Hodge K et al. (2013), Cleaning up the masses: exclusion lists to redu...

XB-ART-46954

J Proteomics 2013 Aug 02;88:92-103. doi: 10.1016/j.jprot.2013.02.023.

Show Gene links Show Anatomy links

Cleaning up the masses: exclusion lists to reduce contamination with HPLC-MS/MS.

Hodge K , Have ST , Hutton L , Lamond AI .

???displayArticle.abstract???
Mass spectrometry, in the past five years, has increased in speed, accuracy and use. With the ability of the mass spectrometers to identify increasing numbers of proteins the identification of undesirable peptides (those not from the protein sample) has also increased. Most undesirable contaminants originate in the laboratory and come from either the user (e.g. keratin from hair and skin), or from reagents (e.g. trypsin), that are required to prepare samples for analysis. We found that a significant amount of MS instrument time was spent sequencing peptides from abundant contaminant proteins. While completely eliminating non-specific protein contamination is not feasible, it is possible to reduce the sequencing of these contaminants. For example, exclusion lists can provide a list of masses that can be used to instruct the mass spectrometer to 'ignore' the undesired contaminant peptides in the list. We empirically generated be-spoke exclusion lists for several model organisms (Homo sapiens, Caenorhabditis elegans, Saccharomyces cerevisiae and Xenopus laevis), utilising information from over 500 mass spectrometry runs and cumulative analysis of these data. Here we show that by employing these empirically generated lists, it was possible to reduce the time spent analysing contaminating peptides in a given sample thereby facilitating more efficient data acquisition and analysis. BIOLOGICAL SIGNIFICANCE: Given the current efficacy of the Mass Spectrometry instrumentation, the utilisation of data from ~500 mass spec runs to generate be-spoke exclusion lists and optimise data acquisition is the significance of this manuscript.

???displayArticle.pubmedLink??? 23501838
???displayArticle.pmcLink??? PMC3714598
???displayArticle.link??? J Proteomics
???displayArticle.grants??? [+]

Species referenced: Xenopus laevis
Genes referenced: prss1

???attribute.lit??? ???displayArticles.show???


	Fig. 1. Dynamic exclusion allows the mass spectrometer to more efficiently identify peptides in a sample. The first scan measures the ions with the highest intensity (most abundant). These masses are added to a temporary ‘exclusion’ list for a period of typically 30–90 s. Once the high intensity peaks have been sequenced and excluded the MS can measure peaks under the threshold, thereby detecting less abundant peptides.
	Fig. 2. The exclusion lists were generated using 527 mass spectrometry runs. The pie chart shows the percentage of each species over the total number or runs. For H. sapiens 253 MS runs were used to generate the exclusion list. The S. cerevisiae exclusion list was generated using 127 MS runs, C. elegans using 105 MS runs and 42 runs for X. laevis. As the graph shows most of the data available to us (48%) was from H. sapiens reflecting the fact that most studies are performed on human cell lines.
	Fig. 3. m/z (mass/charge) plotted against retention time for a whole cell proteome experiment. Contaminant peptide masses appear as vertical ‘lines’ that do not show chromatographic resolution. Peptides were identified for the digestive enzyme trypsin at 421.7584 and keratin at 769.7194.
	Fig. 4. Complex lysate samples were analysed in triplicate with and without the be-spoke exclusion list. Mass-to-charge (m/z) was plotted against retention time. The first graph shows the data run without the exclusion list. The ‘lines’ are apparent in the graph that correspond to common contaminant peptide masses identified during the run such as keratin and trypsin. The second graph shows the data run with the exclusion list. While some ‘lines’ are still apparent, they appear to have decreased in frequency. See supplementary information Fig. 4 for technical replicates.
	Fig. 5. Immuno-precipitation samples were analysed in triplicate with and without the exclusion list. The m/z was plotted against the retention time. Vertical ‘lines’ of contaminant peptide masses were reduced when using the exclusion list. See supplementary Fig. 5 for the technical replicate graphs.
	Fig. 6. Purified protein samples run with and without the exclusion list. m/z values were plotted against the retention time. The ‘lines’ of contaminant peptide masses were apparent and these reduced in frequency when samples were run with the exclusion list. The loss of data upon use of the exclusion list is clearly visible in the lower right graph, which undermines the value of the exclusion list approach when analysing low complexity samples. See supplementary Fig. 6 for technical replicate graphs.
	Fig. 7. Venn diagrams showing the number of unique peptides and proteins identified for complex lysate samples run with or without an exclusion list compared against technical replicates.
	Fig. 8. Sequence coverage diagram for the embryonic protein vitellogenin in C. elegans. Comparison between peptides identified during runs show that there is no loss of protein coverage when using an exclusion list.

References [+] :

Anderle, Quantifying reproducibility for differential proteomics: noise analysis for protein liquid chromatography-mass spectrometry of human serum. 2004, Pubmed