gun
Based on gels Labeling technique
Gels
comparision DIGE Cells Proteins Peptides
13C/12C
15N/14N SILAC ICAT iTRAQ [18O]- water
Free-label
Peak intensity
of peptides
Number of MS/
MS spectra
Absolute
AQUA QConCAT
Proteomic analysis of S. cerevisiae
Figure 2.2. The identification and quantitation methods which have been applied to the proteomic analysis of S. cerevisiae.
With metabolic labeling, the labeling step is performed by incorporation of the isotopes during cell growth. Proteins in cells are labeled with a stable isotope labeled carbon source, or nitrogen source, for example 13C-Glucose [16, 17], 15N-ammonium sulphate [172, 173], or via the SILAC method [18]. In practice, cells are grown under different conditions containing either compounds with the natural isotope abundance, or those containing the heavy isotope. The cells from different conditions are then mixed, and protein extraction carried out (it is possible to do an extraction first, with the proteins mixed after extraction, but this is more likely to create errors) (see Figure 2.3 for details). A traditional 1-DE or 2-
DE gel-based workflow or shotgun proteomics workflow can then be followed. Peptides (usually) are then identified using mass spectrometry. The relative abundance of a peptide is then calculated based on the areas under the heavy and the light versions of the same peptide fragmented by MS.
Peptides labeling
Labeling Labeling
SCX Digestion Mixed proteins
Labeling Labeling
Digestion
SCX SCX
Mixed peptides
Digestion Digestion
Protein extraction
Proteins labeling
Mixed proteins
Protein
extraction Protein
extraction Protein
extraction
Protein
extraction Protein
extraction
Cells in sample 1
Cells in
sample 2 Cells in
sample 1 Cells in
sample 2
In vivo labeling
Metabolic labeling
Cells in sample 2
In vivo labeling Cells in sample 1
Figure 2.3. An overview of the workflows involved in the various labeling methods.
A favourable alternative technique to metabolic labeling is the use of a post-growth labeling strategy. In this case, all cells are grown in normal media with no heavy isotope enrichment. The proteins or peptides are then labeled with tags. A typical method for protein labeling is ICAT [4], see Table 2.1 for detail. For peptide labeling, the labeling step is carried out after the digestion step via a number of potential methods, for example
iTRAQ [11], or during digestion step using 18O labeled water. Then these peptides are fractionated and analysed by tandem mass spectrometry.
2.2.2.2. Comparison of methods
The 2-DE approach is the classical technique for the identification and quantitation of proteins for proteomics applications. 2-DE works by comparing both the positions and intensities of spots on a series of gels. Since the workflow is based on gel-to-gel comparisons it may be affected by spot position and lead to problems in detecting the differences between gels [19]. To meet the requirements of data validity, numerous repeated gel experiments should be performed to ensure the statistical significance of results [20]. To improve the power of this technique, both technical solutions and analysis software platforms have been developed. One of the solutions for the technical aspects is the application of DIGE (difference gel electrophoresis), an approach that allows for the comparison of two samples in the same gel [21]. Briefly, proteins in two samples are labeled with one of two fluorescent dyes and then mixed with a third labeled mixture of two samples as an internal calibration [21]. After separation on 2-DE gels, proteins are detected by different lasers, and then images are overlapped to reveal changes in proteins abundance.
The most significant advantage of this method is increased ease of comparison, since the samples are run together, resulting in the elimination of gel-to-gel variation [22].
For in vivo labeling with 14N and 15N, there is a mass shift in the resultant peptides from the two different (labeled and unlabeled) media observed during MS analysis. The N-label can be found in both the backbone and side chain nitrogen atoms, therefore the mass shift cannot (easily) be used to predict peptides from unknown sequences [23]. As a result, this method is not advantageous for highly complex samples such as cell lysates, since having heavy isotope containing compounds essentially doubles the complexity of the sample submitted to the mass spectrometer. However, this method provides a fractionation step (for example cysteine capture), which is used to reduce the complexity of samples [24].
Table 2.1. An introduction and comparison of methods for quantitation.
Names Reagents - Functions Methods Advantages Disadvantages Release Time
2-DE Gel Two dimensional gel
electrophoresis
DIGE
Difference gel electrophoresis
Coomassie blue stain, or silver stain used as a traditional stains
Fluorescent protein stains which are used to increase the quantitative accuracy
Recently, stains specific for post-translational
modifications have been developed
The mixture of proteins are separated firstly by isoelectric point on gel strip, and then by molecular weight on SDS- PAGE gels
The comparison is based on the position and intensities of spots on 2-DE gels
Accuracy in comparison due to at least triplicate for each phenotype Ease of performance
Sequential, labour intensive Difficult to automate
Not sensitive for basic, hydrophobic, and large molecular mass proteins At least triplicate gels for each sample, therefore, cost time and laborious to analyse and perform
SILAC
Stable isotope labeling of amino acids in culture
Using a heavy or light form of an essential amino acid as a medium, such as [13C6]- arginine, [13C6]-lysine, [D3]- leucine
The isotopic label into peptides via metabolic labeling in
the culture Does not require
multiple chemical processing steps as well as purification of protein samples
Not always suitable for all experimental samples (requires growth on isotope enriched substrate)
2002 by Ong et al.
[9]
ICAT Isotope-coded affinity tags
Two isotopic reagents which are a heavy isotope (e.g.
contains eight deuterium) and a light isotope (e.g. contains no deuterium)
Each reagent includes 3 elements: (see Figure 2.4.A) - An affinity tag (biotin) which is used to isolate ICAT- labeled peptides
- A linker which is incorporated stable isotopes
Number of samples for one ICAT experiment: 2 samples.
Each sample is denatured, reduced, and labeled with isotopic reagent. Protein samples are then mixed and trypsin digested
Cysteine-tagged peptides are enriched by affinity chromatography of the biotin tag using an avidin column Enriched-peptides are then fractionated by RP-HPLC alone or in combination with SCX. Peptides fractions are submitted to the mass spectrometer for identification and quantitation of peptides
The comparison is based on isotopic tagging of cysteine residues which can be seen on MS/MS
Good for complex samples due to only few peptides per proteins being used for analysis
Only two samples can be used for one experiment The excess of biotin in the sample matrix (e.g. serum) may reduce the affinity column because of binding sites competition
Sometimes, the isobaric between ICAT peptides and non-tagged peptides eluting
from affinity chromatography results in
false-positive and false-
1999 by Gygi et al.
[4]
Names Reagents - Functions Methods Advantages Disadvantages Release Time
(e.g. 1H or 2H)
- An iodoacetamide reactive group which is reacted with thiol groups (cysteines) Versions:
- Original: [1H]- and [2H]- isotopic ICAT tags
- [12C]- and [13C]- isotope pairing
negative identification The loss of collision energy because of the ICAT tag during MS fragmentation results in sequencing difficulties
Because of peptide selective strategies, a loss of peptide redundancy of a given protein can be found Possible difficulties for PTM analysis in the case of only one or few peptides containing low-abundance cysteine, which is used as a representative for protein identification
iTRAQ
Isotope tags for relative and absolute quantification
Four reagents: 114, 115, 116, and 117, but coming soon with eight regents: 113, 114, 115, 116, 117, 118, 119, 121
Each reagent include 3 elements: (see Figure 2.4.B) Reporter group (mass 114 - 117 for 4 plex, or 114 - 119 and 121 for 8 plex) which is used quantify their respective samples
Balance group (mass 31-28 for 4 plex, or 92 - 86 and 84 for 8 plex) which is used to
Number of samples for one iTRAQ experiment: 4 samples, and soon 8 samples
Each sample is denatured, reduced, alkylated, and trypsin digested. Subsequently, peptides in each sample are labeled with iTRAQ reagents, and then combined
The labeled-peptides mixture is then fractionated using SCX chromatography. Fractions are introduced to tandem MS/MS for identification and quantitation of peptides and then proteins
The label is cleaved in the MS before quantification
Quantitative comparison can be performed up to 4 samples and soon 8 samples within a single experiment
High validation and good for quantitative comparison of PTM and sub-proteome
The signal intensity is increased due to the
Not so good for identification and quantitation of low abundance proteins, as well as samples with high complexity
Time-consuming
The noise of un-tagged isobaric chemicals may confuse MS sequencing of the labeled-peptides
Problem of protein variants
2004 by Ross et al.
[11]
Names Reagents - Functions Methods Advantages Disadvantages Release Time
maintain each isotopic tag at exactly the same mass
Amine specific peptide reactive group which is used to react with all primary amines, including N-terminus and the ε-amino group of lysine side-chain, to label all peptides
isotopically-labeled peptides are isobaric and all contribute to one ion species that is used for CID and observed in the MS
16O/18O Water with two forms: H218O,
and H216O Two samples are digested with trypsin in two forms of water H218O, and H216O
The atoms of 18O and 16O are attached to the carboxyl termini of trypsin digested peptides
Two samples are then combined, and analysed by MS/MS The quantitation is performed based on the relative intensity of paired peptides with 4 Da mass difference detected by MS/MS
Simple procedure, high efficiency
Less cost for experiments compared to other labeling methods
The 4 Da mass difference is not big enough to detect by ESI ion trap MS/MS
Because of peptide selective strategies, a loss of peptide redundancy of a given protein can be found
1983 by Rose et al.[10]
Label-free quantitation
No reagent is used, the quantitation is based on MS/MS data only
The quantitation can be approached based on:
i) The peak intensity measurement of peptides ii) The number of MS/MS spectra per protein detected
Lower cost for experiments compared to other methods
The quantitation is affected by ionization efficiency and chromatography conditions
i) 2003 by Andersen et al.[6]
ii) 2004 by Liu et al.[7]
Protein- AQUA
An selected isotope labeled (13C, 15N) peptides are used as an internal standard
Each AQUA peptide is a synthetic tryptic peptide incorporating one stable isotope labeled amino acid, forming 6-10 Da in molecular
An optimal tryptic peptide corresponding to an interested protein is selected and then synthesised incorporating stable isotope (13C, 15N) labeled amino acid
The known quantity of AQUA peptide is subsequently added to protein sample extracted from biological sample.
Sample is then proteolysised with enzyme Peptide mixture is analysed by tandem MS/MS
High sensitivity, and more accurate compared to quantitative PCR or Northern blotting
Suitable for study of gene silencing in low abundance proteins
Each protein to be quantified requires at least one stable isotope labeled peptide synthesised independently, therefore high cost is required
Moreover, each peptide must be purified and
2003 by Gygi [13]
Names Reagents - Functions Methods Advantages Disadvantages Release Time
weight
An optimal tryptic peptide corresponding to interested protein can be ordered from the Sigma AQUA Peptide Library
The quantitation is performed by comparing the signal
intensities of native peptide and AQUA peptide Can be used for any silence gene since an AQUA peptide can be formed using only an amino acid sequence Study of quantitative proteomics can be done on specific proteins, as well as specific amino acid modification
quantified
Not suitable for study of global proteins expression since huge amount of synthesised peptides are required
QconCAT Artificial genes are designed de novo to direct the biosynthesis of novel proteins which are assemblies of signature Qpeptides
Qpeptides are arginine or lysine at the C-terminus, and used as internal standards for peptides derived from digestion of proteins samples
This method is performed via the design, synthesis and expression of artificial genes that encode concatenated proteolytic peptides used as surrogates for the interested protein
The artificial genes are transformed into and expressed in a heterologous such as bacterial (E. coli) to create an expression strain
The expression strain is then grown in either media containing isotope labeled source (such as 15N, or 13C) or containing specific stable isotope labeled amino acids so that the artificial proteins are fully labeled
The artificial proteins (QconCAT) is then purified and quantified before being mixed with a complex mixture of proteins
The sample is then proteolysised to peptides, and then these peptides are analysed by tandem MS/MS
The quantitation is performed by comparing the signal intensities of native peptides (proteins) and Qpeptides (QconCAT)
Suitable for absolute quantitation of large numbers of multiplexed proteins
The selection of peptides to use as surrogates is restricted
Protocol is moderately complex
2005 by Beynon [14]
Figure 2.4. Structure of reagents used for iTRAQ (A) or ICAT (B) techniques [4, 11].
To overcome problems inherent in the 2-DE workflow, gel-free (or shotgun) proteomics methods based on orthogonal liquid chromatography have been developed (see [25] for example), allowing for the potential of quantitation on the mass spectrometer. The ICAT shotgun proteomics quantitation technique [4] sought to address the problem inherent in mass spectrometry, in which the detection is based on ionization of peptides. Since ionization efficiency is affected by a number of factors, the peak intensities of the same peptides may be difficult to compare (and thus use for quantitation) across samples. To overcome this, in ICAT, a covalent modification of cysteine residues in the peptide pools derived from the samples of interest with light and heavy isotope labels means that, these peptides can then mixed and injected into the mass spectrometer as a single sample.
Quantitation can be achieved for this duplex sample set using the peak intensities that correlate with peptide abundance. Since this technique is based on labeling of cysteine residues, problems occur since most proteins contain few cysteine residues. Moreover, ICAT is not suitable for post-translational modification (PTM) analysis because of this low abundance of cysteine.
The development of iTRAQ [11] provided a powerful tool measuring differential protein expression changes, because the labels attach to amine groups, removing the limitation of
N O O N O
N
O
Reporter group Reactive group
Balance group
(A)
S NH N H
O
NH O O
O X
X X
X O N
H X
X X
X
O
Biotin tag Labeled linker
X: Heavy or light isotope
(B)
Reactive group
ICAT inherent in labeling cysteine groups only. In the iTRAQ workflow, post-translational modification information contained in proteins is conserved. The iTRAQ reagents also enhance MS/MS fragmentation, therefore yielding more confident results [26]. Compared to ICAT, iTRAQ requires less chromatographic steps, thus minimising sample loss and saving time. While other stable isotopic labeling methods use MS spectra for quantitation, iTRAQ uses the relative abundance of reporter ions obtained from MS/MS spectra for quantitation. A key advantage of this technique is that it allows labeling of up to four (and soon eight) samples in a single experiment. Therefore, this method is very useful for the quantification of proteins from multiplexed samples since it saves experimental time, and allows for enhanced opportunities for adding replicates to generate more statistically confident data [27-29].
Since each method has its inherent advantages and disadvantages (see Table 2.1 for details), the question that arises is which method is useful for studying S. cerevisiae proteomes? If we are concerned about the protein abundance, iTRAQ and ICAT offer great advantages since these techniques do not have the restriction of 2-DE gels’ resolution of difficult proteins [22]. If we are concerned about PTM analysis, 2-DE gels may be a good choice, since iTRAQ and ICAT rarely identify more than a few peptides per protein, and thus this may impact on the ability to detect PTMs.