Dual-pressure linear ion trap mass spectrometer improving the analysis of complex protein mixtures

Abstract

The considerable progress in high-throughput proteomics analysis via liquid chromatography-electrospray ionization-tandem mass spectrometry over the past decade has been fueled to a large degree by continuous improvements in instrumentation. High-throughput identification experiments are based on peptide sequencing and are largely accomplished through the use of tandem mass spectrometry, with ion trap and trap-based instruments having become broadly adopted analytical platforms. To satisfy increasingly demanding requirements for depth of characterization and throughput, we present a newly developed dual-pressure linear ion trap mass spectrometer (LTQ Velos) that features increased sensitivity, afforded by a new source design, and demonstrates practical cycle times 2 times shorter than that of an LTQ XL, while improving or maintaining spectral quality for MS/MS fragmentation spectra. These improvements resulted in a substantial increase in the detection and identification of both proteins and unique peptides from the complex proteome of Caenorhabditis elegans, as compared to existing platforms. The greatly increased ion flux into the mass spectrometer in combination with improved isolation of low-abundance precursor ions resulted in increased detection of low-abundance peptides. These improvements cumulatively resulted in a substantially greater penetration into the baker's yeast (Saccharomyces cerevisiae) proteome compared to LTQ XL. Alternatively, faster cycle times on the new instrument allowed for higher throughput for a given depth of proteome analysis, with more peptides and proteins identified in 60 min using an LTQ Velos than in 180 min using an LTQ XL. When mass analysis was carried out with resolution in excess of 25,000 full width at half-maximum (fwhm), it became possible to isotopically resolve a small intact protein and its fragments, opening possibilities for top down experiments.

Fig. 1

Schematic representation of the Thermo Scientific LTQ Velos mass spectrometer, containing a stacked-ring ion guide (S-lens) source and a dual-pressure ion trap with differential pressure regulation.

Fig. 2

A: Full MS scan from LTQ Velos (top) and LTQ XL (bottom) indicating greater than 10X improvement in sensitivity for the new instrument. Average observed improvement in sensitivity is 5-fold. B: Normal scan resolution for LTQ XL (top) and LTQ Velos (bottom) in full MS achieved at scan rates of 16, 000 amu/s and 33, 000 amu/s respectively.

Fig. 3

Instrument scan lengths as composites of several events in MS and MSn modes on LTQ XL and LTQ Velos, calculated from a single top-ten 60 min LC-MS/MS run using 1 µg of C. elegans digest. Out of seven events, five (six for MS scan) have constant duration and two (injection time and MS/MS mass analysis time) have variable durations. The error bars represent standard deviation and reflect injection time variability for MS and MS/MS events and mass analysis time variability for MS/MS events.

Fig. 4

A: A typical base peak LC/MS trace from 1 µg C. elegans digest recorded on LTQ Velos. Inset shows that the sampling frequency (inversely proportional to the distance between the two adjacent full MS base peaks) is ~ 2.3 times higher for LTQ Velos (inset top) than for LTQ XL (inset bottom, equivalent run). B: Running averages (100 points) for MS peak depth sampling for LTQ Velos and LTQ XL. C: Number of MS/MS events vs. estimated local signal-to-noise ratios of precursor ions. LTQ Velos spent most of the 2x extra scans sampling lower abundance precursors than LTQ XL.

Fig. 5

A: Identified unique peptides and proteins from 1 µg C. elegans digest LC/MS runs reported at <1% FDR (spectral level). B: Venn diagram illustrating the overlap in proteins identified between the two instruments for a 60 minute reverse phase separation of 1 microgram of C. elegans digest. C: Venn diagram illustrating the overlap in proteins identified in three replicated 60 minute separations of 1 µg of C. elegans digest on LTQ Velos.

Fig. 6

A: Identified unique peptides and proteins from 20 ng C. elegans digest LC/MS runs reported at <1% FDR. B: Running average (100 per) of the number of ions recorded in MS/MS scans from these runs (one run per instrument shown). Under injection time limited conditions (as in this experiment) this is proportional to ion beam brightness.

Fig. 7

Identified proteins (A) and unique peptides (B) as a function of their cellular expression levels in whole-cell digests of baker’s yeast (Saccharomyces cerevisiae). Analytical runs were performed in triplicate on LTQ XL and LTQ Velos, using 180 minute LC gradients and 1 µg of digest per run. The results from replicate runs were aggregated for each instrument, and the identified proteins were annotated with their cellular expressions levels.

Fig. 8

High throughput as a function of improved instrument performance. The LTQ Velos identified more unique peptides and proteins in a 60-minute analysis than did the LTQ XL in a 180-minute analysis of the same complex mixture (1 µg of C. elegans digest).

Fig. 9

The increased resolution afforded by the LTQ Velos allows isotopic resolution of small intact proteins. A: Resolution of the 15+ molecular ion of horse heart myoglobin using ultra-zoom scan. B: Theoretical isotopic distribution for this ion at 25, 000 FWHM. C: MS/MS of 15+ molecular ion recorded using zoom scan with fragments up to 8 kDa isotopically resolved. Inset: Larger fragments are resolved using ultra-zoom scan. The mass difference (in units of 1.00235 Da) between the most abundant isotopic peak and the monoisotopic peak is denoted in italics after each M_r value. D: Fragmentation map of horse heart myoglobin (from the output of ProSightPC), showing 10⁻⁸ expectation score for this match.