Top-down protein identification of proteasome proteins with nanoLC-FT-ICR-MS employing data-independent fragmentation methods

Abstract

A comparison of different data-independent fragmentation methods combined with LC coupled to high-resolution FT-ICR-MS/MS is presented for top-down MS of protein mixtures. Proteins composing the 20S and 19S proteasome complexes and their PTMs were identified using a 15 T FT-ICR mass spectrometer. The data-independent fragmentation modes with LC timescales allowed for higher duty-cycle measurements that better suit online LC-FT-ICR-MS. Protein top-down dissociation was effected by funnel-skimmer collisionally activated dissociation (FS-CAD) and CASI (continuous accumulation of selected ions)-CAD. The N-termini for 9 of the 14 20S proteasome proteins were found to be modified, and the α3 protein was found to be phosphorylated; these results are consistent with previous reports. Mass-measurement accuracy with the LC-FT-ICR system for the 20- to 30-kDa 20S proteasome proteins was 1 ppm. The intact mass of the 100-kDa Rpn1 subunit from the 19S proteasome complex regulatory particle was measured with a deviation of 17 ppm. The CASI-CAD technique is a complementary tool for intact-protein fragmentation and is an effective addition to the growing inventory of dissociation methods that are compatible with online protein separation coupled to FT-ICR-MS.

Keywords: Electrospray ionization; Fourier transform-ion cyclotron resonance; Proteasome; Protein LC-MS; Technology; Top-down mass spectrometry.

Figure 1

Schematic of the top-down protein identification strategy using nanoLC FT-ICR MS. The human 20S proteasome complex was separated into its components using a monolithic column followed by intact mass measurement and protein fragmentation using either (A) FS-CAD, (B) CASI-CAD, or (C) selectively enriching the fragments generated by FS-CAD using CASI. The intact protein and product ion masses were then searched against the sequence database for protein identification.

Figure 2

TIC from the LC-MS of the 20S proteasome proteins. All 14 subunits elute within 27–42 min (peaks 1–11, bottom panel). A simulated gel-view of the proteins eluting in each peak is depicted in the top panel for the corresponding elution window shown in the TIC. With the intact and product ion masses, the proteins were identified by database searching and the peaks were labeled with the corresponding protein(s) being eluted (bottom-panel). α and β subunits are labeled in red and green respectively. Peak 7, labeled with an *, was found to contain oxidized form of β2 subunit.

Figure 3

Schematic of Bruker 15-Tesla FT-ICR MS. The declustering voltage of skimmer 1 was increased to facilitate FS-CAD. During CASI-CAD, a window of m/z 800–1200 was chosen by the external quadrupole and all the ions within this range (shown in red) were transferred to the collision cell where it accumulated for a specified time interval and dissociated by CAD.

Figure 4

FS-CAD of the α4 20S proteasome subunit. (A) Full m/z range mass spectrum obtained after FS-CAD, with the inset showing the decharged isotopic distribution of α4 protein. (B) Magnified region of the mass spectrum showing the sequence tag identified from mapping the product ions. (C) Amino acid sequence of the α4 protein with the b- (red) and y- (green) product ions mapped by database searching. The sequence tag obtained by mapping the fragments in (B) is highlighted by the blue box.

Figure 5

CASI-CAD of α4 20S proteasome subunit. (A) Mass spectrum obtained after CASI-CAD, with the inset showing the decharged isotopic distribution of α4 protein. (B) Zoomed-in mass spectrum showing a sequence tag identified from mapping the product ions. (C) Amino acid sequence of the α4 protein with the b- (red) and y- (green) product ions mapped by database searching. The sequence tag obtained by mapping the fragments in (B) is highlighted in the blue box.

Figure 6

MaxEnt deconvoluted zero-charge mass spectrum of the α3 subunit with the singly phosphorylated form present at 5 times greater relative abundance to the non-phosphorylated form.

Figure 7

Intact mass spectrum of the 100 kDa Rpn1 h19S proteasome subunit averaged from the LC-MS peak (TIC shown in inset).