fragmentation of proteins containing disulfide bonds by ... · demonstrate electron capture...

Fragmentation of proteins containing disulfide bonds by electron capture dissociation in a modified quadrupole-time of flight mass spectrometer

Michael C. Hare1, Valery G. Voinov1, 2, Yury V. Vasil’ev1, 2, Joseph Meeuwsen1, 2, Joseph S. Beckman1,2

1e-MSion, Inc., Corvallis, OR 2Oregon State University, Corvallis, OR

IntroductionDisulfide bonds are essential for stabilizing the tertiary structureof many proteins, but the development of unambiguousmethods for locating them is an ongoing challenge. Electron-based fragmentation methods cleave disulfide bondspreferentially, but can be challenging to achieve. We heredemonstrate electron capture dissociation (ECD) resulting indisulfide bond cleavage and extensive backbone fragmentationin a commonly available quadrupole time-of-flight instrument.ECD fragmentation by this method can be usefully carried out ata millisecond time scale, making it compatible with LC-MSmethods.

InstrumentationAn Agilent 6545XT upgraded to include ECD capability was usedfor these experiments. An e-MSion ExD cell is installed betweenthe quadrupole and a shortened collision cell of the Q-TOFinstrument.

The ExD cell uses a combination of electric and magnetic fields to trap low-energy electrons in the path of ions selected by the quadrupole. Electrons are produced within the cell by a rhenium alloy filament.

MethodsIntact proteins were introduced by infusion or via the AgilentLC system. A parent ion was selected by the quadrupole using atargeted acquisition method and fragmented in the EXD cell.Data was processed using MASH Explorer1 or LCMS Spectator2

for fragment identification.

ECD of Bovine InsulinInsulin is composed of two polypeptide chains a and b, bridgedby two disulfide bonds. In addition, there is an intramoleculardisulfide in chain a. The mass of intact insulin confirms that it isfully oxidized. The +5 or +6 charge state of insulin was selectedand subjected to ECD.To calculate possible cleavages, dehydro-cysteine was used tomodel homolytic cleavage of the disulfide bond, and the mass ofthe uncleaved peptide was added to cysteine to model an intactdisulfide. Combining results, cleavage was observed in 27 of apossible 29 bonds (93%) for chain b, and 14 of 20 bonds (70%)for chain a. c-type (blue) and z-type (red) ions are mostabundant and most intense, but complementary ion types arealso observed.

For example, this series of z ions is formed by cleavage of the bchain as well as homolytic cleavage of one or both disulfides.

All ions in this series are doubly charged. The red dots are calculated isotopic distributions for these ions.

ECD of ApaminThe bee-venom peptide apamin contains 18 amino acids, twodisulfide bonds, and three hydrogen bonds. In spite of itscompact structure, using the ExD cell resulted of ECD cleavageof 15 of 16 (94%) of ECD-cleavable bonds.

ECD of a NanobodyNanobodies (single-domain antibodies) are functioningantibodies found in camelid species. In an LCMS experiment, ananobody was isolated from a mixture and subjected to ECDfragmentation, with 76% sequence coverage and revealing adisulfide bond between C28 and C101*.* The C28-C111 disulfide cannot be ruled out.

Conclusions• The Agilent 6545XT LC/MS system gains highly effective ECD

capabilities by installation of the e-MSion ExD cell. • Nearly complete coverage of the sequence of disulfide-

containing proteins can be determined in a single LC run.• Disulfide bonds in complex biomolecules can be confidently

mapped by ECD in the Agilent 6545XT Q-TOF.

More references and information are available at https://e-msion.com/

Cut-away view of the e-MSion ExDcell attached to the collision cell.

Sample was introduced from an Agilent Infinity II HPLC system. Spectra wereacquired in targeted mode and 8 scans from a single LCMS run were averagedfor analysis. Data analysis was performed in MASH explorer using eTHRASHfor deconvolution. Error tolerance for peak matching was 10 ppm. All peakswere visually inspected and verified.

Sample introduced by direct infusion. Data analysis performed in MASHexplorer using eTHRASH for deconvolution. Error tolerance for peakmatching 10 ppm. All peaks were visually inspected and verified.

References1. Cai, W.; Guner, H.; Gregorich, Z.R.; Chen, A.J.; Ayaz - Guner, S.; Peng, Y.;

Valeja, S.G.; Liu, X.; Ge, Y. MASH Suite Pro: A Comprehensive Software Tool for Top - down Proteomics, Mol.Cell.Proteomics Epub ahead of print.mcp.O115.054387.

2. Wilkins, C; Sangtae Kim, and Jungkap Park, http://omics.pnl.gov/Conflicts of Interest and AcknowledgementsJSB, VGV, and YVV are cofounders of e-MSion, Inc., which uses technology developed and licensed from Oregon State University. For employees of Oregon State University, a potential conflict of interest exists. This work was supported by the NIH (GM122131-02 and GM123855).

Sample was introduced by direct infusion. Data analysis was performed inusing LCMS Spectator. Error tolerance for peak matching was 10 ppm.