On Thursday last week, we published a protocol for spike-in SILAC, a method to analyse the proteome of tissues and organisms developed by Yasushi Ishihama & Matthias Mann and co-workers. This is one of many protocols that we have published that uses stable isotope labelling of proteins or peptides prior to separation and analysis by mass spectrometry. A list of these can be found here.
In fact, the very first protocol that I edited was written by John Asara and was on in-gel stable isotope labeling for relative quantification using mass spectrometry (click here for an overview). This is an adaptation of standard in-gel digestion protocols where gel regions of interest are labelled at lysine residues with either light or heavy isotope-labeled reagents. The parallel slices from the samples-to-be-compared are combined before protease digestion and the resulting peptides are analysed by LC/MS to determine relative abundance of light- and heavy-isotope lysine-containing peptide pairs and analyzed by LC/MS/MS for identification of sequence and modifications.
The chances of success are greater in samples where the complexity has been reduced by e.g. immunoprecipitation or chromatographic separation, and while it has the advantage of simplicity, it has been found to have limited utility for looking at post-translational modifications.
In other approaches, stable-isotope labelling is performed after digestion of the sample. Two that look at specific post-translational medications are: Mass spectrometric identification of N-linked glycopeptides using lectin-mediated affinity capture and glycosylation site-specific stable isotope tagging and Chemical derivatization of histones for facilitated analysis by mass spectrometry
Two that are more generally applicable are: Simultaneous analysis of relative protein expression levels across multiple samples using iTRAQ isobaric tags with 2D nano LC-MS/MS and Multiplex peptide stable isotope dimethyl labeling for quantitative proteomics.
IGOT (isotope-coded glycosylation site-specific tagging) is applicable to the identification/analysis of N-linked glycoproteins and the steps are: (1) digestion of protein mixtures, (2) enrichment of glycopeptides by lectin column-mediated affinity capture, (2) purification of the glycopeptides by hydrophilic interaction chromatography (HIC); (3) reaction of the glycopeptides with [^18^O]-H~2~O in the presence of peptide-N-glycanase resulting in an ^18^O tag specifically at the N-glycosylation site (the Asn residue carrying the sugar chain is converted to Asp with concomitant incorporation of the ^18^O from the solvent); and (4) identification of ^18^O-tagged peptides by LC/MS. The protocol typically results in a list of hundreds of glycoproteins and their sites of glycosylation.
Looking at the post-translational modifications of histone proteins can be rather tricky; they contain a disproportionally large number of arginine and lysine residues especially on the N-termini where most of the PTMs occur making it difficult to obtain peptide sequences that can be easily or meaningfully analysed.
Garcia and co-workers’ method involves blocking the lysine residues of extracted histones with propionic anhydride and trypsin digestion (which will now only cleave after arginine). At this stage, you can do stable isotope labelling step where carboxylic acid groups are reacted with heavy or light forms of a freshly prepared “methyl ester reagent” (methyl acetate) made from methanol (CH3OH or CD3OD) and acetyl chloride. This would enable you to measure the relative abundance of a post-translational modification between two treatment types.
In iTRAQ, multiple samples are compared by labelling them individually tags that all have the same molecular mass (isobaric). The reactive group on the tag is N-hydroxysuccinimide, and the tags react with all available amine groups. The samples are pooled, fractionated and analysed by MS. The advantage of this isobaricness is that the initial mass spec peak of the peptide will have an intensity that is the sum of the intensities from all the samples. The amino-acid sequence ions also show this summed intensity, but on further fragmentation a tag-specific reporter ion is released. The relative intensities of these ions represent the relative amount of peptide in the analytes. There is a good article addressing accuracy and precision issues in iTRAQ quantitation.
In triplex stable isotope dimethyl labelling, these free amines are reacted with a mixture of formaldehyde (H~2~CO) and NaBH~3~CN; OR D~2~CO and NaBH~3~CN; OR D~2~^13^CO and NaBD3CN to form light, intermediate and heavy derivatives. This reaction can be performed in-solution, online with LC-MS or on-column using SepPak columns.
This article also includes a table comparing the relative advantages of dimethyl labelling, iTRAQ and SILAC.
This table is interesting in itself, but also because it illustrates that there are two overall approaches to labelling peptides with stable isotopes: non-metabolic and metabolic. And really, when we talk about metabolic labelling we mean stable isotope labelling by amino acids in cell culture. Even spike-in SILAC, the new protocol that we have published, seems to be a method that you would use when “regular SILAC” would not be possible or practical.
Here is a figure showing the differences between SILAC by the normal method and by the “spike-in” method, both methods developed by Matthias Mann and co-workers.
In SILAC, each sample of cells in culture are fed a diet that contain light, medium or heavy forms of arginine and lysine. The samples are then combined before the steps to lyse the cells and digest the proteins are performed. The main advantage of this approach is that multiple cellular states are combined and then processed in a single workflow enabling very accurate measurement of the differences in protein concentrations between the samples. Even elaborate organelle purifications, like nucleolar preparations, can be performed without any adverse effects on quantitative accuracy. Complex experimental designs involving multiplexing several three-state experiments by linking them with a common experimental state can also be performed.
Another protocol from Akhilesh Pandey’s lab demonstrates the use of SILAC for: studying inducible protein complexes, identifying tyrosine kinase substrates, differential membrane proteomics and studying temporal dynamics.
While SILAC can be performed on the organism level, it can take a long time to achieve complete labelling of an organism and it is restricted to systems where the complete proteome can be labelled by proteome turnover (this precludes human tisse samples, for example). The diet restrictions required may also interfere with the ability to answer the specific scientific question of interest, so for the most part SILAC is limited to experiments using cultured cells. These disadvantages can at least be partially overcome by “spike-in SILAC”. In this case, only one cell state (or experimental state of animal) is subjected to SILAC. Processing of this sample results in a reference protein or proteome that can be added to the other experimental states as an internal standard. This is typically done after cell lysis and before protein digestion. By choosing an appropriate mixture of different cell lines, it is also possible to create a reference proteome that could be used for the analysis of, for example, cancer tissue samples.
While these are the protocols we have in our content at the moment, I am sure that there are many that we have not covered. Whole other subject areas (perhaps for future blog posts) are absolute quantitation of proteins and stable isotope labelling for NMR. It is always amazing to me that each field of study is one of many, and yet behind every detail there is probably a whole thesis of complexity.