Improvement of Mitochondria Extract from Saccharomyces cerevisiae Characterization in Shotgun Proteomics Using Sheathless Capillary Electrophoresis Coupled to Tandem Mass Spectrometry

Abstract

In this work, we describe the characterization of a quantity-limited sample (100 ng) of yeast mitochondria by shotgun bottom-up proteomics. Sample characterization was carried out by sheathless capillary electrophoresis, equipped with a high sensitivity porous tip and coupled to tandem mass spectrometry (CESI-MS-MS) and concomitantly with a state-of-art nano flow liquid chromatography coupled to a similar mass spectrometry (MS) system (nanoLC-MS-MS). With single injections, both nanoLC-MS-MS and CESI-MS-MS 60 min-long separation experiments allowed us to identify 271 proteins (976 unique peptides) and 300 proteins (1,765 unique peptides) respectively, demonstrating a significant specificity and complementarity in identification depending on the physicochemical separation employed. Such complementary, maximizing the number of analytes detected, presents a powerful tool to deepen a biological sample's proteomic characterization. A comprehensive study of the specificity provided by each separating technique was also performed using the different properties of the identified peptides: molecular weight, mass-to-charge ratio (m/z), isoelectric point (pI), sequence coverage or MS-MS spectral quality enabled to determine the contribution of each separation. For example, CESI-MS-MS enables to identify larger peptides and eases the detection of those having extreme pI without impairing spectral quality. The addition of peptides, and therefore proteins identified by both techniques allowed us to increase significantly the sequence coverages and then the confidence of characterization. In this study, we also demonstrated that the two yeast enolase isoenzymes were both characterized in the CESI-MS-MS data set. The observation of discriminant proteotypic peptides is facilitated when a high number of precursors with high-quality MS-MS spectra are generated.

Figure 1.

MS data obtained on 100 ng of yeast mitochondrial extract. (A) NanoLC–MS-MS injection: heat map, where the mass-to-charge ratio of the peptides (Da) is plotted against the LC elution time (min). Each dark grey dot is related to a single MS–MS fragmentation spectrum. (B) CESI–MS-MS injection: heat map, where the mass-to-charge ratio of the peptides (Da) is plotted against the CE migration time (min). (C) Number of proteins and peptides identified in either the nanoLC–MS-MS (light grey) or the CESI–MS-MS (dark grey) injection.

Figure 2.

Comparison of peptides MW (A), mass-to-charge (m/z) ratio (B), pI (C) and Mascot score (D) by means of CESI–MS-MS (dark grey) and nanoLC–MS-MS (light grey). Data are obtained from the analysis of 100 ng of a mitochondrial yeast tryptic digest on both types of coupling.

Figure 3.

Effect on the global protein coverage when comparing the CESI–MS-MS (light grey) and the nanoLC–MS-MS (dark grey) data sets. (A) Proteins are ranked according to the total number of MS–MS spectra matching to each sequence (log scale) and five proteins, distributed all along the dynamic range, were further investigated. (B) Comparison of the sequence coverage from the five previously chosen proteins: shared and specific peptides are indicated in each Venn diagram, as well as the total sequence coverage increase (in %). (C) Boxplot showing the sequence coverage distributions for both data sets when considering all the identified proteins.

Figure 4.

Venn diagrams indicating the distribution of mitochondrial proteins identified with Panther and iLoc-Euk databases using nanoLC–MS-MS and CESI–MS-MS approaches. Distribution of proteins between mitochondrion and other sub-cellular compartments using the nanoLC–MS-MS data set (A) or the CESI–MS-MS data set (B). (C) The overlap between the global protein data set and the mitochondrial proteins using the two approaches.

Figure 5.

How is it possible to distinguish two protein isoforms using discriminant peptides and an increased sequence coverage: ENO1 and ENO2 proteins as an example. (A) Comparison of the sequence coverage for each isoform: shared and specific peptides are indicated in the Venn diagrams, as well as the total sequence coverage increase (in %). (B) BLAST-P alignment between both ENO isoforms. (C) Detailed sequence coverage on both isoforms obtained by either the CESI–MS-MS or the nanoLC–MS-MS analysis: identified amino acids are shown with a light grey label, whereas the grey box below the corresponding peptide is indicating the MS–MS spectrum quality (a light grey box is attributed under each amino acid if it was seen with either a b- or a y-ion, represented in the upper and lower section of the grey box, respectively). Proteotypic peptides allowing the distinction between both isoforms in each data set are indicated with purple boxes. (D) Focus on two proteotypic peptides matching on ENO1 and ENO2 sequences: a difference of two amino acids in the composition of a peptide is easily distinguishable when inspecting the MS–MS fragmentation spectra.